My Most Memorable Games (gnv64).pdf

Gelfand

•

My Most Memorable Games

1

Boris Gelfand


Preface by Vladimir Kramnik and introduction by Dirk Poldauf

2005 EDITION OlMS

m 3

Bibliographic Information published by Die Deutsche Bibliothek

Die Deutsche Bibliothek lists this publication in the

Deutsche Nationalbibliografie; detailed bibliographic data is available in the internet at http://dnb.ddb.de.

Copyright © 2005 Edition Olms AG Breitlenstr. 11

•

CH-8634 Hombrechtikon/Zurich,Switzerland

E-mail: [email protected] Internet: www.edition-olms.com All rights reserved. This book is sold subject to the condition that it shall not by way of trade or otherwise,be lent re-sold,hired out or otherwise circulated in any form of binding or cover other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchaser. Printed in Germany Editor and translator: Ken Neat Photographic Acknowledgement: p. 138, 158,235 Archive Abram Gelfand

Typesetting by Art & Satz

•

Ulrich Dire D-80331 Munchen

Printed by: Druckerei Friedr. Schmucker GmbH, D-49624 Loningen Cover: Eva Konig, D-22769 Hamburg

ISBN 3-283-00453-5

4

Contents Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5

C o n ve n t i o n a l S i g n s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6

P reface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7

I nt rod u c t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8

The m a k i n g of a c l as s i c a l g ra n d m aster The boy fro m M i n s k . . . . . . . . . . . . . . . . H i s r i se to wo r l d - c l ass stat u s . . . . . . . S u c c essfu l yea rs . . . . . . . . . . . . . . . . . . .

. . . .

9 10 12 13

M y Favo u r ite Var i a t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

17

Game 1: Gelfand-Dorfman 17

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

Game 2: Gelfand-Ftacnik 21

Game 7: Gelfand-Shirov 41

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

Game 3: Gelfand-Ivanchuk 26 Game 6: Gelfand-Leko 38

Game 5: Gelfand-Kamsky 34

Game 4: Gelfand-Ivanchuk 30

. . . .

M y M ost M e m o ra b l e G a m es. . ............................................... 47 Game 10: Gelfand-Lobron 55

Game 8: Loginov-Gelfand 47

Game 9: Lev-Gelfand 50

Game 14: Gheorghiu-Gelfand 70

Game 15: Gelfand-Kasparov 73

Game 20: Damljanovic-Gelfand 98

Game 21: Polugayevsky-Gelfand 101

Game 13: Gelfand-Ivanchuk 67

Game 12: Gelfand-Malisauskas 62

Game 11: Gelfand-Adams 59

Game 16: Gelfand-Chandler 80 Game 19: Short-Gelfand 94

Game 18: Gelfand-Georgiev 86

Game 17: Short-Gelfand 83


Game 22: Gelfand-Yusupov 103

Game 24: Gelfand-Anand 110


Game 30: Gelfand-Kramnik 133

Game 31: Shirov-Gelfand 139



Game 26: Shirov-Gelfand 117

Game 29: Gelfand-Topalov 129

Game 34: Gelfand-Karpov 151

Game 33: Gelfand-Salov 148

Game 32: Gelfand-Piket 145

Game 37: RUblevsky-Gelfand 163

Game 36: Gelfand-Sutovsky 159

Game 35: Gelfand-Korchnoi 154

Game 40: Gelfand-Georgiev 176

Game 39: Gelfand-Lautier 172

Game 38: Gelfand-Markowski 169

Game 43: Gelfand-Piket 192

Game 41: Gelfand-Ye Jiangchuan 180 Game 42: Gelfand-Anand 185 Game 44: Gelfand-Van Wely 195

Game 45: Gelfand-Delchev 197

Game 50: Shabalov-Gelfand 213

Game 51: Gelfand-Grischuk 218

Game 46: Gurevich-Gelfand 200 Game 49: Gelfand-Acs 210

Game 48: Gelfand-Bruzon 208

Game 47: Grischuk-Gelfand 203

Co m b i n at i o n s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222 .

E n d i n g s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236 Appen d i x . . . . . . . . . . . . . I n d ex of P l ayers . . . . I n d ex of Gam es . . . . I n d ex of O p p o n e nts I n d ex of O p e n i n g s . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

... ... ... ... ...

. . . . .

. . . . .

. . . . .

5

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

252 252 255 260 261

Conventiona l Signs

Symbol Meaning

Sym bol Meaning

ctJ �

King Queen Rook

A � �

Bishop Knight Pawn

+

check captures short castling

# N 0-0-0

checkmate novelty long castling

unclear position Wh ite stands slightly better Wh ite stands better White has a decisive advantage Black resigns equal position

00

+ =+= -+ 0-1 %-%

com pensation for material Black stands slightly better Black stands better Black has a decisive advantage Wh ite resigns draw agreed

!!

interesting move good move brill iant move

?! ? 7?

du bious move bad move blunder

'If

Wh ite to move

..

Black to move

�

x

0-0 00

;j;; ± +1-0 =

!?

6

Prefa ce reflection of his personality. He is not only and this is accessible only to a few - a highly u n iversal player, capable of playi ng equally well in the most varied types of positions. What i m presses me most is his abil ity to create games, where all the moves, from the first to the last, are as though links in a single log ical chain. Th is inexorable consistency in the real isation of his strateg ic conceptions is, in my view, the main trait of Boris Gelfand the chess player.

In 1 994 Boris Gelfand and I were drawn to gether in a q uarter-fi nal Candidates match for the FIDE world championsh ip. That same year the Professional Chess Association (PCA) organised a parallel cycle. I (li ke the majority of players, however) did not see any reason not to play in both. Boris took the dif ficu lt decision not to partici pate i n the PCA world cham pionsh ip, in order to concentrate on one. I remem ber very wel l how, i n an article de voted to the match he had won against me, he quoted Seneca : 'The one who is every where is nowhere' . Years later I can confi dently say that this im portant lesson greatly hel ped me in my su bseq uent chess career. I n my view, that exam ple vividly charac terises the author of this book. Th is is a friendly and intel ligent person (but at the same time very fi rm , when it is a matter of principles) , capable of penetrating deeply into the essence of any problem and finding the correct strateg ic solution. Like all those who devote themselves entirely to chess, Boris's play is in many respects a

The fact that for fifteen years now he has been one of the chess el ite is an add itional ind ication that he is one of the most sign ifi cant of contem porary players. As a person who has a high appreciation of strategy and logic i n chess, I have been given enormous pleasure by the study of this book. I am convi nced that you too will like it.

Vladimir Kramnik Classical World Chess Champion

7

Int ro d uct i on my chess and emotional feelings during the game.

You have before you my first book. I have been playing chess now for thirty years, half of them at the very hig hest level . Therefore I decided that the time had come to compile my most memorable games in one book. Out of the one and a half thousand games that I have played , I faced the difficult problem of choosing the most vivid and interesting.

In my comments on the opening stages of the games, I have normally referred to my own games. In those cases where I have had great experience of playing the specific vari ation , I have g iven n umerous references to my own games, so that the reader can bet ter understand the con nection between the open ing and the m idd legame (in my view, the most im portant and interesting part of a chess game).

Since the size of the book is restricted , I had to excl ude a n u m ber of noteworthy games. In selecting them , I was gu ided by several criteria: firstly, I chose those games that left the deepest trace in my memory. Si nce I have always ai med (and, I hope, succeeded) at being a u n iversal player, in this book I have endeavoured to incl ude games where success came to me as a resu lt of good open ing preparation , successful play in the endgame, a strongly conducted attack or mastery in defence and counterattack, and also as a result of tactical resourcefu lness or deep strategy and subtle manoeuvring.

Creative successes do not always go hand in hand with com petitive achievements. There fore in this col lection I have not incl uded games from some of my best tournaments, preferring those from tournaments where com petitively I was less successfu l. The ma jority of the games presented here were an notated in their time for publications such as New in Chess, Chess Informator and others. Specially for this book I have analysed all the games and extracts anew.

In addition , I hope that this book will interest not only those chess enth usiasts who w i ll, I hope, gain pleasure from playi ng through the games, but also those who primari ly want to i m prove. Therefore in those games where there was especially complicated tac tical play, apart from verbal comments I have given a more detai led analysis of the specific variations. For the analysis of the most com plex positions I have resorted to the help of analytical prog rams, wh ich have hel ped me to correct many evaluations, and sometimes even to rad ically change them . At the same time, I have endeavoured not to abuse the aid of the computer, so as not to overshadow

I should l i ke to express my g ratitude to my fam i ly, and to all my trainers, sparrin g partners and friends, w h o have supported me throug hout my career and continue to do so. I am especially gratefu l to my main fan , my father Abram Gelfand , who for many years has been trying to persuade me to publish a collection of my best games. Many colleag ues mai ntain that one of my strongest professional q ual ities is open ing preparation . I hope that my opening attempt as an author will also prove successful .

Boris Gelfand Rishon-Ie-Zion , Israel

8

The m aki ng of a class i ca l g rand m aster of his pragmatic grandmaster colleagues, for whom only the fi nal resu lt counts, and who will not infreq uently content themselves with playi ng the second -best move in a position to save time and energy. This energy pro b lem has beset Gelfand throug hout h is career. When Gelfand , who d u ring games l i kes to consume /sostar, his favourite energy dri n k, a prod uct for which he wou ld be the ideal promoter, has taken on board sufficient re serves of energy i n the course of focussed preparation for a tournament, he is capable of ach ieving anything. His victories i n two successive I nterzonal tournaments, a series of World Championsh ip Cand idate matches, as well as fi rst prizes in prestigious tour naments such as M oscow 1 992 , Belgrade 1 995, or Polan ica Zdroj 1 998 and 2000 are eloquent proof of this. In contrast are a num ber of fai lures which were primarily brought about by the fact that, following a previous hard -fought match , he had not had suffi cient time to recharge his batteries. 'Should I perhaps adopt a sim pler approach to the game ?' was a question he put to himself on more than one occasion when thi ngs were going badly. It is not just because of his com petitive achievements and resu lts that Gelfand, who is classed more or less reg ularly among the world 's top ten players, has, si nce the n i neties, been regarded as one of the most impressive figures on the international chess scene. Popular with his professional col leagues and having many friends, he is seen by the wider chess public as a sort of moral authority. The words of Boris Gelfand carry weight. A sense of responsi bil ity for the de velopment of their game has latterly become a rare commod ity among leading gran d masters. Gelfand , however, is o n e o f those

I n 1 989 the magazine 64 offered an ori gi nal prize to the partici pants i n one of the now legendary USSR Championsh ips: to be awarded to the player sacrificing the most material in the tournament. It was the twenty one -year-old Boris Gelfand from M i nsk who won this special award and who by his ad venturous play went on to win the bronze medal as wel l . Gelfand 's playi ng style con tin ues to the present day to be characterised by its d i rect and com bative approach. I n contrast to those positional players, who seek to grind down their opponents by long drawn-out manoeuvri ng, Gelfand 's concept of chess has always incl uded a particularly dynam ic approach. H i s dramatic victories with Wh ite, not just but especially agai nst the G rO nfeld Defence, are wel l known. But his precisely judged repertoire with the black pieces is proof of a basic com bative ap proach far removed from classical notions of playing for equal ity. Once an enthusiastic practitioner of the strateg ical ly risky Ki n g's Ind ian Defence, over the years Gelfand has made sign ificant and lasting contri butions to the open ing theory of the double -edged Naj dorf Sici l ian , a defence i n which he ran ks alongside Garry Kasparov as one of the world 's best exponents. His approach to playi ng chess has produced a stri ng of har monious looking games, as the reader of the cu rrent book will soon discover for h i mself. On the other hand, his approach to the game is physically extremely exhausting , and one which does not always choose the quickest route to victory. Boris Gelfand is a seeker after chess truths. H is overriding priority is to fi nd the best possible move in every pos ition . Th is means that he freq uently gets into time -trouble and this has cost h i m many a won game. Certai nly a lot more than so many

9

�

The making of a classical grandmaster � �

V��_S � Q_n_ � , n ,_ _w � _'__ __ __

________

•

The boy from Mi nsk �

____ �_______

people who do not hesitate to express their opinions publ icly and to back them up with sound reasoning. In i nternational special ist publications are to be found a whole series of poi nts of view by Gelfand about important current events. Th us he is one of the most passionate advocates of the classical game of chess, whose very existence he sees as threatened by cu rrent FIDE-led attem pts to red uce time limits ever further. On these and other subjects Gelfand does not argue from reasons of self-i nterest, but rather stays fo cussed on the bigger pictu re. In this sense he always was, and sti l l remains, an ideal ist, who does not regard chess si m ply as a means to an end , but instead sincerely loves the royal game for its own sake. And has done si nce earl iest childhood .

The boy from Minsk When Boris was four years old his father Abram bought him his fi rst chess book. Th is was The ABC of Chess by Averbakh and Beyl i n , as Gelfand sen ior sti l l remem bers to this day : 'I decided that we should look at one diagram a day. In this way we cou ld get through the whole book in a year. And that is how it went in itial ly. We studied chess every day. Throughout the week Boris would wait im patiently for my return from work so that we could get on with learn ing. But after a few months, when we had worked through a third of the book, Boris preferred to work without my hel p. I thought he had lost i nterest i n chess. But soon afterwards I discovered that he had already struggled through on his own to the last chapters and was playi ng through the games of the grandmasters ! ' I t i s almost always parents who sow the idea i n their children's mind that decides the fu ture course of their l ife. Abram Gelfand, l i ke h i s wife Nella, was born before the Second World War in Belarus. They spent five years of their childhood fleeing the war in the east ern part of the USSR, returning later to Minsk.

10

________ __

In their professional lives they both chose scientific careers and were very successfu l in this field . It could never have occu rred to Abram Gelfand, an electro-tech n ical en gineer, who spent forty years travelling be tween construction sites in Belarus, Lithua nia and Russia, that his son would lead a simi lar, if not more intense, nomadic life as a chess professional . But for the g ifted Boris this was al l a long way ahead . By the age of six he was already bei ng coached three times a week by Edward Zel kind . The celebrated trainer must have immediately recog nised the boy's talent, be cause normal ly only schoolch ildren could at tend his trai n i ng sessions, and at that time Boris was sti ll i n n u rsery schoo l . Zel kind is very u n l i kely to have reg retted maki ng this one exception . Th is is borne out by a small i ncident which Abram recal ls. When Boris visited Zel kind for the fi rst time, they exam ined without sight of the board (!), a game which David Bronstein had played i n 1 963 agai nst a computer. Zel kind asked h i m why the knight had gone to g4 at move 4. Boris replied that the kn ight would otherwise have been captured by the pawn . I n short, says Boris's father Abram , it was clear that Boris was not si mply learning the moves by heart, but that he also u nderstood which moves had to be played and why. For five years - from 1 974 to 1 979 Boris remained u nder the wing of his first trainer. From h i m the children learned the general principles of chess and of piece develop ment. An i mportant part of the coaching fo cussed on the study of the most diverse types of combination . The young players concentrated on the elementary principles of pawn end ings and rook endings and mas tered the essential principles of open ing and endgame theory. The endgame technique of chess players from the former Soviet U n ion is legendary. Thus the story that the fam ous Engl ish g randmaster Anthony M i les, at a tournament in the Swiss town of Horgen in -

The making of a classical grandmaster &

'r\_

1

_0lIl1

•

The boy from Mi nsk �;"':"!!!llt

tt)

iI:li!II.dmm;Jlilt.III".�_�.f2'_.Ii __

".1i!P*

sisted Boris at dozens of tournaments, start ing from j u n ior championsh i ps right up to Candidates matches. Boris also considers it i m portant, that between 1 980-1 983 he was able to attend the Tigran Petrosian schoo l , spending a period o f two weeks there o n three separate occasions. ' It was something special , to have direct experience with such a g reat player. To be able to tal k, analyse, and play bl itz games. I remember Tigran Pet rosian saying to me: "Never make a move without there bei ng an idea beh ind it, even when playi ng bl itz. Always think!" I think that this conversation played an im portant role in establishing my way o f playi ng . ' The development of Boris's chess talent was also aided by the hold ing of the 47 th USSR Championsh i p i n his home town of M i nsk i n 1 979 . Th is must have been a key experi ence for the boy, who was able for the fi rst time to meet the stars of Soviet chess i n t h e flesh . Boris's father accom pan ied h i m reg ularly t o t h e tournament. H o w could h e possibly have missed a single round ? After all, at the tender age of 1 1 , Boris had al ready achieved his fi rst Candidate M aster norm . Whi lst Boris devoted his time to a carefu l study of the games, his father wou ld usually go for a wal k around the town or i n a park. One day, outside the M i nsk Ho tel not far from the tournament hal l , Abram met a large group of players returning from that day's rou n d, among them Yefim Geller, M i khail Tal and Alexander Bel iavsky. It was high time to pick up Boris! It was, after all, somewhat late and Boris had to be back i n school t h e next day. Boris's father h u rried to the tournament hal l . There was just one game sti ll i n progress. And there was j ust one spectator - Boris! His son cou ld not be persuaded to leave. He insisted on bei ng al lowed to watch the game to its end . In these circumstances it was hardly surprising that, by the end of the tournament, everybody knew Boris, from the players, the spectators, and u p to and including the ticket sellers.

1 994 lost a rook end ing against Boris, with out putting u p any sort of a fight, and that this was the very rook ending whose prin ci ples had been taught to h i m by Zel ki nd , when Boris was only ni ne, is entirely believ able. In 1 979 Zelkind emigrated to America. Boris became a pupil of Tamara Golovey. She was an outstand ing teacher and a very strong player, who took part eight times i n t h e final o f t h e Soviet Women's Champion ship and won the Championship of Belarus three ti mes. 'She was l i ke a second mother to h i m and loved h i m very much. We never needed to worry about Boris when he went to tournaments with Tamara Golovey, ' re flects Boris's father Abram , when looking back on this time. I n the years when Boris was taki ng his first steps on the 64 squares , Al bert Kapengut was one of the strongest players i n Be larus. The development of chess in the country owed m uch to the fact that Isaac Boleslavsky (1 91 9-1 977) , who i n the early fifties was one of the highest-ranked players in the world, had moved to M i nsk. Albert Kapengut fi rst stud ied with Boleslav sky as a 14-year-old and in the course of time he hi mself became a wel l-respected theoret ician . The Russian G randmaster Yuri Razu vaev once said half-jokingly, half-seriously, that Gelfand was the chess g randch ild of Boleslavsky. Beginning in 1 980, it was great luck for Boris that he was able to work to gether with such a strong trainer, theoret ician and player. Boris was granted access to Kapeng ut's enormous l i brary. H i s new trainer was always ready to help the boy; Boris cou ld ask him any question he wished . They analysed Boris's games, paying par ticular attention to the defeats. During the twelve years they worked together, Boris ac quired first and foremost the methodology of studying chess, and he learned how to pro cess i nformation systematically and invent new ideas. Albert also accompan ied and as-

11

The making of a classical grandmaster

The then 54-year-old Yefim Geller won the championshi p , ahead of the n i neteen-year old Artur Yusupov and the 16-year-old Garry Kasparov. Boris collected the autographs of all eighteen partici pants as wel l as that of the ch ief arbiter Salo Flohr, one of the best players in the world before the war. Boris's father recal ls clearly Kapengut's wife Mari etta Golovey saying to Boris: ' Soon people will be aski ng you for your own autograph . ' As i t happened , the fi rst g reat successes came al most im mediately. In 1 983, at the age of 15, Boris won the Sokolsky Memor ial , which had been held annually in M i nsk si nce 1 970, thus repeating the performance of Garry Kasparov, who five years earl ier had won this traditional tournament u nder the sharp eye, it needs hardly be stated , of the ten -year-old kibitzer Gelfand. He was unde feated i n the 1 983 tournament and fi nished ahead of, among others, two grandmasters. Although Boris achieved the national master norm with two poi nts to spare, drawn-out bureaucratic delays meant that he was not awarded the title until 1 985. In 1 983, too , Boris took part i n the Cham pionship of Belarus for the fi rst time. I n both the su bseq uent Championsh ips of 1984 and 1 985 he won the title! I n 1 985 he be came USSR under-18 champion (half a point ahead of Vasily Ivanchu k) and two years later he won the European J u n ior Championsh ip (again half a poi nt ahead of Ivanchuk) . In 1 988 he shared victory in the legendary So viet 'Young M asters Tournament ' . By 1 989 , when he took the bronze medal in the USSR Championsh ip and won gold with the Soviet team i n the European Team Championsh i p , he had achieved international fame.

His rise to world-class status . . . began in that year, 1 989. At Palma de Mal lorca Gelfand won the qual ifying tournament for the G MA World Cup. Th is was a fearful ly

•

His rise to world-class status

strong Open with more than 150 grandmas ters. After such a success it came as no su rprise that he was invited to take part in the legendary super-tou rnament i n Linares. There, in February 1 990, on his debut in An dalucia he met no less a player than Garry Kasparov in the very first round and held him to an exciting d raw. Gelfand then won six of his remai ning ten games and achieved a brill iant second place in his first Linares, on ly half a point behind the reigning world cham pion . Shortly afterwards he shared victory with Vasi ly Ivanchuk in the Manila I nterzonal Tournament, and also won the Olym piad with the USSR team . At his first attem pt to win the world chess crown he overcame Predrag N i kolic of Bosnia 4%-3% (on the tie-break) but then lost 3-5 to the Engl ish player Nigel Short, who was playi ng the world champion ship cycle of his life. Short went on to defeat both Ti mman and Karpov, and , i n 1 993 he took on Garry Kasparov for the world title. U ndeterred by this setback, Boris won the Investbanka Tou rnament i n Belgrade at the end of 1 991 (ahead of Kamsky and Nunn) and i n the new year tournament at Reg gio Emilia he shared second place with Garry Kasparov beh ind the rising I ndian star Viswanathan Anand. I m mediately after his success i n Italy, Boris went on to tie eq ual fi rst with Valery Salov i n the famous trad i tional tournament i n Wij k aan Zee. I n 1 992 he scored a g reat victory in the Alekh ine Memorial Tournament in Moscow, which led many i n the Russian media to com pare his play with that of the legendary fourth world cham pion . Only Anand managed to keep pace with Gelfan d. But first place went to the player from Belarus, who beat the I ndian in their individual game and whose other vic tims in the eight-strong grandmaster field in cluded Anatoly Karpov, who could only man age sixth place. M oscow 1 992 was one of those tournaments in which the pent-up en ergy of Gelfand was fu lly unleashed . H is in exhaustible fig hting spirit is demonstrated

12


not least by the tremendous average game length of 54 moves. Such success was more than sufficient proof that Boris Gelfand, the autograph -collecting you ngster from M i nsk, had become part of the i nner circle of the world 's el ite chess players.

Successful years In the early nineties Gelfand 's creative part nership with Alexander H uzman began . They had already been analysing together si nce 1 990, even though they lived i n d ifferent parts of the USSR. This was only possible because they were able to meet u p i n a sort of training cam p. And when H uzman left for Israel i n 1 992 thei r creative partnership continued . They met even more freq uently after Boris h i mself emigrated to Israel six years later. Today they train together at least three days a week. I n times when almost all the top grand masters change their sec onds freq uently, the unbroken 14 -year sym biotic chess relationship between Gelfand and Huzman is qu ite remarkable. It bore cre ative fru it not j ust for Boris, but also for his second . Thus Alexander H uzman has contin ually pushed his Elo rating beyond the 2600 barrier and he was the fi rst to defeat Alexander Khal ifman after the latter captured the FIDE world title in 1 999. J u st four years later Gelfand 's faithfu l second de feated Garry Kasparov, and well known play ers such as Peter Svid ler, M ichael Adams, Peter Leko, Valery Salov, and Alexey Dreev were also among his victims. For Boris the relationsh ip with his friend , who is by nature somewhat recl usive, and who accompanies him to every tournament, freq uently putting his own ambitions second, has been a great stroke of luck. I n 1 993 Gelfand managed to achieve a u n ique double by w i n n i ng the I n terzonal , run as in 1 990 under Swiss-system ru les, for a second time. Once again he

•

Successful years

scored 9/1 3, but this time he won outright. An outstanding achievement was his fifth victory in im portant tournaments over Viswanathan Anand (alongside three d raws and one de feat). I n the Candidates matches he first de feated Michael Adams 5-3. I n the spring of the same year 1 994 he underl i ned his good form with victory in the tournament in Dos H ermanas, in southern Spai n , where Ana toly Karpov, among others , fi nished beh ind him. Then followed the duel with Vladimir Kramnik. The Russian , who was then j ust 1 8 -year old , was generally regarded as a fu ture world cham pion , especially after Kas parov had descri bed him as a potential suc cessor. Although Gelfand , at the age of 26, could demonstrate a plus i n experience and achievements, Kramnik was publ icly re garded as slightly the favou rite. Or perhaps Gelfand was bei ng u nderestimated - not for the first time and not for the last. Could this be due to the fact that Boris som e times looks l i ke an absent-m inded profes sor? Gelfand ulti mately won 4%-3%, thereby curbing somewhat his good friend Kram nik's precocious ambitions. I n match prepa ration and in the aggression essential to a sporting contest, Kram nik was not yet the equal of Gelfand. To the question put by the magazine Schach: 'was he himself surprised by his match win over Kramnik? ' , Gelfand replied three years later: ' No, I was better prepared than Kramnik. I learned practically all his games by heart. I knew everyt h i ng about h i m . I knew at what point he would sacrifice material , and which moves he pre ferred i n particu lar positions. For two long months I prepared myself to meet Vlad i m i r alone. ' Th is i s the hard worker from Belarus, speaking from the heart. The man from M i nsk was only two steps away from conquering the FIDE world chess throne, which had been vacant since 1 993 , when Kasparov parted com pany with the World Chess Federation . Gelfand 's match with Anatoly Karpov took place in 1 995 i n

13


•

Successful years

.

the Indian town of Sang h i Nagar, where he had also played his match against Kram nik: paral lel with this was the other sem i-final duel, played between Salov and Kamsky. Gelfand went 2-1 ahead before his oppo nent, the 1 2 th world champion i n history, managed to even the scores. I n the end , after losing a more or less eq ual bishop ending in the seventh game, Gelfand cru m bled and succum bed , far too bad ly, 3-6. Th is was especially trag ic, because only a few months earl ier Gelfand had won a tou r nament i n the French town of Cap d 'Agde, thereby qualifying for a match agai nst Kar pov, which he won 4-2 . Did he reveal his cards there too early, before the more im por tant match agai nst the same opponent ? In cidental ly, France is an especial ly successfu l place for Boris i n rapid -play chess. I n Cap d 'Agde, where today he is sti l l a favourite with the public, he won twice - in 1 994 and 2002 , each time defeating Karpov in the fi nal . The rapid-play tou rnament in Monaco, in which , with the exception of Kasparov, the world el ite partici pates reg u larly, ended in victories for Boris in 2001 and 2002 . I n 2002 he took part in the rapid- play com petition against Russia, where he had the second best score for the ' Rest of the world ' , and in 2003 he outclassed Judit Polgar i n a match by 6-2 . I n that same year Gelfand achieved further sign ificant success by winning once again the Belgrade Investban ka Tournament, com ing eq ual fi rst with Kram nik ahead of Shirov, Topalov, Ivanchuk, Timman , Adams, Leko and Ljubojevic. Belgrade is associated by Gelfand with particularly pleasant memories. For years he has worn the colours of the Serbian capital 's club, Agrouniverzal Zemun. Several years later, i n Apri l 1 999, as NATO bom bs fell on Serbia, he poi nted out, in a courageous col u m n i n the German daily newspaper Die Welt, the special contri bu tion which the former Yugoslavia had made to the development of chess : 'Al l the tou r-

14

naments to which I was invited i n the ter ritories which then constituted Yugoslavia, were organised to an extraordinari ly high level. Every day somewhere between 2 , 000 and 4 , 000 spectators pour i nto the tourna ment hall of the trad itional Belgrade Invest ban ka Tournament. These are figures which other cou ntries hold ing chess tournaments can only dream about. Television daily de votes 45 m i n utes o f prime time t o t h e tour nament ; each of the major daily newspa pers gives over a whole page to chess. The games of my world cham pionship Candi dates match agai nst Predrag N i kolic in 1991 in Sarajevo (today Bosn ia-Herzegovina) were broadcast live, for the first time i n the h is tory of chess. The beg i n n i n g of the war in Yugoslavia caused great anxiety to chess players throug hout the world, because for us Bugojno, Niksic, Banja Lu ka, Pula, Belgrade and others are not just sim ply names on a map, but rather places in which we had been and where we had many acq uai ntances and supporters. ' The year 1 996 also brought two i m po rtant successes for Gelfand . In Vienna, thanks to the better tie- break, he won ahead of Kar pov, whom he defeated i n their ind ividual game, and Topalov, and left among others Kram nik, Shirov and Korchnoi i n his wake. He did the same at Ti lburg to Piket : the de feated field included , among others, Shirov, Karpov and Adams. Gelfand had finally established h i mself among the world 's top chess players. A flood of invitations to all the world -class tourna ments of 1997 was the log ical consequence : Li nares, Dos Hermanas, N ovgorod , Dort m u n d , Biel, Polan ica Zdroj , Belgrade, went the programme up to and including Novem ber. M id -tournament placings were all he could manage, not a single tournament vic tory. ' I sim ply do not have the energy, be cause I play too much , ' he was forced to admit. Fatigue led to a loss of technique and

The maki n g of a classical grandmaster .

many m i ssed chances. Previously he had turned down invitations in order to focus on preparation for the world cham pionsh i p Can didates matches. B ut FIDE had now abol ished the trad itional cycle of Cand idates matches in favour of a World Championsh ip knock-out tournament, which was staged for the fi rst time in 1 997 in the D utch town of Groningen. Gelfand sum moned up the energy to reach the sem i-final of this ex hausting mammoth tournament. There he lost a m i n i -match against Anand by %-1 %. His earl ier outstand ing record of success against the I n d ian player was reversed and his third assau lt on the world title was once again brought to a standstill just a few short steps from his goal . The year 1 998 saw changes in Gelfand 's life. His marriage in Septem ber 1 995 to a Belgian woman , which had meant spending a great deal of time in Brussels, came to an end . He settled permanently i n Israel , i n Rishon - I e Zion , a town south o f Tel Aviv. H is g lobe trotting life -style had tem porari ly abated , not least because his lower Elo rating , after the setbacks of 1 997 , meant fewer i nvitations to the big tournaments. B ut he managed what was a fantastic comeback in Polanica Zd roj , at the Memorial Tournament for Akiba Rubinstei n , one of his most revered heroes. The monograph written by Razuvaev about the great Polish master was descri bed by Gelfand , a lover of classical l iterature, as his favourite chess book. Gelfand 's game with Alexey Shirov, whom he beat by one point and forced i nto second place, is one of the best games ever played on a chess board. The fol lowing year, a s a sort of trai ning ru n for the World Championsh i p i n Las Vegas, Gelfand won a category 14 tournament in Malmo as well as the super-cham pionsh i p of Israel . But his l uck ran out at the knock-out World Championship in the gambler's para dise of Las Vegas and he lost, at the last sixteen stage, to the eventual title-holder, Alexander Khal ifman .

•

Successful years .

The m i l len n i u m began with u neven resu lts at Lvov and Biel, before a tournament vic tory at Polan ica Zdroj fu lly restored his con fidence. Once again he fi n ished ahead of a strong field, which trai led to the fi n ish ing line behind the second-placed Alexey Shirov. It was the same Shirov, who elimin ated Gelfand at the last-sixteen stage of the knock-out World Cham pionsh i p tournament in New Del h i . Earl ier Boris had , of course, managed to get through to the sem i-fi nals of the inaugural and strong ly-contested FIDE World C u p , held i n the Chi nese town of Shenyang . Viswanathan Anand was the op ponent waiting to do battle with h i m . This tournament was nonetheless seen as a suc cess for Gelfand . I n the year 2001 a + 1 re sult i n the six-player tou rnament at Kaza khstan 's capital Astana, in addition to quali fying for the q uarter-fi nals of the knock-out World Championsh i p i n M oscow, stand on the cred it side. Here Gelfand lost out to the R ussian , Peter Svid ler - after a tru ly titan ic struggle. I n the ensuing tie-break, watched by more than a hundred spectators inside the Moscow Krem l i n , determ ined to see the match out to the end , he saved a rook ver sus q ueen endgame. Svid ler, faced by an opponent defending like a computer, felt ab sol utely crushed , but Boris, due to his own exhaustion , cou ld not take advantage of the psycholog ical advantage thus obtai ned and went out in the bl itz-game phase of their match. He has always made it clear, that he is by no means a supporter of this lottery-like deciding phase, which has been used si nce the Moscow tournament, especially si nce it has operated alongside the time reductions i ntroduced by FIDE. Boris has campaig ned for the rei ntrod uction of classical time con trols. More than once he has spoken out in the press, vigorously and convi ncing ly, for the withdrawal of the cou ntless FIDE in nov ations. Although the new system does not suit his style of play at al l , he has managed i n fou r appearances in the knock-out World

15


•

Successful years

Topalov in the second round broke h i s re solve. I n this game Gelfand , next to Kas parov the greatest exponent of the Najdorf in the world, found hi mself playing the Caro Kann for the first time in his life and he man aged to convert a su perior position i nto a loss. This remained for the time being his last attem pt at the world title, because, as Gelfand h i mself said, ' U nfortunately, F I DE organ ised the 2004 world championsh i p i n Li bya, a country where the authorities stated that players from Israel wou ld not be ad m itted . A shamefu l act i n the opinion of many chess players, spectators and organ isations !' But further opportun ities will occur in the future, even if the situation in the chess world remains u npred ictable.

Championships to always make the last six teen . The year 2002 began bri l l iantly with a fabu lous joint victory with Topalov in the category 18 tournament at Cannes (ahead of, among others, Bareev, Karpov, Leko , and Moroze vich). At the Candidates Tournament in Dort mund that summer, a fatefu l time for classical chess, Boris was hoping to launch yet an other bid for the world chess crown, which his friend Kram nik had held for the previ ous two years. Boris prepared himself espe cially precisely. He spent the whole of J u ne with his faithful second , Alexander H uzman , at train i ng sessions i n Russia and Austria. Gelfand was fleeing the u n bearable heat visited on Israel at the time and not the someti mes war-like circumstances in his new homeland . At the end of M ay i n that year a bomb had exploded i n Rishon-Ie -Zion , j ust a kilometre from Boris's house. The bom b went off at n i n e o'clock i n t h e even i n g , i n a park just opened b y t h e mayor, a n d des ignated for the use of chess and draug hts players, who were therefore present in con siderable n u m bers. There were two deaths and many i nj u red . Clearly the series of sui cide bom bings left no one unaffected and they were seriously discussed i n the closed chess circles made up largely of immigrants from the former Soviet Union.

An i nteri m assessment of Gelfand 's chess career to date makes an extremely posi tive impression . The boy from M i nsk has grown up, within a space of only fifteen years, to become one of the leading figu res on the world chess scene and he was the fifth player to overcome the 2700 Elo bar rier. Only a handfu l of players can point, i n t h e period between 1 990 a n d t h e present time, to a g reater n u m ber of first places i n i m portant i nternational tournaments. Boris Gelfand has enriched chess literatu re with so many marvellous games, that choosing a selection for this book must have been very d ifficult for h i m . He has always been, and remains, a leading exponent and defender of the classical approach to the game and as such now finds himself wel l on the way to becoming a classical player himself.

At the Candidates tournament i n Dortm und Boris did all he could to secure the best possible playi ng conditions for h i mself. I n add ition t o Huzman he recru ited t h e for mer Soviet national team trai ner Postovsky, who had emigrated to the USA, to join 'team Gelfand ' . It is probable that his defeat to

Dirk Poldauf (translation by Brian Ings)

16

My Fa vourite Va riation of fashion and I can name j u st my friend (then also a young and promising player) Alexander Khal ifman , who kept on believ ing i n the advantages of this line. You can see us analysing this variation in the picture on page 158.

Very often a chess player has an open ing or a variation in his repertoire, which he enjoys playi ng most of al l . For me this is the 8 '!;b1 line in the G ru nfeld Defence. My attention was drawn to it some time i n 1 982 , soon after it was invented . In 1 982 I played it for the first time and won against Leonid Basin. I analysed it a g reat deal i n the sum mer of 1 983. In fact, from the present-day viewpoint it can hardly be cal led 'analysed ' .

I really enjoyed these positions and the more time I spent analysing them, the more confi dence I had that Wh ite's strong centre - his main trump in this variation - would give him a favourable position .

I selected all the games played i n this line (there were only a few at that ti me) and tried to develop my opinion on them and on the resulti ng positions. I n 1 983 I played it agai n and won a game agai nst Boris Itkis, the first I ever annotated for Chess In forma tor.

My faith i n this variation has not only given me a lot of pleasure during analysis, but has also rewarded me with a number of memor able victories against some of the strongest players in the world, which I hope you will enjoy i n this chapter.

Duri ng the next few years 8 '!;b1 went out

* * *

Game 1

B o r i s G e l fa n d - I o s s i f D o r f m a n M i n s k 1 9 86 GrOnfeld Defence [085J I consider this game to be one of the most important i n my chess career. I was al ready a strong player, I had won the J u nior Championship of the Soviet Union (ahead of Ivanchuk, among others) , but I didn't have an international rating and this was my fi rst in ternational tournament, with almost no hope of playing i n any others i n the foreseeable future. After a shaky start I won five games in a row and I needed 1 V2 poi nts from my last two games to make the grandmaster norm . I had to play Dorfman with Wh ite and 1 M Ste fan Gross with Black. Some of my friends advised me to make a draw with the expe rienced grandmaster and try to win against the outsider i n the last round. However, the tem ptation to cross swords with the strong

theoretician , who at that time was one of Garry Kasparov's seconds, i n the G ru nfeld variation with 8 '!;b1 , was so strong that I didn 't even take any prag matic factors i nto consideration .

1 d4 ti}f6 2 c4 g6 3 ti}c3 dS 4 ti}f3 fig7 5 cxd5 ctlxd5 6 e4 ctlxc3 7 bxc3 c5 8 gb1 0-0 9 fie2 cxd4 10 cxd4 .as+ Th is idea was introduced by Vi ktor Gavri kov in his game with M i khail G u revich i n Riga 1985 (52n d USSR Cham pionsh ip) and imme diately it became clear that it had to be taken seriously. Now I bel ieve that it is the most chal lenging l i ne. Straig ht after the cham pionsh ip, G u revich together with Alexander Chern in gave some lectures to the best So viet j u n iors (Ivanc h u k, Dreev and Smirin to name just a few) . And every even ing I had the honour of joining them in their room to ana lyse this new variation . We tried hard to fi nd any advantage in the endgame after 11 �d2

17

My Favourite Variation

�x d2+ 12 A xd2, but our 10 days of work didn't produce any specific result (and years of practice showed that Wh ite cannot hope for an edge after the q ueen exchange) . On the last day I suggested that we should ana lyse the pawn sacrifice, but the training ses sion was over and we had to do this job separately. Only i n the early 1 990s did I do some joint analysis with Alexander Chern i n , t o t h e great benefit of both of us.

Both 1 7 d5 A xd5 1 8 gxe7 �c6 1 9 A xd5 �xd5 20 ge2 CDc6 21 A xf8 gxf8 and 1 7 A x e6 fxe6 1 8 gxe6 �d5 1 9 gxe7 CD c6 20 gxg7+ c!lx g7 21 A xf8+ gxf8 22 �e3 would have led to an equal position , but you don't start play on a grand scale in order to turn into qu iet waters immediately. 17 . . . fxe6 18 �g5 �h8 1 8 . . . A h6? 1 9 A x e6+ c!l h8 (19 . . . c!lg7 20 d5 CD a6 21 Ac3+ gf6 22 h4) 20 d5 CD a6 (20 . . . Ag7 ?! 21 CDxh7 c!l xh7 22 gb3) 21 Ac3+ Ag7 22 Axg7+ c!l xg7 23 CDx h7 c!l xh7 24 gb3 and wins. 1 8 . . . CD c6 !? seems bad , but some Swed ish players found beautiful a tactical idea and rehabil itated the whole lin e: 1 9 CD x e6 (19 A x e6+ c!l h8 20 CD x h7 CD x d4!) 1 9 . . . c!l h8 20 Ac3 Af6 21 �h6 (21 g4 gfc8 22 g5 CDxd4 23 CD xd4 gxc4 24 g xf6 exf6 oo) 21 . . . gf7 (21 . . . gg8 22 ge1 ! ±, 22 . . . gg7 23 g4 ! +- CDa5 24 A d3 �c6 25 A a1 gf7 26 g5 Ag7 27 d5 �xd5 28 Axg7+ c!l g8 29 Axg6 1-0 Vaisser Pei n , Budapest 1 989) 22 CDc5 �c8 23 Axf7.

11 .td2 .xa2 1 2 0-0 b6 13 .c1 .e6 14 .tc4 .xe4 1 5 Ele1 ! N A new plan , invented over the board . I knew the game Toshkov - Kirov, (Al bena 1 985) 15 A xf7+ gxf7 16 �xc8+ gf8 17 �c4+ e6 18 gb5 �c6 1 9 � b3 CD a6 20 gc1 �d7 21 CDe5 A x e5 22 gxe5 CDc7 23 �g3 with some com pensation for a pawn, but I was young and ambitious and I went for a more chal lenging continuation .

15

...

• b7

1 5 . . . �c6 and 1 5 . . . �f5 were the other op tions.

..

1 6 .tb4 16 A h6 is another possibil ity.

1 6 . . . .te6

8 7 6 5

4

,...� .. i...o-� ,.......0--

�)i:&��d.I

2

L...-_______--::.__-I

'if

1 6 . . . Af6 ?! seemed dubious to me and later practice confirmed this assessment.

17 Elxe6

23 . . . � f5 ! N (23 . . . bx c5 24 A xg6 ±) 24 �c1 (24 A xg6 �xg6 25 �xg6 h x g6 =+=) 24 . . . bxc5 (24 . . . CD x d4! Ivanov) 25 d xc5 CDd4 26 Ac4 �xc5 27 A b4 =+= S. Ivanov- Svensson, (Swe den 1 999). 19 �xe6? Tem pti n g , but not the best. As I found out after the game, both 19 gb3 and 19 �e3 are stronger. I will just g ive some l i nes from my 1 986 annotations and some games where these ideas were put i nto practice. . 19 gb3 : A) 1 9 . . . CDd7 ?! :

18

Game 1

Gelfand - Dorfman , Mi nsk 1 986

A1 ) 20 tD x h7 � x h7 21 .§h3+ (21 A x e6 tDf6 22 .§h3+ tD hS 23 � b1 A xd4 24 .§xhS+ � g7 2S .§gS .§f6 +) 21 . . . � g8 22 A xe6+ .§f7 23 �e4 .§e8 ! 24 Axf7+ �f8 -+ ; A2) 20 Axe7 A21 ) 20 . . . .§fe8 21 .§h3 hS 22 tDf7+ : A211 ) 22 . . . � h7 ?

23 .§x hS+ ! gxhS 24 �b1 + � g8 2S Axe6 tDf8 26 tDgS+ � h8 27 �h7+ ! tD xh7 28 tDf7+ � g8 29 tDeS+ � h8 30 tDg6#.

A212) 22 . . . � g8 23 tD d6 �dS 24 tD xe8 .§xe8 2S .§e3 .§xe4 26 .§xe4 ± ; A22) 2 0 . . . A xd4 2 1 .§h3 tDf6 2 2 A xf8 .§xf8 23 tD xe6 ± ; A23) 2 0 . . . .§fS 2 1 A xe6 .§xgS 2 2 A xgS ;1; ; A3) 20 .§h3 hS (20 . . . tDf6 ? 21 � b1 tD hS 22 .§x hS .§f6 23 .§x h7+ � g8 24 � b3 �e6 2S 'ff h3 1-0 Vaisser-Andrianov, Naberezhnie Chelni 1 988) 21 Axe6 .§ae8 (21 . . . tDf6 22 'ff b1 'ffe7 23 .§e3 �f4 24 �xg6 +-) 22 � b1 tDf6 23 A xe8 !? (23 g4 A h6 !) 23 . . . .§xe8 (23 . . . 'ffx e8 24 Axe7) 24 g4 tDxg4 (24 . . . A h6

2S tDf7+ � g7 26 tD x h6 � x h 6 27 gS+ ! +-) 2S .§xhS+ A h6 26 .§h4 ! +- ; 8) 1 9 . . . .§c8 ! 20 .§h3 (20 tDf7+ � g8 2 1 tD d6 .§xc4 22 �x c4 exd6 23 �x e6+ �f7 -+ ; 20 tD x h7 �dS 21 .§h3 � g8 -+ ; 20 AeS tDd7 21 A x e6 hS 22 � b1 tDf8 23 tDf7+ � h7 24 .§h3 Af6 2S �c1 tD x e6 26 � h6+ � g8 27 'ffx g6+ � f8 28 tD h6 Peek-Timmer, Dieren 1 988) 20 . . . .§xe4 (20 . . . h6 21 tD xe6 +- ; 20 . . . hS 21 .§x hS+ g x hS 22 �e2 +-) 21 .§x h7+ � g8 22 �f4 (22 �xe4 �e6) 22 . . . Af6 23 .§h8+ � g7 24 .§h7+ � g8 = (Gelfand/Kapengut) 2S A x e7 .§e1 + 26 �xe1 A xgS 27 �xgS �e4 (27 . . . � x h7 28 �f6 tD d7 29 �f7+ � h6 30 h4! +-) 28 .§g7+ � x g7 29 �f6+ � g8 30 �f8+ � h7 31 �f7+ � h6 32 Af8+ � gS 33 h4+ ! � x h4 34 � h7+ � gS 3S Ae7+ � f4 36 � h4+ 1-0 Touzane- Kouatly, (French Team Championshi p 1 993) . 1 9 �e3 ! : A) 1 9 . . . tDd7 20 Axe7 tDeS (20 . . . tDf6 21 Axf8 .§xf8 22 �x e6 ±) 21 A xf8 tD x c4 22 A x g7+ � x g7 23 �x e6 .§c8 24 dS ± (24 � h3 h6 2S tDe6+ � h7 26 .§b3 !? (threatening 27 �x h6+ !) 26 . . . hS oo); 8) 1 9 . . . tDe6 20 tD x h7 ! (20 �x e6 (th reaten ing 21 �g8+) 20 . . . .§f6 2 1 � h3 h6 22 A e3 'ffe8 23 Ae6 .§xe6 24 tD xe6 tDd8) 20 . . . � xh7 (20 . . . tD xd4 2 1 tD g S ! tDfS 22 � h3+ tD h6 23 Ae3 .§f6 24 tDf7+ � h7 2S tD x h6 A x h6 26 Ad2 gS 27 A xgS +-) 21 � h3+ A h6 22 Ad2 gS 23 A xgS .§f6 24 Ad3+ � g7 2S Axf6+ exf6 26 'ffx e6 'ffe8 : 81 ) 27 �e4 fS 28 �dS oo ; 82) 27 �d S ! N tD e7 2 8 �e4 � f8 29 .§e1 �e8 (better 29 . . . �d8 - Pein) 30 A bS ! �d8 31 �e6 Ag7 32 Ae4 �e8 33 h4 in tending hS-h6 +- Pei n - M iehaelsen , (Ramat Hasharon 1 988).

19 �d7 19 . . . tD e6 20 Ac3 transposes i nto the 18 . . . tD e6 line (20 tD x g7 tD x b4 21 � h6 �e4 -+ ; 20 tDxf8 .§xf8 21 AdS .§f6 !) . ••.

20 Axe7?

19


20 �xf8 was better, and if 20 . . . �xf8 21 'tWe3 Af6 22 dS 00 . Later 20 . . . �xf8 ! was tried : 21 'tWe3 �c8 22 Ae6 Af6 23 g4 �c7 24 gS Ag7 2S A xe7 (2S dS �f8 Gonzalez Perez, Havana 1992) 2S . . . �eS 26 dxeS �xe7 27 �d1 'tW b8 28 �d6 and I prefer Wh ite.

20 . . . IUc8

6

��

5

4

4

3

3

2

2

24 �d6? 24 AgS was better: A) 24 . . . �xc4 2S �xc4 bx c4 26 �f6 �fS (26 . . . �xd4 27 � xd7 �xd7 28 Af6+ � g8) 27 � xd7 (27 g4 !? �f3 ! -+) 27 . . . �xd7 28 Af6+ � g8 29 �x c4+ �f7 30 �c6 �f8 31 A h4 + ; B) 24 . . . �xe8 2S �xbS �e1 + 26 �xe1 ,§,xe1 + 27 Af1 '§'b8 28 '§'x b8+ � x b8 29 dS '§'d1 +.

6

5

26 'tWf4 !) 26 h4 with counterplay, but 23 . . . �e6 ! was wi nning.

24 . . .•xe7 The game is over. Wh ite should resig n , but both players were i n time trouble and so he conti n u ed playi n g. Probably this is not an excuse, but merely an explanation .

25 �xc8 2S �f7+ �xf7.

21 �xg7?

25 . . .•x b4 26 Ad5 Hb8 27 h4 .xd4 28 .c6 .c5 29 .xd7 gxc8 30 g3 Hf8 31 Af7 .c3 32 h5 .16 33 h xg6 h xg6 34 Ad5 .xf2+ 35 �h1 .f1 + 36 �h2 .f2+

During the game I thought that 21 �gS �eS ! 22 dxeS 'tWxe7 23 �f7+ 'tWxf7 24 Axf7 �xc1 + 2S �xc1 AxeS was easily winning for Black, but now I doubt whether this is so : 26 A e6 as 27 �c8+ �xc8 28 A xc8 with d rawing chances.

36 . . . ,§,f2+ was mate in 2 .

21 . . . •c6? 21 . . . 'tWe4 ! would have won on the spot.

37 �h3 .f5+ 38 .xf5 gxf5 White resigns

22 gb4 b5 23 �e8?! After 23 'tWc3 as ! or 23 'tWa1 bxc4 24 dS 'tWxdS 2S �e8+ c3 B lack is w i n n i n g , but 23 �fS !! was shown by the com puter to be an excellent defence: 23 . . .'�V e4 24 �a1 gxfS (24 . . . �xc4 2S �xc4 bxc4 26 dS+ �eS 27 'tWxeS+ �xeS 28 Af6+ � g8 29 �h6+ �f8 30 A xeS as 31 � g4 a4 32 � e3 a3 33 � f1 ) 2S dS+ 'tWeS (2S . . . � g8 26 Aa2 'tWeS 27 'tWf1 with counterplay) 26 'tWxeS+ � x eS 27 Af1 � g8 28 �x bS �c1 and Black i s better, but Wh ite retains cou nter-chances.

I real ly enjoyed havi ng played a fighting game, fu l l of errors, agai nst such a player and I didn't regret at all the fact that I missed the grandmaster norm . I spent the following month analysing this game alone or together with my trainer Alburt Kapengut. We p u b l ished a com plete version o f t h e analysis i n New in Chess, which was not a practical de cision , as after this for a lengthy period no one played the G rO nfeld against me. B ut I am proud that, in the following 1S years, very few i m provements were found and a n u m ber o f games were w o n on t h e basis o f o u r analysis.

23 . . . •e4?! If 23 . . . bxc4 24 � d6 (24 dS 'tWxdS 2S 'tWc3+ � eS) 24 . . . �cb8 2S �xc4 'tWdS (2S . . . 'tW b6

* * *

20

Game 2

ttJ

Gelfand - Ftacnik, Gideon Barcza Memorial Tournament, Debrecen 1989

Game 2

�xg7 19 0-0 gf8 20 gxf4 gxf4 21 't'fxf4 't'ff6 22 't'fe4 gb8 23 gf1 't'fd4+ 24 't'fxd4+ cxd4 25 gb1 Ad7 26 Af3 b6 27 c5 gc8 28 c6 Axc6 29gc1 Ad7 30 gxc8 Axc8 31 Ac6 �f6 32 d7 A xd7 33 A xd7 e5 34 � f2 e4 35 A c6 � e5 36 h4 ! (it is i m portant to keep the h - pawn on the board) 36 . . . h6 37 g3 g5 38 h5 ! � f5 39 a4 � e5 40 A b7 ! and White converted his advantage i nto a win (Novi kov -Tu kmakov, 51 s t USSR Championsh ip, Lvov 1 984).

B o r i s G e l fa n d - L u b o m i r Ft a c n i k G i d e o n B a rcza M e m o r i a l To u r n a m e n t , D e b recen 1 9 89 GrOnfeld Defence [085] I eagerly awaited this duel with a strong the oretician , who had been playing only the GrOnfeld throughout his career.

1 d4 �f6 2 c4 g6 3 �c3 d5 4 cxd5 �xd5 5 e4 �xc3 6 bxc3 c5 7 �f3 Sig7 8 l:lb1 0-0 9 Sie2 �c6 This plan is the main alternative to 9 . . . cxd4 and it was developed mai n ly thanks to the efforts of Czech and Slovak players. Black's idea is first to blockade and then attack White's central pawns.

10 d5 �e5 Igor Stohl accepted the challenge and took the pawn - see below on page 24.

11 �xe5 Sixe5 12 .d2

1 3 f4 Sig7 14 c4 e5 Black is playing for a blockade, in the spirit of this line. 14 . . . e6 15 A b2 is i n Wh ite's favour, as practice showed i n later years. 1 5 0-0 f5 Trying to provoke a crisis i n the centre. As later games by Ftac n i k showed , B lack can get a good position by 15 . . . exf4 16 't'fxf4 't'fe7 1 7 A b2 A d7 1 8 Ad3 A x b2 1 9 gx b2 f6 20 Ac2 gae8 ! (20 . . . �e5 21 't'fx e5 fx e5 22 gfb1 ! i ntend ing a4-a5 ± ; Wh ite needs both his rooks to break through , while the black rook has noth ing to do on the f-fi le) 2 1 ge1 't'fe5 22 't'fx e5 gxe5 23 a4 f5 with equal chances (Sakaev - Ftacnik, Dortmu nd open 1 992) . 1 6 Sib2 An interesting alternative was 16 d6!?, and if 16 . . . A b7 (16 . . . Ae6 !?) 17 A b2 ! Axe4 18 fxe5 A x b1 (18 . . . ge8 19 Af3 A x b1 20 A x a8 (or 20 gx b1 ) 20 . . . 't'fxa8 21 gx b1 ±) 19 gx b1 with complete domination . 1 6 •d6 . . .

a b c d

12 . . . b6

e 8

Lubom ir chooses a line which had not been explored at the time of the game. I think that this variation was the first one where opening theory developed so deeply with concrete analysis. In 1 984 Igor Novi kov impressed the chess world with a novelty on the 36 th move : 12 . . . e6 13 f4 Ag7 (the attempt to play subtly with 13 . . . A h8 14 c4 ge8 15 e5 f6 is refuted by 16 f5 !! with a strong attack) 14 c4 ge8 15 e5 f6 16 d6 fxe5 17 A b2 exf4 18 A xg7

6 5 4 3 2

�------��- �

21


20 fxe5!

16 . . J::1 e 8 17 d6 ± ; 16 . . . exf4 17 A xg7 � x g7 18 e5 ! .

White has to open up the game if he wants to fight for something. 20 Ac2 e4 21 �g3 �g6, or 20 '8f3 e4 21 '8g3 � h6 (i ntending . . . Ad7, . . . '8f8 and . . . � h8) is unclear.

17 .c3 Wh ite should not al low simpl ification i n the centre.

17 . . . lle8

20 . . . Axe5 21 .d2

Black intends . . . '8e7 and . . . exf4. White must prevent this.

Not 21 �c2 ? A x h2+ 22 � h1 �h6.

21 . . . Axh2+

18 Ad3

21 . . . A x b2 22 '8x b2 would be positional ca pitulation , in view of the weakness at f5.

The only way to fight for an advantage. After 1 8 '8 be1 fx e4 1 9 Ad1 '8e7 20 '8x e4 exf4 21 '8exf4 �xf4 22 �x g7+ '8xg7 23 '8xf4 '8f7 Black stands better, while if 1 8 g4, then 18 . . .fxe4 (18 . . . '8e7 19 gxf5 g xf5 20 � h1 ) 1 9 f5 gxf5 20 g xf5 Axf5 ! .

22 �h1 Ae5 22 . . . Ag3 23 A xf5 A xf5 24 '8xf5 24 . . . '8f8 (24 . . . A h4 25 �c3 +-) 25 '8g5+ � f7 26 Ag7 and Wh ite attacks without even giving up a pawn.

1 8 . . . lle7? A serious mistake, but this was hard to foresee, as now a long , almost forced line beg i ns, lead i ng to an advantage for White. Preferable was 18 . . . fxe4 19 Axe4 '8e7 (threaten i ng . . . exf4) 20 f5 gxf5 21 A xf5 e4 and after the exchange of both pairs of bish ops White can count on only a slight edge.

23 .g5+ Not 23 '8be1 A x b2 24 �g5+ ? '8g7 , while 23 Axf5 Axf5 24 '8xf5 Axb2 25 �xb2 �h6+ is equal .

8

19 exf5! 1 9 '8 be1 ? would be a mistake: 1 9 . . .fxe4 20 '8xe4 exf4 2 1 '8exf4 �xf4 ! with a clear advantage to Black.

6

6

Io--"("---'-=-f'"-:::-"I ,......-.,

5

��

4

19 . . . gxf5

5 4

3

Or 1 9 . . . e4 20 f6 exd3 21 fxe7 ! and wins. After 1 9 . . . A xf5 20 A xf5 g xf5 21 fxe5 A xe5 22 � h3 ± the weakness of the f5 pawn de termi nes the assessment of the position .

23

8

. . .

•g6!

A bri lliant defence ! Other moves lose qu ickly: 23 . . . '8g7 24 A x e5 �x e5 25 �d8+ � f7 26 '8be1 �g3 (26 . . . '8g8 27 �xg8+) 27 �e8+ �f6 28 '8e6+ Axe6 29 �xe6+ �g5 30 '8xf5+ with mate, 23 . . . Ag7 24 '8 be1 , or 23 . . . � h8 24 '8 be1 '8g7 (24 . . . Ad7 25 '8xe5 '8xe5 26 '8e1 ) 25 '8xe5 ! or 25 �xg7+.

24 .xe7 '--_______""""_----' "-

if

24 �xg6+ h xg6 25 '8fe1 '8h7+.

22

Game 2

ttJ

Gelfand - Ftacnik, Gideon Barcza Memorial Tournament, Debrecen 1 989

A) 30 . . . h6 31 �g3+ � h7 32 �f1 (32 �e1 ? � h4+) 32 . . . hS (32 . . . � h4+ 33 gh3 �g4 34 gff3) 33 �xfS ! A xfS 34 A xfS+ � h6 3S gg6+ ; B) 30 . . . hS 31 gg3+ � h7 32 gf1 h4 (32 . . . YWd6 33 �xfS A xfS 34 A xfS+ � h6 3S Ag7#) 33 gg4 Ad7 (33 . . . � h6 34 Ac1 + � hS 3S gg8 ; 3 3 . . . � h6 3 4 A xfS � hS 3S �g8 �d6+ 36 � h3) 34 �xfS A xfS 3S A xfS+ � h6 36 Ac1 + � hS 37 � h3 �f6 38 Ag6+ ! �xg6 39 gxh4#; C) 30 . . . Ad7 31 �bf1 hS 32 �g3+ � h7 33 gxfS ! ; 0) 3 0 . . . � h4+ 3 1 gh3 �f4+ 3 2 gg3+ � f7 33 gf1 .

24 . . .• h6+ 25 �g1 .e3+ Th is is the point. After 2S . . . � h2+ 26 � f2 YWf4+ 27 � e2 Wh ite is winning.

26 �h1 .h6+ 27 �g1 .e3+

8

6

6

5

5

4

4

3

3 2

L..._______ ....._----I ..:;....

'lJ

8

28 11f21 6

Anyway! Even without his queen , Wh ite has a strong direct attack on the black king, which lacks proper cover. 28 � h1 leads on ly to a draw.

5 4 3

28 . . . .th2+ 29 �xh2 .xe7 30 I1f31 The strongest. Th is rook starts an i mmedi ate attack by going t o g3, while the other wi ll join in from f1 . 30 �e2 ?! was less good in view of 30 . . . � h4+ (or 30 . . . �d6+ 31 � g1 Ad7 32 �be1 �f8) 31 � g1 Ad7 32 �be1 �f8 33 d6 (or 33 �e7 �f7 34 d6 YWg3) 33 . . . �f6 ! ' However, 3 0 �bf1 !? was a dangerous alter native: A) 30 . . . Ad7 31 �f3 ; B) 30 . . . �d6+ 31 � g1 h6 32 A xfS A xfS 33 gxfS ± ; C) 30 . . . � h4+ 3 1 � g 1 Ad7 3 2 A xfS A xfS 33 gxfS �e8 (33 . . . h6 34 d6) 34 d6 with an attack.

30 . . .•d6+ As the fol lowing variations show, Black is helpless against the powerful attack of White's four coordinated pieces :

2

c d

e

f

9

h

'lJ

31 11g3+ 1 White need not fear the pin, as calculations show that he gives mate i n time. 31 . . � 32 I1f1 h5 The alternatives were no better: A) 32 . . . f4 33 Ac1 ; B) 32 . . . � e7 33 �e1 + � f8 (or 33 . . . � d8 34 AeS � h6+ 3S � g1 �f8 36 gg7 Ad7 37 gx h7 Ae8 38 Ac7+ � c8 39 Ad6) 34 Ag7+ � f7 3S AeS � h6+ 36 gh3 �d2 (36 . . . �g6 37 Ae2) 37 gee3 ; C) 32 . . . �h6+ 33 gh3 �g6 34 gff3 followed by gfg3 ; 0) 32 . . . Ad7 33 A xfS and now : 0 1 ) 33 . . . A xfS 34 �xfS+ � e7 3S �eS+ � f7 (3S . . . � d7 36 gg7+ � c8 37 � g1 ) 36 � g1 gf8 37 gg7+ � xg7 38 ge6+ ; .

23


02) 33 . . . � e7 34 E1e1 + � d8 3S AeS 'fi' h6+ 36 E1h3 'fi'd2 37 E1x h7 ! 'fi'x e1 38 Af6+ � c7 39 E1xd7+ � b8 40 E1h7 fol lowed by Af6-h4g3 ;

33 .txf5 h4 34 .tg6+ Not 34 Axc8+ ? � e8. Black never gains the tempo needed to win the rook. 34

. • •

3S Af6+ ! � d7 (3S . . . 'fi'xf6 36 E1e3+) 36 Ax h4 � c7 37 E1f7+ (37 AgS 'fi'xg6 38 Af4+) 37 . . . � b8 (37 . . . Ad7 38 AfS) 38 AgS! followed by Af4. 35

.lh7+ ! �xh7 36 IU7+ �h6 37 .tc1 +

Black resigns

�g8

Th is victory provided a good boost to my confidence. I won this tournament and ex ceeded my final g randmaster norm by one point.

I am grateful to Lubomir who allowed to me to demonstrate a beautifu l combination on the board . After 34 . . . � e7 the win is prosaic

*** Black accepts the pawn sacrifice (cf. note to Black's 1 0 t h move)

a b 8 7 6 5 4 3

Boris Gelfand - Igor Stohl I nterpol is, Ti lburg 1 992

6 5 4 3 2

� 1

{(

17 d6! The queen enters the attack.

17 . . . b6 Black cannot protect all his pawns, so h e g ives u p t h e central o n e, hoping t o set u p a blockade on the dark squares.

Position after 10 dS

18 ed5+ �g7 19 exe5+ ef6 20 eh2 h5 21 eg3 �c6!?

10 . . . .txc3+ 11 Ad2 .txd2+ 12 exd2 �a5 13 h4!

Bringing the knight back i nto the game.

Wh ite needs to start a d i rect attack agai nst the king .

22 Dd1 !

1 3 . . . Ag4 14 h5 Axf3 1 5 gxf3

22 E1x hS?! allows 22 . . . E1h8, and if 23 eS 'fi'e6 24 f4? E1x hS 2S AxhS 'fi'fS -+ .

It is i m portant to keep the as knight out of the game ; if 1S A xf3 �c4.

22 . . . �h6 22 . . . 'fi'c3+ ?! 23 � f1 � d4 24 E1x hS +-, or 22 . . . E1ae8 23 E1xhS E1h8 24 d7 ! E1d8

15 . . . e5 16 hxg6 fxg6 (see next diagram)

24

Game 2

tt::J

Gelfand - Ftacnik, Gideon Barcza Memorial Tournament, Debrecen 1 989

83) 24 . . . � c2+ 25 � d1 �d4 26 f4 ! � x e2 27 � x e2 gae8 (27 . . . �xf4 28 ghx h5+ g x h5 29 ge6+ and wins) 28 gd1 with the initiative.

(24 . . . gxh5 25 d x e8� + !) 25 gd6 +- . I nter esting is 22 . . . � d4 23 gx h5 gh8 24 gd5 ! �e6 ! with counterplay, but not 24 . . . gh1 + 25 � d2 ± .

23 Dd5 The rook joins the attack, leaving the bishop to guard the king ; 23 gg1 h4 ! 24 � h3 g5 is unclear.

24 f4! 24 gg1 ?! ge5 ! ; now the fol lowi ng combina tion does not work, as Black's rook is already on the e -fi le: 24 gd x h5+ ? g x h5 25 f4 �a1 + 26 Ad1 gxe4.

24 . . . "a1 + 24 . . . gxe4 25 gd x h5+ � g7 26 gh7+ � g8 27 �b3+ c4 28 �xc4+ ! gxc4 29 Axc4+ gf7 30 gh8+ �xh8 31 Axf7+ � g7 32 gx h8 � xh8 33 A xg6 +- ; 24 . . . � g7 !?

8 7

25 Dd1 25 Ad1 ?! gxe4+ 26 � f1 � g7 ! =t .

4 3 -==�,",--I 2 ='J=-��I

25 . . . "f6 It looks as though Wh ite doesn't have more than a draw by repetition. However, after re peating the position twice i n order to gain time on the clock, he begins a new wave of the attack.

� 1

23 . . . Dae8!?

26 Dd5 "a1 + 27 Dd1 "f6

Black is trying to get to e5 and prevent f3-f4 at the same time. The alternative was 23 . . . � b4 (or 23 . . . � d4 24 �g4! �f7 25 gd x h5+ gxh5 26 f4 � xe2 27 �g5+ � h7 28 gx h5+) and now: A) 24 gd x h5+? is sufficient onl y for a draw after 24 . . . g x h5 25 f4 �c2+ (25 . . . �a1 + 26 Ad1 gf5 ! = , as 27 exf5 ge8+ is bad for Wh ite) 26 � f1 (after 26 � d2 �d4+ 27 Ad3 (or 27 � xc2 �xe4+ 28 Ad3 �a4+ =) 27 . . . gg8 28 gxh5+ � x h5 29 � h3+ � g6 30 e5+ White wins, but 26 . . . �xf4+ !? leaves Black with a slight advantage) 26 . . . �a1 + 27 � g2 �e1 + 28 gxe1 �xe1 = ;

4 3 2 d

e

f

� 1 9

h

{t

28 f3!!

B) 24 ge5 ! and now:

28 gd5 =.

B 1 ) 24 . . . gae8 25 f4 doesn't solve Black's problems ; B2) 24 . . . � c6 allows White to carry out hi s main idea - 25 gex h5+ ! g x h5 26 f4 �a1 + (26 . . . �g6 27 gxh5+ �xh5 28 Axh5) 27 Ad1 gf5 28 gg1 ! �f6 (28 . . . gg8 29 �xg8 �c3+ 30 � f1 ) 29 exf5 � b4 30 gh1 ;

8 7 6

8 7 6 5 4 3

28 . . . �d4 28 . . . �xf4? 29 �xf4+ gxf4 30 A b5 +- . 29 f5 After protecting the i mportant e4 pawn, White gets closer to the black king.

29 . . . g511

25


B 1 ) 31 Ac4 ?! fxe4 ! 32 A x e6 �x e6 ! (32 . . . � xf3+ 3 3 � f1 ! �x e6 3 4 d8 � �c4+ 35 � f2 �xa2+ 36 gd2 +-) 33 f4 �f3+ = ;

Short of time, my opponent blundered . Other replies would have obl iged Wh ite to work hard to gain an advantage. However, from the practical pOint of view B lack's task is very difficult, as al l the time he has to make the only moves and any inaccuracy leads to disaster:

B2) 3 1 f4 !! (by creating a th reat of mate in one, Wh ite forces the exchange of the black knight) 31 . . . � x e2 32 � x e2 � b2+ (32 . . . gxe4+ 33 � f2 ! (33 � f1 gd4 ! 00) 33 . . . � b2+ (33 . . . gd4 34 gxh5+ +- ; 33 . . . �g6 34 gd6 ! +-) 34 � g1 �f6 35 gd6 ge1 + 36 �f2 +-) 33 �f1 �b5+ (33 . . . gg6 34 gxh5+ � g7 (34 . . . � xh5 35 �h3#) 35 gg5) 34 �d3 ! �xd3+ 35 gxd3 gd8 36 e5 ± intend ing gdh3;

A) 29 . . . gd8 30 � f2 gxd6 (30 . . . �xd6 31 f4 i ntending A xh5 +- ; 30 . . . � xe2 31 � xe2) 31 �f4+ � h7 32 e5 �xe2 33 fxg6+ ! (33 exf6 � xf4 34 gxd6 00) 33 . . . �xg6 34 �xf8 gxd1 (34 . . . �g3+ 35 � x e2 �x e5+ 36 � f2 � b2+ 37 � g3 gg6+ 38 � h4) 35 gxd1 ± ;

C) 29 . . . ge5 30 � f2 ! � x e2 (30 . . . �xd6 31 A d3) 31 � x e2 �xf5 32 d7 �f6 33 d8 � gxd8 34 gxd8 gg5 35 �h3 �xd8 36 f4.

B) 2 9 . . . gxf5!? 30 d7 ! (30 �f2 �g5 ; 30 gxd4 �xd4 31 gxh5+ � xh 5 32 f4+ =) 30 . . . ge6 (30 . . . ge7 31 �f2 �g6 32 �f4+ +-) and now:

30 .6xh5+ Black resigns

*** Game 3

a b c d

B o r i s G e l fa n d - Va s i l y I va n ch u k I n terpo l i s To u r n a m e n t , Ti l b u rg 1 9 9 0 GrOnfeld Defence [085]

8 7 6 5 4 3 2

Si nce our first game in 1 985 I have played a good many ti mes with Vasi ly. His extraordi nary talent and real devotion to chess have made him one of the best players of modern ti mes. He has won n u merous g reat tourna ments and beaten all the best players many times. He has a plus score against me, but I am proud that I have managed to beat such a player many times as wel l .

13 Ag5

6 5 4 3 2

After 1 3 gx b7 A xf3 14 A xf3 A xd4 Black does not have any problems. The immediate 13 Ae3 was i ntrod uced by Alexander Cher n in and is the main alternative to 13 Ag5.

1 d4 �f6 2 c4 g6 3 �c3 d5 4 cxd5 �xd5 5 e4 �xc3 6 bxc3 Ag7 7 �f3 c5 8 Ilb1 0-0 9 Ae2 cxd4 10 cxd4 "a5+ 11 Ad2 "xa2 12 0-0 Ag4 This idea of the i nventive Mark Tseitl i n , an expert on the G rOnfel d, q u ickly became the main response to the pawn sacrifice. Black beg ins an immediate cou nterattack agai nst White's centre, without wasting a tem po on 12 . . . b6.

•

e

1 3 . . . h6 Th is was played against me for the fi rst time by Vasi ly i n the 1 990 I nterzonal Tou r nament i n Manila (p. 36), where, i nciden tal ly, we shared first place. In the semi fi nal o f the 1 998 Cap d 'Agde rapid tourna ment Vasi ly invented a new idea : 1 3 . . . Af6 14 Ae3 (14 Axf6 exf6 15 gx b7 ! ? ; 14 A h6 gd8)

26

Game 3

Gelfand - Ivanchuk, I nterpolis Tournament, Tilburg 1990

14 . . . !Dc6 (14 . . . b6 !?) 1S gx b7 gab8 16 gx b8 gxb8 17 h3 Ad7 (17 . . . Axf3 18 Axf3 t) 18 dS �eS (18 . . . !Db4 19 Af4) 19 Af4 VW b2 (19 . . . Aa4 20 VWe1 VW b2 21 A x eS A x eS 22 VWaS VWxe2 23 VWxa4 t ) 20 VWc1 ! g b7 (20 . . . !D xf3+ 21 Axf3 gc8 22 VWx b2 A x b2 23 g b1 ±) 21 VWx b2 gx b2 (21 . . . !D xf3+ 22 A xf3 A x b2 23 gb1 gb6 24 Ac7 (24 Ae3 Ad4) 24 . . . g b4 2S eS Ac3 26 gc1 Ad4 =) 22 AxeS (22 � xeS AxeS 23 A x eS gxe2 24 f3 (24 g b1 f6) 24 . . . fS =) 22 . . . gxe2 23 A xf6 exf6 24 ga1 t (Gelfand - Ivanchuk, Cap d 'Agde 1 998). 14 Ae3 Harm less is 14 A x e7 ge8 1 S g xb7 !Dc6 16 AcS gxe4 17 Ad3 gxd4 18 A xd4 !D xd4 with full com pensation for the exchange, as shown by Kramnik- Kasparov, (Novgorod 1994) . For 14 A h4, see Gelfand - Sh i rov game 7 on page 41 . It seems that Wh ite has lost a tem po, but even after I have played n u m ber of games I can not say whether . . . h7-h6 is a usefu l move or not. On the one hand , White can not use the gS square in his attack. On the other hand , in some l i nes the h6 pawn is hanging and the g6 square is weakened . 14 . . . �c6 1 5 d5 �a5 Protecting the b7 paw n , but now both the kn ight at as and the q ueen at a2 are out of play and can become the targets of attack. 1S . . . !DeS was Ivanchu k's choice i n Mani la. Later Kamsky also played this agai nst me cf. game S on page 34. 16 Ac5 Af6

a b c d

e

8

...._ .::... ----I 'if

L..._ ..______

17 e5!? White must fight for the in itiative without pay ing attention to loss of material . 17 h3 gfc8 !? 1 8 Ae3 A xf3 19 Axf3 is less energetic.

17 . . . Axe5 18 h3 If 18 !DxeS Axe2 19 ga1 Axd1 20 gxa2 !Dc4 21 !D xc4 A b3 . A few years later Vlad i m i r Kram n i k found a better way of implementing White's idea - 1 8 g b4 ! (the wh ite rook comes i nto play with gain of tempo) 18 . . . Axf3 1 9 Axf3 gae8 20 A e3 (th reatening 21 ga4 and attack ing the h6 pawn ; if 20 ga4 VW b3 21 gxaS VWx d1 22 gxd1 b6, regain in g the piece) 20 . . . !Dc4 (20 . . . b6 2 1 A x h6 Ag7 22 A x g7
27


(better is 27 . . . a6 28 g4 gc3 29 c;!? g2 c;!? g7 (0 ) 28 gxa7 b4 29 gb1 gb8 (29 . . . b3 loses to 30 g7+ ! Axg7 31 gx b3 , but 29 . . . gc4!? wou ld have been better) 30 g b3 Ac3 (30 . . . � b5 31 g7+ A x g7 32 gx b4) 31 gc7 with good winning chances (Gelfand - Macieja, M i lan Vidmar Memorial , Portoroz 2001 ).

21 gxb7 Ad6 22 gxa7 'tWd2 leads to a draw) 20 . . . exd6 21 Ad5 �c4 and now : A) 22 gx b7 ? d xc5 23 'tWf3 � d2 (23 . . . 'tWa3) 24 'tWd3 A h2+ ! 25 c;!? h1 (25 c;!? x h2 � xf1 + 26 c;!? g1 'tWa1 ! -+) 25 . . . 'tWxd5 26 'tWxd5 � xf1 27 'tWd1 and White has to fight for a draw; B) 22 'fWd3 'fWd2 23 'fWxc4 dxc5 24 gx b7 'fWf4 ;

18 . . . Axf3

C) 22 A b4 !? and despite being three pawns down , White keeps the i nitiative; D) 22 'tWg4 d x c5 23 'tWx g6+ (23 A xc4 'tWc2) 23 . . . Ag7 24 gx b7 �e5 ! 25 'tWx g7+ c;!? xg7 26 A x a2 with ful l com pensation for the pawn , but not more.

Black had the interesting possibility 1 8 . . . gfd8!? 1 9 h xg4 gxd5 20 Ad4 (20 'tWc1 gc8) 20 . . . Axd4 21 �xd4 gad8 !? (21 . . . e5 22 'tWc2 'tWx c2 23 � xc2 gd2 24 A d1 ±) 22 'tWc2 'tWxc2 23 �xc2 gd2 24 gfd1 (24 Ad1 �c4 (0) 24 . . . �c4 ! with ful l com pensation (Strelbi n Nadan ian , USSR 1 991 ) , but after 22 g b4 ! � c6 2 3 ga4 'tW b2 24 'tW b3 'tWx b3 2 5 � x b3 White retains winning chances.

20 d6! White plays to exploit his trum ps - his d pawn, the bishop pair and the loose position of Black's queen and kn ight.

19 Axf3

a b c d

20 e

...

f

19

. . .

•c4

Th is leads to a d ifficult end in g. How ever, Black is also suffering after 20 . . . exd6 2 1 Ad5 � c4 (21 . . . 'tWxd5 22 'tWxd5 d x c5 23 'fWxc5 b6 24 'fW b5) 22 gx b7 ± , or 20 . . . 'fWe6 2 1 d x e7 ! ? {21 d7 ged8 22 'tWa4 Ac7 ! (22 . . . �c4 23 Ax b7 gab8 24 Axa7 ±) 23 gfd1 A b6 or 23 . . . 'tWe5 !?) 21 . . . gac8 22 Ad5 'tW f6 23 A a3 and the pawn on e7 is extremely strong.

8 7 6 5 4 3 2

. . .

2

21 d7 21 gc1 'tW f4 22 g3 'tWf6 23 d x e7 00 also de served serious consideratio n , bu t t h e text move is stronger.

Bfe8

21

Black faces serious problems. 1 9 . . . Ad67? 20 A xd6 exd6 21 ga1 loses a piece, dubi ous is 1 9 . . . �c4 ?! 20 A xe7 �d2 21 'tWe2 ! ± (21 Axf8 gxf8 22 'tWe2 ge8 23 g3 'tWa5), while after 1 9 . . . Af6 20 d6 ! e6 (20 . . . exd6? 21 Ad5 � c4 22 Ad4 A xd4 23 'tWxd4 gac8 24 'tWe4 c;!? g7 25 gx b7 +-) 21 d7 gfd8 22 Axb7 �xb7 23 gx b7 the d7 pawn determ i nes Wh ite's advantage.

. . .

•xcS 22 dxe8. + Bxe8 23 .a4 ...

a b c d 8 I-:::-.r-=--" 7

3 2

The strongest was 1 9 . . . gae8 !? 20 d6! (with out caring about pawns, White plays for the maximum activity of his pieces ; 20 ge1 �c4

28

8 7 6 5

2

Game 3

Gelfand - Ivanchuk, I nterpolis Tournament, Tilburg 1990

For the moment the material balance is i n Black's favou r, but he is forced t o concede two pawns due to awkward placing of his pieces. Weaker is 23 �d7?! � f8 ! 24 g b5 "fffc7.

23 . . . �c6 Or 23 . . . gc8 24 gb5.

24 1lxb7 Not 24 gfc1 ? ! �d4 25 gc4 �d7 26 gxc6 ?! (26 gcb4 =) 26 . . . bxc6 27 Axc6 �f5 ! 28 Ae4 "fffc8 when Black has a slight advantage.

27 . . . .td4 28 .d5 .tb6 29 g3 h5 30 c!>g2 c!>g7 31 Ub1 .e2 32 .f3 .e6 33 .c3+ c!>h7 34 lle1 .d5+ 35 .f3 .d7 36 Ild1 .e7 37 Ild3 c!>g7 38 .c6 38 �d5 ! was more precise. Here the queen would not only control the greatest area of the board , but it would also put pressu re on f7, thus l i m iting the activity of its black opponent. And after gf3-f4-e4 the q ueen exchange would become unavoidable.

38

25 . . . llxc6 In the event of the queen exchange 25 . . . �xc6 26 �xc6 gxc6 27 gxe7 Ad4 28 gd1 A b6 29 � f1 Wh ite would have to exchange rooks to reach the same ending as in the game. He cou ld ach i eve this by transferring his rook from d1 to f3.

26

. . .

•xe7 27 .xc6

a b c d 8 7 6 5 4 3 2

...

e

8 7 6 5 4

e

f

9

39

. . .

...

h

8 7 6

8 7 6 5 4 3 2

Only a d raw resu lts from 25 gb5 �c3 ! 26 �a6 (26 A xc6 �x c6) 26 . . . gc7 27 A x c6 gxc6 28 �xa7, as with all the pawns on one side Wh ite can never win this position .

Less good was 26 gb5?! "fffd4 27 gb8+ �g7.

•e2 39 Ilf3

a b c d

24 . . . llc8 25 .txc6

26 1lxe7

. . .

4 2

•a2?!

Al lowing Wh ite to exchange queens i m me diately, but also after 39 . . . �e7 40 �d5 ! �g8 41 gf4 � g7 42 ge4 �f8 43 ge2 !? � g8 (43 . . . �a3 44 ge8 ; 43 . . . �c8 44 ge7) 44 �c6 (44 �d7) 44 . . . � g7 45 ge8 � b4 46 �d5 he would have reached the same ending.

40 .c3+ c!>g8 41 .b3 .xb3 42 Uxb3 c!>f8 43 c!>f3 c!>e7 44 Ilb2 c!>e6 45 c!>e4 Wh ite has to advance his king to e5 and threaten f4-f5, in order to provoke the ad vance of the a-pawn.

45 . . . c!>d6?! 46 Ild2+ c!>e6 47 c!>f4

2

I ntending � g5 fol lowed by f4-f5 .

47 c!>f6 48 Ilc2 .td4 49 c!>e4 .tb6 50 Ilc6+ c!>e7 51 f4 c!>d7 52 1lf6 c!>e7 53 Ilc6 c!>d7 54 c!>d5 . . .

White's objective is to exchange queens, but first he has to bri ng his rook into the game.

(see next diagram)

29


a b c d

58 . . . A xg3 59 E!xa7 h4, with a theoretical draw) 55 . . . h xg4 56 h xg4 ci!;> f8 57 E!c8+ ci!;> g7 58 E!c4 fol lowed by the march of the king to e8.

8 7 6

8 7 6 5 4 3 2

54

...

e

55 gc4 a5 55 . . . A b6 56 ci!;> e5 ci!;> e7 was more stu bborn , when White has to follow the plan descri bed in the previous note.

4 2

56 Dc6 AdS 57 Da6 �e7 58 �e4 Ac7 59 �d5 AdS 60 ga7+ � 61 DaSAe7 Or 61 . . . ci!;> e7 62 ci!;> c6.

.lc7 If 54 . . . ci!;> e7, then 55 g4 (but not 55 f5 ?! g xf5 56 ci!;> e5 Af2 ! 57 E!c7+ ci!;> e8 58 ci!;> xf5 (58 ci!;> f4 h4 59 g x h4 A x h4 60 ci!;> xf5 ci!;> f8) . . .

62 Dxa5 Black resigns Th is end ing is easily won , and Vasily decided not to check my techn ique.

*** Game 4

9 . . . 'tWa5 10 0-0 cxd4 11 cxd4 'tWxa2 1 2 Ag5 E!e8 was played in the mid -eighties by Boris Gulko and it found some followers. At a train ing session of the 'Spartak' team before the USSR Cup, in a 15-minute tournament I tried 1 3 A b5 Ad7 14 'tWd3 A x b5 1 5 'tWx b5 , but after 15 . . . � c6 ! 16 d5 � e5 17 'tlt'x b7 � xf3+ 18 gx f3 'tWa3 ! Black had good counterplay (Gelfand - Semeniuk, I rpen rapid 1 984). Re cently I had remem bered this game and it gave an impetus to some thoug hts . . .

B o r i s G e l fa n d - Va s i ly I va n ch u k Dos H e rm a n as 1 9 96 GrOnfeld Defence [085J During the 1 990s I took part i n most of the 'super-tournaments' . That year this tourna ment i n a suburb of the wonderful Spanish city of Seville assem bled ni ne of the world 's top ten players. And most of us were in good shape. I was lead i ng the event after fou r rounds, but then fi n ished with a - 1 score, after m issi ng my best possibility i n the last round of beating Garry Kasparov (the only top player, whom I had never beaten). I blun dered an extra piece on the 40 th move in time -trouble and lost. Anyway, in my opinion that was the most spectacu lar tournament of modern times. And in this case it is re grettable that the tradition of publishing tour nament books of the most important events has disappeared .

10 cxd4 t!ra5+ 1 1 Ad2 t!rxa2 1 2 0-0 b6 1 3 Ag5

a b c d

...

e

8 7 6 5 4 3 2

1 d4 �f6 2 �f3 g6 3 c4 Ag7 4 �c3 d5 5 cxd5 �xd5 6 e4 �xc3 7 bxc3 c5 S gb1 0-0 9 Ae2 cxd4

6 5 4 3 2 a b c d

30

e

9

h

Game 4

Gelfand - Ivanchuk, Dos Hermanas 1 996

�d5 = ; 21 �c2 !? would have kept the tension on the board) 21 . . . tbf6 22 tb xf6+ (22 tbd6 22 . . . tb xd5 23 YWxd5 Ae5) 22 . . . Axf6 23 �f3 Ag5 ! .

Th is move looks strange, as it leads to a pos ition that arises after 9 . . . �a5 10 0-0 cxd4 11 cxd4 �xa2 12 Ag5, only a tem po down . However, things are not so simple. 13 �c1 is the main line.

Th is move destroys the coord ination of the wh ite pieces and a d raw was soon ag reed (Gelfand - Ivanchuk, Novgorod 1 996) .

13 . . . ge8?! This move is not in the spirit of the open ing and it leads to a d ifficult position . A few months later Vasily tried the most chal lenging move 13 . . . A b7, where together with Vladimir Kram n i k I had prepared a new and interesting idea - 14 gc1 !? However, after a long think my opponent found the rig ht re sponse : 14 . . . Axe4 15 Axe7 ge8 16 Ac4 �a5 (if 16 . . . YW b2 17 Ad6 00 - with the idea of tbe5 - 1 7 . . . b5 18 A b3)

14 Ab5! Ad7 1 5 "d3 Compared with the game played in my youth against Semen iuk, the extra tempo plays a negative role - the c6 square is weak and it is hard for Black to bring all h i s pieces i nto play. Moreover, his queen is in danger.

1 5 . . . "a5 Or 1 5 . . . gc8 16 gfc1 .

1 6 Ac4!? Keeping al l the pieces on the board . 16 Axd7 tb xd7 17 ga1 � b4 18 e5 gec8 ! (18 . . . e6 1 9 gfb1 �f8 20 � b5 ; 18 . . . b5 1 9 gfb1 �c4 20 �x c4 bxc4 21 g b7 tb b6 22 A xe7 with the i n itiative) 19 gfb1 �c4 20 �x c4 gxc4 21 A xe7 a5 00 i ntending 22 . . . gac8 would have made Black's task easier.

17 Ac5 !? (this is the point of Wh ite's nov elty; he gives up his bishop to obtain the g5 square from where his knight can join the attack) 1 7 . . . bxc5 18 tb g5 ge7 !? Ivanchuk chooses the safest continuation and soon equalises. After 18 . . . tbc6 19 Axf7+
1 6 . . . �c6 1 6 . . . e6 1 7 A d2 � h5 1 8 d5 and all Black's pieces are misplaced (except the bishop at g7) . 17 Ad2 Wh ite had q u ite a pleasant choice and I am not sure if I chose the most accu rate one. Also tem pting was 1 7 ga1 !? � b4 18 Ad2 �d6 1 9 e5 �c7 20 tb g5 e6, when 21 tb e4 tb e7 is unclear, but after 21 gac1 !? or 21 gfc1 !? he retains the initiative. 17 gb5 was not so clear: 17 . . . YWa4 (1 7 . . . YWa6 18 Axf7+ c;;, x f7 19 gf5+ Axf5 20 YWxa6 Axe4 21 �e2 ±) 1 8 A b3 tb b4 1 9 A xf7+ c;;, x f7 20 YWc4+
1 7 . . . "h5 1 8 gb5 e5 19 �g5

31

My Favourite Variation '"

a b c

20 .txf7+ �h8 21 .txe8 .txe8 After 21 . . . gxe8 22 gdS Ac6 23 gxd4 exd4 24 'ffc 4+ White wins. 22 gb2! ± This is the right square for the rook, as from here it protects the bishop, which is i m por tant, as we wi ll see later. If 22 gdS h6 23 eof3 Ac6. 22 h6 I guess that Black missed my next move, but h is task was difficult anyway: A) 22 . . . 'ffg 4 23 h3 'ffd 7 24 ttlf3 ; B) 22 . . . ttl e2+ 23 � h1 ttlf4 24 A xf4 exf4 2S 'ffd S +- ; C) 2 2 . . . Af6 2 3 ttl e6 ttl x e6 24 'ffd 5 ttl c7 2S 'ffd 6; D) 22 . . . A h6 23 ttle6 ttl x e6 (23 . . . A xd2 24 �xd2 ttl x e6 25 �dS ttl c7 26 � b7 ±) 24 �d5 (24 A x h6 'ffx h6 25 �d5 ttlc7 26 'ff b7 ±) 24 . . . ttlc7 2S 'ff b7 ± ; E) 22 . . . gd8 23 ttlf3 ± . Black did have the i nteresting tactical chance 22 . . . a6 !? at his d isposal . H owever, some of these ideas were found by com puter, and it wou ld have been very hard to find them over the board : A) 23 gx b6 A b5 24 � b1 h6 25 ttlf7+ � h7 26 ge1 'ffg 4 with a strong counterattack ; B) 2 3 'ffc4 h 6 24 eof7+ (24 eoe6 Af7 2 S eo xg7 'ffx h2+ ! 26
8 7 6 5 4 3 2

6 5 4 3 2 a b c d

e

f

9

. • •

h

19 . . . �xd4!? After the game I didn't like Black's decision , as it leads to a clear advantage for Wh ite, but now I think that it is the best practical chance. If 19 . . . ttl d8 20 gd5 Ae6 21 d x e5 and White retains a clear advantage. After 1 9 . . . gf8 White has two strong options : A) There is no need for 20 eo xf7 gxf7 21 dxe5 gd8 ! 22 Axf7+ �xf7 23 'ffd S+ Ae6 24 'ffx c6 gxd2 ; B) 20 d xe5 ! ttl xe5 (20 . . . gad8 2 1 f4 ±) 21 gx eS A xeS 22 f4 Ag7 23 A b4 ! (23 ttl xf7 ! ? ; 23 �xd7 gad8 00) and I don 't see how Black can hold the position ; C) 20 f4 ! ttl xd4 21 gxe5 ! A xeS 22 fxe5 ttle6 23 ttl xf7 ttlc5 (23 . . . Ac8 24 ttl h6+ � g7 2S gxf8 ttlxf8 26 e6) 24 'ffd 6 and wins. 1 9 . . . ge7 is the move I was most concerned about. However: A) 20 'ffa3 Af8 21 dS (21 gx b6 axb6 22 'ffx a8 ttl xd4) 21 . . . ttld4 22 gx b6? gee8 23 gd6 A bS ! -+ ; B) 20 dS ttld 8 ! (20 . . . ttld4 21 d6) 21 gbb1 A h6 (21 . . . h6 !?) 22 'ffg 3 'ffg4 23 'ffx g4 Axg4 24 f4 and Wh ite has sufficient compensation for the pawn , but not more; C) 20 dxeS eo xeS 21 gxeS !! AxeS 22 f4 Af6 23 A b4 (23 ttl xf7 Ae6 24 A x e6 gxe6 2S f5 gxfS 26 exf5 ge2 00) 23 . . . gae8 24 ttl xf7 and Black is defenceless.

and Black can save the game by 27 . . . gf3 !! 28 Ae3 gh3 !! 29 gxh3 eof3+ 30 �g2 (30 � h1

32

Game 4

Gelfand - Ivanchuk, Dos Hermanas 1 996

'l'xh3) 30 . . . � h4+ 31 chg1 �f3+ with perpet ual check;

a b c d 7 6 5 4 3 2

C) 23 ge1 ! . Th is cal m move is best, retain ing an advantage after 23 . . . A b5 24 �a3 gfB 25 �e7.

a b c d

e

8 7 6 5 4 3 2

7 6 5 4 3 2 a b c d

e

a b c d

f

9

h

If

28 gxc7 was a simpler way to win, but I was concerned about 28 . . . Ac6 and I decided to eliminate all the tactics :

If

This unexpected tactical blow gives White a decisive advantage. 23 �f3 � xf3+ 24 �xf3 'l'xf3 25 g xf3 ± . . . .

e

28 11bc2

23 �e6!!

23

8 7 6 5 4 3 2

�xe6 24 .d5 �c7 25 .b7

A) 28 . . . �x b2 29 �x a8 �xd2 30 g3 (30 �x e8+ ch h7 31 h3 �e1 + 32 ch h2 �xf2 33 �x e5 +-) 30 . . . ch h7 (30 . . . �d 1 + 31 ch g2 �a4 32 ge7) 31 �b7 +- ; B) 2 8 . . . Ac6 !? 2 9 �xc6 gfB 3 0 h3 ! (this calm move escaped my attention ; 30 Ae1 �x b2 31 �xg6 gg8 wou ld allow Black counterplay) 30 . . . gxf2 31 Ac3 +-.

28. . . l1d8 29 .xc7 �h7

Black cannot save his knight at c7 .

Or 29 . . . A a4 30 �xd8+ �xd8 31 gc8 and wins.

25 . . . •e2

30 .tc3

Or 25 . . . �g4 26 gc2 .

26 11e1 Wh ite is not going to g ive u p the e4 paw n . 2 6 Etc2 ? gd8 2 7 gxc7 A b5 ; 2 6 �xc7 �xe4.

30 �xa7 gd7 31 �a8 Af7 (31 . . . gd8 32 �a1 ) is also winning, but see the previous com ment.

30 . . .•d7 31 .te1 g5 32 .xd7 l1xd7 33 l1c7

26 . . .•c4 27 I1c1 .d4 (see next diagram)

Black resigns

***

33

�

My

Favourite Variation

.

Game 5

1 7 A a3 A xf3 (17 . . . gc8 1 8 �a4) 1 8 gx f3 ! ± �d7 (better was 1 8 . . . A h6 1 9 gxb7 ! A xc1 20 �xc1 !? ±) 19 A x a6 bx a6 20 �c7 � h3 2 1 ffg3 ! (21 A x e7 gfe8 allows Black coun terplay) 2 1 . . . ffxg3+ 22 h x g3 +- and the e7 pawn fal ls (Gelfand - Kindermann , Dortmund 1 990) ; The alternative is 1 3 . . . �c6 14 �d3 �d6 15 A b4 �d8: A) 16 �a3 tb c6 17 Ac5

B o r i s G e l fa n d - G at a Ka m s ky Dos H e r m a n as 1 9 96 GrOnfeld Defence [085] Th roug hout his career Gata Kamsky was at the centre of attention in the chess world, mainly due to his serious style of play, ex treme tenacity and fine techniq ue, rather than his scandalous father. As his career de veloped , around 1994 his play became more un iversal. He began playi ng strongly i n dy namically un balanced positions as wel l . Un like his father, Gata demonstrated on a num ber of occasions his acceptance of the rules of fai r play. Because of the wi l l of his father, he quit chess at the peak of his career. How ever, his extremely prag matic style (which in some ways resembled that of a com puter) found a n u m ber of fol lowers among the under-20 generation .

1 7 . . . A xd4 ! (as d4-d5 was an u n pleasant threat, Black changes the character of the game by an exchange sacrifice) 18 tb xd4 tb xd4 19 Axe7 tb xe2+ 20 � h1 �e8 21 Axf8 �xf8 22 �e3 a5 ! 23 �xe2 a4 and the strong a-pawn gives Black sufficient cou nterplay (Gelfand - Kamsky, I nterpolis, Ti lburg 1 990) ; B) Later I i m proved with 1 6 d5 tb a6 1 7 A a3 b6 1 8 �e3 tb c5 1 9 gfd1 (1 9 A xc5 bxc5 20 �xc5 Ag4 21 gb7 �d6!=) 19 . . . Ag4 20 e5 gc8 2 1 h3 A xf3 22 A xf3. I n this position , which is typical of the entire system , I prefer Wh ite, as the extra pawn is not felt at al l , whereas Wh ite's domi nation in the centre is the main factor (Gelfand - Kamsky, I nvest ban ka, Belgrade 1 991 ). 1 3 Ag5 h6 14 Ae3 �c6 1 5 d5 After 1 5 gx b7 gab8 1 6 gx b8 gx b8 1 7 h3 Wh ite regains the paw n , but he can only count on a symbolic advantage: 1 7 . . . A xf3 1 8 A xf3 gd8 (18 . . . e5 19 d5 tb d4 20 ffa1 ! ; 18 . . . e6 19 e5 !? ;!; Sakaev) 1 9 d5 tb e5 (19 . . . e6 20 �c1 !) 20 �c1 ! (ai m i n g at the c7 and c8 squares, rather than attack ing the h6 pawn) 20 . . . � h7 21 ffc7 gd7 22 �c8 �c4 23 �e8 ! (White's idea is to play

1 d4 �f6 2 c4 g6 3 �c3 d5 So, Kamsky chooses the Grunfeld for the fifth time agai nst me. Our theoretical discus sion enriched the theory of this open ing.

4 cxd5 �xd5 5 e4 �xc3 6 bxc3 Ag7 7 �f3 c5 8 gb1 0-0 9 Ae2 cxd4 10 cxd4 Oa5+ 1 1 Ad2 Oxa2 12 0-0 Ag4 Before switching to the most testing line, Kamsky preferred 1 2 .. . '�e6. It is natural that Black shou ld try to return his queen to safety. However, it allows Wh ite time to develop his in itiative with 13 �c2 . N ow 1 3 . . . �d6 was Stefan Kinderman n 's choice: A) 14 d5 b6 (14 . . . tba6 !?) 15 A b4 �d8 16 gfd1 tb a6 1 7 A a3 tb c5 1 8 tb d4 A xd4 1 9 gxd4 �d6 20 �c3 �f4 21 A b5 with advantage to Wh ite (Gelfand - Ki ndermann, Gideon Barcza Memorial, Debrecen 1 989) ; B) 14 A b4 ! �d8 1 5 d5 Ag4 1 6 gfc1 (16 Aa3 b6 1 7 gfc1 tb d7 1 8 �c7 tb c5 !) 1 6 . . . tb a6 ? (16 . . . A xf3 1 7 g xf3 !? and Wh ite's position looks im pressive ; 17 Axf3 tba6 18 Aa3 b6 ;!;)

34

Game 5

Gelfand - Kamsky, Dos Hermanas 1996

his rook to the 8 th rank, fol lowed by mate) 23 . . . .§.b7 (23 . . . �c7 24 gc1 �d8 2S �xd8 .§.xd8 (2S . . . Ctl xf3+ 26 g xf3 '§'xd8 27 gc7 e6 28 .§.xf7 exdS 29 Ad4 +-) 26 Ae2 ! ± threaten ing Axa7, f2-f4 and .§.c7) 24 .§.c1 , and now:

A) I was more afraid of 24 . . . �d3!?

A 1) as 2S .§.c8 allows a beautifu l save 2S . . . .§.b1 + 26 � h2 (26 Ac1 Ctl xf3+ 27 g xf3 .§.xc1 + ! 28 gxc1 �xf3 29 .§.c8 hS 00 in tend ing . . . Ad4) 26 . . . Ctl xf3+ 27 g xf3 AeS+ 28 f4 Axf4+ !! 29 A xf4 (29 � g2 ? �xe4+) 29 . . . .§.h1 + ! ! 30 � x h1 � f1 + with perpetual check; A2) White can set a n ice trap with 2S � h2 ! .§.b1 ! (2S . . . Ctl c4 ? 26 �xf7 Ctl xe3 27 .§.c6 ! ! -threaten i ng 28 �xg6+ � h8 29 gc8+ 27 . . . �f1 + 28 � g1 g b8 29 gxg6 �d4 30 .§.x h6+ !! 'i!? x h6 31 �hS# ; 2S . . . �bS is sim ilar to 24 . . . � bS) 26 .§.x b1 �x b1 27 �xe7 . It looks a s though Wh ite is simply a pawn up, but 27 . . . � b8 ! forces a d raw - 28 d6 Af8 29 Axa7 �xd6 30 �xd6 A xd6 31 � g1 eLl xf3+ 32 g xf3 and Wh ite has no winning chances at al l ; 8) 24 . . . �bS, and now : 81 ) it was more challeng ing to keep the queens on the board by 2S �a8 !?, for exam ple 2S . . . Ctl xf3+ (2S . . . aS 26 Ae2 ! ; 2S . . . gd7 26 Ag4 �xg4 27 h xg4 ±) 26 g xf3 as 27 .§.cS (27 �d8 27 . . . �d7) 27 . . . � b1 + (27 . . . �d7 28 .§.xaS) 28 � g2, and if 28 . . . .§. b3 29 �xaS .§.xe3 30 fxe3 �b2+ 31 'i!? f1 �b1 + 32 �e1 +- ; 82) 2S �x bS .§.x bS 26 Axa7 (26 .§.c7 ? .§.b1 + 27 'i!? h2?? � xf3+ 28 g xf3 AeS+, or 26 Ae2

.§. b4 27 f3 (27 .§.c7 .§.xe4 28 .§.xa7 Af6) 27 . . . gS ! 28 A xa7 (28 .§.c7 � g6 29 d6 exd6 30 .§.xf7 as) 28 . . . �g6 00 with fu ll control of the dark squares) 26 A x a7 fS ! and Wh ite failed to prove that h i s advantage was sufficient for a win (Gelfand - Kamsky, Dos Hermanas 1 995). Later Judit Polgar demonstrated the stronger plan 17 . . . Ad7 18 dS Ctl eS ! 19 Af4 (after 1 9 Ctl xeS A xeS 20 A x h6 as !? Black has fu l l com pensation for t h e pawn) 1 9 . . . � b2 ! . Judit finds the best plan - to keep the kn ight on eS as long as possible. Otherwise Wh ite's strong centre will tel l . If 1 9 . . . A a4 20 �c1 �xe2 21 � xeS , or 19 . . . � xf3+ 20 Axf3 .§.b2 21 eS A bS 22 .§.e1 .§.xf2 23 d6 exd6 24 exd6 Ad7 (24 . . . A c3 2S d7 A xd7 26 �xd7 A x e1 27 �e8+ � h7 28 �xe1 .§. b2 29 � h4 ±) 2S .§.e7 ± . 20 �c1 gS! The only move. If 20 . . . � xf3+ 21 A xf3 .§.c8 22 �x b2 A x b2 23 .§. b1 ± , or 20 . . . .§. b4 21 �c7 �xe2 (21 . . . .§.xe4 22 �d8+ � h7 23 A xeS A xeS 24 Ad3) 22 A xeS (22 Ctl xeS �xe4) 22 . . . A xeS 23 Ctl x eS A bS 24 .§.c1 .§.xe4 (24 . . . �d2 2S Ctlf3 +-) 2S �d8+ � g7 26 .§.c8 +- . 21 A xeS (21 Ag3 .§.c8 22 �x b2 Ctl xf3+ 23 A xf3 A x b2 ; 21 � xeS A xeS 22 �x b2 .§.x b2 23 A x eS gxe2 24 f3 fS =) 21 . . . A xeS 22 Ctl xeS �xeS = (Gelfand - J . Polgar, Nov gorod 1 996) . 1 5 . . . �e5 Ivanchu k played this move against me (Manila I nterzonal 1 990, p. 36) i n practically the first game where 12 . . . Ag4 occurred and he drew easily. Strangely enough , it was completely forced out for a number of years by 1 S . . . � aS , which was used by Ivanchuk agai nst me a few months later i n Ti lburg 1 990(p. 26). 16 Iixb7 e6 17 d6 Iifd8 18 Iie1 j,xf3 18 . . . '�a3 19 �xeS Axe2 (1 9 . . . .§.xd6 20 �xg4 '§'xd1 21 .§.xd1 +-) 20 �xe2 AxeS 21 �f3 ± . 19 gxf3 "a3 (see next diagram)

35


The only move, as after 22 . . . gd7 23 gb3 the knight is trapped in the very middle of the board !

2

23 e5 �d4

6 5 4 3 2

a b c d

e

8 7 6 5 4 3 2

�------��- �

1 9 . . .'�Va 5 !? is an idea of M iguel i l iescas, by which he tried to repair the entire 1 5 . . . tb e5 line (20 gf1 Af8 21 d7 �a2). It was voted the best novelty of In forma tor No.68 . It is quite a u n ique case when two novelties i n a narrow line received such a n award with in one year! However, later Alexander Khal if man won an impressive game against Peter Leko in Linares 2000, so for the moment the ball is back in Black's court.

�------��-

�

24 Ilcc7 !! This strong intermediate move is the point of my novelty. 24 exd6? tb xe2+ 25 � f1 tb xc1 26 A xc1 Af6, or 24 Af1 gc6 ! 25 gd1 tbf5 26 gdd7 gf8 =.

24 ... �xe2+ 25 �1 Ildd8 26 Ilxf7 !

20 14!

26 � xe2 gdb8 = and Black exchanges one pair of rooks.

20 d7 �e7 21 A b5 �h4 ! 22 Af1 (22 . . . �e7 =) %-% (Gelfand - Ivanchuk, Manila I nterzonal 1 990). A few days later I was discussing this game in a swimming pool with Mikhai l Gure vich . He told me something l i ke 'You were on the right track, you just needed a bit more imag i nation ' . Th is h i nt was enough for me to find out what he meant. I was waiting all these years for an opportun ity to demon strate my improvement.

26 . . . £18 I n 1 990 I considered 26 . . . A h8 27 � xe2 a5 28 gh7 gdb8 29 gx b8+ gx b8 30 ge7 ga8 31 gxe6 � f7 32 gc6 a4 33 Ad4 ± to be the main line: 33 . . . a3 34 � f3 a2 35 A a1 , and if 35 . . . ga4? 36 e6+ .

27 �xe2 Ildb8 After the dangerous 27 . . . a5 28 gf6 a4 Wh ite gets there in time: 29 gxg6+ � h8 30 gxe6 a3 31 g b1 (31 A a7 !?) 31 . . . a2 32 ga1 gdb8 33 gb6, or 29 gxe6 g5 (29 . . . a3 30 gxg6+ � h8 31 gg1 (31 A a7 a2 32 ga6 Bologan) 31 . . . a2 32 ga1 ) 30 fxg5 a3 31 g x h6 a2 32 Ag5 ! (32 gg6+ � h8 33 Ag5 a1 � 34 Af6+ Ag7) 32 . . . a1 � 33 h7+ � h8 34 Af6+ Ag7 35 Axg7+ � x h7 36 gh6+ � g8 37 gh8#.

20 . . . Ilxd6 21 "Oc1 !! N This is the key move of Wh ite's plan . This idea was voted the best open i ng novelty of In forma tor No.66 ! Black managed to survive after 21 �c2 gc6 22 �d2 gd6 23 g b3 �x b3 24 �xd6 tb c4 25 �c6 tb x e3 26 �xa8+ � h7 = (Solozhen kin - Oan i l i u k, Russian Championship, Elista 1 995) .

28 Ilxb8 Ilxb8

21 . . . "Oxc1 22 Ilxc1 �c6

(see next diagram)

36

Game 5

a

b c d

Gelfand - Kamsky, Dos Hermanas 1 996

a

e

8 7 6 5 4 3 2

8 7 6 5 4 3 2 a

b c d

e

9

h

{r 33

Black has exchanged one pair of rooks, so Wh ite's only chance is to obtain a passed pawn on the kingside, which should be more dangerous than the a-pawn. 29 ,§,xa7 does not promise any winning chances.

29 . . . ge8 After 29 . . . a5 30 ,§,xg6+ � f7 31 '§'f6+ Wh ite wins.

30 gxg6+ �h7 31 gf6 a5 32 f5!? Fol lowing the idea described in the note to White's 29 th move. Another good try was 32 �f3 a4 33 '§'f7+ (33 f5 exf5 34 � f4 Ag7) 33 . . . � g6 34 ,§,a7 a3 35 � e4 '§' b8 36 ,§,a6, and now the only way to hold the game is 36 . . . i> h5! (36 . . . i> f7 37 f5 '§'b4+ 38 �f3 exf5 39 '§'f6+ � e8 40 '§'xf8+ � xf8 41 Ac5+, or 36 . . . ,§,b4+ 37 � f3 � f5 38 ,§,a8 Ae7 39 ,§,a7) 37 f5 (37 ,§,xe6 37 . . . '§'b4+ 38 � f3 ,§,a4 and Wh ite has to give up his bishop for the a pawn.) 37 . . . '§'b4+ 38 � f3 exf5 39 ,§,xa3 ,§,e4 and Black should not lose. 32 '§'f7+ � g6 33 ,§,a7 A b4 34 A b6 (34 � f3 gc8) 34 . . . ,§,b8 35 A xa5 A xa5 36 ,§,xa5 � f5 leads to an easy draw, as Wh ite cannot make any progress. (see next diagram)

e

8 7 6 5 4 3 2

29 1U6!

32. . . exf5 33 f4

b c d

. . .

l1a8?

Fol lowing the standard ru le, which recom mends placing t h e rook behind a passed pawn in rook endings. However, here there are also bishops on the board and this turns out to be a decisive mistake. The pawn end ing is lost after 33 . . . Ae7 34 ,§,e6 � g7 35 Ac5 � f7 36 '§'xe7+ ,§,xe7 37 A xe7 � xe7 38 � d3 � d7 39 � c4. 33 . . . a4 34 '§'xf5 a3 (for 34 . . . ,§,a8 - see the game) 35 '§'f7+ � g6 is also insufficient : A) 36 ,§,a7 � f5 37 � f3 ,§,c8 38 Ad2 (38 ,§,a6 Ae7) 38 . . . ,§,b8 ! (38 . . . Ac5 ? 39 ,§,a6 Ae7 40 A b4 +-) ; B) 36 e6 ! a2 37 f5+ � h5 38 Ad4 and Wh ite is winning, for example 38 . . . � g4 39 � d3 � f4 40 '§'xf8 ! '§'xf8 41 e7 '§'b8 42 f6. However, Kamsky missed an excel lent op portunity to beg i n pursuing the opponen t's rook: 33 . . . � g7 ! 34 '§'xf5 (as indicated by Viorel Bologan , after 34 ,§,a6 A b4 35 A b6 ,§,c8 36 A x a5 ,§,c2+ 37 � d3 ,§,a2 38 ,§,a7+ � g8 39 A b6 ,§,xa7 40 A x a7 � f7 41 Ae3 i> e6 White has hardly any winning chances) 34 . . . Ae7 ! (th reatening to trap the rook by . . . � g7-g6 ; if 34 . . . a4 35 '§'f6 ,§,a8 36 Ad4 a3 37 e6) 35 '§'h5 (it takes too much time to bring the rook back i nto play; after 35 e6 Af6 36 ,§,xa5 ,§,xe6 Black has excellent chances of saving the game, while if 35 i> f3 a4 36 Ad4 � g6 37 � g4 '§'d8) 35 . . . � g6

37


47 � f7 ga7+ 48 � g8 ga8+ =) 38 . . . a2 39 Aa1 � gS 40 h4+ � x h4 41 f6 � gS 42 e7 A xe7 43 gxe7.

36 gh3 a4 37 � f3 g b8 ! 38 � e4 (38 � g4 gb4) 38 . . . gb4+ 39 �dS a3 40 �e6 a2 41 fS+ � h7 42 gx h6+ � g8 43 gg6+ � h7 44 � f7 (44 gh6+ � g8 =)

36 11a7 ! +- 11xa7 37 Axa7 Aa3 Or 37 . . . a3 38 fS a2 39 Ad4 AcS 40 A a1 .

38 Ad4 Ac1 39 f5 a

b c d

8 7 6 5 Analysis diagram after 44 i?f7

34

8 7 6 5 4 3 2

3 2

44 . . . Ag S ! ! , forcing 4S ga6 A xe3 46 gxa2 gb7+ with a draw.

...

e

11xf5 a4 35 11f7+ c;!?g8?!

The bishop end ing is easily won . How ever, Black's position is hopeless anyway: 35 . . . � g6 36 e6 a3 37 fS+ � hS 38 Ad4 ! (38 e7 A x e7 39 gxe7 a2 40 Ad4 a1 � 41 A xa1 gxa1 42 gg7 ga2+ 43 � e3 ga3+ 44 � e4 ga4+ 4S � d5 gaS+ 46 � e6 ga6+

39

. . .

00

Or 39 . . . a3 40 � d3.

40 c;!?d3 a3 41 c;!?c2 Af4 42 h3 h5 43 c;!?b3 Ac1 44 Ac5 Ab2 45 e6+ ct>f6 46 Axa3 Ae5 Black resigns

*** G am e 6

Peter Leko began partici pating in top events at a very early age. However, already then he was wonderfu l ly prepared theoretically and he possessed an excellent technique, which he kept on im proving. His knowledge of the GrOnfeld Defence was already very deep. He brought back i nto practice the rare variation 1 2 . . . aS, with which he won against Joel Lau tier i n Ti lburg and against Anatoly Vaisser in Cap d 'Agde the day before this game. I decided to take up the challenge.

B o r i s G e l fa n d - Peter Leko E u ropean C h am p i o n s h i p (rap i d ) , Cap d ' Ag d e 1 9 96 Gr{jnfeld Defence [085] The southern French resort of Cap d 'Agde has been stag ing top-level rapid chess events since 1 994. I have played there five times. Every four years (1 994, 1 998 and 2002) I faced Anatoly Karpov in the final and i n 1 994 and 2002 I managed t o w i n the event. It was extremely pleasant to play in such a friendly atmosphere, when hundreds of spectators-come to support their heroes. Al though someti mes they behaved more l i ke soccer fans than chess fans.

1 d4 �f6 2 c4 g6 3 �c3 d5 4 cxd5 �xd5 5 e4 �xc3 6 bxc3 Ag7 7 �f3 c5 8 Hb1 0-0 9 Ae2 cxd4 10 cxd4 'Oa5+ 1 1 Ad2 'Oxa2 1 2 0-0 a5 1 3 Ag5 a4

38

Game 6

a

b

c

d

Gelfand - Leko, European Championship (rapid), Cap d 'Agde 1996

b5 (24 . . . a3 25 Axa3 ; 24 . . . A b5 25 f!d1 f!ae8 26 h4 ±) 25 �f7+ f!xf7 26 't!Vxf7 tDf3+ 27 gxf3 A x b2 28 't!V b7 ! f!e8 (28 . . . f!g8 29 f!x a4) 29 f!e1 a3 30 't!Vf7 f!a8 31 f!xe2 a2 32 f!e7 a1 't!V + 33 � g2 +- ;

e

8 7 6 5

6 5 4 3 2

�------�--�

tt:J

B) 18 . . . 't!Vf4

�

This pawn advance is a popular plan i n the 8 .§.b1 variation . The idea beh ind it is that this pawn is qu ite dangerous and White has to assign some of his pieces to combat it. The drawback is that it takes time and gives Wh ite some valuable temp i . A simi lar plan was tried in Gelfand -Shirov (game 7 on page 41 ).

1 9 Ac4 !! (the point of White's idea - his queen is u ntouchable ; 1 9 A xg4 f!xe1 + 20 't!Vxe1 �c6 ! is fi ne for Black). Now Black has :

14 1le1 ! I fou nd this idea with Alexander Cherni n i n his house i n Budapest during one cold win ter even ing i n December 1 992 . Actual ly, I wanted to use the prepared novelty agai nst Peter in Vienna a few months earl ier, but just before the game I realised that it req u i red some polish i n g , and so I played something else. I n the meantime I spent some hours on this position together with my friend Yuri Shulman and I was ready to play it in Tilburg , but there, u nfortunately, I had t h e black pieces against Peter.

B 1 ) 1 9 . . . Ae5 !? (this move came to m y mind just hours before the game i n Vienna; how ever, it doesn 't solve Black's problems) 20 A xf7+ (20 �f3 � c6 + ; 20 't!Vxg4!? 't!Vxg4 2 1 A xf7+ � g7 22 A x e8 't!Vxg5 (22 . . . Af6 !?) 23 f!xe5) 20 . . . �g7 (20 . . . � h8 21 dxe5 Axd1 22 e6 ! 't!Vf6 23 f!axd1 +-)

14 . . . "e6?! We considered 14 . . . Ag4 to be the main line and a few years later Shulman was able to use the fruits of our analysis. I find it q u ite impressive, so I will include it here. 15 Axe7 .§.e8 16 f!a1 't!Ve6 1 7 A a3 't!Vxe4 1 8 �g5 and now : A) 18 . . . 't!Vx e2 1 9 f!xe2 A x e2 20 't!Vc2 ! �c6 (20 . . . h6 21 � xf7 � xf7 22 f!e1 ±) 21 't!Va2 tD xd4 22 't!Vxf7+ � h8 23 A b2 f!f8 24 't!Va2 !?

Analysis diagram after 20 . . . 'i!lg7

2 1 d xe5 !! A x d1 (21 . . . 't!Vxg5 22 't!Vd5) 22 f!axd1 (22 A x e8 A b3 23 �e4 �c6 24 Axc6 bxc6 25 e6 Axe6 26 A b2+ � g8 oo) (see next analysis diagram)

39


1 5 d5 15 tWc2 tWc6 16 tWa2 ± , a typical idea in the 8 gb1 variation , was also very tempting.

1 5 . . .•d6 15 . . . �xe4?! 16 A b5 �f5 17 Axe7 a3 18 Axf8 Axf8 1 9 Ac4 � f6 20 d6 ± . b

c

d

e

a

b

c

d

e

7 6 5

Analysis diagram after 21 dxe5!! fi.xd1 22 lfa xd1

White has on ly two pieces for the queen , but as analysis shows, Black is helpless : B11 ) 22 . . . �c6 23 gd7 �xg5 24 Axe8+ - see 22 . . . �xg5 ;

3 2

B 1 2) 22 . . . gxe5 23 �e6+ (23 A b2 tWxg5 24 gxe5 tWxe5 25 A x e5+ � xf7 26 gd8 a3 27 gxb8 gx b8 28 Axb8 �e6 29 f4 b5 30 �f2 b4 31 � e3) 23 . . . gxe6 24 Axe6 �c7 25 A b2+ � h6 26 gd3 ga5 27 g4 gc5 28 Ad4 tWa5 29 gh3+ gh5 30 ge5 ;

f

9

�

h

1 6 e5!? This type of pawn sacrifice also often occurs i n this l i ne.

B 1 3) 22 . . . tWxg5 23 A x e8 �c6 24 gd7+ (24 Axc6 bxc6 25 gd7+ � g8 26 e6) 24 . . . � h8 25 Af7 +- ;

16 . . . .txe5 17 �xe5 .xe5 18 .d2 a

B2) 1 9 . . . � c6 2 0 � xf7 ! +-

b

c

d

•

e

8 7 6 5 3 2

20 . . . gxe1 + (if 20 . . . A xd1 21 � d8+ ge6 22 gxe6 � h8 23 ge8+ Af8 24 �f7+ � g7 25 A xf8+ � f6 26 ge6+ � f5 27 A h6 tWxd4 28 �d6+ � g4 29 ge4+ +-) 21 tWxe1 A xd4 22 Ac1 ! tWf6 (22 . . . A xf2+ 23 tWxf2 tWx c4 24 � h6+ �g7 25 A b2+ �xh6 26 � h4+ A h5 27 tWxc4 +-) 23 � h6+ � h8 24 � xg4 Axf2+ 25 �xf2 �xa1 26 � b2+ 1-0 (Shulman -Oral , Ostrava Cup 1 998).

a

6 5 4 3 2

For the two sacrificed pawns White has the two bishops (whereas Black lacks his im portant dark-sq uare one) and a big lead i n development. Also, the e7 pawn i s weak and it is difficult to protect it.

18 . . .•d6 19 .tc4 De8 20 .114 .d8 21 .c3 �d7 22 .1b5 e5 23 d x e6 Dxe6 24 Ac4 .16 25 .d2

40

Game 7

Gelfand - Shirov, Akiba Rubinstei n Memorial Tournament, Polanica Zdroj 1998

ttJ

36 YWc4 would have maintained White's ad vantage.

White keeps on playing for an attack. If 25 YWxf6 gxe1 + 26 gxe1 tD xf6.

25 . . . �f8

36 . . . �c5 37 g4?! .td7 38 .d5

25 . . . gxe1 + 26 gxe1 tD f8 27 ge8 (27 Ae5 YW h4) 27 . . . b5 28 Ae5 YWx e5 29 gxe5 b xc4 30 ge8 A b7 31 gxa8 A xa8 ± .

26 .txe6 �xe6 2 7 .te5 a

b

c

d

...

e

8 7 6 5

6 5 4 3 2

3 2

38 . . b6? .

Black misses his chance: 38 . . . gc6 ! with counterplay (39 gxc5 ? Ae6) .

Black has two pawns for the exchange, but his king is weak.

39 !lxc5 bxc5 40 .xd7 Be6 41 .xa4 �h7 42 .d7 !le7 43 .d2 .c8 44 .f4 .e6 45 .tc3 .xe1 + 46 .txe1 !lxe1 + 47 �g2 !le7 48 .d6 Be6 49 .xc5 �g8 50 f4 �g7 51 �3 �g8 52 .c8+ �g7 53 .c3+ �g8 54 f5 g xf5 55 g xf5 !lb6 56 �g4 Bd6 57 �h5 �h7 58 .c7 B16 59 .e5 !lc6 60 16 Bc4 61 .e3 Black resigns

27 •d8 28 .c3 28 YWe3 !? . . .

28 . . . h6 29 Bbd1 .f8 30 .td6 .g7 31 .te5 .f8 32 .f3 32 gd6 !?

32 . . . !la6 33 !lc1 �g5 34 .f4 .tf5 35 h4 �e6 36 .f3

*** Game 7

B o r i s G e l fa n d - Al exey S h i roy Aki b a R u b i n ste i n M e m o r i a l To u rn a m e n t , P o l a n i c a Zd roj 1 9 98 Griinfeld Defence [085J I am proud that this won the prize for the best game i n Informator Volume 73 , where seven out of the nine judges gave it maxi mum pOints.

So, I had a chance to try the 8 gb1 variation again . Actually, before the game I had some doubts about whether I should go in for it, as it was obvious to me that Alexey would have analysed this variation i n depth be fore his match with Vladimir Kramnik, which he had won a few months earlier. (However, Vladimir tried some different set-ups in their encounter).

8 . . . 0-0 9 .te2 cxd4 10 cxd4 .a5+ 11 .td2 .xa2 1 2 0-0 .tg4 1 3 .tg5 h6

1 d4 �16 2 �f3 g6 3 c4 .tg7 4 �c3 d5 5 cxd5 �xd5 6 e4 �xc3 7 bxc3 c5 8 !lb1

(see next diagram)

41


a

b

c

d

a

e

7 6 5 4 3 2

8 7 6 5 4 3 2

�------��-

7 6 5 4 3 2 �

b

c

d

e

8 7 6 5 4 3 2

�------�--�

14 Ah4 For some reason , this idea of Alexander Chernin had never been popular. It was only when , a few months before this game, Vishy Anand joined the '8 �b1 Club' , that this move received a new boost. The idea is to keep an eye on the e7 square, and if Black drives away the bishop by . . . g6-g5, his kingside is seriously weakened .

14 . . . a5 The alternatives are: A) 14 . . . �d8 1 5 d5 g5 1 6 Ag3 b6 ?! 1 7 �e1 Axf3 18 A xf3 �d7 19 e5 ± (Anand - I l lescas, Madrid Magistral 1 998) ; B) 14 . . . g5 ! 15 Ag3 �c6 1 6 d5 �ad8 17 �x b7 f5 ! (17 . . . e6 1 8 Ac7 ± Chern i n -J . Horvath , H u ngarian Championshi p 1 992) 1 8 �e1 (18 Ac7 fxe4 !) 18 . . . A xf3 1 9 A xf3 �d4 20 exf5 �xf3+ 21 gxf3 �xd5 22 �xa7 �xf5 = (analysis by Chernin).

1 5 llxb7 g5 1 6 Ag3 a4 17 h4 a3 A few months later Black managed to revive this line, than ks to the efforts of Ivan Sokolov - 1 7 . . . � c6 ! N (learn ing a lesson from the Gelfand - S h i rov game, Black immediately bri ngs his knight i nto play) 18 h x g 5 h xg5 1 9 d5 A xf3 20 A xf3 �d4 21 d6 e5 22 d7 �e6 23 Ag4 �c6 24 �b1 �xe4 25 �e1 �d5 26 Af5 ! �ad8 (Shi pov - I . Sokolov, Hastings 1 998) .

Both the players defi n itely knew they were following the Lautier- Sokolov game (see be low) which had taken place a few months earl ier, but they had different opinions about it. I believed that it was extremely dangerous for Black to rely only on his a-pawn , whereas Alexey thought that this was a strong trump and it wou ld hel p Black to hold the position . 19 gc7 ! N No, this was not an elaborate home prepa ration, as many people suggested later, but an over-the-board decision . This is a m u lti purpose move: Wh ite wants to win a pawn by �c2 and �xg5 or place his bishop on the i m portant c4 square. I spent more than an hour tryi ng to make this idea work. 19 � b5 also looked very tempting, but Black man aged to survive : 19 . . . � c6 20 �xg5 A xf3 21 g xf3 � b2 22 A c4 �fd8 23 f4 �xd4 24 A xf7+ c;;, x f7 25 � h5+ c;;, g 8 26 e5 � b3 27 �g6 �f7 28 �x c6 �ad8 29 �a6 �d3 30 f5 �f3 31 �a1 �xf5 32 �xf5 �xf5 33 �xa3 e6 34 �e7 �f8 35 �d6 �g6 36 �d1 �f5 37 � b8+ c;;, h7 38 �d4 �xe5 39 �h4+ �h5 40 �xh5+ %-% (Lautier- I . Sokolov, Sige man & Co. , Malmo 1 998) . 19 �a6 Black goes for the most forcing l i ne, which , however, doesn 't solve his problems. 19 . . . Axf3 20 A xf3 doesn 't give him any ad d itional opportun ities, but I was more con cerned about two other options : A) 1 9 . . . �d7, with several possibilities : . • •

18 hxg5 hxg5

42

�

Game

7

Gelfand - Shirov, Aki ba Rubi nstein Memorial Tournament, Polanica Zdroj i

A 1 ) 20 Ac4 'f¥ b2 2 1 !!xd7 Axd7 22 ttl xg5 is too aggressive - 22 . . . 'f¥ b6 ! 23 e5 (23 'f¥ h5 � h6 24 Axf7+ � h8) 23 . . . 'f¥g6;

A2) 20 !!e1 'f¥ b2 2 1 !!c2 'f¥ b6; A3) 20 !!c2 ?! is too straightforward 2 0 . . . 'f¥ b3 2 1 ttl xg5 A xe2 (21 . . . a2 !?) , and now : A31 ) 22 'f¥xe2 ? ! a2 (22 . . . A xd4 23 e5 'f¥xg3 24 'f¥h5 'f¥d3 25 !!d2 'f¥g6 � ) 23 !! b2 'f¥a3 24 !!a1 A xd4 25 !! b x a2 'f¥ xg3 26 !!xa8 �xg5 -+;

B22) 20 . . . 'f¥ b3 2 1 � xg5 a2 22 !!xa2 'f¥x d1 23 !!xd1 !!xa2 2 4 A x g4 � th reatening � e6. During the game I thought that this position was critical , and I sti ll bel ieve it! A simi lar position could also occu r i n the 19 . . . �d7 line. Wh ite is defi n itely better. He has two plans : to create a passed d - pawn or try to harass the black ki ng. I n my opi nion, Black should try to exchange rooks , which would minimise Wh ite's advantage.

A4) 20 e5 ! (sh utting both the Ag7 and ttl d7 out of the game with ju st one move) , and now :

A42) 2 0 . . . 'f¥ b2 2 1 !!c2 'f¥ b3 (21 . . . 'f¥ b6 22 ttlxg5 Af5 23 !!a2 ±) 22 � xg5 a2 (22 . . . Axe2 23 'f¥x e2 a2 24 !!a1 'f¥ b1 + 25 !!c1 , and if 25 . . . !!fb8 26 e6 'f¥xc1 + 27 !!xc1 !!b1 28 exf7+ � f8 29 � e6+ � xf7 30 ttJg5+ +-) 23 !!xa2 (23 !!c1 'f¥xd1 24 Axd1 Axd1 (24 . . . a1 'f¥ 25 !!xa1 !!xa1 26 A xg4 !!xf1 + 27 � xf1 � b6 � ) 25 !!cxd1 (25 !!fxd1 A h6 26 Af4 !!a4 27 Ae3 !!fa8 28 !!a1 ttJ b6 (0) 25 . . . � b6 (0) 23 . . . 'f¥xd1 24 !!xd1 gxa2 25 Axg4 � b6 � ; B) 19 . . . 'f¥b2 , with two possi bilities : B 1 ) 20 Ac4 (th i s deserves consideration , but I prefer 20 !!c2) 20 . . . a2 (20 . . . e6 21 Ad6 gd8 !? (21 . . . a2 22 'f¥a1 'f¥ b6 23 A xf8 'f¥xc7 24 A xg7 'f¥xc4 25 �e5) 22 e5 �d7) 2 1 'f¥a1 �xa1 22 !!xa1 Axf3 23 gxf3 ttJa6 (23 . . . Axd4 24 gxa2 gxa2 25 Axa2 e5 26 Ac4 �) 24 !!xe7 tb b4 25 !!d7 ;

20 gxe7 20 A xa6?! is pointless : 20 . . . !!xa6 (20 . . . Axf3) 2 1 'f¥d3 A xf3. 20 A c4 'f¥ b2 2 1 !!xe7 wou ld have forced matters, but in any case I don't see how Black cou ld avoid this position.

20

. . .

• b2

20 . . . Af6 2 1 !!b7.

21 J.c4 .b4 Black fol lows the main line, as otherwise he gets mated after 21 . . . Af6 22 !!xf7 !!xf7 23 e5 ! (23 A xf7+ � xf7 24 � xg5+ A xg5 25 'f¥xg4 +-) 23 . . . Ae7 (23 . . . 'f¥ b7 24 exf6 A xf3 25 'f¥d2 Ae4 26 !!e1 ) 24 A xf7+ � xf7 25 ttJ xg5+, or 21 . . . a2 22 !!xf7 !!xf7 23 Axf7+ � h8 (23 . . . � xf7 24 � xg5+ +-) 24 'f¥a1 or 24 Ad5.

22 J.xf7+ After 22 !!xf7 ?! 'f¥x c4 23 !!xg7+ � x g7 24 'f¥d2 A xf3 25 'f¥xg5+ � h7 26 gx f3 'f¥ f7 Wh ite could hardly hope even for a draw.

22

B2) 20 !!c2 , and now : B21 ) 20 . . . 'f¥ b6 21 � xg5 'f¥xd4 (21 . . . A xe2 22 'f¥xe2 �d7 23 e5 !!a4 24 e6 ± ; 21 . . . a2

tt)

!It!} fti'J!"'_'S"'�

22 !!xa2 !!xa2 23 A xg4 A xd4 (23 . . . 'f¥xd4 - cf. 2 1 . . . 'f¥xd4) 24 Ae6 ! fxe6 25 'f¥ h5 !!fxf2 26 'f¥g6+) 2 1 . . . 'f¥xd4 22 'f¥xd4 A xd4 23 A xg4 a2 24 !!xa2 !!xa2 25 � e6 � c6 26 ttJ xfB � xfB � . Perhaps this was one of the best options, as it is very difficult for Wh ite to win this ending (although Black cou ld suffer for about 1 00 moves !) ;

A32) 22 !!xe2 'f¥xd1 23 !!xd1 !!a4 (23 . . . a2 24 !!a1 Axd4 (24 . . . !!a5 25 e5 !!fa8 26 !!b2 ±) 25 !!axa2) 24 ttlf3 ttl b6 with fu l l compensa tion ;

A41 ) 20 . . . !!fc8 2 1 Ac4 'f¥xc4 (2 1 . . . 'f¥ b2 22 Axf7+ � h8 23 !!xc8+ !!xc8 24 'f¥d3 ±) 22 !!xc4 !!xc4 23 'f¥ b3 !!aa4 24 ttl xg5 ± ;

1 998

>l'��_.

. . .

�h8

N ot 22 . . . !!xf7 23 !!xf7 A xd4 24 Ae5 A x e5 25 'f¥d5 +- . Now it looks as though White's

43


rook is trapped and that his attack was in correct, but I had prepared a surprise for my opponent.

A21 ) 28 gxg5+ � h7 29 gxd1 a2 30 � h2 (30 gd3 a1 r!i + 31 � h2 r!ix e5+ 32 d x e5 �xe5+) 30 . . . �e4 31 � g3 � b1 32 gh5+ �g8 33 gdh1 � f7 00 ; A22) 2 8 gd7+ r!ix e5 2 9 d x e5 . I was really amazed when my opponent showed me this line in the post-mortem . During the game Alexey was not sure about Black's chances here, but, as analysis shows, a d raw is the most likely result. 29 . . . Ac2 (29 . . . Ag4 30 gd4 A xf5 31 ga4 /f;) c7 32 gxa8+ /f;) x a8 33 ga1 /f;) c7 34 gxa3 /f;) e6 t) 30 g4 /f;) c5 31 gc7 a2 32 � g2 a1 r!i 33 gxa1 gxa1 34 gxc5 Ae4+ 35 f3 ga2+ 36 � g1 A xf3 ;

8 \---1---'

6 5 4

3 2 c

d

e

f

9

h

A1 ) 27 gb7+ gxe5 28 gx b5 Ae2 (28 . . . gxd5 29 gxd5 00) 29 dxe5 Ax b5 = ; A2) 27 exf5 r!ixd5, and now :

�

23 Eld7!! This is the main move of the game, and I am proud to have foreseen it qu ite a long way in advance. For his part, Shirov cal led 23 gd7 a 'p rosaic' move (I have to agree with h i m !) and he was afraid of an even more i mag i native idea : 23 Ae6 ! A xf3 (after 23 . . . r!ix e7 24 A xg4 the a-pawn sti l l has a long way to go and White already has a material advan tage) 24 gxg7 Axd1 25 Ae5 !!

B) 26 d 5 ! (I believe t h i s is stronger than 26 A d5) 26 . . . r!i b2 ! (fi nally reaching the bishop) 27 gg8+ � h7 28 A x b2 gxg8 (28 . . . axb2 29 gxf8 b1 � (29 . . . gxf8 30 gxd1 ) 30 gxa8) 29 Axa3 ! Ac2 30 Af5+ (no, we are not going to exchange our nice bishop for a passive rook!) 30 . . . � h6 31 gc1 A a4 32 e5 and Wh ite is better, but it is hard to claim anything more, seei ng as he is a rook down !

23

. . .

Axd7

After fol lowing the difficult route a1 -b1 -b7c7-e7-d7, the rook has nowhere to go, but 24 Ad6 is threatened , and so Black has no choice but to take it. If 23 . . . Af6 24 Ad5 (24 Ad6 r!i b5 ; 24 Ae6 !) 24 . . . A xd7 25 /f;) xg5, 23 . . . r!i b5 24 gd5, or 23 . . . a2 24 A x a2 (24 Ad6) 24 . . . gxf3 (24 . . . Axd7 25 /f;)xg5) 25 gxf3 Axd7 26 � g2 Axd4 27 gh1 + � g7 28 Ae5+.

The point of Alexey's idea. Now mate in two is threatened , and 25 . . . r!i b5 is the only de fence :

24 �xg5 .b6 25 Ae6! This is the point. Black has to g ive u p h i s qu een , t o prevent a decisive check on the h -fi le.

A) 26 Ad5 (the bishop tries to protect its more i m portant col league) 26 . . . gf5 ! ! (26 . . . r!ixd5 27 exd5 A a4 (27 . . . A e2 28 ga1 +-) 28 ga1 gf5 29 gxg5+ gxe5 30 d x e5) . N ow Wh ite has several possibilities, but it would appear that he does not have any advantage:

25 •xe6 25 . . . Ae8 26 r!ig4 A xd4 (26 . . . gf6 27 Ae5 , and if 27 . . . gxe6 28 /f;)f7+) 27 r!i h4+ � g7 28 r!i h7+ � f6 29 e5+ � xg5 (29 . . . A x e5 . . .

44

Game 7

30 'fff 5+ � e7 31 'ffx e5 +-) 30 'ffg7+ Ag6 31 A h4+ � f4 32 'ffx g6 +-.

26 �xe6 Axe6 a

b

ttJ

Gelfand - Shirov, Akiba Rubinstein Memorial Tournament, Polanica Zdroj 1998

c

d

e

f

28 0h5+ 28 d5 A d7 29 'ffd 4 Ifi1c7 30 ga1 Ifi1 b5 31 A x g7+ gxg7 32 'ffe 3 was worth considering. 28. . . cc!1g8 29 Og6 Ad7 Bad is 29 . . . Ac4 30 'ffc 6, or 29 . . . A b3 30 A xg7 gxg7 31 'ff b 6. 30 Axg7

5 4 3 2 L...-_______...._ .:::... ----I

'lJ

The hard -to-find move 30 'ffg 3, suggested by the computer, was more practical (but perhaps not stronger) . Very often you see the opposite picture - the computer's sug gestion is stronger, but m uch less practical . 30 . . . a2 31 ga1 , or 30 . . . lfi1 b4 31 A x g7 gxg7 32 'ff b 3+ � h8 33 'ffx b4 A h3 34 ga1 .

27 Ae5!?

30 . . . gxg7 31 0d6

27 Ad6 was equally strong, but I instinctively wanted to deprive the a3 pawn of its poten tially main supporter. After 27 Ad6 Black has two options :

8 .1. 7

A) 27 . . J �fd8 28 Ae5 Ac4 (28 . . . ga7 29 'ffh 5+ � g8 30 ga1 a2 31 d 5 ; 28 . . . gd7 29 'ff h 5+ �g8 30 'ffg 6 A b3 31 A xg7 gxg7 32 'ff b6 +-) 29 � h5+ �g8 30 'ffg 6 ga7 31 ga1 a2 32 'ffc6 Af7 33 Axg7 � xg7 34 d5 Ifi1c7 ± ;

5 4 2

B) 27 . . . a2 2 8 A xf8 gxf8 (28 . . . A xf8 2 9 d5 Ag7 30 'ff h 5+ � g8 31 d x e6 +-) 29 'ff h 5+ � g8 30 'ffa 5 Ac4 31 'ffa4 gc8 32 gc1 and wins.

27 . . . gf7 27 . . . Ac4 was the alternative :

A) 28 'ff h 5+ � g8 29 'ffg 6 ga7 30 ga1 a2 31 'ffc6 Af7 32 A x g7 � xg7 33 d5 Ifi1 c7 34 gxa2 gxa2 35 'ff x c7 . I bel ieve that this type of position , which can arise from a num ber of lines, is won for White in the long run ; B) 28 'ffc 1 A x e5 2 9 'ffx c4 Ag7 3 0 ga1 {3D gc1 a2 31 ga1 gfc8 32 'ffd 3 Ifi1 b4 ! (32 . . . tD c5 33 'ff h 3+ � g8 34 gxa2 A xd4 35 e5 !) 33 'ff h 3+ � g8 34 'ff b 3+ � h7 35 'ffx b4 gcb8) 30 ga1 gfc8 31 'ffd 3 Ifi1 c5 (31 . . . Af8 32 e5) 32 'ff h 3+ � g8 33 e5 and Black is helpless.

•

8 7 6 5 4 3 2

31 . . . cc!1h7 Alexey misses an excellent practical chance, which is strange, as he is, in my opinion , per haps the best defender in the chess world. 31 . . . gf8 abandons the a3 pawn too cheaply, as we already know that it is more dangerous than any piece : 32 'ffx a3 (32 'ffx a6 A h3) . 31 . . . lfi1 c7 !! was the best try, when after 32 'ffx c7 A b5 33 'ffc 5 A xf1 34 'ffd 5+ � h7 35 'ff x a8 A xg2 36 'ff x a3 A xe4+ 37 � f1 Wh ite looks to be winning, as Black's pieces are very poorly coordi nated , but 32 . . . A h3 would have set me a tough choice, taking into account the fact that I had less than 10 minutes left on my clock:

45


A) 33 �c4+ � h7 34 �d5 ga6 ;

Sergey Rublevsky joked that it is unusual to see a position , where the n u m ber of wh ite pawns equals the n u m ber of black pieces, and Black has no pawns at al l !

B) 33 �xg7+ �xg7 34 gxh3 looks to be eas ily winning, but then 34 . . . ga4 ! ! (an excellent move ; after 34 . . . a2 35 ga1 ga4 36 f3 the three extra pawns should be enough) 35 ga1 gxd4 36 gxa3 gxe4. Theory considers such a position to be drawish , but as far I know, many strong grandmasters question this as sessment. I n Monaco 2001 this endgame occurred in my blindfold game with Alexey. I managed to lose it with the stronger side(!), by forgetting the position and blundering my rook;

33

. . •

lile6 34 d5 lilg5 35 f4 lilh3+

35 . . . eDf7 36 f5 .

36 �h1 fta2 36 . . . gag8 37 g x h3 gg3 38 gf3 .

8 1.---.,7 6 5 4 3

C) 33 �c6 ga5 34 gc1 ! . The most precise solution, but I have to ad m it that I am not sure I would have found it, bei ng very short of time. As the following lines show, Wh ite is winning : C 1 ) 34 . . . A xg2 35 �c8+ � h7 36 gc7 gag5 37 gxg7+ gxg7 (37 . . . � x g7 38 �c7+ +-) 38 � h2 ;

8 6 5 4 3 2

1-.-_______---=-__-'

{f

37 f5!

C2) 34 . . . a2 35 � h2 A xg2 (35 . . . Ad7 36 �c4+ � h 8 37 �c3 gag5 38 g3 gh5+ 39 � g2 ggh7 40 d5+ � g8 41 � f3) 36 d 5 ! � h7 37 � f6 ga8 3 8 d6!? +- ;

Avoiding the last trap : 37 gx h 3? ggg2 ! = .

37 . . . lilg5 38 f6 ftg6 39 17 40 f8eD is threatened, and so Black resigns. I was very happy to play this game i n a me morial tournament to one of my favou rite players , Akiba Rubinstei n . I don 't think that it was played i n h i s style, but I bel ieve it is worthy of his memory!

C3) 3 4 . . . gxg2+ 3 5 � h1 a2 3 6 �e8+ � h7 37 �e7+ � h6 38 gc6+ (38 � h4+ gh5 39 � f6+ gg6) 38 . . . gg6 39 � f8+ � g5 40 �d8+ � f4 41 �xa5 gxc6 42 �xa2 .

32 Oxa3 lilc7 33 Oe3

***

46

My M o st Me m o ra b l e G a m e s I present here a further 44 of my most memorable games, arranged i n chronological order. In each case I have tried to set the scene, by saying a few words about either my opponent, or the event in which the game was played . Game 8

Va l e r y L o g i n ov - B o r i s G e l fa n d U S S R C h a m p i o n s h i p F i rst Leag u e , Sverd l ovsk 1 9 87 Queen 's Pa wn Opening [A4 1] The USSR Championsh ip Fi rst Leagues were extremely toug h grandmaster events, from which only a few winners q ual ified for the finals. These 17 -round events lasted about a month and thus req u i red good physical and psychological preparation by the partici pants.

a5 1 7 �a3 ! ± ) 1 6 f4 �c7 1 7 gdd1 a5 1 8 Af3 ttl c5 1 9 �c2 g b6 20 gb1 ± 1/2-1/2 (Gelfand Ivanchuk, I nvestbanka, Belgrade 1 997) . 3 . . . .th5 3 . . . A xf3 !? was tem pti n g , but even after Wh ite had wasted a tem po on 3 h3, I was reluctant to give up my bishop. 4 e4 thf6 5 the3 e6 6 d5 e5 6 . . . exd5 7 exd5 would give Wh ite a risk-free edge due to his space advantage. 7 .te3 .te7 B .te2

1 thf3 d6 2 d4 .tg4 Th is system was completely u nexplored when this game was played . I had already successfu l ly used it a few rou nds before against H uzman , and now i n the last round I chose it agai n , as only victory wou ld give me some chances of qual ifying for the 1 988 USSR super-championsh ip.

6 5 4 3

3 h3 Ten years later I gained an advantage with the wh ite side after 3 c4 A xf3 4 exf3 c6 5 ttJ c3 ttJ d7 6 Ae2 g6 7 0-0 Ag7 8 Ae3 ttJgf6 9 d5 0-0 10 � b3 � b8 1 1 gfd1 gc8 12 E!ac1 a6 13 gd2 cxd5 (13 . . . ttJ c5 14 �d1 c x d5 15 ttJ xd5 ttJ xd5 1 6 cxd5 ± ; 1 3 . . . b 5 14 d xc6 bxc4 1 5 �x b8 ttJ x b8 1 6 ttJ a4 gxc6 1 7 Axc4 �f8 (17 . . . d5 1 8 ttJ b6 E!x b6 1 9 Axd5 tD xd5 20 gc8+ Af8 21 gxd5 gc6 22 ge8 +-) 1 8 ttJ b6 ga7 1 9 ttJ d5 g b7 20 gdc2 ttJ bd7 2 1 b4 ttJ e5 22 Af1 ±) 14 ttJ xd5 ttJ xd5 15 gxd5 gc6?! (15 . . . b5 16 cxb5 gxc1 + 17 A xc1 ttJ c5 ( 1 7 . . . e6 1 8 gd1 ttJ c5 1 9 � b4 ax b5 20 a3 ±) 18 gxc5 ! ? dxc5 19 bxa6 ± ; 15 . . . ttJ c5 1 6 't'# b4

9

h

B . . . .tg6 8 . . . 0-0 9 0-0 ttJ e8 ! ? 10 ttJ d2 Axe2 1 1 't'#xe2 Ag5 1 2 ttl c4 ttl d7 leads to an eq ual pos ition (12 . . . a5 merely weakens the queenside - 13 a3 ! ± ) , as after 13 ttl a5 A x e3 1 4 �x e3 Black is not forced to weaken his pawn structure : 1 4 . . . gb8, and if 15 �xa7 ga8 16 ttJ x b7 't'#c8. 9 thd2 e6 A more classical approach wou ld be 9 . . . 0-0 10 0-0 ttJ e8 11 ttJ c4 Ag5 (11 . . . a5 12 a3 Ag5 with cou nterplay) 1 2 A xg5 �xg5, and if

47


15 gb3 �d7 (15 . . .f4? 16 A b6), and if 16 gxb7 f4.

13 tLla5 f5 ! (but not 13 . . . tLlf6 14 Ad3 tLl bd7 15 �c1 ;1;; ) .

1 5 . . . �d7 1 6 Db3 �ef6

10 0-0 0-0 1 1 a4 a5 I preferred to weaken the b5 square (after the u navoidable pawn exchange on d5), rather than give up some space : 11 . . . cxd5 12 exd5 tLl bd7 13 a5 a6 ; t

12 Af3 1 2 tLlc4 is premature : 1 2 . . . tLl xe4 1 3 tLl xe4 A x e4 14 tLl b6 ga7 1 5 tLlc8 (15 f3 c5) 15 . . . � xc8 ! ? (15 . . . ga8 16 tLl b6 =) 16 A x a7 Axd5 oo .

1 7 g 3 1Jc8 18 Ae2 a

1 3 exd5 �e8 c

d

e

e

•

f

6 5 4 2

2

7 6 5

2

d

3

8

3

c

7 6 5

1 2 . . . tLl bd7 13 tLlc4 cxd5 14 exd5 ;1;; .

b

b

8

1 2 . . . cxd5

a

1 6 . . . f4 ? ! is weak positionally and does not achieve its aim tactical ly: 1 7 A b6 �c8 18 �e2 ! (18 Ae2 ? f3 ! 19 gxf3 �x b6 20 � xb6 � x h 3 -+ ; 1 8 Ag4!? � xc4 1 9 A xd7 f3 with counterplay) 18 . . . A x c2 (18 . . . tLlef6 19 gb5) 1 9 gb5 ! ± .

6 5 4

a

c

d

e

f

9 h

The bishop is moved away from the defence of its king, so it is time for Black to press forward . 18 �e2 f4 ! (18 . . . e4 19 Ag2 A h5 20 f3 (0 ) 19 gxf4 Axc2 20 �xc2 (20 ga3 e4 ! +) 20 . . . �xc4 + .

f-.J....,p.-'LJ�� .,. k=��

b

2

�------��-� �

18 . . . f4! 1 9 g xf4 exf4 2 0 Axf4 �c5 21 Da3 1Jxh3

I have managed to obtain a pawn structure that is very famil iar to me from the Najdorf Variation , the only difference being that here all four minor pieces are still on the board (in contrast to the Najdorf Variation , where one is exchanged on d5).

a

b

c

d

e

f 8

7 6 5

14 Da3?! Th is careless move hands Black the i n itia tive. Prophylaxis was cal led for: 14 Ae2 ! Ag5 (14 . . . f5 15 f4 !) 15 tLlc4 Axe3 16 fxe3 ;1;; .

3

3

2

2

14 f5 • . .

14 . . . Ag5 1 5 g b3 A x e3 1 6 fxe3 is s i m i lar to the previous note.

22 �xd6?!

15 �c4

White is playing with fire.

48

6 5 4

Game 8

22 �e4 ?! is also not good : 22 . . . � h4 23 � xc5 �xf4 24 �e6 � h4 25 �xf8 (25 gg3 A e4 26 Af3 � h5 -+) 25 . . . Ae4 26 f4 (26 Af3 �g4 ! -+) 26 . . . gxf8 ;

8 7 6 5

22 �b5 ! was the strongest move, with an unclear game after both 22 . . . � h4 23 Ag3 't\V h6 (23 . . . �g5 24 �d2) 24 �d2 �xd2 25 �xd2 A xc2 26 ge3 gf7 27 gc1 , and 22 . . . �d7 23 A h2 ! (23 tD b6 � f5 24 A xd6 Axd6 25 �xd6 �f4) .

3 2 --=--

22 . . . �h5 23 �cb5

L..._ .._ _ _ _ _ _

Or 23 �ce4 � h4 (23 . . . tD xf4 !? 00 is possi ble, but there is absol utely no need for any sacrifices) 24 Ag3 tD xg3 25 fxg3 � h6 -+. a

b

c

tD

Loginov - Gelfand, USSR Championship First League, Sverdlovsk 1987

d

e

f 8

7 6 5

----' 'lf

_

I real ly enjoyed making this move ! The two kn ights are hanging over the position of the wh ite king, and Wh ite has no defence. 25 . . . tD e5 26 Ae6+ h8 27 A d4 was u n clear.

26 .te6+

2

2 a

b

c

d

e

f

9

26 fxg3 �xg3+ (26 . . . gxf1 + 27 �xf1 �xg4 !?) 27 h1 gxf1 + 28 �xf1 �xg4 29 gxd3 (29 cxd3 � h4+ 30 g2 gf8 -+) 29 . . . gf8 30 �g2 � h4+ 31 � h2 �e1 + 32 Ag1 A xd3 33 cxd3 gf6 -+.

26 . . . cc!.?h8 27 fxg3 exg3+ 28 cc!.?h1 .th5! -+

h

23 . . . �d3?! Being short of time, I was tempted by a beautifu l , but not the strongest option . The simple 23 . . . � h4 24 Ag3 (24 A h2 tDf4 25 Ag4 ? Axc2 ! -+) 24 . . . � xg3 25 fxg3 �g5 would have given Black a strong attack.

24 £g4? Wh ite returns the favou r. He could have beaten off the attack by 24 A h2 ! tD hf4 (24 . . . gxf2 25 gxf2 A h4 26 gg2 +-) 25 A xf4 gxf4 26 gxd3 �d7 (26 . . . A xd3 27 �xd3 +-) 27 gg 3 (27 gd4 A xd6 28 Ag4 Af5 29 gxf4 Axf4 with counterplay) 27 . . . Axd6 28 Ag4 ± .

The point of the combination .

29 ftxd3 29 �xd3 gxf1 + 30 �xf1 Af3+ 31 �xf3 �xf3+ 32 h2 wou ld have posed more problems, but Black is still winning : 32 . . . gf8 33 gc3 ! (33 �f7+ gxf7 34 A xf7 A x a3, 33 A h3 Af6, or 33 gd3 �e2+ 34 h1 �f1 + 35 h2 Af6) 33 . . . h6 34 Ag1 �f4+ (34 . . . �f1 35 gg3 gf4) 35 g2 gf6 (35 . . . Axd6 36 �xd6 �xd6 37 Ac5) 36 gg3 (36 �f7+ gxf7 37 Axf7 �xf7 38 gd3 �g6+) 36 . . . �f1 + 37 h1 (37 h2 gf4) 37 . . . gf4 with a decisive at tack.

24 . . . eh4 25 .te3 �g3!!

29 . . . .txd1 30 £g1

49

My Most Memorable Games ...

7 6 5

6 5 4 3 2

3 2

8 1____7 6 5

8

3 2

3 2

�------��- �

30 Af4 was also hopeless: A) 30 . . . '� h4+ 31 gh3 gxf4 ? 32 gxf4 ! �xf4 33 �f7+ �xf7 (33 . . . � g8? 34 � e5+ � f8 35 �g6+ !) 34 A xf7 with counterplay; B) 30 .. J:txf4 ! 31 gxg3 gxf1 + 32 � h2 (32 gg1 gxg1+ 33 � xg1 A xd6 34 � xd6 A xc2) 32 . . . gaf8. 30 • h4+ 31 gh3 gxf1 32 Dxh4 Axh4 33 �f7+ �g8 34 �e5+ �h8 35 �f7+ gxf7 36 Axf7 Axc2 37 Ae6 gd8 38 Ab6 Ae4+ Black gets rid of Wh ite's only tru m p - h i s passed d-pawn . 39 �h2 gxd5 40 Axd5 Axd5 41 Axa5 Ac6 42 �c3 Ae1 43 Ac7 g5 44 Ae5+ �g8 45 b4

45

. . .

�

Black's position is easily won , so there is no need even to consider 45 . . . A xc3. The opposite -colour bishop ending after 46 A x c3 A x a4 47 Af6 h6 48 � g3 � f7 49 Ad8 � g6 50 � f3 Ac6+ 51 � e3 � f5 52 Ae7 � g4 53 Af8 h5 54 Ae7 h4 55 � f2 � h5 56 � e3 100ks drawish .

. . .

46 b5 Af3 47 a5 �e6 48 Ad4 h5 49 a6 bxa6 50 bxa6 h4 51 �g1 g4 52 a7 Or 52 � b5 g3 53 Ae3 Ae4.

52 . . . Axc3 53 Axc3 g3 White resigns

***

Game 9

had to win a qual ifying tournament. I played five times in such tournaments, without once managing to win. Th is was not surprising , as my opponents i ncluded Bareev, Khalifman , 011, Ivanchuk, Dreev, Shabalov and Serper, among others.

R o n e n Lev - B o r i s G e l fa n d E u ro pean J u n i o r C h a m p i o n s h i p , Arn h e m 1 9 87 King 's Indian Defence [E84] Th is was my first tournament i n a Western country (actual ly, even in Eastern Europe I had played only once, a few months before). Because of the I ron Curtain it was impossible to travel freely to tournaments, so the only chance was to qual ify for the World Under1 6 or Under-20 Champions h i ps or the Eu ropean Under-20 Championsh i p . However, it was an extremely h ard task, as you fi rst

Final ly, in 1 987 I finished second beh ind Ser per (but ahead of Dreev, Shirov, Khenkin , Smiri n , Alterman etc.), which gave me the opportun ity to play in the European J u n ior Championship. But when I arrived in Arnhem I was surprised to fi nd that the best players from other countries had ignored this tour nament. It meant that the fight for first place

50

Game 9

Lev - Gelfand, European Junior Championship, Arnhem 1987

was going to be between myself and the cur rent cham pion Vasi ly Ivanchuk. After I lost to the U krai n ian in the 5 t h rou n d , I real ised that I would probably have to win all my re maining games(!) , as a result of 1 1 out of 1 3 was unlikely t o b e sufficient t o w i n t h e tou r nament. (And the d ifference between win ning this tou rnament and finishing second was huge - besides the title itself, it also meant direct q ual ification to the World J u nior Cham pionsh ip . It also brought me in vitations to Oakham and the O HRA tourna ment in Amsterdam.) I remem bered the ad vice of my friend Alexander Khal ifman , who had won such an event i n 1 985. He recom mended looking for a tough fight i n each and every game, as opponents who had not gone through the tough and someti mes even cruel qual ification system inside the Soviet Union would be unlikely to be able to match such an attitude. As a resu lt, I managed to win eight games in row (and eleven in the tournament) and to win the title.

1 d4 �f6 2 c4 g6 3 �c3 Ag7 4 e4 d6 5 f3 0-0 6 Ae3 �c6 I used to play both 6 . . . c5 and 6 . . . �c6. I be lieve that both are good antidotes against the Samisch Variation .

7 fld2 a6 8 �ge2 Hb8 a

b

c

d

e

8 6 5 4 3 2 �------��-

9 h4 and the classical 9 �c1 are more ambi tious.

9

• • •

Ad7 10 b4 b5 11 c5

Theory recom mends 1 1 c x b5 a x b5 1 2 d5 � e5 13 � d4, going after the b5 pawn .

11

• • .

e6

1 1 . . . dx c5 1 2 bxc5 e6 1 3 e5 ! � .

1 2 cxd6?! Wh ite prematurely releases the tension in the centre. Better was 12 �c1 � h5!? (1 2 . . . dxc5 13 bxc5 � h5 !? ; 1 2 . . . e5 13 d5 �d4 14 c6 Ac8 15 Ad3 (15 �3e2 !? ge8 !) and due to the poor placing of the A c8 and g b8 Wh ite stands better) 1 3 � b3 (13 g4 d x c5 14 bxc5 e5 ! ; 1 3 cxd6? cxd6 14 g4 e5 ! 1 5 dxe5 (15 g x h5? exd4) 1 5 . . . � xe5) 1 3 . . . f5 ! ? , when White has not yet developed his kingside and the pawn structure in the centre is not determ i ned , so therefore Black has counterplay.

1 2 . . . cxd6 1 3 �c1 13 g3 �e7 14 � c1 �c8 15 � b3 � b6 would lead to an eq ual game. 13 . . . e5 Here too 1 3 . . . � h 5 ! ? came into considera tion .

14 �b3 White underestimates the dynam ic re sources. Safer was 14 d5 � d4 1 5 Ad3 (15 A xd4? exd4 1 6 t'fxd4 � x e4 ; 1 5 �3e2 ? ge8 !) 1 5 . . . � h5, but as Black has the c -fi le, this changes the assessment of the position in his favour. ...

8

8 7

6 5 4 3 2

5 4 3 2

�

9 a3 This rare plan is not without venom . Wh ite gains space on the q ueenside. The s harp

a

51

b

c

d

e

9 h


14 . . . d5!

a

8 7

Black beg ins fighting for the initiative, paying secondary consideration to material. He has to develop it quickly, before Wh ite can man age to castle. Less energetic was 14 . . . exd4 15 tD xd4 tD h5 (15 . . . d5! ?) with chances for both sides.

5 4 3 2

15 exd5 Dangerous was 15 dxe5 tD xe5, and now : A) 1 6 tD xd5 tD xd5 1 7 exd5 (1 7 �xd5 � h4+ 18 Af2 �f4) 17 . . . ge8 (the X-ray on the e file i s very unpleasant for the wh ite monarch) 18 �f2 (18 �d4 � xf3+ 19 gxf3 � h4+ 20 �d1 gxe3 ! 21 �xe3 A xd4) 18 . . . tD g4+ 19 fxg4 Axa1 20 � xa1 �f6+ 21 � g1 gxe3 22 �xe3 ge8 ! 23 �f2 �xa1 =+= ; B) 1 6 exd5 ge8 1 7 0-0-0 ( 1 7 Ae2 tD c4 18 Axc4 A h6 +) 17 . . . gc8 18 � c5 as and the weakness of the wh ite king is the main factor determ ining the position . Or 15 � xd5?! � xd5 16 exd5 � xd4 17 � xd4 exd4, and if 18 Axd4 �h4+ 19 Af2 �e7+ .

1 5 . . . �xd5! Another surprise for my opponent ! However, a strong alternative was 1 5 . . . exd4 1 6 tD xd4 �e7 ! (in the event of 16 . . . � xd5 17 � xd5 ge8 1 8 0-0-0 gc8 1 9 tD xc6 gxc6+ 20 Ac5 the compensation for the piece is clearly insuffi cient), for exam ple 17 Ae2 tD exd5 18 � xd5 � xd5 1 9 Af2 �f6 20 0-0 �f4 and Black has the i n itiative.

1 6 �xd5 Ae6 Black regains the piece than ks to the pin.

17 dxe5 Axd5 1 7 . . . �xd5 1 8 �xd5 A xd5 1 9 tD c5 tD xe5 20 0-0-0 leads only to equal ity.

18 Ag5 Wh ite also tries to fight for the i nitiative. It is he who has to play carefu lly to equalise after 1 8 tD c5 tD xe5 1 9 0-0-0 Ac4.

b

c

d

...

e

8 6 5 4 3 2

18 . . . Axb3!! Played in good style ! I felt that, despite being h i m a qu een down , Black's threats should not be underestimated . However, as is usual in such instances, the defending side should try to bai l out by giving back some material (the exchange in our case) and bring his king to safety. Weaker was 1 8 . . . �d7 1 9 tD c5 (1 9 gd1 �f5 20 �xd5 �xg5 2 1 �x c6 �e3+ 22 Ae2 �x b3 23 �e4 =) 1 9 . . . �f5 20 f4 ! A x eS 21 Ad3 �xg5 22 fxg5 Axa1 23 0-0 ± . Writing i n Shakhmaty v SSSR, Ratmir Khol mov was sceptical about 18 . . . f6 19 exf6 A xf6 20 A xf6 gxf6 21 tD c5 gd6 (21 . . . �e7+ 22 Ae2 gd8 23 0-0, and if 23 . . . A xf3 ? 2 4 � b2) 22 O-O-O ! ? , b u t after 22 . . . �f6 ! Black has excel lent com pensation for the pawn. 19 Axd8 ftbxd8 20 ee3 The attem pt to cling on to the material by 20 �c1 !? was worth serious consideration . The cruel computer refutes 20 . . . gfe8 ? : A 1 ) 2 1 f4 tD x e5 2 2 fxe5 and White has to fight for a draw : A11 ) 22 . . . A x e5 23 A x b5 ! ax b5 (23 . . . Ag3+ 24 � f1 ) 24 0-0 Ad4+ (24 . . . A x a1 25 �x a1 ge2 26 �c3 gdd2 27 �c8+ =) 25 � h1 Axa1 26 �xa1 ge2 =+= ; A1 2) 2 2 . . . gxe5+ 2 3 Ae2 Ac4 24 0-0 gxe2 , and now : A1 21 ) 2 5 ge1 ? Ad4+ 2 6 � h1 gxe1 + 2 7 �xe1 A xa1 28 �xa1 A b3 ;

52

Game 9

Lev - Gelfand , European Junior Championship, Arnhem 1987

A1 22) dubious is 25 �g5?! Ad4+ 26 h1 f6 27 �g4 A xa1 28 gxa1 gdd2 29 �c8+ g7 30 �c7+ Af7 31 gg1 (31 � b7 ? 31 . . . gxg2 32 �xg2 Ad5) 31 . . . g5 =t , threaten ing . . . g6 and . . . Ad5 ; A1 23) 25 gd1 ! A d4+ (25 . . . gxd1 + 26 � xd1 Axa1 27 � x a1 Ad5 28 � f6 ! =) 26 h1 Ad5 27 �g5!? (27 gxd4 A x g2+ 28 g1 gxd4 29 �c8+ g7 30 �c3 ged2 31 ga2 31 . . . g2d3) 27 . . . f6 28 �g4 ge4 29 �g3 Axa1 30 gxa1 ge2 31 gg1 + ;

A2) 2 1 � xc6 ! A xe5 2 2 A x b5 ! ge6 2 3 0-0 Ad4+ (23 . . . a x b5 24 � x b5) 24 h1 gxc6 25 A xc6 Axa1 26 gxa1 ± . Better is 20 . . . A x e5 21 g b1 tb d4 22 f4 (22 A d3 tb xf3+ 23 g xf3 gxd3) 22 . . . Ag7 23 A d3 tbf3+ 24 g xf3 gxd3 25 0-0 and , as White has completed his development, he is no longer in danger of losing (nor is Black, of course).

20 . . . Axe5 21 .xb3 Bad is 21 � x c6 A x a1 22 � x a6 (22 Ae2 gfe8 ! ; 22 �c1 Ad4) 22 . . . gfe8+ 23 f2 gd2+ ! (23 . . . Ad4+ 24 g3 gd5 25 A x b5 gg5+ 26 � h4 gh5+ 27 g3 =) 24 g3 ge1 25 h3 (25 h4 A e5+) 25 . . . Ae6+ 26 g4 h5 when there is nothing protecting the ki ng.

23 g3 tb d4! 24 � b1 A b2 , but 22 Ad3 gfe8+ 23 A e4 Ad4 transposes i nto the game.

22 Ad3 The bishop joins the defence. The alterna tives were weaker: 22 gc1 Ag3#, 22 Ax b5? A x a1 + 23 f2 Ad4+, or 22 f2 ?! gd2+ 23 e3 Ac3+ 24 f4 A x a1 25 g3 (25 A x b5 a x b5 26 gxa1 gxg2 27 �c3 g5+ 28 f5 ge6) 25 . . . A b2 ! .

22 . . . Axa1 + When I had already completed these an notations, I sudden ly realised that Black had another way of continuing the attack : 22 . . . Ac3+ !? 23 f1 (23 f2 g x d 3 2 4 � b1 gd2+ 25 g3 gee2 26 h3 A xa1 27 � xa1 gxg2 -+) 23 . . . A x a1 24 Ae4 gd2 ! ? , invad ing the second rank, even at the cost of another piece. After 25 A x c6 (25 g4 A d4) 25 . . . gee2 26 Ad5 Ad4 27 A xf7+ f8 Wh ite can save h i s skin only with com puter assistance : 28 �d5 !? (28 Ag8 ge7 29 Ae6 gc7 30 e1 gcc2) 28 . . . Af2 (28 . . . ge5 29 �d8+ xf7 30 �c7+ f8 31 �d6+ g7 32 �d7+ h6 33 � h3+ = , and if 33 . . . gh5 34 �g4) 29 �d8+ ! xf7 (29 . . . gxd8 30 x e2) 30 �c7+ f6 31 �c6+ = , and if 31 . . . ge6 32 �c3+ .

23 Ae4

21 ... IUe8!? a

8 7

b

c

d

e

Thanks to the pin on the f7 pawn, Wh ite has managed to beat off the first wave of the attack (23 d2 ge2+).

f 8

23 . . . Ad4!

6 5 4 3 2 L...-_______-=--_----'

But the second wave is coming ! Black pre vents Wh ite from castling and underl ines the weakness of his ki ng.

24 �1

'0'

A nice finesse. It is pleasant to threaten mate in one move ! - 22 . . . Ag3. After 21 . . . A x a1 dubious is 22 f2 gd2+ (22 . . . gfe8 23 g3)

Wh ite tries to take his king to safety in the vicin ity of h3. The alternatives were dubious: A) 24 d2 tb e5 25 c1 tb c4 26 gd1 Ae3+ (26 . . . A b2+ 27 c2) 27 c2 (27 b1 gd2) 27 . . . gc8 28 b1 tbd2+ 29 gxd2 Axd2 =t ; B) 24 e2 A b6 2 5 �a2 tbd4+ 26 f2 tbf5+ 27 e2 tbe3 28 g3 tbc4 =t ;

53


C) 24 g4 �e5 25 g5!? gc8 26 � f1 ged8 27 � g2 gc3 + .

30 fxe4 gf6, with the idea of 31 gc1 Ae5+ 32 � h3 gf3#) 29 . . . gh6+ 30 � g3 Ae5+ 31 � f2 �c4 with a continuing attack.

24 . . . �e5 25 g3

28 g4 h5!

After 25 g4 � xg4 26 fxg4 gxe4 according to Kholmov 'Wh ite simpl ifies the position, but Black restores material eq ual ity with an unceasing attack'.

Th is was my idea. Black further exposes the opponent's king .

29 Axg6

25 ... �c4 26 �g2 �e3+ 27 �h3

There were two alternatives : A) 29 g x h5 gxh5+ 30 � g3 gg5+ 31 � h4 (31 � h3 � g7 ; 31 � f4 gg2 32 Ac6 �c4) 31 . . . gg2 32 f4 gb2 -+ (Kholmov) ;

8

B) 29 g5 gxg5, and now :

5 4 3 2

B 1 ) 30 gc1 �g2 31 gc7 �f4+ (the Sim ple 31 . . . gf8 is also good enough) 32 � h4 ..

After the game I wrongly assu med that Wh ite could have offered more resistance by 27 �f2 ?! gc8 28 h4 (28 �e2 gc3 29 �a2 (29 �xc3 A x c3 30 � x e3 f5 31 � d3 fxe4+ 32 � x c3 exf3) 29 . . . �c4) 28 . . . gc3 29 � b1 , but the sim ple 29 . . .f5 puts an end to the story. 27 . . . A b6! ? is also sufficient : 28 � e2 (28 gc1 gd2+ 29 � e1 ged8) 28 . . . � c4 29 gd1 f5 30 gxd8 Axd8 +.

32 . . . gdd5 !! 33 A xd5 (33 �xd5 � xd5 34 gc8+ � g7 35 � xg5 Af6#) 33 . . . Af6 34 A xf7+ � h8 35 gc8+ � h7 followed by . . . gg4# ;

27 . . . ge5 Again I was tempted by a threat of mate in one. Unfortunately, I didn't even consider the natural alternative (or fortunately from the practical point of view, as I didn't waste pre cious time choosing between two approxi mately eq ual options) 27 . . . g5!? 28 g4, and now: A) 28 . . . ge6 29 � g3 Ae5+ 30 � f2 gd2+ (30 . . . �c4 31 gd1 gxd1 32 �xd1 gd6 33 �e1 gd2+ 34 � g1 ) 31 � e1 (31 � x e3 ? Af4#) 3 1 . . . �c4 32 � xc4 bxc4 33 � xd2 = ; B) 2 8 . . . gd6 ! 29 ge1 (29 � g3 and you need to be made from silicon to notice 29 . . . gxe4!!

B2) according to the computer Wh ite can defend by 30 f4 ! gg4 31 ge1 gxf4 (31 . . . �c4 32 �f3 ge8 33 ge2 +) 32 gxe3 ge8 ! 33 gf3 gexe4 34 gxf4 gxf4 35 �d5 + .

29 . . . hxg4+ 30 �g3 30 fxg4 �d 5 ! 31 Ad3 �f4+ 32 � g3 � xd3 (Kholmov) . After the computer suggestion 30 � h4 �c4 31 f4 ge3 32 � b1 gd6 Black maintai ns his attack.

3O . . . �c4

54

Gelfand - Lobron, OHRA Open Tournament, Amsterdam 1 988

Game 10

a

b

c

d

e

33 A x b3 g xf3 34 � xf3 A b2 35 � e2 A x a3 36 �a1 + the draw is with in reach) 31 . . . g xf3 32 A x c4 bxc4 33 �x c4 f2 and n ow the computer's suggestion 34 �d1 ! (Khol mov gives only 34 �U1 Etd6, and rightly notes that White wou ld be forced to give up his rook for the f2 pawn) saves the game after 34 . . . Ete1 35 �xd4 Etxd4 36 �c8+ � g7 37 � xf2 = .

h

8 7 6 5

8 7

3 2

31 ... gxf3 32 Oxf3

L--_a______e__..;;9__

---'

Or 32 A xf3 32 . . . Ete3 33 �c2 � d2 34 'tWxd2 A e5+ 35 � f2 Etxf3+ , as indicated by Khol mov.

11

32 f5 33 Oh5?

31 Ae4

• • .

I n his annotations Ratmir Kholmov foun d a fine way to save the game. Even though he did not work everything out to the end , the brill iant defender's i ntu ition did not betray h i m : 31 Ad3 ! (after 31 . . . �e3 32 A xc4 �x b3

A blunder in time trouble, but in any case Wh ite was lost. 33 �f4 Ae3 34 'tW h4 Etd4 35 'tW f6 �d6 -+ , or 33 �d1 �xe4 34 'tWxf5 �e7 -+.

33 . . . f4+ White resigns

*** Game 1 0

B o r i s G e l fa n d - E r i c L o b ro n O H RA Open Tournament, Amsterdam 1 988 Benoni Defence [A43] This was my first ad ult tournament i n the West. It was held i n the Krasnopolsky Ho tel , i n the very centre of Amsterdam. I fel l in love with this city. I was ai m i ng to win the grandmaster group, i n order to q ual ify for the mai n tournament and have a chance of testing myself against the best western play ers. I managed to tie for 1 sL 3 rd places with my opponent and Boris Gulko, who won the event due to the superior tie-break. I like this game because of its rich tac tical content. Black violated the basic strateg ic rules and neglected his devel opment, in order to com p l icate matters from the very fi rst move. Wh ite had a big choice on every move and had to play energetically i n order to keep the ini tiative. I was su rprised to see annota tions to this game in Informator and New

in Chess magazine, clai ming that White's at tack was not totally sound and that on move 22 Black could have turned the tables. How ever, my assessment of the position was correct - the complications were in Wh ite's favour.

1 d4 e6 2 c4 c5 3 d5 exd5 4 cxd5 d6 5 e4 g6 6 �c3 Ag7 7 �f3 a6 7 . . . �f6 would transpose into normal Modern Benon i lines, but Eric Lobron is always look ing for u ntried options. He also wanted to avoid 8 h3 0-0 9 Ad3, which was becoming popular (and is still popular today).

8 Af4!? 8 a4 Ag4 9 Ae2 �f6 (9 . . . �d7 !?) would get back to the Modern Benon i , but where Black would have avoided the variation with h3 and Ad3. 8 h3!? b5 9 Af4 was also possible.

8 . Ag4 . .

8 . . . �f6 9 'tWa4+ !? Ad7 1 0 � b3 wou ld lead to yet another line of the Benon i . The most

55


1 2 . . . A x e5 1 3 A x e5 � x e5 14 �e2 and Wh ite has a clear lead in development, or 12 . . . �xb2 13 gc1 �xe5 (13 . . . Axe5 14 Axe5 �xe5 15 0-0 ±) 14 0-0 and the black king is in danger of being mated .

logical (seeing as Black has already played . . . a7-a6) was 8 . . . b5 9 h3 �f6 10 Ad3 0-0 11 0-0 ge8 1 2 ge1 ga7 13 a3 gae7 14 �d2 � b6 1 5 b4 � bd7 16 gac1 A b7 with an unclear position (Gelfand -Topalov, M elody Am ber, Monaco blind 2001 ).

1 3 1!!ra4+ 13 A x e5 A x e5 14 0-0 � f8 (14 . . . �x b2 ? 1 5 ge1 +-) 1 5 gc1 ;t; seemed too prosaic to me, even though it was good enough for an advantage. 13 . . . �d7

9 Ae2 9 �a4+ �d7 !? 1 0 Axd6 (10 �d2 b5 1 1 �c2 �e5 1 2 h3 Ad7 13 A h2 g5 !?) 10 . . . � b6 11 � b5 gc8 12 e5 with a messy game.

a b c d

e

•

f

5 4 3

a b c d

6 5 4 3 2

5 4 3 2 a b c d

e

f

9

e

h �------�--�

9 . . . Axf3?!

�

14 �e4 Again the most am bitious move ! Wh ite is tread ing a narrow pat h , where one m i stake can lead to a loss. Other continuations would allow Black a chance to complete his devel opment, with only slightly the worse game. However, as the reader will understand , this becomes clear only after prolonged home analysis, and an over-the-board player nor mally takes h i s decisions on the basis of intuition . The alternatives were : A) 14 Ag3, and now : A1 ) 14 . . . �x b2 is bad in view of 1 5 0-0 �xc3 (15 . . . Axc3 16 gab1 �d2 17 gx b7 �gf6 1 8 gd1 � h6 1 9 A d6) 1 6 gae1 + ! (preventing Black from castli ng) 16 . . . � d8 17 gb1 ! ga7 1 8 gb3 (18 gfc1 b5) 1 8 . . . b5 1 9 gx b5 ; A2) 14 . . . �gf6 1 5 0-0 0-0 1 6 �c2 ;t; ; B) 14 Af4 !? � b4 (14 . . . �x b2 1 5 0-0 �x c3 1 6 gae1 + � d8 1 7 gb1 1 7 . . . ga7 1 8 gfc1 �f6 1 9 A b8 �x b8 20 �a5+) 1 5 �x b4 cxb4 ;t; .

The most ambitious course. Black eliminates the e4-e5 breakthroug h , but he fal l s be hind i n development. The alternatives were 9 . . . �f6 !? 1 0 �a4+ ! ? , 9 . . . b5 1 0 0-0 and if 10 . . . �f6 1 1 e5 , or 9 . . . �d7 10 �d2 !?

10 Axf3 �d7?! Black provokes compl ications, but it is hard to count on success with the �g8 and gh8 on their i n itial squares and h i s king stuck in the centre. 1 0 . . . �f6 1 1 0-0 0-0 12 e5 d x e5 1 3 A x e5 would have left Wh ite with a slight edge. If 1 0 . . . �e7 , then the i ntermed i ate check 11 �a4+, also typical of this open ing, is very unpleasant.

1 1 Axd6!? 11 0-0 �e7 (11 . . . �e5 12 a4 or 12 Ae2 �e7 ;t;) leads to a typical Benoni-type position, but I was ai ming for more.

1 1 . . . 1!!r b6 1 2 e5 �xe5

56

Game 10

14 . . . f5 1 5 .txe5 After 1 5 0-0 fxe4 1 6 �x e4+ c;!? d8 1 7 gfe1 �xd6 18 �e8+ c;!?c7 19 �xa8 /!i)e7 the queen is trapped .

15

. . .

•xb2 a

b

c

d

ttJ

Gelfand - Lobron, OHRA Open Tournament, Amsterdam 1 988

e

25 A xd7 Ad4 and Black has good d raw ing chances) 19 . . . gc8 (1 9 . . . �xd5 20 gxe4+ /!i) x e4 21 A x e4 �d2 22 g3 and despite h i s extra rook, Black cann ot defend his king) 20 A a3 gc7 2 1 A x e4 c;!? d8 (21 . . . gx b7 22 A x g6+) 22 �a5 /!i) x e4 23 A b2 �d6 24 A xg7.

18 .xe4+ .e5 19 .b41 •

8 .i. 7

'''-_ -_ _ _ _ _ _ --=--_---'

6 5 4 3 2

If

16 0-0! Bringing the rook i nto the game. 16 /!i) d6+ was i ncorrect, as after 16 . . . c;!? f8 17 �xd7 (17 /!i)c4+ /!i) x c5 -+) Black can choose be tween forcing a d raw (17 . . . �c3+) and play ing an unclear position - 17 . . . �xa1 + 18 Ad1 (18 �e2 � b2+ 19 c;!? f1 � b1 + 20 c;!? e2 t'fc2+ -+) 18 . . . �e5+ (18 . . . /!i) e7 1 9 0-0) 19 � f1 /!i)e7 00 .

16 . . . fxe4 After 16 . . . b5 1 7 /!i)d6+ c;!? f8 1 8 �a5 /!i) x c5 19 t'fc7 /!i) e7 20 �xc5 Wh ite maintains his attack for free.

c

d

e

f

9

h

19 . . . He81? An excellent resource. Black gives u p his queen , but tries to block the c - and e -files. The alternative was 19 . . . �c3 (1 9 . . . c;!? f7 20 gfe1 �c3 transposes) 20 gfe1 + c;!? f7 (20 . . . c;!? d8 21 Ae7+ �c8 22 gbc1 +-) 21 �f4+ �f6 (other moves also do not hel p : 21 . . . Af6 22 gbc1 , or 21 . . . /!i) gf6 22 ge7+ c;!? g8 23 gc1 �d3 24 A e4 �e2 25 Af5) 22 �c7 �d8 (22 . . . gd8 23 d6 +-)

17 1lab1 1 17 t'fxe4+ � d8 =t .

17 . . . b5 During the game both of us thought that this was forced , and Eric even attached an 'only move' sym bol to it i n his comments i n In formator. But when annotating the game for this book, I noticed that B lack had the al ternative defence 17 . . . �e5 1 8 g xb7 /!i)gf6, when Wh ite has to play precisely : 1 9 ge1 ! (19 Ad4 �d6 20 A xf6 A xf6 2 1 Ag4 gd8 22 ge1 0-0 23 gxd7 gxd7 24 �xd7 �xd7

23 d6! �x c7 (23 . . . gc8 24 A d5+ - forcing the king u nder the X-ray of the bishop at c5 - 24 . . . c;!? f8 25 A b7 !! +- /!i) xc5 26 A x c8

57


'flIxc7 27 dxc7 file7 28 A h3) 24 Ad5+ ! (an im portant intermediate move ; 24 d xc7 ? fil xc5 25 Axa8 fileT t) 24 . . . �f6 (24 . . . �f8 25 dxc7+ fil x c5 26 A xa8 fil e7 27 f!bd1 +-) 25 Ad4+ �f5 (25 . . . �g5 26 dxc7 Axd4 27 Axa8 filgf6 28 A b7) 26 dxc7 (26 Axg7 'flIxd6 (26 . . . 'flId8 27 h4, threatening Ae6+) 27 Axa8 is also in White's favour, but there is no need to com plicate matters) 26 . . . Axd4 27 A xa8 +-.

20 Iife1 20 A a7 ?! was too su perficial : 20 . . . f!c4 21 'flIa3 f!a4 22 'flIc1 'flIc3 23 f!e1 + � d8, but also strong was the simple suggestion of my silicon assistant - 20 Ad6 'flId4 21 f!fe1 + �f7 22 Ag4 'flIx b4 23 Ae6+ � e8 24 A x b4 f!d8 25 Axg8+ fil e5 26 Ae6. 20 . . . Iixe5 21 Iixe5+ Axe5 22 Iie1

(23 . . . f!c4 24 Ae6+ � g7 25 'flIa3) 24 'flIe4 !! (piece activity first and foremost ! - 24 'flIf4+ ttlgf6 wou ld slow down the attack and al low Black to complete his development), and now : A1 ) 24 . . . fil e5, and the poor coord ination of Black's pieces prevents h i m from keep ing all the material - 25 Ae6+ (25 'flIe3 fil xg4 26 'flIxc5 Axe1 -+) 25 . . . �e7 (25 . . . �g7 26 'flIe3 ±) 26 'flIe3 � d6 (26 . . . f!c4 27 f!d1 ) 27 f4 ;

A2) 24 . . . A x e1 25 A xd7 (25 'flIe6+ is also strong, as after 25 . . . � g7 26 'flIxd7+ � h6 27 'flIe8 f!c4 28 'flIf8+ � g5 29 Ae2 Black is defenceless, despite his h uge material advantage, e. g . 29 . . . f!c1 30 h4+ � x h4 31 'flIf4#) 25 . . . Ac3 26 g4 ! (26 'flIe3 f!c4 27 d6 Ad4 +). As Black's knight and rook are un able to come into play, I don't see how he can save the game, e. g . 26 . . . g5 (26 . . . � g7 27 g5) 27 Ae6+ (27 h4) 27 . . . � g7 28 'flIe3 . B) 23 'flIa3 is also good - 23 . . . Ac3 24 'flIxa6 (24 f!e2 ttl gf6 25 'flIx a6 f!e8 + ; 24 f!e6 !?) 24 . . . Axe1 25 'flIe6+ �f8 26 'flIxd7 f!c1 27 g4 ! and despite his big material advantage, I be l ieve that Black cannot save the game, as his pieces, including the king , are too poorly placed .

23 'tJd6! That's it. Black loses material .

22 . . . Iie4?

23 . . . �e7

Losing on the spot. I considered 22 . . . � f7 ! to be best, but, being short of time, Eric was unable to find this defence : A) 23 Ag4!? (23 f!xe5 f!c4 -+) 23 . . . Ac3

The knight comes i nto play too late. If 23 . . . ttlgf6 24 f!xe5+ fil x e5 25 'f!Ix e5+ � f7 26 'flIe6+ � g7 27 'flIe7+ � h6 28 h4 +-.

24 'tJxa6 +Or 24 'flIe6 � d8 25 'flIxa6. 24 . . . Iif8 25 d6 �e8 25 . . . filf5 26 Ac6 +-.

26 Ae6 �d8 27 Axd7 Ae3 27 . . . � xd7 28 'flIx b5+ f!c6 29 'flIxe5 +-. 28 lie1 Ad4 29 lixe4 Axf2+ Black re signs

58

Game 11

Gelfand -Adams, Hakoah Open Tournament, Sydney 1 988

Game 1 1

14 ghe1 0-0-0 15 Aa6 is an alternative plan , but I wanted to play my knight to e5 , where it puts pressu re on Black's position .

B o r i s G e l fa n d - M i ch a e l Ad a m s Hakoah Open To u rn a m e n t , Syd ney 1 9 88 Caro - Kann Defence [B 1 7J

14 . . . 0-0-0 1 5 Ilhe1 c!>b8 Wh ite also had to pay attention to 1 5 . . . c5 16 d xc5 A x e5 (16 . . . A x c5 is safe, with a typical position, slightly better for White) 1 7 �x e5 �x e5 1 8 gxe5 ttlg4 (18 . . . ttl d7 19 ge3 tD xc5 20 Af1 ;t) 19 ge2 Axg2 20 cxb6 axb6 21 gg1 Af3 22 gee1 with advantage.

This was my first ever game with M ichael Adams, to be followed by many more, in cluding a Candidates Match. But the unique position at the end of the game was the rea son I decided to include it in this book.

1 e4 c6 2 d4 d5 3 �d2 dxe4 4 �xe4 �d7 This variation has been very popular si nce the late -1 980s, mainly thanks to its main ad vocate, Anatoly Karpov. M ickey Adams has also employed it for many years. In the 21 s t century, thanks to the efforts of Bareev and Dreev, 4 . . . Af5 has also been revived at high level .

5 �g5 �gf6 6 Ad3 e6 7 �1 f3 Ad6 The poi nt of Wh ite's idea is 7 . . . h6 8 ttl x e6 'ffJe7 9 0-0 fxe6 1 0 Ag6+ with an attack, as in the famous game Deep Blue- Kasparov, New York 1 997 .

16 c!>b1 Hhe8 17 f3 It would have been a mistake to play 17 c4 c5 18 Ac3 cxd4 19 Axd4 A b4, but serious con sideration should have been given to 17 tDc4 Af8 (1 7 . . . Af4 1 8 Axf4 � xf4 1 9 �e5+) 1 8 g3 tD d5 1 9 tD e5 , simi lar to the game.

17 . . . �d5?! M issing White's 1 9 th move. It was correct first to take the king i nto safety by 1 7 . . . � a8 18 a3 ;t , but not immediately 17 . . . c5 18 dxc5 A xc5 1 9 Af4 Ad6 20 A b5.

18 a3 f6 a

8 1!re2 h6 9 �e4 �xe4 10 1!rxe4 �f6

b

11 1!re2 b6 12 Ad2 Ab7 1 3 0-0-0 1!rc7 b

c

d

19 c4!

5 4 3 2 �� 9

14 �e5

h

e

3 1""'""""'-"'2

e

6 L-.J=::.I 5 4 3

d

8 7 6 5 4

Nowadays 10 . . . �c7 1 1 �g4 �f8 1 2 0-0 c5 is the main line. Despite the fact that the black king is stuck in the centre, so far Wh ite has failed to show a convi ncing way to gain an advantage.

a

c

With this strong i ntermed iate move Wh ite gains the in itiative.

19 . . . fxe5 20 dxe5 Af8 21 cxd5 exd5

\r

Clearly the best practical decision , giving B lack cou nter-chances on the q ueenside. He would have been condemned to passive defence after other repl ies : 21 . . . gxd5 ± , or 21 . . . cxd5 22 Ag6 ge7 23 gc1 Ac6 ± .

59


22 14 cS 23 Ag6!

31 AhS There is noth ing for the bishop to do at g6, so it returns to the defence. If 31 �e2 c3 !? 31 . . . c3 32 A13 gbS 33 ge2 White has a winning position , but to fin ish the game he sti l l has to repel Black's attack on the b -file. 33 . . . a6 34 Axb7+ l:lxb7 3S .dS .eS 36 .c4 �a7 37 �a2

Wh ite increases the pressure.

23 . . . ge6? Better was 23 . . . ge7 24 e6 d4 with a defend able position.

24 1S gc6 2S e6 Black faces serious problems i n trying to stop the wh ite pawns.

2S

. . .

•e7

8

25 . . . Ae7 26 g3 � a8 27 Af4 'ffc8 28 Ae5 +-.

8 7 6 5 4 3

26 .eS+ �aS

7 t=� 6 5 4

5 4 3

8 7 �-=-,I 6 5 4 =I �

a b c 37 . . . gb4! Black was pinning all his hopes on this trick. 38 .d3 Th is does not yet throw away the wi n , but also strong was 38 �d5 ! (not 38 a x b4? �a4+) 38 . . . gb7, an d now : A) 39 gde1 � b5 (39 . . . Ae7 40 �c4, and as the rook has moved from d1 , 40 . . . g b4 no longer works) 40 b x c3 Ae7 !! with counter chances (but not 40 . . . d x c3 41 e7 A x e7 42 �xb7+ ! � xb7 43 gb1 , or 40 . . . �a4 41 gb1 (41 e7 A x e7 42 g b1 ) 41 . . . gx b1 42 �d7+ �xd7 43 exd7 +-) ; B) 39 b x c3 �a4 40 e7 ! �x d1 (40 . . . c4 41 �xb7+ (or 41 �xd4+) 41 . . . � xb7 42 gb1+, or 40 . . . A x e7 41 g b1 , and if 41 . . . gx b1 42 gxe7+, as indicated by Adams) 41 �x b7+ (41 �xc5+ g b6 42 g b2) 41 . . . � x b7 42 g b2+ � c7 43 exf8 � . 38 . . . ga4! Focusing on a new target - the a3 pawn. After 38 . . . �a4 39 e7 �b3+ 40 � a1 ! (40 � b1 �xa3) 40 . . . cx b2+ 41 � b1 White wins.

2 �------�--�

�

27 Axh6! +Th is l ittle tactical shot increases Wh ite's ad vantage.

27 . . . c4 Black's only chance is a counterattack (27 . . . g x h6 28 f6 +-) .

2s Ad2 If 28 h4, then 28 . . . c3 ! . I was not satisfied with 28 f6 g x f6 29 A xf8 'ff x f8 ! (29 . . . gxf8 30 'ffx d5 +-) 30 'fff5 'ffe7 (or 30 . . . gc7 31 gf1 'ffd 6 32 'ffx f6 d4 with counterplay) 31 gxd5 gcd6 ;t , when White faces techn ical difficul ties.

2S . . . d4 29 Ab4 gcS! 30 Axcs bxcS At the cost of the exchange Black has avoided immediate disaster. Now he is pin ning his hopes on the b -fi le.

60

Game 11

Gelfand - Adams, Hakoah Open Tournament, Sydney 1988

39 .e4 But not 39 b xc3 ?? gxa3+ ! (40 c!> x a3 c4+) , while if 39 f6 g xf6 40 � h7+ A e7 41 � f7 �d8. But simple and strong was 39 e7 ! �f7+ (39 . . . Axe7 40 �e4 �f7+ 41 �e6) 40 ge6 Axe7 41 �c2 gaS 42 � b3 gb5 43 gxa6+ , winning.

39 . . . Ad6 40 e7 It was time to complete the queen 's tour d5-c4-d3-e4-d5 with 40 �d5 � b5 41 gf2 cxb2, and now:

A) not so clear is 42 �xd6 b1 � + 43 gx b1 �xa3+ 44 c!> x a3 �a5+ 45 c!> b3 � b4+ (45 . . .' ffc3+ 46 c!> a4 �c4+ 47 c!> a3 �c3+ 48 gb3 �a1 + 49 ga2 �c1 + 50 c!> a4 �c4+ 51 c!> a5) 46 c!> c2 �c3+ 47 c!> d1 �d3+ 48 � e1 �e4+ ! (48 . . . �x b1 + 49 c!> d2 � b2+ 50 �d3 +-) 49 c!> f1 �x b1 + 50 c!> e2 �e4+ 51 �d1 �d3+ and I don 't see a win ;

(41 . . . �b5 42 gx b2 gxa3+ 43 c!> b1 ) 42 gde1 b1 � + 43 gxb1 .

41

. . .

• b5!

An unexpected resource! And I failed to see the point of it.

42 eS.?? It was time to agree to a draw with 42 gx b2 gxa3+ 43 c!> x a3 (43 c!> b1 �a4 -+) 43 . . . c4+ 44 c!> a2 �a4+ 45 c!> b1 �x d 1 + 46 c!> a2 = , or t h e com puter-style line 4 2 ge3 !? d x e3 43 �xe3 b1 �+ 44 gx b1 �c4+ 45 gb3 �c2+ 46 g b2 ge4 47 e8 � gxe8 48 �x e8 �c4+ 49 c!> b1 �d3+ = .

42 . . . gxa3+ ! 43 �xa3 If 43 c!> b1 ga1 + 44 c!> c2 b1 �+ (d iagram) with mate.

B) 42 gx b2 ! (42 � b3 !? i s also possible) 42 . . . gxa3+ 43 c!> b1 ga1 + 44 c!> xa1 �a4+ 45 'ffa2 (that's why the queen h as to be on d5 !) 45 . . . �xd1+ 46 gb1 +-.

40 . . . cxb2 a

b

c

8 7

d

e

f

9

h 8 7 6 5 4 3 2

5 3 2 a

b

c

43

. . .

c4+

One of the most amazing positions I have ever had ! Despite being a queen , rook and the exchange up, Wh ite's position is hope less and he gets mated in 22 moves, accord ing to the com puter. a

b

c

41 .e6?

5 4 3 2

Being short of time, I lost the thread of the game. Wh ite could sti ll have won by 41 gxb2 gxa3+ 42 c!> x a3 (42 c!> b1 �x e7 ±) 42 . . . c4+ 43 g b4 A x b4+ 44 c!> b2 ! Ac3+ (44 . . . �b5 45 gxd4 A xe7+ 46 c!> c2) 45 c!> c1 �a4 46 gxd4 +-, or again 41 �d 5 ! A x e7

a

61

e

f

9

h 8 7 6 5 4 3 2

8 7

{r

d

b

c

{r

My Most Memorable Games 44

�a2 b1 0+ ! 45 Ilxb1 Oa5+ 46 Oa4 The only attempt to delay mate (46 � b2 46 . . . �c3+ 47 � a2 �a3#). 46 . . . 0xa4+ 47 �b2 a

b

c

d

..

e

8 7

now my opponent had to seal a move. I wanted to resign without resuming, but a friend of mine convinced me to go and check the sealed move, and he was right.

47 . . . £a3+11

8 7 6 5 4 3 2

5 4 3 2

M ickey failed to notice the winning 47 . . . c3+ 48 � c1 Af4+ .

48 �a1 Now Wh ite can hold the position .

48 . . . £b4+ 49 �b2 £a3+ 49 . . . �a3+ 50 � c2 �d3+ (50 . . . d3+ 51 � d1 �a4+ 52 � c1 A a3+ 53 gbb2 A x b2+ 54 � xb2 �b3+ 55 � c1 ) 51 � c1 , or 49 . . . c3+ 50 � c1 A a3+ 51 geb2 with a d raw in every case.

The tournament was played with the strange time control of 46 moves in two hours, so

50 �a1 £b4+ Draw agreed

*** Game 1 2

we were youngsters and were both trying (successfu lly) to qualify for the fi rst time for the Cand idates M atches : 5 . . . Ag7 6 Ae3 �f6 7 �c3 0-0 8 Ae2 d6 9 0-0 Ad7 10 �d2 (the alternatives here are 10 �c2 and 10 � b3 , avoiding exchanges wh ile maintain ing the space advantage; however, I didn 't like moving my knight to an awkward square) 10 . . . �xd4 1 1 Axd4 Ac6 1 2 f3 a5 13 b3 �d7

Boris Gelfand - Vidmantas Malisauskas G MA O p e n To u r n a m e n t , M oscow 1 9 89 Sicilian Defence [836] 1 �f3 c5 2 c4 g6 3 d4 cxd4 4 �xd4 �c6 5 e4 Normal ly this position arises by another move order: 1 e4 c5 2 �f3 � c6 3 d4 cxd4 4 � xd4 g6 5 . c4 . The M ar6czy Bind is con sidered to be a passive, but very solid set up for Black. In contrast to the Hedgehog System , he doesn 't have a potential weak nesses on b6, but on the other hand he has l ittle prospect of making the . . . d6-d5 break. I have always liked and I sti l l like play ing Wh ite's side i n this system , as I believe he enjoys a space advantage without much counterplay for his opponent.

5 . . . �f6

14 A e3 (I d i d n 't see the need to exchange bishops, as it would g ive more scope to Black's q ueen) 14 . . . � c5 15 gab1 � b6

Against Vishy Anand I faced another vari ation in our first Interzonal Tournament, when

62

Game 1 2

Gelfand - Malisauskas, G MA Open Tournament, Moscow 1989

16 gfc1 �b4 (as Black has no active play at all, he tries to provoke some weaknesses i n White's position) 1 7 gc2 gfc8 1 8 �c1 � b6 19 a3 �d8 20 �d2 e6 21 Af1 Ae5 22 b4 (22 �e2 !?) 22 . . . a x b4 23 a x b4 �a4 24 �e2 'l' h4 25 g3 �e7 26 �d4 A e8 27 g bc1 ;1;; , keeping a long -term edge (Gelfand -Anand, Manila I nterzonal 1 990).

ttlg4 (13 . . . b5?! 14 b4 ! - the point of White's 1 2 th move - 14 . . . �c7 1 5 e5 ! d x e5 1 6 � x b5 and the b - and c - pawns pose more of a threat than the doubled e -pawns) 14 ttld5 (I had to choose between the game move and the alternative 14 Ad4 A xd4 15 �xd4 ttlf6 !? ;I;; , but not 1 5 . . . �e5 ?! 1 6 �x e5 ttl x e5 1 7 f4 ttlc6 1 8 f5 Ad7 1 9 f6 ± Ivanchu k) 14 . . . �xd2 1 5 A xd2

6 �c3 d6 7 J.e2 �xd4 I showed q u ite a good understanding of such positions when I was stil l a kid : 7 . . . Ag7 8 Ae3 �xd4 9 Axd4 0-0 10 0-0 Ae6 11 gc1 'l'a5 1 2 �d5!? Axd5 13 exd5 a6 14 Ac3 �c7 15 b4 �d7 (15 . . . a5 !?) 16 Axg7 � xg7 17 c5 !? gac8 18 �d4+ � g8 19 Ag4 d x c5 20 b x c5 f5 21 Ae2 ttl xc5 22 d6 ! ttl e6 (22 . . . �xd6 23 Ac4+) 23 A c4 +-, and I won my fi rst game against one of my contem poraries, with whom I was to play dozens of games in the forthcom ing decade, and who was my team colleague in numerous events, first for the Belarus team and then for Israel. l Iya be came a very strong player, who once even crossed the magic 2700 mark (Gelfand Smirin, Grodno 1 980).

a 1txd4 iLg7 a

b

c

d

e

6 5 4 3

L..._ ._ _ _ _ _ _ -=--_--I

lf

9 0-0 Here I think that White has a number of de cent plans, which I have also tried : 9 Ae3 Ae6 10 0-0 0-0 11 �d2 �a5 1 2 gab1 (a subtle prophylactic move) 12 . . J Uc8 13 b3

ttJ

Analysis diagram after 15 fi.xd2

1 5 . . . � f8 (15 . . . A x d5 !? 1 6 exd5 �f6 1 7 Ad3 �d7 with a position similar to the main game) 1 6 Ag5 f6. Black delays the decision about what to do with the knight at d5. I n my opinion, it was better to do this now, as in any case he cannot tolerate it for long : 16 . . . A xd5 17 exd5 ;1;; , or 16 . . . �f6 17 � xf6 Axf6 18 Axf6 exf6 19 gbd1 � e7 when only Wh ite has any winning chances. H owever, Black has only one weakness on d6, which can easily be protected . 1 7 Ad2 ttl e5 (17 . . . ttl h6!? i ntending . . . f6f5) 18 f4 �c6 (18 . . . ttl d7 !?) 19 f5 ! (giving u p t h e e 5 square, b u t shutting t h e Ag7 o u t of the game) 19 . . . Af7 (1 9 . . . gxf5 20 exf5 Axd5 21 cxd5 ttld4 22 Ad3 gc5 23 Ae3 gxd5 24 Ae4 ge5 25 Ax b7 ±) 20 g4 � e5 21 gbc1 (21 h4 b5 !?, and if 22 cxb5 gc2 with counter play) 21 . . . Axd5?! (Black feels exceptionally squeezed , as his pieces are slowly bei ng suffocated) 22 exd5 � f7 23 Ae3. Black's problem is that he is u nable to ex change his bishop : 22 . . . ttlf7 23 h4 ! A h6 (23 . . . g5 24 h5 ttle5 25 Ae3) 24 g5 Ag7 25 Ag4 (25 fxg6 25 . . . h xg6 26 h5 g x h5

63


27 A xh5 ttl e5 28 g6) 25 . . . ttl e5 26 A h3 ± or 26 fxg6 !? ttl xg4 27 g x h7. 23 . . . b6 24 h4 and the famous tech nique of the Swed ish player (whose games had a really big i nfluence on me when I was a teenager) d i d n 't help h i m in the game Gelfand- Andersson, (Akiba Rubinstein M e morial , Polanica Zd roj 1 997). Or 9 Ag5 0-0 1 0 �d2 Ae6 1 1 gc1 �a5 12 f3 gfc8 13 b3 a6 14 ttla4 �xd2+ 15 � xd2 ttl d7 16 h4 gc6 17 h5 Af6 18 Ae3 g5 19 g3 gd8 20 ttl c3 (20 h6!?) 20 . . . A xc3+ 21 gxc3 f6 (21 . . . h6 !) 22 h6! (an instructive moment : Wh ite fixes the weakness on h7 , which will always restrict Black) 22 . . . ttl c5 23 A xc5 gxc5 24 f4 � h8 25 b4 gc7 26 f5 Af7 27 ghc1 with a serious endgame advantage (Gelfand - Dautov, Klai peda 1 985).

9

• . .

0-0 10 1Je3

My trainer for many years, Al bert Kapengut, recommended this system to me. He played the Mar6czy Bind with Black for many years and he was very fami liar with its weak points.

ttl xd6 20 ttl a4 t ) 1 9 exd6 ttl xd6 20 gfd1 (20 ttld5 was better) 20 . . . a6 (20 . . . A xc3!? was probably the lesser evil) 21 ttl d5 ± and the strong knight and the potential advance of the queenside pawns make Black's task a difficult one (Gelfand -Topalov, Melody Am ber, Monaco blind 2002) . 12 Elac1 1Jb6 In cram ped positions it is recom mended to exchange pieces. 1 3 b3 1Jxe3 14 .txe3 �c5 14 . . . gfc8 was worth consideri n g, i ntend ing to meet 15 ttld5 with 15 . . . � f8 . 1 5 �d5! N 15 f3 was also possi ble, but I jumped at the chance to force an exchange which I hoped wou ld be favourable for me. 15 . . . .txd5 Or 15 . . . ttl xe4 16 ttl xe7+ � h8 17 Af3. 1 6 exd5 a5 a

b

c

d

e

8 7 6 5 4 3

10 . . . .te6 An expert on this system chose agai nst me a plan that keeps the queens on the board : 10 . . . ttld7 11 Ad2 a5 1 2 gad1 ttlc5 13 b3 Ad7 14 ttl d5!? e6 15 ttl c3 (a target on d6 has been created) 15 . . . Ac6 16 A e1 'f!ie7 17 f3 gfe8 1 8 Ad3 (Black's problem , as usual i n this variation , is that h e doesn 't have any ac tive counterplay) 18 . . . b6 19 Ac2 Ae5 20 Ag3 gad8 21 a3 A xg3 22 h xg3 'f!ic7 23 b4 ttl d7 24 'f!id4 axb4 25 axb4 ttlf8 26 f4 ± (Gelfand Pigusov, Sverdlovsk 1 987) .

6 5 4

17 Elc2! ;t A strong prophylactic move. Black was threaten ing to play his bishop to a3 via b2 , followed by . . . a5-a4 , either i m mediately or after . . . ttl c5-a6 and . . . A a3-c5. Now Wh ite can combine ideas of a breakthrough on the queenside after gb1 , a2-a3, b3-b4 and c4c5 etc. with play on the kingside. As he has more space, he is able to transfer his pieces from one flank to another more easily than his opponent. 17 f5

1 1 .td2 �d7 11 . . . 'f!i b6 12 b3 'f!ixe3 13 Axe3 gfc8 14 gac1 � f8 15 f4 ! (Black has delayed playing his knight to c5 and I decided to place my pieces more actively) 1 5 . . . Ad7 (15 . . . ttl g4 16 A xg4 A xg4 17 f5 and the bishop is in danger) 16 Af3 A c6 17 e5 ttl e8 (17 . . . A xf3 ? 1 8 exf6) 1 8 A xc6 gxc6 ( 1 8 . . . b x c6 1 9 exd6

. . •

64

Game 1 2

Black apparently does not sense the danger of his position. He should have looked for counterplay on the queenside by 1 7 . . . .§.fc8 18 .§.b1 (or 1 8 .§.fc1 ; after 18 g4 �a6, threat ening . . . b7-bS-b4 or . . . as-a4 , Black gains good play) 18 . . . � a6 19 a3 bS 20 b4 a x b4 21 axb4 �c7 22 .§.d1 ;t ; or taken prophylac tic measures on the kingside - 1 7 . . . hS!?, and if 18 f4 fS.

The alternative was 24 Ad3!? �cS 2S Af1 .

24 . . . Ag7 25 Ilf1 I1f8 26 Ag4 'i!1h8

8 7 6 5 4 3

18 g3 Ae5?! Intending . . . fS-f4 and thus provoking f2-f4 . However, Wh ite has no reason to avoid it. 18 . . . hS was better. Too subtle. The im mediate 20 g4 ! was better.

e

9

h

�

27 f5?!

20 . . . Ilfe8?

27 Ae6 and only then f4-fS was better.

e

27 . . . gxf5?

8 7 6 5 4 3

6 5 4 3 2 a b c d

19 f4 Af6 20 Ilb1 ?!

a b c d

ttJ

Gelfand - Malisauskas, G MA Open Tournament, Moscow 1 989

6 5 4

�------��- �

20 . . . hS ! was much stronger, as Black would get the h-file if White were to play h2-h3 and 93-g4 (21 h3 � f7 intending . . . '§'h8).

21 g4! fxg4?! Black was afraid of being left with a weak f5 pawn, but in any case this was his best chance : 21 . . . � f7 22 gxfS gxfS 23 Af1 ± .

22 Axg4 Ileb8 Threatening . . . b7-bS .

Both players missed the tactical trick 27 . . . gS ! 28 Af3 (after 28 .§.g2 h6 Wh ite retains a definite plus) 28 . . . .§.xfS ! 29 Axe4 .§.eS =, re gain ing the piece.

28 Axf5 �e5 29 Ilef2 The first half of the plan has been com pleted : White has opened lines on the kingside. Now he will combine an attack on the h7 pawn with threats of a pawn advance on the oppo site wing. Th is is the pri nciple of two weak nesses - as N imzowitsch taught us !

29 . . . Ilf7 30 Ae2 Rook exchanges favour Wh ite, as then his king is able to come into play.

30 . . . Ilxf2 31 Ilxf2 Af6 32 �1 ! 32 .§.g2 .§.f8 would al low the king to be to cut off, si nce if 33 A h6?? Ad4+.

32 . . . Ilg8 33 Ilg2 Ilxg2

23 Ae2! White is carefu l to prevent Black's counter play.

33 . . . .§.f8 34 � e2 and in any case Black can not make use of the f-file. 34

23 . . . �e4 24 Ilbe1

65

'i!1xg2 'i!1g7 35 �3 Ae5 36 h3


a b c d

e

f

9

•

h

8 7 6 5 4 3

Wh ite's extra pawn, his king is too active. The threat of transposi ng i nto a won pawn ending should prove decisive.

8

40 .lc8

6 5 4

Avoiding the last trap 40 � d5?? �f6 41 Ax h7 ti)c3#.

40 . . . �c3+ 41 �d3 .......-/-,

a b c d

e

f

9

2

41 � f5 !? ti) xa2 42 Ax b7 +-.

41 . . . �xa2 42 .lxb7

h

a b c d 36 . . . e6 Black gets rid of his e7 pawn, but this al lows the white king i nto his position . The fol lowing line shows Wh ite's plan : 36 . . . Af6 37 Ad2 b6 38 � g4 � g8 39 � h5. First of all he has brought his king i nto play, and now he begins an advance on the other side: 39 . . . � g7 40 A h6+ � g8 41 A c1 � g7 42 a3 Ac3 43 Ag5 Af6 44 A h6+ � g8 45 Ad2 ti) a6 46 b4 axb4 (46 . . . A b2 47 bxa5 bxa5 48 A xa5 A xa3 49 Ad8 +-) 47 axb4, i ntend ing 48 Ad3 ti) c7 49 A e3 and Black cannot protect both his b6 and e7 pawns. The most tenacious defence was 36 . . . h5 37 Ad2 b6 38 a3 (th reaten i ng b3-b4) 38 . . . ti) d7 ± (if 38 . . . A b2 39 � g3 A xa3 40 � h4).

e

f

9

•

h

8 7 6 5 4 3 2

8 6 5 4 2 a b c d

e

f

9

h

The bishop pair together with his active king guarantee White a win.

42 . . . �b4+ Losing i mmediately. Black could have pro longed the resistance by 42 . . . A b2 (or 42 . . . A c3) 43 Ad2 (43 A b6 Ac3 44 � c2) 43 . . . ti) b4+ 44 � e4 � f6 45 h4.

37 dxe6 �xe6 38 .lfS �d4+ 38 . . . � f6 !? is no more than a trap : 39 Ag4 ! (39 Axh7? ti)g5+ 40 Axg5+ �xg5 and White cannot win such a position).

43 �e4 Threatening A b6.

39 �e4 �e2 39 . . . ti)xf5 40 �xf5 �f7 41 Ad2 a4 42 bxa4 is unlikely to save the game, si nce, apart from

43 . . . .lc3 44 .lf4 .leS 45 .lxeS+ dxe5 46 cS Black resigns

***

66

Game 13

Game 1 3

B o r i s G e l fa n d - Va s i ly I va n ch u k U S S R J u n i o r Tea m C h a m p i o n s h i p , Kram ato rsk 1 9 89 King 's Indian Defence [E94] This was my farewell game i n j u n ior chess, and it became a very memorable one. It was played in the USSR Junior Team Cham pion ship, in which I was playi ng for Belorussia. We had a very strong team (myself, Smiri n , Atlas, Sagalch ik a n d Aleksandrov) . Four of these five players are now g randmasters, with an average rating approaching 2650 ! We had played with more or less the same team for several years, but had never man aged to finish among the medals. Now it was our last chance and we had to score at least 3112-1 % against the strong U krai n ian team , led by Vasily Ivanch uk. Vasi ly had to leave for a tournament in Biel, (which he won), so our game took place on the rest day, one day before.

1 d4 �f6 2 c4 g6 3 �c3 �g7 4 e4 d6 5 Ae2 0-0 6 �f3 �bd7 7 0-0 e5 8 �e3 c6 9 d5 c5 10 �e1 �e8 a

b

c

d

e

8 7 6 5 4

tt:J

Gelfand - Ivanchuk, USSR Junior Team Championship, Kramatorsk 1 989

6 5 4 3 2

I...-_______...._ .;;.... ----I

round of the USSR Under- 1 8 Championship, Yurmala 1 985 (p. 68). White's idea is to fight against . . .f7-f5 , as I demonstrated in my game against Romero Hol mes. 11 �d3 had been played in nearly all the previous games.

11

. . .

• h4?

Vasi ly is following my game w ith Savchen ko (he was an eye -witness to it) , but I had made a deep analysis of this game (as was the case with most of my games) with my trainer Al bert Kapengut and we had prepared an i m portant improvement. In fact, my work with this wonderful trai ner consisted mainly in goi ng through the most complicated of my games, especially those where I experienced difficu lties. 11 . . . f5 (11 . . . � h 8 !? ;t ) 1 2 gxf5 g xf5 13 exf5 � b6 14 � h1 A xf5 1 5 E!g1 (Wh ite has cov ered his king and i ntends to use the open g -fi le to his advantage) 15 . . . �f6 (15 . . . e4 16 �d2 �f6 17 E!g3 �bd7 18 �g2 ;t , intend ing E!g1 and �f4) 16 �f3 ! (threatening 17 �h4 or 1 7 �g5) 1 6 . . . �g4 ?! (16 . . . �e4 1 7 � x e4 Axe4 1 8 E!g4 ! Af5 1 9 E!g3 Ae4 20 � g1 ± in tending �g5, or 16 . . . � h8 17 E!g3 ;t ) 17 �g5 � x e3 (17 . . . �xg5 18 E!xg4 �e7 19 E!g3 ±) 18 fxe3 e4 ?! (18 . . . � h8 19 Ad3 ! e4 20 �cxe4 Axb2 21 E!b1 ) 19 �e6 ! ± Axe6 20 dxe6 �e7 21 �xe4 �xe6 22 Ad3 ! � h8 23 E!g3 ! +- and Wh ite has a mating attack (Gelfand - Romero Holmes, Hoogovens, Wij k aan Zee 1 992) . c

d

e

8 7 6 5 4 3

lf

I had played a n u m ber of games with this system for Black (for a certain time it was my main weapon) and I was famil iar with this position.

11 94!?

6 5

I...-....--..:. .::.... ---. ... _ .:.. ___...._ .::... ----'

This strong plan was introduced by Stanislav Savchen ko against me i n the pen ultimate

1 2 �h1 !

67

lf


Wh ite carries out the same idea as in the game against Romero, whereas Black has wasted time on his queen sortie. 12 g5 � b6 1 3 �g2 � h3 14 �d2 Ag4!? 1 5 f3 A h5 16 b4 led to unclear play in the Savchenko Gelfand game. Even though Wh ite is prob ably better here, Black always has counter play associated with . . . f7-f6, as White has already pushed his pawn to g5. I managed to win this game, which brought me first place in the afore -mentioned Under-18 Champion ship. 12 . . . �h8 12 . . . A h6? 13 g5 Ag7 14 �g2 and the queen is trapped . 13 Ilg1 Threatening to cut off the queen by g4-g5 . 1 3 •e7 14 a3 14 �d3 !? was also possible, trying to save a tem po. 14 . . . �dI6 If 14 . . . f5, then 15 g5 ! ± (15 gxf5 gxf5 16 exf5 e4 ;t ). Th is is a typical idea - the g5 pawn excludes both knights and the g7 bishop from the game. And si nce Black has al ready played . . . f7-f5 , there is not the . . . f7-f6 breakthrough. 15 b4 b6 1 6 �d3 �g8 17 a4 15 1 7 . . . a5 1 8 bxc5 bxc5 1 9 gb1 ± and Wh ite penetrates on the b-file. With 17 . . . A h6 Black could provoke g4-g5, but two tempi are too high a price : 1 8 g5 Ag7 1 9 a5 gb8 20 ax b6 axb6 21 �b5 ± . . . .

18 g5?! An inaccuracy, which allows Black a chance to gain counterplay. With 1 8 a5 !? gb8 1 9 ax b6 axb6 20 f3 ± White would have re tained all his advantage.

18

. . .

•17?

Black misses his chance : 1 8 . . . h6! (18 . . . f4 1 9 Ad2 h6 20 g x h6 � x h6 21 gxg6 ±) 1 9 a5 g b8 20 a x b6 a x b6 21 g x h6 (21 h4?! is not so good because of 21 . . . f4 22 Ad2 h xg5 23 h xg5 Af6 ! 24 Ag4 Axg5 and Wh ite can not win material : 25 A x c8 gxc8 26 �g4? � h7+ ! 27 � g2 f3+ -+) 21 . . . A x h6 22 A xh6 � x h6 23 f3 !? with only a slight advantage, as Black has managed to exchange his bad g7 bishop and bring his g8 knight into play (if 23 gxg6 �h7 with counterplay) .

19 a5 llb8 20 13 ± a

8 7 6 5 4 3 2

b

c

d

...

e

8 6 5 4 3 2

The position is strateg ically won , as the d5 and g5 pawns shut all Black's pieces out of the game and Wh ite is going to break through on the queenside.

20 . . . h6 21 a x b6 a x b6 22 h4 14 23 £12 hxg5 24 hxg5 £16 25 Ilg2!

6 5 4 3 2

Wh ite is not going to blu nder into a mate: 25 g xf6?? � h7+.

25 . . . £d8 All Black's pieces, apart from his queen , are on the back rank, which says a lot.

�------�--� �

68

Game 13

26 �b5 +- �e7 27 flh2+ 27 bxcS bxcS 28 �xd6! �xd6 29 �xeS �g7 30 AxcS ! would have won more quickly, but when you have a strateg ically won position you are not so keen to find a tactical solution .

�g8 28 £f1 28 bxcS bxcS 29 �xd6!? +-. 27

28

32 �xc5 �c8 32 . . . Ac8 33 �xc8 gxc8 34 �e6 �f7 3S cS. 33

�xd7 .xd7 34 £h3 flc7 35 £e6+ �g7 3S . . . gf7 36 � xc8 gxc8 37 �f1 followed by �h3.

. . .

. . .

•g7 29 �a7

36 �xc8 flxc8 37 .g1 ! Black resigns

29 ga7 !?

29

. .

. £d7

30 bxc5 bxc5

8

a

b

c

d

..

e

8 7 6 5 4 3 2

8 7 6 5 4 3 2

8 7 6 ..;;;;;;n_-. 5 """"'".. "�-J""'''' --I 4 3

a

�------��- �

31 £xc5!? A typical sacrifice. Black can do nothing against the c- and d-pawns, especially as he is unable to avoid the exchange of his key light-square bishop.

31 dxc5 31 . . . �c8 32 Af2 A xgS 33 cS and Wh ite is winning with material equal . . . .

ttJ

Gelfand - Ivanchuk, USSR Junior Team Championship, Kramatorsk 1 989

b

c

d

e

9 h

After 37 . . J 3f7 (or 37 . . . � b6 38 gh7+ ! � x h7 39 �h2+) 38 �g4 �d6 39 Axc8 �xc8 40 cS White wins. I am happy that this victory inspired my team-mates to win our match 4-1 and we thus finished among the medals. This game also brought me the best score on top board, ahead of Shirov, Bologan , Dreev, Ivanchuk, Akopian etc.

***

69


Game 1 4

F l o r i n G h e o rg h i u - B o r i s G e l fa n d G M A O p e n To u rn a m e n t , P a l m a de M a l l o rca 1 9 89 King 's In dian Defence [E8 1]

8

The G MA Open Tournament in Palma de Mal lorca was very memorable for me, as it was my fi rst big i nternational success (I won it ahead of more than 1 50 grandmasters) and it enabled me to receive invitations to such big tournaments as Li nares and Ti lburg .

1 d 4 �f6 2 e4 g 6 3 �e3 Ag7 4 e4 d6 5 f3 0-0 6 Ae3 e5!? I began employing this idea of Igor Glek i n 1986 and I liked i t immed iately. I have used it many times with g reat success, and I am surprised that it has never become really pop u lar. For exam ple, Garry Kasparov did not u se it at all against Anatoly Karpov in thei r 1 990 World Championsh i p M atch, but instead tried nearly all the other variations against the Samisch .

7 dxe5 White often transposes into a Benoni by 7 d5 or 7 tLlge2 tLlc6 8 d5 tLle5 .

7 . . . dxe5 S "xdS The attem pt to avoid the exchange of q ueens did not succeed after 8 A x c5 tLl c6 9 Ae3 tLld7 ! 10 !!c1 �a5 11 tLlh3 !!d8 ! 12 tLlf2 tLl c5 1 3 Ad2 A xc3 14 b x c3 Ae6 1 5 �c2 tLl e5 16 Af4 tLl xc4 and Black is winning (Dlugy - Gelfand , G MA Open , Palma de Mal lorca 1 989) . 8 e5 tLlfd7 9 f4 f6 1 0 exf6 exf6 1 1 Ae2 , an idea of Portisch , was popular for some time, u ntil the following game: 1 1 . . . �e8 ! N 12 Af2 tLla6 13 tLlf3 tLlb6 14 0-0 A h6 15 �c1 tLl a4 = (Beliavsky - Gelfan d, Mephisto, M u n ich 1 991 ).

7 6 5 4 3

6 5 4 3

�------�--� �

One may wonder why Black has given up a pawn , in addition exchanging the queens. H owever, he has strong cou nterplay on the dark sq uares and Wh ite has problems in de veloping his kingside pieces.

10 �d5 I have also faced other continuations, for example: A) 1 0 A a3 b6 1 1 tLl ge2 e6 12 !!d1 tLl d7 1 3 b3 (13 tLlf4 A x c3+ 14 b x c3 A a6 1 5 c5 =) 1 3 . . . A b7 14 g3 tLl de5 1 5 Ag2 tLld3+ 1 6 � f1 a6 !? 1 7 f4 b5 18 Ac1 b x c4 1 9 b x c4 f6 !? (1 9 . . . tLl a5 20 e5 A x g2+ 2 1 � xg2 tLl xc4 22 tLl e4 =) 20 A e3 ? (20 f5 ! e5 2 1 tLl d5 oo) 20 . . . tLl a5 ! =+= (Bren n i n kmeijer - Gelfan d , Eu ropean U nder-20 Championsh ip, Arnhem 1 987) ; B) 10 tLlge2 tLld7 1 1 Ae3 tLlde5 1 2 tLlf4 tLl b4 1 3 � f2 A e6 ! 14 tLl cd5 A xd5 1 5 tLl xd5 tLl c2 1 6 !!c1 tLl x e3 1 7 � x e3 e6 1 8 tLl c3 A h6+ 19 f4 g5 ! 20 g3 tLlg6 0-1 (Gunawan -Gelfand, M i nsk 1 986).

10 . . . �d7 Obviously it was not Black's i ntention to regain the pawn immediately : 10 . . . tLl xd5 1 1 cxd5 A x b2 12 !! b1 A c3+ 1 3 � f2 and Wh ite stands better.

1 1 Axe7 Th is was a novelty at the time, but later this line became popular, occurring in some games of Anatoly Karpov. Neither 11 tLl xe7+

S . . . IlxdS 9 Axe5 �e6

70

Game 14

fi1 xe7 1 2 Axe7 A x b2 nor 1 1 Aa3 e6 1 2 fiJ c7 .§.b8 poses any particular danger for Black.

a

13 fi1 xc8 ?! Axb2 14 Etb1 Ac3+ 15
13 . . . Axb2 14 gb1 Ag7 White has an extra pawn and a strong kn ight on d5, but I believe that the complete control of the dark-square bishop and the potential weakness of the a2 and c4 pawns give Black full compensation . b

c

d

e

f

c

d

e

f 8 6 5 4 3 2 'If

Al l the pieces have reached thei r best squares. It was too early for 1 7 . . . fiJ a4 18 Etxb7 Etab8 19 Etb3 ! ± .

8

18 Ae2?!

6 5 4 3 2 '---_______...._ .;;;... ----'

b

8 7 6 5 4 3

11 . . . �xe7 12 �xe7+
a

ttJ

Gheorghiu - Gelfand, G MA Open Tournament, Palma de Maliorca 1989

Safer was 1 8 fiJ xc5 gxc5 1 9 gb3 ! (19 gxb7 ga5) 1 9 . . . b5 (1 9 . . . Eta5 20 a3) 20 Etx b5 Etx b5 21 cxb5 Axd5 22 exd5 Etxd5 with a probabl e draw.

18 . . . �a4! 'If

The c3 square is critical for Black's counter play.

19 �3f4!

15 �e2!? Almost ten years later this position again oc curred in one of my games : 1 5 h4 fiJ b6 !? N 16 h5 Ae6 (16 . . . Ad7!?) 1 7 fiJ h3 Etac8 (17 . . . A xh3!? 18 Etxh3 Etac8 19 f4 (1 9 fiJ xb6?! ax b6 20 Etx b6 A c3+ !? is i n Black's favou r) 19 . . . fiJxc4 20 h6 00) 1 8 fiJ g5 A xd5 1 9 cxd5 .§.c2 with counterplay (Eh lvest - Gelfand , Ak iba Rubinstei n M emorial , Polanica Zdroj 1997). 15 fiJ h3 was Karpov's choice.

15 . . . �c5 Black doesn 't want to attack the c4 pawn, but rather the a2 pawn by . . . Etd6-a6 after he has developed his pieces. 15 . . . fiJ b6 16 fiJef4 Ad7 17 Ae2 gac8 1 8 o-o ;t .

G h eorg h i u rightly avoids 1 9 Etx b7 gxc4 20 Etxa7 Axd5 21 exd5 tDc3 with advantage to Black.

19

g5!

Keeping up the pressure. If 19 . . . fiJ c3 ?! 20 fiJ xe6+ fxe6 21 fiJ xc3 Axc3+ 22
20 �h5 Ac3+! 21
16 �c1 16 tDef4 Ad7 17 Ae2 Etac8 50 ; 16 fiJec3 !?

•. .

21 . . . Ad4 The best square for the bishop.

16 . . . Ae6 17 �d3 gac8!

22 h4

71

My Most Memorable Games r----:---:---�-__;:____, ..

8

2S . . . �xa2

8

The a-pawn decides the fate of the game.

26 fld3

6 5

The only move. If 26 gxb8 gxb8 27 g4 �c3 intending . . . as-a4 -+.

26 . . . Db1 + 27 Ad1

3 2

Or 27 gd1 gdb8 ! . a

The only chance for counterplay. 22 gx b7 gb8 would play into Black's hands.

22 . . . AxdS!? Securing the i m portant c3 sq uare for the knight. 22 . . . � c3 ?! 23 � x c3 A xc3 24 hxgS oo , or 22 . . . h6 23 h xgS hxgS 24 � hf6 with counterplay.

b

c

d

..

e

8

8 7 6 5 4 3 2

6 5 4 3 2

23 exdS? 27 . . . Ab6 ! - +

Probably the decisive mistake. 23 cxdS wou ld open the c -fi le for Black's rook, but would at least maintai n White's strong pawn centre : 23 . . . � c3 (23 . . . gc2 24 h xgS gdc8 2S gd1 ) 24 g b4 (the only move ; if 24 gx b7 � xe4 ! +) 24 . . . AeS (24 . . . A b6 2S gx b6 ! axb6 26 �f2 00) 2S hxgS and I don 't see a substan tial advantage in any of the following lines :

The final touch ! Wh ite could have com plicated matters after 27 . . . � c3 ?! 28 gxd4 � x d1 29 � e2 ! (or 29 tD g3 !?) 29 . . . � c3+ 30 � d3 gxh1 31 � xc3 with counterplay.

28 �f6 �b4 29 �e2 More stubborn was 29 gd2 Ae3 30 � e2 Axd2 31 � xd2 as ! (31 . . . g x h4 32 gx h4 �e7 33 tDe4) 32 hxgS a4 33 gxh7 a3 34 g6 fxg6 3S ga7 gd6 36 Ac2 gg1 , winning a piece for the a-pawn .

A) 2S . . . � xe2 ? 26 � xe2 gc2+ 27 � e3 ; B) 2S . . . � xe4 26 g3 ! ; C) 2 S . . . � xdS 2 6 exdS gc1 + 2 7 � f2 gxh1 28 f4 00 ;

2 9 . . . �xd3 3 0 �xd3 a s 3 1 h x gS a4 32 �e4

D) 2 S . . . � x a2 2 6 gb1 � c3 27 ge1 � x e2 28 gxe2 gc1 + 29 ge1 gdc8 30 � f2 (30 gh4 as 31 f4 gxe1 + 32 � x e1 Ac3+ 33 � d1 a4) 30 . . . g8c2+ 31 � e3 gc3+ 32 � f2 gxe1 (32 . . . g3c2+ 33 � e3 =) 33 gxe1 as 34 gd1 ! 00.

The last chance was 32 � c2 ga1 (32 . . . gb3 33 gx h7) 33 gxh7 a3 , when Black must win anyway.

32 . . . a3 33 �c3 Da1

23 . . . �c3 24 flxb7 flb8!

33 . . . A aS ! 34 tD b1 a2 was more aesthetic.

Inaccurate is 24 . . . ge8 ?! 2S �g3 =.

34 �c2 Ad4 35 �bS a2 36 �b3 Da8 White resigns

2S flb3

***

2S gx b8 2S . . . gxb8 26 Ad3 gb2 -+ .

72

Game 1 5

Gelfand - Kasparov, Linares 1 990

Game 1 5

B o r i s G e lfa n d - G a r r y Ka s p a rov L i n ares 1 9 9 0 King 's Indian Defence [E99] Linares was my first big tournament (or super-tournament as they are now called) and I was i nvited as a last-m i n ute substi tute. I guess that Luis Rentero (the tourna ment director and for many years the main force beh ind the Linares tou rnament) d e cided to give a young player h i s chance after my win in the G MA Open i n Palma de Mal lorca. The d rawing of lots gave me some tough opponents in the first few rounds (Kas parov, Salov, Beliavsky etc). And in the fi rst round I had to face Garry Kasparov hi mself. An encounter with a world champion is al ways a memorable event for a young chess player, especially the first time. And I ap proached it with full responsibil ity, intending to give of my best and to play an interesting game.

1 d4 ctsf6 2 e4 g6 3 ctse3 Ag7 4 e4 d6 5 Ae2 0-0 6 ctsf3 e5 7 0-0 ctse6 8 d5 f1le7 In the early 1 990s the King's I ndian Defence became very popular, than ks to the efforts of both players. And I enjoyed playing it with either colour.

12 . . . �h8 The entire plan with . . . h8, refrain i ng from an immediate . . . f5-f4 and vacating the g8 square, became extremely fash ionable after the game Karpov- Kasparov, G M A, Skelleftea 1 989 (although that was i n the 9 tDd2 variation). In our next game, played on New Year's Eve 1 992 , Garry chose the more 'ancient' way of playing : 1 2 . . . f4 13 g4!? (13 c5 is the more classical approach, but here too I tried to stop Black's play more radical ly) 1 3 . . . g5 14 b4 h5 1 5 h3 f7 (Black has to occu py the h -fi le and then th reaten to sacrifice in the vicinity of f3 or g4) 16 Ae1 gh8 17 g2 tDg6 (17 . . . gh6 !?) 18 c5 (safer was 18 Af2 h xg4 1 9 h xg4 tD h4+ 20 A x h4 gxh4 21 gh1 �h8 (0) 18 . . . h xg4 1 9 h xg4 •

9 f1le1 ctsd7 10 ctsd3 f5 1 1 Ad2 ctsf6 I already had some experience of this line with the black pieces : 1 1 . . . h8 12 b4 tDf6 13 f3 �eg8 14 c5 Ah6 15 Ek1 Axd2 16 �xd2 f4 17 tDf2 g5?! (better is 17 . . . h5 00 , prevent ing g2-g4) 18 g4! h5 19 h3 gf7 20 gc2 gh7 21 a4 hxg4 (21 . . . �f8 !?) 22 hxg4 �f8 23 g2 lfiJ.e7 24 gh1 and Wh ite has advanced quite far on the queenside, whereas as yet Black has fai led to create any real threats agai nst the opponent's king (Malaniuk- Gelfand, 56 th USSR

Championsh ip, Odessa 1 989) .

12 f3

19 . . . tDh5! (an incred ible tactical resource ! - if 1 9 . . . tD h4+ 20 Ax h4 gxh4 21 gh1 t) 20 gh1 . I didn't dare to accept the piece sacrifice with 20 gxh5: A) 20 . . . A h3+ !? is interesting :

73


A1 } 21 c;!? x h3 loses to 21 . . . gxh5+ 22 c;!? g2 �h8 23 A h4 tD x h4+ 24 c;!? f2

I n my game with Christopher Lutz (M u n ich 1 992) I rei ntroduced at high level the move 13 . . . c6, maintaining the tension in the centre and i ntending . . . b7-b5 . Probably because of the successfu l outcome of that game for Black, I had to face it with Wh ite (success fully as wel l !) the very next day! In forthcom ing years I had to uphold my faith in Wh ite's position i n games with Topalov (VSB, Ams terdam 1 996) and N ij boer (Hoogovens, Wij k aan Zee 1 998).

14 94 24 . . . tD g2 !! (keeping the king encaged) 25 �a4 (25 gh1 gh2) 25 . . . gh2 26 �d7+ c;!?f8 ; A2} 2 1 c;!? g1 ! gxh5 22 gf2 � h8 23 Af1 Ad7 with an attack; B} 20 . . . gxh5 and now: B 1 } 2 1 tD f2 tD h4+ ! 22 c;!? g1 � h8 23 A c4 (23 tDg4 tDf5 !) 23 . . . tDg2 !! (we are already fa miliar with this typical combination) 24 c;!? xg2 gh2+ 25 c;!? g1 � h4 26 tD g4 gh1 + 27 c;!? g2 �h3+ 28 c;!? f2 gh2+ 29 tD x h2 �x h2#; B2} 21 gh1 gxh1 22 c;!? x h1 � h8+ 23 c;!? g2 �h5 00 with a strong attack. 20 . . . tDg3 21 Axg3 fxg3 (21 . . . gxh1 22 �x h1 fxg3 23 � h3 (23 c;!? xg3 tD f4 00) 23 . . . �f6 ! ?, i ntending . . . Ad7 an d . . . g h 8 ; i f 23 . . . tD f4+ ?! 24 tD xf4 exf4 25 e5 !? with counterplay} 22 �d2 gh4 23 c;!? xg3 Ad7 24 gxh4 g x h4+ 25 c;!? h2 and Black has full compensation for the pawn, but not more (Gelfand - Kasparov, Reggio Emilia 1 991 /1 992) . 1 3 Hc1 c5 a

b

c

d

On the basis of the games Ftacnik- Geller, (Sochi 1 977) and Ftacnik- N u n n (Vienna 1986, see below), this position was assessed as favourable for White. He is trying to keep the black pieces and pawns as far away as possible from his monarch. As the fi rst world champion Wilhelm Stein itz wrote : ' If the nec essary precautionary measures are taken, the king can defend itself' .

1 4 . . . a6 Intend ing . . . b7-b5 . Kasparov's concrete play enables Black to avoid a bind on the kingside and breathes new life i nto the development of this variation . If 14 . . . h6 1 5 h4 fxg4 1 6 fxg4 tD eg8 (16 . . . g5 17 h 5 !) 1 7 c;!? g2 tD h7 1 8 gh1 Af6 1 9 Ae1 and Wh ite succeeds i n halting the opponent's kingside attack. White's strategic ideas are well illustrated by the following game: 14 . . . tDeg8 15 c;!?g2 ! tDe8 (15 . . . f4 1 6 h4 tD xg4!? 1 7 fxg4 �xh4 1 8 Ae1 �g5 1 9 tD f2 h5 (19 . . . tD h6 20 tD h3 �f6 ? 2 1 g 5 ! A x h3+ 22 c;!? x h 3 �xg5 23 A h4 +-) 20 tD h3 �f6 2 1 g x h 5 g5 22 tD f2 +-} 1 6 g5 f4 17 h4 gf7 18 gh1 Af8 19 �g1 (the pawns on d5 and g5 take too m uch space from the black pieces ; the reader has already seen this in Gelfand - Ivanchuk, Game 1 3 , p . 67) 1 9 . . . tD g7 20 A d1 ! tD h5 2 1 tD e2 h6?! (21 . . . Ag7 22 A a4 tD e7 23 b4 b6 ;t Nunn) 22 c;!? f1 Ae7 23 A a4 h xg5?! 24 h x g 5 A xg5 25 Ae8 !! with advantage (Ftacn i k - N un n , Vienna 1 986) .

e

8

_ _ _ _ _ _ ...._ .;;... ----' L..._ .-

lf 74

Game 15

Gelfand - Kasparov, Linares 1990

15 �f2

20 . . . hxg5 21 h5 tlfe8

How else can the e4 pawn be protected ? 15 a4 a5 = offers no chances, while after 15 a3 b5 !? 16 g5 ttJ h5 17 cxb5 axb5 18 ttJ xb5 A a6 � Wh ite has made a big concession (he has released the tension in the centre pre maturely b y g4-g5) , and s o Black has fu ll com pensation for the pawn.

8 7 6 5 4

4 3

15 . . . h6!? 15 . . . �d7 16 a3 ttJ eg8 17 b4 ;t .

��L� 2

16 h4 16 b4 ? cxb4 17 ttJ a4 b5 and if 18 cxb5 ax b5 19 �xb5 �a5 -+ . 1 6 g b1 fxg4 1 7 fxg4 g5 18 b4 b6 1 9 bxc5 bxc5 would al low Black to carry out the manoeuvre . . . ttJ e7-g6 -f4.

16 . . . fxg4 17 fxg4 �eg8 18 �g2 White has to protect his h4 pawn .

18 . . . �h7 19 llh1 £f6 From the fifteenth move onwards everything has been forced . Now the result of the inclu sion of 14 . . . a6 15 ttJf2 is evident : the pawn cannot be protected by �d2-e1 .

�------�--� �

21 . . . tD e7 is also possible, when Wh ite has a choice between 22 �g4 with play on the light sq uares, followed by �e2 and the dou bling of rooks on the h-file, and 22 �e3 !? (or 22 gb1 !?), consolidating the position and intending the standard King's In dian plan of gb1 , b2-b4 etc. In the post-mortem the world champion suggested an original possi bil ity of cou nterplay: 22 . . . ttJf5 !? 23 exf5 gxf5 with chances for both sides.

20 g5!

22 b4!?

20 h5?! �h4 + . It is interesting that both dur ing the game, and in later annotations (Kas parov in Chess Informator and myself in New in Chess magazine) 20 � g3 was condemned because of 20 . . . �xh4+ 21 gxh4 gxf2 : A) 22 �x h6? ttJ x h 6 (22 . . . g 5? 23 gh5) 23 gx h6 �f8 24 �d2 A xg4 25 gch1 A h5 and Black wins; 8) possible is 22 �h1 �f8 ! (22 . . . gf7 23 �xh6 � xh6 24 gxh6 and if 24 . . . �g5 25 gf1 ) 23 g5 (23 �e3 gf4 ! 24 A xf4 exf4+ 25 � f2 g5 26 gh3 tD hf6 27 gg1 ttJ d7 =+= and Black has more than sufficient compensation , accord ing to Kasparov) 23 . . . h5 24 �g1 gf7 and the extremely poor placing of Black's knights make the chances unclear;

B) 22 . . . ttJ e7 23 �x e8 gxe8 24 tD a4 and the queen exchange favours White, as now Black has no counterplay;

C) 22 �f3 !? This move, winning the ex change, was overlooked . However, Black can gain good com pensation by 22 . . . � g7 23 gh1 gxd2 ! 24 �xd2 ttJ g5.

C) the natural 22 . . . Ad7? loses to 23 h x g 6 ! ttJ h6!? (23 . . . �xa4 24 gx h7#) 24 gxh6 �xa4 25 �g4 !! - the simplest, after which Black is helpless against gx h7+ and �e6+ ;

This move indicates my desire to play very concretely and energetically. I now th ink that positional play was more to the point. Kasparov's recommendation 22 hxg6 �xg6 23 �h5 �g7 24 �g4 gives Wh ite good play on the light sq uares, which compensates well for the missi ng pawn. 22 �a4 is tem pting : A) hard ly sufficient is 22 . . . b5 23 c x b5 Ad7 24 �d1 !? a x b5 25 A x b5 , as 25 . . . A x b5 26 ttJ x b5 �x b5 27 h xg6 ga7 28 �h5 is bad for Black;

75


0) 22 . . . �f7 23 gcf1 (23 b4 !?), bringing the rook into the game.

24 . . . Ad7 24 . . . b5 25 Axd6!? gxf2+ (25 . . . gf6 26 hxg6) 26 � xf2 bxa4 27 h xg6 �xg6 28 A x e5+ Af6 29 �d4 looks to be in Wh ite's favou r. 24 . . . gxf2+ !? 25 � xf2 �gf6 is interesting here as wel l .

22 Ag4 Axg4 23 �xg4 �h6 24 �e2 is simi lar to 22 h xg6.

22 . . . cxb4 Obviously, Garry also fights for the in itiative. 22 . . . b6 23 gb1 is a concession .

a

b

c

d

e

23 �a4 Now 23 �a4 ?! fails to achieve its goal : 23 . . . bxc3 24 �x e8 gxe8 25 h xg6 �e7 ! 26 gx h7+ � g8 27 gxc3 �xg6 =t .

5 r--__ 4 3

23 . . . Ad8! From here the bishop not only takes con trol of the b6 sq uare (from where it can also come into play later), but it also clears the f fi le for the rook and the f6 square for a knight. Bad was 23 . . . b5 24 �b6 gb8 25 �xc8 gxc8 26 Ag4 gc7 27 hxg6 �xg6 28 Af5 , or 23 . . . a5 24 � b6 gb8 25 �xc8 gxc8 26 Ag4.

L...-_______----=-__.....

{r

25 hxg6?! In this extremely com plicated position both players prefer not to take half measures, but crucial decisions. However, from g6 the q ueen takes a d i rect part in the attack (the e4 pawn is weak!).

24 Axb4 One of the favou rite questions I hear from chess amateurs is: ' How many moves ahead can you calculate ? ' I try to explain that sometimes you can not foresee more than 2-3 moves, as both you and your opponent have too wide a choice of possibil ities. In this case it is far more important to see as many of the avai lable options as possible and to trust your intuition . This is one such position and even after a lengthy home analysis it is impossible to make a correct j udgement.

25 A xd6?! leaves Wh ite a piece down after 25 . . . Axa4 26 hxg6 't'#xg6 (26 . . . gxf2+ !? 27 � xf2 �gf6 28 't'#d3 't'#xg6 29 Axe5 � g8) 27 Axe5+ Af6 =t . 25 � c3 seemed rather passive t o m e : 25 . . . Ae7 (25 . . . A b6 2 6 �d3) 2 6 �d3 g4!? 25 �b2 , with the idea of playing the knight to d3, seems obvious, but 25 . . . gxf2+ 26 � xf2 �gf6 00 allows Black to take the in itiative, due to the weakness of the e4 pawn : 27 �d3 A b6+ ! (it is i m portant to bring the bishop into play) 28 � e1 g x h5 29 A xd6 Ad4 with counterplay.

An alternative was the 24 c5 breakthroug h , when i t i s hard to say what is the best reply: A) 24 . . . � gf6 25 A x b4 (25 c6 !? ; 25 cxd6 g4!?) 25 . . . a5 26 A a3 b5 27 cxb6 Ad7; B) 24 . . . Ad7 25 � b6 A xb6 26 cxb6 00 ; C) 24 . . . dxc5 25 �xc5 ;

25

0) 24 . . . �f7 !? ; E) 24 . . . gxf2+ !? 25 � xf2 �gf6 26 cxd6 (26 gc4 d x c5 27 � xc5 b6 28 �d3 b5 with counterplay) 26 . . . �xe4+ 27 �g2 Ad7 with a crazy situation .

4 3 2

. . .

•xg6 26 c5

Intending c5-c6. Now Black has no less than five (!) tempting ways to develop his attack.

26 . . . g4?! The alternatives were :

76

Game 1 5


A) 26 . . . gxf2+ 27 � xf2 �x e4 28 gc4 �f5+ 29 � g2 �gf6 30 � c3 and Wh ite consoli dates; 8) 26 . . . gc8 !? 27 Ag4 (27 � b6 Axb6 28 cxb6 g4 ! with the initiative) 27 . . . Axg4 (27 . . . gc7 !?) 28 Wfxg4 gc7 29 cxd6 gcf7 , and if 30 �d3 ? (30 Ac5 gf4 31 �e2 � gf6 with an attack) 30 . . . gf4 ;

27 g3 27 . . . Ag5 28 gc3 . 28 �d3 bxc6 The computer refutes 28 . . . gc8 - 29 � ac5 ! d xc5 30 cxd7 gc7 31 Ac3. Dubious is 28 . . . b5 ?! 29 cxd7 b x a4 30 gc8, while 28 . . . Ag5 29 gc3 is unclear. 29 dxc6 . . .

C) 26 . . . Axa4!? 27 �xa4 g4 and now :

a

8 .1. 7

C1 ) less concrete is 28 Ac3 dxc5 (28 . . . b5 !? ; 28 . . . Ag5!? ; 28 . . . � gf6 !? 29 cxd6 A b6) 29 A xe5+ (29 Wfd7 � h6 (29 . . . gxf2+ 30 �xf2 'i' xe4 31 �x g4) 30 A x e5+ Af6) 29 . . . Af6 30 Ag3 Ad4 and it is m uch easier to play this position from Black's side ; C2) 28 Wfd7 !? gxf2+ !? (28 . . . �gf6 29 'tWxd6 ; 28 . . . g3 29 �x h7+ �x h7 30 gx h7+ � x h7 31 �d3 with sufficient compensation for the exchange) 29 � xf2 �gf6 30 Wfxb7 (30 gx h7+ tb xh7 31 �xg4 (31 �x b7 �xe4) 31 . . . � h6) 30 . . . Wfxe4 ! 31 �xa8 (31 Ad2 g3+ 32 � xg3 'i'xe2) 31 . . . g3+ 32 � xg3 Wfe3+ with perpet ual check; 0) 26 . . . �gf6 ! (analysed by Kasparov i n In formator) 27 c6 (27 �c3 g4 28 c6 g3 29 �d3 tbg4!) 27 . . . � xe4, and now: 0 1 ) White loses after 28 cxd7 (or 28 � x e4 'i'xe4+ 29 Af3 �x b4 30 cxd7 g4 -+) 28 . . . gxf2+ (28 . . . � xf2 ? 29 �c2 �x c2 30 gxc2 � xh1 31 A xd6 oo) 29 � g1 �f5 30 Ae1 (30 gc3 gg2+ 31 � xg2 �f2+ 32 c;!? h3 g4+ 33 � xg4 �ef6+ 34 � h3 �g5#) 30 . . . Wff4 ! 31 �d3 (31 gh3 g4 -+) 31 . . . �xc1 32 gx h7+ � g8 -+ (Kasparov) ;

02) 28 � d3 ! (suggested by the com puter, this is the only move that al lows Wh ite to continue the fig ht) 28 . . . bxc6 29 d x c6 �f5 30 Ae1 (30 cxd7 �f2) 30 . . . Ae6 31 c7 A e7 32 �b6 and I wou ldn't even dare to try and assess this position .

27 c6 Only forward ! 27 �c3 �gf6 transposes into the 26 . . . �gf6 variation .

b

c

d

e

...

f 8 7 6 5 4 3 2

5 4 3

f

9

h

At this moment I estimated my chances very high ly, thinking that Black's play had come to a dead end , whereas Wh ite had broken through on the queenside. The world cham pion 's reply came like a cold shower to me. 29 gc8! 29 . . . Ag5 also came into consideration . Here I realised that I had to fight for a draw, but although I thought for a long time, I cou ld not find the right sol ution. . . .

30 Af3 Not 30 cxd7? gxc1 31 � x c1 gf2+ 32 � g1 �x e4 -+ , or 30 � x e5 gf2+ 31 � g1 �x e4 32 Af3 �e3 -+ , but as Kasparov showed after the game, 30 Axd6! was stronger: A) 30 . . . �x e4+ 31 Af3 gxf3 32 �xf3 A xc6 33 A x e5+ (33 � x e5 ? �e2+ -+ ; 33 gxc6 �xc6 34 � x e5 �xf3+ 35 � xf3 Af6 =) 33 . . . Af6 34 A xf6+ �g xf6 35 gxc6 gxc6 36 �ac5 with a draw ; B) Black can avoid the draw by 30 . . . �xd6 31 cxd7 gxc1 32 � xc1 ! (I didn't see this move ; 32 Wfxc1 ?! Wfd4!) 32 . . . Wfg6 (32 . . . Wfc6 33 Af3 oo ; 32 . . . gf2+ 33 � xg3 Wff6 34 Af3 +-)

77


33 Af3 tLl h6 with an unclear game, but his chances are not better here.

30 . . J lxf3!? The post-mortem analysis showed that 30 . . . A xc6?! 31 tLl xe5 ! (31 l'!xc6 l'!xc6 32 tLl xe5 d x e5 33 A xf8 tWf7 ! +) 31 . . . d x e5 32 A xf8 A xa4 33 tWd6 (33 tWxa4 ? l'!xc1 34 l'!xc1 tLl xf8) 33 . . . tWxd6 34 A xd6 l'!xc1 35 A xe5+ Af6 36 A xf6+ tLl g xf6 37 l'!xc1 leads to a draw.

B3) 34 . . . A xc1 35 A x e5+ tLl gf6 36 tLl xc1 tWxc6 37 tLl c3 l'!g8 38 l'!h6 Af5 39 A xf6+ tWxf6 40 tWxg8+ � xg8 41 l'!xf6 tLl xf6 with a total elimi nation of the fighting material . 33 . . . Af3 !? was one of the most dangerous attem pts : A) 34 l'!h3 tWxd3 35 tWxf3 tWd4+ 36 � g2 tWx b4 (36 . . . l'!c7 ? 37 a3 ! a5 38 l'!d1 or 37 . . . l'!g7+ 38 � h1 ) 37 l'!x h7+ � x h7 38 tWf7+ � h8 39 l'!h1 + A h4 40 tW h5+ � g7 41 l'!x h4 tWd2+ and White is not guaranteed a draw ; B) 34 A xd6 (here this 'standard ' move is not sufficient) 34 . . . tWe3+ 35 � h2 tWxd3 36 A xe5+ Af6 ; C) 34 tLlf2 !? tWf5 35 Axd6 (after 35 l'!h3 gxc6 36 ge1 Ad5 I prefer Black) 35 . . . Ac7 ! (intend ing to bring the rook across along the 7 th ran k ; now Wh ite has to look for an escape) 36 gx h7+ !? (36 gh3 ? A xd6 37 tWxf3 tWg5+ -+ ; 36 A xc7 gxc7) 36 . . . � x h7 37 gc3 and this looks good enough for Wh ite to draw.

31 Oxf3 £g4 32 Oxg3 32 tWe3 ? Ag5 -+.

32 . . . 0xe4+ 33 ct>g1 In Chess Informator my opponent suggested that 33 � h2 was stronger, but it loses to 33 . . . Af3 ! (33 . . . tLl gf6 34 A xd6 tWxa4 35 c7) 34 tLlf2 (34 l'!hf1 tLl g5) 34 . . . tWx b4 35 tWxf3 tWxa4 36 � g2 l'!c7 when Wh ite has no coun terplay for his material deficit. a

b

8 7 5 3

..

c

to Wh ite's poorly placed pieces) 36 . . . tWd2 37 l'!xg4 oo ; B) 34 Axd6. As i n most li nes, it is advisable to take on d6; furthermore, here Black can not win material : B 1 ) 34 . . . tWx a4 and now the strongest is the com puter suggestion 35 A x e5+ tLl gf6 36 tLlf2 ! A xc1 37 tWd3 ! (37 A xf6+ � g8 38 tLl xg4 also wins easily) ; B2) 34 . . . tWd4+ 35 tLlf2 ;

8 7 6 5 4 3 2

Wh ite also has to be precise after 33 . . . gc7 !? 34 A xd6 (there is no other reasonable way of cou nteri ng . . . gc7-g7), and now: A) 34 . . . gg7 35 A xe5 Af6 (35 . . . tLlgf6 proves u nsuccessfu l after 36 c7 A x c7 37 gxc7 Ad7 38 tWxg7+ ! � x g7 39 gxd7+) 36 tLl ac5 (or 36 l'!h4 !? A x h4 37 tWx h4 ±) 36 . . . tWxc6 37 � f2 and Wh ite parries the i m mediate threats, remai ning the exchange u p; B) 34 . . . tWd4+ ! 35 � h2 (35 � g2 !? tWd5+ 36 � g1 tWxd6 37 tWx e5+) 35 . . . tWxd6 36 tWx e5+ tWx e5+ 37 tLl x e5 and the d raw is with in reach.

33 . . . �gf6!? Pointless was 33 . . . Ae7 ? 34 tLlf2 tWx b4 35 tWxg4 or 33 . . . Ac7 ? 34 l'!h4 tLlgf6 35 l'!f1 , but Black had several other promising con tin uations at his disposal . With 33 . . . Ag5 Black can hard ly pose any serious problems: A) 34 l'!e1 tWd4+ 35 tLlf2 tWx b4 (35 . . . tLl gf6 36 tWc3) 36 l'!e4 (36 tWxg4 tWxg4+ (36 . . . tWx e1 + 37 � g2 tWa5 38 tWxc8 tWxa4 39 c7) 37 tLlxg4 l'!xc6 and Black is clearly better due

78

Game 1 5

a b c d


e

8 7

7 6 5 4 3 2 a b c d

and now 38 . . . �xc6 ? is a mistake due to 39 �x h7+ �x h7 40 �h2+ tbh5 41 �x h5+, when if 41 . . . �g7 ?! 42 tbf5+ +-.

37 . . . •xa4 38 .xeS

8 7 5 4 3

e

A �______________�____� u

34

..

a b

.txd6!

8 7 6 5 4 3 2

At this moment I felt that 34 . . . Ac7 and 35 . . E!.g8 had to be prevented at any price. .

34 . . .•d4+

After 34 . . . �xc6 ? ! 35 �xc6 �xc6 36 Axe5 ! (36 E!.x h7+ �x h7 37 �h4+ Ah5 38 Axe5 =) 36 . . . �xa4 37 tbf2 �g8 38 �h4 it is Black who has to make a d raw from the 'weaker side' : 38 . . . �xa2 39 tbxg4 Ab6+. After 34 . . . �xa4 35 Axe5 Ab6+ 36 tbf2 E!.xc6 (36 . . . �g8 37 �h6 ! �xc6 (Black is in trou ble after 37 . . . �f7 38 �h4 !) 38 �xc6 �xc6 39 E!.xf6 tbxf6 40 Axf6 �xf6 41 �xg4+ =) 37 E!.xc6 �xc6 38 Axf6+ �xf6 39 �b8+ �g7 40 'tWb7+ �f8 41 �b8+ Wh ite gives perpetual check.

35 �f2 .xd6 36 �xg4 It would appear that 36 �x h7+ tbx h7 (36 . . . �x h7 37 tbxg4 �d4+ 38 tbf2 , and if 38 . . . �xa4 ? (38 . . . tbe4 39 :tl¥h3+ �g6 40 �e6+ =) 39 �h3+) 37 �xg4 �xc6 (37 . . . E!.c7 38 �d1 ) 38 �xc6 �xc6 39 �e4 ! + was sufficient for a d raw. Perhaps this was a more practical way of playi ng, as we will see.

36 . . .•d4+ 37 �f2 It is so natural to place the pieces as safely as possible, but the sil icon monster sug gests 37 tbe3 �xa4 (better is the concrete 37 . . . Ab6 ! 38 tbx b6 �g8 39 �xg8+ �xg8 40 tbbc4 �c5, el i m i nating the c6 pawn and thus any winning attem pts) 38 �xe5,

38 . . J �c7?! After this series of forced moves I ex pected 38 . . . �xc6, when I didn't see a d raw after 39 �d1 ? Ab6 ! . Here is a possible l i ne : 40 �b8+ �g7 41 �g3+ �f7 42 �d7+ (42 �x h7+ tbx h7 43 �d7+ �e8) 42 . . . �e8 43 � h x h7 (43 �d x h7 Axf2+ 44 �xf2 �g4+ 45 �h2 tbx h7) 43 . . . tbxh7 44 �x h7 Axf2+ ! 45 �xf2 �d1 + ! (45 . . . �c1 + 46 �h2 �c2 47 �h8+ and the king cannot escape from the checks) 46 �g2 �c2 -+ 47 �h8+ �d7 48 �h7+ �c6 49 �h6+ �b5 50 �h5+ �b4 51 �h4+ �a3. But Garry saw more : 39 �xh7+ ! �xh7 40 �h2+ tbh5 ! (even after the rook blun der 40 . . . �g7 ? 41 �g2+ �f7 42 �xc6 �xa2 it is u n l i kely that White would achieve more than an ending with E!.+tb v. A+tb) 41 �x h5+ �g7 with a draw. But in any case the game continuation doesn 't pose any problems for Wh ite, so I am surprised that the world cham pion didn't try such an excellent prac tical chance.

39 J::l h 2 I decided to secure the position of my king. After 39 �d1 !? �g7+ 40 �f1 �c4+! 41 �e2 �g8 oo it is Black who is more l i kely to de velop an attack.

79


42 .c8+

39 . . . 11g7+ 40 11g2

Forcing a draw.

Hardly anyone would venture 40 �h1 on the last move before the time control , but apparently this does not lose : 40 . . . Ac? (40 . . . ffxa2 !?) 41 ffxf6 Ax h2 42 ffd8+ !'!g8 43 c? Axc? 44 ffxc?

42 •g8 43 .xg8+ �xg8 44 Ilxg7+ �xg7 45 �d3 Draw agreed . . .

Th is fascinating game gave us a good boost, as we both won our next four games ! I can not remember a simi lar occurrence in a tour nament of such cali bre! It was also awarded a prize for the best game of the tournament an extremely heavy statue. We had to toss a coin to decide who would take it. And it was a rare case when I was happy to lose, as I had no idea how I would be able to transport it back home!

40 . . . .tc7 41 .f5 41 ffe6? fff4 -+.

41

. . .

•xa2

a b c

d

e

8 l---r-__�7 6 5 4 3 2

5 4 3 2

�------�--�

Only in the last m i n utes of the tournament did Kasparov manage to overtake me, but it was still a great success for me. My score of +4 in Linares (I won six games) lifted me from a promising j u nior to one of the strongest players i n the worl d . I n the next rating list I was eq ual 3 rd , behind only Kasparov and Karpov.

�

*** Game 1 6

9 . . . g5 !? 10 Ag3 eDh5 was the most natural plan.

B o r i s G e l fa n d - M u r ra y C h a n d l e r G MA Qualifying Tournament, Moscow 1 990 Queen 's Indian De fence [E 1 2]

10 d5

a b c

1 d4 �f6 2 c4 e6 3 �f3 b6 4 a3 .tb7 5 �c3 g6

7 6 5 4 3 2

The double fianchetto against the Petrosian Variation was often employed by players from M i nsk, so I was fami liar with the ideas and I twice played it with Black agai nst Jeroen Pi ket in 1 988. It is a kind of hy brid of the King's Ind ian and Queen 's Ind ian defences. Black concedes the centre, but keeps pressing on it.

6 .tg5

d

...

e

8 7 6 5 4 3 2

10 . . .•e7 More in the spirit of the position was 10 . . . g5 11 Ag3 eDh5, transposing i nto one of my

6 d5 and 6 ffc2 are other common plans.

6 . . . .tg7 7 .c2 h6 8 .th4 d6 9 11d1 0-0

80

Game 16

games with Piket : 1 2 �d4 (1 2 e3 !?) 12 . . . VWe7 1 3 e3 �xg3 14 h xg3 cS ! with counterplay (Piket- Gelfand, OHRA, Amsterdam 1 988).

11 e3 c6? This is a mistake, breaking up Black's pawn structure. If Chandler didn't like the plan with . . . g6-gS and . . . �hS, he should have played 11 . . . cS !? 1 2 h 3 ! ? (1 2 Ae2 exdS 13 cxdS gS 14 Ag3 �hS ; 12 d xe6 VWxe6 1 3 �bS ttle8 14 Ag3 dS!? with an unclear game) 12 . . . exdS 13 cxdS with a kind of Modern Benoni .

1 2 dxe6 exe6 1 3 Ae2 �bd7 After 13 . . . dS 14 ttld4 !? (14 cxdS ttlxdS 15 �xdS cxdS 1 6 ttld4 �e4 =) 14 . . . �c8 15 cxdS �xdS 16 �xdS cxdS Black's coun terplay doesn 't compensate ful ly for his iso lated pawn, but I think this was a better so lution .

14 0-0 �e5

a b c d

ttJ

Gelfand - Chandler, GMA Qualifying Tournament, Moscow 1 990

16 �d5?! This allows the penetration of White's knight to d6. Better was 1 6 . . . gfd8 1 7 �a4 gxd1 (17 . . . gS 1 8 gxd8+ gxd8 1 9 �xa7 !? or 1 9 Ag3 ;!;) 18 gxd1 gS 1 9 Ag3 ;!; . 17 �e4 Not 1 7 Ac4 ? �xe3 (17 . . . VWg4 1 8 �e4) . 1 7 e4 ttlxc3 1 8 VWxc3 gS 1 9 gd6 ;!; was also strong , but I decided to keep the knights on the board , since White's is heading for d6, whereas Black's on dS is vulnerable to attack (by a possible e3-e4) . . • •

17

• . •

g5 18 Ag3 gab8

18 . . .fS 1 9 �d6.

19 �d6 bxe5 1 9 . . .fS 20 e4 ! ± . 20 �f5 20 VWxcS Af6 (20 . . . Aa8 21 e4 ±) would allow Black some counterplay. 20

. . .

e

Ae8

a b c d 8

7

7

6

6

5

5

4

4

3

3

2

2

L...-_______---=-__.....

'If

15 �xe5! I think that (together with the next move) this is a strong strategic decision . 1 S ttld4 VWe8 16 e4 was more natural , but after 1 6 . . . a6 ;!; Black has counterplay associated with . . . c 6 c5 and . . . b6-bS.

15 . . . dxe5 16 c5! This is an instructive case of transforming an advantage. Black has got rid of his weak d6 pawn, but both of h i s bishops are shut out of the game.

e

f

h

8

8

7

7

6

6

5

5

4

4

3

3

2

2

L..-_______...._ .:::.... ----J

'lf

21 �xg7 ! A concrete decision. Not so clear was 21 e4 �f4 22 Ac4 (22 gd6 VWb3 !) 22 . . . VWe8 ;!; . 2 1 . . . �xg7 2 2 exe5 22 Ac4!? ± . 2 2 . . . gxb2 23 Ae4 23 Ag4 !? was suggested by Chandler after the game. I missed this possi b ility, as I saw that the text move was strong enough . How ever 23 Ag4 is not as decisive as it seems: 23 . . .fS (23 . . . VWf6 24 Axc8 gbS 2S gxdS

81


.!3.xc5 26 .!3.xc5 +-; 23 . . . �xg4 24 Axe5+) 24 e4 ! fxg4 25 exd5 cxd5 26 .!3.xd5 .!3.e2 and Wh ite still has to work hard to win the game.

White had a wide choice of options : 30 .!3.b1 !? or 30 Ad6.

30

23 . . . Dc2

• • •

Aa6 31 Aa4

31 �xc6 ?? .!3.c8.

23 . . . .!3. b7 24 e4 �e7 25 �xe7 (25 Axe5+ �xe5 26 exd5 cxd5 27 .!3.fe1 �f4 ;;\;) 25 . . . �xe7 26 Axe5+ ± .

31 . . . Ab7 32 h3! 016

24 Dc1 ! E lim i nating all counterplay. 24 e4 ? Aa6 (24 . . . �f4 25 �xf8+) 25 exd5 cxd5 26 .!3.fe1 .!3.xc4 27 Axe5+ f6 28 �xa7+ .!3.f7 leads to a draw.

8 7 6

7 6 5 4 3 2

24 . . . Dxc1 25 Dxc1 e4 Or 25 . . .f5 26 e4 fxe4 27 �xa7+ .!3.f7 ? 28 �b8.

2

26 0xa7

a b c d 8 7 6 5 4 3 2

..

e

a b c d

8 7 6 5 4 3 2

e

f

9

h

�

33 �h2! Avoiding the last trap : 33 .!3.b1 �c3 34 .!3.x b7? �e2+ 35 �h2 �a1 -+. 33 . . . h5

33 . . . �d8 34 �d4 +-. 34

Db1 h4

34 . . . �c3 35 .!3.x b7 �xa4 36 �b4.

35 Dxb7 hxg3+ 36 fxg3 Of2 37 Db3

As Wh ite has the advantage of the two bish ops and a passed a-pawn, the rest is just a matter of technique.

37 Axc6 ? �xe3 38 �xg5+ �h8 39 �h5+ �g8 40 �xf7+ 'tWxf7 41 .!3.xf7 .!3.xa5 and White has still to demonstrate some techn ique.

26 . . . �g8 27 Oc5 .td7 28 a4 Da8 29 a5 Ac8 30 Ab3

37 . . . De8 38 Axc6 Black resigns

* * *

82

Game 1 7

Short - Gelfand, I nterpolis Tournament, Tilburg 1 990

Game 1 7

N i g e l S h o r t - B o r i s G e l fa n d I nterpo l i s To u r n a m e n t , Ti l b u rg 1 9 9 0 Sicilian Defence [892J The Interpolis tournaments were held i n the south of The Netherlands for 15 years. They were some of the best tournaments of their time, not only as regards the strength of the players, but also the level of organisa tion . The players had noth ing to worry about, apart from the quality of their games. In this, my first Ti lburg event, I had lost my first two games, and I was hoping to get back i nto the tournament. N igel Short has been one of the most awk ward opponents for me throughout my ca reer. The Engl ishman is especially strong with the wh ite pieces. I n our games N igel has tried almost all the possible moves against the Najdorf Variation . In the first two games I managed to w in, but later on he gained re venge many times. Th is was our first game.

1 e4 e5 2 �f3 d6 3 d4 exd4 4 �xd4 �f6 5 .tie3 a6 6 f4 �bd7 7 £e2 e5

a b c d

e

(Kasparov - Gelfand , Cred it Suisse Masters, Horgen 1 994) .

8

..•

�e5 9 �g3 1tb6

Black tries to prevent Wh ite from castling and to gain some tempi for the development of his pieces.

1 0 Db1 .ld7 1 1 fxe5 dxe5 1 2 .le3 1tc6 1 3 0-0

a b c d

e

6 5 4

5 4 3

d L...-_a _b c _ _

__

e__....: 9_ ::: h_...J _

11

N igel leaves a pawn en prise. This was also the main l i ne in open ing books at that time, but i n 1 990 I was not yet collecting chess books, so I had no idea about this. 13 Af3 was safer.

1 3 . . . �exe4 N 1 3 . . . 0-0-0 14 b4 tbcx e4 1 5 tbcxe4 tbxe4 1 6 tbxe4 �xe4 1 7 Ab6 ge8 1 8 Af3 'tlfg6 1 9 Ax b7+ �b8 20 Ae3 +- (F. Olafsson Sax, Novi Sad 1 976) was the main reference game.

6 5 4 3 2 L...-_______....::..-. ...; _ . ---I

14 �exe4 �xe4 15 �xe4

11

8 �f5 8 fxe5 later became fash ionable and was played against me three times. The most crit ical game for opening theory was the latest one: 8 . . . d xe5 ! 9 tbf5 'tlfb6 10 tbd5 ?! tbxd5 11 �xd5 �c5 ! N 12 �b3 tbf6 13 Ac4 �b4+ 14 �x b4 Ax b4+ 15 c3 Af8 ! 16 Ad3 Ae6 =

1 5 Af3 is the most chal lenging, but as the fol lowing game shows, Black has noth ing to fear: 1 5 . . . Ac5 1 6 tbxe4 Axe3+ 1 7 �h1 0-0 1 8 tbc3 �e6 19 tbd5 e4 20 Ah5 Ab5 21 tbxe3 Axf1 22 �xf1 �h6 23 �e2 f5 + (Mack- Howell, London 1 991 ) .

15 . . . 1txe4 1 6 1td2 .le6 17 Ah5 Wh ite has to act very energetical ly if he wants to demonstrate that he has compen sation for the pawn .

17 . . . g6

83


a b c d

e

Th is allows the exchange of queens. Bet ter was 24 �d3 ! �f7 (24 . . . b4 2S Ag7 gg8 26 Axf6) 2S �x bS b6 26 a3 (26 a4 gaS) with a balanced position.

f

24 . . . nc8

5 4 3

24 . . . �f7 ? 2S gxf6+ ! ± .

25 .xb7 .c6 + 26 .xc6+ nxc6

a b c d a b c d

8 �-I'-7 6 5 4 3

e

A �______________�__� v

18 A13 I th i n k that 1 8 g be1 !? was stronger, but it is sufficient only for a draw: 18 . . . Ag7 (danger ous is 1 8 . . . g x hS?! 1 9 Ad4 �xg2+ 20 �xg2 Axg2 21 gxeS+, or 18 . . . �b4 19 �f2 with an attack; 18 . . . gd8 1 9 �f2 �xg2+ 20 �xg2 Axg2 21 �xg2 g x hS 22 AgS with sufficient compensation) 19 Ad4 �dS (19 . . . �xg2+ 20 �xg2 Axg2 21 �xg2 g x hS 22 AxeS =) 20 Af3 (20 AxeS �xd2 21 Axg7+ �d7 22 gd1 �xd1 23 gxd 1 + �c7 24 AeS+ =) 20 . . . �d7 21 AxeS �xd2 22 Axg7+ �d7 23 gd1 �xd1 24 gxd 1+ �c7 =.

18

. . .

L...-______________---=-____.....

Normally one extra pawn in the centre is sufficient - I won a n ice game against l Iya Smirin in Riga 1 987 (see on the facing page).

•c4

27 h4

19 nbd1 Ae7 20 Ae2!?

27 c3 gS 28 h4 gg8 and the bishop is trapped .

20 Axc6+ �xc6 21 Ah6 gd8 =t . . . .

1i

Black's advantage is obvious, as he has two extra pawns i n the centre and his b -pawn stops all three pawns on the queenside.

Protecting the f7 pawn. If 1 8 . . . �b4 1 9 �f2 with an attack.

20

e

•e6 2 1 .c3 Ab5

27

I decided to get rid of the im portant bishop, even though I had a pleasant choice : 21 . . . 0-0 !? 22 Ac4 �g4 23 AdS ! (23 gd2 gad8 24 gdf2 Ah4 (24 . . . �e4 2S Ah6 AcS 26 Axf7+ gxf7 27 �xcS ±) 2S ge2 AdS =t) 23 . . . gad8 24 h3 with some com pensation for the pawn.

. . .

tM1

27 . . . gg8!? =t .

28 nd5 Ac5+ 29 ct>h2 ct>e6 30 nd3 15 30 . . . Ae7 -+.

31 nb3 nb6 32 nd1 lia8! The rook is aim i n g for its best post on c4. 32 . . . b4 was also winning.

22 Axb5+ axb5 23 Ah6 Wh ite's only hope is to keep the black king in the centre.

33

a3 lia4 34 g3 lic4 35 lid2 lid6 36 c3 nxd2+ 37 Axd2 ne4 38 Ag5 ne2+ 39 ct>h1 Af2 40 ct>g2 nc2

23 16 24 .c7? . . .

84

Game 1 7

Short - Gelfand, I nterpolis Tournament, Tilburg 1 990

41 gxb5

a b c 8 7 6 5 4 3 2

�'-'

8

41 �f1 �d6 42 gxb5 Axg3 -+ was equally hopeless.

6 5 4

41 . . . h6 42 Axh6 Ae3+ 43 @f1 Axh6 44 h5 Ae3 45 h xg6 Hf2+ 46 �e1 Hg2 47 g7 Hxg3 48 �e2 f4 49 g b8 Hxg7 50 �d3 Hf7 51 · �e4 f3 52 �xe3 f2 53 gb6+ �d5 White resigns

L...-_______--=--_----I

'lf

I was very pleased when Yuri Razuvaev, the author of one of my favou rite books Ak iba Rubinstein , told me that this end ing re minded him of the games of the great Akiba. * * *

A Similar Ending (cf. note to Black's 26t h move) lIya Smirin - Boris Gelfand World Junior Qualifying Tournament, Riga 1 987

a b c d

e

8

Position after 20 I1fd1

20 . . . gad8 21 b3 �f7 22 tDa5 tDxe3 23 fxe3 Ac5 24 �f2 gxd1 25 gxd1 �e6 26 tDc4 f5 27 �e2 e4 28 ga1 gc8 29 gd1 g5 30 gh1 Ae7 31 g3 h5 32 �f2 gd8 33 �e2 gh8 34 �f2 a5 35 gd1 a4 36 �g2 axb3 37 cxb3 Ac5 38 ge1 gd8 39 ge2 gd3 40 ge1 gc3 41 ge2 �f6 42 �f1 �e6 43 ge1 h4 44 g x h4 g x h4 45 h3 f4 46 exf4 �f5 47 gd1 �xf4 48 gd5 Ae3 49 gd7 �g3 50 ge7 Af4 51 gxe4 gf3+ 52 �e2 gf2+ 53 �d3 �x h3 54 ge7 gxa2 55 gb7 �g4 56 tDe3+ �f3 57 tDd5 gd2+ 58 �c4 Ag5 59 gg7 gg2 60 tDx b4 h3 61 tDd3 h2 White resigns * * *

85

My Most Memorable Games G am e 1 8

B o r i s G e lfa n d - K i r i l G e o r g i e v O l y m p i ad , N ov i S a d 1 9 90 Queen 's Gambit [03 7J Chess Olympiads have always held some magic for me, as wel l as for the majority of players. It is a unique com bination of a high level event for professionals and a reunion for amateur players. I n N ovi Sad I was a mem ber of the winning team . I n fact, at that time it was m uch more d ifficult to qual ify for the Soviet team than to win the Olympiad with it. I remember how i n 1 988 Vasi ly Ivanchuk was selected for the first time t o play for the Soviet team . During the long fl ight back from the World Junior Championsh ip in Australia, I said I had dreamed that one day we would play together in this team . Vasi ly instantly replied : 'Sure, on the first two boards ! ' I took this as a joke, but just two years later Vasi ly's words came true. And as we had same rating (we were shari ng 3r d_4t h places in the world list) the team captain had to toss a coin to establish the board order.

White tries to combine castling with the de velopment of h is rook. I became very in terested in this sharp l i ne, stud ied it q uite deeply and played n u m ber of entertaining games with it. Mai n ly because of this, the team decided to g ive Ivanch uk a rest and allow me to play this variation on board one. 10 gd1 was the main line before q ueenside castling came i nto use. At the beg i n n ing of the 2 1 s t century 1 0 �d2 Ab4 1 1 cxd5 has become popular. 10 gd8 Not the most chal leng ing move. Black puts his own queen i n trouble. More often I had to face other moves. I. 10 A.d7 1 1 g4 (one of the main ideas beh ind 1 0 0-0-0 ; 1 1 cxd5 exd5 1 2 �xd 5 ? �xd5 1 3 gxd5 Axe3+ +) 1 1 . . . gfc8 12 'i!?b1 (12 g5 �h5 1 3 cxd5 �xf4 14 exf4 Axa3 ! ? , and if 15 bxa3 �d4 16 �xd4 gxc3). . . .

. • •

1 d4 �f6 2 c4 e6 3 �f3 d5 4 �c3 Ae7 The Queen 's Gambit was the main choice of the Bulgarian grandmaster for many years.

5 Af4 0-0 6 e3 c5 6 . . . �bd7 is a passive but solid alternative.

7 d x c5 Axc5 8 "c2 �c6 9 a3 "a5 10 0-0-0

a b c d

Th is position was the subject of a theoretical d iscussion between myself and Alexander Bel iavsky . 12 A.fB 1 3 g5 (13 �d2 �e7 !?) 1 3 . . . �h5 14 Ag3 (14 cxd5 �xf4 15 exf4 exd5 16 gxd5 VJ!Jc7 g ives Black ful l com pensation for the pawn) 14 . . . �e7 (Black has managed to hold the d5 sq uare, but his pieces are awkward ly placed ; 14 . . . b5 1 5 c x b5 �e7 1 6 VJ!Ja4 ±) 15 �e5 Ae8 1 6 Ae2 f6 (du b ious is 16 . . . g6 1 7 �g4 (17 Ax h 5 !? g x h5 1 8 VJ!Jd2) 1 7 . . . Ag7 18 Ae5, underlining the weakness of the dark squares) 17 gxf6 gxf6 (17 . . . �xf6 would leave all the strong pOints i n Wh ite's hands after

...

e

8 7 6 5 4 3 2

. . •

6 5 4 3 2

86

Game 1 8

Gelfand - Georgiev, Olympiad , Novi Sad 1 990

18 �b3 ±) 18 �f3 (18 ghg1 ? fxe5 1 9 Axe5+ tilg7).

18 . . . Ag6. Insufficient was 1 8 . . . b5 19 c x b5 �xg3 (19 . . . a6 20 Ad6 ! ±) 20 h xg3 Ag6 21 Ad3 a6 22 �d4 ± . The most chal lenging was 18 . . . dxc4 1 9 Ad6 19 . . . Ag6 20 ghg1 with a messy position (20 e4 ? �d5 ! =t). 19 e4 dxe4. Black does not choose the strongest contin uation :

A) 1 9 . . . b5 20 �h4 �xg3 21 h xg3 d x c4 22 tilxg6 �xg6 23 gh5 155 ; B) 1 9 . . . �xg3 20 h xg3 d xc4 21 �h4 Af7 (21 . . . b5 22 �xg6 hxg6 23 gd6 !?) 22 f4, and if 22 . . . b5 23 e5 fxe5 (23 . . . f5 24 g4) 24 �e4 with an attack, when in each case Wh ite has full compensation . 20 �h4 �xg3 (20 . . . f5 21 �xg6 h x g 6 2 2 Ax h5 g x h5 23 g h g 1 with an attack) 21 hxg3 f5 22 g4 ! ± and Wh ite's chances are better on account of the weakened pos ition of the black king (Gelfand - 8eliavsky, Linares 1 990) .

8) 1 3 �x b5 a6 and now : 8 1 ) 14 Ac7 gxc7 1 5 �xc7 �xc7 1 6 cxd5 �b6 !? 17 dxc6 Axc6 with an attack ; 82) 1 4 �bd4 �xd4 1 5 �xd4 (15 exd4 Axa3 16 bxa3 �xa3) 15 . . . Aa4 1 6 �d2 �b6 17 gc1 �e4 =t ; 83) 1 4 g 5 �e4 ; 84) 14 �c3 Axa3 ! 1 5 b x a3 �xa3 1 6 �b2 gab8 17 Ax b8 gx b8 18 �b5 �a5 with the initiative ; C) 13 g5 �h5 14 cxb5 (14 cxd5 �xf4 15 exf4 b4 !) 14 . . . �xf4 15 exf4 (15 bxc6 Axc6 16 exf4 d4 17 �e4 Axa3) 1 5 . . . �e7 16 �e5 Ae8 =t . 13 . . . �e7 14 �d2 !? (as the main action i s go ing to take place on the q ueenside, Wh ite moves his knights there; 14 gc1 !?) 14 . . . �d8 ! ( 1 4 . . . Axa3 ?! is interesting, b u t hard ly suf ficient.) 1 5 �b3 (15 Ad3 d4 !) 1 5 . . . �e4 ! (15 . . . Ab6 1 6 Ad3) 1 6 �xc5 (16 �xe4 Axe3 17 �x c8 gxc8 18 Axe3 d x e4 =t) 16 . . . gxc5 17 Ae5 �xc3+ 18 Ax c3 Ax b5 19 Ax b5 (1 9 b4 Aa4 !) 19 . . . gx b5 with approximately equal chances (Gelfand - 8el iavsky, Linares 1 991 ). II. 10 dxc4 My fi rst experience with the 1 0 0-0-0 variation was also entertain i ng : 10 . . . d x c4 1 1 Axc4 Ae7 1 2 g4 a6 ( 1 2 . . . e5 transposes into 10 . . . Ae7) 13 g5 �h5 14 Ad6 Axd6 1 5 gxd6 �e5 1 6 Ae2 �xf3 1 7 Axf3 �x g5 (Gelfand - 8eliavsky, 56t h USSR Champion ship, Odessa 1 989) and now 1 8 ghd1 ! was tem pting, as 18 . . . e5 (18 . . . �f6 1 9 gd8 �a5 20 �d3 �c7 21 �d6 �xd6 22 g1 xd6 ±) loses to 1 9 h4 ! (open ing the h -fi le with gain of tempo) 1 9 . . . �x h4 20 gh1 �e7 . • • •

A year later the Lvov grandmaster chose 12 b5!!. This move was awarded the prize for the best novelty in Informator Volume 51 . Black does not waste time but beg ins his play against the wh ite king. . . .

13 cxb5. Other options are too dangerous: A) 13 cxd5 b4 ! ;

87


21 �dd1 !! (I was real ly amazed when I found this combination in home analysis; 21 �xh5 g6 22 ttle4 f5 ! or 22 . . . Ae6 23 �xe6 �x e6 is far from clear) 21 . . . g6 (21 . . . ttlf6 22 ttld5) 22 Ax h5 g x h5 (22 . . . Af5 23 e4) 23 �dg 1 +
The main l i ne 10 .l.e7 1 1 g4 (I have also tried 11 h4 !? and 11
IV.

For 10 �e4, see the game Gelfand Yusupov analysed below (p. 92). • . .

1 1 �d2!? 11 �b5 !? deserved serious consideration , as later games have shown.

2) In such endgames it is favourable to have the king on the q ueenside, as it is closer to the action and it controls at least one file against the potential penetration of the op ponent's rook. White does not fall for the trap 15 �a4 ?! Ae7 1 6 ttlab6?? a x b6 1 7 ttlxb6 �a5 -+ 1 8 ttlxc8 �c5+. 15 . . . Ad7 1 6 gd1 Ac6 Black has developed his bishop, but now his d8 knight is shut out of play. 16 . . . �c8 !? was better, as in any case the rook belongs on this fi le, and if 17 ttld6 Axd6 18 Axd6 �d5. 17 f3 I nten d i ng to gain control of the d5 square by e3-e4. 17 g4 !? also came into considera tion , or 17 b4 Ae7 18 f3 �d5 1 9 �xd5 exd5 20 ttld6. 17 . . . �d5 1S �xd5 exd5

a b c d

1 1 . . . dxc4

8

Forced , in view of the threat of �b3.

7 6 5 4 3 2

12 �xc4 1 2 Axc4 !? Ae7 was also tempting, but I de cided to steer towards a better end ing.

12 . . . gxd1+ 1 3 .xd1 'OdS 14 .xdS+ �xdS 15 Ae2 ;t

a b c d

e

...

e

3 2

8 7 6 5 4 3 2

�------�--� �

6 5 4 3 2

1 8 . . . Axd5 1 9 e4 Axc4 20 Axc4 ;t would leave Wh ite with the advantage of the two bishops. 19 Ad6! The only way to fight for the in itiative ! Less good was 1 9 ttle5 ttle6 = , or 1 9 b4 Ax b4 ! 20 ax b4 d x c4 21 Axc4 Aa4 ! 22 �d6 �c8 23 � b2 ttle6 24 Axe6 (24 �xe6 �xc4) 24 . . .fxe6 =. 19 . . . b5?! Kiri l counterattacks as wel l . Black would have been condemned to passive defence after 1 9 . . . Axd6 20 ttlxd6 �f8 21 b4 a6 22 ttlf5 ;t .

Two factors determ ine White's advantage: 1 ) Black needs time to bring his d8 knight and c8 bishop into play. Moreover, they have only the c6 square, so one of these pieces will stil l be m isplaced .

88

Game 18


20 .txe5 bxe4

a b c d

24 . . . Aa4 (24 . . . gxa7 25 gd8+) 25 gd5 tDb3+ 26 �b1 g6 27 Ae3 gc8 28 f5 ! would not have solved Black's problems.

e

8 7 6 5 4 3 2

25 .te3 gxf4 26 .txf4±

a b c d 6 5 4 3 2

1...-_______---=-__.....

..

e

8 7 6 5 4 3 2

{f

6 5 4 3 2

21 e4! 21 Ad4 tDe6 = .

21 . . . dxe4 21 . . . tDb7 22 Ae3 d x e4 (22 . . J::! e 8 23 Axa7) 23 f4 would have transposed .

White's advantage is clear, but Black's next move is simply a blunder.

26 . . . f5? 27 Ild6 .ta4 28 Ild5 +- �b3+ 29 ct>b1 ge8 30 Ilxf5 �d4 31 Ilg5+ rm-r 32 .th5+ ct>f6 33 gd5 �b3 34 .tg4

22 f4!!

34 g4!?

A subtle positional decision. White fixes his strategic trumps (advantage of the two bish ops, open d -file) and as in any case he w i l l regai n his paw n , he will have a pawn ma jority on the q ueenside, whereas Black's passed pawn on e4 1acks any chance of be com ing active. As our team captain Sergey Makarychev said to me after the game, at first he thought I had blundered and it took him some time to realise what was going on.

34 . . . lle5 35 Ild6+ ct>t7 36 .te3 lle5 37 ga6

Also sim ple and strong was 37 Ae2 Ab5 38 �c2 , bringing the king into the game, as indicated by Artur Yusupov.

37 . . . .tb5 38 lla7+ ct>g6

a b c d 8 7 6 5 4

22 Axa7 tDe6 (22 . . . exf3 23 Axf3 Axf3 24 gxf3 tDe6 25 gd7 �f8 26 gb7 with the ini tiative) 23 gd6 !? exf3 24 gxf3 Axf3 25 Axf3 Etxa7 26 gc6 00 was far less convincing.

e

8

� I ,"-,

22 . . . �b7 22 . . . tDe6 23 Ae3 Ab5 24 f5 and Wh ite dom inates the board .

� {f

23 .te3 �a5 24 .txa7 White has restored material equ i l i bri u m , while keeping h i s positional trumps.

6 5 4 3 2

I...-______________��__

39 gb7? A mistake in time trouble. 39 a4 was winning.

39 . . . e3!

24 . . . g5

89


An excel lent resource ! Black gives up an other pawn, but starts an unpleasant coun terattack against the white king . 40 Db6+ 00 41 bxc3 41 �c2 !? would have made things easier for White. Here one extra pawn is enough for a win, and the king would have got out of difficulties. 41 . . . .td3+ 42 �a2 �a5 43 .tf4 Ac4+ 43 . . . .!:k 5 44 Ae6+ +-. 44 �a1 Dc5 45 .te3 De5 46 Dh6 Db5 Black does not pay any attention to one more pawn, but pins his hopes on a coun terattack. 46 . . . �g8 !? . 47 Dxh7+ �g8 48 Dh5 �b3+ 49 �b2 Db7 49 . . . �c5+ 50 �c2 Ad3+ 51 �d2 g b2+ 52 �e1 +-. 50 �c2 �a1 + 51 �c1 .td3 51 . . . Aa2 52 Af4. 52 .te6+ �g7 Black's pieces have achieved maximum ac tivity and so White has to be very accurate. 53 .ta2 Also possible was this long , but quite forced line: 53 Ad4+ �g6 54 g4 g b1 + 55 �d2 g b2+ 56 �e3 �c2+ 57 �f4 �xd4 58 cxd4 gf2+ 59 �g3 gf3+ 60 �h4 e3 61 gg5+ �h7 (61 . . . �f6 62 ge5 e2 63 g5+ �g7 64 Ag4) 62 ge5 e2 63 Af5+ Axf5 64 g xf5 gxf5 65 gxe2 gf4+ 66 �g5 gxd4 67 ga2 with an easily won rook end ing. 53 �c2 54 .tf4

54

•. .

�xa3?

Black cou ld have kept defending by 54 . . . e3 55 ge5 e2 56 gxe2 (56 Ag3 !?) 56 . . . Axe2 57 �xc2 ± , when Wh ite sti l l has to demon strate some technique.

55 Dg5+ �6 56 h4? I relaxed too early and the fight started again. 56 gc5 would have deprived my opponent of any counterplay.

56

•••

.tc4! 57 .txc4 Db1 +

I completely missed this intermed iate check. 58

�d2 �xc4+ 59 �e2 Db2+ 60 �e1 Db1 +

Or 6 0 . . . e3 6 1 gc5 g b1 + 6 2 �e2 g b2+ 63 �d3 ! (63 �f3 ? 63 . . . gf2+ 64 �g3 gxf4! 65 �xf4 e2) 63 . . . e2 64 Ag3 and Wh ite should win.

61 �2

a b c d 8 7 6 5 4 3 2

..

e

8 7 6 5 4 3 2

•..

a b c d

e

9

..

h

8 7 6 5 4 3 2

8 7 6

H ere the game was adjourned . I was sure that the position was won , but White has to overcome some technical obstacles. A l ittle later my team-mate Artur Yusupov returned from the birthday party of our m utual friend Lev Psakhis in order to help me. I was really amazed to see with what accuracy and seriousness Artu r analysed this simple -looking position. It was a great lesson for me!

4 3 2 a b c d

e

9

61 . . . Db2+ 62 �g3

h

90

Game 18

Gelfand - Georgiev, Olympiad , Novi Sad 1990

Wh ite's task is to win the e - pawn and not give up more than one pawn i n the mean time. 62 . . . e3 We also analysed in detaiI 62 . . . �c2 63 h5 : A) 63 . . . e3 64 h6 �c1 (64 . . . e2 65 �f2 tbb2 66 gd5 +-) 65 h7 (or 65 Axe3 tbxe3 66 �h5 �f5+ 67 �g4 tbe7 68 h7 tbg6 69 �h6) 65 . . . gh1 66 �f3 �xh7 67 �c5 ; 8) 63 . . . gxc3+ 64 �g4 �c2 (64 . . . e3 65 h6 e2 66 h7 +-) 65 h6 �xg2+ 66 Ag3 tbe3+ 67 �h3 gg1 68 h7 �h1 + 69 Ah2 +-. 63 He5 .tid2 64 ne5 ne2 A precise win was also found after 64 . . . tbc4 65 ge4 etJd2 66 �e8 �f7 67 ge5 tbc4 68 gc5 tLld2 69 �h3 tbe4 70 �c7+ +-. Or 64 . . . tbf1 + 65 �h3 and now : A) 6 5 . . . �f2 6 6 �e4 ! (66 Axe3 �e2 ; 6 6 g 3 e2) 66 . . . e2 (66 . . . �f5 67 Axe3 +-; 6 6 . . . tbd2 67 Ag5+ +-; 66 . . . �c2 67 Axe3 �xc3 68 gf4+ +-) 67 Ag5+ ! (67 Ag3 ? tbd2 with counterplay) 67 . . . �f5 68 �e8 followed by c4-c5 +-; 8) 65 . . . �c2 66 �e8 �xc3 (66 . . . �f7 67 �e4 gxc3 (67 . . . e2 68 �g4 ; 67 . . . �g6 68 g4 �xc3 69 h5+ followed by �h4) 68 g3, threatening �g2) 67 Ae5+ �f7 68 Axc3 �xe8 69 �g4 �f7 70 �f3 �g6 71 g4 tbh2+ 72 �f4 e2 73 h5+ �h6 74 �g3 tbf1 + 75 �h4 +-. 65 �h3 .tie4 66 ne4 e2 67 �g4 e1 n 67 . . . etJb2 68 Ag5+ �g6 69 �e7 . 68 nxe1 Ilxg2+ 69 Ag3 lle2 70 nf1 + �g7 71 Hf3

a b c d

e

f

9

...

h

8

8

7

7

6

6

5

5

4

4

3

3

2

2

Here we term inated our analysis. White has fu lfi l led his objective - the dangerous e pawn has been eliminated .

71 . . . ne2 72 Af4 ne8 72 . . . tbe5+?! 73 Axe5+ �xe5 74 c4 +- didn't promise any chances of survival .

73 Ild3 �6 74 nd4 lle8 75 lle4 �7 76 h5 .tib6 77 Ae5 .tid5 78 nd4 wasn 't too concerned about g iving u p my c - pawn - th e strong placing o f Wh ite's pieces should enable his h-pawn to promote.

78 . . .tixe3 .

78 . . . �e6 79 Ag7 !? +-.

79 �5 ne6 80 nd7+ �e8 80 . . . �g8 81 �g7+ �f8 82 �g6 �xg6 (82 . . . tbd5 83 �xc6 tbe7+ 84 �g5 tbxc6 85 Af6 �g8 86 �g6) 83 �xg6 tbd5 84 Ad6+.

81 nd3 Here the game was again adjourned . Black's only chance is to g ive u p his knight for the h - pawn to reach a theoretically d rawn endgame.

81 . . . .tib5 81 . . . tba4 82 �h3 +-.

82 nh3 82 Af6 ?! tbd6+ 83 �g6 tbf7 (83 . . . �f8 !?) 84 �g7 �e6 and if 85 h6? �e1 =; 82 Af4 �c5+ 83 �g4 �f7 84 h6 +-.

82 . . . Hh6 82 . . . �f7 83 h6. Strangely enough, for a long time I couldn't find a clear win and I went to sleep. Then around 4.00 a.m. I woke up with a clear win in my head !

83 �g5 83 Ag7 tbd6+ ! = ; 83 Af4 �h8 84 h6 tbd4+ !?

83 . . . ne6

91


a b c d

e

8 7 6 5 4 3 2

8 7 6 4 2

�------��-

�

86 1ld3!

84 £f4!! The point. I was reluctant to remove my bishop from the long diagonal . However, it is much more important to prevent the knight from coming back into play via d6 to f7. 84 . ctm 84 . . . ttld6 8S Axd6 gxd6 86 h6. 85 h6 c!>g8 U nfortunately, Black didn't al low the n ice finish 8S . . . ttld4 86 h7 gg6+ 87 �hS ttlfS 88 h8ttl+ ! +- (d iagram).

The fi nal finesse. Again , the kn ight should not be allowed to join the fight.

86 . . . 11e7 87 c!>g6 lle6+ 88 em5 lle7 89 £g5 Black resigns

..

And Black resigned , as he cannot prevent the h - pawn from q ueeni n g. Rather a long game; however, earl ier that year Kiril and I played a game which lasted even longer 91 moves !

*** 10

A quick but exciting draw

. • •

�e4 11 �b5!?

11 ttlxe4 d x e4 12 ttld2 hard ly promises much .

Boris Gelfand - Artur Yusupov Linares 1 992

11 ..

•. .

a6 12 �c7 e5!?

Black was relying on this counterattack

8 �--=--,

1 3 �xd5 Other ideas are:

6 5

A} 13 AxeS ?! ttlxeS 14 ttlxa8 ttlg4 + ; B} 1 3 ttlxa8 exf4 1 4 cxdS (14 gxdS ttlf6) 14 . . . ttlxf2 ! ; C} 1 3 ttlxeS ! ? ; O} 1 3 gxdS ! fS ( 1 3 . . . �xc7 14 �xe4 exf4 1 S gxcS fxe3 5O) 14 gxeS ttlxeS 1 S AxeS ± led to a quick win for Wh ite i n Kasparov Vaganian , (European Team Championship, Oebrecen 1 992). After this game the entire

2

2

Position after 10 0-0-0

92

Game 1 8


1 5 . . . exf4 16 .xf4 tile7 !

10 . . J��e4 line disappeared from tournament practice.

Forcing a draw; 1 6 . . . Ag6 17 h 4 �e7 00 .

13 . . . tilxf2!

1 7 tilxe7+

The best move ! - 13 . . . exf4 14 �xe4 ± .

1 4 tilg5 14 �xf2 exf4 15 �xf4 ge8 00). Now I expected 14 . . . f5 15 �xf2 exf4 16 �xf4 h6 (16 . . . b5 !?) when both 1 7 h4!? b5 and 17 �f3 Ae6 00 are difficult to evaluate.

a b c d

Or 1 7 Ad3 �xd5 1 8 �xf5 (18 cxd5 Axa3 ! 1 9 �xf5 �c3+ 20 Ac2 g6 21 �d3 �x b2+ 22
1 7 . . . Axe7 18 l1d5!

a b c d

•

e

8 .. 7 .l 6 .l 5 4 3 2

8 7 5 4 3 2

6 5 4 3 2

14 . . . Af5!! But this move came l i ke a bolt from the blue! - Black develops his bishop on the right diagonal .

15 .xf2 15 �xf5 g6 1 6 �f6+ (16 �f6 exf4 1 7 �xf4 etJxd1 1 8 �f6+
•

e

8 6 5 4 3 2

Both sides have to repeat moves ; 1 8 Ad3 ? Axa3 1 9 �xf5 �c3+ 2 0 Ac2 g 6 2 1 �e4 Ax b2+ 22 �b1 �a3 .

18 •e1 + 1 9 l1d1 .a5 20 l1d5 1re1 + 21 l1d1 Draw agreed . . .

After this game both players received a letter from the tournament d i rector Lu is Rentero with a warn ing that we would be financially penalised for making a quick d raw. I sti l l don't know i f t h i s was a publicity stu nt o r i f he simply wrote i t before looking at the game.

***

93

My Most Memorable Games Game 1 9

N i g e l S h o r t - B o r i s G e l fa n d Ca n d i d ates Q u arter- F i n a l M at ch (7) , B ru s s e l s 1 9 91 Pe tro ff Defence [C42] Th is q uarter-fi nal candidates match of the last unified world championship cycle, where my opponent was N igel Short, was one of the most memorable events i n my ca reer. It was excel lently organ ised by Bessel Kok in the Rad isson SAS hotel in Brussels. N igel and I were young and ambitious stars then , and both of us were especially strong with the white pieces. After winning the 6t h game and celebrating M i khail Botvin n i k's 80t h birthday on 18t h August, I had the fol lowing day to prepare some su rprise for N igel. I was one point down with two games to go. The m i l itary coup in M oscow that morn ing strongly affected the working mood of my seconds, but my good friend Eduardas Rozentalis, who came for a few days to watch the match, generously shared with me his ideas on the Petroff Defence, i n which he is a great special ist.

After 45 minutes' thought N igel introduces a strong novelty. The game Dvoi rys - Rozenta lis (Lvov Zonal 1 990) went 1 5 lixd5!? �xd5 16 h xg4 00 �c6 (the only move) 17 �g5 h6 18 �e4 (threatening both 19 �f3 and 19 g5) 1 8 . . . liad8 ! 1 9 g5 (1 9 �f3 ? �e5) 1 9 . . . life8 20 lie1 �xd4 ! 21 cxd4 �xd4 22 �c2 ! (22 �f6+? g xf6 23 lixe8+ lixe8 24 Ah7+
1 5 . . . axb5 16 1tc2 g6 16 . . . h6?! 17 g5 ± would merely give White an extra target to attack.

17 .th6 Later theory developed in the d i rection of 17 Ax b5 !? �d7 (17 . . .'�c8 ; 17 . . . �d6 has also been tried) 1 8 Ah6 !? (18 �b3 �f6 1 9 Ah6 �xg4 ! 20 Axf8
17 . . J �e8 18 .txb5 ge4 18 . . . lie6?! 1 9 c4 ; 18 . . . �d7 would transpose into the above -mentioned line.

1 e4 e5 2 �f3 �f6 3 �xe5 d6 4 �f3 �xe4 5 d4 d5 6 .td3 .td6

8

Ten years later I returned to this variation.

7 0-0 0-0 8 c4 c6 9 cxd5 cxd5 10 �c3 �xc3 11 bxc3 .tg4 12 Db1 b6 13 Db5 .tc7 14 h3 a6

a b c d

5 4 3

e

8

5 4 3

6 5 4

L..._ .._ _ _ _ _ _ �_----I

�

Wh ite has a pawn for the exchange, while Black's king is in danger, his d5 pawn is weak and h is knight on b8 is out of play. So, I be lieve that generally speaking Wh ite has to be better here, but this is far from easy to prove in practice.

2 �------��-

3 2

�

1 5 hxg4!

19 95

94

Game 1 9

A very dari ng move. White boxes i n the op ponent's king , but on the other hand he shuts his bishop out of the game. I was more wor ried about 1 9 c4 !? (if 1 9 �gS?! �d6) , which was later tried twice against Rozentalis: 1 9 . . . gxg4 20 cxdS Af4 21 �eS gh4 22 Axf4 gxf4 23 �c4 (23 �d2 gf6 24 �g4 gd6 25 �f4 ga7 26 ge1 ge7 27 geS �d7 28 gxe7 Wixe7 29 Axd7 �xd7 30 �f6+ gxf6 31 �xf6 WixdS = V2-V2 Kotronias - Rozentalis, Euro pean Team Championsh ip, Debrecen 1 992) 23 . . . �g7 24 gc1 ga3 2S gc2 ge4 26 �b4 gaS 27 a4 �a6 28 �b1 �xdS 29 Ac6 Wixd4 30 �f3 �d3 31 �b2+ �h6 32 �c1 + �g7 33 �b2+ �h6 34 �c1 + �g7 3S �b2+ �h6 36 �c1+ V2-V2 (Wahls- Rozentalis, Bun desliga 1 991 /92).

19

. . .

tt:J

Short - Gelfand, Candidates Quarter-Final Match (7), Brussels 1991

•d6

After the game I thought that Black could have gai ned an edge by 19 . . . ga3 !? , pre venting c3-c4 . But Wh ite can cool ly re ply 20 gd1 ! (20 c4 ? gxf3 21 g xf3 �d6 -+ ; 20 �d2 �d6 21 g3 ge7 =+= followed by . . . �c6 ; 20 g3 �c8 ! =+=) 20 . . . �e7 (20 . . . �d6 21 �eS �c6 (21 . . . gxeS 22 d x eS �xeS 23 g3 gxc3 24 �d2 ±) 22 Axc6 gxeS 23 d xeS �xc6 24 c4 +-) 21 Ad3 ge6 22 c4 with an unclear game. Sti l l , this was the strongest continua tion.

A) 21 f3?! gxeS 22 Axc6 ge7 ! -+ ; B) 2 1 Axc6 gxeS 22 f4 (22 Ax a8 ? ge2 ! -+ ; 22 d xeS �xc6 23 f4 ga3 ! =+=) 22 . . . ge2 23 �xe2 �xc6 with counterplay; C) 21 f4 ! �xeS (21 . . . �e6 transposes i nto the game) 22 fxeS gxeS 23 d xeS �xeS 24 g4 ! and Black does not have sufficient compen sation for the piece, as he can never ex change queens because of the poor position of his kin g , for example: 24 . . . Ad6 2S gf3 �e1 + 26 Af1 AcS+ 27 �h1 ge8 28 �a4.

21 f4! 21 f3 was the alternative : A) 2 1 . . . ge3 ? ! 22 f4 !? ± (22 �f2 ?! AxeS 23 �xe3 Ag3+ 24 �d2 �c6) 22 . . . ga3 23 Ad3 �c6 24 �c1 ; B) 21 . . . gxeS ! 22 dxeS �xeS 23 f4 . Th is pos ition was analysed after 20 . . . geS . It doesn 't g ive White such an advantage as in the game, but leads to a much sim pler position . Now it looks as though White is dominating the centre and there is no defence against 22 Ad3, but. . .

a b c 7 6 5 4 3

20 . . . gxeS 21 d x eS �xeS fai ls to solve all Black's problems:

B) 22 g3 ! Ad6 (22 . . . �a6? ! 2 3 Axa6 gxa6 24 Wid2 +-) 23 �d2 AcS 24 �g2 ± . I fou nd the right idea o n e move later, but after the best reply by Wh ite it would have led to the same position anyway. 20 . . . �c6 !? and now:

e

..

f

8

20 �e5 .e6

A) 22 f4 �e3+ 23 �f2 (23 �h1 �a6 00 ) 23 . . . Wixf2+ ! (23 . . . Axf4 ?? 24 ge1 ! +-) 24 gxf2 �a6 2S Ac6 (2S Axa6 gxa6 26 g3 (26 ge2 gaB 27 g3 bS) 26 . . . bS =) 2S . . . gd8 26 g3 �cS 27 ge2 ;!; ;

d

21

. . .

6 5 4 3 2

�c6!!

An excel lent chance, clearly the best (if 21 . . . AxeS 22 fxeS �c6 23 �f2 , threat ening Ad3 ±). Now Wh ite has to play ex tremely precisely to keep an advantage. When I was a j u n ior I remember bei ng i m pressed by a game Averbakh - Spassky, (23 r d USSR Championsh ip play-off, 1 9S6) , where Spassky played . . . �b8-c6 with a

95


wh ite pawn on d5, just to change the charac ter of the game i n an u nfavourable position . He s ucceeded and even gained the better position. During the game I didn't remember it, but probably my subconscious helped me here!

22 �xc6 22 23 26 25

32 �e4 (32 �h3 32 . . . .E!f2 33 �h1 (33 �e4 cf. 32 �e4) 33 . . . �e2 34 g4 �e3+ 35 �h4 .E!e2 36 Axf7+ ! =) 32 . . . .E!f2+ 33 �h3 �f1 + 34 �h4 .E!h2+ 35 �g4 �h3+ 36 �f3 (36 �f4 .E!f2+ -+) and now Black should be satisfied with a d raw by 36 . . . �f1 +, as 36 . . . �h5+? 37 �f4 .E!f2+ 38 �e5 .E!e2

f5 leads nowhere : 22 . . . �d6 ! (22 . . . g xf5 Axc6 Axe5 24 Axa8 Af4 25 .E!xf4 .E!xf4 �d2 +-) 23 Axc6 .E!xe5 24 d xe5 �xc6 e6 fxe6 26 fxg6 �c5+ 27 �h1 �d6.

After 22 Ad3 �xe5 23 fxe5 .E!xe5 24 d xe5 �xe5 25 .E!f3 �h2+ 26 �f2 .E!e8 (26 . . . �h4+ 27 g3 �h2+ 28 �e1 ) 27 �a4 b5 28 �d4 Ae5 Black has sufficient counterplay, as Wh ite hasn 't solved the problem of his king. I now believe that 22 Axc6 ! was best and probably the only way to win. However, after analysing the game for hours i n 1 991 , I thought that it led only to a d raw after 22 . . . Axe5 23 Axa8.

39 �f6!! leads to a beautifu l mate : 39 . . . .E!f2+ (39 . . . .E!xe4 40 Axf7+ �h8 41 Ag7#) 40 �e7 (40 �f4) 40 . . . .E!e2 41 Axf7+ �h8 42 �xe2 ! �xe2+ 43 �f8 +-. 26 c4 ! (White needs to bri ng his bishop into play at any cost ; if 26 �f2 �e4) 26 . . . d x c4 27 g 3 ! and Black has no defence : 27 . . . .E!h3 (or 27 . . . �g4 28 �e1 �xd4+ 29 �g2 �b2+ 30 �f1 .E!xh6 31 g x h6 �f8 32 Ad5 b5 33 �f2 +-) 28 �g2 c3 29 �c2 (29 �xc3 .E!xh6 30 gxh6 �xa2+ 31 �g1 �xa8 32 d5 ±) 29 . . . .E!x h6 30 g x h6 �e3 31 Ad5 �xd4 32 Ab3.

23 fxe5 is sufficient only for a draw after 23 . . . �xc6 24 �f2 �e6 25 �f6 �xf6 26 .E!xf6 .E!e2 (26 . . . .E!e3 27 .E!c6) 27 .E!x b6 .E!ex a2 28 .E!d6 .E!c2 29 .E!xd5 .E!xc3 30 .E!d7 .E!c2 31 d5 .E!e8. 23 . . . Axf4 24 .E!xf4 .E!xf4 25 �d2 .E!h4. The most stu b born defence. I thought that 25 . . . �f5 saved the game, but 26 Axd5 (26 �e1 ! +- refutes Black's idea.) 26 . . . .E!f1 + 27 �h2

However, it is possible to find all these l i nes only after prolonged home analysis, making use of analysing mod ules and so on. And it is extremely u n l i kely that they could be fou nd over the board , when there is a choice of other tempting possibilities.

22

. . .

lixf4!

It is more i mportant to keep the bishop on the board , rather than the rook : 22 . . . Axf4? 23 .E!xf4 .E!xf4 24 �e5 ± .E!e4 25 a4 .E!e1 + 26 �f2 .E!xe5 27 d xe5 �xe5 28 �d2 and White has won several tempi compared with the game.

27 . . . �g4 !! (27 . . . �xd5 28 �e2 +-) 28 �e3 �h4+ 29 �h3 �f4+ 30 g3 �f2+ 31 �g2 �e1

23 lixf4

96

Game 19

Short - Gelfand, Candidates Quarter-Final Match (7), Brussels 1 991

23 'tWd2 gxf1 + (23 . . . gh4 is also sufficient, transposing into 23 'tWe2) 24 �xf1 Ag3 2S a4 �fS+ (2S . . . 'tWe4 26 Ad3 'tWh4 27 'tWe3 ! ±) 26 �g1 'tWb1 + 27 Af1 , and now:

A) 27 . . . ge8 28 tileS AxeS 29 d x eS gxeS (29 . . . 'tWe4 30 'tWd4 'tWxeS 31 'tWx b6 'tWxc3 32 'tWf6 +-) 30 'tWd4 'tWe4 31 'tWx b6 'tWe3+ 32 �xe3 gxe3 33 as gxc3 34 �f2 +-;

26 a4 26 'tWd2 ga3 ! 27 Ac4 gxc3 28 Axd5 gc2 = ; 2 6 'tWf2 'tWxc3 2 7 'tWx b6 'tWa1 + ! 2 8 Af1 'tWxa2 29 'tWf6 'tWa1 = .

26 . . . •e1 + 26 . . . gc8 ? 27 �f2 gxc3 (27 . . . gcS 28 'tWf6 +-) 28 'tWx b6 +-.

27 �h2 licS

8) 27 . . . 'tWe4 28 AbS 'tWb1 + (28 . . . Af4 29 'tWe2 Ae3+ 30 �f1 'tWb1 + 31 'tWe1 'tWfS+ 32 �e2 ±) 29 Af1 =.

27 . . . 'tWh4+ =.

2S .d3 .eS+

23 �e2 ge4 24 'tWd2 gh4 2S ge1 (2S gf6 'tWe4 26 Ad3 gxa2 27 Axe4 gxd2 -+) 2S . . . Ah2+ 26 �f1 'tWfS+ 27 'tWf2 'tWxf2+ 28 �xf2 gxa2+ 29 ge2 gf4+ =.

28 . . . gxc3 29 'tWd4 'tWg3+ 30 �g1 'tWe3+ =.

29 .g3 .xc3 30 .14 30 'tWxc3 gxc3 31 �g1 ge3 ! =.

23 ... Ax14 24 �eS 24 �e2 ? gxa2 ! =+= 2S 'tWxe6 fxe6 26 tDd8 Ag3 27 tDxe6 ga1 + 28 Af1 gc1 .

24 . . . AxeS 2S dxeS .xeS =

30 . . .•c7 31 �g3 .c3+ ! 31 . . . 'tWxf4+ 32 �xf4 ± . Black must not allow the king to cross the e -fi le.

32 �h2 32 �g4 'tWc2 ! =.

8 7 6 I-� 5 4 3

tt:J

32 . . . •c7 33 �g3 .c3+ Draw agreed

....·.. _- 1

5 4 3 2

L...-_______----"-__....

{f

Neither side has grounds for playing for a win, because of the poor position of both kings. However, the match situation forced us to go on.

So, the resu lt of the match was decided in the last game, where an extraord inary de fence by Nigel enabled him to turn th e tables and 'snatch victory' in this really fascinating match , where only two of the eight games ended in a d raw (this was one of them) . I should add that Short followed up with vic tories over Karpov and Timman to become the chal lenger for the world cham pionsh ip title. After reaching h is peak, N igel lost some of his motivation . However, he quite often re minds the chess world what a strong player he is.

***

97

My Most Memorable Games Game 2 0

9 a3 h6

B ra n ko D a m lj a n ov i c - B o r i s G e lfa n d I n vest ban ka To u r n a m e n t , Be l g rade 1 9 91 English Opening [A D 7]

The immediate 9 . . . a5 !? was also not bad .

10 .c2 as 1 1 e3!?

I real ly enjoyed playing i n the I nvestban ka tournaments in Belgrade. Not only because the organisers provided excel lent playing conditions, but mostly because the people in Belgrade really love chess. Every round was watched by thousands of spectators (someti mes as many as 4000 !) , who had a good understanding of what was happening on the boards. I n add ition , every round was covered by a special 45-minute TV program at prime time. I think it was this atmosphere that gave me the necessary i nspiration to win this tournament in 1 991 (alone) and i n 1 995 (jointly with Vladimir Kramni k). A n d i n 1 995 I managed to beat Ivanch uk, Adams, Shirov, Leko and Lautier in one tournament !

A useful move, consideri ng that 11 . . . �d4 could be part of Black's plan .

11

12 . . . gd8 13 £d2 fS 1 4 �cS

a b c d

8 7 6 5 4 3 2

0-0 6 �bd2 �c6 7 c4

c d

N ow a Dragon Variation with reversed colours is reached , in which White has put his knight on d2, where it is less wel l placed than on c3 . 7 e4 e5 wou ld have led to a Pirc Defence.

14

•f7 !

1 S gab1

e

8 6 5 4 3

e

The e7 square is vacated for the knight, where it is ideally placed . Moreover, on f7 the queen is correctly positioned on the a2g8 diagonal .

7 . . . eS 8 cxdS �xdS

a b c d

. . .

...

e

7 6 5 4 3 2

Branko Damljanovic played most of his white games in this tournament as if he were Black, al beit with an extra tempo. . . .

•e7 12 �e4?!

The in itial cause of Wh ite's problems. 12 �b3 ! was much more solid, retaining the possibil ity of both �c5 and e3-e4 fol lowed by d3-d4 : 1 2 . . . gd8 (1 2 . . . a4 1 3 �c5 �b6 14 gb1 followed by b2-b4) 13 e4 �b6 14 Ae3 a4 15 �c5 oo .

1 �f3 dS 2 93 96 3 £92 £97 4 0-0 �f6 S d3

S

. . .

White could have exchanged a pai r of knights and gai ned some manoeuvring space by 15 gac1 b6 (but 15 . . . �de7 is similar to the game, and if 16 b4 axb4 17 ax b4 ga2) 1 6 �e6 Axe6 1 7 �xc6 gac8 1 8 e4 (18 b4 axb4 1 9 axb4 �e7 +) 18 . . . fxe4 1 9 dxe4 �e7 20 �b5.

6 5 4

1 S . . . b6 1 6 �a4 �de7 ! 17 �e1

2 L...-_______----"-__--'

{f

Other contin uations also do not completely solve Wh ite's problems:

98

Game 20

ttJ

Damljanovic - Gelfand, Investbanka Tournament, Belgrade 1991

B) 17 � h4?! A a6 ! 1 8 A x c6 A xd3 1 9 VWc1 gab8 !! 20 Ag2 A x b1 ;

8 7

7 6 5

C) 17 .l':'!fd1 Ae6 1 8 b3 (18 b4 a x b4 1 9 a x b4 e4 ! 20 d x e4 fxe4 21 � e1 .l':'!f8 !? 22 f4 exf3 23 �xf3 A a2 +) 1 8 . . . .l':'!d7 1 9 �c3 .l':'!ad8 20 Ac1 g5.

'-:"::':'�=-I""=-r-===;�+-=�

17 . . . .tb7 18 �c3

a b c d

..

e

8 7 6 5 4 3 2

7 6 5 4 3 2 a b c d

e

9

..

h

A) 17 b4? axb4 1 8 axb4 e4 ! -+ ;

h

c d

3 2

e

22 . . . f1b8! Analogous to one of the Spassky - Geller games in their 1 968 matc h . Wh ite cannot make use of the a-fi le.

23 �fd1 23 b5 would allow Black the choice of either acting as in the game, or exploiting some tactical possibilities : A) 23 . . . g4 24 �e1 Axg2 25 � xg2 �e6 ; B) 23 . . . �e6 24 .l':'!a7 �c5 (24 . . . Axf3 25 A xf3 .l':'! bd8 26 A c1 .l':'!xd3 27 A e2 e4 00 ) 25 � a4 e4 + .

18 . . . �d7! Vacating the d8 square for the kn ight, so that after the exchange of the light-square bishops it can take part in the attack via the weakened light squares. Black cou ld have won the exchange, but spoi led his position , by 18 . . . e4 19 dxe4 Aa6 20 exf5 VWxf5 21 Ae4 'tlVd7 22 �f3 .

2 3 . . . g 4 2 4 �e1 .txg2 2 5 �xg2 �e6 + 26 �g1 ?! Losing a tem po, but there is nothing useful that Wh ite can do.

26 . . . �g5 27 "a2 �f3+ 28 �g2 28 �xf3 g xf3. 28 . . . e4

19 b4 axb4 20 axb4 �d8! 21 �f3?!

a b c d

e

8 8 7 7 6 6 5 5 t----=_ 4 4 3 3 """"�� " ""' ':;;:';��-I 2 2

Dam ljanovic had only ten minutes left on his clock and he wastes some temp i . I should add that it was time troubles that only j ust prevented h i m from q ual ifying for the can didate matches a year earl ier i n the Manila Interzonal . 21 A x b7 � x b7 + was stronger, luring the knight away from kingside.

..

21 . . . g5 22 f1a1

�------��- �

(see next diagram)

99


The triumph of Black's strategy. He has fu l fi lled all his dreams, without allowing White any counter-chances.

29 Oxf7+ If Wh ite delays the exchange of q ueens by 29 d4 c5 ! 30 et)c2 , Black can chose between 30 . . . c4 !? (threatening . . . � h5) and 30 . . . cxd4 31 et) xd4 et) xd4 32 exd4 gxd4 33 Af4 gbd8.

29

. • .

�xf7 30 d4 c5! 31 �xf3

This is tantamount to capitulation , but other moves also do not hel p : 31 ttl b5 et) xd2 32 gxd2 cxb4 ! , intending . . . et) d5 and . . . Af8 -+, or 31 ttlc2 cxd4 32 ttl xd4 A xd4 ! (32 . . . et) xd4 33 exd4 gxd4 34 Af4 =F) 3 3 exd4 gc8 ! .

41

42 .te3 �d5 43 �c8 �xe3+ 43 . . . b5 44 ttla7 ttl c3 45 Ad2 A d4 46 et) c6 � d5 was an alternative.

44 fxe3 Af8 45 �xb6 .txb4 46 �2 �d6! Avoiding the trap 46 . . . Ac5 ? 47 ttlc4 � d5 48 ttl d2 A b6 49 ttl xf3 g xf3 50 � xf3 � e5 51 h3, and after g3-g4 Black wou ld be left with an h -pawn and the wrong bishop.

47 �c4+ �d5 48 �b6+ �e4 49 �a4 ..

8 7 6

38 �c7 Dd7 38 . . . gd1 39 ttl b5 A e5 40 g b7 h5 41 gx b6 et) d5. 39 . . . gd1 !.

40 �xa7 �e6 41 Ad2 f

9

8 7 6 5 4 3 2

8 7 6 5 4 3 2 a b c d

e

f

9

a b c d

..

h

4 3 2

3 2

39 �b5 Uxa7

e

h5?!

41 . . . et) d 5 ! -+ was simpler, not allowing the attack on the b - pawn . B ut I made a m is take, which unfortunately is a common one I didn't make a pause to calm down after the time control.

31 . . . exf3+ 32 � 1 cxd4 33 exd4 Uxd4-+ 34 .te1 Ubd8 35 Uxd4 Uxd4 36 Ua7 Ud3 37 �b5 Ud5 37 . . . A e5 ! , and if 38 g b7 gd1 39 gxb6 et) d5 40 gxh6 et) x b4 -+.

a b c d

• • •

e

f

9

h

49 . . . h4!? Or 49 . . . Ae7 50 ttlc3+ � d3 51 ttl d5 Ag5 52 h4 A h 6! 53 ttlf4+ A xf4 54 exf4 � d2 55 � f1 , and now 55 . . . � e3 56 � e1 f2+ 57 �f1 � d3 ! (57 . . . �f3 ?? leads to stalemate) or 55 . . . f2 ! -+.

50 g x h4 Ae7 51 h5 Ah4+ 52 �1 Ag5 53 �c5+ �xe3 54 �e6 Ah6 55 �d8 f4 56 �f7 g3 57 h x g3 fxg3 58 �xh6 g2+ 59 �g1 12+ White resigns

h ***

100

Game 21

Polugayevsky - Gelfand , Reggio Emilia 1 991 /1992

Game 21

8 . . . �g4!?

Lev Po l u g ayevs ky - B o r i s G e l fa n d Regg i o E m i l i a 1 9 91 / 1 9 92 GrOn feld Defence [002]

I considered Wh ite's last move to be inaccu rate and I tried to exploit it im mediately.

9 �ef3

Lev Polugayevsky was one of the g reatest players of his very strong generation . H i s games, his books (especially Grandmaster Preparation) and his attitude to chess had a big influence on my chess style. Also i n t h e Reggio Emilia tournament (the fi rst ever category 18 tournament !), although by far the oldest partici pant, Lev showed i m pres sive play and scored 50 % . This was also a memorable tou rnament for me, as I was i n contention for fi rst place until t he last round and I fin ished equal second with Garry Kas parov, just half a point beh ind Vishy Anand.

1 d4 �f6 2 �f3 g6 3 g3 Ag7 4 Ag2 0-0 5 0-0 d5 The most solid reply. Wh ite normally o b tains slightly t h e more comfortable game, but Black is OK.

Admitting his inaccuracy. However, there is l ittle for the knight to do on g4.

9 . . . Af5 I didn't see any reason to return immediately with 9 . . . �f6 !?

10 b3 1 0 ge1 is a pointless move. Black replies 10 . . . �f6.

10 . . . �f6 Black cou ld also have made use of his knight on g4 by 1 0 . . . e5 !? 1 1 d x e5 (11 h3 � xf2 !? 12 � xf2 e4 1 3 � e5 c5 oo ; 11 cxd5 e4 !? ; 11 A a3 e4 1 2 A xf8 �xf8 oo) 1 1 . . . � xe5 1 2 �d4 Ag4!? with chances for both sides.

11 Ab2 11 a4 �a6 = ; 11 a3 !?

6 �bd2 Too passive, but Lev probably didn't want to complicate the game after 6 c4 d xc4 .

6 . . . a5 Anticipating b2-b3 and preparing . . . a5-a4

7 c4 c6 8 �e5

11

. . •

a4

Black is fighting for counterplay. After 11 . . . �e4 1 2 �h4 �xd2 13 �xd2 Ae6 White wou ld play e2-e4 in favou rable circum stances.

1 2 �h4

This is not the best time for activity. 8 b3 is more natural, with equal chances.

After 1 2 bxa4 �a5 13 cxd5 cxd5 14 �e5 � bd7 15 � xd7 A xd7 16 � b3 �xa4 17 �c5 � xd1 18 gfxd1 a draw was agreed in Polu gayevsky- Kasparov a few rounds later.

12 . . . Ae6 1 2 . . . Ag4?! 13 h3 Ae6 14 f4 with an attack.

1 3 1tc2 Preparing to play e2-e4 and take control of the centre. If Black should allow this, he will find himself in difficu lties. The immed iate 1 3 e4 d x e4 14 � xe4 � x e4 1 5 A xe4 allows 15 . . . c5 ! with counterplay.

101


a b c d

..

e

a b c d

e

8 7 6 5

7 6 5 4 3

6 5 4 3 2 a b c d

9 h

1 3 . . . a3! 13 . . . tWc8 14 bxa4 ! ± ; 13 . . . tDa6 14 bxa4 ;t .

14 J.e3 e5! Si nce Black has decided to advance his pawn a long way (to a3 i n this case), he has to act energetically to make use of it.

15 Bad1 Black also has a comfortable game after White's alternatives :

Wh ite wants to make use of the pin on the d -file, but he misses a nice refuta tion . Exact play was cal led for: 18 cxd5 A xd5 (18 . . . tDd4 ! ? , and if 19 tWc5 gc8 ; 1 8 . . . tD b4?! 1 9 tWc4 tD xd5 20 tD b1 ) 1 9 A xd5 (1 9 �e4 A xe4 20 A xe4 tDd4 21 tWd3 tWb6 =+=) 1 9 . . . tWxd5 20 tDe4 tWe5 (20 . . . tD b4 !?) 21 �f3 tW b2 22 tD xf6+ tWxf6 (22 . . . exf6 23 tWx b2 axb2 24 gd2 ! =) 23 gd2 intending tWe4 with a probable draw.

18 . . . dxe4! 19 Iixd8 Bfxd8 +

A) 15 cxd5 tDxd5 16 Axd5 cxd4 ! (16 . . . tWxd5 17 e4 tW h5 18 d5 A xc3 19 tWx c3 A h3 20 tD g2 ;t ) 1 7 A x e6 (1 7 A x b7 ga7 =+=) 17 . . . dxc3 =+= ;

8 7 6 5 4

B) 1 5 d xc5 d4 1 6 A b4 tDc6 1 7 A x c6 b x c6 18 tDdf3 A h3 1 9 gfd1 tWb8 20 A e1 e5 00 ; C) 1 5 e3 !? cxd4 (15 . . . tD c6 1 6 d xc5 d4 17 exd4 tD xd4 1 8 tWd3 tD d7 19 tD e4 ;t ) 16 exd4 tDc6 17 c5 b6 18 b4 bxc5 1 9 bxc5 00 .

6 5 4

1 5 . . . exd4 1 6 Axd4 �e6 Black has completed hi s development and has a strong a3 pawn (the a2 pawn is a po tential target) , so it is White who has to play carefully to keep the balance here.

17 J.xf6 After 1 7 A a1 tWa5 1 8 cxd5 tD xd5 1 9 tD c4 tWc5 =+= it is White who may have problems due to the weakness of his a2 pawn .

17 . . . Axf6 18 �e4?

�------�--� �

Despite h is extra q ueen , White is u nable to defend his a2 pawn and prevent the appear ance of a new queen .

20 ttxe4 20 A xe4 tDb4 21 tWb1 gd2 22 tWe1 Ac3 + .

102

Game 22

Gelfand - Yusupov, SKA-Mephisto, Munich 1992

20 . . . Ild2

8

20 . . . tO b4 !? 21 'ffx b7 tO xa2 22 'ff b 6 =t .

8

21 f4 An attempt to confuse matters, but it is too slow. If 21 'ff b 1 .E!xe2 !?

21 . . . Ilxa2 22 f5 g xf5 23 �xf5 Ila1 23 . . . Axf5 !? 24 'ffx f5 .E!a1 -+.

24 �xe7+ �xe7 25 Ilxa1 £xa1 26 Oxb7 J.d4+ 27 e3 J.xe3+ 2S �1 Ila7 29 ObS+ �cS White resigns

h __....;9 11 ::...-_ L...-_a______e ----J

b c d

*** Game 2 2

B o r i s G e l fa n d - Ar t u r Yu s u p o v S KA- M e p h i st o , M u n i c h 1 9 92 Bogo - Indian Defence lE 1 1]

7 6

You may wonder why this simple game was included in the book. First of al l , it is a rare instance for such a great player with a solid style like Artur Yusupov to find h i mself i n a lost position after j ust 1 5 m oves. Artur played three ti mes i n Candidates sem i-fi nal matches, which says everything . I should like to add that Artur has an i mpeccable repu tation among his colleagues for his honesty, moral principles and friendl iness. Another reason why the game was included here wil l be mentioned later.

1 d4 �f6 2 c4 e6 3 �f3 Ab4+ 4 �bd2 d5 Black is fig hting for the centre. 4 . . . b6 was played in Gelfand - Korchnoi (Vienna 1 996) Game 35 on page 1 54.

5 0a4+ An important check. White forces the knight to c6, thus preventing the attack on the cen tre by . . . c7-c5. Black wil l be forced to waste a few tempi to achieve this.

6 5 4 3 2 9

h

6 . . . £e7 Black could also try giving u p his bishop i n order to establish his knight in the cen tre : 6 . . . Axd2+ 7 Axd2 tO e4. However, since he is unable to break up White's centre by . . . c7-c5 or . . . e6-e5 , I think that this must favour White, as for example in my first ever game with Alexey Shirov : 8 .E!d1 0-0 9 e3 tOe7 10 'ffc2 f5 11 Ad3 b6 12 0-0 A b7 13 c5 Ac6 14 tOe5 Ae8 15 Ae1 tOg6 16 tO xg6 hxg6 1 7 f3 tOg5 1 8 Ag3 and Wh ite stands m uch better (Gelfand - Shirov, USSR Junior Cham pionship, Oaugavpils 1 986).

7 e3 0-0 S Oc2 Preparing to meet . . . a7-a5 with b2-b3 .

S . . . a5 9 b3 £d7 10 £b2 IlcS

5 . . . �c6 6 a3

103


The 9 th world champion failed to achieve m uch after 10 . . . tL'l a7 11 Ad3 h6 1 2 0-0 (1 2 eL!eS !?) 12 . . Jk8 13 eL! eS cS 14 dxcS AxcS 1 S tL'l df3 (1S tL'l xd7 !? t ) 1 S . . . d x c4 1 6 b x c4 Ae8 1 7 E!fd1 'f!Jc7 1 8 Ac3 Ae7 1 9 'f!Jd2 A a4 20 E!db1 b6 21 Ad4 AcS %-% (Petrosian Gipslis, Moscow 1 967) .

1 1 Ad3 �a7 Prepari ng . . . c7-cS . However, I feel that this entire set-u p is rather d u bious and it is no wonder that after this game the entire vari ation went out of use i n top-class tourna ments.

12 �e5!? Black wants to play . . . c7-cS under favoura ble circumstances.

12

• • •

1 3 g4!? I t h in k that the wh ite pieces are excellently placed to launch such an attack after Black has weakened himself with . . . h7-h6. 13 cxd5 exdS 14 tL'l xd7 tL'l xd7 1S 0-0 tL'lf6 is too dull, wh ile 13 0-0 transposes into the Petrosian Gipslis game.

13

T he morn ing before the ga me I had some stomach problems and after play started I began to feel even worse. So emergency doctors were summoned and when this pos ition was on the board they entered the play ing hal l . Despite feeling u nwel l , I decided to play in accordance with the demands of the position .

a b c d

14 . . . a4? The decisive m i stake. Black does noth ing to counter the opponent's threats. He should have exchanged the eS knight in order to slow down White's attack: 14 . . . eL!c6 (if 14 . . . cxd4 1 S gS d x e3 1 6 fxe3) 1 S g5 (1S tL'l xd7 tL'l xd7 1 6 cxdS exdS 00 ) 1S . . . tL'l xe5 16 d x eS tL'l g4 17 g x h6 ! (it is i m portant to open the g -fi le; if 1 7 Ae2 hS 18 E!g1 a4 with counterplay) 1 7 . . . tL'l x h6 1 8 E!g1 , sti l l with an attack.

8 7 6 5 4 3 2 9 h

15 g5! +- axb3 16 "xb3 dxc4 17 Axc4

a b c d

...

e

8 7 6 5 4 3 2

7 6 5 4 3 2

e

8 7 6 5 4 3 2

c5 14 h4

After the doctors had checked me and given me some med icine, I returned to the board after around 40 minutes and decided to play in the most natural way. Also interesting was 14 d xcS A x cS (or 14 . . . tL'l c6 1 S E!g1 ) 1 S gS! h x g S 16 h4 d x c4 17 b xc4 g4 (17 . . . g x h4? 18 E!x h4) 18 hS with a strong attack.

h6?!

The i mmed iate 1 2 . . . cS !? 13 d x cS d x c4 leaves Black a pawn down : A) 14 tL'l d x c4 A xcS! (14 . . . E!xcS 1 S Ad4) 1S tL'lxd7 'f!Jxd7 16 Axf6 g xf6 17 Ax h7+ � g7 18 Ae4 bS with counterplay; B) 14 b x c4 ! ? , and if 1 4 . . . A xcS!? 1 S tL'l xd7 'f!Jxd7 1 6 A xf6 g xf6 1 7 Ax h7+ ±.

. • •

11 1

9 h 'if

17 . . . b5

104

Game 23

Gelfand - Shirov, I mmopar Rapid , Paris 1992

A desperate attempt to gain counterplay at the cost of a piece. 1 7 . . . ttld5 1 8 g x h 6 ! , or 17 . . . ttlh5 18 g6 1 .

8l 7 6 5 4 3 2

-__�·--��·�

__ _

1 8 gxf6 Axf6 19 �xd7 "xd7 20 Ae2 c4 20 . . . cxd4 21 ttle4 Ae5 22 f4.

21 "c2 c3 21 . . . 'fic7 22 Ac3.

22 Axc3 Ae7

8 7 6

3 2

22 . . . 'fic6 23 ttle4.

23 Elg1 "c7 24 �e4 f5

_ _ _ _ _ _ ----'''--_----' L..._ .-

(see next diagram)

1J

28 "xc3 Elxc3 29 �xb5 �xb5 30 Axb5 f4 31 exf4 exf4 32 a4 Dfc8 33 Dg4 f3 34 a5 D3c5 35 El b4 Dd8 36 a6 Dc7 37 Elb3 Black resigns

24 . . . 'fih2 25 � d2 (25 gf1 ) 25 . . . f5 26 d5 +-.

25 d5! e5 26 d6 Axd6 27 �xd6 "xc3+ 27 . . . 'fixd6 28 'fi b3+ � h8 29 A b4.

But not 37 a7 7? ge7+ 38 � f1 gxa7 1 . * * *

Game 23

B o r i s G e l fa n d - A l exey S h i rov I m m o p ar Ra p i d , Paris 1 9 92 Ca talan Opening [E04] The I m mopar tournaments were the first high-level rapid chess events. All the best players took part and the organ isation was perfect. Play took place on the prestigious Cham ps- E lysees and was popular among the Parisian public. It was impossible to ob tain a seat during the decisive games, and all this put the players i n a creative mood . The only drawback was at that time the Fischer clock (with an i ncrement after every move) didn't yet exist, and so the quality of play was general ly lower than after the introduction of this device (at least in my games). However, this is a fascinating game anyway. But please don't judge the players too strictly, as it was only a rapid game.

transposes into a Catalan-type position . The drawback is that in this type of position Black can capture the c4 pawn in more favourable circu mstances than in the Catalan , as the knight is already comm itted to c3 .

5 . . . dxc4 After 5 . . . ttl bd7 6 Ag2 d x c4 7 a4 Wh ite has fu ll com pensation for the pawn, but hardly more than that. 6 Ag2 b5

1 d4 d5 2 c4 c6 3 �c3 �f6 4 �f3 e6 5 g3 I used to employ this idea from time to time. Wh ite avoids a theoretical discussion in the Botvin n i k or M eran Variations and

�_______----,,,--_----,

1J

Black has g rabbed the pawn, so it is a real gambit now!

105


7 �e5 Wh ite gained good com pensation for the pawn after 7 Ag5 � bd7 B �e5 � xe5 9 Axf6 g xf6 1 0 d x e5 A d7 1 1 exf6 tWxf6 1 2 0-0 .!:!dB 13 a4 (Khal ifman -Sveshni kov, Moscow 1 9B7).

7 . . . �d5 8 a4

As Yuri Razuvaev expresses it in such cases, White's attacking coefficient (the number of pieces taking part in the attack, against the n u m ber of pieces defendi ng) is too high. Black has three log ical defences and it is amazing that Wh ite has to react differently in al l three cases !

18 . . . �c2

Wh ite provokes a weaken ing of Black's pos ition .

8 . . . f6 Giving u p the im portant bishop by B . . . A b4 9 Ad2 A xc3 10 bxc3 is not to everyone's taste.

9 �f3 b4 10 �e4 £e7 11 0-0 0-0 12 1!rc2 £a6 1 2 . . . f5 13 �c5 (13 � ed2 c3) 13 . . . A xc5 14 d x c5 g ives Wh ite fu l l com pensation for the pawn.

I. 18. . �d3 .

(bad is 1 B . . . AcB? 1 9 tWe2 or 1 B . . . �d7 19 � e6) is the most natural move, but it is hardly sufficient to save the game. 19 � x h7 ! (the point of Wh ite's concept he not only destroys the protection of the black king, but also secures a post for its colleag ue on g5 in the immediate vicin ity of the king ; if 1 9 .!:!e6 AcB) 1 9 . . . � x h7 20 � g5+ � g6 (20 . . . � gB 21 tWh5)

13 £h3 f5?! Th is allows Wh ite a d i rect attack. 13 . . . tWcB is safer.

14 �eg5 b3 1 5 1!rd1 Hf6 Not very logical is 15 . . . AcB?! 1 6 e4.

16 e4 �b4 17 exf5 exf5 18 He1

a b c d

..

e

Analysis diagram after 20 LZJg5+ \!;g6

8 7 6 5 3 2

21 .!:!e6 ! ! .

6 5 4

A n incred ible idea! It is i m portant t o d ivert the rook at f6 from the protection of the f5 pawn. Other tries are unsuccessful :

2

B 1 ) 21 . . .fxg4 is good enough for a d raw : 22 �xg4 �a5 (22 . . . � xf2 23 � h4 �gB 24 .!:!xe7 with an attack ; 22 . . . � xe1 23 �e4+ .!:!f5 24 �e6+ � h5 25 �xf5 A xg5 26 � h7+ A h6 27 � f5+ =) 23 � e6+ (23 �f3+ � f7 24 .!:!xe7+ � x e7 25 �x g7+ � dB 26 �xf6+ � cB leaves Black with an extra piece) 23 . . . � f7 24 �x g7+ � eB 25 � c7+ �xc7 26 Ag5 � xe1 27 .!:!xe1 .!:!f7 2B �gB+ .!:!fB = ;

A) 2 1 g4?! f4 ; B) 21 Ag4 - an orig inal , but insufficient idea:

So, White has burned his boats. He has ne glected the q ueenside and concentrated all his forces i n the d i rection of the opponent's king . I thought that White had a strong attack and deep analysis confirms this assessment.

106

Game 23

Gelfand - Shirov, Immopar Rapid, Paris 1992

B2) 21 . . . tD xc1 ! 22 tDe6 gxe6 23 A xf5+ (23 gxe6+ Af6 24 A h5+ h7 25 Af7 g6) 23 . . . xf5 24 �f3+ g6 25 �g4+ h7 26 gxe6 tDd7 !? 27 tWe4+ h8 -+ and Black's material advantage is too great ; C) 21 tDe6. I n contrast to the game, here this continuation doesn 't promise success : C1 ) weak is 21 . . . tW h8 (or 21 . . . gxe6 22 Axf5+ ! xf5 23 tW h5+) 22 A xf5+ ! and now : C11 ) 22 . . . f7 23 Axd3 gxe6 24 gxe6 cxd3 25 �xb3 and Black cannot hold the position ; C12) 22 . . . gxf5 23 �g4+ Ag5 24 tD xg5 (24 Ax g 5 ? tD xf2) 24 . . . tD xf2 (24 . . . tD x e1 25 tDe4+ f7 26 tWxf5+ g8 27 tD g5) 25 ge6+ gf6 26 gxf6+ g xf6 27 �e6 � f8 28 xf2 (28 tDf3 Ac8 29 tDe5+ h7 or 29 tD h4+ h7) 28 . . . A c8 29 �e4+ Af5 30 tWh4 fxg5 31 �xg5+ h7 32 Af4 and the attack continues ; C13) 2 2 . . . xf5, and now : C131 ) 23 tWf3+? g6 24 tWe4+ f7 25 tDg5+ � g8 26 tWx e7 � h5 (or 26 . . . tD d7) 27 tD e6 gxe6 28 gxe6 tDd7 ! 29 tWxd7 tWd1+ 30 g2 gfB 31 tWxg7+ xg7 32 A h6+ h7 33 gxd1 gxf2+ 34 g1 tD x b2 35 g b1 tD d3 36 gxc6 A b7 37 gc7+ xh6 38 gxb7 b2 and Wh ite is losing ; C132) 23 tD xg7+ g6 (23 . . . tWxg7 24 tW h5+) 24 gxe7 (Wh ite is only a piece down , and since three of the black pieces on the queen side are out of the game, Black shou ldn't be able to repel Wh ite's threats) 24 . . . tD xf2 (24 . . . tW h3 25 �e2 (25 tWd 2 ? tDf4 ! 26 g xf4 'l'g4+ =) 25 . . . gf7 26 tDe8) 25 tWe1 ! (25 tWd2 � h3+ ; 25 �e2 c3 -+) 25 . . . Ac8 (25 . . . tD h3+ 26 g2 gf2+ 27 h1 gx h2+ 28 x h 2 �g5+ 2 9 g2 tWh3+ 3 0 f2 tWh2+ 3 1 e3) 26 h4 +-;

f7 27 tWf5+ Af6 -+) 26 . . . xg5 (26 . . . f7 27 gxe6 xe6 28 tWe4+) 27 gxe6 Af6 28 h4+ h6 29 g4, mating ; C 1 332) 23 . . . Ad6 24 h 3 ! ! (weaker is 24 g4+ g5 25 h4+ ! �xh4 26 tD g2+ Af4 27 tD x h 4 x h4 with a dangerous attack, b u t Black can resist) 24 . . . tD xe1 25 tWg4+ e4 26 tDg2+ Af4 27 tD xf4 tDf3+ 28 g2 tDe1 + 29 h1 , threatening tWe2+ and tDg6+ ; C 1333) 23 . . . tD x e1 (best) 24 tWx e1 Ad6! (24 . . . tDd7 25 g4+ g5 26 tWe4) 25 g4+ ! g5 (25 . . . xg4 26 f3+) 26 h4+ and now : C 1 3331 ) 26 . . . x h 4 ? 27 tDg2+ xg4 28 �e4+ h3 (28 . . . h5 29 tW h4+ g6 30 tWx h 8) 29 �e3+ g4 30 tWg5+ h3 (30 . . . f3 31 tD e1 + e4 32 f3+ xd4 33 Ae3#) 31 Af4 ! +-; C 13332) Black should run away with 26 . . . h6 27 g5+ h7 28 gx f6 A xf4 (28 . . . g x f6 29 �e4+ g7 30 �g6+ f8 31 tD h5 tDd7 (31 . . . e7 32 tD xf6) 32 A h6+ e7 33 ge1 + d8 34 tDg7 00) 29 �e4+ g8 30 A xf4 � h5 ! (30 . . . �x h 4 31 �e8+ h7 32 Ag3 ! (32 f7 tD d7 33 tWxa8 tWxf4) 32 . . . tWxf6 33 g2 +-) 31 �g2 � f7 32 ge1 and after the cool-headed 32 . . . c3 ! (32 . . . tWxf6 33 ge8+ f7 34 Ae5 with an at tack) 33 ge8+ h7 34 tWg5 c x b2 35 h2 tDd7 (35 . . . b1 tW ? 36 fxg7 �x e8 37 g 8 tW + ! tWxg8 3 8 tW h6#) 3 6 gxa8 tD xf6 Black counter-attacks successful ly; C2) 2 1 . . . �d 5 ! (here Black can play this, as his knight protects the f4 square) 22 ge5 !? (the only possibility of continuing the attack).

C133) it is also interesting to try and trap the king on f5 with 23 tDf4 !? : C1331 ) bad is 23 . . . tD xf4 24 A xf4 ge6 25 tWb1 + (or 25 d 5 ! ? gxe1 + 26 tWxe1 cxd5 27 tWxe7 +-) 25 . . . f6 26 Ag5+ ! (26 Ae5+

107

...

Analysis diagram after 21

. . .

t!fd5! 22 J"fe5!?


And in this very messy position : C21 ) 22 . . . 'tWxe6 23 .§.xe6 .§.xe6 24 A xf5 + ! (24 'tWf3 .§.e1 + 25 � g2 Ac8) 2 4 . . . � xf5 25 'tW h5+ � e4 26 Ad2 ! with the i nitiative ; C22) 22 . . . 'tWxe5 23 dxe5 .§.xe6 and now:

A) 19 . . . hxg5 20 Axg5! (20 et) xg5 wou ld al low 20 . . . � d7 ! 2 1 A xf5 �f8) 20 . . . tDc2 2 1 A xf6 A xf6 22 A xf5 tD x e1 23 .§.xe1 and again the black pieces are unable to protect their ki ng; B) 1 9 . . . tDc2 , when :

C22 1 ) 24 'tW f3 !? .§.xe5 25 Af4 .§.d5 26 'tWe2 � f7 (26 . . . �d7!? 27 'tWxe7 �f6 28 Ag2 00 ) 27 'tWh5+ �f8 (27 . . . � e6 28 'tWe8 00 ) 28 'tW h8+ = ;

B1 ) 20 'tWxe7 'tWxe7 21 .§.xe7 hxg5 (21 . . . et) xa1 22 tDe6) 22 A xg5 � x a1 23 A xf6 g xf6 24 tD h4 ! (threaten ing A xf5 and Ae6) also looks very dangerous; B2) 20 Axf5 (the most direct way) 20 . . . et) xe1 (or 20 . . . h x g 5 21 � xg5 .§.xf5 (21 . . . � xe1 22 Ae6+ � f8 23 tDh7+ � e8 24 tD xf6+ Axf6 25 'tW h5+) 22 'tWe6+ � h8 23 'tWxf5 A xg5 24 A xg5 'tWc8 25 'tWg6 tD x e1 26 .§.xe1 ) 21 Ae6+ � h8

C222) Wh ite can also draw t h e game by 24 f4 tDd7 (24 . . . Ac5+ 25 �g2 et)f2 26 Axf5+ � xf5 27 'tW h5+ � e4 28 'tW f3+ =) 25 A xf5+ (25 g4 � h7 26 gxf5 .§.h6) 25 . . . �xf5 26 W# h5+ (26 g4+ � g6 27 f5+ � h7 28 fxe6 �7 x e5 with the i n itiative) 26 . . . � e4 27 'tWe2+ � d5 28 'tWf3+ � c5 29 Ae3+ = ; C23) 22 . . . 'tW h1 + ! (the best way t o simplify the game and transpose i nto a favourable ending ; however it is q u ite hard to spot !) 23 � x h1 tDxf2+ 24 � g1 tD xd1 25 tDc7 Ad6 26 � xa8 c3 27 bxc3 � xc3 and Black is in the driving seat. After 21 .§.e6 !! Black must continue 21 . . . 'tWx d4 (21 . . . Ac8 22 A xf5+ � xf5 23 'tW f3+ ; 21 . . . .§.xe6 2 2 A xf5+ � xf5 2 3 'tWf3+) 2 2 Ae3 (22 A xf5+ � xf5 23 'tW f3+ (23 Ae3 !?) 23 . . . � g6) and now : A) 22 . . . 'tWd8 23 Axf5+ ; B) 22 . . . .§.xe6 23 A xf5+ ! (noth ing can d istract Wh ite from his attack ; 23 A xd4 .§.e1 +) 23 . . . � xf5 24 'tW f3+ 'tW f4 (the only move ; if 24 . . . � g6 25 'tW f7+ � h6 26 'tWx e6+) 25 g xf4 Axg5 26 fxg5+ �g6 27 h4 and Black has no defence ; C) 22 . . . 'tWd5 23 .§.xe7 � x b2 is perhaps the most stu bborn , but I don 't bel ieve it can save the game, for exam ple 24 A xf5+ .§.xf5 (24 . . . 'tWxf5 25 'tWd4) 25 'tWg4 .§.xg5 26 Axg5. II. 18. . . h6

19 'tWe2 ! (now Axf5 is a real threat ; bad is the direct 19 et)e6 ?! 'tWd7 20 tDe5 'tWxe6 ! 21 tDg6 .§.xg6 22 .§.xe6 .§.xe6 23 A xf5 .§.f6 -+) and now:

22 �e5 !! (there are far more wh ite pieces attacking than black pieces defending, and this decides the outcome) 22 . . . 'tWxd4 (22 . . . �f3+ 23 'tWxf3 and if 23 . . . .§.xf3 24 �g6#) 23 � ef7+ � g8 24 � x h6+ � f8 25 �f5 .§.xf5 26 A xf5 c3 27 'tW h5 A xg5 28 Axg5 tDd3 29 Ae6.

19 �xh7 ! A s also after 1 8 . . . � d3, t h i s is t h e o n l y way, but a strong one, to start a direct assault.

19

. . .

�xh7

Or 1 9 . . . .§.d6 20 � hg5 (th reaten i ng Af4 ; 20 Ag5!? et) xe1 21 'tWxe1 Axg5 22 et) hxg5 00) 20 . . . tD xa1 21 Axf5 ! g6 22 Ae6+, mating.

20 �g5+ �g6 20 . . . � g8 21 'tW h5 .§.h6 22 'tW f7+ � h8 23 .§.xe7 +-.

108

21 �e6

Game 23

Gelfand - Shirov, I mmopar Rapid, Paris 1992

Or 21 ge4 fxe4 22 't'fg4 e3 23 �e6+ � f7 . Here, in contrast to the 18 . . . �d3 l ine, 21 ge6 can be wel l parried by 21 . . . � xd4 ! , protecting the f5 pawn (21 . . . 't'fxd4 22 A xf5+ � xf5 23 't'f h5) 22 gxe7 't'fx e7 23 't'fxd4 't'fe1 + 24 � g2 c3 and it is Black who is going to give mate.

21

. . •

• h8

21 . . . gxe6 22 A xf5+ � xf5 23 't'f h5+ g5 24 'tWf7+, mating.

8 7 �-J 6

whether Black can save any of these pos itions.

23 .f3+71 Wh ite loses the thread . Obviously, I was very excited after such a game, and while wal king the streets of Paris late that night I found 23 � xg7+ ! � g6 (23 . . . 'tWxg7 24 'tWh5+) 24 gxe7 with mating threats. I told this to my opponent the n ext day and after some thought he proposed 24 . . . 't'f h3 (24 . . . A c8 25 't'fd2 +-) 25 't'fd2 ? (stronger is 25 't'fe2 ! gf7 26 't'fe4+ � xg7 27 'tWe5+ ! , mating - Shi rov) 25 . . . 't'ff1 + ! ! 26 � xf1 c3+. 23 �f4 � xe1 transposes into a position ana lysed after 18 . . . �d3. 23 g6 24 .e4+ t7 25 �g5+ g8 26 .xe7 .h5? . • •

B lack should have g iven up some of his extra material and developed his pieces : 26 . . . �d7 ! 27 ge6 gxf2 28 h4 gaf8 and Wh ite's position is hopeless.

3 I'-i..=. 2 L..._ .- ______...._ .;;.... ---l iJ

27 �e6 gxe6 28 gxe6 �xa1

a b c d

22 Axf5+ ! xf5

f

8 7 6 5 4

Other replies were more tenacious, but would sti l l lose agai nst best play by White. 22 . . J �xf5 23 't'fg4+ � f6 (23 . . . Ag5 24 �xg5) 24 Ag5+ gxg5 25 't'fxg5+ � f7 26 't'ff5+ Ci!> e8 (26 . . . Af6 27 � g5+ � f8 28 't'fc5+, or 26 . . . � g8 27 � g5) 27 't'fe5 ! (this comp uter suggestion is the clearest way to win, but the human 27 't'fg6+ � d7 28 � x g7 � x e1 29 gxe1 is also good enough) 27 . . . Af6 (27 . . . �xe1 28 gxe1 A b4 29 �c5+) 28 'tWc7 ! , completing the net around t h e black ki ng. It wou ld seem that from the analytical point of view the best defence was 22 . . . � f7 23 � g5+ � e8 , although after 24 gxe7+ (or more sim ply 24 A xc2 b x c2 25 't'fe2 �f8 26 � h7 't'ff7 27 � xf6+ g xf6 (27 . . . 't'fxf6 28 h4) 28 Af4 ±) 24 . . . � xe7 25 't'fe2+ � f8 (or 25 . . . � d8 26 't'fe5 �d7 27 't'fa5+ � b6 28 Axc2 bxc2 29 't'fx a6) 26 't'fe5 � xa1 27 �h7+ (27 Af4 !?) 27 . . . 'tWx h7 28 Ax h7 �d7 29 'tWc7 ge8 30 't'fxd7 c3 31 � g2 ± I doubt

e

3 r--� 2 �

� iJ

�__

_ _ _ _ _ _ _

Black is acting very material istically. Sim pler was 28 . . . � d7 29 't'fxd7 gf8 30 A h6 'tWf7 (30 . . . gxh6 31 ge7 (0 ) 31 ge7 (31 'tWxf7+!? � xf7 32 gxc6 � x a1 33 Ad2 (0 ) 31 . . . 't'fxf2+ 32 � h1 'tWf3+ with perpetual check.

29 Ah6! Wh ite has only a few pieces left and he has to make maxi m u m use of their activity. He would have lost after 29 't'fd8+ � h7 30 ge8 � d7 ! 31 't'fx a8 't'fd 1 + 32 � g2 �f6 33 gh8+

109


�g6, but 29 'Se5 "fif7 30 "fid8+ � h7 31 "fih4+ was sufficient for a draw.

29 . . . •17 29 . . . g x h6 30 "fid8+ � g7 31 "fif6+ � g8 32 'Se7 +-.

30 .g5 �h7

a b c d

e

8 7 6 5 3 2

I overestimated my chances. After 31 'Se7 "fixe7 32 "fixe7 � xh6 33 f4 !? (33 "fie3+ leads to perpetual check) 33 . . . c3 ! , despite having a h uge material advantage it is Black who has to play accurately to gain a d raw, for example 34 f5 Ae2 35 "fi h4+ A h5 36 g4 g6 37 fxg6 c2 (quite an amazing position of the kn ight on a1 !) 38 "fix h5+ � g7 39 "fi h7+ �f6 40 "fif7+ � g5 and White must force perpet ual check.

31 . . .•xe6 32 J.e5 .g6!

5 4 3 _.....---_-1 2 �_--l

L..._ .______--'�_---J

31 J.xg7?1

I missed this move. Now the game is over.

33 .d8 33 "fie7+ � g8 34 "fid8+ � f7 35 "fic7+ � e8. 33 . . . �d7

34 .xd7+ ct>g8 35 f4 c3 36 f5 cxb2 White resigns

if

*** Game 2 4

8 J.d3 J.b7 9 a3

B o r i s G e l fa n d - V i s wa n at h a n A n a n d I n terzo n a l To u r n a m e n t , B i e l 1 9 93 Semi-Sla v Defence [04 7]

a b c d

...

e

8 7 6 5 4 3 2

The Biel I nterzonal was one of the g reat est achievements of my career. I played a number of entertaining games and I was for tunate enough to add this victory to my suc cess in Manila 1 990, thus winning what were probably the last two Interzonal Tournaments in the history of chess.

6 5 4 3 2

1 d4 d5 2 c4 c6 3 .tlc3 .tlf6 4 e3 e6 5 .tlf3 .tlbd7 6 J.d3 dxc4 7 J.xc4 b5 Another i nvention of one of my favou rite players, Akiba Rubinstein, which he fi rst in troduced in Meran in 1 926, which gave it the name of the M eran Variation. I enjoy the re sulting positions and I have played this vari ation several ti mes with both colours. It was also one of the main components of Vishy's repertOire for many years.

I have successful ly employed this harmless looki ng idea on several occasions.

9 . . . b4 10 .tle4 1 0 a x b4 A x b4 1 1 0-0 is not dangerous, as was shown by 1 1 . . . c5 1 2 �a2 a5 1 3 � x b4 a x b4 14 'Sxa8 "fix a8 1 5 Ae2 0-0 1 6 d x c5 � xc5 1 7 "fid4 A a6 1 8 A x a6 � x a6 1 9 Ad2

110

Game 24

Gelfand -Anand, I nterzonal Tournament, Biel 1993

�d8 20 �c4 � e4 21 A x b4 %-% (Gelfand Shirov, USSR Championsh i p First League, Klaipeda 1 988) .

After 1 6 . . . � xc5? 1 7 A x b7 � x b7 1 8 �a4+ the king is forced to move and White gains attacking chances.

10 . . . �xe4

17 .a4

10 . . . c5 is too ambitious, as it leads to too many weaknesses i n the black pos ition : 11 � xf6+ g xf6 1 2 0-0 gg8 ?! 13 �e2 � b6 14 a x b4 cxd4 15 exd4 �c6 16 c;;, h 1 Axb4 1 7 Af4 gg4 1 8 Ag3 gxd4 19 gg1 ?! �e4 ? 20 A x e4 �x e4 21 �x e4 A x e4 22 ggc1 ± %-% (Gelfand - Sh i rov, M elody Amber Monaco rapid 2002) .

1 1 Axe4 .c7 Nowadays 1 1 . . . bxa3 is considered to be the strongest : 12 0-0 Ae7 13 �c2 �c7 14 bxa3 c5 15 gb1 Axe4 16 �xe4 0-0 1 7 gb7 �c8 = (Gelfand - Bareev, Russia-Rest of the World rapid, Moscow 2002) .

12 axb4! N I

failed to achieve anything after 1 2 0-0 bxa3 13 bxa3 Ad6 14 gb1 0-0 15 �c2 h6 1 6 Ad2 gab8 17 A b4 c5 18 d x c5 A x e4 19 �x e4 CDxc5 20 �g4 ttld3 (Gelfand - Bareev, Linares 1993).

Trying to keep the black king in the centre. H owever, after the s imple 17 �xc5 � x c5 1 8 A x b7 � x b7 1 9 c;;, e2 Black had to suffer to make a draw in Bareev- Kramnik, (Linares 1 994) .

17 . . . Hb8 The only defence. 1 7 . . . �c7 looks tem pti ng, but then 18 �a3 ! (18 0-0 Axe4 19 gac1 �b7 2 0 � x e4 0-0 2 1 � d6 � b6 leads to an al most eq ual position) poses serious pro b lems : 1 8 . . . �f6 1 9 A x b7 �x b7 20 � c4 ! (20 0-0 �e7 =) 20 . . . �xg2 21 �d6+ and now : A} 2 1 . . . c;;, f8?? leads t o a smothered mate after 22 �f5+ c;;, g 8 23 � e7+ c;;, f8 24 � g6+ c;;, g 8 25 �f8+ gxf8 26 �e7# ;

12 . . . Axb4+ 1 3 Ad2 Axd2+ 13 . . . c5 14 A x b4 c x b4 15 gc1 ;t and Wh ite keeps an advantage. 14 �xd2 c5 1 5 .c2! .b6 1 6 dxc5 16 A xb7 �x b7 1 7 dxc5 �xg2 18 O-O-O !? 00 leads to a sharp game (if 1 8 �e4 �xe4 19 ttl xe4 f5 with cou nterplay) . 16

. . .

•xc5

a b c d

e

7 6

6

5

5

4

4

3

. B } 2 1 . . . c;;, d 7 2 2 O-O-O ! with a decisive attack, for example 22 . . . �d5 (22 . . . �c6+ 23 c;;, b 1 �d5 24 � xf7 +-) 23 e4 ghc8+ 24 c;;, b 1 ! (but not 24 � x c8 gxc8+ 25 c;;, b 1 �x e4+ 00) 24 . . . gc6 25 exd5 �g6+ 26 c;;, a 1 gxd6 (26 . . . exd5 27 � b7 +-) 27 dxe6+ �xe6 28 �a4+ c;;, c7 29 gxd6 c;;,x d6 30 �b4+ c;;, c7 31 gc1 + c;;, d 8 32 �f8+ �e8 33 gd1 + c;;, c7 34 �d6+ c;;, b7 35 � b4+ c;;, c7 36 gc1 + c;;, d 8 37 �a5+ and wins.

18 0-0

2

L...-_______....:_ ::... ----I

lf

None of White's other options achieves anything : 18 �xa7 Axe4 19 ttlxe4 �b4+ 20 ttld2

111


0-0 (and if 21 ffxd7 gfd8), 18 � b3 ff b6, or 18 gd1 Axe4 19 �xe4 ff b4+ =.

a b c d 8 7 6 5 4 3 2

e

a b c d

f

8

8 7 6 5 4 3 2

...

f

e

8 6 5 4 3 2

6 5 4 3 2 c d

'if

e

20 Axh7+ ! 1B . . . 0-O! Again the best, if not the only reply. The endgame after 18 . . . A xe4 19 � xe4 ff b4 20 ffx b4 gx b4 21 gxa7 ± is very d ifficult, if not lost. Now I had a number of alternatives and I spent q u ite a lot of time calculating them .

19 1txd7 Wh ite's i n itiative would have evaporated after 1 9 gac1 ffd6 20 � c4 ffe7 = , but he could also have tried 19 A x h7+ � x h7 20 ffxd7 : A) 20 . . . A xg2 ? 2 1 � xg2 g bd8 22 � e4 ff f5 23 ffa4 ; B) 20 . . . gfd8 21 ffxf7 gxd2 22 ga4 +- leads to the game continuation ;

This is White's last chance. Blac k can hold the position after either 20 gac1 ff b4 ! (the only move), and if 21 gc4 ffx c4 22 ffxd8+ gxd8 23 A x h7+ � x h7 24 � x c4 A a6 25 b3 gb8, or 20 b4 ff b6 (the only move) 2 1 �c4 (21 ffe7 gxd2 22 Ax b7 ffx b7 23 gxa7 ffxe7 24 gxe7 g b2 =) 2 1 . . . gxd7 22 � x b6 axb6 23 Ax b7 gd x b7 =.

20 . . . �xh7? Vishy took the bishop instantly, thereby fallin g i nto a trap. The famous speed of play of the 2000 world champion has turned against h i m q u ite a few times in his ca reer. After the simple 20 . . . �f8 ! 21 ffa4 gxd2 22 Ae4 (22 gac1 ffd5 ! 23 e4 ffd4) 22 . . . Axe4 23 ffx e4 g6 Wh ite would have only a sym bolic advantage.

C) 20 . . . ffd5 2 1 ffxd5 A xd5 22 e4 gx b2 !? 23 exd5 gxd2 24 d x e6 fxe6 25 gxa7 gc8 ;!;; and White is a pawn up, but the most prob able resu lt is a draw; D) 20 . . . ffg 5 ! ? 21 ffd3+ ;!;; (21 g3 gfd8 22 ffxf7 ffd5 !), when Black has some com pensation , but I am not sure if it is enough ; E) 20 . . . Ad5 ! , and Wh ite cannot achieve more than a 4 v. 3 ending, which should be a draw.

19 . . . IUdB

112

21 1txf7 Hxd2? e

8 7 6 5 4 3 '---''-0-. 2

f

9 h 8

L """",,- , -

��==9-_��__'��1

6 5 4 3 2

Game 25

Gelfand - Shirov, Chalkidiki 1993

This move was also played very q u ickly. 21 . . . �d 5 ? was hopeless in view of 22 �f3 gf8 23 �c7 (or 23 �x e6 +-) 23 . . J !xf3 24 �x b8 gxe3 25 f3 +-, but, as Anand pointed out to me the next day, 21 . . . A xg2 was the most stu bborn defence : 22 c;!? xg2 (22 gxa7 ?? �g5 -+ ; 22 ga4 ? �g5 ;t ) 22 . . . gxd2 (i nterposing 22 . . . �g5+ 23 c;!? h1 gxd2 loses to 24 gg1 �d5+ 25 f3 �e5 26 gg3 gx h2+ 27 c;!? x h2 �x b2+ 28 c;!? h3 'ffx a1 29 c;!? g4 ! +-), and now :

A} 23 ga4 looks l i ke a blunder, but after 23 . gb4 ! (23 . . . �c6+ 24 c;!? h3 �xa4 25 gg1 and mate is unavoidable) 24 gx b4 �x b4 25 �xe6 �x b2 Wh ite's chances of winning and Black's of d rawing are approximately equal ; . .

B} 23 gxa7 �g5+ 24 c;!? h1 �d5+ = (25 f3 ? �g5) ; C} 23 �xe6 gbxb2 (23 . . . gb6 24 �e4+ gg6+ 25 c;!? h1 ± ; 23 . . . �g5+ 24 c;!? h1 ±) 24 �e4+ (24 ga4 gxf2+) 24 . . . c;!? g8 25 c;!? g1 and the chances of a win and a draw are again 'fifty fifty' .

22 Ha4! Og5 23 g3! +This pawn move, defending Wh ite's king and creating i rresistible threats to its black col league, was missed by my opponent. How ever, it was also missed by some of the other top players , as after the game they told me they thought I was losing ! Now it is all over.

23 . . . e5 24 Hh4+ Oxh4 25 g x h4 Hd6 26 h5 Ae4 27 Oe7 Hbb6 28 Oxe5 He6 29 Of4 Black resigns

*** Game 2 5

q ueenside. When these notes were written in 2003 , no one had even come close to re futing it. On the contrary, it had g rown in popu larity an d even Garry Kasparov used it a number of times, once against me.

B o r i s G e l fa n d - A l exey S h i rov C h a l k i d i k i 1 9 93 Sla v Defence [0 1 5J Unfortunately, the Chal kid iki tournament has a very short history, as only two versions of it took place. But all the players who took part in them remem ber with pleasure the warm hospital ity of the organisers of this sea resort event. Because of the awkward tournament formula, I had to play fou r games against Alexey there, and I won two of them in good style.

a b c d

e

6 5 4 3

1 d4 d5 2 c4 c6 3 �c3 �f6 4 �f3 a6 This variation was invented by the splen did Moldavian trai ner and theoretician Vya cheslav Chebanen ko and has become very pop u lar of late. Despite its strange appear ance, it has logic beh ind it. Black is prepar ing . . . b7-b5 and all of Wh ite's repl ies have their drawbacks : e2-e3 blocks in the bishop, £g5 al lows . . . �e4, and a2-a4 weakens the

�------��- �

5 �e5 Traditionalists don't recommend movi ng a piece twice in the open ing, and this move cannot pose Black any serious problems. H owever, it has its log ic. If Wh ite should

113


manage to play Af4 and e2-e3 (as in the game) he can be happy with the outcome of the open ing. Moreover, it is never pleasant to be faced with a novelty, as you have to work out at the board what you r opponent has analysed at home

5 . . . �bd7 6 cxd5!? After 6 Af4 d xc4 7 � xc4 b5 8 �e5 A b7 fol lowed by . . . e7-e6 and . . . c6-c5 Alexey achieved a good position against Korchnoi (Vienna 1996) and Bel iavsky (Olympiad, Yere van 1 996).

6 . . . cxd5 Th is standard reply to the novelty leaves Wh ite with some advantage. I think that 6 . . . � xe5 ! ? 7 d x e5 � xd5 was better, and this assessment was confirmed i n a n u m ber o f games, e. g . 8 e 4 � x c3 9 �xd8+ c;!? xd8 1 0 b x c3 e6 1 1 a4 c;!? c7 12 h4 Ad7 1 3 a5 h6 14 .E!h3 Ae7 1 5 .E!f3 .E!af8 = (Oiz dar- Schandorff, Olympiad , Moscow 1 994) .

7 Af4 e6 8 e3 b5 9 Ad3 Ab7 10 0-0 Ae7

a b c d

5 4 3

A} 1 2 �e2 !? .E!c8 1 3 a5 0-0 14 �c1 � xe5 1 5 A x e5 � e4 1 6 � b3 ;t; (H O bner- Pieper Emden , Bundesl iga 1 991 ) ; B} 1 2 � b1 N 0-0 1 3 a5 ;t; � x e5 ( 1 3 . . . �e4 !? Bel iavsky) 14 A x e5 Ad6 1 5 A xd6 �xd6 1 6 �d2 e5 1 7 � b3 Ac8 1 8 Ae2 .E!e8 1 9 .E!c1 h6 20 dxe5 �xe5 21 tLld4 (Bel iavsky- Kallai , Hungarian League, Budapest 1 996) .

12 . . . 0-0 1 3 �c1 �xe5 14 dxe5 A very ambitious move. Wh ite gains the d4 sq uare for his knight and d reams about an attack on the kingside. Better was 14 A x e5 ! ? , with a slight advantage, similar to the H O bner game.

14 . . . �d7 14 . . . � e4 !? was an alternative, but Alexey wanted to keep open the possibility of at tacking the e5 paw n . 1 5 �b3 flc8 1 5 . . . � c5 ? 1 6 tLl x c5 A x c5 1 7 A x h7+ c;!? x h7 18 'fWc2+ c;!? g8 19 'fWxc5 leaves Wh ite a pawn up. 16 .e2 After 16 a5 tLlc5 the a5 pawn is a weakness.

e

1 6 . . .• b6 1 6 . . . a5 1 7 � d4 tLl c5 1 8 A b5 ;t; and White maintai ns some pressu re.

6 5 4 3 2

�------�--�

17 a5 17 .E!fc1 !? ;t; was more subtle, as the pawn on a5 is more of a weakness than an asset.

17 . . .•a7 18 flfc1 .a8! = �

Both sides have almost completed their development. Wh ite stands slightly better than ks to the more active placing of his pieces.

11 a4 b4 12 �a2 The knight is head i ng for b3 , but it can go there by different routes :

A strong defensive move, prepari ng to ex change rooks on the c -fi le. 1 8 . . . � c5 !? was also i nteresting, as 1 9 � xc5 (1 9 �d4) 1 9 . . . .E!xc5 20 .E!xc5 �x c5 ! (we both m issed this possi bil ity; if 20 . . . A x c5 2 1 � h5 g6 22 � h4 ! intending Ag5 ±) 21 A x a6 (21 'fW h5 g6 22 �g4 is worth considering) 2 1 . . . A xa6 22 'fWxa6 'fWc2 gives Black sufficient counter chances.

19 .h5 19 Ag3 .E!xc1 + 20 .E!xc1 .E!c8 =.

114

Game 25

Gelfand - Shirov, Chalkidiki 1 993

23 . . . .tdB

19 . . . g6 20 .g4 Intending h2-h4.

Black begin s a d i rect attack on the white pawns. He had a wide choice of alternatives :

20. . . Dxc1 + 21 flxc1 DcB

A) 23 . . . �cS 24 � xcS EtxcS 2S AgS ;t ;

8

B) 23 . . . Etc4 !? was a tempting exchange sacrifice, as after 24 Axc4 dxc4 2S �d4 �cS 00 Wh ite's attack is over, h i s pieces are awk ward ly placed , and Blac k will soon have a far-advanced passed pawn ; it is better to stick to the attacking plan 24 h4 �d8 2S AgS "(but not 2S hS gS) ;

5 4 3 2 ____

L..._ .._ _ _ _ _ _

.....

_ _

C) d u ring the game I was m ostly worried about the prophylactic 23 . . . � g7 ! 24 h4 h6 oo , restricting White's attack ;

lf

22 fla1 Trying to avoid fu rther simplification . The rook can be useful protecting the back rank and the as pawn, a nd Black's rook h a s no points of entry. 22 Etx c8+ �x c8 23 h4 hS 24 VWh3 (24 �g3 �d8 ! 2S AgS AxgS 26 hxgS �c7 +) 24 . . . � g7 (24 . . . � cS 2S � xcS �x cS 26 �g3, threatening AgS with the i n itiative) 25 g4 � h8 26 Ag3 was good enough for equal ity.

22 • bB 22 . . . �cS ? 23 � xcS EtxcS 24 AgS allows White control of the dark squares.

D) 2 3 . . . Af8 24 h 4 Ag7 2 S hS is similar to the next note.

24 �d4 A worthy alternative was 24 h4!? Ac7 2S hS � x eS (2S . . . A x eS 26 h xg6 h xg6 27 A xg6 fxg6 28 �xg6+ � f8 29 �x e6 with danger ous threats) 26 �cS � xd3 27 � xd3 and Wh ite's control of the dark squares g ives him fu l l compensation for the pawn.

24 . . . �c5

. . .

23 .g3!? If 23 �e2 �a8 (i ntend ing . . . � cS) 24 �g4 �b8 = .

a b c d 8 7 5 4 3 2

..

e

8 6 5 4 3 2

24 . . . EtcS deserved serious attention, as the tempting 2S � xe6 fxe6 26 A xg6 leads only to a d raw after the cool-headed 26 . . . �c7 ! 27 A x h7+ (27 Etf1 �f8) 27 . . . � x h7 28 � h3+ �g8 29 VWxe6+ � h7 30 VW h6+ � g8 31 VWe6+. 24 . . . Ac7 is too provocative : 2S � x e6 ! fxe6 26 A xg6 �f8 (26 . . . A xeS 2 7 Ae8+ +-) 27 A hS+ � h8 28 Af7 et:)g6 29 �h3 (29 Axg6 Etg8) 29 . . . � x eS 30 A x e6, or 26 . . . � x eS 27 Ac2+ � f7 28 � h3, in both cases with a strong attack for the piece.

25 Ab1 2S Ac2 b3 ! with counterplay.

25 . . . .txa5? Too g reedy. B lack does not sense the dan ger:

115


A) 25 . . . et)e4 26 Axe4 dxe4 27 h4 ± ;

29 hxg6 hxg6 30 Axg6! fxg6 31 flxg6+ �hS

B) 2 5 . . . et)a4 2 6 h 4 gc1 + 2 7 � h2 et)c5 28 h5 when Wh ite has dangerous threats, but Black has a fine resource - 28 . . . gx b1 ! (28 . . . et)e4 29 A xe4 gxa1 30 h x g 6 d x e4 31 g xf7+ � xf7 32 et)f5 +-) 29 gx b1 et) e4 30 �h3 (30 �f3 et)d2) 30 . . . et)xf2 31 �f3 et)e4 with compensation ;

a

b

c

8 7 5 4 3 r--ioo...-2

C) 25 . . . � g7 ! ? , as i n the note to Black's 23r d move, deserves consideration : 26 h4 (26 Ag5 et)d7 +) 26 . . . h6 00 .

d

e

8 7 6 5 4 3 2

26 h4! 26 gxa5?? et)e4 27 A xe4 gc1 #.

'--_______----=-__--'

26 . . . AdS 27 h5 flc7 a

b

c

d

8 I- ��::::;.Ji"= "" 7

If

32 lic1 ! The key move of Wh ite's attack. The sacri fice on c5 decides the game in a n u mber of cases.

e

8

..

L--_______----=-__...J

32 . . . Dc7

If

27 . . . et)e4 28 A x e4 d x e4 29 h xg6 h xg6 30 et) xe6 +- destroying the black position .

2S Ah6!

32 . . . et)e4 33 gxc8 A xc8 34 et)c6 +-. Per haps 32 . . . �h7 was more stubborn : 33 �e8+ (33 't'f h5 A b6) 33 . . . �g8 34 � h 5 A b6 (34 . . . � h7 35 gxc5 ! gxc5 36 et) x e6 gc1 + 37 � h2 (threatening 38 �e8+ �g8 39 Ag7+) 37 . . . Ac6 38 et)xd8 +-) 35 g3 ! (trying to bring the rook to the h -fi le) 35 . . . �e8 (35 . . . � h7 36 � g2) 36 't'fg4 't'f f7 (36 . . . gc7 37 gxc5 A x c5 38 et) x e6 ±) 37 � g2 et)e4 (37 . . . 't'f h7 38 't'f h 5) 38 Ag7+ �x g7 39 gh1 + � g8 40 �xe6+ � f8 41 �x b6 +-.

33 f4

Creating a mating net around the black king. 28 hxg6 hxg6 29 Axg6 fxg6 30 �xg6+ �g7 31 �e8+ �f8 =.

2S . . . fle7? Pro bably Black missed my 32 n d move, or else he was d issatisfied with the position arising after 28 . . . et)e4 29 Axe4 dxe4 30 Ag5 A xg5 31 't'fxg5, and if 31 . . . 't'fd8 32 't'fxd8+ gxd8 33 h6! ± , when , despite being a pawn down , White keeps the i n itiative due to Black's weak king and the numerous weak nesses in his position.

The black pieces are too pinned to be able to defend their king.

33 . . . flh7 Or 33 . . . A c8 34 Ag5 't'ff8 35 Af6+ ! A xf6 36 exf6 and now : A) 36 . . . 't'fg8 3 7 � h5+ ! (37 't'fxg8+ � xg8 38 et) b3 a5 ! 39 gxc5 gxc5 40 et) x c5 � f7 ±) 37 . . . gh7 (37 . . . �h7 38 �e5) 38 �e5 or 38 f7 ; B) 36 . . . a5 37 et) b5 gd7 38 't'fc2 't'fxf6 39 �xc5 and Black's position is hopeless.

116

34

fleS+ figS

Game 26

a b c d

Shirov - Gelfand , Chalkidi ki 1993 35

Af8! Hf7 36 Oxd8 ctld3 37 Hc7 Hxc7 38 016+ ! Hg7 39 ctlxe6 Black resigns

e

8

3 2 L...-_______�__.....

{f

***

Game 2 6

Al exey S h i rov - B o r i s G e l fa n d C h a l k i d i k i 1 9 93 Sicilian Defence [890] 1 e4 c5 2 ctlf3 d6 3 d4 cxd4 4 ctlxd4 ctl16 5 ctlc3 a6 6 Ae3 e5 This line has been played ten times between Alexey and myself. It is hard ly an overesti mation to say that a large part of the cur rent theory of this variation is based on our games.

7 ctlb3 Ae6 8 f4 8 f3 (or 8 'tWd2 and 9 f3) is stronger and was tested in seven of our encounters.

8 . . . exf4 9 Axf4 ctlc6

a b c d

e

5 4 3 2 � {f

L...-_______��__

10 0e2 This idea of the strong Latvian theoret ician Zigurds Lanka (who has helped Alexey for many years) looks slightly artificial , but White is trying to prevent . . . d6-d5, which is strong after 1 0 'tWd2 d 5 ! 1 1 exd5 et) xd5 12 et) xd5 'tWxd5 1 3 'tWxd5 A xd5 14 0-0-0 0-0-0 15 Ae2 g6 with a very drawish position (Yudasin - Gelfand , I nterzonal , Siel 1 993) . N ow 1 0 . . . d5 doesn't work because of 11 exd5 (11 0-0-0 Ag4) 11 . . . et) xd5 12 0-0-0 with an unpleasant pin. 10 . . . Ae7 Th is was the third time that I had this pos ition with Alexey. In our two previous games I played 10 . . . a5 11 O-O-O! (Shirov finds a strong improvement over our game in Paris, sacrificing the exchange for the in itiative ; 1 1 et) b5 Ae7 1 2 et)3d4 et) xd4 1 3 et) xd4 0-0 14 et) x e6 fxe6 15 0-0-0 'tW b6 d i d n 't bri ng White any advantage in Shirov- Gelfand, Im mopar, Paris rapid 1 992) 1 1 . . . Ag4 12 'tW b5 A xd1 (1 2 . . . a4 13 et)d4 A xd1 14 et) xc6 bxc6 1 5 'tWx c6+ et)d7 1 6 et) xd1 ) 1 3 'tWx b7 Ag4 ! (13 . . . 'tWc8 ? 14 'tWx c8+ �xc8 1 5 � xd1 ± and Wh ite has more than sufficient compensa tion for the exchange, as his bishops are strong and the a5 pawn is weak) 14 A b5

117


Ad7 15 A xc6 Ik8 ! 16 � xa5 A xc6 17 � xc6 �d7! 18 �a5 (White has won a second pawn for the exchange and it's time to change the course of events) 1 8 . . . E!.xc3 ! 19 b x c3 �x b7 20 � x b7 � x e4 21 E!.e1 d5 (Black is only a pawn down, Wh ite's queenside pawns are weak, the kn ight at e4 is excellently placed , and this gives Black sufficient com pensation) 22 �d6+ (22 c4 A b4) 22 . . . A xd6 23 Axd6 �d7 24 A b4 f5 ! 55 (Shirov-Gelfand, I nterzonal , BieI 1 993) .

a b c d 8 7

1 1 . . . �d7 1 2 0-0-0 �ce5! 1 2 . . . �de5 is not very log ical, as the c6 kn ight blocks the line of the rook.

1 3 �d5 Antici pating . . . E!.c8 x c3 , a typical positional threat in the Sicil ian Defence, Wh ite goes ahead with his plan .

1 3 . . . .txd5 14 Elxd5 14 exd5 0-0 + and I don't see a plan for White.

14 . . . 0-0 15 h4 Dc8 15 . . . �c7 !?

1 6 g4 Wh ite is fig hti n g for the f6 sq uare - it is i m portant to leave the e7 bishop out of play. From f6 it could have become an i m portant part of Black's attack (16 g3 � b6 17 E!.d1 Af6 +).

1 6 . . . 't!fc7 17 g5 �b6 18 Eld4?

8 6 5 4 3 2

5 4 3

a b c d

1 1 h3 To make it possible to castle. Despite the satisfactory resu lt of the open i n g duel, I was looki ng for something better. And one even i n g , when I was fal l i n g asleep, an idea suddenly appeared i n my head . Black should put his kn ights o n e5 and d7, followed by . . . E!.c8 w ith the threat of . . . E!.xc3 .

...

e

�

e

1

9 h

It is u nderstandable that Wh ite should try to prevent the expansion of Black's knight to a4, but he obviously missed my reply. 18 E!.d1 �a4 ! 19 E!.h3 ! was better. White prepares 20 �d4, as 20 . . . . � b6 would be met by 2 1 E!.b3. If 19 �d4 � b6, or 19 h5 b5 20 A h3 E!.ce8, while 19 A h3 E!.cd8 20 �d4 � c4 21 b3 allows the n ice tactical shot 21 . . . d5 ! , and if 22 Axc7 Aa3+ 23 � b1 �c3+ 24 � a1 A b2#. Now Black has the following options : A) t h e immed iate 1 9 . . . � c4 20 �x c4 �xc4 21 Axc4 E!.xc4 22 �d4 gives White an edge, as Black's pieces are very poorly coordi nated ; B) 1 9 . . . b5 seemed tem pting d u ri n g the game, but after 20 �d4 (20 A x e5 d xe5 21 E!.hd3 E!.fd8 =) 20 . . . E!.fe8 (20 . . . �c4 21 b3 and I was u nable to fi nd sufficient com pen sation for the piece) 21 h5 Af8 22 g6 Wh ite is the first to beg i n an attack ; C) 1 9 . . . E!.fe8 ! 20 h5 �c4 (after the inclusion of . . . E!.fe8 this is stronger) 21 �x c4 �xc4 22 A x c4 E!.xc4 23 � d4 � c5 24 �f5 (24 b3 E!.xd4 25 E!.xd4 �e6 is w i n n i n g for Black) 24 . . . �xe4 w ith an equal position.

18 . . . a5! Threaten ing . . . a4-a3.

19 Dh3 The pawn can be won only tem porari ly: 19 � x a5 E!.a8 20 � b5 (20 � b3 E!.xa2 and

118

Game 26

Shirov - Gelfand , Chalkidiki 1 993

White is helpless against the threats of . . . eDa4 and . . . EUa8) 20 . . JUc8 ! 21 c3 gxa5 ! 22 f#xa5 eDc6 23 f#c5 dxc5 24 Axc7 gxc7 + . After 19 A h3 a4 20 eDa1 gce8 2 1 a3 d5 Black stands better, and 19 � b1 a4 20 eDc1 a3 se riously damages the wh ite ki ng's defences, but 19 a3 !? a4 20 eD a1 + was worth consid ering, intending to bring the knight back into play by c2-c3 and eD c2-b4.

19

. • .

a4 20 �a1

24 . . . Axh4? wou ld have given Wh ite a strong attack on the h -fi le: 25 � h 5 ! (25 � b5 Af2 ! ; 2 5 e5 gxe5 2 6 A x b6 Ag5+ 27 Ae3 gxe3 28 gxe3 f# b6 ; 25 Axb6 f#xb6 26 f#c4+ � h8 27 �x a4 +) and if 25 . . . Af6 26 A x b6 �x b6 27 Ac4+ � h8 28 gh1 h6 29 �g6 +-. 2S .bSI �d7 After 25 . . . A x c3 ? 26 A x b6 Black loses ma terial . 26 ged3 26 ga3 eDc5 27 A x c5 d x c5 28 gxa4 eD d4 =t doesn 't solve White's problems. 26 . . . �eS 27 gxd6 �xe4 27 . . . a3 ? is sim ply a b l u nder because of 28 gxf6 ! . 28 1l6d3 Or 28 gd7 �g3 29 g7d3 eD e5 30 ga3 Ae7 31 Ac5 �x h4.

a b c d 8 7 6 5

20 gc3 ? a x b3 ! 21 gxc7 b x a2 -+ and d e spite t h e extra q ueen , nothing can stop the a2 pawn . N ow the knight at a1 is out of the game, so Black must act energetically to achieve something before it gets back.

3

..

e

8 6 5 4 3 2

20 . . . f61 21 ge3 I

th ink that Wh ite missed a good practical chance - 21 g6 ! ? , blocking the f8 rook and the e7 bishop. After 21 . . . eD xg6 22 A h2 eDe5 23 gc3 he has defi nite com pensation for the pawn.

21 . . . �e6 21 . . . f#d8!?

22 gd1 fxgS 23 .txgS gee8 Another strong possibil ity was 23 . . . Axg5+ !? 24 hxg5 gce8 25 g6!? d 5 ! (25 . . . h6 !?) 26 gx h7+ � h8 27 gf3 gxf3 28 f#xf3 d x e4 with a clear advantage.

24 .te3 .tf6

28 •eSII + The key move of the game. After the ex change of queens Black's great advantage is indisputable. As Sergey Dolmatov once ex plained to me, it is usefu l to exchange pieces when one of your opponent's pieces is out of play. He com pared it with ice -hockey, where a 5 against 4 advantage is hard to convert i nto a goal , with 4 against 3 it is easier, and with 3 against 2 it is sim ply a piece of cake. 29 1la3 29 �x e5 eD xe5 30 ga3 eDg4 + . 29 gd5 �g3 30 g5d3 transposes into the line with 28 gd7.

119

. . .


29

. . .

36 .lxe4+

•xb5 30 .lxb5

36 Ad7 ga8 37 Ae6+ � h8 38 a4 A x h4.

36 . . . g3xe4-+ Wh ite cannot maintain the material balance and the bishop is a m uch better piece than the knight.

37 h5 gh4 38 gd5 .le3 39 c!>b1 39 ge7 gh1 + 40 gd1 A b2+.

39 ... gh1 + 40 �e1 h6 Making use of the pin on the back rank, Black switches his bishop to gS.

41 a4 .lf6 42 Ded2 .lg5 43 Dd1

30 . . . �e3!

8 7 6

8 7 6 5

31 bxe3 31 !;.xc3 A xc3 32 AcS !;'fS -+ , or 31 Ac4+ � h8 32 !;'d3 bS ! -+.

31 . . . gxe3 32 gxa4 32 Axc6 bxc6 33 !;.xa4 !;'h3 +.

4 3 2

3 2

32 . . . gxe3

. 1

32 . . . !;'h3 33 tD b3 !;'xh4 34 !;.xh4 A x h 4 3 S tDcS allows Wh ite counter-chances.

9 h

33 �b3 After a long absence the knight is back i n the game, but Wh ite's weak pawns a nd the continuing possibility of an attack against his king leave him with few chances. If 33 !;'d7 !;.cS .

33 . . . �e5! A precise m ove. As often happens, the knight is excel lently placed i n the centre. 33 . . . !;'h3 allows counterplay by 34 tDcS, while after 33 . . . !;.c8 34 !;'d3 ! (34 !;.c4 ? tD b4 ! -+ ; 34 Ad 3!?) Black has only a slight advantage. 34

..

h

An im portant nuance.

ge4 gfe8 35 ge2 �e4!

43

. . .

gh2?

This over-su btle move could have created a headache for B lack. The s im ple 43 . . . gxh5 would have ensured an easy win. 44

gb5 gexe2 45 �d3 .lf6 46 gf1 ?

46 gh1 ! was an excellent chance. After 46 . . . gcg2 47 gx h2 gxh2 48 as White forces the exchange of one pair of rooks, captures the b7 pawn and pushes his a-pawn with good drawing chances.

46 . . . Da2 Now it is all over.

3S . . . gh3 36 gde1 !? with counterplay, but not 36 gxeS A xeS 37 Ad7 gcc3 -+.

*** 120

47 Dxf6 g xf6 48 Dxb7 Dxa4 49 D b8+ c!>f7 50 D b7+ c!>e8 51 Db8+ White re signs

Game 27

Gelfand - Adams, FIDE Candidates Match (4), Wijk aan Zee 1994

Game 27

B o r i s G e l fa n d - M i c h a e l A d a m s F I D E C a n d i d ates M at c h (4) , Wij k aan Zee 1 9 94 Queen 's Gambit A ccepted [020J The refined positional style and excellent calculating abil ity of M ichael Adams made him an extremely dangerous opponent right from h is yout h , and h is abi lity to concentrate and keep cool d u ring a game has become legendary. I lost to him on the tie -break in the final of the Ti lburg I nterpolis tournament in 1992 and a few weeks before our match he won the P CA Qual ifying Tournament. So, this was a match between the winners of the Qualifying (or Interzonal) Tournaments in both of the world championsh ip cycles.

1 d4 d5 2 c4 dxc4 3 e3 My opponent's first move was already a sur prise, as i n those days M ichael normally played the Ben ko Gambit or the N i mzo Indian Defence. So, in order to avoid his preparation I deviated from 3 e4 and 3 ttlf3, as I normally used to play.

tt:J

The pawn structure has been determ i ned . Wh ite has an isolated pawn and some space advantage, but he also has a big tru m p : his lig ht-sq uare bishop is clearly stronger than its black cou nterpart. Black will have to play extremely accurately if he wants to neutralise Wh ite's initiative.

Robert Hu bner, a great expert in the hand ling of these positions, preferred 8 . . . tDbd7 9 ttlc3 ttl b6 10 A b3 tDbd5 as the knight looks more sensible on d5 than on c6.

9 clilc3 Th is is a well- known theoretical position from . . . the Petroff Defence. It arises via the fol lowing move order: 1 e4 e5 2 ttlf3 ttlf6 3 ttl xe5 d6 4 ttlf3 ttl xe4 5 d4 d5 6 Ad3 Ae7 7 0-0 ttl c6 8 c4 ttlf6 9 ttl c3 0-0 1 0 h3 d x c4 1 1 A x c4. And , as also as i n the 6 th game of our match (p. 1 25) , such transpositions favoured me, as I was more famil iar with the resulting positions.

9 clila5 1 0 Ad3 Ae6 11 ge1 clilc6 12 a3 Od6 . . .

3

. . .

e5

Black usually transposes i nto the main line after 3 . . . ttlf6 4 Axc4 e6, but, lacking experi ence in this opening, my opponent narrows down White's choice.

4 Axc4 exd4 5 exd4 clilf6 6 clilf3 Ae7 7 0-0 0-0 8 h3

a b c d

e

..

h 8

7 6

6

5

5

4

4

Preventing the bishop from alighting on f4. After 1 2 . . . a6 1 3 Ae3 (13 Af4 occurred in the last, 48 th ga me o f t h e u nfi n ished Kas parov - Karpov, World Championsh ip Match, M oscow 1 984/85 - 1 3 . . . �d7 14 ttl e5 ttl x e5 1 5 d xe5 ttl d5 1 6 ttl xd5 A xd5 1 7 �c2 g6 1 8 gad1 with advantage to Wh ite) 1 3 . . . ttld5 14 �c2 h6 1 5 gad1 Ad6 1 6 A c1 ttl ce7 1 7 ttl e4 ttlf4 1 8 Af1 Wh ite stands bet ter (Ivanchuk- Rokhmanov, World U nder-20 Qual ifying Tournament, Klai peda 1985). Both of these games orig inated from the Petroff Defence.

13 Ae3 clild5 14 Oc2 h8

3 2

14 . . . h6 15 gad1 ;!;; followed by Ac1 was simi lar to the afore -mentioned little-known game of the young Vasily Ivanchuk.

121


a b c d

C) 1 7 . . . g6 1 8 d 5 ! (18 A x g6? A b3) 1 8 . . . Af5 (18 . . . A xd5 1 9 Etxd5 � x h7 20 Ac1 �a6 2 1 �d2 with a dangerous attack) 1 9 �d2 (19 �e2 � x h7 20 d xc6 �a6 21 �xa6 bxa6 22 � d4 is probably even stronger) 1 9 . . . �a5 (or 1 9 . . . � x h7 20 d x c6 b x c6 21 � d4 with a clear advantage)

e

6 5 4

5 4 3 t-=""-"'"-I-"'2

2 9 h

tr

1 5 gad1 The alternative was 15 �e4 !? �d7 (15 . . . �d8 16 � g3 ;t ) 16 �eg5 (16 � g3 f5 with coun terplay) 1 6 . . . A xg5 (16 . . . f5 1 7 � x e6 �x e6 18 Ag5 ;t ) 1 7 � xg5 h6 1 8 � xe6 fxe6 ;t .

1 5 . . . f51 Black has a very cramped position , so it is understandable that he wants to control some more squares. But the weakening of the a2-g8 d iagonal will soon become i m portant. 15 . . . � xc3 1 6 bxc3 �xa3 was more consistent.

Analysis diagram after 19 . . . tZJa5

20 Ac5 !! (vacating the route to h6 for the queen) 20 . . . �xc5 21 �h6 +-; 0) 17 . . . A b3 (this is the most com p l icated possibility) 1 8 �f5 g6 (18 . . . A x d1 1 9 � h5) 1 9 �f4 ! � x h7 20 �h6+ � g8, and now : 0 1 ) 21 �g 5 ? does not work: 2 1 . . . A xg5 22 A xg5 �d6 23 Ete4 A x d1 24 Eth4 A h5 25 g4 (25 Etxh5 g x h5 26 Af6 �xf6 27 �xf6 Etfe8) 25 . . . Etae8 ! 26 g x h5 Ete1 + 27 � g2 �d5+ -+ ; 02) hard ly sufficient is 21 Eta1 �d6 22 Af4 �f6, or 22 . . . �d7 23 � e5 �e8 ; 03) 21 Ag5 ! Axd1 (21 . . . f6 10ses to 22 �xg6+ � h8 23 Ete4 ! fxg5 24 � xg5 A xg5 25 � h5+ � g7 26 �xg5+ with mate in three moves)

I had intended playing 17 Etb1 �d6 (17 . . . �d8 18 c4) 18 Etx b7 Etab8 w ith some advantage. After the game my seconds for this matc h , Alexander H uzman a n d Valery Atlas, found the agg ressive 1 7 A x h7 ! , when the play becomes forcing and Black has several choices : A) 1 7 . . . �a2 1 8 �xa2 A xa2 1 9 Ae4 ± ; B) 1 7 . . . � b3 1 8 Et b1 ! �x c2 1 9 A x c2 and White's advantage is obvious ;

122

Analysis diagram after 21 2g51 2xd1

Game 27

tt)

Gelfand -Adams, FIDE Candidates Match (4), Wijk aan Zee 1 994

22 gxe7 ! (I think that this is stronger than 22 A x e7 !? 'ffix e7 23 f1:) g S ! (23 gxe7 A xf3 ! 24 gxc7 AdS 00 ) 2 3 . . . 'ffix gS (23 . . . 'ffix e1 + ? 24 � h2 +-) 24 'ffix gS Ac2 2S dS ! with the initiative) 22 gxe7 ! 'ffid 6 (the only move ; if 22 . . . f1:) x e7 23 Af6) 23 gd7 ! (after 23 ge4 ? Axf3 24 gh4 (24 g xf3 fS ! 2S gh4 � f7 -+) 24 . . . A hS 2S gxhS g x hS 26 Af6 'ffix f6 27 'ffix f6 gfe8 Wh ite is fighting for a d raw) 23 gd7 ! 'ffie 6 (again the only move ; if 23 . . . 'ffix d7 24 Af6) 24 dS 'ffifS 2S f1:) h4 ! (Wh ite keeps fighting for the f6 square) 25 . . .'�· eS (2S . . . ttld4 26 cxd4) 26 dxc6 (White also has winning chances after 26 f1:) xg6!? 'tWg7 27 ttl xf8 f1:) eS 28 ge7 gxf8 29 'ffix g7+ (29 gxeS f6) 29 . . . � xg7 30 gxeS ;t ) 26 . . . Ac2 27 ttlf3 ! 'ffix c3 (27 . . . 'ffig 7 28 'ffi h 4 ! (28 c x b7 gab8 29 'ffix g7+ �xg7 30 ttld4 Ae4 31 ttle6+ � g8 32 ttl xf8 � xf8 ;t ) 28 . . . AfS 29 gd4 ! fol lowed by Af6, but not 29 Af6 gS or 29 gxc7 bxc6 30 Af6 'ffi h7) 28 'ffi h 4 ! (28 gd4 is simpler) 28 . . . 'ffi h 8 (forced) 29 'ffix h8+ � x h8 30 Af6+ � g8 (30 . . . � h7 31 ttlgS+ +-) 31 gd4 g5 32 ttlxgS.

17 . . . AfS 18 Dde1 Ae8 18 . . . Ag8? 1 9 f1:) bS +- 'ffid 7 20 AxfS.

1 9 Ae4! Gaining control of the a2-g8 diagonal . Only a temporary advantage would have been achieved by 19 ttlbS ?! 'ffid7 20 ttleS ? (20 Ac4 a6 21 AxdS 'ffix dS 22 ttlxc7 ttl xd4 23 ttl xd4 'ffix d4) 20 . . . AxeS 21 dx eS a6 22 e6 'ffie7 .

1 9 . . . �xe3 1 9 . . . f1:) b6 ? 20 f1:) bS 'ffid 7 2 1 Ae6 +-, or 19 . . . ttlf4 20 Axf4 'tWxf4 21 ttldS 'ffid 6 22 ttlxf6 'ffix f6 23 dS ± and again the wh ite pieces dominate the board .

20 bxe3 ± h 6 2 1 a4?! Played with the idea of 'ffia2 and A a3 in mind . But the immediate 21 h 4 was stronger, as it renews the threat of f1:) gS. I n gen eral Black's pieces are bad ly placed , which should lead to Wh ite's attack being decisive.

21 . . . b6 22 h4! Return in g to the right pat h . 22 'ffia2 ?! f1:) aS 23 Aa3 cS does not achieve its goal .

22 . . . �a5 23 Aa2 e5

a b c d

e

8

4 3 2

A picturesque position . Black has no defence against gh4. Al l these variations, although they are not ind isputable, allow us to assess the position as better for Wh ite.

...;:;....

L...-_______

15 . . . gad 8! ;t was a reasonable alternative, when Wh ite has some advantage.

16 Ae1 gad8 17 ge2! The e6 bishop has been deprived of support, which Wh ite exploits to seize control of the e-file.

8

----I 11

_

24 �g5! Launching an attack on the king. 24 f1:)eS AxeS 2S dxeS 'tWe7 allows Black to stabilise the position .

123

24 . . . Aa6?!


Th is loses qu ickly. The alternatives were : A) 24 . . . cxd4? 25 tilf7+ E!xf7 26 A xf7 d3 27 E!e8+ � h7 28 h5! +- and if 28 . . . d x c2 29 Ag6#; B) 24 . . . hxg5? 25 h xg5 cxd4 26 �d3 ! +-;

The point of White's idea. In view of the threat of 28 A b1 Black is forced to liqu idate into an absol utely hopeless endgame.

27

. . •

•f7

27 . . . E!f6 28 E!e8+ E!f8 29 A b1 , or 27 . . . Axh4 28 A b1 A xf2+ 29 � h1 .

C) 24 . . . c4 25 tile6 Axe6 26 E!xe6 ± ;

28 .xf7 Dxf7 29 hxg5 cxd4 30 cxd4

0) 24 . . . �c7 ! (the strongest) and now : 01 ) 25 d5 c4 ! 26 til e6 A x e6 27 d x e6 E!d3 28 Ad2 ;t (28 E!e3 A x c3 29 E!1 e2 E!fd8 with cou nterplay, or 28 A a3 E!e8 29 A b4 til c6 30 Axc4 til x b4 31 cxb4 E!c3) ; 02) 25 d x c5 ! bxc5 26 til e6 A x e6 27 A x e6 A x h 4 (27 . . . f4 28 h5 ±) 28 A xf5 til c4 and White's advantage, although clear, is not de cisive.

25 De6 .d7

In view of time trouble I chose the safest con tinuation . Wh ite is two pawns up and now it is only a matter of time. Also possible was 30 gxh6!?, or 30 g6 !?, and if 30 . . . E!ff8 (30 . . . E!f5 31 E!e8+ E!f8 32 E!8e7 til c6 33 A x h 6 til x e7 34 E!xe7 E!g8 35 c xd4) 31 Ax h 6! gxh6 32 E!e7.

30

. • .

Ac4 31 De8+

31 A xc4 til xc4 32 g x h6.

a b c

Dxe8 32 Dxe8+ cc!?h7 33 Ab1 + g6 34 g x h6 �c6 35 Ae3 De7 36 Dc8 Ad5 37 Ad3 �b4 38 Ae2 Ae6 39 Dd8 �d5 40 Ag5 Dd7 41 De8 Af7 42 D b8 Ae6 43 .1f3 Df7 44 Dd8 Df5 45 Ad2 �f6 46 Da8 g5 47 Dxa7+ cc!?xh6 48 Da6 �d7 49 a5 Db5 50 axb6 Black resigns 31

8 7

3 2 L...-_______....:_ :.... �

. • •

..

'lf

26 .xf5! This simple but elegant tactical blow decides the outcome of the game.

26 . . . Axg5 Forced : if 26 . . . hxg5 27 h xg5 +-.

27 .g6! +* * *

124

Game 28

ct:J

Gelfand -Adams, FIDE Candidates Match (6), Wijk aan Zee 1994

Game 2 8

B o r i s G e lfa n d - M i ch a e l Ad a m s F I D E Can d i d ates M at c h (6) , Wij k aan Zee 1 9 94 Pirc - Ufim tse v Defence [B0 7J like this game, because on a n u m ber of occasions the course cou ld have been changed by a positional sacrifice of the ex change or the queen .

1 d4 d6 2 e4 �f6 3 f3 Black wants to play a Pirc-Ufimtsev Defence, but I prefer to go into a King's I nd ian after 3 . . . g6 4 c4.

3 . . . d5!? Black doesn't m ind losing a tem po, as the pawn on f3 doesn 't help Wh ite. 3 . . . eS 4 dS is a different story, and it occurred in another game between us in the Fontys Tou rnament, Tilburg 1 996.

4 e5 �fd7 5 f4 c5 6 �f3 �c6

6

6

5

5

4

4

3

3 2

�------�--� �

7 . . . � b6 8 ttlc3 e6 would have transposed into a French Defence, but it has never been part of Adams's repertoire. In modern chess it q u ite often happens that the fig ht i n the open ing is not for an advantage, but for the obtain in g of a position that is fami l iar to a player and less fami liar to his opponent.

8 �xd4 �xd4 9 Axd4 �b8!? The kn ight is head ing for c6, where it will attack t h e centre. 10 �c3 �c6 10 . . . e6. 11 Ab5 Trying to gain control of the dark squares. A sl ight but lasting advantage could have been ach ieved with the less ambitious 11 Af2 !? e6 1 2 Ad3 ;!;; , which was tried later in the game Sadler- Hodgson, (Bundesl iga 1 998/99).

1 1 . . . e6 1 1 . . . AfS was i nteresting . However, at some point the bishop cou ld have come under at tack by g2-g4. 1 2 a3!? 12 �d2 wou ld have led to double -edged play : 1 2 . . . �aS 1 3 a3 Ad7 14 A x c6 bxc6 1 S 0-0 cS 1 6 b4 �c7 ! (16 . . . c x b4 1 7 a x b4 �xb4 18 E{fb1 +-) 17 bxcS AxcS 18 E{f3 00 . Or 1 2 0-0 Ad7 13 fS �h4!? 00 . 12 . . . Ad7 1 2 . . . A e7 13 �d2 0-0 14 A xc6 bxc6 1S ttla4 ± and the bishop at c8 is shut in. 13 Axc6

a b c d

e

7 Ae3! N This idea was invented d u ring my prepa rations for the match together with Evgeny Agrest, who is now a Swed ish g randmas ter. Previously the weaker 7 c3 cxd4 8 cxd4 etJ b6 was played , when Black activates his light-square bishop, which would be more problematic i n the game. 7 c4 is another possibil ity.

7 . . . cxd4

125

6

6

5

5

4

4

3 2

1"""--=""' ....,--

3 2


17 . . . hS

1 3 . . . .txc6? Now Wh ite's positional advantage is un q uestioned . Strong measures were cal led for: 13 . . . bxc6 ! 14 �a4 (intending �cS) 14 . . . �aS+ 1S c3 cS 16 �xcS AxcS 17 b4

The bishop exchange favou rs Wh ite : 17 . . . AcS 1 8 .Etc3 Axd4+ 1 9 �xd4 ± .

1 8 a4! Starting to exert pressure agai nst a new weakness - the b6 pawn . 18 .Eth3 !?

18 . . . .tcS 18 . . . as 19 �f2 or 18 . . . Ae7 19 as !? would not have solved Black's problems.

1 9 Dc3 .te7 Bad is 1 9 . . . A b4 ? 20 .Etxc6 ! �xc6 21 �xb4.

20 b4! 1!fb7 20 . . . A x b4 21 .Etxc6 �xc6 22 �x b4 +-. 1 7 . . . A xd4 ! ! (17 . . . A x b4 1 8 ax b4 ± is hard ly sufficient) 18 bxaS Axc3+ 1 9 �f2 O-O ! (Black needs his dark-square bishop to d isturb the opponent's king ; 19 . . . A xa1 20 �xa1 ± ; 1 9 . . . AxaS !?) 20 .Etc1 AxaS followed by . . . f7f6. Black has just two bishops and a pawn for the queen , but Wh ite's king is misplaced and may come under attack.

21 gb3 .td7 22 �e3 gc8 22 . . . a6 allows 23 fS ! .

8 7

8 7 6

14 0-0 Threatening f4-fS .

3 2

14 . . . 96 1 S 1!fd2 ± 1!fc7 1 6 D13 b6 a

b

c

d

e

4 3 2

f _ ..... '-----------""-

�

23 as?!

a

b

c

d

e

f

9 h

�

17 �d1 ! White regroups his pieces. The place for his kn ight is on e3 , while h i s rooks will aim for the queenside.

White has rearranged his pieces and it is time to begin a breakthrough . However, con fronted with a wide range of possibil ities, I failed to choose the best contin uation. I was unable to decide on the best move after 23 fS ! AgS (bad is 23 . . . g x fS 24 � xfS ! exf5 2S e6 .Etg8 26 exd7+ �xd7 27 .Etf1 +-) 24 fxg6 (24 f6 !? deserves attention, as it ham pers Black even more) 24 . . .fxg6 2S lWd1 � f7 ±, although the wh ite position is very menacing. 23 �d3!? is also possible, maintaining both threats, 24 fS and 24 as.

126

Game 28

Gelfand -Adams, FIDE Candidates Match (6), Wijk aan Zee 1994

ttJ

23 . . . .tb5! 23 . . . bS 24 fS ! +-

24 axb6 a6!

8 7

7 6 A} 28 . . . � e8 29 g4 ! (29 c4 ? d xc4 30 tLl xc4 A x c4 31 tf xc4 tf x b6 ! =) 29 . . . Ad8 (29 . . . h4 30 fS) 30 g x hS g x h S 31 c!> f2 +- followed by gg1-g8 ;

4 3 2 ---""

L...-_______

--' 'If

_ _

I

overlooked this reply. At the cost of a pawn , Black stabilises the position on the q ueen side, and if he should manage to win it back (the pawn on b6 is hard to defend) , his position will be defensible. After 24 . . . a x b6 25 gba3 White seizes the a-file.

25 flc3 �d7?

a b c d

B} 28 . . . gxcS !! (a fine resource) 29 bxcS tfc6 30 c!> h1 (Wh ite is u nsuccessful with 30 b7 c!> c7 ! (30 . . . A xcS ? 31 b8tLl +) , or 30 c4 A xc4 31 b7 c!> c7 32 tf b2 A bS} 30 . . . AxcS 31 tfd2 tfxb6 ;t with chances of defend ing success fu l ly.

26 flc5! Having created an outpost on cS, White can open up the game with help of his c -pawn.

26 . . . flxc5 e

8 7 5 I-�::' 4 I--�""",-"=-F''''"","",,,,� 3 2

After 26 . . . AxcS both 27 bxcS !? and 27 AxcS are strong .

8 7 6

27 bxc5 27 A x cS ! +- fol lowed by 28 c4 was even stronger.

4 3 2

L--_______--=-_---I

27 . . . flc8 28 c4 dxc4 28 . . . A xc4 29 tLl x c4 d x c4 30 Af2+ c!> e8 (30 . . . tfdS 31 tfc2 c!> e8 32 tfa4+ +-) 31 tf b4 and White should win easily. 'If

29 flc1 �e8 30 �xc4 Od5

Too frivolous. Black should have calculated a l engt hy variation : 2S . . . gxc3 ! 26 tfxc3 c!> d7 (not, of cou rse, 26 . . . 0-0 27 g4 ! with a mat ing attack) 27 AcS (27 tfd2 gc8 and if 28 c4 dxc4 with counterplay) 27 . . . gc8 (27 . . . g b8 28 �d4 Ad8 29 c4 Ac6 30 bS +-) 28 tfd4 and now :

30 . . . gxcS 31 tLl d6+ A xd6 32 A xcS A x cS+ 33 gxcS tf x b6 was hardly enough for sur vival .

31 �a5 31 tLl d6+ A xd6 32 exd6 seems obvious, but the position after 32 . . . Ac6 ± fol lowed by . . . a6-aS was not completely clear to me.

127


8 1_"'-'-7

8 7 6

3 2

31 . . . Acs 31 . . . A xc5 32 gxc5 gxc5 33 b7 +-, or 31 . . . gd8 32 b7 Ac6 33 �b2 Ax b7 34 � x b7 �xd4+ 35 �xd4 gxd4 36 c6 +-. From the practical point of view 31 . . . gxc5 !? was stronger: A) 32 b7 ? looks to be easily winning. H ow ever, Black has an incred i ble defensive re source - 32 . . . gxc1 + 33 �xc1 �xd4+ 34 c;!? h1 ..

36 �xg2 gxd1 + 37 Ag1 A xg2+ 38 c;!? xg2 g b1 39 A a7 Ad 8 ! 40 b8 � gx b8 41 A xb8 Axa5) 35 . . . �xc6 36 gb1 gb8 37 �b2 +- fol lowed by 38 A a7 ; B2) 32 . . . A a4 !? 33 b7 gb5 34 gc1 A b4 35 � b2 !? (probably even stronger than 35 �f2 A x a5 36 gc8+ c;!? d7 37 b8 f# gx b8 38 gx b8 with good winning chances) 35 . . . c;!? d7 (35 . . . A x a5 36 b8 � + Ad8 37 � 2 x b5+ �x b5 38 �d6 +-) 36 � c6 Ad6 (36 . . . �x c6 37 gxc6 c;!? x c6 38 �a1 c;!? xb7 39 �x a4) 37 b8� + ! ? (or 37 �a1 ) 37 . . . c;!? e8 (37 . . . gx b8 38 � x b8+ A x b8 39 Ae3) 38 exd6! +- gx b2 39 d7+ c;!? f8 40 d8 � + �xd8 41 � xd8.

32 .te3 Oe4 33 �xcS OxcS 34 h3?1 34 �e2 fol lowed by � f3 was s impler (the same manoeuvre was also possible a little later), but, being slightly short of time, I over looked this possibil ity. 34 . . . 0b5

35 Oc2 OcS 3S �h2

36 �e2 ! .

3S . . . a5 37 ga1 37 �e2 ! . 37 . . . gaS 38 ga4 h4 39 Oa2 Ob5 40 Oc4 OcS 41 Oa2 Ob5

a b c d 8 X 7 34 . . . � b6 !! 35 �c8+ A d8 36 b8 � (36 � c6 A xc6 37 b8 � �x b8 38 �x b8 as ±) 36 . . . �xa5 00 with good chances of surviving. Despite his extra q ueen , White h as no tar gets to attack. An analogy with the position exami ned i n the com ment on Black's 1 3 t h move inevitably comes to mi nd ! ;

e

f

9

h

8 7 6 5 4

�------�--� �

42 cSI

B) 32 gd1 ! i s stronger: B 1 ) 32 . . . gc8 (passive) 33 b7 gd8 34 c;!? h1 Ac6 (34 . . . A a4 35 A b6 A xd1 (35 . . . �xd2 36 gxd2) 36 �xd5 exd5 37 Axd8) 35 � xc6 (avoiding the trap 35 A b6 ?? �xg2+ !

By exchanging one of his passed pawns, White activates his pieces and this, together with the b-pawn , decides the outcome of the game.

128

Game 29

42 •xe6 43 Hxa5 Hb8 44 Ha7 .e4 44 . . . Ad8 45 �a3 ! , and Wh ite controls the a3-f8 d iagonal .

a b c d

. . .

45

.b3 �8 46 Hd7

46 ... g5 47 Hd4! Now Black's kingside is weakened as wel l , cutting short his resistance. . . .

e

9 h 8 7 6 5 4

8 7 6 5 4 3 2

Activating the rook and threatening b6-b7.

47

CD

Gelfand - Topalov, Dos Hermanas 1 994

•f5

2 a b c d

47 . . . �b7 48 .E!c4 g xf4 49 .E!xf4 +-.

e

'lJ

9 h

48 He4! Hb7 51 Ad4! .e4 52 .f3 Black resigns

48 . . . gxf4 49 A xf4 +-.

49 He8+ �g7 50 He7 g xf4 50 . �e4 51 .E!x b7 gxf4 52 Af2 (52 .E!xe7 fxe3 is far from clear) 52 . . . �x b7 53 �f3. . .

In view of 52 . . . �xd4 53 .E!x b7 (53 �x b7 ? ! �x e5 !) 5 3 . . . Ac5 5 4 �g4+ a n d mate, Black resigned .

***

Game 2 9

B o r i s G e l fa n d - Ve s e l i n To p a l ov Dos H e r m a n as 1 9 94 Du tch Defence [A 89]

8

Although I have a negative score with Veselin, I have won number of games against him as wel l , and of these this is the one that I like most. It enabled me to take the lead in this prestigious event, which I was able to maintain right to the end. Vesel i n was already a dangerous opponent and it was clear that he was going to keep on improving. And for already ten years now he has almost per manently been ranked among the best 5- 1 0 players i n the world. H e never hesitates to play as riski ly as possible, an � this is excel lently combi ned with a deep understanding of both dynamic and strateg ic positions.

1 d4 f5 2 g3 �f6 3 Ag2 g6 4 e4 Ag7 5 �e3 0-0 6 �f3 d6 7 0-0 �e6 8 d5 �a5

7 6 5 4 3

6 5 4 3 2 a b c d

e

9 h

'lJ

M y opponent has chosen the sharpest l i n e i n t h e m a i n system o f t h e Len ingrad Vari ation . That is Veselin - he always goes for the sharpest possibilities. Here Black's plans are similar to the Yugoslav Variation of the King's I ndian Defence, but the position of the pawn on f5 (compared to f7) is in White's favour, in my opin ion , as it seriously weakens Black's position (the plus, on other han d , is control of the e4 sq uare) . However, as Tigran Pet rosian once joked : ' If you r opponent wants

129


to play the Dutch, you should never prevent it ! ' .

9 .a4 cS 10 dxc6 Th is trail was first blazed in the classic game Keres - Korchnoi .

10 . . . bxc6 1 0 . . . � xc6 1 1 gd1 � a5 1 2 c5 Ad7 1 3 f#a3 � e8 14 Ag5 Ae6 15 cxd6 � xd6 1 6 gxd6 f#xd6 1 7 f#xa5 with a decisive material ad vantage (Keres - Korchnoi , 20 th USSR Cham pionsh ip, Moscow 1 952).

1 1 cS!

After 1 2 . . . f#c7 Wh ite has a pleasant choice between 1 3 Af4 � h5 (13 . . . g5 14 A xg5! ±) 14 � xg6 � xf4 1 5 � xf4 .E!b8 1 6 gab1 , and 13 � c4 � x c4 (13 . . . � b7 14 Af4 (14 f#xc6 f#x c6 1 5 A x c6 ±) 14 . . . � d6 1 5 � xd6 exd6 1 6 f#x c6 f#x c6 1 7 A x c6 g b8 1 8 A xd6 +-) 14 f#xc4+ � h8 15 f#xc5 ± .

1 3 .c2 Threatening 14 �a4. 1 3 Ae3 ! also deserved attention, but here too Black has counterplay: A) 1 3 . . . � g4 14 � xg4 fx g4 1 5 � e4 ± leaves Black with too many weaknesses ; B) 1 3 . . . A e6 14 gac1 ! (the threat is 1 5 �d3 �d7 16 b4) 14 . . . f#c7 15 �xc6 �xc6 16 Axc6 gac8 (16 . . . gab8 1 7 Af4) 1 7 Af3 ;t ; C) 1 3 . . . gb8! 14 � c4 � x c4 1 5 f#x c4+ � h8 1 6 A x c5 A a6 ! (16 . . . f#a6 1 7 f#x a6 A xa6 1 8 gab1 g b7 1 9 gfd1 ± ; Black has weak pawns at a7, c6 and e7) 1 7 A x b6 A xc4 18 A xa7 gxb2 .

13

I think that this is an important novelty, intro duced over the board . It is thought that the top grandmasters have everything analysed at home, but nowadays you cannot man age to do this even if you use all the search engi nes. Chess is sti l l an unexplored gam e ! Wh ite i s trying to isolate the kn ight at a5 and perpetuate Black's pawn weaknesses. After this game the popu larity of 8 . . . � a5 rapidly waned . Previously the main l ine was 1 1 �d4 Ad7 12 �xc6 �xc6 13 Axc6 Axc6 14 f#xc6 gc8 1 5 f#a4 f#d7 1 6 b3 d5 ! , regaining the c-pawn .

•c7 14 �d3 After 14 Af4 g5 15 � xc6 g xf4 16 � xa5 gb8 Black has got rid of his bad knight and has good cou nterplay, despite his poor pawn structure. . . .

14 . . . c4 14 . . . � d7 15 A d2 ! (15 � a4 c4 16 �dc5 � xc5 1 7 � x c5 gb8 +) 15 . . . c4 (15 . . . �c4 ? 1 6 �d5 ! +-) 1 6 �f4 �f6 1 7 �a4 ± .

1 1 . . . dxcS After 1 1 . . . d5 1 2 �e5 Black cannot mai ntain the material balance, e. g . 1 2 . . . �e4 13 �xc6 Ad7 14 � xe7+ ! � h8 15 c6.

1 2 �eS .b6

130

1 S �cS a

b

c

d

e

,..

f

8 7 6 5 4 3

6 5 4 3 2

Game 29

Gelfand -Topalov, Dos Hermanas 1 994

17 �3a4

15 . . . IlbS?! Topalov misses a chance to get rid of the principal weakness i n his position - the knight at as - by 1 S . . . � b7 ! (1S . . . � e8 !? 16 Ad2 � d6 1 7 gad1 � was another pos sibility; if 1S . . . gd8 16 �3a4) and now: A) 16 � x b7 �x b7 ! (16 . . . A x b7 ? ! 17 �a4 � ) 17 'tWa4 �a6 ! (Black should keep his c4 pawn, as it pins down the b2 pawn and c1 bishop) 18 �xa6 (18 A xc6 g b8 ; 18 �xc6 �b8) 18 . . . Axa6 1 9 Af4 �e4 ! 20 Axe4 Axc3 ! (with the opposite -colour bishops, Black should be able to hold the position , as White can not take control of the open files ; if 20 . . . fxe4 21 gac1 � ) 21 A xc6 (21 b x c3 fxe4 =) 21 . . . Ax b2 22 !'!ab1 c3 23 AdS+ � g7 (23 . . . e6 24 A xa8 !'!xa8) 24 AeS+ (24 A xa8 �xa8 2S Ac1 !'! b8 26 A x b2 c x b2 27 !'!fd1 A x e2 28 gd2 � ) 24 . . . � h6 2S A xa8 !'!xa8 26 !'!x b2 c x b2 27 A x b2 with a draw ;

17 !'!ad1 !?

17 . . . �d5 1S Ilad1 ± It turns out that Black's play has come to a dead end ; meanwhi le the weaknesses in his position persist and the knight at as remains out of action .

1S . . . e5 Creating fresh weaknesses, but what better course is there? In reply to 18 . . . � b4 Wh ite has the strong move 19 'tWc1 ! ± and the pawn is u ntouchable: 19 . . . � x a2 ?? 20 A x aS + �xaS 21 �xc4+ � h8 22 'tWxa2 .

1g e4 1 9 AxaS? 'tWxaS 20 'tWxc4 e4 00 wou ld regain the pawn but lose Wh ite his advantage. 19 . . . �b4 1 9 . . .fxe4 20 Axe4 ± .

a b c d

B) 16 �3a4 � d6 (here the knight is m uch better placed and Wh ite can count on on ly a smal l edge) 17 Ad2 (17 Af4 ?! does not work in view of 1 7 . . . � dS ! , and if 18 A xd6? exd6 19 'tWxc4 d x cS 20 A xdS+ cxdS 21 �xdS+ � h8 22 �x a8 Ad7 and Black is w i nni ng) 17 . . . �dS 18 �c3 (18 !'!ac1 !,!b8; 18 e4 is pre matu re before Black has played . . . e7-eS 18 . . . fxe4 1 9 � xe4 AfS 00) 1 8 . . . !'! b8 (18 . . . e6 19 �3a4 ; Wh ite doesn 't hesitate to waste two tempi, as now the e-pawn blocks one of the bishops - on e6 the Ac8 and on eS the Ag7) 19 !'!fd1 and I prefer Wh ite.

16 Ad2 White com pletes his development and tar gets the knight at as.

16 . . . IldS?! Black persists with his mistaken plan . In reply to 16 . . . � dS Wh ite should mai ntain course by 17 !'!ad1 and not regain the pawn, as 17 �xdS cxdS 18 A xdS+ � h8 1 9 � a4 !'!d8 gives Black good piece play. 16 . . . �e8 !? intending . . . � d6 would have been more tenacious.

e

8

8 7 6 5 4 3

6 5 4 3 2 a b c d

e

'i1

20 Axb4!? Wh ite exchanges his opponent's active pieces, thus increasing his advantage. 20 �c1 is not as strong here as in the note to Black's 1 8 th move, as Wh ite has al ready played e2-e4 and weakened the d4 sq uare : 20 . . . !'!d4 ! , and noth ing is achieved by 21 exfS gx fS 22 A x b4 !'!x b4 23 !'!xd4 exd4 24 'tWe1 !'!b8 2S �e8+ Af8 .

20 . . . Ilxb4 21 IlxdS+ .xdS 22 Ild1 .fS 22 . . . 'tWc7 23 A h3 ! fxe4 24 Ae6+ and wins.

131

23 .d2


After the game my opponent suggested 23 exfS !? A xfS 24 �d2, but I was reluctant to activate the black pieces.

23 . . . IlbS 24 "dS .tf6 24 . . . tD b7 2S �c7 c3 (2S . . . tD xcS 26 .§d8 tD x a4 27 �x c8), tryi ng to get counterplay on the b -file: 26 bxc3 tD x cS 27 .§d8 .§ b1 + 2 8 Af1 A a6 2 9 tD xcS A xf1 3 0 h 4 fxe4 31 � h2.

2S "c7 .te7

a b c d 8 6

26 . . . tD b7 27 tDd7 A xd7 28 '§xd7.

27 exfS gxfS 2S "xeS+ cc!;Ig6 29 "c71 a6 The knight at cS cannot be touched: 29 . . . A xcS 30 tD x cS .§xcS 31 .§d6+, or 29 . . . .§xcS 30 � xcS AxcS 31 �xaS (31 '§d8).

30 �d7 30 Ag2 may have been even better.

30 . . . .txd7 31 "xd7 "f6 31 . . . .§eS was more stu bborn , taking control of the e -file.

e

8

32 Ile1 Ilb7 32 . . . .§eS 33 .§xeS �xeS 34 A xfS+.

6 5

33 "cS cc!;Ig7 34 .txfS "d4 3S .tb1

a b c d

e

8 7

L-_______---=:__....J

f

•

9 h 8 6 5 4

11

2S . . . f4 ! ? was a better practical chance, demanding accurate play on Wh ite's part : 26 g xf4 (26 �xa7 Ad8) and now : A) 26 . . . A h4 27 fS ! g x fS 28 � h1 ! tD b7 29 tDxb7 Axb7 (29 . . . '§x b7 30 �xc6) 30 exfS �c8 31 �d6;

8) 26 . . . exf4 27 eS Ae7 28 tDe4 ! (the key move!) 28 . . . Ag4 (28 . . . f3 29 tDf6+ � h8 30 A xf3 +-) 29 .§d2 (29 f3 !? AfS (29 . . . .§dS 30 .§xdS cxdS 31 tDf6+ A xf6 32 exf6 +-) 30 tDf6+ � h8 31 tD c3 .§ b6 is not so clear) 29 . . . f3 (29 . . . .§ b7 30 �x aS A b4 31 �a6 +-) 30 tDf6+ A xf6 31 exf6 .§dS 32 .§xdS cxdS 33 �x aS fxg2 34 �xdS+ and Wh ite should win.

A s impler cou rse was 3S Ac2 ! �d2 36 �g4+ +-, or 3S Axh7 3S . . . �xh7 36 �f5+, but the text move does not spoil anything.

3S . . . .tb4 36 Ilf1 Ilf7 37 "h3 h6 38 "h5 Ilf6 39 �c3 .txc3 40 bxc3 "xc3 41 I1d1 In terms of material the two sides are equal, but White's attack is irresistible.

41 . . . "f3 42 Ild7+ �S 43 IldS+ cc!;Ig7 44 Ild7+ �S 45 "cS+ cc!;IgS 46 .th7+1 cc!;IhS 47 "a7 IlfS

26 .th31 +E l i m i nating any attem pt at counterplay by Black. Another possibil ity was 26 tDd7 Axd7 27 '§xd7 .§b7 28 �xeS ± .

The alternatives 47 . . . tD b7 48 Ae4 ! and 47 . . . cS 48 .§d8+ .§f8 49 .§xf8+ �xf8 SO Ae4 requ ire no explanation .

26 . . . cc!;Ig7

4S .tc2 Black resigns

132

Game 30

Gelfand - Kramnik, FIDE Candidates Quarter-Final Match (4), Sanghi Nagar 1994

a b c d

The fi nal position is h igh ly eloquent : the knight at a5 never did join i n the action . If 48 . . . �c3 49 gd8 ! .

e

8

ttJ

8

5 4 3

�------�--� �

*** Game 30

7 . . . cxd4 8 exd4 £e7

B o r i s G e l fa n d - V l a d i m i r K ra m n i k FIDE Candidates Quarter-Fi nal Match (4) , Sanghi Nagar 1 994 Queen 's Gambit [04 OJ My match with the future world champion Vladimir Kramn i k is one of the h i g h l ig hts of my career. Although only 19 years old, Vladimir was already rated among the top five players in the world and it was clear that further successes were bound to follow. A friend of mine, M ark Kogan , invited me to prepare at his hotel in Karelia (in the north of Russia) at the time of the white nights. There, together with my seconds Alexander H uz man and Valery Atlas, I made a deep study of Vladimir's games. And we discovered that Vladimir's main weakness at that time was the defence of unfami l iar and slightly inferior positions. So we decided to avoid theoret ical battles in the Slav Defence and to aim for less well-stud ied positions after 1 c4.

1 c4 c5 2 �f3 �c6 3 �c3 �f6 4 e3 e6 5 d4 d5 6 a3 a6 7 b3 Played with the same idea - of avoiding the oretical d iscussions. 7 d xc5 A x c5 8 b4 is the main line here, described in the books as being slightly more promising for White.

5 4

�------��- �

9 c5!? Th is appeared to be a novelty, but i n my opinion it is the most chal lenging move, gain ing space on the queenside. The alternative was 9 Ae2 .

9 . . . b6 9 . . . 0-0 10 b4.

10 cxb6 U nfortunately, White is not wel l enough de veloped to maintain the pawn on c5 : 1 0 b4 bxc5 1 1 b x c5 (11 d x c5 e5 12 Cba4 Ag4 00) 11 . . . Cbe4 1 2 Ad2 (1 2 �c2 �a5 13 Ad2 Cb xd2 14 �xd2 0-0 15 Ae2 gd8 16 0-0 e5 =) 1 2 . . . Af6 with counterplay.

133


The struggle now becomes very concrete if Black should manage to play . . . a6-a5 fol lowed by . . . A a6 and get rid of his light square bishop, he will be fi ne, so Wh ite must do his best to prevent this.

move. Better was 14 . . . �f6 (14 . . . Aa6 15 �c7) 15 ik1 Ad7 16 �e2 gfc8 17 gc2 �e4 18 gfc1 Af6 1 9 �e3 ;!; .

1 5 �c7 llb8

a b c d

10 . . . �d7

8

10 . . . �xb6 11 �a4 ;!; was more log ical , for the knight has to return to f6 anyway.

e 8

7 6

1 1 J.d3

5

11 �a4 � x b6 1 2 � x b6 �x b6 13 b4 0-0 14 Ae2 Ad7 1 5 0-0 �a7 16 � e5 A b5 17 A x b5 �x b5 is an instructive l i ne, where Black is fine.

4 3 2

1 1 . . . a5 L..-_______----"__......

11 . . . 0-0 1 2 0-0 a5 !?

12 �b5!

11

16 b4!

The most forcefu l move. There is l ittle Wh ite can hope for after 1 2 0-0 A a6, or 1 2 �a4 � x b6 13 �c2 (13 A b5 Ad7 14 �c5 � xd4 !) 1 3 . . . � x a4 14 �xc6+ Ad7 1 5 �c2 gc8 16 �d2 �c3 with counterplay.

1 2 . . . t!!fx b6 1 3 J.f4 0-0 14 O-O! As Black is not yet ready to play . . . Aa6, it is time to castle. If 14 Ac7 , then 1 4 . . . � b7 1 5 �e2 �f6, and if 16 �e5 � x e5 (or 16 . . . � a7 ! 17 gc1 � x b5 18 Ax b5 A xa3) 1 7 dxe5 �e8 . 14 gc1 A a6 ! (14 . . . �f6 1 5 Ac7 � b7 1 6 �e5 �xe5 17 Axe5 ;!; ) 15 �c7 is weaker because of 15 . . . Axa3 ! 16 � xa8 gxa8, and now: A) 17 Ax h7+ ? x h7 18 �c2+ (18 �g5+ g6 19 �c2+ f6 20 �h7+ e7 21 �xc6 A b4+) 1 8 . . . g8 1 9 �xc6 A xc1 , and if 20 �x a8+ h7 21 Axc1 � b5 ; B) 1 7 g b1 A b4+ 1 8 f1 (18 e2 � xd4+ 19 � xd4 �xd4) 1 8 . . . � xd4 19 Ae3 A xd3+ 20 � xd3 � b5 and Black stands better.

14 . . . �a7? In this critical position Black makes an un fortunate attem pt to force matters . It is probable that Vlad i m i r mi ssed Wh ite's 1 6 th

Wh ite has to play energetically to exploit the poor coordination of the black pieces and to avoid his knight at c7 bei ng trapped . 16 �c2 �x b3 1 7 A x h7+ h8 was unclear.

1 6 . . . J.b7 Several other moves were possible, but they too would not have solved Black's problems: A) 16 . . . �f6 17 bxa5 �xa5 18 � xd5 +-; B) 1 6 . . . g b7 1 7 �a8 !? (I enjoy using the corner sq uare ; I also did so in my game against U l i b i n i n 1 985 - see below p. 1 38) 1 7 . . . �d8 18 �c2 axb4 (18 . . . �f6 1 9 Ac7 �e8 20 bxa5 or 20 Axa5 +-) 1 9 ax b4 � b5 20 Ax h7+ h8 21 Ad3 ± ; C) 1 6 . . . Ad8 1 7 �c2 g b7 (1 7 . . . e5 1 8 � xd5 �d6 19 � xe5 �xd5 20 � xd7 A b7 21 Ax h7+ h8 22 Ae4 +-) 18 gfc1 e5 1 9 �a8 +-; D) 16 . . . axb4 1 7 ax b4 Ax b4 18 �a6 ± ; E) the most reasonable defence was 1 6 . . . Ad6 1 7 A xd6 �xd6 1 8 �c2 A b7 (18 . . . h6 1 9 b x a5 A b7 20 � b5 ±) 1 9 A x h7+ h8 20 Ad3 gfc8 2 1 gfc1 a x b4 (21 . . . Aa8 22 A a6 A b7 23 �g5 +-) 22 a x b4 � c6 23 � b5 �x b4 24 gab1 �e7 25 �d2 ± .

134

1 7 bxa5!

Game 30


The most natural way to play. By tak ing the q ueen away from the e6 sq uare, Wh ite secures an enduring advantage. An interesting alternative, suggested by Kram nik, was 17 �a4 !? Ac6 18 �c2 (18 �x a5 �xa5 1 9 b x a5 g b3 with fi ne cou nterplay) 1 8 . . . gbc8 ! (18 . . . mc8 1 9 gfc1 ax b4 20 axb4 A b5 21 Ax h7+ � h8 22 Ad3 ±), and now :

A) 1 9 gfb1 ?! a x b4 (1 9 . . . A b5 20 A x h7+ � h8 00) 20 ax b4 A b5 21 Ax h7+ � h8 22 gxa7 �xa7 23 ttl x b5 gxc2 24 ttl xa7 � x h7 ; 8) 1 9 gfc1 ! Ad8 ( 1 9 . . . e 5 20 ttl x e5 gxc7 2 1 b5 ±) 20 ttla6 t .

1 7 . . . 0xa5 1 8 Oe2 It is hard to defend agai nst the threat of etl xe6.

ttl x e5 (22 . . . gf6 23 ttlf7+ gxf7 24 �xf7 +-) 23 �xe7 ttled7 ± .

19 Axe5? I fai led to notice the simple 19 ttl x e5 ! �x c7 (1 9 . . . ttlf6 2 0 ttl d7 ttl xd7 2 1 �x e7) 20 gac1 +-, as pointed out by Kramnik, and after the forced 20 . . . �d6 21 ttl xd7 �xf4 22 �x e7 me8 23 � b4 Wh ite wins on the spot. One may wonder about the sou rce of such mistakes at high level. I have only one explanation - the very intense nervous pres sure.

19 . . . �xe5 20 Oxe5 Now Wh ite is a pawn u p , but he sti l l has to convert it into a win.

20 . . . �c6 21 Of4 Ilbc8 22 �b5 Aa6 23 a4 �b4 24 Af5

18 . . . e5?

a b c d 8

7

e

Faced with the prospect of mutual time pres sure, I decided to return the extra pawn i n order t o gain time t o concentrate m y pieces for an attack on the black ki ng. The sim ple 24 �d2 ± was also possible, retaining a considerable superiority.

f

��--��'��-�

8

6 5 4 3 2 L..-_______-=-__.....

ttJ

24 . . J �cd8 25 Hfe1 ? Simply a waste of a tempo. By contrast, the i m m ediate 25 ttle5 would have confronted Black with insurmountable problems: 25 . . . h6 (25 . . . A x b5 26 a x b5 �x b5 27 ga7 ! +- Af6 28 ttld7)

'lf

Simply a blunder. The alternatives were :

A) 18 . . . Ad8 1 9 ttl x e6 ge8 20 A x b8 ttl x b8 (20 . . . gxe6 21 Ae5) 21 �b2 +-; 8) 18 . . . Af6 1 9 ttl xe6 gbe8 20 �b2 +-; C) 18 . . . g bd8 1 9 ttl x e6 fxe6 20 �x e6+ gf7 (20 . . . � h8 21 Ad2) 21 A x h7+ � f8 22 ttlg5 Axg5 23 Ad6+ +-;

D) 18 . . . ttl c6 1 9 gfc1 e5 20 ttl x e5 ttld x e5 2 1 Axe5 gbc8 22 ttl b5 ;

Analysis diagram after 25 ttJe5 h6

E) 18 . . . � h8 (Black's best chance) 1 9 ttl xe6 ! fxe6 20 A x b8 ttl x b8 21 �xe6 ttlac6 (21 . . . Af6 22 � h3 g6 23 Axg6 �c7 ±) 22 ttle5

26 A b1 ! (threaten i ng �f5) 26 . . . Ac8 (26 . . . � b6 27 ttld7 ! +- gxd7 28 �f5 ; 26 . . . A x b5 27 a x b5 �x a1 (27 . . . � b6 28 ttld7 gxd7

135


29 �f5) 28 �f5 or 28 A h7+ +-) 27 ga3 (the rook joins the attack) 27 . . . Ag5 28 �g3 +-.

a b c d

e 8

25 . . . .tf6 26 �e5 6

With the same idea of attacking. White could have transposed i nto a good ending by 26 �c7 �xc7 27 fiJ xc7.

5 4

26 . . . .txe5 2

26 . . . Ax b5 was the alternative : A) interposing 27 fiJg4 is not strong , for sev eral reasons: A1 ) 27 . . . A xd4 28 �xd4 Ad7! (28 . . . Ac6 29 fiJf6+ � h8 30 fiJ x h7) 29 A xd7 (29 fiJe3 fiJc6) 29 . . . fiJc2 (29 . . . gxd7 30 fiJf6+ ! +-) 30 fiJf6+ gxf6 31 �g4+ � h8 32 �f5 (32 Af5 fiJ xa1 ) 32 . . . fiJ xe1 33 �xf6+ = ; A2) 2 7 . . . � b6 !? 2 8 a x b5 (28 fiJ xf6+ �xf6 29 a x b5 fiJ d3 +) 28 . . . �xd4 (28 . . . A xd4 29 ga4 g6 30 A b1 (30 gxb4 Ac3) 30 . . . Ac3) 29 fiJ xf6+ �xf6 30 Axh7+ �xh7 31 ffxb4 d4 with counterplay;

Allowing a decisive piece sacrifice, but salvation was no longer possi ble: 29 . . . h6 30 gg3 � h8 31 gh3 +-, or 29 . . . �b6 30 gb3 +-.

30 .txh7+ ! �xh7 31 .f5+ �g8 32 Hh3 IUe8 The only move : 32 . . . g6 33 �f6 +-, or 32 . . . �c2 33 �h5 +-.

a b c d

B) 27 a x b5 �x b5 28 ga3 g6 (28 . . . A xe5 is careless - 29 A x h7+ � x h7 30 gh3+ � g8 31 �h4 +-) 29 gh3 Ag7 30 � h4 h5 (30 . . . h6 31 fiJ xf7 +- gde8 32 fiJ x h6+ Axh6 33 Ae6+) 31 gb1 ± (31 fiJ xf7 gde8) maintaining the bet ter chances.

27 . . . .txb5 28 axb5 .xb5 29 Ha3! The most vigorous move. 29 � h4 h 6 ! ;t (29 . . . g6 30 e6 !) ; 29 gab1 !? d4 (29 . . . gb8 !? 30 gb3) 30 �d2 gb8 31 �xd4 ±.

•c4

8

7

7

6

6 -'--

4

5 4

3 2

27 gxe5 ?! would have forfeited White's su periority after 27 . . . Ax b5 28 Ax h7+ (28 ga3? h6 or 28 . . . fiJ d3 29 A xd3 A xd3 30 gxd3 �xa4 =) 28 . . . � x h7 29 gh5+ � g8 30 �h4 f6 31 ge1 (31 gh8+ �f7 32 �h5+ � e7) 31 . . . � f7 32 gh7 (32 gh6 gg8 ! -+) 32 . . . Ad3 33 �g4!! A x h7 34 �e6+ =.

. . .

f

8

5

27 dxe5

29

e

33 .h7+? I n extreme time pressure, Wh ite almost throws away the fru its of his efforts. After 33 � h 5 ! � f8 (33 . . . f6 34 � h7+ � f7 35 gg3 � e6 36 �xg7 d4 37 �xf6+ � d5 38 ga1 +-) 34 �g5 � g8 (34 . . . f6 35 gh8+ � f7 36 �h5+ g6 37 e6+ �e7 38 �h7+ � d6 39 gxe8 gxe8 40 �d7+ +-) 35 gg3 ! g6 36 gh3 �c6 37 �h4! (37 �h6? �f6 !!) 37 . . . � f8 38 e6 ! (38 �x b4+ � g7 ±) Black would have had no defence (38 . . . gxe6 39 �xd8+).

136

Game 30


33 . . . �8 34 "h8+ �e7 35 "xg7 d4?

a b c

37 . . . lld7?

d

In the time -shortage lottery, Vlad i m i r d rew the far from best ticket. Nor would his problems have been solved by 37 . . . ga8 ? ! 38 gh6+ � c7 3 9 �g3+ � b7 40 �d6 � b5 41 gb1 .

8 7

7

6

6

5

5

4

4

H owever, 37 . . . gc8 ! , as pointed out by my opponent immediately after the game, would have led after 38 �e5+ to extreme compli cations:

3 2

The last ten moves before the time control were made instantly, and it is hardly sur prising that Kramn i k missed the opportunity 35 . . . �f4 !? 36 e6 (36 gf3 ? �xf3 37 gxf3 gg8) 36 . . . � d6 37 gf3 �d2 (37 . . . �xf3 38 �e5+) 38 �e5+ � c6 39 gxf7 (39 gc3+ � b5 40 gec1 ? (40 gf1 00 is better) is refuted by 40 . . . gc8 ! 41 gxc8 gxc8) 39 . . . gd6 40 �a1 ! with a complete mess on the board .

36 e6 36 gh6 � d7 37 e6+ � c8 38 �xf7 was also possible, but Black is out of danger here. 36 . . . �d6 37 e7

A) 38 . . . � c6 ? 39 gh6+ � b7 40 gd6 ! d3 (40 . . . gc5 !?) 41 �a5 (41 gd4 d2 42 �e4+ is simpler) 41 . . . �c5 42 gd7+ �c6 (42 . . . � b8 43 �x c5 gxc5 44 gd8+ gc8 45 gxe8 gxe8 46 g b1 � a7 47 gx b4 d2 48 gb1 gc8 49 gf1 +-) 43 �a4+ � b6 (43 . . . � b5 44 gc1 + +-) 44 �a7+ � b5 45 � b7+ � b6 (45 . . . � c4 46 gc1 + tb c2 47 �e4+ +-) 46 gd5+ tD xd5 47 �xd5+ +-; B) 38 . . . � d7 ! 39 �f5+ � c7 (39 . . . � c6 40 gh6+ � b7 41 �d7+ +-) 40 �a5+ � d7 (40 . . . � b7 41 gh5 +-) 41 gh6 (perhaps White does best to settle for a d raw by 41 �f5+) 41 . . . �c5 ! 42 �a4+ tbc6 =+= and the attack is over.

38 Oe5+ �c6 39 Ilh6+ �b7 40 Oa5! + Ildxe7

The last critical moment.

a b c

d

e

f

..

9 h

ttJ

There was no escape from mate.

8

41 1lxe7+ 6 5 4

It is never too late to make a mistake: 41 g b6+ ? � c8 42 �a8+ � d7 43 �a4+ � c8 =.

41 ... llxe7 42 "b6+ Black resigns 2

9 h

Despite the many m utual mistakes, which can be explai ned (but not excused !) by the high stakes, this was a fascinating encounter.

***

137

Most Memorable Games

Piece in the Corner (cf. note to Black's 16 t h move)

Th is was one of the best games of my ch ild hood . Wh ite wants to play h4-h5, but at the moment this is not possible.

Boris Gelfand - Mikhail Ulibin USSR U nder- 1 8 Championsh ip, Yurmala 1 985

21 0h1 !! So, despite the fact that everyone tel ls you pieces should be placed i n the centre, I found this move and I was very proud of it! The move is a m u lti-purpose one. White vacates the f3 square for his knight and in tends h4-h5-h6 (if immed iately 21 h5 Ag4).

8

6 5

5

4

4

3

_.r---;

21 . . . Elh5?! 21 . . . Ag4 22 �f3 t .

22 .ig5!

3

Now the rook is trapped .

2

'---_a ______ e__....; 9_ ;; _--'

2 2 . . . 0c5 23 c3 �e5 24 .ie2 Elxg5 25 h xg5 .id6 26 Elad1 �g4 27 .ixg4 .ixg4 28 �b3 Oc7 29 Oxd5 Eld8 30 Oxd6 Black resigns

If

* * *

Monaco 200 1 : Boris 's father Abram watches the start of the Gelfand- Kramnik game.

138

Game 31

Shirov - Gelfand , Dos Hermanas 1995

Game 31

a b c d

8 1.

Al exey S h i rov - B o r i s G e l fa n d Dos H e r m a n as 1 9 95 Sicilian Defence [896J

e

7 6 5

This was another classic Sicil ian clash be tween attack and defence against Alexey Shirov. M ost of our encounters have been fascinati ng, as the reader may have noticed . Both of us have succeeded many times and Alexey presented a number of our games in his book. It is clear that Shirov is the succes sor to the magician from Riga, M i khail Tal , and he was strongly influenced by joint anal yses with this g reat player during the early stages of his career. Li ke Tal , Shirov aims for the in itiative at any price with both colours, and you have to be u p to the mark i n order to com pete with him.

1 e4 c5 2 �f3 d6 3 d4 cxd4 4 �xd4 �f6 5 �c3 a6 6 Ag5 Th is came as the fi rst surprise for me. Nor mally Alexey plays 6 Ae3 . I guess he was in fluenced by his second for this tournament, Alexander Shabalov, who is also an ex- Riga player.

6 . . . e6 7 f4 �bd7 Not a very popular variation , but I use it quite often . 7 . . . � b6 is the choice of most players, includ ing Garry Kasparov.

8 1!!fe2 Played with a concrete idea. 8 �f3 is the main l i ne.

8 . . . 1!!fc7 9 0-0-0 b5 10 g3 10 g4 is more often played .

10 . . . b4

----

4

5 4

3

d _e__....:9_ :: h_....J {( L...-_a__b_c__ 1 1 �d5 Th is is the point of the 8 �e2 plan .

1 1 . . . exd5 1 2 Ag2!? N A new and origi nal idea. I had previously faced 12 exd5+ Ae7 1 3 CDf5 (13 CDc6 CD b8 14 Axf6 CD xc6 15 Axg7 gg8 16 dxc6 gxg7 +) 13 . . . CD c5 ! (coveri ng e6 ; if 13 . . . CD b6 14 CDxg7+ � d8 15 ge1 with the threat of 16 Axf6 A xf6 1 7 �e8+ gxe8 1 8 gxe8+ � d7 1 9 A h3#) 14 CD xg7+ � d8 1 5 ge1 ? (the wrong move order; 15 Ag2 gg8 16 gde1 was correct) 1 5 . . . gg8 ? (15 . . . CDfe4 ! 1 6 Ag2 f5 ! 1 7 CD h5 A xg5 1 8 fxg5 b3 ! 1 9 a x b3 CD x b3+ 20 � d1 CD d4 ! -+) 1 6 Ag2 CDfe4 1 7 A x e4 (17 A x e7+ �x e7 1 8 �c4 gxg7 1 9 A x e4 g b8 20 �d4 �f8 21 �f6+ �c7 22 ge3 +) 17 . . . gxg7 + and I agreed a draw in order to g uarantee my first qual ification to the Candidates matches (yudasi n - Gelfand, Manila I nterzonal 1 990).

1 2 . . . Ae7 13 �f5 1 3 exd5 O-O ! 14 �x e7 ge8, trapping the queen for insufficient com pensation .

Com pared to my game with Yudasin, there is pawn tension i n the centre and that is why the knight has to control the d5 square. 1 3 . . . h6!? also deserved serious considera tion .

14 �xg7+

(see next diagram)

139


a

b

c

d

e

..

f

C2) 1 8 gde1 ! Axg5 (18 . . . f6 !? 19 f5 !?) 19 fxg5 'tlVc6 !? (vacating the c7 sq uare for the king ; after 1 9 . . . � c8 20 'tlV f6 � b7 2 1 ge7 Ad7 22 'tlVd4 gac8 23 A xd5+ ttl xd5 24 'tlVxd5+ � b8 Wh ite has some com pensation for the kn ight, but I wou ld prefer to play this position with Black) 20 'tlVf6+ � c7 21 ge7+ Ad7 and White hardly has sufficient compensation.

8 7 6 5

5

4

4

3

3

1s Ah6

2

a

14

. . .

b

c

The attack comes to an end after 15 ttJ h 5 ttl x h 5 1 6 'tlVx h5 A xg5 1 7 fxg5 Ae6 ( 1 7 . . . dxe4 ? 1 8 'tlV h6+ � g8 1 9 gxd6) 1 8 exd5 gc8 1 9 gd2 Af5.

d

c!>f8

The king goes to the opposite side, com pared with the Yudasin game. 14 . . . � d8!? was also possible and was probably stronger. After 1 5 e5 Ag4 Wh ite faces a choice : A) 16 exf6? Axe2 17 fxe7+ �d7 18 A h3+ (18 gde1 Ad3) 18 . . . � c6 1 9 gde1 Af3 -+ ; B) 1 6 Af3 A xf3 1 7 'tlVxf3 ttl e8 (1 7 . . . gc8 !? 18 'tlVe2 dxe5 19 fxe5 ttJe8 +) 18 ttJf5 (18 exd6 ttl xd6 +) 18 . . . A xg5 19 fxg5 gc8 (19 . . . d xe5 20 gxd5+ � c8) 20 gd2 'tlVc4 21 � b1 'tlVe4 ; C) 16 W#f2 !? (clearly strongest) 16 . . . ttJe8 ! (es tablishing coordination between the black pieces ; if 16 . . . d x e5 17 fxe5 A xd1 1 8 gxd1 with a dangerous attack, despite being a rook down , or 1 6 . . . A xd1 ? 1 7 exf6) 1 7 exd6 (17 Axe7+ �xe7 + , or 17 ttJ xe8 gxe8 18 Axd5 A xd1 19 gxd1 ( 1 9 exd6 'tlVxd6 20 A x a8 A h5 21 Af3 Axf3 22 W#xf3 �c8 +) 1 9 . . . ttJ xd5 20 gxd5 A xg5 21 gxd6+ � e7 22 fxg5 � f8 -+) 17 . . . ttJ xd6 (or 1 7 . . . 'tlVxd6 1 8 h3 Ad7 ! (the bishop is excel lently placed here ; 18 . . . Axd1 1 9 ttJf5) 1 9 Axe7+ � xe7 2 0 ghe1 + � f8 21 ttJ xe8 gxe8 +), and now: C 1 ) 1 8 A xd5?! is i nsufficient : 1 8 . . . A xd1 1 9 gxd1 ttl xd5 20 g xd5 h6 2 1 A xe7+ (21 gxd6+ 'tlV xd6 22 ttlf5 'tlVd5 23 A x e7+ � d7) 21 . . . � xe7 22 'tlVe3+ (22 'tlVd4 ghc8 -+) 22 . . . � f8 23 'tlVe5 (23 'tlVd4 gg8) 23 . . . gc8 24 ttle6+ fxe6 25 'tlVx h8+ � e7 + ;

1S

c!>g8 Now the � g8 and gh8 are dubiously placed , and this g ives Wh ite a strong attack. But if 15 . . . 'tlVc4 ? 16 ttl h5+ � e8 1 7 'tlVx c4 d xc4 (17 . . . ttl xc4 1 8 ttJ xf6+ A xf6 1 9 e5) 1 8 ttl xf6+ Axf6 19 e5 or 19 gxd6!? with a clear advan tage. . . .

1 6 .tlhS 16 e5 was a weaker alternative : A) 16 . . . d xe5 1 7 fxe5 , when Black has : A1 ) 1 7 . . . ttlg4 (this is what I i ntended playing d u ring the game) 1 8 Af4 'tlVc4 ! 1 9 'tlVxc4 d xc4, and now : A 11 ) 20 ttJ h5 ttJf2 21 Axa8 ttJ xd1 (21 . . . ttJ xa8 22 A h 6 ! (sh utting in the king) 22 . . . ttJ xd1 23 gxd1 Ag4 24 ttJf6+ Axf6 25 gd4, winning) 22 gxd1 (22 Af3 ttJf2 23 gf1 ttJ g4 (23 . . . ttJ h3 24 A h6 Ag5+ 25 A xg5 ttl xg5 26 gd1 ! with cou nterplay) 24 A xg4 A xg4 25 ttJf6+ A xf6 26 exf6 h5 and Black keeps his extra piece) 22 . . . Ag4! (22 . . . ttl x a8 23 A h6 ttJ c7 24 ttJf6+ Axf6 25 exf6 +-) and Black is out of danger; A1 2) 20 ttl e8 !? ttlf2 (20 . . . ga7 21 ttlf6+ � g7 22 h3 ttl xf6 (22 . . . ttJf2 23 Ae3) 23 exf6+ A xf6 24 Ae3 , restori ng the material bal ance thanks to the pin , or 20 . . . A b7 21 ttJf6+ (21 Ax b7 gxe8 22 Ac6 gc8 23 A b7 gb8 -+) 2 1 . . . A xf6 22 A x b7 ttl x e5 23 gd6 with ad vantage to Wh ite) 2 1 A x a8 ttl x a8 22 A h6 ttl xd1 23 gxd1 f5 ! 24 exf6 � f7 ! and Black retains winning chances ;

140

Game 31


A2) 17 . . . Ag4! (this is stronger) 18 Af3 �k8 ! 19 Axg4 ( 1 9 exf6 ? A xf3 20 \Wxf3 \Wx c2#) 19 . . . ttl xg4 20 ttlfS ttl x h 6 21 ttl x h6+ f8 22 b1 ttl c4 and I th i nk that White's attack will not ach ieve its goal ; 8) 16 . . . tDg4 17 tDe8 !? (17 exd6 \Wxd6 18 tDe8 �d8) 17 . . . \Wc6 18 exd6 \Wxe8 (it is important not to blunder with 18 . . . Axd6??

19 e51? A fine move, precisely in Shirov's style ! Wh ite creates n umerous tactical possibilities for his pieces. If 1 9 exdS ttl ge3 20 gde1 Ag4. However, he also had avai lable the strange looking move 19 gde1 ! ttlf2 ( 1 9 . . . ttlce3 20 Af3 d x e4 21 A xe4 dS 22 Af3, or 19 . . . ttl ge3 20 Af3) 20 exdS ttl x h1 21 gxe7 Ag4 22 A x h 8 x h 8 23 ttlf6 ttlf2 with fu l l compensation for the piece.

19

• . .

�ge3 20 exd6 �xd6

20 . . . Ag4!? was my original intention :

19 \Wxg4+ ! A x g4 20 ttlf6#) 19 AgS AfS 20 dxe7 gc8 and Black wins. 16 . . . �g4 Trying to destroy the u n pleasant 'cage' around the black king. Bad was 16 . . . ttl x e4 17 gxd S ! ttl xdS (17 . . . fS 1 8 gxfS A xfS 19 Axe4 A x e4 20 \Wx e4) 1 8 A x e4 +-, or 16 . . . dxe4 1 7 Axe4 ttl xe4 18 \Wxe4. 17 Ag7 .c4 After the exchange of queens the mating threats should not be so dangerous. 18 .xc4 18 gd3 \Wx a2 1 9 b3 ttlc4 ! ( 1 9 . . . ttl eS is also strong) 20 exdS (20 bxc4 dxc4) 20 . . . ttla3 -+. 18 . . . �xc4 8

.i.

8

7 6 5

5

4

4

3

3 2

if

A) i n the event of 2 1 d x e7 A x h S 22 A x h 8 A x d1 23 Ad4 Black can h o l d b y 23 . . . ge8 ! (23 . . . tD xg2 24 gxd1 ge8 2S AcS ± ; 23 . . . Ag4 24 ge1 ge8 2S A x e3 +-) 24 gxd1 ttl x d1 2 S AcS (2S xd1 gxeT += ) 2 S . . . tDde3 2 6 A h3 fS 27 b3 ttla3 28 Axe3 gxe7 with an unclear ending ; B) but then I found 21 A xdS!

(a rare instance, where nearly all the pieces of both(!) sides are under attack!) 21 . . . AxhS (21 . . . A xd1 22 A x a8 A x h S 23 A x h 8 A xd6 24 Ad4 ;t) 22 Ax h8 (22 Axa8 xg7 23 dxe7 A x d1 24 Ac6 Af3 ! 2S A xf3 ge8 -+) 22 . . . tD xdS 23 gxdS Af3 (23 . . . tD xd6 24 gxhS xh8 2S ge1 ;t) and now : B 1 ) 2 4 d x e7 A xdS 2S ge1 (2S g d 1 ttle3 26 ge1 x h 8 27 gxe3 ge8 28 geS +) 2S . . . ge8 (2S . . . Ae6 26 gd1 ) 26 Ad4 (26 Af6 tDd6 27 gd1 tDe4) 26 . . . fS 27 AcS with coun terplay;

141


B2) 24 �d3 !? Axd6 (24 . . . Ax h1 25 dxe7 �e8 26 Af6) 25 �xf3 Ci!? x h 8 and normally in the endgame rook and two pawns are slightly better than bishop and knight.

�c7) 24 . . . � de3 when White has insufficient compensation for the piece.

23 Axh8 23 �xd5 � x g7 24 �xe7 Ae6 ! (24 . . . Ci!? f8 !) 25 �g5 (25 �a7 A xd5) 25 . . . Ci!? f8 -+. 23 . . . �fe3

8 .1

A serious alternative was 23 . . . � de3 24 Af6 (24 �d3 Ci!? x h 8 25 g4 Ae6 26 g xf5 � xf5) 24 . . . � xd1 25 g4 (25 A x e7 � x e7 26 �xe7 Ci!? f8 27 �e5) 25 . . . A b7 26 g xf5 Axf6 27 � xf6+ Ci!? g7 28 � h5+ Ci!? h6 29 �f6 and everywhere the advantage is on Black's side, but it is not easy to convert it into a win.

7

�------�--�

24 �f6+ 24 Af6 � xd1 25 A xe7 Ag4.

�

24 . . . Axf6 25 Hxd5!

21 Hhe1 !!

The poi nt of Wh ite's play!

Played in fi ne style ! Instead of protecting his attacked pieces, Alexey attacks the oppo nent's! Bad is 21 Axh8 Ci!?xh8 22 �de1 � xg2 23 �xe7 Ae6 + , or 21 �de1 � df5 22 A x h8 � xg2 23 �xe7 � xe7 24 Ae5 � c6 + .

25 . . . �xh8 25 . . . � x d5 ?? would have led to a mate of rare beauty: 26 �e8#.

2 1 . . . �df5!? Black was already short of time (it is easy to guess that this position was far from easy to defend) , but this time my i ntuition d i d n 't betray me. I n the event of 21 . . . Ag4 22 �xe3 (22 �xd5? � xd5 23 A xd5 �e8) 22 . . . A xd1 23 Ci!? xd1 (23 �xe7 A x h 5 24 A x h8 Ci!? xh 8 25 �e5 Ag4 26 �xd5 �d8 with good winning chances) 23 . . . �f5 24 �xe7 � xe7 25 Af6 Wh ite has fu ll compensation for the two ex changes.

22 Axd5 22 A x h8 � xg2 23 �xe7 � x e7 24 Ae5 (24 Af6 Ag4) 24 . . . � c6 and the advantage remains in Black's hands.

22 . . . �xd5?! I didn't notice the simple 22 . . . � xd1 23 Axh8 (23 A xa8 � xg7 24 � x g7 Ci!? x g7 (24 . . . Af6 ? 25 � h5 A x b2+ 26 Ci!? b1 !) 25 �xe7 �d8 +) 23 . . . �a7 24 Ae5 (24 �xd1 Ci!? x h 8 25 A xf7

26 Hxe3 Ae6 27 Hd6 �g7? With the flag on his clock hang i n g , Black commits a big inaccuracy, but White fails to exploit it ! Correct was 27 . . . h5! (28 f5 ? Ag5), fixi ng the kingside pawns and securing the important f6 and f5 squares for the bishops.

28 a4? Stronger was 28 g4 ! h6 29 b3 a5 00 , when the bishops are restricted .

28 . . . h5 Correcting the mistake.

142

29 a5 Af5 30 c3 bxc3

Game 31


30 . . Jk8 !? was more precise, limiting Wh ite's possibilities.

a

c

d

e

f

9

h 8

31 bxe3 geS 32 cc!?d2?! 32 � b2 !?

32 . . . ge5?!

7

7

6

6

5 4

Another time trouble m ove. Stronger was 32 . . . g b8 !? 33 gb6 gd8+ 34 � c1 (34 � e2 Ag4+) 34 . . . Ad3 =+= . 33

b

�r-�....--' -

3

----'1_

4 3 2

gxa6 gb5 34 gaS?

Why give up a pawn ? 34 h4 ! would have led to a draw. 34 . . . gb2+

35 cc!?e1 gxh2

'--_______---"-__......

{f

43

ggS+ 43 c5 Ad4 44 ge7 A xc5 -+. 43 . . . cc!?xgS 44 geS+ cc!?g7 45 as'O J.g4+ 46 cc!?e1 ?

35 . . . gc2 ?! 36 � d1 A xc3 37 ge2 .

36 a6 ga2 Black has good winning chances here, as the wh ite pawns are not th reatening to queen and are liable to fal l .

37 a7 ga1 + Gaining time on the clock i n order to reach the time -control (37 . . . A h3 !? ; 37 . . . � h7 !?).

38 cc!?d2 ga2+ 39 cc!?e1 cc!?h7 40 e4 The only chance (40 � d1 A h3 =+=).

40 . . . ga1 + 40 . . . Ad4 41 ge7 � g7 came into considera tion :

A) 42 � d1 � f6 ; B) it appears that Wh ite can not save the game by 42 g b8 A x a7 43 gbb7 Ad4 44 gxf7+ � g6, and if 45 gxf5 (45 gfe7 gg2) 45 . . . � xf5 46 gb5+ � e4 47 gxh5 � d3 48 gb5 gg2 ;

C) 42 gg8+ � xg8 43 ge8+ � g7 ; 0) 42 gd8 ! A xa7 43 ge5 with counterplay.

41 cc!?d2 ga2+ 42 cc!?d1 cc!?g7 42 . . . Ad4 !? 43 ge7 � g7 44 gg8+ � xg8 45 ge8+ �g7 46 a8 � Ag4+ -+ with an extra tem po com pared with the game (47 � e1 ? Af2+ 48 � f1 A h3+) .

Wh ite clearly gains more d rawing chances by 46 � e1 !? Ac3+ 47 � f1 Ae2+ 48 � g2 gxa8 49 gxa8 Axc4 =+= when the winning plan wou ld be to put the bishop on g4, followed by . . .f7-f5 and the penetration of the king.

46 . . . gxaS 47 gxaS J.d4 4S cc!?d2 J.12! Cutting off the king from the g3 pawn.

49 ga3 cc!?t6 50 gd3 cc!?e6 50 . . . Ae6 !? followed by . . . � f5-g4 was strong. White's pieces are too poorly coordi nated to prevent this. 51 cc!?e3 J.e2 52 gdS cc!?e7 53 Iid5 16 54 cc!?d2 J.g4 54 . . . A xc4?! 55 gx h5 A xg3 56 � e3 =t . 55 gd3 cc!?e6 Wh ite is in zugzwang .

56 ga3 56 � c3 Ae2 -+. 56 . . . cc!?d7 Not so clear is 56 . . . � d6 57 ga6+ � c5 58 gxf6 �xc4 59 gg6 (59 gd6 Axg3 60 � e3 h4 61 gg6 h3 62 gxg4 h2 -+) 59 . . . � d4 !? (59 . . . A xg3 60 � e3) 60 f5 ! � e5 (60 . . . A xf5 61 gf6) 61 ga6 � xf5 62 ga3 � e4 63 gb3.

57 ga6 57 c5 !? was another chance:

143


A) 57 . . . A xc5 58 ga6 f5 (58 . . . Af2 ? 59 gxf6 A xg3 60 � e3 � e7 61 ga6 h4 62 ga7+ =) 59 � e1 � c7 -+ is similar to the note to White's 46 th move; B) 57 . . . � c6 58 ga6+ (58 gc3 f5 and Wh ite is in zugzwang : 59 � c2 (59 � d3 A xg3) 59 . . . Axc5 (59 . . . Ad4 60 gc4 A xc5) 60 � d2 � d5 -+) 58 . . . � xc5 59 gxf6 � d4 and this also should be won , but Black must be care fu l not to allow the exchange of the rook for his light-square bishop.

that the position was not as simple as it ap peared .

62

. . •

�!

It took me an hour and a half of home analy sis to fi nd a precise win. It was easy to fall into a trap with the naturaI 62 . . . �d7? 63 gc1 �c6 64 � e4 Ae2 (64 . . . Af2 65 gh1 Ag1 66 f5 A h3 67 � f3 A xf5 68 � g3) 65 � e3 A h5.

57 . . . .txg3 58 �xf6 �e7?! Simpler was the immediate 58 . . . h4! 59 � e3 h3 60 gh6 h2 61 f5 � e7 (61 . . . Axf5? 62 � f3 Ae5 63 gh4) 62 f6+ � f7 63 c5 Ad1 64 c6 A a4.

59 �a6 h4 I d i d n 't want to have to demonstrate my technique after 59 . . . Axf4+ 60 � e1 , although theory says that this position is won .

60 �e3 h3 61 c5 h2 62 �a1 62 gh6 � f7, or 62 ga7+ Ad7. a

b

c

d

It looks as though Wh ite is i n zugzwang, but h e is saved by 66 ga1 ! = (66 � e4 ? Af2 67 gh1 Ag1 68 f5 Af7 -+) and if 66 . . . � xc5 67 ga5+. That is why it is necessary to lure the pawn to c6 or c7 and block it there.

'"

e

8

8

7

7

6

6

5

5

4

4

3

3

2

2

63 �b1 Black is winning after 63 � e4 � g6 64 c6 � h5 65 c7 (65 g b1 � h4 66 gh1 � h3 67 f5 � g2 68 gx h2+ A x h2 69 f6 A e6 70 � d4 � f3 71 � c5 Ac7 72 � b5 � e4 73 � a6 �d5 74 � b7 � d6) 65 . . . � h4 66 � e3 � h3 67 gc1 � g2 68 gc2+ � f1 -+ . The main line was 63 c6 � e7 64 gc1 � d8 65 c7+ � c8 66 � e4 Ae2 67 � e3 A h5 and now the zugzwang is real : 68 � e4 (68 f5 Axc7) 68 . . . Af2 69 gh1 Ag1 70 f5 Ae8 -+.

This was the sealed move in the last ad journed game I ever played . It turned out

***

144

63 . . . �5 64 c6 .txf4+ 65 �2 .th3 66 �3 .tc7 67 �b5+? � 68 �b1 .te6 69 �d1 White resigns

Game 32

Game 32

with the 1 1 d5 line and shared some ideas with me, and we spent several hours to gether, analysing this position . 1 1 . . . �f6 1 2 bxc4 b4 1 3 .txf6 "xf6 14 "a4+ �d7 15 �d4 e5 16 �b3 rtJe7

B o r i s G e l fa n d - J e r o e n P i ket M ax E u we M e m o r i a l To u r n a m e n t , VS B A m sterd a m 1 9 96 Sla v Defence [03 1J

a

I was a last-m i n ute replacement i n the final V S B Tournament and I real ly enjoyed playing in this perfectly organ ised event in one of my favourite cities. Although I made a poor result, this game was a kind of consolation. My opponent, one of my contem poraries, was (u nfortunately, he quit chess a few years ago) a player of classical style. However, he had an extremely broad open ing repertoire. It therefore made no sense to prepare deeply for him, as it was hard to guess what open ing you wou ld have to face.

1 d4 d5 2 c4 c6 3 �c3 e6 4 �f3 dxc4 Jeroen chooses a sharp variation that is named after his com patriot Dan iel Note boom , a talented player from the early part of the century, who sad ly d ied very youn g . Black grabs t h e pawn and tries t o keep it.

5 a4 .tb4 6 e3 b5 7 .td2 a5 8 axb5 .txc3 9 .txc3 cxb5 10 b3 .tb7 a

b

c

tt:J

Gelfand - Piket, Max Euwe Memorial Tournament, VSB Amsterdam 1996

d

e 8

6 5 4

L--_______----:;.__.....

1}

11 d5 Whereas the preced ing moves occurred i n the overwhel m i ng majority o f games, this move is a rare choice, compared to 1 1 bxc4 b4 1 2 A b2 ttlf6. But my good friend Yuri Shulman had played a n u m ber of games

b

c

d

e

7 6 5 4

6 �""'-::-

5 4

L--_______----:;.__.....I

1}

17 .te2 A new trend. I n the 1 980s 1 7 � b5 used to be played , but after 17 . . . A a6 18 �x a5 ghb8 1 9 d6+ � e8 Black has the advantage. White's idea is not to force events, but to try to exploit three long -term factors : his pow erfu l pawn centre, the position of the black king in the centre, and the bad bishop on b7 . N atu rally, B lack too has his tru m ps, i n par ticular the a5-b4 pawn pai r. Wh ite can only hope that, as in this game, his opponent will be short of one tempo . . . 17 . . . Hhc8 Black also has another possible continuation 17 . . . �d6, for exam ple: 18 f4 ! ghc8 19 0-0 ttl c5 (1 9 . . . � f8 20 ttl x a5 ttl b6 21 ttl x b7 !) 20 ttl x c5 gxc5 2 1 gad1 ! e4 22 �a1 � f8 23 �e5 �x e5 (23 . . . �e7 24 � h5 � g8 25 d6 +-) 24 fxe5 and Wh ite's pawns proved to be the stronger (Shu lman - San Segundo, Pamplona 1 996) . 18 Hd1 Another, more concrete possibility was cho sen by White in Shulman -V. Ivanov (Moscow 1995) : 18 Ag4 �d6 19 � xa5 ! gxa5 20 �xa5 ga8 2 1 �x a8 A x a8 22 gxa8 �g6 with an unclear game. 18 . . . �c5

145


An alternative is 18 . . . �d6. Now only a draw resu lts from 1 9 0-0 (evidently White should choose 19 f4 !? i n analogy with the Shul man - San Seg u ndo game, with a position that is hard to evaluate) 19 . . . ttl b6 20 �a1 (20 � b5 A a6 21 c5 .;!xc5 22 �xc5 A xe2 +) 20 . . . ttl xc4 (20 . . . ttld7!? 2 1 f4 a4 22 ttl d2 00 ; 2 0 . . . .;!xc4 !?) 2 1 A xc4 .;!xc4 2 2 ttl x a5 .;!c3 (or 22 . . . .;!c5 23 ttl x b7 .;!xa1 24 .;!xa1 �xd5 25 ttlxc5 �xc5 26 ';!fb1 �b5 27 ';!b3 � d6 =) 23 ttl x b7 .;!xa1 24 ttl xd6 ';!xd1 25 ttlf5+ � f6 (25 . . . � d7? 26 ';!xd1 b3 27 ttlg3 ! b2 28 ttle2 .;!a3 29 ';!b1 .;!a1 30 � c3 .;!a3 31 ';!x b2 .;!xc3 32 � f1 ±) 26 ';!xd1 � xf5 = .

19 �xc5 1 9 d6+ � f8 20 � b5 is not good , if only be cause of 20 . . . Ac6 21 �xc5 Axg2 +.

19 ... l1xc5 20 0-0 8 7 6

6 5 4

3

3

2

2

The key poi nt of the game. Wh ite is ready to beg i n an attack (�c2 , f2-f4 etc.) and on Black's next few moves will depend whether he is able to do anything to counter White's plan . I now believe that he should be able to do so.

C) 20 . . . ';!d8 2 1 f4 exf4 22 ';!xf4 �e5 , and if 23 ';!df1 �x e3+ 24 � h1 f6 25 �c2 Axd5 26 �x h7 A x g2+ 27 � xg2 �x e2+ 28 � h1 � d7 (28 . . . ';!d1 29 .;!e4+) 29 �x g7+ �c8 when Black is doing fine; D) 20 . . . .;!e8 2 1 f4 exf4 (21 . . . e4 22 Ag4 �f8 23 Ad7 ';!d8 24 A b5 00) 22 ';!xf4 �e5 23 gdf1 �x e3+ 24 � h1 f6 (24 . . . � d6 25 .;!xf7) 25 A h 5 ! ? (25 �c2 A x d5) 25 . . . ';!d8 26 ';!4f2 with an u nclear game; E) 20 . . . �f8 ! (I think that this is the strongest) 21 f4 (21 �c2 � g8 22 f4 exf4 23 ';!xf4 'tWe5) 21 . . . ';!d8 (21 . . . e4 22 ';!d4 .;!e8 00) 22 'tWa1 ge8 23 'tWb1 �g8 24 Ad3 h6 25 fxe5 'tWxe5 26 gf3 (26 A h7+ � h8 (26 . . . � f8 27 �g6 00) 27 ';!xf7 Aa6 ! 28 ';!df1 .;!cc8 00) 26 . . . Aa6 and I saved this game only by luck (Gelfand-Vaisser, Eu ropean Rapid Championsh i p , Cap d 'Agde 1 996). 21 0c2! Wh ite first i m proves the position of his queen . If 21 f4 exf4 22 �c2 00 (22 ';!xf4 'tWe5 23 ';!df1 ';!f6). 21 . . . a4 B lack does n 't have time to protect his h7 pawn : 2 1 . . . g6 22 f4 Ac8 23 � b2 ! exf4 24 �d4 �xd4 25 exd4 .;!c7 26 c5 ± , or 21 . . . �g6 22 � b2 . 22 14 Clearly, Wh ite's first aim is the black king, and not the h7 pawn . I believe that Wh ite's attack is decisive. 22 . . . b3 a 8

b

�r-::-

c

d

e

f

9

h 8

7 6

20 . . . l1a6?

5

In my opinion, the rook is not well placed here. Let us consider Black's other options : A) 20 . . . �d6 21 �c2 !? (21 f4 transposes i nto the Shulman -San Segundo game examined earl ier; p. 145) ; B) 20 . . . � d6 21 f4 with the i nitiative ;

4 3 2

L...-._______----=-__.....

146

tr

Game 32

ttJ

Gelfand - Piket, Max Euwe Memorial Tournament, VSB Amsterdam 1 996

23 'fJe4!

27 'fJb3!

An important interposition. The black king is drawn away from the f7 pawn. If 23 �x h7 a3 00 , or 23 �c3 e4 ! (23 . . . exf4 24 d6+ (24 �b4 � b6 25 d6+ � d8 26 �c3 gg5) 24 . . . gxd6 25 � b4 gc7 26 gxf4) 24 � b4 (24 d6+ gxd6 25 �b4 gc7) 24 . . . �b6 25 d6+ � d8 26 �c3 f6 00 .

The best solutio n . After 27 e4 Black would gain good counterplay: A) 27 . . . � d7 28 e5 � b6 29 � h1 (29 � h3+ � c7 30 d6+ � b8 31 d7 �d8 32 � b3 +- is also strong) 29 . . . a2 30 e6+ � e7 31 � h7 ! ...

23 . . . �d6 24 'fJxh7 24 fxe5+ �xe5 25 �x e5+ � x e5 26 gxf7 Ac8 leads nowhere, as the black pawns are too close to the queening square.

24 . . . e4 Jeroen tries to stem Wh ite's attack at the cost of another pawn (of what sign ificance would this be, if another black q ueen were shortly to appear at a1 I). The weakness at f7 tel ls in the variations 24 . . . b2 25 fxe5+ �x e5 26 gxf7 �x e3+ 27 � h1 +- and 24 . . . exf4 25 gxf4 �e5 26 gxf7 �xe3+ 27 � h1 +-. 24 . . . �c7 !?, suggested by Hans B6hm, gives White a pleasant choice between 25 fxe5 �xe5 26 gxf7+ � b8 27 �d3! gd6 (27 . . . gb6 28 d6 ; 27 . . . b2 28 �a3) 28 �d2 ± , and 25 �e4 � d6 26 fxe5+ �x e5 27 �d3 f6 28 gf5 �e7 29 e4 +-.

31 . . . A xd5 (31 . . . a1 � 32 �xg7 �d8 33 d6+ ; 31 . . . b1 � 32 gx b1 axb1 � 33 gx b1 ) 32 cxd5 a1 � (32 . . . b1 � 33 gx b1 a x b1 � 34 gx b1 ) 33 �xg7 gc1 34 exf7 +-; B) 27 . . . � c7 ! 28 e5 � b6 29 d6+ � b8 (the king hides in q u ite a safe place ; if 29 . . . � d8 30 � h1 a2 31 � h7 �c6 32 � h4+ f6 33 � h 8+) 30 �d4 gc8 (30 . . . a2 31 d7 a1 � 32 d8 � + �xd8 33 �xd8+ gc8 00 ; 30 . . . Ac8 !?) 31 c5 �c6 00 .

27 . . . .tc8 27 . . . � c7 28 d6+ gxd6 29 �x a3 +-, or 27 . . . a2 28 �x b7 a1 � 29 e4 +- (diagram). ...

25 'fJxe4 a3 26 'fJd3 b2 Or 26 . . . a2 27 �x b3 a1 � 28 gxa1 gxa1 29 �xb7. a 8 7

b

1-,--::-

c

d

e 8 7

5 4

I especially l i ke this position as, despite Black's extra q ueen , he has no defence against e4-e5+. Or 27 . . . �e7 28 e4 � c7 29 e5.

2

'"-_______....;:;.-._--1

11

28 'fJb8+ �d7 29 .tg4+ �d8 30 d6

147


30 . . . a2

..

I n the event of 30 . . . b1 � I was planning 31 .§.x b1 .§.xd6 32 .§.bd1 +-, but the brutal computer suggests 31 d7 ! with mate in 7(!) after 31 . . . �xd1 32 dxc8 �+ .§.xc8 33 '§'xd1 +.

31 d7 Black resigns

*** Game 33

B o r i s G e l fa n d - Va l e r y S a l oy M ad ri d 1 9 96 Queen 's Gambit A ccep ted [028J 1 d4 dS 2 c4 d x c4 3 �f3 a6 4 e3 �f6 S Axc4 e6 6 Oe2 a

b

c

d

e

An over-ambitious move. I discovered back i n 1 988 that this idea is not sound an d I had to wait eight years for an opportunity to demonstrate this. 9 . . . b5 is considered stronger, for example 10 e5 tDd7 (10 . . . bxc4!? 11 exf6 �d 3 !) 11 A b3 A b7 12 .§.d1 �b6 13 �c3 �e7 (Gelfand - Balashov, Minsk 1 986).

10 eS! �d4?! 10 . . . �c7 11 Af4 f6 would give Black an extra pawn, but after 1 2 � bd2 � g x e5 1 3 tD xe5 fxe5 14 Ag3 it is clear that the pawn is irrel evant and that Wh ite has the better game.

11 �xd4 Oxd4

L...-_______---"'-__-'

As the e5 pawn cannot be defended, it looks as though Black has seized the i n itiative, but. . .

(f

I have used a n u m ber of variations against the Queen 's Gam bit Accepted and here I preferred the Furman Variation . It makes the opponent's preparation harder if he h as to be ready to face different plans. White antic ipates . . . c7-c5 and removes his queen from a possible exchange on the d -file. Recently, thanks to the efforts of Kramn i k and Bareev, 6 0-0 c5 7 dxc5 has become popular, where Black also has to be careful .

6 . . . c S 7 d x cS Axcs 8 0-0 �c6 9 e4 �g4?!

148

5 4 3

L...-_______---"'-__-'

{f

Game 33

Gelfand - Salov, Madrid 1996

12 �d2! White sacrifices a pawn , but Black is left ter ribly behind in development.

1 7 �xg7 would allow B lack good counter play : 1 7 . . . 'ffe 5 (17 . . . Ae5 1 8 � h6 .§.g8 1 9 f3) 1 8 'ff h 6 .§.g8 1 9 .§.ad1 Ad7.

17

12 . . . �xe5 1 2 . . . b5 1 3 Ad3 � x e5 14 Ae4 (14 � b3 ±) 14 . . J �b8 15 � b3 is simply bad for Black.

13 �b3 .xc4 14 .xe5 Af8 15 Ae3 In 1988 I term inated my analysis here, cor rectly judging that Wh ite h as more than enough com pensation for the paw n. In the early n ineties, however, this position oc curred a few times i n practice, beari ng out this assessment, and so I gave up hope that I wou ld be able to profit from my oid analy sis. Strangely enough, my opponent allowed me to do so. 1 5 .§.d1 !? has also been tried .

15

. . .

•d5 N

Black managed to survive after 1 5 . . .f6 1 6 'ff h 5+ g6 1 7 �f3 Ae7 1 8 .§.ac1 �d5 1 9 'ffx d5 exd5 20 .§.fe1 Ad8 i n Lerner Sorokin, (Alekhine Memorial Open , Moscow 1 992) . However, his position was m uch worse d u ring the course of the game. 15 . . . Ad7 1 6 .§.fd1 Ac6 1 7 �d4 ! ± (or 1 7 �a5 'i' b5 18 � x c6 bxc6 1 9 'ffg 3) si mi larly does not allow Black to develop his kingside.

b

c

•e4

1 7 . . . 0-0 was n ot sufficient, because of 1 8 .§.fd1 'ffc 6 1 9 .§.ac1 'ffd 7 20 � c5 �e7 (20 . . . Axc5 21 Axc5) 21 �e4 A b8 and now : A) premature is 22 Ac5 A a7 23 A xa7 .§.xa7 24 f5 exf5 25 'ff b 8 fxe4 26 .§.xc8 'ffc 5+ ! ; B) 2 2 �f2 b 6 (22 . . . .§.d8 2 3 .§.xd8+ �xd8 24 A b6 'ffe 8 25 .§.d1 A d7 26 � c5 +-, or 22 . . . 'ff b 4 23 A c5 �x e4 24 A xf8 A xf4 (24 . . . 'ffx f4 25 �xf4 A xf4 26 .§.d8 +-) 25 .§.d8 +-) 23 Ax b6 Aa7 24 �d6 ± ; C) 2 2 h1 ! (the strongest ; Black's pieces are so awkward ly placed that it is no wonder he loses material) 22 . . . A a7 23 f5 exf5 24 A h6 g6 25 'ffc3 f6 26 Axf8 xf8 27 � xf6 +-.

18 Elad1 Ae7

5 4

16 .g3 Ad6!? a

. . .

3 d

e

5 ....� .. ."""""'...t--...., _.,....,...."

4 3 2

f

L...-_______...:_ ::... ----I

'\J

19 f5! 5

Played i n the spirit of the position . Wh ite should keep the black king in the centre and attack it, rather than restore the material bal ance. 19 Ad4 'ffg 6 20 'ffx g6 h xg6 21 A x g7 .§.h5 = would let Black off the hook.

4

L...-_______----=-__.....I

19 . . . exf5?!

'\J

The right idea. Black is ready to give back his extra pawn i n order to develop his pieces.

17 f4!

Underestimating White's attack. There would have been more possibil ities to defend after: A) 1 9 . . . 'ff h 4? 20 'ffx g7 Af6 21 .§.d8+ xd8 2 2 'ffx f7 and wins;

149


B) 1 9 . . . h5?! 20 Ad4 tffg 4 21 tffc7 Ad8 22 tffc3 with a dangerous attack; C) 19 . . . A h4 20 tff h 3 (20 tffx g7 tffx e3+ 21 � h1 .!:U8 22 f6 Ad7 00) 20 . . . Ae7 21 Ad4 ±.

20 Ilfe1 ! Premature is 20 tffx g7 tffx e3+ 2 1 � h1 .§.f8 22 .!:Ue1 tffb 6.

more qu ickly, but the difference in the plac ing of Wh ite's pieces and Black's was so great that I didn't want to force matters.

24 . . . �8? Losing by force. 24 . . . tfff6 25 A e5 �g5 26 h4 ± . a

20 . . . "g4

b

c

d

e

8 7

20 . . . A e6 21 tffx g7 E!f8 ± was more tena cious, although Wh ite has a big advantage.

6

21 "c7 Ae6 22 Ild4 Now Black's position is hopeless, as he can not castle and he has no defence against the mating attack. Less good was 22 etlc5 ?! (22 tffx b7 tffe4) , and now not 22 . . . E!c8 23 etl x e6 E!xc7 24 etl xc7+ � f8 25 Ag5 (25 E!d8+ A xd8 26 Ac5+ A e7 27 E!xe7 g6 28 E!e4+ � g7 29 A d4+ f6 30 E!xg4 fxg4 31 etl d5 E!f8 32 � f2 � f7 00 ) 25 . . . f6 26 E!d7, but 22 . . . E!d8 ! .

22 . . . "g6

8

5

5

4

4

3

3 2

�------�--� �

25 Ild7 ! This double exchange sacrifice proves deci sive.

25 . . . Axd7 26 "xd7 "b6+ 27 ct>h1 Ilea

22 . . .f4 23 tffx b7 E!c8 24 Axf4.

27 . . . �d8 28 E!xe7 �xd7 29 E!xd7 +-.

23 "xb7 Ilc8

28 Hxe7 ! Hxe7 29 Ad6 "xd6 30 "xd6 h5 31 �d4 Ilh6 32 "c5 Ilf6 33 ct>g1

23 . . . E!d8 24 E!xd8+ A xd8 25 Ac5.

24 Af4 Perhaps 24 E!c1 !? E!xc1 + (24 . . . E!d8 25 E!xd8+ A xd8 26 Ac5) 25 A xc1 would have won

Destroying Salov's last hopes : 33 etlxf5 E!xf5, or 33 etlc6 E!xc6.

***

150

Black resigns

Game 34

Gelfand - Karpov, Vienna 1996

Game 34

B o r i s G e l fa n d - A n a t o l y K a r p o v V i e n n a 1 9 96 Queen 's Indian Defence lE 1 7] Unfortunately, top-level tournaments in Aus tria are extremely rare nowadays. However, the 1 000 th anniversary of the city was cele brated with a great chess festival , which con sisted of a category 18 su per-tou rnament, an extremely strong grandmaster open , and eight more tournaments. The games took place i n Vien na's marvellous Rathaus. I had to face a legendary opponent, who had been the best player i n the world for ten years. Karpov seemed to be in good form - he had just won his F I DE world championship match against Gata Kamsky. I am happy that I had the chance to play dozens of games (as wel l as many rapid games) against Anatoly in the 1 990s when he was sti l l one of the very best players in the world.

1 �f3 �f6 2 c4 b6 3 g3 Wh ite is trying to delay d2-d4 and to avoid the main lines of the Queen 's Ind ian Defence.

3 . . . .tb7 4 .tg2 e6 5 0-0 .te7 6 �c3 0-0 7 11e1 d5 8 cxd5 exd5 9 d4 This position is quite typical . If Wh ite reaches this position via the Petrosian Variation of the Queen 's I ndian, he will have a pawn on a3 instead of a rook on e1 . But it is not clear which of these moves is more usefu l .

9 . . . �a6 10 .tf4 c5 a

b

c

d

e

1 1 dxc5 Other moves are also possible, but I wanted to define the pawn structure immediately. 1 1 . . . �xc5 So, Black decides to play with an iso lated pawn . I guess that Anatoly d i d n 't l i ke 1 1 . . . b x c5 1 2 ttl e5 (with the idea of e2-e4) 1 2 . . . ttl c7 1 3 ttl c4 ttle6 14 Ae5 and the d5 pawn comes u nder attack. 1 2 11c1 a6 After 1 2 . . . ttlfe4 13 A e5 the bishop goes to d4 and it is easier to play this position with Wh ite. The following year Yan n ick Pelletier achieved a satisfacto ry position with 1 2 . . . ttl ce4 N 1 3 A e5 Ac5 14 Ad4 "ffJe7 1 5 A h 3 ! gfd8 (15 . . . ttlxc3 16 gxc3 ttle4 (16 . . . A b4 ?! 1 7 ge3 ttle4 1 8 gf1 ) 1 7 gc2 ;t; Pelletier) 1 6 a3 ttl x c3 (16 . . . a5 ? ! 1 7 ttla4 ! Pelletier) 1 7 gxc3 ttl e4 1 8 gc2 a5 1 9 "ffJd 3 ( 1 9 "ffJc 1 !? i nten d ing "ffJf4 ;t; Pelletier) 1 9 . . . "ffJd 6 20 "ffJe 3 ge8 21 "ffJf4 gad8 !? (21 . . . "ffJx f4 22 gxf4 ;t; Pel letier) 22 "ffJx d6 gxd6 23 e3 ;t; (Gelfand - Pelletier, Credit Suisse, Biel 1 997) . 1 3 a3 A usefu l move, as in some cases after b2-b4 the b4 pawn needs to be protected . 1 3 . . . 11e8 14 �d4 .td6 B lack exchanges bishops in order to gain some space. If 14 . . . "ffJd 7 15 b4 ttle6 16 A h3, or 14 . . . ttlfe4 15 b4. 15 .txd6 "xd6 1 6 "d2 Had8 17 11ed1 g6 18 "f4! a

f

b

c

d

8

8

8

..

e

7 6

6 5

5

5

4

4

4

3

3

2

2

�------��-

�

a

151

b

c

d

e


I took this decision q u ite easily. S u bcon sciously I felt that this move was strong. Re cently, browsing through my oid games, I discovered that a simi lar idea brought me success agai nst Leonid Basin many years ago i n the Belorussian Championshi p (see below, p. 154).

1S

. . •

•xf4?!

I don 't l i ke this decision, as without the q ueens Wh ite has more opportun ities for manoeuvring his pieces. Also, he gains con trol of the e5 sq uare. It wou ld seem that Ana toly didn't feel comfortable in this situation , as he had won qu ite a number of such pos itions with the stronger side and normally it is d ifficult to fight against you r favourite weapon . 18 . . . fNe7 was better.

24 gdc2 gd8 25 &Dc6 Axc6 26 gxc6, as then B lack's cou nterplay after 26 . . . d4 27 &D xd4 (27 gx b6 d3 with counterplay) 27 . . . &D xd4 28 ex d4 gxd4 29 gx b6 gxf4 30 gxa6 ge2 31 gf1 is hardly sufficient.

23 . . . �dS 24 a5 Ae6 24 . . . b5 Ieaves Black without any counterplay and with his pieces bad ly placed .

25 �e3 AeS 26 Af1 bxa5 26 . . . b5 is s i m i lar to the previous note. If 26 . . . ga7 ? 27 b5 ! .

27 bxa5 Hb7 8

8

7

19 gxf4 ;t 'i!?fs 20 e3 Hd6?! The rook is bad ly placed here. Better was 20 . . . � e7 (Black has to bri ng his king to the centre) 21 b4 &De6 22 &D ce2 gc8, or 20 . . . &Dfe4 !?

2

21 b4 �e6 21 . . . &Dce4 22 &Da4, intending f2-f3 .

L..-_______---"'__--'

22 �ee2 He7 a

b

c

2S Ha1 ! d

e

8

The most difficult move in the game and one that I am proud of find ing ! Again I fol lowed one of N i mzowitsch's principles and over protected the a5 paw n , which is the key to Wh ite's success. Anatoly spent a lot of time here, but he failed to fi nd the rig ht defence.

8

7 6 5

5

4

4

3

3

2

2

a

b

c

1f

d

e

9

h

2S . . . De7?!

1f

23 a4! Wh ite has to create a second weakness (the a6 pawn), as N imzowitsch wrote in his My System . H e could also have taken control of the c -fi le by 23 gd2 , and if 23 . . . &D e8

Blac k does not find the pro per response to Wh ite's idea. 28 . . . &D e4 is refuted by 29 &D d b5 ! ! +- (H uzman ; if 29 &D de2 g bd7) 29 . . . A x b5 30 &D x e4. Better is 28 . . . � e7, or 28 . . . &De6 29 &Dxe6+ (29 gdb1 gx b1 30 gxb1 &D xd4 31 exd4 � e7 32 gb6 gx b6 33 axb6 Ac6 34 A x a6 &Dd7 35 b7 &D b8) 29 . . .fxe6 30 gdb1 ;t .

152

29 �a2 D b7 30 f3 �e6 31 Ddb1 He7 32 �b4 ± �e5

Game 34

Gelfand - Karpov, Vienna 1996

Sad ly necessary, as Black loses after 32 . . . � xd4 33 exd4 ga7 34 � x a6 gaxa6 35 Ax a6 gxa6 36 gb6 gx b6 37 a x b6 Ac6 38 gc1 .

a

b

c

d

8

8

d

e 8

7

'"

e

c

8

33 Ilc1 Ilb7 34 Ilab1 a

b

7

6

6

5

5

4

4

3

3

2

2

6 5

5

4

4

3

3

2

2

--' l}

L...-_______---=-__

49 �c2 The simplest decision . N ow Wh ite wins the rook for the a-pawn.

Black's pieces are extremely passive here and his time trouble doesn't help. 34

. • .

49 . . . Dxe3 50 �b2 De2+ 51 �b3 Ilxh2 52 a6 Dh1 53 �b4 Da1 54 Da5 Db1 + 55 �c5 Ilb8 56 a7 Da8 57 �c6 h5 58 �b7 58 ga6 ! was simpler.

�fd7 35 �bc6 Ilc7

58 . . . Dxa7+ 59 �xa7 �d6 60 �b6 h4 61 Dc5 h3 62 Dc2 �e5 63 Dh2 �xf5 64 Dxh3 �4 65 �c5 f5 66 �d4 g4 67 fxg4 fxg4 68 Ilh8 �3 69 �d3 Black resigns

35 . . . f6 !?

36 �e5 �e7 37 Dc3 f6 38 �xd7 �xd7 39 Ilb8 39 gbc1 �e6 40 � x e6 gxc3 41 �f8+ � e7 42 � xg6+ h xg6 43 gxc3 d4 ± allows Black some counterplay.

39 . . . �e6 40 Ilxc7+ �xc7 41 �2 �e7 42 f5 g5 43 �e1 The black pieces have no moves and White is threaten i n g to take his king to c5. There fore Black decides to force matters. 43

Ab5 44 Axb5 �xb5 45 �xb5 axb5 46 Ilxb5 Ilc6 47 Ilxd5 Ilc3 48 �d2 Ila3 . . .

I especially l i ke this game, because I beat the g reat Anatoly Karpov i n his own style. Obviously, he normally used to win i n re fi ned positional style, rather than lose ! Th is game remains up till now the o n ly decisive one after our Cand idates sem i-final match in 1 995 ! This victory enabled me to win the tournament on tie -break ahead of Topalov and Karpov. It is also memorable for the fact that I managed to beat both of the best play ers of the mid -seventies to early-eighties, Karpov and Korchnoi, in one tournament.

* * *

153

My Most Memorable Games A

Similar Idea (cf. note to White's 18 th move) Boris Gelfand - Leonid Basin Belorussian Championshi p 1 985 a

b

c

d

21 . . . Ad6 22 Axd6 "xd6 23 "d2 ge7 24 "f4! "xf4 25 g xf4 �e8 26 e3 �d6 27 b4 b5 28 a4 a6 29 Ha1 g5 30 axb5 a x b5 31 Hxa8+ Axa8 32 Ha1 Ab7 33 �e2 ge8?! 34 fxg5 hxg5 35 �c1 Has?! 36 Ha5 �c4 37 Hxa8+ Axa8 38 �d3 ± and I managed to win this better ending.

e 8 7 6 5 4

Position after 2 1 tWc2

***

Game 35

5 a3 Axd2+ 6 "xd2

B o r i s G e l fa n d - V i c t o r Ko rch n o i V i e n n a 1 9 96 Bogo- Indian Defence [E 1 1] I don't t h i n k that Victor Korchnoi needs any further introduction from me. Apart from any thing else, Dims have recently published his excel lent books. I would just l i ke to add that even now, at the age of 73, he remains a dangerous opponent for everyone and he has more fighting spi rit than almost anyone else on the circuit ! I faced him for the first time in a s i m u ltaneous d isplay in M i nsk i n 1 975, six months before h e defected and his games became ' half-forbidden ' in the USSR.

White is pinning his hopes on his long -term trum ps - the pair of bishops , strong centre etc. Black, on the other hand , has no weak nesses and he can start an assault on the kingside by . . . � e4, . . .f7-f5 etc. A compli cated strateg ic battle n ow ensues. 6 Axd2 A b7 7 Ag5 is an alternative way to play.

1 d4 �f6 2 c4 e6 3 �f3 Ab4+ 4 �bd2 This move is less popular than 4 Ad2. How ever, I have played it consistently for many years and generally with good resu lts. 4 . . . b6 G iving up the centre by 4 . . . 0-0 5 a3 Ae7 6 e4 d5 is not to everyone's liking. 4 . . . d5 was played i n Gelfand -Yusupov (Game 22 , p. 1 03).

154

6 . . . Ab7 7 e3 0-0 8 Ae2 a

b

c

d

•

e 8

7 6

6

5

5

4

4

3

3

2

2

a

8 . . . a5

b

c

d

e

9

h

Game 35

Gelfand - Korchnoi , Vienna 1996

A rare plan , the idea of which becomes clear on the next move. The majority of my oppo nents have chosen 8 . . . d6 9 0-0 � bd7 10 b4 �e4, when Wh ite faces a choice. In fact, he is sim ply a tempo up, compared with a line in the N i mzo-I n d ian with 4 �c2 (there the pawn is on b2) , which was popular i n the early eig hties and is sti l l used from time to time by Vladimir Kram nik. I therefore believe that, bei ng a tempo up , White should stand better, but things are far from clear:

fxe4 26 A a3 (Gelfand - Ionov, USSR Cham pions h ip First League, Klai peda 1 988) and White is clearly better. A2) 1 2 . . . gf6 ! (this d i rect attack is the best solution) 13 dS

A) 11 �d3 is tem pti n g , as from here the queen is better placed to su pport d4-dS. After 1 1 . . . fS 1 2 A b2 Black has : A1 ) 1 2 . . . �f6 1 3 � d2 gae8 14 f3 (the attack comes to an end , leaving Wh ite with an enduring advantage) 14 . . . �gS (14 . . . � xd2 1S tWxd2 ;;!;;) 1S a4 ! (intending to bring the rook into play on the a-file after a4-aS) 1S . . . �g6

16 tWc2 ! (a strong prophylactic move ; if 16 as eS 1 7 a x b6 (1 7 dS e4 00) 1 7 . . . a x b6 1 8 ga7 e4 with counterplay) 16 . . . cS (16 . . . eS 17 dS) 17 bxcS b x cS 1 8 � h1 ! (again prophylaxis against 1 8 . . . cxd4 19 exd4 � h3+ 20 � h1 �f4) 18 . . . tW h6 1 9 as �f6 (19 . . . cxd4 20 exd4 �f6 ;;!;; ) 20 d x cS � hS 21 � g1 d x cS 22 f4 ! (White changes the character of the game, exchanging Black's good pieces ; if 22 � b3 eS 23 � xcS � h3+ 24 g x h3 �xe3+ 2S � h1 tWxcs 26 Aa3 tWd4 with defi nite compensa tion) 22 . . . � e4 (22 . . . �f7 23 � b3 ai m i n g at the weak cS pawn ; 22 . . . � h3+ 23 gxh3 tWg6+ 24 � f2) 23 Af3 �g6 (23 . . . � xd2 24 �xd2 Axf3 2S g xf3 ±) 24 � xe4 A xe4 2S A xe4

13 . . . gg6 ! (the rook has taken up its best pos ition , opposite the wh ite king) 14 d xe6 ? (I u nderestimated the opponent's attack and opened the long d iagonal for his bisho p ; 14 �e1 !? tW h4 1S f4 00 I liescas ; 14 g3 !? Razu vaev ; 14 .§.ad1 ) 14 . . . �f8 1S cS � xe6 16 cxd6 cxd6 17 .§.ad1 � h8 and nearly all Black's pieces take part i n the attack (Gelfand I I lescas, Madrid 1 996) ; B) 1 1 �c2 cS (11 . . . fS 1 2 dS and if 1 2 . . . exdS 1 3 � d4 00 is another theoretical l ine) , and now : B 1 ) 1 2 � d2 was tried i n another game of m i n e : 12 . . . � xd2 1 3 �xd2 (13 A xd2 !?) 13 . . . �gS 14 f3 .§.fd8 1S .§.d1

X

1 S . . . d S ! (realising that the bishop pair may become a dangerous factor, Black starts concrete counterplay in the centre) 16 � b2 ! c x b4 1 7 �x b4? ! (as shown by Anand, White's only chance to fight for an edge was 17 e4 ! tWe7 18 cxdS exdS 19 eS bxa3 20 tW b3

155


�f8 21 Axa3 (0) 17 . . . Aa6 18 e4 �g6 1 9 �a4 Axc4 20 Axc4 dxc4 21 �xc4 with an equal game (Gelfand -Anand, Dortmund 1 996) ;

without wasting a tempo on moving the bishop twice.

B2) 12 A b2 Ek8 13 gad1 dS ! (13 . . . fS 14 dxcS b x cS 1 S �d2 �gS 1 6 f4 l!fg6 1 7 Af3 �df6 1 8 A xf6 gxf6 1 9 A xe4 fxe4 20 � b3 ±) 14 d x cS bxcS 1 S Ad3 fS ! (1S . . . l!fe7 1 6 bS fS (16 . . . h6 17 A xe4 d x e4 18 �eS ;I; ) 17 � eS � x eS 18 A xeS �d6 1 9 f3 ;1; ) 16 bS � d6 1 7 � eS ;I; and here the advantage of the two bishops is very relevant (Gelfand - Macieja, Akiba Rubi nstein M emorial , Polanica Zdroj 1 998) .

1 3 . . . �e4 14 l!fe1 � x eS 1 S d xeS l!fgS (1S . . . l!f h4 1 6 gd3) 1 6 Af1 l!f hS (16 . . . l!fg6 1 7 Ad3) 1 7 f3 � gS 1 8 l!fg3 (18 h4 � xf3+ 19 g xf3 dxc4 (0) is in White's favour.

9 b3 d5!? N A new approach to this position . Black puts direct pressure on the centre.

13

• • .

14 ctlxd7 I decided to exchange a pair of knights. 14 �e1 a4 ! (14 . . . cxd4 1S Axd4 ;1; ) gives Black good counterplay, for exam ple: 1 S � xd7 gxd7 16 b4 c x b4 17 l!fx b4 l!fx b4 18 axb4 dxc4 1 9 Axc4 gc7.

14 . . . ctlxd7 14 . . . gxd7 1S l!fe1 ;1; .

15 11ac1

10 0-0 ctlbd7 11 J.b2 1 1 gd1 l!fe7 1 2 a4 cS 1 3 A a3 was an alter native.

11

. • •

a

b

c

d

..,

e

8

c5 a

l1fd8

7 b

c

d

6

6

5

5

4

4

6

3

3

5

2

2

e

8

4 e

3 2

f

9

h

'if

15

h

• • .

a4

Th is is the accurate move order. Bad is 1S . . . cxd4 16 exd4 a4 17 cS ! and if 17 . . . bxc5 18 dxcS � xcS 19 l!fd4.

1 6 bxa4 dxc4

1 2 11fd1 1 2 cxdS A xdS 1 3 b4 � e4 14 l!fe1 a x b4 1 S axb4 gxa1 1 6 Axa1 l!fa8 = . . . •

9

This allows B lack to simpl ify the position. 1S a4 !? gac8 16 gac1 or 16 Aa3 !?

Now B lack's idea becomes clear. Because of his weakness on b3, Wh ite cannot clear the centre.

12

f

•e7 1 3 ctle5

Another possibility was 13 a4 gfd8 14 A a3 � e4 1S l!f b2 . However, it would have been more log ical to do this on the 11 t h move,

1 6 . . . gxa4 1 7 cxdS AxdS 1 8 A bS.

17 J.xc4 ctle5 18 J.e2 Definitely not 18 Af1 71 �f3+ ! -+, or 18 A bS? Af3 ! (18 . . . �f3+ 1 9 g x f3 A xf3 20 � f1 l!fg5 21 l!fc3 �g2+ 22 � e1 ) 1 9 ge1 l!fgS.

156

18 . . .•g5?!

Game 35

Gelfand - Korchnoi, Vienna 1996

This leads to a better end ing for Wh ite. It appears that Black could have equalised by 18 . . . cxd4 1 9 A xd4 ( 1 9 exd4 tb g6 00 with the idea of . . . tb h4) 1 9 . . . tb c6 20 � b2 tbxd4 21 gxd4 gxd4 22 �xd4 (22 exd4 h6 =, but not 22 . . . �d6?! 23 gb1 ) 22 . . . �xa3, but after 23 ga1 (23 gd1 Ac6 24 A b5 !?) 23 . . . �b3 24 f3 ! (24 Ad1 �d5 25 �xd5 Axd5 26 .§.b1 .§.a6 =) his problems are not yet over: 24 . . . �c2 (24 . . . Ac6 25 A d1 �d5 26 �x b6) 25 A d3 �c5 (25 . . . �c7 26 g b1 ) 26 �xc5 bxc5 27 a5 ± .

1g e4! a

b

c

d

..

e 8

7 6

6

5

5 4

3

3

2

2

The only move, but a strong one anyway. Now the game transposes i nto an ending.

19

. . .

•xd2 20 Hxd2 .txe4 21 Hcd1

An inaccuracy. The same idea would h ave been stronger after 21 f3 ! A b7 (21 . . . A c6 22 a5 ! gxa5 23 Ac3 gxa3 24 d xe5 gxd2 25 A xd2 ga2 26 gd1 A a4 27 Ac4 +-) 22 gcd1 (22 a5 b x a5 (22 . . . gxa5 23 A c3) 23 gxc5 tb g6 =) 22 . . . tb c6 23 d x c5 gxd2 24 .§.xd2 bxc5 25 .§.d7 ±.

21

. . •

�c6

21 . . . cxd4 22 A xd4 ± , or 21 . . . tb g6 22 d x c5 .§.xd2 23 gxd2 bxc5 24 f3 Ad5 25 A b5 ± .

22 dxc5 Hxd2 22 . . . b x c5 23 gxd8+ gxd8 24 gxd8+ tb xd8 25 a5 ± .

23 Ilxd2 bxc5 24 .tb5

a

b

c

d

e 8

2

24 . . . �a5? B lack misses a chance to exchange one of the bishops : 24 . . . tb d4 25 A xd4 c xd4 26 '§'xd4 A d5 when he is very close to a d raw, as his king goes to the centre and White cannot push h is a-pawn any further. 25 Hd7 This end ing is definitely in Wh ite's favour be cause of the bishop pair and the passed a pawns. The fact that they are doubled didn't bother me too much. And it is always en joyable playing when your opponent doesn 't have any counterplay. 25 . . . h5 I think it was more accurate to put the pawn on h6 and not create any extra weaknesses. 26 h4 26 h3 !? 26 . . . .td5 27 Hc7 �c4 27 . . . tb b3 28 f3 ;t and White i m proves his position by playi ng his king to e3 fol lowed by g2-g4 etc. 28 .tc1 I decided to avoid going i nto an opposite colour bishop end ing by exploiting Black's weakness on the kingside. 28 A x c4 A x c4 29 gxc5 gxa4 (29 . . . Ad5 is the lesser evil , with a tenable endgame after 3 0 a5 ;t ) 30 gg5 ! g6 31 gc5 � f8 32 Af6 � e8 33 gc7 e5 (33 . . . A d5 34 ge7+ � f8 35 gd7 ga8 36 gd6 A b3) 34 ge7+ � f8 35 gd7 ga8 36 Axe5 ± .

157


28 . . . .tid6?! Th is came as a surprise to me, as I felt that the h4 pawn was irrelevant to waste so many tem pi on winning it. I was more expecting 28 . . . � b6 29 �xc5 � xa4 30 �c7 ±. 29 Hxc5 .tif5 30 .ld2 I ' m not sure if this was the right move, as Wh ite had other nice options such as 30 Ad3 � xh4 31 f3 (31 a5 Axg2) 31 . . . �g6 32 a5 A b7 33 �xh5 ± , or 30 Ac4 !? 30 . . . .tixh4 31 a5 a

b

c

d

e -.r.=�-I

8

7

7

6

6

31

. . •

16?

The decisive mistake. 31 . . . �f5 (31 . . . Axg2 ? 32 Ag5 +-; 31 . . . � xg2 ? 32 �xd5 +-) was more tenacious :

A) 32 Af1 � d6 33 f3 � b7 with counterplay; 8) 32 �c7 � d4; C) 32 a6 � d4 (32 . . . � d6 33 Ad3 �e4 34 Axe4 Axe4 35 �a5 ;1;) 33 Af1 � b3 34 �c2 � xd2 35 �xd2 f6 36 �c2 �a7 ;1; ; D) 32 Ad3 A b7 (32 . . . � d4 33 Ac3) 33 �c7 (33 A xf5 exf5 34 �xf5 �d8 35 A b4 �d5 =; 33 Ac3 !?) 33 . . . A a6 34 Ae4 �d8 35 A b4 ± .

32 .lf1 +-

5

Now nothing can stop the a-pawn.

4 3

3

2

2

32 . . . .tif5 33 a6 Ha7 34 .la5 .tid6 35 .lb6 Has 36 Hc7 Hc8 37 Hxc8+ .tixc8 38 .lc5 00 39 a7 .tie7 40 .le2 Black resigns Ae2-f3 is unavoidable.

***

Minsk 1986: Boris Gelfand and Alexander Khalifman analyse the 8 l1b 1 variation of the GrOnfeld Defence .

158

Game 36

Gelfand - Sutovsky, Fontys, Tilburg 1 996

Game 36

B o r i s G e l fa n d - E m i l SutoYs ky F o n tys , Ti l b u rg 1 9 96 King 's Indian Defence [E94J It was excel lent that, after a one -year break, the tradition of stag ing famous chess tour naments i n Tilburg was restored , albeit with new sponsors. The organ isers were able to maintain the extremely high standards. How ever, the new sponsors - a group of universi ties - preferred to have mainly young players in their tournament. And again I was i nvited just a few weeks before the start, to replace a player who had withdrawn . Th is time I per formed better i n this role than I had i n Am sterdam earl ier that year, and I managed to tie for first place with Jeroen Piket. My opponent i n the present game was then the world under-20 champion and this en sured his invitation to the tournament. Su tovsky is a player with a sharp attacking style who has won a n u m ber of beautifu l tactical games. I have been a victim once and have witnessed some of the others, as we are colleagues in the Israeli national team . Dur ing the last few years Emil has improved his endgame technique (I hope, u nder my i nflu ence) and this has enabled him to get very close to the magic 2700 mark.

The plan with . . . tDa6 against d ifferent vari ations of the King's Ind ian Defence became popular in the late 1 980s, than ks to g rand master Igor Glek. I was one of the players who tried it regularly, and so I hoped that this experience would help me when playing with the opposite colour.

7 0-0 e5 8 £e3 c6 9 d5 �g4 1 0 £g5 16 1 1 £h4 11 Ad2 f5 12 tDg5 tDf6 13 exf5 gxf5 14 f4 e4 1 5 A e3 c5 did not bri ng me any advantage in my encounter with Veselin Topalov earl ier that year in Novgorod (in fact I was happy to draw).

1 1 . . . c5 12 �e1 !? N 1 2 tDd2 is the alternative plan and it is simply a matter of taste which one is preferred . I n fact, this i s one of the key qu estions i n the Ki n g 's I ndian Defence. However after this game 12 tDe1 became the main move. a

b

c

d

..

e 8

7 6

6

5

5

4

4

3

3 2

1 d4 �16 2 c4 g6 3 �c3 £g7 4 e4 d6 5 £e2 0-0 6 �f3 �a6 1 2 . . . �h6 8 7 6

6

5

5

4

4

3

3 2

if

The following year Topalov played 12 . . . h5 !?N against me. The bishop at h4 is more vulnerable and the play becomes sharper. 1 3 a3 "ffe7 14 g b1 b6 (14 . . . tD h6!? 15 f3 g5 16 Af2 f5 1 7 b4 b6 with counterplay) 1 5 b4 Ad7 16 tDd3 (both sides have made usefu l moves on the q ueenside - Black has run out of them , so he forces matters on the kin g side, b u t t h i s i nvolves a pawn sacrifice ; if 16 h3 tD h6 17 f3 g5 18 Af2 f5 19 exf5 tD xf5

159


with cou nterplay) 1 6 . . . g5 1 7 Ag3 f5 1 8 h3 �f6 19 bxc5 � xc5 (19 . . . bxc5 wou ld have left the knight at a6 out of the game 20 A xh5 fxe4 21 �e1 ; 1 9 . . . f4 !? 20 c6 Ac8 21 A h2 (21 A x h 5 fxg3 22 fxg3 g4! 23 hxg4 tLl xh5 24 gxf8+ Axf8 25 gxh5 �g5) 21 . . . g4 22 � h1 A h6 23 a4 ± and Black's threats are not so dangerous) 20 � xc5 bxc5 21 A xh5 � x h 5 (21 . . . g4 22 A h4) 22 �xh5 Ae8 23 �f3 f4 (23 . . . g4 24 hxg4 f4 25 A h2 Ad7 26 g b7 !) 24 A h2 Af6, and now 25 �d3 !? (25 g4?! �h7 � Gelfand -Topalov, Dortmund 1997) 25 . . . �g7 26 f3 ± would have left Black with insufficient compensation for the pawn .

13 �d3 1Je8 1 3 . . . �d7 was recommended by Sutovsky, but I don't like this move as it leaves Black's pieces very poorly coordinated .

14 f3 �f7 Black could have begun with 14 . . . f5 but this would have given White an additional possi bility: 15 g4 �f7 16 gxf5 (1 6 � h1 fxg4 17 fxg4 A h6) 16 . . . g xf5 17 � h1 � h8 18 � b5 !? 00 .

15 Hb1 Th is is qu ite a typical position for the King's I ndian Defence. White has to break through on the b -fi le, while B lack h as to prepare a kingside attack. But as it will take him a con siderable time, I prefer Wh ite's position here. Now 15 g4 does not make much sense, be cause of 1 5 . . . �e7 , intending . . . A h6.

1 5 . . . f5 1 6 b4 b6 17 bxc5 17 a4 !? A h6 (threaten ing . . . Ae3) 18 Af2 Ad7 1 9 a5 �e7 20 a x b6 a x b6 21 b x c5 b x c5 22 �a4 00 was an alternative that I con sidered , but Black can gain good counter play by 22 . . . A x a4 23 �x a4 fxe4 24 fxe4 �g5.

There were two alternatives : A) 1 7 . . . dxc5 !? and now : A1 ) 1 8 � b5 Ad7 1 9 a4 allows Black to set u p a blockade with 1 9 . . . A x b5 20 c x b5 � c7 2 1 � b2 � d6 22 � c4 �d7 23 tLl xd6 �xd6, and as Wh ite no longer has his knight, the q ueen can not be d riven away from the blockading sq uare d6 and can later be re placed by the knight ; A2) 1 8 a4 !? � d6 1 9 � b5 (normally White should try to exchange B lack's c8 bishop, but here it is more im portant to get rid of the blockad ing kn ight) 1 9 . . . �xc4 20 �xc5 tLle3 2 1 �c1 � x c5 (21 . . . � xf1 22 � d6 loses the queen) 22 �xe3 ± ; B) 1 7 . . . � xc5 1 8 � b5 �d7 1 9 � x c5 bxc5 20 �a4 and the exchange of the bad tLla6 hasn 't solved Black's problems.

18 1Ja4! This queen exchange elimi nates all counter play by B lack, after which he will be con demned to defence.

18 . . . 1Jxa4 19 �xa4 Ad7 20 �c3 Hfb8 21 �b5! ;t B lack h as no active plan at all after this m ove. 21 .§.b3 .§.x b3 22 a x b3 .§.b8 23 ga1 .§.x b3 24 .§.xa6 .§.xc3 25 .§.xa7 Ae8 wou ld have al lowed an u n necessary simplification of the position and given Black counterplay.

17 . . . bxc5

21 . . . Ah6 22 Af2

160

Game 36

Gelfand - Sutovsky, Fontys, Ti lburg 1 996

2S ga2 Ac3

The bishop has fulfilled its duties on the h4d8 diagonal , and now it is time to cover the e3 sq uare. Black equalised after 22 gb3 Ax bS 23 gx bS /!i) c7 24 g b3 gx b3 2S a x b3 Ae3+ 26 Af2 Axf2+ 27 � xf2 fxe4 28 fxe4 as 29 ga1 i n the game Kiriakov - Isupov, (Orel 1997).

8

.1

8

7 6 5

22 . . . ct>fS

4 3 8

...._ .;;.... ---I {f

L..._ ._ _ _ _ _ _

5

29 exf5!

4

An i m portant decision . White rules out any possibility of his e4 pawn becom ing a target, and he begins an attack on the fS pawn and on the kingside in general .

3

L..-_______----=-__....J

{f

29 . . . g xf5 30 ct>f1 30 . . . tila6 31 Ad1 tilb4?!

23 gb3?! 23 g b2 ! was another, perhaps more accu rate possibil ity, but I wanted to provoke the exchange on bS. After 23 . . . g b7 (23 . . . A x bS 24 cxbS /!i) c7 2S a4) 24 gfb1 � e7 ;t; it is hard to suggest the next usefu l move for Black.

23 . . . Axb5 24 gxb5 tilc7 25 gb3 2 S gaS abandons the b -file and al lows 2S . . . gb7 with the idea of . . . gab8 and . . . Ad2. 2S gbb1 gb6 26 /!i)xcS dxcS 27 AxcS+ � e8 leads to an unclear position .

25 . . . Iixb3 26 axb3 a5 27 ga1 Ad2 27 . . .fxe4 28 fxe4 /!i) gS was an alternative, but it would allow Wh ite to break through with 29 /!i) xcS !? (29 Af3 /!i) xf3+ 30 g xf3 ;t; ) 29 . . . d xcS 30 A xcS+ � e8 31 A b6 /!i)a6 32 Ad3 (32 gxaS /!i) xe4 33 Ad3 /!i)d2 34 cS (34 b4 gb8 3S gxa6 gx b6 36 gx b6 A e3+ 37 � h1 A x b6 38 cS Ad8 39 d6 � d7) 34 . . . Ae3+ 3S � h1 /!i) x b3 36 gxa6 gxa6 37 A x a6 /!i) xcS =) 32 . . . /!i) b4 33 A b1 Af8 34 � f1 , although this position is far from clear.

I t h i n k that this exchange is in Wh ite's favou r. After 31 . . . � g7 32 ga4 (the i m m e diate 3 2 Ac2 is met b y 3 2 . . . e4) and Ac2 h e has only a sl ight advantage.

32 tilxb4 Axb4 33 Ac2 Forcing a weaken ing of Black's kingside.

33

• . .

tilh6

33 . . . e4 34 fxe4 f4 is too elaborate for an ending : 3S g3 ± . 34

Ae3 f4 34 . . . � g7 3S A x h6+ � xh6 36 AxfS leads to a type of opposite -colour bishop ending that is extremely d ifficult to defend , as Fischer and several other players have shown many ti mes, although it is also very hard to win after 36 . . . � gS ! ' 35 Ac1 �g7 36 g3 White gains space for his bishops.

36 . . . fxg3 36 . . . gfB 37 gxf4 exf4 38 Ad3 ± .

161

37 hxg3 gfS 38 �g2 tilf5 39 f4! tild4


39 . . . h5 40 fxe5 dxe5 41 Axf5 gxf5 42 Ae3 ± , or 3 9 . . . A e1 !? 40 A xf5 (40 g4 tiLl d4 41 f 5 h5 (41 . . . g b8 ! ?) 42 � h3 h x g4+ 43 � xg4 tiLl xc2 44 gxc2 g b8) 40 . . . gxf5 41 g4 gf7 42 f5 h5 43 � f3 ;t .

56 � e1 tiLl e6 (56 . . . e4 57 � f1 ) 57 gf5+ �e4 58 gf6 tiLld4.

8

8

40 fxe5 dxe5

7

7

a

6

6

b

c

8

5

8

4

4

6

3

3

5

2

2

7 6

4 a

56 ______...._ L..._ ...::... ----l

lf

41 £95! By seizing this i m portant diagonal , Wh ite makes the d -pawn very powerfu l . Black will soon be forced to g ive u p the exchange for it.

41 . . . h6 42 £h4 1lf7 42 . . . � f7 43 Ad1 . 43 d6 Ild7 44 £e7 cm7 45 £d1 J:be7 46 Ilf2+?!

A tempti ng check, but it takes away an i m portant sq uare from t h e k i n g a n d makes Wh ite's task more d ifficult. 46 d x e7 � x e7 47 � f2 +- and the king goes to e4.

46 . . . �e6 47 dxe7 �xe7 48 cm1 �e6 49 Hh2 49 g4 e4 50 gh2 � e5 51 gx h6 � f4 52 gf6+ � g5 53 gd6 � f4 54 gd5 was also possible. 49 . . . cm5! The king becomes active and creates a lot of problems. If 49 . . . e4 50 gx h6+ � e5 51 gh5+ � f6 52 g4 e3 53 gd5 � g6 54 g5 +-.

50 Ilxh6 �e4 51 94 £d2 52 Hh2 £f4 53 Hh3 £d2 54 cm2 cm4 55 Ilh5 £e3+ 56 cm1

• • .

b

c

d

e

f

9

h

£c1 ?

Losing control of the g - pawn . 56 . . . tiLle6 ± was the best chance, tryi ng to block the pawn . In this case after 57 gf5+ �e4 I would have faced serious technical difficulties : A) 58 gf6 tiLl d4 59 � e1 and there is no win after: A1 ) 59 . . . Af4 60 ga6 ! (60 gd6 � d3 61 gd5 Ag3+ (61 . . . a4 62 b x a4 � xc4 63 gd7 � b4 64 g b7+ ±) 62 � f1 � d2 63 g5 � x d1 64 g6 e4 65 g7 e3 66 � g2 e2 67 gh5 e1 'tW 68 gh1 �c2 69 gxe1 Axe1 70 g8 'tW � xb3) 60 . . . �d3 61 gg6 Ag3+ (61 . . . � e4 62 g5 � f5 63 ga6 A xg5 64 gxa5 ±) 62 � f1 � d2 63 � g2 Af4 64 Af3 tiLl x b3 ;

A2.) o r 5 9 . . . � d3 6 0 gg6 e4 (60 . . . Ad2+ 61 � f1 Af4 62 g5) 61 g5 Ad2+ 62 � f1 �f5 (62 . . . e3 63 gd6) 63 Ae2+ � c2 64 gf6 �g3+ 65 � f2 A xg5; B) 58 Af3+ ! (it is i m portant to bri ng the bishop back i nto the game, otherwise it is d ifficult for the g - pawn to cross the g5 sq uare) 58 . . . � d3 (58 . . . � d4 59 Ad5 tiLl g5 60 �e2 Af4 61 gf8 � c3 62 gb8) 59 Ad5 �d4 60 gxe5 tiLl x b3 61 g5 a4 62 g6 a3 63 ge7 a2 64 ga7 a1 'tW + 65 gxa1 tiLl xa1 66 g7 +-.

162

57 95 �f5?

Game 37

Rublevsky - Gelfand , Akiba Rubinstein Memorial Tournament, Polanica Zdroj 1998

57 . . . � f5 58 g6+ � xg6 59 gxe5 A a3 would have put up more of a resistance, but would not have saved the game.

58 g6 �e7 58 . . . tD g3+ 59 � g2 tD x h5 60 Axh5 +-.

ttJ

59 g7 e4 60 Ilxc5 Ab2 61 !lc7 Black resigns Soon after the tournament I was very pleased to receive a call from Vladimir Kram n i k, in which he highly praised this game.

*** Game 37

Sergey R u b l ev s k y - B o r i s G e l fa n d A k i b a R u b i nste i n M e m o r i a l To u r n a m e n t , P o l a n i c a Zd roj 1 9 98 Sicilian Defence [852J Sergey Rublevsky is relatively l ittle -known to the chess public outside of Russia, but his strength is obvious to even the best play ers in the world . H is rating has hard ly ever been below 2650 during the last five years. It is both easy and difficult to prepare against him. H is open ing repertoire is very narrow, so normal ly you can easily guess which vari ations he is going to choose. On the other hand , he polishes them u ntil they are shin ing and he knows all their ins and outs. So, even though I was sure about the position we were going to reach , he sti l l managed to come out of the opening with an advantage. This game was also i m portant for the tour nament standings. The previous year Sergey had won this tournament with a bri l l iant 7 out of 9, leaving me in second place a point behind. Th is year too he was in contention for the top places until the last round.

have also occu rred in my games, but the variation chosen here is the most sol id.

4 Axd7+ "xd7 5 c4 �c6 6 �c3 �f6 7 0-0 Wh ite can try to keep his knight on d4 by the fi nesse 7 d4 c x d4 8 tD xd4 g6 (8 . . . �g4 is probably the reason why Wh ite prefers 7 0-0, but even after the exchange of queens the game is not drawn : 9 �xg4 tD xg4 10 tD xc6 bxc6 11 Af4 etc .) 9 f3 Ag7 1 0 Ae3 0-0 1 1 0-0 �d8 1 2 tD de2 ( 1 2 �d2 is more log ical , but Black is alright after 1 2 . . . �a5) 12 . . . �a5 13 !'k1 a6 14 tDf4 (14 a4 transposes i nto the Kram n i k - Gelfand game, analysed below, p. 164) 14 . . . tD d7 15 a3 ?!

1 e4 c5 2 �f3 d6 3 Ab5+ Ad7 I

analysed this variation i ntensively du ring my candidates match with Vladimir Kram nik in 1 994, si nce when it has served me wel l for many years. This has happened with me (an d , I wou ld guess, not only with me) many times - the level of preparation be fore and during such matches is so intensive and deep, that you can use such analysis for many years to come. 3 . . . tD d7 and 3 . . . tD c6

15 . . . e6 !

A deep strategic idea: Black sacrifices the d6 pawn , but i n return he keeps the wh ite knight out of the im portant d5 sq uare and gains some tem pi for cou nterplay. Th is re m i nds me very m uch of the Ki n g's Indian Samisch Variation with 6 . . . c5 . Some famil iar ideas can be seen in the notes to the game Gheorghiu - Gelfand (No. 14, p. 70) .

163


16 'tWxd6? (accepting the Greek gift; 1 6 �d3 was safer and better) 16 . . . � de5 17 'tWc5 gfc8 ! 18 � ce2 (18 gfd1 'tWxc5 19 A xc5 � xc4 +) 1 8 . . . 'tWa4 ! (18 . . . 'tWxc5 19 A xc5 � a5 !? +) 19 Af2 ( 1 9 �c3 'tW b3) 19 . . . Af8 20 'tWe3 � xc4 and Black stood better in Shi rov- Gelfand , (Akiba Rubi nstein M emorial , Polan ica Zdroj 2000).

7

• . •

his q ueen . I n my fi rst game in this line I failed to play well and was deserved ly pun ished : 1 2 . . . gfc8 13 b3 'tWd8 14 � h1 �d7 15 Ag5 'tWa5 16 'tWd2 �c5 17 gab1 e6 18 gfd1 (18 'iWxd6? � x b3 1 9 g x b3 'tWxg5 20 gx b7 �a5) 1 8 . . . gab8 1 9 A h4 ! 'tW b6 (19 . . . AfS 20 f4 ;t ) 20 'tWxd6!?

g6 8 d4 cxd4 9 �xd4 Ag7 a

b

c

d

e

f

6 5 4 3

20 . . . Ae5 ?

2

L...-_______---=-__--'

'if

10 �de2 I n Tkachiev - Gelfand , (F I DE World Cham pionsh i p , G ron ingen 1 997) the rare m ove 10 � c2 led to a d raw after 10 . . . 0-0 11 'tWe2 a6 1 2 Ad2 gab8 13 gac1 b5 ! 14 c x b5 a xb5 15 �d5 (15 � xb5 � xe4 16 'tWxe4 gxb5 17 b4 ga8) 1 5 . . . e6 1 6 � cb4 exd5 1 7 � x c6 gb6 1 8 exd5 � xd5 1 9 'tWf3 %-% .

10 . . . 0-0 The very concrete 10 . . . 'tWe6 became popular after the Kasparov -The World game in 1 999.

11 f3 a6 1 2 a4 A Mar6czy-style position has arisen where Wh ite has some space advantage. How ever, Black has exchanged his worse piece in such a structure - his light-square bishop, so I believe that his position is very secure here.

12

. . .

•d8!?

The m ost precise move. In the fi rst in stance Black must i m prove the position of

I m m ediately after the game I fou nd the strong sacrifice 20 . . . gd8 ! (20 . . . � xb3? 21 a5) 21 Axd8 (21 'tWg3 !? H ickl) 21 . . . gxdS, and now : A) 22 'tWg3 � d3 23 gf1 {23 � d5 gxd5 (23 . . . �f2+ 24 � g1 � h3+ 25 � f1 ) 24 cxd5 �f2+ 25 'tWxf2 'iWxf2 26 d x c6 b x c6 27 �f4 A h6 =) 23 . . . 'tW b4 24 �d1 (24 � d5 exd5 25 cxd5 �a5 oo ; 24 � a2 'iWd2 +) 24 . . . 'tWd2 with counterplay; B) 22 a5 'iWx a5 23 b4 gxd6 24 bx a5 gxd1 + 25 � x d 1 (25 gxd1 � x a5 26 gd8+ Af8 +) 25 . . . � xa5 with fu ll compensation for the ex change. 2 1 'iWd2 � x b3 22 'tW b2 ! (Vlad i m i r gains the i n itiative with a beautifu l combina tion) 22 . . . � ca5 23 � d5 ! exd5 (23 . . . Axb2 ? 2 4 � x b6 gc6 25 � d7) 2 4 'iWx e5 � xc4 (24 . . . dxe4 25 �f4 !? � xc4 26 �d5 !) 25 'tWxd5 (25 'tW f4 !? oo Kramn i k) 25 . . . 'tWe3 ? {25 . . . �e3 ! 26 'tWd3 (26 'iWd6 'tWxd6 27 gxd6 �c5 28 �c3 !? ;t ) 26 . . . � x d1 27 'tWxd1 'tWc6 ! 28 'tWx b3 'iWc2 29 � d4 !? ;t ) 26 ge1 ! ± (Kram nik- Gelfand , F I DE Candidates Quarter-Final, Sang h i Nagar 1 994(3)) .

164

1 3 Ae3 .a5 14 �h1

Game 37

Rublevsky - Gelfand , Akiba Rubinstein Memorial Tournament, Polanica Zdroj 1998

Black d rew comfortably after 14 gc1 �d7 1 5 b3 �cS 16 � a2 gac8 17 g b1 b5 !? 1 8 c x b5 (18 b4 't'fx a4) 1 8 . . . a x bS 19 ax b5 �xb5 20 �ac3 't'fd3 ! 21 �d5 gb8 ! = (Kram nik- Gelfand , F I DE Candidates Quarter-Final , Sanghi Nagar 1 994(7)) .

14 . . . IUd8? Faced with a novelty, I i m mediately made a mistake. The rook on d8 takes away an important square from the q ueen . 14 . . . �d7 15 gb1 has been tested a few times in prac tice. Wh ite keeps a slight edge, but Black has a safe position . 14 . . . gac8 was the nor mal move, as 1 5 � d5 �d7 1 6 b4 doesn 't work because of 16 . . . 't'fd8, and the c4 pawn becomes really weak. Worthy of serious consideration was a pawn sacrifice, sim ilar to my game with Shirov : 14 . . . e6 !? 1 5 't'fxd6 gfd8 1 6 't'fa3 (16 't'fc5 �d7) 16 . . . gd3 <55 . In fact I found this idea dur ing an analysis of this game and I was happy to put it into practice in a simi lar position two years later.

15 ctld5! 15 �b3 � b4, or 15 b3 e6 intending . . . d6-d5.

15 . . . ctlxd5 Unfortunately, this is forced , as bad is 15 . . . �d7 16 b4 � x b4 (the rook has taken away the d8 square from the queen) 17 Ad2.

16 exd5 ctle5

tt:J

Wh ite has to prevent . . . b7-bS and . . . e7-e6 if he wants to keep an advantage. He needs to play precisely and energetically to do this.

17 b3?! With the stronger 17 Ad4! (if 17 't'f b3 � d7) Wh ite could have consolidated his advan tage after 1 7 . . . 't'fb4 : A) 18 gc1 and now: A1 ) 18 . . . A h6 1 9 f4 � xc4 20 b3 (or 20 Ac3 !? �c5 21 b4 �c8 22 a5 ! (22 �d4 f6 23 �e4 !? e5 24 dxe6 d5) 22 . . . gd7 23 �d4 f6 24 �d3 ±) 20 . . . � b6 (20 . . . �aS 21 Ac3 't'f b6 22 't'fd3 ! (th reaten in g b3-b4) 22 . . . gac8 23 't'f h3 ± with the threat of A x a5) 2 1 Ac3 't'fc5 (21 . . . �e4 22 Aa5 � xd5 23 �g3 ±) 22 't'fe1 ! (22 Ag7 Axg7 23 gxc5 dxc5 ; 22 Ad4 � b4 =) 22 . . . � xd5 23 A aS ± , and Black hard ly has sufficient com pensation for the exchange, as control of the c -fi le is very important ;

A2) 1 8 . . . gac8 1 9 b3 �d7 2 0 A x g7 xg7 21 �d4 �c5 22 f4 e5 23 d xe6 fxe6 ;J;; ; A3) 1 8 . . . bS 1 9 Ac3 (1 9 a x bS a x bS 20 Ac3 't'fa4 21 b3 't'fa6 =) 1 9 . . . 't'fx a4 (19 . . . 't'fc5 20 c x b5 , and if 20 . . . a x bS 2 1 A x eS) 20 b3 't'fa3 21 cS ! with the th reat of c5-c6, and Black's q ueen is in danger - 21 . . . d xc5 22 ga1 ; A4) B lack has to settle for a slightly worse position after 18 . . . �d7, as he cannot free hi mself by tactical means; B) 18 b3, when :

8

6 5 4 3 2

�------�--� �

B 1 ) Black can not free h imself by 1 8 . . . b5 19 f4 �g4 (19 . . . �d7 20 Axg7 xg7 21 �d4 !) 20 A x g7 x g7 21 't'fd4+ (21 �d4 't'fcS ! (21 . . . � e3 22 't'fe1 ) 22 �e6+ fxe6 23 't'fxg4 exd5 24 't'fe6 d xc4 25 't'fx e7+ g8 26 f5 't'feS 27 't'f h4 gf8 00 ) 21 . . . �f6 because of 22 g4 ! (22 c x b5 't'fxd4 (22 . . . 't'fx b3 23 b6 't'fxd5 24 a5 (55 ) 23 � xd4 � xd5 with coun terplay) 22 . . . g8 23 g5 �d7 24 cxb5 �xb3 25 gab1 ± ;

165


B2) 18 . . . tLl d7 ! 19 A xg7 � xg7 20 tLl d4 tLl c5 21 f4 e5 22 d x e6 fxe6 t with a defensi ble position , although White stands clearly bet ter.

17 . . . �d7 18 Ad4 After 18 �b1 b5 19 �d4 Axd4 ! 20 Axd4 bxc4 21 bxc4 �ac8 the weak a4 and c4 pawns give Black good cou nterplay i n exchange for his dark-sq uare bishop. As White cannot gain an attack here, Black is alright. 18 . . . �f6 After 1 8 . . . A xd4 1 9 tLl xd4 ± Wh ite wou ld achieve his ideal set-up - he has prevented both . . . b7-b5 and . . . e7-e6 and he can choose between increasing the pressure on the e -fi le and a d i rect attack on the king by f3-f4-f5 .

19 0d3 Probably slightly better was 1 9 Ac3 'ffIc7 20 a5 �e8 (20 . . . e6) 21 A d4 e6 , or 20 'ffId 3 e6 21 d x e6 fxe6 t and although Black has managed to break the grip by . . . e7-e6, his pawns are vulnerable, which gives Wh ite the edge.

B) 20 Ac3 - cf. 1 9 Ac3 ; C) 20 dx e6 fxe6 and now : C1 ) 21 'ffIe3 (21 Ac3 !?) 21 . . . �e8 22 A b6 (22 �fd1 d5) 22 . . . 'ffI h 5 23 �ad1 d5 24 tLlf4 '@'h6 25 cxd5 g5 with counterplay; C2) 2 1 tLlf4 �e8 22 �fe1 (22 �ad1 '@'c7) 22 . . . e5 (22 . . . � f7 23 �ad1 ) 23 Ac3 '@'c5 24 tLl h3 d5 (24 . . . b5 25 ax b5 a x b5 26 �xa8 �xa8 27 c x b5 �b8 28 'ffIc 4+ tLld5 29 Axe5! 'ffIx c4 30 bxc4 A x e5 31 cxd5 �x b5 32 �d1 +-; 24 . . . �ac8 !?) 25 A x e5 dxc4 26 '@'xc4+ '@'xc4 27 bxc4 tLld7 (27 . . . �ac8!?) 28 Axg7 � xg7 t .

20 �c3?! Wh ite centralises his pieces. Black has coun terplay after 20 �fe1 (20 Ac3 !? - cf. 19 Ac3) 20 . . . e6 21 d x e6 �xe6 (21 . . . fxe6 22 tLlf4 cf. 1 9 . . . e6) 22 Ac3 'ffI b 6 23 tLl d4 �xe1 + 24 �xe1 d5 25 �e7 �c8, or 20 �ae1 '@' b4 (20 . . . e6 21 dxe6 �xe6 22 �d1 ), but stronger was 20 b4 ! 'ffIc7 (20 . . . 'ffIx b4 ? 21 �fb1 '@'a5 22 A b6) 2 1 a5 �ac8 22 �ac1 e6 23 dxe6 fxe6 (23 . . . �xe6 24 tLlf4) 24 �fd1 e5 25 Ab6 'ffIf7 !? t (25 . . . e4 26 fxe4 tLl xe4 27 tLlf4 i).

19 . . . Iie8

20 . . . e6 21 dxe6 fxe6 e

f

h

21 . . . �xe6, i ntending to double rooks on the e -fi le, deserved serious consideration, as from c3 the wh ite knight cannot attack the rook at e6 and the d6 pawn cannot be con sidered weak, as no piece can attack it. After 22 f4 (22 �ad1 �ae8) 22 . . . 'ffIf5 !? 23 '@'xf5 gxf5 24 �ad1 tLlg4 Black has sufficient coun terplay on the e -file.

8 7 6 5

5 4

3

3

2

2

22 Iiad1 22 tLle4 tLl xe4 23 fxe4 Ae5 ! = .

22 ... Iiad8 23 Oe3 Black can open the game i mmediately with 19 . . . e6, but this still does not solve his prob lems completely: A) 20 b4 'ffIc7 21 a5 exd5 22 A b6 (22 cxd5 �e8) 22 . . . 'ffIx c4 23 'ffIx c4 d x c4 24 A xd8 �xd8 (55 ;

An alternative was 23 �fe1 'f!! b 4 ! (here too Black must play actively and not be afraid of losing his queen ; if 23 . . . d5 24 cxd5 tLl xd5 25 A x g7 � x g7 26 tLl e4 ±) 24 tLla2 (24 Axf6 A xf6 25 tLl e4 Ae7) 24 . . . 'f!!a 3 (24 . . . '@'a5 25 'f!!e3 �d7 26 tLl c3 t) 25 'f!!c 2 b5 ! (25 . . . d5

166

Game 37

Rublevsky - Gelfand, Akiba Rubinstein Memorial Tournament, Polanica Zdroj 1998

26 b4 ±) 26 ax b5 (26 A b2 �c5) 26 . . . a x b5 27 cxb5 e5 with counterplay. 23 tDe4 tD x e4 24 A xg7 tD g3+ (24 . . . tD c5 25 Ac3) 25 h xg3 �h5+ 26 � g1 � xg7 leads to an equal position .

23 . . .• b4!

t[)

A 1 ) 26 A b2 �xa2 (26 . . . �c5 27 �xc5 dx c5 28 gxd8 gxd8 29 A x e5 t tD h 5 ! 30 A x g7 tD x g7 31 tD c3 gd3 32 gc1 tDe6 with coun terplay) 27 ga1 A h6 28 �xh6 �x b3 ;

A2 ) 2 6 A b6 gd7 2 7 Aa5 �c5 2 8 �xc5 dxc5 29 gxd7 tD xd7 30 gd1 t ; B) 25 . . . d5 26 Ac5 , with these possibilities :

8

B 1 ) 26 . . . d4 27 A xd4 ± (27 �x e6+ gxe6 28 Axa3 t ) ;

8

7

B2) 26 . . . A h 6 ! (the key move o f Black's defence, enabl ing h i m to maintai n eq ual chances) 27 �xh6 (27 �d4 �xb3 (or 27 . . . e5 28 A x a3 exd4 29 gxd4 d x c4 30 gxc4 tDd5 with cou nterplay) 28 �xf6 �x c4 ! (28 . . . A xd2 !? 29 A d4 e 5 3 0 A x e5 gxe5 31 �xd8+ �f7 =) 29 gc1 �f4 =t , or 27 f4 tDg4 28 �xe6+ gxe6 29 Axa3 tDe3 with counter play) 27 . . . �xc5 = ;

6 5 4 3

3

2

2

�------�--�

�

It req u i red precise calcu lation and confi dence to put the queen i n a dangerous place (d uring a practical game, when you are pressed for time, you can never be sure that you have not missed somethi ng) . How ever, i ntu itively I felt that Black should have enough resources to save his q ueen , which i n the meantime wi ll disrupt the coordi nation of the wh ite pieces.

24 �a2 Or 24 gb1 d5.

C) as indicated by the computer i n 2003 , Black has the strong reply 25 . . . b5 ! 26 A b2 �c5 (or 26 . . . tDd5!?) and he is at least equal .

25 . . .•xa2 26 gd2 Not 26 gf2 ? tDe4 27 fxe4 �xf2 , or 26 A xf6 A xf6 27 �xf6 �x b3 28 g b1 (28 c5 d5) 28 . . . �xa4 29 gx b7 gd7 =t , but also possible was 26 ga1 tDd5 ! 27 cxd5 (27 �d3 �x g2+ 28 � xg2 tDf4+) 27 . . . A xd4 28 �xd4 �x b3 29 gab1 �xd5 30 �xd5 exd5 31 gx b7 with a probable draw.

26 . . .•a3!

24 . . .•a3 Around here Sergey spent a lot of time, try ing to find a way to trap the queen . It cost him dearly, as he lacked this time later on.

25 .c3 More chal lenging was 25 gd2 , when it wou ld have been more d ifficult for me to find the correct way. The l ines given below are obvi ously the result of home analysis; during the game both of us merely saw some ideas, but obviously not all the l ines : A) 25 . . . e5 , and now :

Diverting the opponent's pieces from the centre. Now Black has an easy game and it is Wh ite has to play accurately to make a d raw. I n fact, this is a very instructive mo ment - psychologically it is difficult to switch from having an advantage to playi ng pre cisely t o make a draw. Rublevsky under stood this wel l , but he failed to cope with it, most probably due to lack of time. Less good was 26 . . . tDd5 27 cxd5 A xd4 28 gxd4 exd5 (28 . . . gc8 29 gc4 t ) 29 gxd5 �e2 =, or 26 . . . tDe4 27 fxe4 Axd4 28 gxd4 t .

167


27 11a1 �dS 27 . . . �e4 28 fxe4 A xd4 29 gxd4 ;t and it is Wh ite who is pressing.

28 1td3 2 8 cxd5 A xd4 29 gxd4 gc8 ! =+= (a strong i ntermediate move, turning the tables in Black's favour), and if 30 �xc8 �xa1 + .

28 . . . �f4 29 11xa3 �xd3 30 l1xd3 dS So, an equal ending has arisen . H owever, due to the poor position of the rook at a3, Black is ' more equal ' .

31 Axg7 Also possible was 31 A b6 !? gc8 32 c5 a5 33 ga2 Af8 34 gc2 =.

31 . . . �xg7 31 . . . dxc4 32 bxc4 =.

34 gxd8 gxd8 35 b x c4 gc8 36 g b1 gxc4 37 gx b7+ =.

33 ... � 34 exdS? Choosing the wrong moment for this move. 34 �f2 d4 35 b4 ! (White was afraid of having to play the i nferior end ing after 35 ga2 e5 36 ge2 g5 37 ge4 � f5 38 h4 h6 =+=) 35 . . . gd7 36 ga1 gc8 37 gc1 axb4 38 gb3 e5 39 gx b4 should be enough for a draw. 34 . .

.

flxdS 3S l1e3

Now 35 b4 ax b4 36 gxd5 bxa3 37 gd3 ga8 38 gxa3 is bad because of 38 . . . b5 39 a5 � e5 40 a6 � d4. 35 gxd5 exd5 36 b4 doesn 't solve Wh ite's problems either after 36 . . . ge2+ ! 37 �f1 gb2 38 bxa5 d4 =+= and his pieces are too passive, so that, despite bei ng a pawn u p , he is in trouble.

3S . . . l1d2+ 36 �h3 11ed8 37 fla1

32 g3

..

32 � g1 a5 33 �f2 = was better, bringing the king closer to the centre.

8

8

7

32 . . . aS a

1-1'---"

b

c

d

6

6

5

5

e 8

8

4

7

3

6 5 4

2

a

33

Finally the rook comes back i nto play, but now the king is Black's target.

2

b

c

d

e

f

9

h

37 . . . 118dS!

�

A strong intermediate move, provoking more weaknesses. If 37 . . . gb2 38 f4 gdd2 39 gh1 .

�g2?!

Th is natural move creates problems for Wh ite. He had to play concretely, but he was already in time trouble : 33 cxd5 gxd5 34 b4 a x b4 35 gxd5 bx a3 (35 . . . exd5 36 g b3) 36 gd3 a2 37 ga3 ga8 38 gxa2 b5 39 a5 = and Wh ite succeeds i n exchang ing his a pawn for the b - pawn , or 33 ga1 d x c4

38 g4 l1b2 39 fle3? 39 gh1 gdd2 40 ge3 g5 leaves Wh ite completely paralysed . The last chance was 39 � g3 gdd2 40 h4 gg2+ (40 . . . e5 41 gg1 ) 41 � f4 g be2 42 g5+ � e7 43 ge3 gxe3 44 � x e3 g b2 45 gc1 gx b3+ 46 � e4 � d6, or 42 ge3 e5+ 43 � e4 gd2 44 gee1 gd4+

168

Game 38

tt:J

Gelfand - Markowski , Akiba Rubinstein Memorial Tournament, Polanica Zdroj 1 998

45 � e3 h 5 ! with a serious advantage for Black in both cases. However, White would retain some drawing chances.

a

39 . . . Ddd2 40 Dae1

b

c

d

...

e

8

8

7

7

6

6

5

40 Eth1 g5 -+.

4

40 . . . llxh2+ 41 �g3 Dbg2+ 42 �4 g5+ 43 �e4 Dd2! Cutting off the king from the queenside and starting a new wave of the attack on it.

3

3

2

2

44 Dd3 47 . . . Dh4+

If 44 f4 Etd5 intending . . . Ethd2 .

44 . . . llxd3 45 �xd3 h5! -+ The black rook secures a post on the 4 t h rank, from where it will attack t he b 3 pawn.

Avoiding the last trap : 47 . . . Eth3 ? 48 � b5 Etxf3 49 � x a5 Etx b3 ? (49 . . . g4) 50 Etxe6+ � f5 51 Ete5+ � f4 52 Etf5+ ! � g4 53 Etxg5+ � xg5 with stalemate.

48 �e5 Db4 49 lle3 �5 50 �d6 �4 51 lle3 e5 52 �d5 e4 53 fxe4 Dxe4 White resigns

46 gxh5 Dxh5 47 �e4 47 Etc1 Eth3 48 � e2 Eth4. * * *

Game 38

B o r i s G e l fa n d - To m a s z M a r kow s k i Ak i ba R u b i nste i n M e m o r i a l To u rn a m e n t , P o l a n i c a Zd roj 1 99 8 King 's Indian Defence [E94] I think that this was one of the best tourna ments in my career, not only from the com petitive point of view (I won this strong event, a point ahead of Alexey Shirov) , but also cre atively. That is why I am presenting a third game from this tournament. After the Wij k aan Zee Corus Tournament in January, I had a com plete lack of invitations and played only fou r tournament games before I was invited to Polanica Zdroj just a few weeks before the start of the tournament. So, my motivation was extremely high.

i n 1 998 and 2000. On both occasions it was n ice to return to form after a series of mediocre resu lts.

1 d4 d6 2 �f3 �f6 3 e4 g6 4 �e3 J.g7 5 e4 0-0 6 J.e2 �a6 7 0-0 e5 8 De1 8 Ae3 is another variation , one which I have played with both colours.

Over the course of three memorial tourna ments of my favou rite player Akiba Rubin stein i n 1 997 , 1 998 and 2000 I d i d n 't lose a single game and I won this tournament

169

8 . . . e6 9 J.f1 exd4 10 �xd4 �g4 11 h3 'ftb6 12 hxg4 'ftxd4 a

b

c

d

e 8

6

6

5

5

4

4

3

3 2

L...-_______----''---_--'

1J


13 .f3 13 g5 tLlc5 14 'tWf3 'tWe5 15 Af4 'tWe7 1 6 gad1 Ae5 1 7 'tWg3 is the regular move order, but I wanted to limit Black's choice, as other 13 th moves are possible. However, a few years later I tried 1 3 . . . f5 ! ? (13 . . . '�xd1 !?) with the black pieces, but without success : 14 Af4 ! fxe4 15 'tWxd4 Axd4 1 6 gxe4 A xc3 1 7 bxc3 tLl c5 1 8 gd4 tLl e6 1 9 A xd6 tLl xd4 20 A xf8 tLl c2 21 gc1 tLl e3 22 fxe3 � xf8 23 e4 ! (Wh ite's pawn structure looks strange and I thought that Black was fine despite bei ng a pawn down , but Wh ite can gain pressu re on the b - and h -fi les and his king q u ickly reaches the centre after 23 . . . Ae6 24 � f2 gd8 25 e,!;> e3 e,!;> e7 26 c5, or 23 . . . Ad7 24 gd1 (24 gb1 !? b6 25 c5) 24 . . . e,!;> e7 25 e,!;> f2) 23 . . . a5 (trying to fix the pawn on a2 and bring the rook into the game via a5 ; however, this idea appears to be ineffective) 24 e,!;> f2 a4 25 e,!;> e3 a3 26 g b1 ga5 27 � f4 � g7 (27 . . . � e7 28 Ae2 e,!;> d6 29 gh1 ) 28 Ae2 and Wh ite con verted his advantage i nto a win (Kramn i k Gelfand, Melody Amber, Monaco blind 2001 ).

17 . . . f6 1 7 . . . tLl e6 1 8 A x e5 d xe5 1 9 'tWx e5 'tWxg5 20 'tWxg5 tLlxg5 ;t; (Krasenkow - Kaula, Poland 1 991 ).

18 gxf6 Hxf6 1 8 . . . 'tWxf6 1 9 A x e5 d x e5 20 b4 tLl a6 21 a3 tLl c7 22 c5 ;t; (Komarov - Di m itrov, Yugoslav Team Championsh ip, N i ksic 1 997) .

19 Axe5 .xe5 1 9 . . . d x e5 20 b4 tLl d7 21 c5 � g7 22 gd2 gf8 23 ged1 tLlf6 24 gd6 tLl h5 25 'tWe3 e,!;> h8 26 b5 ± (Shariyazdanov-Alexi kov, Swidnica Open 1 997) .

20 b4 a

5 4

d

.. 8

7 6

6

5

5

4

4

3

3 2

..

f

6 5

1-___......

I

4

....--,..;: ... ;;;; :=. ;.,

3

3

2

a

e

e

-

This position had already been tried in a few games, where Wh ite was successfu l . My op ponent goes for the most testing line . c

d

6

1 6 Af4 Ae5 17 Had1

b

c

8

1 3 . . .•e5 14 g5 .e7 1 5 .g3 �c5 15 . . . A e5 1 6 f4 (16 Af4 f6 !?) 1 6 . . . Ag7 is the main possibil ity of diverg i n g from the wel l known path. I am sure i t will b e tested i n the future.

a

b

b

c

9

h

20 . . . •xg3? Th is is the main moment in the game. What could be more natural than to exchange queens and double Wh ite's pawns? How ever, it leads to a very u npleasant position. Stronger was 20 . . . tLl d7 ! 2 1 'tWe3 (21 'tWxe5 tLl x e5 22 c5 d x c5 23 b x c5 b5 ; 2 1 c5 'tWxg3 22 fxg3 - see the note to White's 22 n d move) 21 . . . a5 22 b5 (22 a3 g5) 22 . . . tLlc5 with coun terplay.

21 fxg3 �d7 (see next diagram)

170

Game 38

a

b

Gelfand - Markowski , Akiba Rubinstein Memorial Tournament, Polanica Zdroj 1998

c

d

tt:J

24 . . . �g7 25 Ac4 .6e7 26 a4

e

As none of Black's pieces can move, his only idea is to play . . . b7-bS or . . . b7-b6, so I de cided to prevent the fi rst one. 26 !3.d6 bS (26 . . . as !? 27 bS b6 2B cxb6 ttlx b6 29 Af1 ±) 27 A b3 A b7 28 !3.ed1 tDf8 29 !3.d8 also looks fine, as it is hard to imagine how the black bishop can come back i nto play.

8 7 6

6

5

5

4

4

3

3

a

2

b

c

d

...

e 8

..... 11

7

L...-_______---"'-__

22 e5!! Th is is an excellent example demonstrating that the activity of the pieces is more impor tant than pawns, even i n a sem i-endgame. 22 cS ?! looks tempting, but after 22 . . . ttl eS ! (22 . . . d x cS 23 eS EUB 24 e6 is virtually win ning for White) 23 cxd6 Ag4 24 Ae2 A x e2 2S !3.xe2 !3.dB although White is a pawn u p , the advantage i s with Black, since h i s knight at eS controls the entire centre, whereas the knight at c3 is out of play. After 22 !3.d2 ttleS 23 !3.ed1 � fB 24 !3.xd6 !3.xd6 2S !3.xd6 � e7 the assessment is simi lar to 22 cS .

22 . . . dxe5

24 bS !?

5

5 4

3

3

2

2

26 . . . .6e8? Black gives up the game. M uch more stub born was 2 6 . . . b 6 2 7 !3.d6 b x cS (27 . . . A b7 2B as) 2B b x cS ttlfB 29 !3.xc6 A b7 30 !3.d6 !3.cB 31 AdS A xdS 32 !3.xdS ± .

27 .611 .6e7 28 .6d2 b5 After 2B . . . aS 29 !3.df2 White switches to a mating attack : 29 . . . a x b4 30 !3.f7+ !3.xf7 31 !3.xf7+ � h6 (31 . . . � hB 32 ttlgS) 32 g4 gS 33 Ad3 !3.xa4 34 ttld6 +-.

Or 23 . . . !3.fB 24 ttld6 ttlf6 2S !3.xeS.

24 c5

6

Or 27 . . . !3.fB 2B !3.xfB � xfB 29 !3.f1 +.

Black takes the paw n , so that at least he has something for his sufferings. 22 . . . ttl xeS 23 tDe4 !3.f8 24 tD xd6 and Wh ite is much bet ter after 24 . . . ttlf7 2S ttl xc8 !3.axcB 26 !3.d7.

23 �e4 .6f7

6

29 axb5 cxb5 30 Ad5 Black resigns, as after 30 . . . !3.b8 31 !3.df2 he simply has no moves.

***

171

My Most Memorable Games Game 39

B o r i s G e l fa n d - J o e l L a u t i e r S i g e m a n & Co . , M a l m o 1 9 99 Semi-Sla v Defence [04 6] After winning an extremely strong World Ju nior Championsh i p (on tie - break ahead of Ivanchuk, Serper and myself) at the age of 15, Joel Lautier became a real star in France and was very welcome in a n u m ber of big events. He is one of the few players who has a positive score agai nst Garry Kasparov. According to statistics, he has a disastrous score agai nst me, but, to tel l the truth, I have many times had to display m i racles i n de fence, and when this was not enoug h , l uck was o n my side. I have won a number of nice games against the Frenchman , including a six and a half hour battle in the last round of the I nvestbanka Tou rnament i n Belgrade 1 995, which enabled me to catch u p with Vladimir Kram nik and win a beautiful cup on the tie-break! I also have good memories of the charm ing city of Malmo and a tourna ment where all the five games I won were de cided by mating attacks on the opponent's king .

1 d4 d5 2 c4 c6 3 �f3 �f6 4 �c3 e6 5 e3 �bd7 6 .c2 Ad6 7 Ad3 0-0 8 0-0 a

b

c

d

e

...

h 8

7 6

6

5

5

4

4

3

3

8 . . . dxc4 9 Axc4 a6 (or 9 . . . b5) against Ana toly Karpov. 8 . . . e5 9 cxd5 cxd5 10 e4 ! dxe4 (10 . . . exd4 11 etJ xd5 etJ xd5 1 2 exd5) 11 etJ xe4 etJ xe4 1 2 Axe4 gives Wh ite sl ight pressure in a symmetrical position . 8 . . . �e7 ?! would al low 9 c5 Ac7 10 e4, and if 10 . . . e5 (10 . . . dxe4 1 1 tD x e4 h6 1 2 �e1 leaves Black's position too cram ped) 11 exd5 cxd5 12 Ag5.

9 b3 Pointless now is 9 c5 Ac7 10 e4 on account of 10 . . . e5 with good counterplay, as Ag5 is not possible. 9 Ad2 was played in the game Gurevich - Gelfand (No. 46, p. 200).

9 . . . b6 9 . . . e5 10 cxd5 cxd5 11 tD b5 A b8 1 2 dxe5 tD x e5 1 3 tD x e5 A x e5 14 A b2 ;t leads to a typical position, where Black is condem ned to passive defence.

10 Ab2 Or 10 cxd5 cxd5 11 tD b5 Ae7 .

10 . . . Ab7 1 1 llac1 Wh ite is faced with a common decision how best to place his rooks. It was difficult to foresee how the game would progress Black has a number of plans - so decisions are normally made (as also in this case) on the basis of experience and intuition . A little more than a month later, I faced Joel in the F I DE World Championsh ip (knock-out) i n Las Vegas and i n our tense match this position occurred agai n three times ! 11 �fd1 �e7 1 2 �e2 (1 2 c x d 5 cxd5 13 etJ b5 �fc8 14 �e2 A b8 15 Aa3 �d8 16 �ac1 �xc1 17 �xc1 etJe8 18 g3 a6 19 etJ c3 Ad6 20 Axd6 tD x d6 2 1 tDe5 �c8 also does not promise m uch , Topalov - Ivanchuk, M elody Amber, Monaco rapid 1 997) 1 2 . . . �fe8 13 cxd5 exd5 14 A a6 tDf8 15 �ac1 tD g6 16 A x b7 �xb7 17 �c2 �e6 18 �dc1 and a draw was agreed as Wh ite has no promising plan (Gelfand Lautier, 2 n d game).

2

8 . . . h6!? A usefu l move. Black avoids reveal ing his cards. As Black I have several times chosen

On ly i n the tie-break (rapid) after 1 1 �ad1 !? �c7 1 2 e4 dxe4 13 etJ xe4 etJ xe4 14 Axe4 did

172

Game 39

Gelfand - Lautier, Sigeman & Co. , Malmo 1 999

I manage to get close to an open ing advan tage: A) 14 . . . gad 8 1 S cS ? (premature ; 1 S gfe1 is better) 1 S . . . Ae7 16 b4 A a6 ! 17 gfe1 ttlf6 18 ttleS ttl x e4 1 9 �xe4 gdS + (Gelfand Lautier, 4 th game) ; B) 14 . . . Ae7 1 S gfe1 gfe8 1 6 �c3 gad8 17 A b1 Af6 18 ttl eS cS 19 �c2 ttlf8 20 f4 cxd4 2 1 A xd4 gd6 22 cS b x cS 23 �xcS �xcS 24 A x cS gxd1 2S gxd1 a6 26 b4 ;t (Gelfand - Lautier, S t h game).

1 3 d5! The most daring move and in my opinion a strong pawn sacrifice.

1 3 . . . exd5 After 13 . . . �eS 14 ttlxeS AxeS 1S e4 White is clearly better thanks to his strong centre and the bad A b7. 14 �xd5 �xd5 1 5 cxd5 Axd5?! Accepting the challenge. Black could have tried to defend an i nferior position after 1S . . . �e7 16 e4 ttleS ;t .

11 . . . dxc4?!

1 6 1lfd1

Possibly i nspired by the i nert nature of Wh ite's last move, J oel carelessly opens up the centre. After 11 . . . cS 12 cxdS ttl xdS 13 �xdS AxdS 14 A h7+ � h8 1S Ae4 !? Axe4 16 �xe4 ;t White has unpleasant pressure. It is understandable that Lautier should avoid this position , as he lost a similar position (after the unavoidable exchange of the d and c -pawns) t o me i n o u r first game i n the Manila I nterzonal , 1 990.

Due to the overcrowd ing of Black's pieces on the d -file, he is forced to exchange his i mportant bishop, opening the g -file for the wh ite rooks.

16 . . . Axf3 U n satisfactory was 1 6 . . . ttlf6 1 7 A xf6 (or 17 e4 c4 18 exd5 cxd3 19 �xd3 ± Shipov) 17 . . . �xf6 1 8 A h7+ � h8 19 gxdS g6 20 gcd1 Ac7 21 gd7 gac8 22 Axg6 �xg6 23 �c4 ± .

17 gxf3

11 . . . gc8 was better, and only after 12 �e2 12 . . . d xc4 13 bxc4 cS :

... 8

A) 14 dS (here this is not so strong) 14 . . . exdS 15 cxdS � xd5 16 � xd5 Axd5 17 e4 (17 gfd1 Axf3 1 8 �xf3 �e7 1 9 �fS g6 20 � h3 AeS 21 Axg6 fxg6 22 gxd7 �f6) 17 . . . A b7 ;

,'""""-='-1 ""

7 6

B) 14 gfd1 cxd4 1 S exd4 with chances for both sides.

12 bxc4 c5

5

5

4

4

3

3 2

8 7 6 5 4 3

L...-_______-=--_----'

1I

Now it is clear why Wh ite sacrificed the pawn : his two bishops are exerting strong pressure on the black king and they are go ing to be joi ned by a rook on the g -file. The pieces on the d -file can easily become targets for attack and for the moment the queenside pawn majority is not i m portant, as all the events wi ll take place on the oppo site flank.

173


17 . . . "c7?! B lack is hoping to transfer his bishop onto the long d iagonal at e5 or f6 (via e7) , but he probably missed Wh ite's next move. 17 . . . '�e7 was much more stubborn : A) 1 8 f4 tDf6 1 9 � h1 � b7+ ; B) 1 8 � h1 al lows 1 8 . . . Ae5 ! 1 9 A h7+ � h8 20 gxd7 �xd7 21 Axe5 f5 ! 22 gg1 (22 Ag6 �d5) 22 . . . � x h7 23 gxg7+ �xg7 24 A xg7 � xg7 and Wh ite has to fight for a draw ; C) 1 8 Ac4 ! , and now: C1 ) 18 . . . tDe5 (18 . . . Ae5 ? 19 gxd7), with these possibilities : C 1 1 ) 1 9 Ad5 (the most natural , but not the best) 19 . . . gad8 20 � h1 (20 f4 tDg4 21 �g6 tDf6 22 �g3 (22 �g2 b5 23 � h1 � h 8 24 gg1 gg8 25 � h3 A xf4 , or 22 � h1 Ae5 23 A xe5 gxd5 24 A xf6 gxd1 + 25 gxd1 �xf6) 22 . . . Ac7 with sufficient com pensa tion , but not more) , when : C111 ) 20 . . . � h8 21 f4 tDg4 22 gg1 f5 23 Af3 ! (23 f3 Ae5 ! 24 A xe5 (24 fxg4 A x b2 , or 24 fxe5 tDxe3) 24 . . . tDxe5) 23 . . . h5 24 h3 �h4 25 gg3 followed by 26 � g2 and wins; C 1 1 2) 20 . . . � h4! (ai m i n g for cou nterplay) 21 �c3 (21 gg1 tDg4 22 gg2 Ae5) 21 . . . � h8 (21 . . . b5 22 Aa1 b4 23 �b2 �h5 24 f4) 22 f4 f6 ! with an unclear game; C 1 2) 19 Ae2 ! (only this move enables Wh ite to mai ntai n a dangerous attack ; Black can not gain a tem po by attacking the bishop, as he can with it on d5) 19 . . . gfd8 (1 9 . . . c4 20 �c3) 20 f4 tDd7 21 � h1 tDf6 22 gg1 and it is hard to withstand the pressure on the g -fi le; C2) 1 8 . . . tDf6 1 9 �f5 (19 f4 is premature - 1 9 . . . gfd8 20 �f5 gab8 21 a4 �e4) 1 9 . . . gad8, when :

L-L-..J..:;� ;;; :::.....L. ..I.....J.::::. .::: -..I-l

'if

C21 ) 20 � h1 (20 f4 �e4, or 20 gd3 Ac7) 20 . . . �d7 21 �xd7 (beautifu l but incorrect is 2 1 �xf6 ? gx f6 22 gg1 + Ag3 ! 23 gxg3+ � h7 24 A xf6 gg8) 21 . . . gxd7 22 A xf6 gxf6 23 gd5 gb8 (23 . . . gfd8 24 gcd1 �f8 25 A b5 � e7 26 A xd7 gxd7 27 e4 c4 28 f4 gd8 29 �g2) 24 gcd1 (24 a4 �f8 25 gcd1 �e7 =) 24 . . . b5 ! 25 Ae2 gb6 26 gxc5 Axc5 27 gxd7 and Wh ite has only a sym bolic advantage; C22) 20 e4 ! Ac7 (20 . . . gfe8 2 1 A b5) 21 e5 tD h7 22 f4 and Wh ite stands better, as all Black's pieces are poorly placed and the ex tra pawn is not felt.

18 .tc4! Preventing the manoeuvre of the bishop onto the a1-h8 diagonal (18 f4 Ae7), and thus ob tain ing a decisive attack.

18 . . . gad8 18 . . . gfd8 19 f4 tDf8 20 �f5 ai m i n g at f7 , or 1 8 . . . Ae7 19 gxd7 �xd7 20 �g6.

19 f4 I hesitated between this and another strong move : 1 9 �f5 !? A e5 (1 9 . . . Ae7 20 gxd7 gxd7 21 �g6 gd4 22 Axd4 cxd4 23 Axf7+) 20 gxd7 A x h2+ 21 � g2 gxd7 22 �g4 ! g5 23 � h5 (23 �f5 !? gd6 24 gh1 ) 23 . . . � h7 24 gh1 or 23 . . . gd6 24 �x h2 +-.

19 . . . �h8 Avoiding the pin on the a2-g8 diagonal .

20 "f5 f6 Or 20 . . . Ae7 21 Ad3.

21 �h1

(see next diagram)

174

Game 39

Gelfand - Lautier, Sigeman & Co. , Malmo 1999

The rook joins the attack. There was no im med iate win - 21 Ad3 g6 22 �xg6 !!g8.

21 ... E:lfe8 22 E:lg1 E:le7 23 E:lg6?!

B) 24 Ad3 c4, when : •

8

��--��

7

!!h7 29 �f6+ !!g7 30 � h4+ (30 !!g1 ?! �e7) 30 . . . !!h7 31 �f6+ = ;

B 1 ) 25 !!g2 (25 A b1 � b7+) is tem pting, but 25 . . . tDf8 (25 . . . cxd3 26 !!xc7 Axc7 27 'tWxd3) and now :

8 7

B 11 ) 26 Axf6 allows Black to attack too many pieces simu ltaneously with 26 . . . A a3 ! =+= (but not 26 . . . gxf6 27 �xf6+ !!g7 28 �x h6+ � g8 29 !!cg1 !!dd7 30 Ae4) ;

6 5 4 3

B 1 2) 26 A b1

Despite a long think, I fai led to fi nd the sim plest way to win. 23 Ad3 � b7+ 24 !!g2 tDf8 25 A xf6 g xf6 (25 . . . !!f7 26 Ae4 �d7 27 !!xg7 !!xg7 28 !!g1 ) 26 �xf6+ (26 !!cg1 .!'!dd7 27 �xf6+ !!g7) 26 . . . !!g7 27 !!cg1 !!dd7 was only good enough for a draw. The best way to execute Wh ite's idea was 23 !!g2 ! , when Black i s helpless : 2 3 . . . �b7 (23 . . . �c6 24 Ad3 tDf8 25 Axf6 !!f7 26 Ae4, or 23 . . . b5 24 Ad3 tDf8 25 Axf6) 24 Ad5 �c8 25 !!cg1 .

23 . . tt)f8? .

Black fails to seize his chance. Perhaps Joel decided to make use of a rare opportu nity to put all(!) his pieces on dark squares, as in draughts ! If 23 . . . �c6+ 24 Ad5 � b5 25 Axf6 ! � xf6 26 !!x h6+ g x h6 27 �xf6+ � h7 28 Ae4+ !!xe4 29 !!g1 and White wins. But after the strongest move 23 . . . b5 ! Wh ite has to play precisely to demonstrate an ad vantage : A) 2 4 Ad5 c 4 (24 . . . !!f8 25 !!cg1 c 4 26 � h5, or 25 . . . �d8 26 !!6g2 (threatening A e4) 26 . . . !!ff7 27 A xf7 !!xf7 28 !!d1 ; 24 . . . � b6 25 !!x h6+) 25 A xf6 (this only leads to a draw, but if 25 !!cg1 c3 , while after 25 A e4 Wh ite has to go in for the same position as after 24 Ad3) 25 . . . � xf6 26 !!x h6+ g x h6 27 �xf6+ !!g7 (27 . . . � h7 28 !!g1 ) 28 �x h6+

26 . . . !!dd7 !! (overprotecting the 7 th ran k ; if 26 . . . �c5 27 �xf6, or 26 . . . � b7 27 !!cg1 !!dd7 28 � h5) 27 � h5 (27 �g4 !!f7 28 Ae4 b4 29 Ad5 c3 30 Axf7 !!xf7 =+= , 27 �xb5 �b7 28 �x b7 !!x b7 29 Ad4 �e6 with a proba ble d raw, or 27 A xf6 !!f7) 27 . . . �c6 28 Af5 (28 !!cg1 c3) 28 . . . !!c7 (28 . . . b4 29 Axf6 gxf6 30 �x h6+ � h7 31 A xd7 �xd7 32 !!cg1 ) 29 A xf6 g xf6 30 �x h6+ � h7 31 !!cg1 !!g7 32 Axh7 !!x h7 33 �xf6+ !!hg7 34 'tWh6+ �g8 35 �e6+ = ; B2) 25 Ae4 ! , threatening 26 !!cg1 , followed by a devastating sacrifice. As the following lines show, White should wi n : B21 ) 2 5 . . . !!xe4 2 6 �xe4 ; B22) 25 . . . !!de8 26 !!cg1 (26 !!xh6+ g x h6 27 Axf6+ tD xf6 28 �xf6+ !!g7 29 �xh6+ !!h7 30 Axh7 �xh7 31 �xd6 �e4+ 32 �g1 !!g8+ 33 � f1 ±) 26 . . . c3 27 !!6g2 �f8 (27 . . . !!xe4 28 !!xg7)

175

(see next diagram)


825) 2 5 . . . �c5 2 6 gxh6+ ! g x h 6 2 7 A xf6+ ttl xf6 28 �xf6+ gg7 29 �xd8+ ± and White remains a pawn up with an attack, but Black avoids an immed iate loss. a

b

c

d

e

f

8

8

7 6 Analysis diagram after 27 f!.6g2 !fJf8

5

28 �xf6 ! ! ttlg6 (28 . . . ttl h7 29 �xc3 �xc3 30 A x c3) 29 A xc3 ! g xf6 30 A xf6+ (or 30 gxg6 � h7 31 Af5 �c5 32 Ae5) 30 . . . � g8 31 gxg6+ � f8 32 Ad5 ge6 33 Ag7+ ; 823) 25 . . . b4 26 gcg1 c3 (26 . . . A xf4 27 gx h6+ g x h6 28 A xf6+) 27 gxg7 gxg7 28 gxg7 � xg7 29 �g6+ � f8 30 �x h6+ ; 824) 25 . . . A b4 26 gcg1 (26 gg2 gxe4 27 �x e4 ttlc5 2 8 �g6 �e7 +) 26 . . . c3 27 gxg7 gxg7 28 gxg7 � xg7 29 � h7+ � f8 30 Ad5 ttlc5 31 �xc7 gxd5 32 A xc3 A xc3 33 �c6 ± ;

4 3

�------�--�

�

24 Dxh6+ ! gxh6 25 Axf6+ 25 gg1 was equally strong.

25 . . . Dg7 26 Axg7+ "xg7 27 Dg1 Black resigns

***

Game 40

B o r i s G e l fa n d - K i r i l G e o rg i ev French Team Championsh ip, Orange 2000 Sla v Defence [0 1 5J Th is was my first ever game i n the French League. I managed to outplay an experi enced opponent without him making an ob vious mistake.

1 d4 d5 2 c4 c6 3 �c3 �f6 4 �f3 a6 5 c5 Th is move relieves the pressure in the centre. Sti l l , I th i n k that Wh ite can count on some edge here.

5

. • .

Af5

5 . . . ttl bd7 is the main alternative and was played against me by Sergey Movsesian a few months later: 6 Af4 ttl h5 (trying to pre vent Wh ite from keeping his bishop on this d iagonal ; 6 . . . g6 7 h3 Ag7 8 e3) 7 e3 96 8 Ad3 Ag7 9 0-0 f6 ? (9 . . . 0-0 10 Ag5 ;t ) 1 0 h 3 ! (Wh ite is not afraid of the com pl ications, in view of the fact that all(!) Black's pieces are poorly placed) 10 . . . tbxf4?! (10 . . . e5 1 1 A h2 e4 was the most ambi tious, but Wh ite keeps a strong advantage by 12 g4 ! exd3 13 gxh5 tbf8 14 e4 Axh3 15 ge1 � f7 1 6 exd5 cxd5 1 7 � b3 �d7 1 8 tbxd5 Ae6 19 gxe6 ttl x e6 20 ttlc7 ± Yermoli nsky Xu J u n , F I D E World Championsh i p , New Del h i 2000) 1 1 exf4 e5 (11 . . . �c7 1 2 f5 ! tbf8

176

Game 40

ttJ

Gelfand - Georgiev, French Team Championship, Orange 2000

13 ge1 ! g x f5 14 � h4 ± Dreev) 1 2 f5 ! (the c5 and f5 pawns completely restrict the black pieces ; White gains noth ing after 12 fxe5 fxe5 1 3 � x e5 A x e5 14 d x e5 0-0 with counterplay) 12 . . . e4 ? (playi ng i nto Wh ite's hands; 1 2 . . . 0-0 1 3 fxg6 e4 14 g x h7+ � x h7 15 A x e4+ d x e4 1 6 � x e4 ± ; 1 2 . . . g5 1 3 ge1 0-0 14 b4 ± e4 15 Axe4 dxe4 16 �b3+ � h8 17 tLlxe4) 13 Axe4 d xe4 14 �xe4 +-

a

b

c

d

e

f

6 5 4 3

��--I��

...

9

h

2

11

10 g4!? N

(despite Black's extra piece, his position is hopeless, as his pieces have hard ly any moves) 14 . . . 0-0 (14 . . . Af8 1 5 ge1 � f7 16 � b3+ � g7 1 7 �d6 A xd6 1 8 cxd6 ge8 19 gxe8 �xe8 20 ge1 ) 15 �b3+ � h8 16 tLld6 A h6 1 7 gfe1 � g7 (1 7 . . . b6 1 8 fxg6 h xg6 19 �f7 ! +-) 18 ge4 � xc5 19 d x c5 g x f5 20 ge8 (20 gd4 VJ!ic7 21 gad1 was simpler) 20 . . . �xe8 (20 . . . gxe8 21 �f7+ � h8 22 tLlxe8 Ae6 23 �xe6 �xe8 24 �xf6+ Ag7 25 �xf5) 21 �xe8+ gxe8 22 g3 ! and White soon won (Gelfand - Movsesian , Akiba Rubinstein Me morial , Polanica Zdroj 2000).

S Af4 �bd7 7 e3 eS S �d2 This multi-purpose move was introduced by Kramn i k against Topalov in a bri l l iant game in Dortmund 1 999 . White prevents the ex change of knights on e4 and also prepares for the typical manoeuvre . . . �c8 followed by . . . Ae7-d8-c7. Then he will exchange on c7 and play f2-f4 , preventing . . . e6-e5.

S . . . Ae7 9 Ae2 'ttcS Allowing White an extra optio n . Normally Black plays 9 . . . Ag6 or 9 . . . h6.

The most principled decision. Wh ite gains space and also some temp i , by kicki ng the knight and the bishop. Also, it looks pro b able that both sides will proceed with their kings i n the centre. After 1 0 b4 Ad8 1 1 f3 Ac7 1 2 Axc7 �xc7 13 e4 dxe4 14 fxe4 Ag6 15 0-0 0-0 16 e5 tLld5 17 tLlxd5 cxd5 18 �b3 b6 19 �c3 bxc5 20 bxc5 a draw was agreed in Sakaev -Georgiev, (yugoslav Team Cham pionship, Vrnjacka Banja 1 998) .

10

. . .

AgS 1 1 g5 �gS

Now we can see the usefu lness of 8 �d2 , as bad is 1 1 . . . �e4 1 2 �c x e4 A x e4 (or 12 . . . dx e4 13 �c4) 13 �xe4 dx e4 14 �c2 f5 (14 . . . �d8 1 5 �xe4 A xg5 1 6 Ag3 ±) 1 5 g xf6 � xf6 1 6 Ae5 winning the e4 pawn .

12 h4 1 2 e4 h6.

12 . . . hS?! Black bri ngs his pieces i nto play (rook or kn ight, depending on White's choice) , but I thi n k that it was stronger to fol low the old maxim : respond to a flan k attack with a counterblow in the centre! - 1 2 . . . e5 ! 13 Axe5 (13 Ag3 exd4 14 exd4 h6!? Pel letier; 13 dxe5 �xc5 14 tLl b3 �e6) 13 . . . �xe5 14 dxe5 V!!Jc7 (14 . . . Axc5 15 h5 Af5 16 e4 with the initiative) 1 5 f4 A xc5 16 h5 Af5 1 7 e4 dxe4 1 8 �dx e4 gd8 with counterplay, as indicated by Yan n ick Pelletier in Die Schachwoche.

177

1 3 gxhS!


White allows the knight back into the game, but he gains pressure on the g-file in return . If 13 gg1 hxgS 14 hxgS Ad8 1S e4 Ac7 (1S . . . eS 16 dxeS d4 17 tDa4) 16 Ad6!? and Wh ite has the initiative, but Black can gain counterplay with 14 . . . eS , as compared with the 1 2 . . . eS line he has control of the h -fi le.

1 9 � eS 'tl¥x b3 20 a x b3 � x eS 21 A xeS as 22 tDa4 :t (22 h6 � xh6 23 gxg7 Ac2).

19 �xb6 20 Hc1 The rook moves to an open file, prevent ing . . . c6-cS at the same time: 20 . . . c5 21 dxcS AxcS 22 tD xdS tD xdS 23 gxcS tDxf4 24 exf4 ± . •. .

13 . . . �xh6 14 h5 �f5 14 . . . A h7 1S gg1 is merely a transposition of moves.

20 . . . 0-0 a

8

15 Hg1 .th7 a

b

7 c

d

e

d

e

8 7

5

7 6 5

4

4

3

3 2

4

4

3

3

� 'if

�_______...._ ..;:;. _

2

..... 'if

L...-_______---"'__

1 6 �f3! The knight returns, as it is im portant to take control of eS. . . .

c

6 8

16

.!.

b

•d8

Prepari ng . . . b7-b6. If 1 6 . . . Af6 1 7 e4 d x e4 1 8 � x e4 . B lack does not succeed i n ex changing the bishops with 16 . . . Ad8 17 e4 (17 A d3 Ac7) 17 . . . d xe4 18 � x e4 A aS+ 19 � f1 Ac7 20 � d6+ A xd6 21 cxd6 and , although all White's pawns are weak, he stands better due to the big difference i n the activity of the pieces.

17 .b3! White exchanges queens, thus ensuring the safety of his king. If 17 e4 ?! d x e4 18 � x e4 tDf6 1 9 tDegS Ag8 ! , protecting the important f7 pawn and eyeing the weaknesses on d4 and hS. I n the future the bishop would have a chance to get back i nto the game.

17 . . . b6 18 cxb6 .xb6 19 .xb6

The alternative was 20 . . . f6 2 1 A c7 ! (an i mportant m ove - Wh ite d rives away the kn ight from the dS square ; after 21 e4 dxe4 22 � xe4 tDdS 23 Ad2 � f7 ! (23 . . . c;, d7 24 �cS+ A x cS 2S gxcS :t ) 24 gxc6 ghc8 Black gets rid of his weak c -pawn and gains counterplay on the c -file, or 21 tD b1 � d7 22 � bd2 as , again with counter play) 2 1 . . . �d7 (21 . . . Ad8 22 A x b6 Axb6 23 tDa4 ±) , and now : A) 22 � a4 � f7 ! (22 . . . A b4+ 23 � d1 , but not 23 � f1 gc8 24 gxc6 �e7 2S Ax a6 ga8 26 gxe6 gxa6 -+) 23 gxc6 ghc8 with coun terplay - if 24 � d2 A b4+ ; B) 22 e4 ! d x e4 23 � x e4 � f7 (23 . . . A b4+ 24 � f1 ±) 24 gxc6 ghc8, and now : B 1 ) the tem pting 2S Ac4 �f8 leads to big complications : B 1 1 ) 26 dS A b4+ 27 � e2 �e7 ! 28 dx e6+ c;, e8, and if 29 gxa6 gxa6 30 A bS+ (30 Axa6 gxc7 31 A bS+ � d8 32 gd1 + � c8 33 Aa6+ � b8) 30 . . . gc6; B 1 2) 26 �cS A x cS (26 . . . ga7 27 A xe6+ � x e6 28 � x e6, or 27 � xe6 � e8 28 gxa6)

178

Game 40

Gelfand - Georgiev, French Team Championship, Orange 2000

27 gxc5 � d7 ! 28 gc6 ga7 29 Axe6+ (29 d5 ,§,cxc7 (29 . . . exd5 30 A xd5+ ; 29 . . . gaxc7 30 d x e6+ � e7 31 exd7 gxc6 32 d xc8 'tW ,§,xc8 33 Axa6 gc2 34 Ad3 ±) 30 dxe6+ �e7 31 �d4 � xd4 32 gxg7+ � d8 33 gxc7 gxc7 34 gx h7 gxc4 35 gxd7+ � e8 =} 29 A x e6+ � e8 30 Axf5 (30 Axd7+ �xd7 31 A b6 gxc6 32 A x a7 gc2) 30 . . . A xf5 31 gxg7 (31 h6 g5) 31 . . . gcxc7 (31 . . . gaxc7 32 gg8+ �f8 33 gxf6 gf7) 32 gxc7 gxc7 33 � h4 A b1 34 h6 gc1 + (34 . . . � f8 35 � d2) 35 � d2 gh1 (35 . . . gc2+ 36 � e3 gx b2 37 h7 A x h7 38 gx h7 gxa2 39 �f5 t) 36 h7 gxh4 37 gg8+ tDf8 38 h8 'tW gxh8 39 gxh8 Axa2 = ;

ttJ

23 . . . � h4 !? deserved attentio n , but Wh ite's advantage is still i ndisputable.

24 dxe5 gxe5 25 �e2 25 A xf5 !? exf5 26 � e2 gac8 t and com pared with the game Black has counterplay on the c -file.

B2} 2 5 gc4 !? ± and Black does not have suf ficient com pensation for the pawn.

21 h6! Creating a new target at g7 . After 2 1 Ac7 tD d7 22 � a4 gfc8 23 � b6 gxc7 24 � x a8 ,§,a7 (24 . . . gc8 25 Axa6 gxa8 26 gxc6 A b4+ 27 � e2 � e7 28 gc7 00) 25 gxc6 gxa8 Black has counterplay.

21 . . . gfe8 After 21 . . . f6 (21 . . . � d7 22 Ad3) 22 h xg7 (22 Ac7 � d7 23 h x g7 gfc8 24 Af4 c5 (24 . . . � b6) 25 e4 !?) 22 . . . gfc8 23 Ad3 (23 � d2 c5 with counterplay, but not 23 . . . � x g7 ? 24 A h6 Af8 25 Ag4 ±) 23 . . . c5 24 � e2 c4 (24 . . . cxd4 25 A xf5 A xf5 26 � xd4) 25 Ac2 (threatening gg2 and gh1 , so Black has to take the g7 pawn, which in a way is protecting his king) 25 . . . � xg7 26 A x h7+ � x h7 27 gg2 Wh ite has a strong attack, for exam ple 27 . . . A b4 28 � h2 ! .

22 Ad3! 22 h xg7 c5.

22 . . . g6 There is no other way to protect the paw n , but now look at the h7 bishop ! On t he other hand , Black hopes to surround and win the h6 pawn, but in real ity he will never be able to do this.

25 . . . a5? Black does not sense the danger and he makes a decisive m istake. After 25 . . . �d6 26 A xg6! fxg6 27 � xg6 A xg6 28 gxg6+ �f8 29 gcg1 Wh ite wins, but 25 . . . � h4 wou ld have forced him to play energetically: 26 e4 ! f6 (26 . . . d4 27 � b1 ; 26 . . . dxe4 27 Axe4 gac8 28 A b7) 27 Ae3 ! (a tactical finesse) 27 . . . fxe5 28 Axc5 Axc5 29 � xd5 and Black's pieces are too poorly placed to create counterplay: A} 29 . . . � xd5 30 exd5 Ad4 31 d x e6 gf8 32 e7 gxf2+ 33 � d1 � f7 34 gc8 or 34 gc7 !? +-; B} 29 . . . � d7 30 b4 Axf2 31 � c7 +-. 25 . . . d4 was the most stubborn defence : 26 A xf5 d xc3 27 Ae4 ! (27 A xg6 fxg6 28 � xg6 Axg6 29 gxg6+ � f7 30 gg7+ �f6 00 is far from clear) 27 . . . � d5 28 A xd5 gxd5 29 gxc3 Af6 30 ggc1 ± and the extra pawn should tel l .

26 Axf5 exf5 2 7 �d3 gee8 27 . . . gc4 28 � xd5.

28 Ae5! +-

23 �e5 e5

179


a 8

b

c

d

...

e

.1

29 £d4 llcbS Or 29 . . . �d7 30 �xd5 gxc1 31 gxc1 Axh6 32 gc7 .

8

7

7

6

6

5

5

4

4

3

3

30 �f4 £xh6 31 �h5! 31 �fx d5 �xd5 32 �xd5 Ag7 ± and the game sti l l goes on .

31 . . . �d7 32 �xd5 mS 33 �hf6 �xf6 34 �xf6 £gS 35 llh1 £g5 36 �xg8

2

White regrou ps his pieces. H i s bishop will dominate on d4, wh i le his knight will go to f6 with damaging effect.

What a poor career for this bishop ! After be ing developed at f5, it retreated to g6, then h7 , then g8, and finally it was devoured there!

36 . . . 1ldS 37 �f6 Or 37 gh8.

2S . . . £fS

37 . . . �e7 38 llc7+ �e6 39 ftc6+ IId6 40 IIxd6+ �xd6 41 f4 Black resigns

28 . . . Ab4 29 �f4.

*** nese chess. However in my huge database I found only a few games where he had faced one of my favourite variations.

Game 41

B o r i s G e lfa n d - Ye J i a n g c h u a n Wo r l d C u p , S h enyang 2 0 0 0 King 's In dian Defence [E9 7J

7 0-0 �c6 S d5 �e7 9 �d2

The Chinese men have not yet reached the pinnacle of the chess world, as their women have done, but their successes are l i kely to come. They are traini ng very hard and they regularly have opportunities to face tough in ternational opposition in strong tournaments in their homeland . The World Cup attracted a very powerful field, but my opponent sti l l managed to w i n a strong qualification group. And our quarter-final match was a tough test for me, which I managed to win only after the tie-break. Th is game looks dry, but I am happy that I managed to convert a tiny ad vantage i nto a wi n , i n a style which resem bles that o f my favourite player, Akiba Ru binstein .

I regularly choose other variations against the King 's Ind ian and I had n 't played 9 ttJd2 for years, but I noticed that my opponent d i d n 't have much experience with this vari ation and so I decided to give it a try.

9 . . . a5 10 a3 �d7 1 1 ft b1 f5 1 2 b4 �h8 13 1fc2 �f6 a

b

c

d

e

7 6

1 d 4 �f6 2 c 4 g 6 3 �c3 £g7 4 e4 d6 5 £e2 0-0 6 �f3 e5 The King's Ind ian Defence is the main and al most the only weapon of the Patriarch of Chi-

180

�------�--� �

Game 41

Gelfand - Ye J iangchuan , World Cup, Shenyang 2000

14 Ab2 Th is idea was i ntroduced by Lajos Portisch and I trusted the tastes of the great Hungar ian player.

14 . . . axb4 In our tie -break game Ye deviated with 14 . . . c6 15 gbd1 fxe4 16 ttld x e4 ttlxe4 1 7 ttlxe4 cxd5 1 8 cxd5 ax b4 1 9 ax b4 Af5 20 Af3 �b6 and ach ieved good play.

1 5 axb4 fxe4 After a long think Ye decided to simplify the game, as his pupil (he has seconded the women 's world champion i n most, if not all of the matches that she has played) got into trouble after 15 . . . c6 16 gbd1 !? cxd5 17 exd5 Ad7 18 ttldb1 ! f4 19 ttla3 Af5 20 �b3 g5 21 c5 ttlg6 22 ttlc4 d xc5 23 b x c5 ± %-% (Portisc h - Xie J u n , Flamenco, Veterans v. Lad ies, Marbella 1 999).

dominates the only open fi le? The answer is that Black's rook is the main com ponent of his counterplay, and without it he can only sit and wait. If 21 ga7 Af5 ! 22 Ad3 ge8 23 f3 �b8 (23 . . . Ah6 24 gx b7 Ae3+ 25 c!>h1 ) 24 �a4 gfB intend ing . . . Ah6-e3+ with coun terplay.

21 . . . ee7 22 Ad3 Af5 23 I:Ixf8+ Axf8 24 g3 Axe4 It would appear that i n any case Black cannot avoid this exchange, as otherwise the supremacy of the wh ite knight over the g7 bishop would be too obvious. But perhaps he could have carried it out in more favou rable circumstances : 2 4 . . . Ah 6 25 h 4 Axe4 (25 . . . c!>g8 26 f4 Axe4 27 Axe4) 26 Axe4 Ad2 27 b5 and at least Wh ite is forced to move his pawn to b5, which re stricts his possibil ities.

25 Axe4

1 6 �cxe4 �xe4

a

16 . . . c6 with counterplay, as proposed by my second Alexander Huzman in ChessBase Magazine, is also worthy of consideration .

8

18 . . . Ad7 19 gxa8 �xa8 20 c5 and Black cannot use the a-fi le.

19 I:Ixa1 �d4 20 Axd4 exd4

An im portant moment. Why does White aim for the exchange of his active rook, which

c

d

..

e 8

7

17 �xe4 �f5 18 �a1 I:Ixa1

21 l:Ia8!

b

6

6

5

5

4

4

3

3

2

2

Both sides had aimed for this position , but to judge by the speed with which my opponent made his moves, I g uess that he assessed it too optimistical ly. I thought that White had serious winning chances, as opposite -colour bishops normally favour the attacking side and the black king is much more vul nera ble. However experience should have taught me a lesso n . I failed to win an end ing with the same material (but obviously, a d iffer ent pawn structure) against Istratescu in the 1 996 Yerevan Olympiad and I was extremely

181


lucky to beat Schebler in the Bundesliga the same year.

25 . . . b6?! A careless move, also played instantly. I think that it was much more accurate to start with 25 . . . g5!?, blocking the wh ite pawns. As we wi ll see, this was the best defensive concept.

34 . . . g5 !? deserved serious attention , al though White has available the plan of mov ing his king to the q ueenside: 35 �f1 �h5 (35 . . . �g7 36 �e2) 36 �f5 �g7 37 �e1 .

8

6

4

30 . . . "d7

B) 32 f4 h5; C) 32 �f5 ! g x h4 33 g4 ! (33 �h7+ al lows B lack to exchange some more pawns after 33 . . . �f8 34 �xc7 (34 �x h6+ �e7 35 g4 �g8 36 �h3 Ag5) 34 . . . �d8 35 �h7 h x g3 36 �x h6+ Ag7 37 �f4+ �f6 38 �xg3 Ah6 and the liqu idation is obviously i n Black's favou r, as now he can safely exchange queens) 33 g4! �f7 34 �h7+ �f8 35 �x h6+ �g7 36 �xg7+ �xg7 37 f4 ! (j ust in time) 37 . . . �h6 38 �f3 with an easy win.

31 . . . "f7 32 h x g6 h xg6 33 "e4 �h6 34 "g4 "e8

1-1<=

4

2

35 e5!? I decided to exploit the opportu n ity of ad vancing my pawn to c6. It has its advan tage - very often the d7 square can be used by the wh ite queen , but also its d isadvan tage - it will be m uch more d ifficult for the king to advance to b7 , as it can be met by checks on the b -and a-fi les. But after 35 b5 �f7 36 f4 �e8 (or 36 . . . �g7 37 �f3 �e8 38 �e6 �f7) 37 �e6 White is one tempo short : 37 . . . �xe6 38 d x e6 g5 39 �f3 gxf4 40 gxf4 (40 �xf4 �g7 41 g4 �f8 42 g5 Ae7 43 g6 Af6 44 �e4 �e7 45 �d5) 40 . . . �g7 41 �e4 �f8 = .

35 ... bxe5 36 bxe5 �g7 36 . . . d xc5 37 �f4+ Ag5 38 �xc7 .

37 e6 "17 38 �1 "e8 39 lLe4

31 h5?! White is in too much of a hurry. It made more sense for h i m to i m prove his position first with 31 �e4 �e7 32 �g4 fol lowed by f2f4 and only then h4-h5, i n order to prevent Black's strongest defensive plan of . . . g6-g5, after which he can hold on .

5

3

30 . . . h5 is too risky, as after 31 �f1 Black cannot exchange queens, si nce after playi ng f2-f4 White takes his king to c6, so the queen is tied to defending the g6 pawn . 30 . . . �a4 ?! is pointless in view of 31 h5.

A) 32 h5 �f7 (32 . . . Ad8 33 �f5) 33 �f5 �f8 34 �c8+ �g7 35 Af5 �e7 and I don 't see a way to improve Wh ite's position ;

1-.1-=

5

Now Black has to define his pawn structure.

30 . . . h6 is critical , but i nsufficient - 31 �g4 (31 h5 g x h5 32 �f5 h4 33 �h7+ �f8 leads to the same position) 31 . . . g5, and now :

7

7

26 h4! ILg7 27 �g2 ILf6 28 "d3 "e8 29 "f3 �g7 30 ILd3

White is planning �e2-d3-c4, but it wi ll still be d ifficult to make progress afterwards. H owever, my opponent decided to gain some scope for his bishop at the cost of a pawn. Also, after this Black can nearly al ways exchange queens, as the c7 pawn can easily be protected by his bishop.

(see next diagram)

182

39 . . . d3 40 ILxd3 lLe3 41 �g2 41 . . on 42 lLe4 1Lf6 ..

Game 41

Gelfand -Ye J iangchuan , World Cup, Shenyang 2000

42 . . . Ad4 43 f4 Ab6 1eads to the same pos ition. 43

"ca Ad4 44 f3 Ab6 45 f4 Ad4

45 . . . �f6 !? 46

"g4 Af6 47 @f3 Ad4

Black sticks to Capablanca's advice to keep your pawns on squares of the opposite colour to that of your own bishop. With 47 . . . g5!? 48 �e3 �h6 ! 49 �d3 gxf4 50 gxf4 (50 �xf4+ �g7 51 �f5 Ae5) 50 . . . �h5 Black could also have retained good drawing chances.

4a c!>e2 Ab6 49 "d7 49 f5 �f6. 50 �d3 �f6. ...

b

8

8

7

7

6

6

5

5

4

4

3

3

2

2

But with 50 . . . �f6 ! 51 �g4 �g8 Black could have constructed a fortress, as I don't see any way for White to make progress.

51 g4 51 �g4?! Ad4 52 �xf7+ �xf7 53 �g5 Af6+ 54 �h6 Ag7+ 55 �h7 g5 is a draw.

51 . . . g5 A blunder, but after 51 . . . Ad4 (or 51 . . . �f8 52 g5 �g7 53 �h3) 52 g5 Black was lost anyway : A) 52 . . . Ab2 53 �g4 Ad4 (53 . . . Ac1 54 Axg6 �xg6 55 f5+ �g7 56 g6) 54 �d8 (found by Sergey Shipov; after 54 f5 g xf5+ 55 Axf5 �f8 56 Ae6 �xd7 57 cxd7 �e7 58 �f5 Ae5 White has a pawn less com pared with the game, so I don't see a win) 54 . . . Ac3 55 f5 ! g xf5+ 56 Axf5 Ae5 57 �c8 ! Ah2 (57 . . . Ac3 58 Ae6) 58 �d7 �f8 (58 . . . Ae5 59 Ae6 !) 59 �d8+ �e8 60 �f6+ �f7 61 �h8+ ;

49 . . . @fa 50 @f3

a

d5 63 c7 �d7 64 ctJf5) 56 . . . ctJg7 57 �d7+ �f8 58 c7 �c3+ (58 . . . �f1 + 59 �g3 �e1 + 60 �h2 �f2+ 61 �h3) 59 �g2 �c2+ 60 �g3 �c3+ 61 �h4 �h8+ 62 �g5 �g8+ 63 �h6 �h8+ 64 �h7 �f6+ 65 �h5 +-.

B) 52 . . . �f8 53 �c8+ �g7 54 �h3 �g8 55 �e2 and Black's pieces are too badly placed for h i m to have any defensive chances.

52 Af51 Ad4 Or 52 . . . g xf4 53 Ae6 �xd7 54 cxd7 c6 55 d xc6 �f6 56 Ab3 �e5 57 Ac2.

50 . . . c!>g7? 50 . . . Ad4 would also have lost, although after 51 g4 Ab2 White would have had to find the d ifficult win by 52 Axg6! (52 g5 al lows Black back i nto game: 52 . . . Ac1 ! 53 �g4 Ad2 54 �h4 �g7 ±) 52 . . . �xg6 (52 . . . �xd5+ 53 �g3 Af6 54 g5 Axg5 55 fxg5 �e5+ 56 �f3) 53 �d8+ �e8 (53 . . . �g7 54 �xc7+ �h6 55 g5+ �h5 56 �e7 �d3+ 57 �e3 �xd5+ 58 �g3 �xc6 59 �e2+ �g6 60 �x b2 +-) 54 �xc7 Aa3 55 �h7 �e1 56 �f5+ (or the study- l i ke 56 �h6+ �e7 57 �e6+ �xe6 58 dxe6 �xe6 59 f5+ ctJe7 60 f6+ �e6 61 g5 Ab2 62 �g4

183

53 fxg5 @fa 8 7 6

6

5 4 3 2

....;:;...

L..._ ._ _ _ _ _ _

----I 'lJ

_


54 �e4 After I made this move I became very ner vous, as I noticed an i ncredible resource for my opponent. But after home analysis I re alised that it was the easiest way to win, as the alternative 54 �e2 Ab6 (54 . . . '�e7+ 55 Ae6) 55 �f1 !! (55 g6 �f6) 55 . . . Aa5 (55 . . . �e7 56 Ae6) 56 �d8+ �g7 57 �f2 ! ! (57 �e2 Ab6) 57 . . . Ab6+ 58 �e2 (Black is in zugzwang) 58 . . . Ad4 59 �d3 Ab6 60 �c4 etc. is too subtle.

54

. • .

Ab2

54 . . . Af2 ! was the best chance : A) 55 �d3 �xd5+ 56 �e2 �e5+ 57 �xf2 �h2+ 58 �f3 �h1 + and I don't see where White can hide his king ;

and Black cannot hold the position : 57 . . . Ah4 58 �d5 Af6 59 �c4 �e7 60 � b5 d5 61 �a6 d4 (61 . . . �d6 62 �b7 Ae5 63 �c8 d4 64 �d8 Af6+ 65 �e8 d3 66 �f7 d2 67 Ac2) 62 �b7 �d6 63 e7 . After 54 . . . Ab6 55 Ae6 �g6+ 56 �f4 Ae3+ 57 �f3 ! (Wh ite can now avoid taki ng the black bishop, so there is no stalemate) 57 . . . Axg5 58 �c8+ �e8 59 �xc7 .

55 g6 1Jxd7 55 . . . �e7+ does not help after 56 Ae6, for ex ample: 56 . . . Af6 57 �d3 �g7 58 �c4 �xg6 59 � b5 �g7 (59 . . . �d8 60 �xd8 Axd8 61 �a6, or 59 . . . �xd7 60 cxd7 Ad8 61 �c6) 60 �a6 Ac3 61 g5 +- �xg5 62 �xg7+ Axg7 63 �b7.

56 cxd7 �e7 57 g5 Ag7 58 �d3 �d8 59 �c4 �e7 60 �b5 �d8 61 �c6

B) 55 Ae6 �g6+ 56 �f4 (this was my original thought, but then I realised that Black can force stalemate ; if 56 �f3 �d3+ 57 �xf2 �e3+ =, but White can sti l l retrace his steps with 56 Af5 �f7) 56 . . . Ag3+ !!

a

C) 5 5 g6 ! (this appears t o b e the only way to win) 55 . . . �e8+ 56 �e6 �xe6+ 57 d x e6

c

d

e

f

9

h

8

8

7

7

6

6

5

5

4

4

3

3

2

2

a

57 �xg3 �d3+ 58 �h4 �h3+ ! ! ' I remem ber a simi lar idea from my game with Yuri Balashov in M i nsk 1 986 (see below, p. 1 85) ;

b

b

c

d

e

f

9

h

�

Black resigns, as Wh ite has a q u ite long, but simple winning plan : 61 . . . Ah8 62 Ac2 Ag7 63 Aa4 Ah8 64 � b7 Ag7 65 Ac6 Ah8 66 �a6 Ag7 67 �b5 Ah8 68 �c4 Aa1 69 �d3 Ah8 70 �e4 Ag7 71 �f5 �e7 72 d8� + �xd8 73 �e6.

***

184

Game 42

Gelfand -Anand, World Cup, Shenyang 2000

Stalemate idea (cf. note to Black's 54t h move)

46 Axe6+ was the most natural move and should have led to a win, but I was surprised by a stalemate idea and I deviated from the right path. 46 . . . ttlxe6 47 'tWxe6+, and now :

Boris Gelfand - Yuri Balashov M insk 1 986 8

A) 47 . . . �h8 (this also does not help) 4 8 Wfxe7 (48 Wff6+ �g8 49 Axf4 ttlf5 and it is not easy to win) 48 . . . Wfg6+ 49 �h2 Wfh6+ (49 . . . Wfh5+ 50 �g2 'tWg4+ 51 �f1 'tWh3+ 52 �e2 'tWg4+ 53 �d3 'tWf3+ 54 �d4 'tWxf2+ 55 �c3 +-) 50 �g2 'tWg6+ 51 �f3 'tWd3+ 52 �g4 'tWxd2 53 'tWf6+ �g8 54 e6 ;

8

1---.1'--'

7

7

4

B) 47 . . . �f8 48 Ab4 'tWb1 + (48 . . . 'tWh7 49 Wfxe7+ 'tWxe7 50 �g2) 49 �h2 Wfh7+

3 2

2

L...-_______--:::.-_---I

'lf

44 Ad2! Intend ing Ag5, when the knight at g6 will be threatened . If 44 'tWg5 �h7 ! 45 Ab4 ttlf4 ! ! 46 'tWxe7+ �h6 =.

44 . . . f4 The only move. 45

Ag4! �f8

45 . . . ttlxe5 46 Axe6+ ttlf7 47 Ac3 'tWg6+ 48 'tWxg6+ ttlxg6 49 �g2 is also hopeless.

46 "xf4?

50 �g2 ? (50 'tWh3 ! +- refutes Black's trap) 50 . . . 'tWg6+ !! 51 'tWxg6 f3+ = .

46 ... "b1 + 47 �h2 "xb2 And a draw was agreed a few moves later.

*** Game 42

B o r i s G e l fa n d - V i s wa n at h a n A n a n d Wo r l d C u p , S h enyang 2 0 0 0 Ca talan Opening [E05J I have played a number of memorable games with Vishy Anand and this was the second game of our World Cup semi-fi nal match. Vishy does not need a ny special i ntroduc tion as a chess player. I should j ust l i ke to add that he is an extremely kind, honest and friendly person . I always enjoy the company

of him and his wife Aruna and we also share a passion for Japanese food.

1 d4 �f6 2 c4 e6 3 g3 d5 4 �f3 Ae7 5 Ag2 0-0 6 0-0 dxc4 Black chooses the most solid system against the Catalan . 7 "c2 I have put a lot of effort into reviving the risky variation 7 ttle5 ttlc6 8 Axc6 bxc6 9 ttlxc6 'tWe8 1 0 ttlxe7+ 'tWxe7 11 'tWa4. White wins a pawn , but, as he has given up his important bishop, Black has good compensation :

185


A) 1 1 . . . e5 1 2 d x e5 �xe5 1 3 �xc4 Ae6 14 �d3 ! (the theory books give 14 �c2 Af5 15 �c4 Ae6 with a draw, but White can keep on playing) 14 . . J �ad8?! (14 . . J 3ab8 !?) 15 �e3 �h5 1 6 f3 ! gfe8 (16 . . . Ac4 1 7 �c3 gfe8 18 �g5 ! ±) 1 7 �g5 �h3 (17 . . . Ac4 1 8 �x h5 �x h5 1 9 �c3 Axe2 20 gf2 Aa6 21 Ag5 ! ±) 1 8 �c5 ! ± (against all the rules, White does not develop his pieces, but i m proves the position of the only developed one! - how ever, he puts an end to Black's counter play) 1 8 . . . Ad5 (18 . . . Ac8 !? with the idea of . . . Aa6 - Timoshchenko ; 18 . . . Axa2 1 9 �c3) 1 9 e4 ! (19 �c3 ? �g4).

B) 1 1 . . . c5 , and now : B 1 ) 1 2 �xc4 cxd4 1 3 �xd4 e5 14 �h4 �e6 !? 15 �c3 Ab7 (Black has good piece play and sufficient com pensation for the pawn) 16 e4 (16 Ag5 �d5 !? 00 ; the exchange of knights is in Black's favour) 1 6 . . . gfc8 (16 . . . �e8 (intending . . . f7-f5) 1 7 g4 ! �d6 1 8 f3 ; 1 6 . . . �d7 !? 1 7 g4 �c5 1 8 f3 �d3 oo) 1 7 f3 (17 Ae3 ?! �g4 i ntending . . . f7-f5 ; 17 Ag5? gxc3 ! 18 bxc3 �xe4 with an attack) 1 7 . . . �b6+ 1 8 gf2 h6!? 1 9 �g2 gd8 20 g4 gd3 21 g5 h xg5 22 Axg5 �h7 23 gg1 ge8 ! with an unclear position (Gelfand -Aseev, USSR Championship First League, Klaipeda 1 988) ; B2) 1 2 �a3 (an almost untried idea, which I thought would be interesting for rapid chess) 1 2 . . . �b7 ! 13 �xc5

Analysis diagram after 19 e4 !

White is not afraid of ghosts. Black can sac rifice a piece i n various ways, but none of them is sufficient : A1 ) 1 9 . . . Axe4? 20 fxe4 �xe4 21 �xc7 +-;

A2) 19 . . . Ab7 20 �c3 ! Aa6 21 gf2 ± ; A3) 1 9 . . . �xe4 2 0 fxe4 Axe4 (20 . . . gxe4 21 �f2 +-) 21 �f2 Ad3 ! 22 �xf7+ �h8 23 �c3 Axf1 24 �xf1 ± ; A4) 1 9 . . . gxe4 ? ! 2 0 fxe4 �g4 (20 . . . �xe4 21 �e7 gf8 22 �h4 ! �g5 ! 23 gf2 ! �e6 (23 . . . ge8 24 Ad2 �x h4 25 g x h4 �h3+ 26 �f1 Ac4+ 27 �g2 �xf2 28 �xf2) 24 Axg5 �e1 + 25 gf1 �e2 26 �h3 �e4 27 �d2 +-) 21 gf2 �xh2 22 �xc7 ! gc8 23 gxh2 (transposing into an easily won end ing) 23 . . . gxc7 24 gxh3 gxc1 + 25 �f2 Axe4 26 g4 g5 27 ge3 Ax b1 28 b4 and White soon won (Gelfand -Timoshchenko, USSR Cham pionship First League, Sverd lovsk 1 987) ;

13 . . . e5 ! (a brill iant reply! Black is looking for an attack and he does not pay any attention to pawns) 14 gd1 (14 d xe5 Ah3) 14 . . . Ah3 (14 . . . exd4 15 �xd4 Ah3 16 f3) 15 d5 (in my preparations I had pinned my hopes on this move, which blocks the long diagonal , but Black can continue his attack) 15 . . . gac8 1 6 �a5 �e4 ! 17 f3 gc5 1 8 �e1

186

Game 42


18 . . J �xd5! (after the normal 18 . . . �d6 19 �c3 f5 20 Ae3 '8c7 21 '8ab1 Black has some com pensation for the pawn, but Vishy chooses the most challeng ing continuation) 19 '8xd5 �xd5 20 fxe4 �d4+! (an im portant intermed iate move ! - if 20 . . . �xe4 21 �f2 '8d8 22 �c3) 21 e3 �d3 (Black tries to keep as many Wh ite pieces out of the game as possible; if 21 . . . �xe4 22 �e2 followed by �c3) 22 �c3 (the only possibil ity to avoid a draw was 22 �d2 ! '8d8 23 a4 �c2 24 '8a3 '8d3 25 '8xd3 cxd3 26 b4 �xa4 00 ) 22 . . . '8d8 23 g4 (Black was threatening . . . '8d6-f6 f1 , so this move is forced ; 23 �e2 �xe2 24 �xe2 '8d1 + 25 �f2 '8f1 #) 23 . . . '8d6 24 �e2 '8g6 (now i n the event of 24 . . . �xe2 25 �xe2 '8d1+ 26 �f2 the king can escape to g3) 25 �xd3 cxd3 26 Ad2 (White could keep his extra piece by 26 �f2 ? '8xg4 27 �e1 , but he would be unable to stop Black's king side pawns after the simple 27 . . . '8g2 28 '8b1 '8x h2 fol lowed by . . . g7-g5 etc.) 26 . . . '8xg4+ 27 �h1 with perpetual check (Gelfand Anand, Melody Amber, Monaco rapid 2001 ).

a

...

b 8

6 5 4 3

t--"'-"..J-Io-

2

�:r=:!',,",-,� ...-� -.., �=.a-�

3 2

14 . . . g6 1 5 llac1 This natural move was considered a novelty here. Another plan was tested in the fol low ing game: 15 h4 '8c8 16 '8d2 Ag7 17 '8ad1 t �e7 18 �g5 h6 19 �f3 b6 20 d5 Ad7 21 �d4 exd5 22 exd5 Ag4 23 �c6 ! �xc6 24 �xg4 �e5 25 �e2 (Lputian - J . Polgar, eorus, Wij k aan Zee 2000).

15 . . . 11e8 Or 1 5 . . . '8c8 1 6 �e1 Ag7 1 7 �c2 �xc2 1 8 '8xc2 �e7 19 b3 '8fd8 20 '8cd2 Ae8 t .

7 ... a6 8 a4

16 d5

8 �xc4 b5 9 �c2 Ab7 10 Ad2 is a serious al ternative, which I used in a number of games in 2002-2004.

8 . . . Ad7 9 1txc4 Ac6 10 Ag5 a5 I adopted another plan against Vladimir Kramnik: 1 0 . . . Ad5 1 1 �c2 Ae4 12 �d1 c5 13 d xc5 Axc5 14 �xd8 '8xd8 1 5 �bd2 Ac6 1 6 �b3 �bd7 1 7 '8fc1 Ab6 1 8 �fd2 Axg2 1 9 �xg2 '8dc8 with a fairly eq ual position , which I later m isplayed (Kramn i k - Gelfan d , Astana 2001 ).

After prolonged thought I decided to go for ward . White could have kept a pleasant edge by 16 �e1 !? t .

1 6 . . . exd5 1 7 exd5 Ad7 1 8 �d4 llc8? This i nnocent looking move is the source of Black's subsequent troubles. 18 . . . Ae5 ! was more precise, combining two tasks : protect ing the c7 pawn and moving the bishop to a better position Then after 1 9 �cb5 '8c8 20 �c6

1 1 �c3 �a6 1 2 Axf6 Axf6 1 3 e4 �b4 14 11fd1 (see next diagram)

Th is is the mai n theoretical position of this line and both side's trumps are clear: Wh ite has the centre and a space advantage, but Black has the two bishops and no weak nesses in his position .

187

...


20 . . . bxc6 ! 21 d x c6 Ae6 !? (21 . . J 3e7 !?) 22 gxd8 gcxd8 23 'tlfe2 Ab3 Black has good compensation for the queen . 1 8 . . . 'tlfb8 was recommended by Sergey Shi pov, with the idea of transferri ng the queen to a7 in the style of Akiba Rubinstein , but White can stil l keep a slight edge: A} 1 9 ttldb5 Ax b5 (1 9 . . . c6 20 d xc6 Axc6 21 ttld6 'f!"e7, or 21 . . . 'f!"e6 22 ttlce4 Axe4 23 ttlxe4 Ax b2 24 'f!"b1 Ag7 25 gd7 with attacking chances) 20 'tlfxb5 ge7 = ;

29 d7 Axd7 30 gdxd7 and White is winning) 25 Af3 ttla2 (25 . . . c6 26 �x b7 ±) 26 gxa2 Axa2 27 �xa5 �a6 (27 . . . Ab3 28 �b5) 28 �xc7 ± .

20 �c5 b6 2 1 �c6! The fourth knight move i n a row ! 21 ttlde6 �e7 22 ttlxd7 �xd7 23 Ah3 �e7 leads nowhere.

21 . . . .txc6 22 dxc6 a

b

c

8

7 6

6

5

5 4

C} 1 9 ttlb3 Ag5 (19 . . . Axa4 20 ttlxa4 b5 21 'tlfxc7 bxa4 22 ttld4 �) 20 ga1 Af5 21 ttld4 Ag4 22 ttlf3 Axf3 23 Axf3 � .

3

3

2

2

19 �e4!

c

I guess that Vishy underestimated this idea. 1 9 . . . Axa4 was the alternative: A} 20 ttle6 'tlfe7 (20 . . . fxe6 21 d x e6 'tlfe7 22 gd7 Axd7 23 exd7+ �h8 24 ttlxf6 �xf6 25 'tlfe2 ! �d8 26 �b5 c6 27 d xe8'tlf+ �xe8 28 �xa5 ± Shi pov) 21 b3 fxe6 (21 . . . Ab2 22 ttlxc7 Axc1 23 d 6 ! Shi pov) 22 ttlxf6+ �xf6 23 bxa4 exd5 24 Axd5+ ttlxd5 25 �xd5+ �h8 26 �x b7 'f!"f8 and Black should be able to hold ; B} 20 b3, and now : B1 } 20 . . . Ad7 21 ttle6 (21 ttlc6 bxc6 22 dxc6 Axc6 23 'f!"xd8 Axd8 24 'f!"d1 ±) 21 . . . fxe6 (21 . . . 'tlfe7 22 ttlxc7) 22 d xe6 Axe6 23 gxd8 Axd8 24 'tlfd4 and Black has some compen sation for the queen , but it is not sufficient ; B2} 20 . . . Axd4 21 'tlfxd4 gxe4 22 Axe4 Ax b3 23 gd2 'tlfd6 24 'tlfa7 ! (this is the move I was counting on) 24 . . . ge8 (24 . . . c5 25 �x b7 ge8 26 'f!"e2 ! �f8 27 'tlfa7 ± ; 24 . . . b6 25 �b7 'f!"e8 26 �xc7 �xc7 27 'f!"xc7 'f!"xe4 28 d6 Ae6

•

e

8

B} 1 9 ttle4 Ag7 !? (1 9 . . . Ae5 20 ttlc5 Ag4 21 ge1 'tlfc8 22 ttlx b7 !) 20 'tlfxc7 (20 ttlc5 Ag4 21 gd2) 20 . . . 'tlfxc7 21 gxc7 Ag4 22 gd2 gad8 and Black obtains quite good compen sation for the pawn (Shi pov) ;

19 . . . .te5

d

d

e

f

9

h

22 . . . bxc5! Surprising ly, my opponent sacrificed his queen instantly, but it is clearly the best prac tical solution. Actually this is not first time that Anand has used such a sacrifice, so perhaps he should apply for copyright! After 22 . . . �e7 23 'f!"d7 �f8 24 ttle4 Ax b2 25 gcd1 White has total domi nation. If 22 . . . �f6 23 ttld7 (23 ttle4 �e6) 23 . . . �e6 24 �xe6 (or 24 ttlxe5 �xe5 25 Ah3 gcd8 26 gd7 ±) 24 . . . gxe6 25 Ah3 ! (25 ttlxe5 gxe5 26 gd7 ttla6 (26 . . . �f8 27 Ah3) 27 Ad5 ttlc5 28 Axf7+ �f8 with counterplay} 25 . . .f5 26 Af1 he has a clear advantage.

23 gxdS gcxdS 24 ge1 ! It is important to exchange one pair of rooks to limit Black's counterplay: 24 Ah3 ?! Ax b2 25 'f!"b1 gd2 .

24 . . . .td4 25 gxeS+ DxeS 26 1tb5 26 Af3 does not prevent Black's counterplay after 26 . . . ge1 + (26 . . . ge6 !?) 27 �g2 gb1 .

188

26 . . . �g7

Game 42

a

b

c

d

e

Gelfand -Anand , World Cup, Shenyang 2000

30 b3 c4 31 bxc4 ge1 + 32 �xe1 �xe1 33 �xe1 should be win n i ng) 30 AdS (30 �d2 .E!e1 + 31 �xe1 �xe1 32 �xe1 Ax b2 33 �d2 =) 30 . . . ge1 + (30 . . . ge7 !?) 31 �xe1 (31 �g2 .E!e2+ 32 �h3 Ag1 with counterplay) 31 . . . �xe1 32 �xe1 Ax b2 and it is not clear whether this endgame is won . It requ i res a special investigation, but I rather think that it is a draw.

f

8

8

7 6

6

5

5 4

3

3

2

2

L...-_______----=-__.....

27 . . . 11e2 28 1!!fx c7 llxf2 29 �h1 29 h4 is not a good idea: 29 . . . �d3 30 �d8 .E!d2+ 31 �h2 .E!d1 32 AdS .E!d2+ =.

lf

29 . . . �d3

Suddenly all three of Black's pieces are ready for an attack against the white ki ng.

a

26 . . J �e1 + 27 Af1 would not have hel ped :

8

A) 27 . . J �e6 28 �x aS eDxc6 29 �a8+ �g7 30 �g2 +-;

7

B) 27 . . J � b1 28 �x aS (28 �b8+ �g7 29 �xc7 gx b2) 28 . . . .!�x b2 29 �xc7 �g7 30 AbS .E!xf2 31 �h1 �dS 32 �d8 �e3 33 c7 .E!c2 34 h4 eDg4 3S �f8 + ! ! �xf8 36 c8�+ �g7 37 �xg4 +-.

b

c

d

e 8

6

6

5

5

27 1!!fx a5 With 27 �f1 !? it looks as though Wh ite could have destroyed the coord i nation of his op ponent's pieces, but : A) 27 . . . c4 28 �xaS (28 �xc4 Ab6 should also be won in the long run) 28 . . . �d3 29 AdS .E!eS 30 �a6 ! (I obviously missed this com puter style move) 30 . . . .E!xdS 31 �xc4 ; B) 27 . . . �c2 28 AdS .E!eS 29 Ac4 ; C) 27 . . . geS !! An i ncred i ble resource ! Now White has to chose between : C1 ) 28 �xaS �d3 29 Af3 .E!fS ; C2) 28 Af3 .E!fS 29 �g2 c4 30 �xc4 Ab6 31 g4 and White will probably be able to break into the fortress by exposing the black king with g4-gS and h4-hS after i mproving the position of his queen , but it is far from clear; C3) 28 f4 ge3 29 �xaS �c2 (29 . . . g b3 30 �xc7 .E!x b2 31 Ae4, or 29 . . . eDd3

'--_______-=-_----'

If

30 1!!fd 8? This move throws away the win . However, at the board it was almost im possible to find the road to success . As shown by Shipov, 30 Ae4 would also not have won after 30 . . . �eS 31 h3 (31 �c8 fS or 31 . . . hS 32 c7 eDg4) 31 . . . .E!f1 + 32 �h2 .E!f2+ 33 Ag2 .E!x b2 (33 . . . �f3+? 34 �h1 �e1 3S �f4 ! .E!xg2 36 �f1 +-) 34 g4 �d3 3S �g3 (3S �d6 ? �e1 !) 3S . . . .E!b1 36 Ah1 (36 �xd3 AeS+ 37 �g3 Axg3+ 38 �xg3 .E! b8 39 as 39 . . . .E!c8 ! 40 a6 .E!c7) 36 . . . .E!b2+ =. Correct was 30 AdS ! eDeS (30 . . . �e1 31 �xf7+ .E!xf7 32 Axf7 AeS 33 as and one of the pawns will queen) and now : A) t h e natural 31 a s is not good enough to win : 31 . . . eDg4 (31 . . . gS !?) 32 �xf7+ gxf7 33 Axf7 eDf2+ 34 �g2 �e4 3S c7 (3S a6 c4 36 Axc4 eDd6) 3S . . . �d6 36 a6 c4 37 Ae6

189


�f6 38 c8� ttlxc8 39 Axc8 �e5, for ex ample 40 �f3 �d6 41 �f4 �c7 42 Ae6 h6 43 Axc4 Ax b2 44 Af7 g5+ 45 �g4 � b6 46 Ac4 Ad4 47 �h5 Ag1 48 h3 Ae3 49 g4 Af2 50 �xh6 Ah4 = ;

35 �g2 ge2+ 36 �f1 , or 34 . . . ttle3 35 g4 gx b2 36 Ae4) 35 �g1 ttld3+ 36 �xd4+ cxd4 37 c8� ge1 + 38 �g2 ge2+ 39 �h3 gS 40 Af3 g4+ 41 Axg4 ttlf2+ 42 �g2 ttlxg4+ 43 �f1 ± and Wh ite retains wi nning chances;

B) 3 1 �c8. This was m y original i ntention when I played 27 �a5, but when consid ering my 30t h move I found that Black has counter-resources :

C) 31 h4!!

B 1 ) 31 . . . f5 32 �g8+ ! (32 c7 gf1 + 33 �g2 gf2+ 34 �h3 (34 �g1 gx b2+ 35 �f1 g b1 + 3 6 �e2 gb2+) 34 . . . ttlg4, o r 3 2 �b7+ �h6 33 c7 (33 Ag8 gf1 + 34 �g2 gf2+ 35 �h3 ttlf3 36 �x h7+ �g5 37 �e7+ with perpet ual check) 33 . . . gf1 + 34 �g2 gg1 + 35 �h3 g5 36 Af3 g4+ 37 Axg4 ttld3 ! ! (an extraor d i nary resource! - 37 . . .fxg4+ 38 �h4 or 37 . . . ttlxg4 38 �f3 loses instantly) 38 �f3 fxg4+ with counterplay) 32 . . . �h6 33 �f8+ �h5 34 �e7 +- and the centralised queen is strong enough to prevent Black from setting up a mating net ; B2) 31 . . . h 5 ! 32 �h3 ! (another study-l i ke possibility for White, found by com puter; if 32 h3 gf1 + 33 �h2 gf2+ 34 Ag2 gxb2 , 3 2 �a6 ttlg4 3 3 Ag2 gxb2 34 h 3 g b1 + 3 5 Af1 ttle5 with cou nterplay (36 c7 ? loses after 36 . . . c4) , or 32 c7 ttlg4 33 �xg4 gf1 + (33 . . . h xg4 34 h4) 34 �g2 gf2+ =, as bad is 35 �h3? h x g4+ 36 �xg4 f5+ 37 �h3 g5 38 g4 Ae5 -+) 32 . . . ttlg4 (32 . . . gx b2 33 �f1 ) 33 c7 ge2 (33 . . . gd2 34 �xg4 h xg4 35 h4)

Analysis diagram after 33 c7 l1e2

34 �h4 !! (34 c8� ttlf2+ 35 �g2 ttld3+ ! with perpetual check) 34 . . . ttlf2+ (34 . . . ge1 +

First White has to force Black's pawn to h5, after which his kingside pawns will be vul nerable: C1 ) 31 . . . ttlg4 32 Ag2 h5

33 �f4!! (it is important to stop the counter attack and now the pawns will decide) 33 . . . gxf4 34 gxf4 ttlf2+ 35 �h2 Ae3 36 as ! (36 �g3 ttld3 ! 37 �f3 Axf4 38 Af1 c4 39 Axd3 cxd3 40 a5 g5 =) 36 . . . Axf4+ 37 �g1 ttld3 (37 . . . ttlg4 38 a6 c4 39 a7 Ae3+ 40 �f1 Axa7 41 c7) 38 a6 +-; C2) 31 . . . h5 32 a5 ttlg4 (32 . . . gx b2 33 a6 g b1 + 34 �g2 g b2+ 35 �h3, or 32 . . . gf1 + 33 �g2 gf2+ 34 �h3 ge2 35 �e7 +-) 33 �xf7+ gxf7 34 Axf7 ttlf2+ 35 �g2 (35 �h2 ttle4) 35 . . . ttle4 36 c7 ttld6 37 a6 c4 38 Ae6 (38 Axc4 �f6 39 �f3 �e7) 38 . . . �f6 39 c8lt ttlxc8 40 Axc8. Now this ending is

190

Game 42


won , as the g6 and h5 pawns are fixed on light squares and the black king cannot both protect them and prevent its white opponent from moving across to support the a-pawn , for example 40 . . . �e5 41 Ad7 �e4 42 Ae8 �f5 43 �f3 +-.

30 . . . .6c2! 30 . . . '!;x b2 31 c7 is i nsufficient : A) 31 . . . �f2+ 32 �g1 �g4+ 33 �xd4+ cxd4 34 c8� '!;b1 + 35 Af1 �e3 36 �a6 �f8 37 a5 �e7 (37 . . . ,!;xf1 + 38 �xf1 �xf1 39 �xf1 ) 38 �f2 ,!;xf1 + 39 �xf1 �xf1 40 a6 ; B) 31 . . . '!; b1 + 32 Af1 ,!;xf1 + 33 �g2 ,!;g1 + 34 �f3 (34 �h3 �f2+ 35 �h4 ,!;g2 36 �xd4+ cxd4 37 c8� h6) 34 . . . �e5+ 35 �e4 �c6 36 a5 ,!;a1 37 �d6 ± .

31 £f1 31 Af3 also leads to a d raw : 31 . . . '!;c1 + 32 �g2 '!;c2+ 33 �h3 �f2+ 34 �h4 �d3 ! (34 . . . Af6+ 35 �xf6+ �xf6 36 c7 g5+ 37 �h5) 35 �h3 = (35 h3 '!;c4) .

40 a5 h5+ 41 �h4 Af6+ 42 �xf6+ �xf6 43 a6 �f3+ 44 �h3 �g5+ 45 �g2 �e6 ; B) 35 �f1 '!;f2+ 36 �e1 �d3+ 37 �d1 �x b2+ =. 33 . . . �e5+

Black could have forced a draw with 33 . . . f5 34 �e7+ �h6 35 �h4+ �g7 , but Vishy wants to play on. 34

ct>e4 �xc6 35 flc7 �b4

I could hardly bel ieve that White's king on e4 could be in the middle of a mating net !

36 g4?!

31 . . . �f2+ 32 ct>g2 �g4+ 32 . . . �d3+ ? 33 �h3 �f2+ (33 . . . �e5 34 c7 �f3 35 c8� �g1 + 36 �h4 ,!;xh2+ 37 Ah3 Af6+ 38 �xf6+ �xf6 39 �h8+) 34 �h4 Af6+ 35 �xf6+ �xf6 36 c7 +-.

A careless move; I should have settled for a draw with 36 b3 '!;f2 37 Ac4 �f8 38 �d8+ �g7 39 �e7 �h6 =.

36 . . . .6f2 37 £b5 Fortunately, there is sti l l a way to save the game.

33 «i!n3 Or 33 Ae2 �e5 ! (33 . . . ,!;xe2+ 34 �f3 ,!;xh2 35 c7 �f6 36 �xd4 cxd4 37 c8� ,!;x b2 with counterplay) 34 c7 ,!;xe2+, and now : A) 35 �h3 (it i s incredible, but after this move Wh ite cannot even make a draw) 35 . . . �f3 36 c8� �g1 + 37 �h4 ,!;e4+ 38 �g4 ,!;xg4+ 39 �xg4 Ae5 !!

37 . . . g5 Or 37 . . . �h6 (threatening . . . f7-f5+) 38 �d6 �g5 39 �g3.

38 a5 Draw agreed After 38 . . . '!;f4+ 39 �xf4 g xf4 40 a6 = White regains a piece.

* * *

191

My Most Memorable Games G am e 4 3

B o r i s G e l fa n d - J e ro e n P i ket F I D E Wo r l d C h a m p i o n s h i p , N ew De l h i 2 0 0 0 Ragozin Defence [D38J Th is game was played i n the 2000 F I D E world knock-out championship i n New Delhi. Normally I advanced q u ite far i n such tour naments but failed to prod uce any good games, as I think the formula is too stressful to encourage players to be creative.

1 d4 �f6 2 c4 e6 3 �f3 d5 4 �c3 Ab4 The Ragozin Defence is quite a rare guest in tournament play, but it is a solid set-up.

5 Ag5 h6 S . . .fi:lbd7 leads to another type of pos ition and is cal led the Westphal ia Defence. S . . . d xc4 6 e4 transposes i nto the Vienna Variation .

6 Axf6 'ttxf6 7 e3 0-0

A) 8 . . . �d8 9 a3 Axc3+ 1 0 �xc3 dxc4 1 1 �xc4 c6 1 2 Ad3 ttld7 1 3 0-0 (13 eDeS ttlxeS 14 d xeS 'ffIe7 , i nten d i ng . . . Ad7-e8) 13 . . J �e8 N (13 . . . eS ?? 14 dxeS eDxeS 1S eDxeS 'ffIx eS 1 6 Ah7+) 14 Ab1 eS 1 S d x eS eDxeS 1 6 eDxeS �xeS 17 �f4 'ffIe7 18 �d4 Ae6 and Black almost eq ualised (Gelfand - M ilov, Cred it Suisse, Biel 1 997) ; B) 8 . . . eDc6 ?! (I don't l i ke this move as Black will be u nable to attack the centre by . . . c7cS , but this is a popular plan i n this opening and it is probably just matter of taste) 9 Ad3 d x c4 1 0 Axc4 eS 1 1 0-0 Axc3 ?! (allow ing White to strengthen his centre; stronger is 1 1 . . . exd4 1 2 ttldS 'ffId 6 1 3 ttlxd4 eDxd4 14 ttlx b4 eDfS = or 14 'ffIx d4 AaS with a solid game) 12 bxc3 Ag4 1 3 h3 Ae6 (13 . . . AhS 14 AdS ! �ad8 1S Axc6 bxc6 16 g4 e4 17 tileS Ag6 1 8 'ffIa4 cS 1 9 'ffIx a7 ±) 14 Axe6 fxe6 1S �b1 �ab8 16 �bS exd4 17 cxd4 a6 1 8 �cS ± (Gelfand - G reenfeld , Wydra Me morial , Haifa rapid 2000).

9 Axc4 c5 10 0-0 cxd4 1 1 exd4

Genna Sosonko, a great expert on this open ing, once wrote that 7 . . . cS is more in the spirit of the variation .

a

b

c

d

..

e 8

7

7

8

6

6

7

5

5

6

6

4

4

5

5

3

3

4

4

3

3

"::;;;;:;'��-I

2

9

_ L--_______---:_ ::: ....J

'If

S gc1 Antici pating . . . dxc4, White fights for a tempo and refrains from developing his bishop.

S

•..

dxc4

I have also had to face other replies :

h

This type of position with an isolated pawn often arises from different openings, for ex ample the Tarrasch Defence, the N imzo I n d ian Defence etc. I think it is i n White's favour that he has exchanged his bishop for the knight at f6, as the knight wil l not be able to blockade the pawn from dS. Now the main q uestion is whether Blac k can bring his c8 bishop i nto play.

192

Gelfand - Piket,

Game 43

FIDE

World Championship, New Delhi 2000

11 ttlxd4 is harmless in view of 11 . . . Ad7 ! 12 f!Vb3 ttlc6 ! 13 ttlxc6 Axc3 14 f!Vx b7 Axc6 15 f!Vxc6 Ax b2 .

1 1 . . . �c6 12 �e4 "f4 The alternative was 1 2 . . . f!Ve7 or 1 2 . . . f!Vd8 , but Black wants to disturb White's pieces.

13 "e2 gd8 14 g3 Avoiding 14 gfd1 Ae7 15 d 5?? exd5 0-1 (Vandevoort - Sosonko, Brussels Zonal 1993), or 16 Axd5 gxd5 17 gxd5 f!Vxc1 +.

21 Af3 !? oo) 21 . . . f!Vx b2 22 gb1 f!Vc2 23 Ad3 f!Va2 as he cannot trap the queen ; B} 17 gfd1 f!Vf5 18 �c5 (18 a3 e5 !) 18 . . . f!Vf6 ! and now : B 1 } 1 9 a 3 Axc5 20 d x c5 (harm less i s 20 gxc5 ttle7 followed b y . . . b7-b6) 20 . . . gxd1+ 21 gxd1 f!Vx b2 and it is not clear whether White has fu ll com pensation for the pawn ; B2} 1 9 �e5, and now :

14 . . . ..g4 To be honest, I was surprised by my op ponent's choice of opening i n such an im portant game (he had a broad repertoire of open i ngs and he used almost all of them against me in our n u merous meetings) . The entire concept is one that I i nvented over the board during a game with Joel Benjami n and si nce then 1 1 ex d 4 has been success fu lly tried by a number of players. 14 . . . f!Vf5 !? followed by . . . Aa5-b6 deserved considera tion. As we will see later, the position of the queen on g4 gives Wh ite an extra tempo. a

b

c

d

e 8

7

7

6

6

5

5

4

4

3

3 2

a

b

c

d

e

11

1 5 gfd1 The alternative was 1 5 f!Ve3 Aa5 1 6 Ae2 Ab6 : A} White should avoid the tem pting 1 7 ttle5 f!Vf5 18 ttlxc6 (18 Ag4?? f!Vxe5 !) 18 . . . b x c6 19 �c5 f!Vd5 20 gfd1 f!Vxa2 21 Ac4 (better is

B21 } 1 9 . . . ttlxe5 and White has a choice be tween 20 dxe5 gxd1 + (20 . . . f!Ve7 t Korchnoi) 21 gxd1 (21 Axd1 f!Ve7 22 Af3 gb8) 21 . . . f!Ve7 22 b4 a5 23 a3 axb4 24 axb4 Axc5 ! 25 bxc5 (25 f!Vxc5 25 . . . f!Vxc5 26 bxc5 �f8 ! t and Black has good chances of saving the game thanks to the l i m ited amount of material left) 25 . . . Ad7 26 Af3 Ac6 27 Axc6 bxc6 28 gd6 ± , and 20 f!Vxe5 !? f!Vxe5 21 d x e5 gxd 1 + 22 Axd1 Axc5 23 gxc5 Ad7 24 Af3 gc8 25 ga5 ± ; B22} 19 . . . �b4 ? 20 a3 (20 Ah5 !? was tempt ing and equally strong : 20 . . . g6 21 f!Vx h6 f!Vg7 (21 . . . ttlxa2 22 gc4) 22 f!Vxg7+ �xg7 23 Af3 ttlxa2 24 ga1 (24 gc4 !?) 24 . . . ttlb4 25 �c4 ±} 20 a3 �d5 21 f!Ve4 Ac7 22 Ad3 g6 23 f!Ve2 ± and Black has failed to develop his c8 bishop, whereas Wh ite has put his pieces on their best squares (Gelfand - Benjami n , Cred it Suisse, Horgen 1 994).

1 5 . . . Ad7 The plan with 15 . . . Aa5 was used by Ben jamin.

193

1 6 a3 Ae7 17 "e3


White is playing for a spatial advantage and trying to make use of the fact that several of Black's pieces (queen, knight and bishop at d7) lack good squares.

17 . . . .te8 18 b4 a6 If 1B . . . gacB 1 9 bS �aS 20 Ae2 ;t , shutting the as kn ight out of the game followed by �eS.

19 .te2 1 9 dS exdS 20 AxdS gacB = is not i n spirit of the position, as Black does not have any weaknesses.

19

..•

22 .td3 .f6 23 .te4! 23 h4 would allow 23 . . . �e7 24 �eS Aa4 with counterplay.

23 . . . Dxd1+?! Black fails to find a good defensive plan . Cor rect was 23 . . . �e7 !? 24 h4 gxd1 + 2S gxd1 gdB 26 gb1 ;t .

24 Dxd1 Dc8 24 . . . gdB ! ? 2S g b1 (2S gd6 !?) 2S . . .' tWe7 26 a4.

25 h4 m8 a

• h3?!

Wasting a tempo, as in any case the queen will have to return to fS. 19 . . . �fS 20 �cS AxcS 21 dxcS �f6 22 Ad3 ;t .

b

c

d

e

8

8

7

7 6

20 �c5 20 �c3 fol lowed by d4-dS also came i nto consideration , but I didn 't want to deviate from the chosen plan.

5

5

4

4

3

3

2

2

20 . . . .txc5 a

Neither would 20 . . . bS 21 Ad3 or 20 . . . gabB 21 Ad3 have solved Black's problems.

21 dxc5 Now White's plan is to play his bishop to e4, followed by a3-a4 and b4-bS. If 21 gxcS b6 22 gcc1 �e7.

21

. . .

•f5

Al l Wh ite's pieces are excel lently placed , so it is time to start an attack!

26 . . .•e7 26 . . . gdB !? 27 ge1 .

8

27 g5 h xg5 28 hxg5 Dd8 a

b

4

4

3

3

2

2

d

e 8

7

7 6

6 5

c

8

7

5

'lr

c

26 g4!

21 . . . gxd 1 + 22 gxd1 gdB 23 gxdB �xdB 24 �eS ± , or 21 . . . �e7 22 �eS �dS 23 �e4.

7

b

5

5

4

4

3

3

2

2

'lr 'lr

29 De1 ! +-

194

Game 44

Gelfand - Van Wely, European Team Championship, Leon 2001

The attacking side should normally avoid un necessary exchanges. Black can not make use of the d -fi le, whereas the rook on e1 exerts pressure on eS. However, 29 gxd8 �xd8 30 �f4 ± was also strong.

29

. . .

30 . . . fS 31 AxcS AxcS 32 ttld4 +-.

31 �g5 �d4 31 . . . Af? 32 AxgS, or 31 . . . �e? 32 AxcS AxcS 33 ttlxe6+.

32 .xd4 Black resigns

•c7

In 1 993 I worked for several weeks with the H ungarian G M Peter Lukacs on positions with an isolated pawn, after which I won a n u m ber of games i n this type of position i n d ifferent open ings.

Or 29 . . . g6 30 �f4 �g? 31 �g2 .

30 g6! Destroying the king's residence.

30 . . . fxg6

*** Game 44

B . . . e6 9 0-0 �fd7 10 f4

B o r i s G e l fa n d - L o e k Va n We l y European Team Championsh ip, Leon 2001 GrOnfeld Defence [0 79] 1 d4 �f6 2 c4 g6 3 g3 Already Wh ite's third move requ i res some com mentary. This was the last round of the European Team Championsh ip, which began at 9 a. m . - too early for chess players ! And as I had a long game the even ing before, I decided not to enter i nto a theoretical dis cussion in the GrOnfeld Defence but to steer the game in a quieter direction .

6

6

5

5

4 3

="--'-...1'--1 =",--_1

4 3 2

9

h

10 . . . �b6?!

3 . . . c6 The most solid reply, but it hard ly su ites Loek's aggressive style. I had some neg ative experience from Black's side during my world championsh i p sem i-fi nal match with Anatoly Karpov, and so I was fami l iar with a num ber of finesses in this type of position .

4 Ag2 d5 5 cxd5 cxd5 6 �c3 Ag7 7 �f3 0-O B �e5 Wh ite occupies the centre with his knight before castl i n g , i n order to prevent . . . ttlcS. However, as Ken Neat rightly poi nted out, 8 . . . ttlcS 9 ttlcS bxcS is q u ite possible, as Wh ite cannot profit much from the weakened pawn on cS. I n a cou ple of games where Black had tried this, he satisfactorily solved his open ing problems.

The fi rst inaccu racy, and a serious one. 1 0 . . . ttlcS (intending . . . ttld xe5) 1 1 Ae3 ttlbS was better, as now Wh ite's bishop will be wel l placed on a3.

1 1 b3 f6 12 �d3 lif7 1 2 . . . ttlcS 13 e3 ;t .

13 a 4 1 3 . . . �c6 14 e3 libB Van Wely is reluctant to fix the pawn struc ture by 14 . . . f5 as it wo uld weaken his e5 square. However, in any case he lacks space and good squares for his pieces.

15 g4 Gaining more space and creating the poten tial threat of f4-f5.

195

1 5 . . . a5 16 Aa3 AfB


a

b

c

d

21 . . . �a6?!

e

8

Too slow. Black should have aimed for a counterattack by 21 . . . b6 !?

8

7 6

6

5

5

22 e4 dxe4 Playing into Wh ite's hands 22 . . . b6 was stubborn .

4 3

3

2

2

more

23 Axe4 Ild7 24 .c3 b6 A sense of danger has never been the main tru m p of the strongest Dutch player of the 21 st century. H owever his position was al ready poor:

�------��- �

A) 24 . . . tDd4 25 gce1 ;

17 Ac5 17 Axf8 gxf8 18 gc1 was more natural . How ever, I wanted to exert more pressure on the queenside.

17 . . . �g7 18 Ilc1 The tempting 18 f5 exf5 1 9 Axb6 �x b6 20 tDxd5 �d8 21 gxf5 Axf5 22 e4 Ac8 is a mistake, as White gives up his bishop with out getting something significant i n return .

B) 24 . . . �g8 25 gce1 ± , and if 25 . . . f5 26 gxf5 g x f5 (26 . . . exf5 27 Axc6 b x c6 28 tDd6) 27 ge2 (27 ge3 !? gg7+ 28 Ag2) 27 . . . gg7+ (27 . . . tDe7 28 gg2+ tDg6 29 tDe5 fxe4 30 tDxg6 h xg6 31 gxg6+ �h7 32 gg5 +-) 28 gg2 ; C) 24 . . . tDcb4 25 tDx b4 axb4 26 �e3 +-; D) 24 . . . tDe7 25 g5 tDd5 26 Axd5 exd5 27 �d4 h6 28 c6 bxc6 29 gxc6. a

18 . . . Axc5

b

c

d

e

8

Other moves do not change the character of the position and leave White with his pos itional tru m ps : 18 . . . f5, 18 . . . tDd7 19 Axf8+ gxf8, or 18 . . . Ad7.

8

7 6 5

19 dxc5!?

4

I decided to open up the position . Black's pieces are badly placed , so if White is able to break with a move such as e3-e4, he should have a strong initiative. 19 tDxc5 ;t .

20 �b5 �c7 21 .e1

3

2

2

�------��-

19 . . . �a8 19 . . . tDd7 does not look natural , as it would destroy the coordination of the black pieces.

3

25 �e5! It is natural that White should have a tactical win in this position. If 25 g5 tDd4.

25 . . . �xe5

Preparing e3-e4. 21 tDd6 gfB 22 f5 exf5 23 g xf5 would have been met by 23 . . . g5. However, serious consideration should have been given to 21 g5 ! ± , weakening the dark squares sti ll further, and if 21 . . . fxg5 22 fxg5 �xg5 23 gxf7+ �xf7 24 tDxc7.

�

25 . . .fxe5 26 Axc6 gd3 27 �xe5+ .

26 fxe5 �xc5 26 . . . f5 27 c6.

27 exf6+ � Or 27 . . . �g8 28 gcd1 ! gxd1 29 f7+.

196

Gelfand - Delchev, FIDE World Championship, Moscow 2001

Game 45

30 Ae8+ !!

28 "h3 "g8 29 Ac6 Dd3 Black makes the only moves to avoid losing instantly, but it allows a beautiful fin ish . a

b

c

d

e 8

8 7 6

6

5

5

For the next three moves Wh ite decoys the opponent's pieces and then begins collect ing the harvest. I n fact, a few days earl ier Genna Soson ko told me a story of a journal ist who asked h i m : ' I s it possible in chess first to g ive up fou r draughtsmen and then get nine bac k ? ' In a way I managed to do this!

3O . . . �xe8 31 f7+ "xf7 32 �d6+ ! Dxd6 33 I1xf7 �xf7 34 "xh7+ �e8 35 "c7 !

4 3

The final touch.

2

L...-_______-""__-'

35 . . . l1d4 36 "xb8 �d7 37 h3 Black re signs

{f * * *

Game 45

liquidates i nto a slightly worse end ing with 8 . . . �xc3 9 bxc3 �xd4 10 'tWxd4 'tWxd4 11 cxd4 Ab4+ 1 2 Ad2 Axd2+ 13 �xd2.

B o r i s G e lfa n d - A l e x a n d e r D e l ch e v FIDE World Championsh ip, Moscow 2001 Queen 's Gambit [04 1J Th is game was selected for this book for two reasons: I managed to i ntroduce two strong novelties in one game and I think it is very instructive for Catalan-type pOSitions.

7

1 �f3 �f6 2 c4 e6 3 �c3 d5 4 d4 c5 5 cxd5 �xd5 6 g3 �c6 6 . . . cxd4 deserves serious attention here, as after 7 �xd5 'tWxd5 8 'tWxd4 �c6 9 'tWxd5 exd5 Black is only sym bolically worse.

6

6

5

5

4

4

3

3 2

� {f

7 Ag2

�_______�__

This position has arisen many times from the Eng l ish Opening, for example in the games of Vladimir Kramnik.

7 . . . cxd4 7 . . . ttldb4 or 7 . . . �b6 is playable. 7 . . . Ae7 would transpose i nto a Tarrasch Defence.

8 �xd4 �db4?! (see next diagram)

Th is continuation had been tried only once before in tournament play. Normally B lack

9 �xc6! N Vishy Anand tried more a com plicated ap proach, but I l i ke my move. 9 �db5 a6 (9 . . . 'tWxd 1 + 10 �xd1 �a6 11 a3 i nten d ing b2-b4 ;t) 1 0 'tWxd8+ � x d 8 1 1 �a3 (the king is badly placed at d8) 1 1 . . . e5 12 0-0 Ae6 13 Ae3 �c7 14 b3 gc8 1 5 gfc1 �b8 1 6 �c4 �d4 ! 17 gab1 ! (th reaten in g �xe5) 17 . . . Axc4 (17 . . . f6 18 �f1 ) 18 bxc4 gxc4 19 Ad2 ! b5, and now 20 e3 ! �xa2 (20 . . . �e6 21 Af1 ! i ntend i n g a2-a4 ±) 21 �xa2 �e2+

197


22 �f1 �xc1 23 �xc1 e4 24 �e1 wou ld have g iven White a clear advantage, al though after 20 �f1 ? Ae7 he nevertheless won (Anand - Krasenkow, Madrid 1 998).

9

. . .

a

b

c

d

e

8 7

•xd1 + 10 c!>xd1

1 0 �xd1 , as proposed by Anand, is i l log i cal : 10 . . . !i)xc6 (after 10 . . . !i)c2+ 1 1 �d2 !i)xa1 12 !i)eS Wh ite follows up with !i)c3 and !i)d3, and the knight at a1 will eventually be lost) 11 0-0 Ad7 =.

6

6

5

5

4

4

3

3 2

�------��- �

10 . . . .t1xc6?! I wou ld prefer 10 . . . bxc6!? ;!; , securing the dS square for the kn ight and trying to gain coun terplay against the white king. But the pawn on c6 should give Wh ite some advantage.

11 .te3 The most natural move, but it was better to keep the black king i n the centre : 11 !i)bS ! �d8 1 2 Ae3 Ad7 1 3 gc1 ± . 1 1 Axc6+ b x c6 1 2 �c2 eS = was harmless.

11 . . . .td7 12 Hc1 .te7 13 �c2 0-0 14 Ilhd1 IltdS 15 c!>b1 This type of position arises q u ite often, so I would l i ke to explain the ideas beh ind it. White's bishop is exerting strong pressure on the opponent's queenside and his rooks con trol the central fi les. Another i m portant fac tor is that Wh ite's king is on the queenside, closer to the potential field of action . But if B lack should succeed i n exchanging the lig ht-square bishops, he will reduce White's advantage to the minimum.

15 . . . .t1a5 I ntend ing to exchange bishops, but now the poor position of the knight becomes an i mportant factor. After 1 S . . . gac8 16 !i)bS a6 17 !i)d6 (17 !i)a7 !i)xa7 18 Axa7 Ac6 19 gxd8+ Axd8 20 Axc6 gxc6 21 gxc6 b x c6 and White's advantage is symbol ic) 1 7 . . . Axd6 18 gxd6 eS 1 9 Ae4 the two bish ops ensure White an advantage.

1 6 .tt3!! The most d ifficult move of the game! White protects his bishop and thus makes its ex change more d ifficult. 1 6 h4 Ac6 1 7 gxd8+ gx d8 1 8 Axc6 !i)xc6 ;!; would have left him with merely a sym bol ic advantage. As I d iscovered afterwards, this position was reached in a game between two great play ers (via a different move order) . Wh ite failed to achieved anything and the game ended quickly: 16 !i)e4 AbS 17 gxd8+ gxd8 18 Ac5 �f8 1 9 Axe7+ �xe7 20 gcS a6 21 !i)c3 Ac6 %-% (Keres - Korchnoi, USSR Team Cham pionsh ip 1 963).

16 . . . .tc6 Blac k insists on carrying out his plan, but now the knight at as is left out of the game. However, in any case White wou ld have kept a slight edge: 1 6 . . . �c4 1 7 Af4, 16 . . . �f8 1 7 !i)e4, 1 6 . . . Ae8 1 7 gxd8, or 1 6 . . . fS 17 g4! Ac6 1 8 g xfS Axf3 19 exf3 exfS 20 gxd8+ Axd8 21 !i)bS .

17 IlxdS+ J.xdS 17 . . . gxd8 18 Axa7 Axf3 19 exf3 leaves Black a pawn down .

1S .t1e4 Now b2-b4 is threatened .

1S . . . .te7 1 8 . . . AdS 1 9 !i)d6 Axf3 20 exf3, and if 20 . . . Ae7 21 !i)x b7 ! !i)x b7 22 gc7 , winning a pawn .

198

Game 45

a

b

c

d

Gelfand - Delchev, FIDE World Championship, Moscow 2001

21 �d6 Axf3 22 exf3 I1d8

e

_"--='_---.,.;=-0-1 8

After 22 . . . b6 23 'Sc? 'Sd8 24 �c8 White wins a pawn, but sti l l this was the best chance, as the rook ending that arises in the game is completely hopeless.

6 5

5

4

4

23 I1xa5 Hxd6 24 I1xa7 Hd1 + 25 c2 Hf1 26 I1xb7 I1xf2+ 27 d3 I1xh2

3

Or 2? . . 'Sxf3+ 28 �e2 trapping the rook.

28 a4 Hh1 L..-_______....__ .;;:. --1

{f

8

8 i--�'"

19 Ac5!

7

The only way to maintain the pressure. If 1 9 �c5 Axf3 20 exf3 Axc5 21 '8xc5 �c6, or 19 Ad2 f5 20 �c5 Axf3 21 exf3 Axc5 22 'Sxc5 �c6 and the knight is not worse than the bishop.

5

5

3

-"--l

2

If 1 9 . . . �f8 20 b4. I considered the most stubborn to be 19 . . .f5 20 Axe? fxe4 21 Ag4 Ad5 22 'Sc? �c6 23 Ac5 ± and Wh ite has the advantage of the two bishops, but at least the knight is back in the game. 20 . . . b6 21 'Sc3 f5 (21 . . . Ad5 22 �d6 Axf3 23 exf3 'Sd8 24 'Sc? f6 25 �c8) 22 �g5 Axf3 23 �xf3 and Black's d ifficulties are far from over.

6

--...--,

19 . . . Axc5

20 I1xc5 f5

6

4 3 2

...:_ :--J {f

L-_______

29 11b3! Taking the opportun ity to put the rook be h in d the paw n , as the win would be more complicated after 29 a5 'Sa1 30 b4 'Sa3+.

***

199

29 . . . � 30 a5 l1a1 31 l1a3 I1d1+ 32 cc9c4 Black resigns


9 Ad2

M i k h a i l G u revi ch - B o r i s G e lfa n d Corus To u r n a m e n t , Wij k a a n Zee 2002 Semi-Sla v Defence [D45J After my victory in 1 992 (together with Valery Salov) and my candidates match i n 1 994, I have not played wel l or achieved good re su lts i n Wij k aan Zee. This is very unfortu nate, as I really enjoy playing i n one of the oldest festivals in the chess world (it has al ready been taki ng place for 66 years !) which is wel l-organised and has its own un ique at mosphere. Around a thousand people come every day from all over the N etherlands, as wel l as from other countries, to watch the games. I didn't start wel l i n this tournament either, so I desperately wanted to win this game in order to get back on track. But how to do this with Black against a strong and solid 2650 player who wants to play for a win according to his style in a safe mode, without burning his boats ? I decided to choose the M eran Variation , which I have played many times with both colours.

Basically I think it is better to fianchetto the bishop in this type of position , but in the Sem i-Slav M i khai l l i kes to place his bishop on d2, as he has done a few times against Alexander Morozevich . For 9 b3 see the game Gelfand - Lautier (No. 39 , p. 1 72).

9

• . .

Ilea

Wh ite was hoping for 9 . . . e5 1 0 cxd5 cxd5 1 1 tDb5 Ab8 12 d xe5 tDxe5 1 3 tDxe5 Axe5 14 Ac3 ;!;; with a smal l but risk-free advan tage.

10 Ilad1

1 d4 d5 2 c4 c6 3 .t1c3 .t1f6 4 e3 e6 5 .t1f3 .t1bd7 6 1!!fc2 Ad6 7 Ad3 The most popular plan since the famous Kar pov - Kramn i k game i n Vienna 1 996. I had recently been successful with the Shabalov Shirov Attack 7 g4, including a win over Joel Lautier a few rounds earl ier. But this clearly requires a lot of homework in order to imple ment it.

7

. . .

0-0 a 0-0 h6

I had to face th is move four times against Joel Lautier i n 1 999, and I chose it as it leads to a complicated game and should n 't have been wel l known to my opponent, as it had never occurred in his games with either colour. 8 . . . e5 9 cxd5 cxd5 10 e4 !? leads to a sl ight advantage for White where Black can hope only for a draw. 8 . . . d x c4 9 Axc4 a6 is a good line, one chosen many times by both myself and my opponent.

10 . . . dxc4! N Using a standard idea at the right moment.

11 Axc4 e5 = Black is threatening . . . e5-e4 , while the bishop at d2 is misplaced .

1 2 dxe5 Wh ite releases the tension i n the centre. 12 Ad3 ! ? leads to an approximately equal game.

1 2 . . . .t1xe5 13 .t1xe5 Axe5 14 f4?! Safer was 14 �e2 �e4 15 Ac3 with an equal game.

200

14 . . . Ac7 15 c!>h1

Game 46

a

b

c

d

Gurevich - Gelfand , Corus Tournament, Wijk aan Zee 2002 ...

e

White's pieces. Safer was 16 . . . Ae6 17 Axe6 .l;xe6 1 8 e4 �g4 1 9 Ac1 'f!! h 4 20 g3 'f!! h 3 21 'f!!g 2 'f!! h S 22 .l;d1 with a complicated game.

8 7 6

6

5

5

4

4

3

3

17 Ab3 'ttd 7 17 . . . b4 18 �e2 (18 �a4 as) 18 . . . 'f!!d 6 1 9 �g3 ;1;; , or 1 7 . . . as 18 a3.

18 e4 b4?

2

a

b

c

d

e

9

h

White is preparing e4-eS . The next few moves are critical, as Black has to prevent this advance at any cost.

1 5 . . . Ag4?!

Now the knight has to move to a poor sq uare, but on the other pawn Black's pawn chain will lack mobil ity. Stronger was 18 . . . cS !? 19 eS c4 20 Axc4 (20 exf6 AfS 21 'f!!c1 cxb3 22 axb3 b4 oo) 20 . . . AfS 21 'f!!c1 bxc4 22 exf6 with an u nbalanced position , where both sides have their trumps. a

Playing into the Wh ite's hands. There is noth ing for the bishop to d o on g4. 1 S . . . �g4 ! (or 1 S . . . 'f!!e7 , preventing 1 6 e4 i n view of 16 . . . �xe4 17 �xe4 AfS !?) was stronger:

b

c

d

e -��-1

7

��""-=-l

6

A) 16 'f!!g 6 'f!!e7 (16 . . . 'f!!f6) 17 e4 ? (this loses quickly, but after 1 7 h3 �f6 White cannot push his pawn to eS , and so Black stands better)

5 4

8 7 6

1---=_ I-�,"",-",�

3

5 4 3 2

�------�--�

�

19 �a4? Surprisingly, my opponent made this move instantly. It is probable that he assessed in correctly the position after the 22n d move. The knight will remain on a4 until the end of the game. 17 . . . f8 18 'f!! h7 (18 'f!! h S g6) 18 . . . 'f!! h4 19 h3 'f!!g 3 20 'f!!h 8+ e7 ; B) 1 6 'f!! b 3 'f!!e7 1 7 h3 b S (17 . . . Ab6 !?) 18 �xbS c x bS 19 AdS .l; b8 20 Ab4 'f!!f6 21 hxg4 Axg4 22 .l;c1 with a defensible pos ition .

16 gde1 b5!? Black chooses a plan of counterplay on the q ueenside, gaining tempi by attacking

The correct move for the knight was 19 ttlb1 ! . Despite its strange appearance, from here the knight could have hoped to get back i nto the game via d2 and , what is most i m portant, all the tactics i nvolvi ng the e4 pawn would be removed . 1 9 . . . aS (1 9 . . . Ab6 20 Ax b4 .l;xe4? 21 .l;xe4 �xe4 22 'f!!x e4 .l;e8 23 'f!!g6) 20 Aa4 ! (I didn't see this move dur ing the game ; 20 Ac1 �xe4 2 1 .l;xe4 .l;xe4 22 'f!!x e4 AfS ; 20 a3 !?), and now :

201


A) 20 . . J ;ac8 21 Axc6 (21 e5 Af5 22 �xc6 ged 8 ; 21 Ac1 !) 21 . . . Axf4 22 Axd7 gxc2 23 Axe8 Axd2 24 �xd2 �xe8 25 �b3 a4 26 �d4 gx b2 27 gb1 ;

of the knight on a4 were stronger factors than the temporary lack of coordin ation of the black pieces.

B) 20 . . . ga6 21 Ac1 ! (21 gc1 ? Ae2 22 gfe1 Ad3 23 Axc6 �d8 24 �c5 Ab6 (24 . . . Ax b1 25 Axe8 �xe8 26 gx b1 �xd2 +) 25 �f5 gf8 26 Ab7 ga7 27 gc8 gx b7 28 gxd8 gxd8) 21 Ac1 ! gd8 (21 . . . �xe4 '!? 22 gxe4 gxe4 23 �xe4 Af5 24 �e2 Ad3 25 gd1 ) 22 �d2 and the kn ight comes back i nto play, giving White a serious advantage.

23 �b7 �e2 24 gg1 24 . . . �f2 + .

19 . . . Hxe4 20 Hxe4?!

23 .g6 (24

gc1

�e3)

23 . . . .te2! The point of Black's concept.

24 .xf7+ �h8 25 .f5 .txf1 26 .xc8+ .td8 27 .a8 27 �b7 Ab5.

27 . . . .tb5!

White fails to sense the danger and follows the wrong track. He could have mai ntained the balance by 20 �c5 ! gxe1 21 gxe1 �d6 22 �b7 ! �d4 (22 . . . �d7 23 �c5 =) 23 Ae3 00 (23 ge7 Axf4 24 �g6 Ah5).

20 . . . .tixe4 21 .xe4 21 Ax b4 ge8 =+= was the lesser evil .

2 1 . . .•xd2 22 .xc6

8

8

7

7

6

6

5

5

4

4

3

3

22 �e7 Ae2 .

2

22 . . . Hc8 + a

b

c

d

a

e

8

8

7

7

6

6

5

5

4

4

3

3

b

c

d

e

c

d

e

f

1f

Now things are clear. Black's queen and two bishops are attacking the king and White's position is hopeless.

28 h3 28 30 tal 33

2

a

b

�xa7 Ac6 29 �g1 Af6 ! (29 . . . �xf4 �c5 Ab6 31 �d3) 30 �f1 Ad4 (to domination !) 31 h3 g5 32 fxg5 Axg2+ �xg2 �e1 + 34 h2 Ae5+.

28. . .•c1 + 29 �h2 .xf4+ 30 �h1 ?

1f

The critical position of the game. I was told that the grandmasters in the press room thought Black was i n danger. However I felt that the two bishops and the poor position

Losing i mmediately, but 30 g3 �d6 31 �f3 Ac7 32 �f2 Ac6 was also hopeless.

3O . . .•c1 + White resigns Since if 31 h2 �c7+ 32 h1 Ac6.

*** 202

Game 47

Grischu k - Gelfand , Corus Tournament, Wijk aan Zee 2002 1. 9

Game 47

A l exa n d e r G r i s ch u k - B o r i s G e l fa n d Co rus To u r n a m e n t , Wij k aan Zee 2002 Pe tro ff Defence [C42] Despite his youth , my opponent in this game has already shown hi mself to be one of the strongest players in modern chess. I l i ke his natural and energetic style and he is proba bly the last player whose development has not been strongly i nfl uenced by com puter programs.

1 e4 e5 2 �f3 �f6 After many years of playing only the Najdorf Variation agai nst 1 e4, I added the Petroff Defence to my repertoire. I remember Victor Korchnoi saying something l i ke ' if a player wants to develop, he has to include new open ings i n his repertoire' . The Petroff has the reputation of being a dull open ing, where Black is fighting on ly for a draw. However, as my experience has shown , it can be no less exciting than the Sicilian.

cxd5 cxd5 1 0 ttlc3 �e8 (10 . . . ttlxc3 1 1 bxc3 £g4 is the main theoretical line - see Short Gelfand, Game No. 1 9, p. 94), and now : A) 1 1 £xe4 d x e4 1 2 ttlg5 £f5 1 3 f3 (this straightforward approach doesn 't pose any big problems; 13 �e1 ttlc6 14 d5 ttlb4 15 ttlg xe4 £xe4 16 ttlxe4 ttlxd5 17 £g5 £e7 1 8 tWh5 g6 1 9 £xe7 g x h5 20 £xd8 �axd8 with a sym bolic advantage for White, Fritz Anand , Frankfurt 2000) 1 3 . . . e3 14 ttlge4 £f4 1 5 ttle2 ?! (it was time to settle for a d raw by 1 5 �e1 £xe4 (15 . . . ttlc6 1 6 £x e3 £xe3+ 1 7 �xe3 tWx d4 =) 1 6 ttlxe4 £x h2+ 17 �xh2 �xe4 18 fxe4 tWh4+ = ; after 1 5 d5 ttld7 1 6 tWe2 tWb6 Black is fine) 1 5 . . . £h6 1 6 tWb3 £g6 !

r--.I--.--r- --,.-,-..-r------,..,...---. .t.

3 �xe5 d6 4 �f3 �xe4 5 d4 d5 6 £d3 £d6 Since the year 2000 I have also tried 6 . . . £e7 7 0-0 ttlc6 many times.

7 0-0 0-0 8 c4 c6 a

b

c

d

e 8

6

6

5

5

4

4

3

a

b

c

d

e

9

h

A �______________��__� v

9 1!rc2 During the next few months after this game I had to face other continuations :

��==�L-� 'lf

(the refutation of Wh ite's idea! - if 1 6 . . . £e6 1 7 tWx b7 £d5 18 tWb5 ttlc6 1 9 ttl4c3 ± , o r 1 6 . . . ttlc6 1 7 £xe3 £e6 1 8 tWc3 £xe3+ 19 tWxe3 £c4 =) 17 tWx b7 ttld7 (the c1 bishop, the e2 knight and neither of the rooks are taki ng part in the game, and this is for j ust for one pawn ; in addition the e3 pawn is ex tremely dangerous) 18 b3 (18 tWb3 ttlf6 with the i n itiative) 18 . . . ttlf6 19 ttlxf6+ ? (this ex change of the only active piece is the deci sive mistake; if 19 ttlc5 ttld5 =t or 19 . . . tWb6!?, but better was 19 ttl4c3 =t) 19 . . . tWxf6 20 £b2 £d3 21 �fe1 �ac8 ! -+ and I didn 't man age to spoil this won position (Morozevich Gelfand, NAO, Cannes 2002) ; 8) more problems can be posed by 1 1 �e1 ttlxc3 1 2 �xe8+ tWxe8 13 bxc3 £g4 (13 . . . ttlc6 allowed Black to equal ise after 14 tWb3 tWd7 15 g3 £f8 16 £a3 £xa3

203


17 �xa3 �d8 18 ge1 Ae6 19 �c1 h6 20 tDh4 �k8, Kasparov- Ehlvest, Moscow Grand Prix rapid 2002) 14 gb1 ! N (this strong novelty created serious problems for me) 14 . . . �e7 ? (14 . . . �d7 1S h3 Axf3 1 6 �xf3 ;;1;; ) 1 S h3 AhS 16 gbS �d7 (16 . . . Axf3 17 �xf3 �e1 + 18 Af1 �xc1 1 9 �xdS +-) 1 7 c4 d xc4 (17 . . . tDc6 18 gxdS Axf3 19 �xf3 tDb4 20 gbS tDxd3 21 �xd3 ge8 ±)

other hand, White has already exchanged on dS, which allows Black to gain counterplay on the c -fi le. 1 S . . . gc8 (1S . . . AxeS 16 d xeS tDcS 17 tDe2 ;;1;; leaves White with a smal l , but risk-free ad vantage), and now : A) 1 6 f3 'ff h 4 1 7 g3 tDxg3 1 8 h xg3 'ffx g3+ (18 . . . 'ffx d4+ 19 �g2 AxeS 20 Ae3 +-) 1 9 �f1 00 ; B) 1 6 Af4

Analysis diagram after 17 c4 dxc4

1 8 Ax h7+ ! �x h7 1 9 tDgS+ �g6 20 �c2+ fS 21 �xc4 tDc6 (21 . . . tDa6 22 gdS gfB 23 Af4) 22 g4 tDxd4 (22 . . . Axg4 23 h xg4 tDxd4 24 gdS) 23 g x hS+ (missing a beau tifu l win by 23 gdS, when Black is helpless against gd6) 23 . . . �xhS 24 gdS (24 �xd4? Ah2+) 24 . . . gc8 (now, fortunately, there is no longer a win) 2S 'ffx d4 gxc1 + 26 �g2 �xgS 27 gxd6 'ffe7 28 gd7 'ffe 4+ 29 'ffx e4 fxe4 30 gxg7+ �f6 31 gx b7 gc2 32 gxa7 e3 33 ga3 gxf2+ 34 �g3 �fS 3S gxe3 gxa2 1/2-1/2 (Topalov - Gelfand , Melody Am ber, Monaco rapid 2002) . 11. 9 [je1

AfS 1 0 'ff b 3 tDa6 1 1 cxdS cxdS 1 2 tDc3 Ae6 13 a3 tDc7 14 'ffc 2 (14 'ffx b7 g b8 1S �xa7 ga8 16 �b7 is a way to draw) 14 . . . fS 1S tDeS !? N . After this novelty the position strongly re sem b les the main l i ne, but with two d iffer ences. The black bishop is on e6, rather than c8, and as a conseq uence it prevents the knight from going to e6, an i m portant ma noeuvre that I used in the main game. On the

1 6 . . . gS I . Black has to take action to build u p counter play. If Wh ite has time to play f2-f3 and expel the kn ight from e4, he will clearly be better in view of the weakness of the eS square. Now White has : B 1 ) 1 7 Ad2 ?! tDa6 1 8 'ffd 1 (18 tDf3 g4 1 9 tDgS Af7) 1 8 . . . AxeS (18 . . . 'ff b 6 1 9 Axe4 fxe4 20 AxgS 'ffx b2 (20 . . . AxeS 21 d xe5 �xf2+ 22 �h1 ) 21 tDe2 AxeS 22 dxeS �xe5 23 tDd4 ! 55 with excellent counterplay due to the vulnerable position of the black ki ng) 1 9 d x eS tDacS 20 Ac2 tDxd2 21 'ffx d2 d4 with sl ightly the better game for Black, which I converted i nto a win (Fressinet - Gelfand, NAO, Cannes 2002) . Two wins (against Mo rozevich in the 7t h round and this one in the last rou nd) in the Petroff brought me victory i n this strong (category 18) NAO Masters ; B2) 1 7 Ae3 !? - the critical move, on which the assessment of the position depends. 9 �a6 10 a3 f5 Th is plan was rei ntrod uced i nto top-level chess by my Dutch friend Jeroen Piket against M ichael Adams one round earl ier

204

. . .

Game 47

Grischuk- Gelfand , Corus Tournament, Wijk aan Zee 2002

and against Alexander Morozevich the same day. Black tries to establish his knight on e4 i n the style of Harry Nelson Pillsbury. 10 . . J �e8 11 �c3 Ag4 was popular in 2001 , but there Black is fighting only for a draw.

1 1 �c3 �c7 12 1le1 I fail to understand the idea beh ind this move, as the f2 pawn is weakened and there is noth ing for the rook to do on the e-file. But my opponent thought it was a good idea. Practice has shown that 12 �e2 or 12 b4 gives White better chances of gaining an ad vantage.

1 5 cxd5 White accepts the challenge. Weaker was 15 tf:}e2 tf:}6g5 ! 16 tf:}e5 Axe5 17 dxe5 f4 with a strong attack, for example 18 �d4 (18 f3 �xf3+ 1 9 g xf3 �g5+ 20 �g3 fxg3 21 fxe4 �f2) 1 8 . . . f3 1 9 h4 �h3+ 20 g x h3 �x h4.

15 . . . cxd5 16 �xd5 Ilc8 A strong i ntermed iate move. Wh ite has to decide where to go with his queen .

17 'ftd1 a

b

c

d

..

e

8

8

7

12 . . . �h8 It is useful to remove the king from the a2-g8 diagonal .

13 b4 Ad7 14 Ab2

6

6

5

5

4

4

3

3

2

2

Here at least the queen won 't be subject to an attack. If 1 7 �b3 �6g5 , 1 7 �c3 �f4, or 17 �e2 Ac6.

17 �6g5! Th is strong move came as a su rprise to my you nger opponent. He was counting on 17 . . . Ac6 18 �e3 �f4 19 d 5 ! and Wh ite is better. After the text move all Black's pieces are taking part in the attack. . • •

14 . . . �e6! It was also good enough to follow a game from the books, but I think that my m ove is stronger. This novelty secured second place in the competition for the best novelty of In formator 83. And, incidentally, the game itself tied equal 3rd in the competition for the best game in the same Informator. After 14 . . . Ae8 15 c5 (15 �e5 �e6) 1 5 . . . Ae7 1 6 �e5 Ah4 17 �f1 Ag5 18 tf:}e2 Ah5 19 tf:}g3 Ag6 20 �fe1 Af4 21 Ac1 Axe5 22 dxe5 �h4 23 Ab2 tf:}e6 24 �ad1 f4 Black was better i n the game G ufeld - Kochiev, (U SSR Team Cup, Rostov 1 980).

205

18 �e5 a 8

b

c

d

..

e 8

7 6

6

5

5

4

4

3

3

2

2


I thought that White was playi ng with fi re, but this is probably his best option . 1 8 tLle3 tLlxf3+ (18 . . . Aa4 1 9 �e2 tLlc3 20 Axc3 tLlxf3+ 21 �xf3 �xc3 22 �e2 ±) 19 �xf3 was the alternative: A) 1 9 . . . tLlgS (this is what I had i n m i nd) 20 �hS f4 (20 . . . Ae8 21 �d1 f4 22 tLlfS Ag6 23 tLlxd6 �xd6 24 dS, and if 24 . . . f3 2S h4 fxg2 26 hxgS �f4 27 f3 �h4 28 Axg6 hxg6 29 AeS) 21 tLlc4 (21 tLlg4 AfS) and it is not easy to continue the attack; B) 1 9 . . . �h4 ! (this is stronger) 20 g3 �h6 (now the queen has joined Black's forces and he has numerous threats) 21 �g2 (otherwise . . . tLlgS and . . . fS-f4 would fol low; if 21 Axe4 fxe4 22 �g2 �f3 +) 21 . . . tLlxf2 22 �xf2 f4 23 gxf4 �xf4 24 �g2 �cf8 and I don 't see a defence.

18 . . . �h3+ !! The kn ight is sacrificed not even for a pawn , but White's king comes under a strong at tack. I should poi nt out that this was q u ite a lengthy journey by the kn ight (b8-a6-c7e6-gS-h3) . After 18 . . . Ac6 1 9 tLlf4 B lack's play comes to a dead end.

19 gxh3 Og5+ 20 Ci!?f1 20 tLlg4 tLlxf2 ! loses on the spot. a

b

8

c

d

8

7 6

6

5

5

4

4

A) 21 �e2 AxeS 22 d x eS Ac6 23 Ac4 (23 Axe4 fxe4) 23 . . . �x h3+ (the reason why I rejected 20 . . . �h4 was that after 23 . . . tLld2+ 24 �xd2 �xc4+ 2S �g1 Ax dS 26 �ac1 Wh ite seizes the initiative, but Black can play more strongly on his 23rd move) 24 �g1 tLld2 ! (on ly now !) 2S �xd2 �g4+ 26 �h1 (26 �f1 �xc4+ 27 �e2 AxdS 28 �c1 �bS 29 �xc8 �xc8 30 �g1 +) 26 . . . �xc4 + and in view of the vulnerabil ity of the wh ite king, Black's chances must be preferred ; B) 21 �e2 �x h3+, and now : B 1 ) Alexander i ntended to play on a piece up, but I believe that after 22 �e1 Axe5 23 d xeS �x h2 24 tLle3 �h1 + 2S tLlf1 tLlg5 26 �c2 �cd 8 ! (26 . . . tLlf3+ 27 �e2 �xc2+ 28 Axc2) Wh ite can not save the game. Clearly, it is hard to prove this 100% with variations, but this was my feel ing about the position both during the game and after spend ing a n u m ber of hours analysing it later. If 27 �e2 tLlf3+ 28 �d1 AbS ;

2 1 dxe5 Ac6!?

3 2

20 . . . �h4 was my origi nal intention :

B2) 22 �g1 (I think Wh ite has to be satis fied with a d raw) 22 . . . AxeS 23 d x eS �g4+ (23 . . . tLlgS 24 �e3) 24 �f1 �h3+ (24 . . . Aa4 2S �xa4 tLlcS 26 bxcS (26 �d1 �h3+ 27 �g1 =) 26 . . . �xa4 27 e6 with the initiative) 2S �g1 = .

...

e

told that in the press room 20 . . . Aa4 21 �xa4 �h4 22 �e2 was analysed , but I cou ldn't be lieve that it would work.

2

20 . . . Axe5 B lack has a wide choice of possibilities and u nfortunately Wh ite can hold in all the lines. 20 . . . Ac6 21 tLlxc6 �h4 22 �e2 �x h3+ 23 �e1 ± doesn 't look too promisi ng. I was

Black could have won the queen , but this is what Wh ite was hoping for: 21 . . . tLld2+ 22 �e2 Ae8 23 tLle3 AhS+ 24 �xd2 Axd1 2S �axd1 and Black's attack is over, whereas Wh ite launches a cou nterattack after 2S . . . �fd8 (2S . . . f4 26 tLlc4 f3+ 27 �e3 �h4 28 �e1 �xc4 29 Axc4 �xc4 30 e6) 26 �g1 �h6 27 �e1 . 21 . . . Ae6 !? was another strong continuation, which would have g iven Black a powerfu l attack:

206

Game 47

Grischu k - Gelfand , Corus Tournament, Wijk aan Zee 2002

A) 22 Ac1 ? 'tWh4 23 Axe4 fxe4 24 �f4 E!xc1 25 'tWxc1 E!xf4 and wins; B) 22 E!xe4 fxe4 23 Ax e4 Axh3+ (23 . . . gcd8 !?) 24 �e2 'tWh4 25 f3 Af5 ! (25 . . . Ae6 !?) 26 Axf5 E!xf5 27 E!c1 'tWx h2+ 28 �d3 E!cf8 (28 . . . E!d8 29 �e4) 29 Ad4 E!xf3+ with an unbalanced position where I prefer Black; C) 22 f4 'tWh4 23 Axe4 fxe4 24 �g1 E!cd8 (24 . . . Axd5 25 'tWg4 (25 'tWxd5 E!c2) 25 . . . 'tWxg4+ (25 . . . 'tWh6 26 Ad4 E!c6 27 f5 E!xf5 28 'tWxf5 E!g6+ 29 �f1 ) 26 h xg4 E!xf4 27 h3 =) 25 Ac1 E!xd5 26 'tWe2 'tWx h3 + and Black has regained his piece whereas Wh ite's king is open ;

22 .tc1 ! Probably the only way to save the game. Worse was 22 f4 'tWh6 23 Axe4 fxe4 24 �g1 gcd8 25 'tWg4 E!xd5 26 Ac1 gd3 +, or 22 ttle3 E!cd8 23 'tWc2 E!xd3 24 'tWxd3 'tWf4 25 tDd1 tDd2+ 26 �g1 tDf3+ 27 �f1 'tWx h2 28 �e2 'tWx h3 with an attack,

22

a

The proposal of my good friend Zohara Olpi ner 21 . . . E!cd8 can be refuted : A) 22 E!e3 Ae6 (22 . . . f4 23 E!f3 Aa4 24 'tWe2 +-, or 22 . . . Ac6 23 Axe4 (23 f4 'tWh6) 23 . . . fxe4 24 'tWg4) or 23 tDc3 'tWh4 24 'tWe2 ttlg5 25 f4 'tWxf4+ 26 �g1 00 ;

..

c 8

7 6

6

5

5

4

4

3 2

24

• • .

2

•xh3+

N ot 24 . . . E!cd 8 ? 25 'tWg4 . An i nteresting try would be 24 . . . E!f3 25 b5 ! Axb5+ 26 �g1 gxh3 (26 . . . E!d8 27 e6 E!xh3 28 e7 ge8 29 Af4 E!d3 30 Ag3 'tWg5 31 tDc7) 27 Af4 E!d3 28 Ag3 'tWd8 29 'tWg4 E!xd5 30 E!xe4 with chances for both sides.

25 �g1 .txd5 Black has to force a draw. If 25 . . . E!cd8 26 tDf4 'tWh4 27 tDg2 .

26 .xd5 .g4+ 26 . . . E!c6 ? 27 Af4 +- (27 'tWxe4 E!g6+ 28 �h1 E!f5 29 Af4 E!xf4 30 E!ad1 ±) .

27 �1 .h3+

B) 22 f4 'tWh4 ; C) 22 Ac1 ! 'tWh4 (22 . . . f4 23 'tWf3) 23 'tWf3 Ac6 24 Axe4 fxe4 25 'tWg3 (25 gxe4 ±) 25 . . . 'tWxg3 26 hxg3 gxd5 ± .

b

8

01 ) 24 . . . E!fd8?! 25 �f1 'tWg1 26 E!c1 ! (26 h4 E!d7) 26 . . . E!xc1 27 Axc1 �c3 (27 . . . Ac4 28 E!c2) 28 'tWc2 �xe2 29 Axe2 and I prefer Wh ite here;

03) 24 . . . 'tWh1 + 25 ttlf1 ttlg5 26 E!c2 E!cd8 (26 . . . If)f3+ 27 �e2 E!cd8 28 ttld2 +-) 27 f4 ttlf3+ (27 . . . 'tWx h3!? 28 fxg5 E!xd3 29 'tWe2 E!fd8 with ful l compensation for the piece) 28 �f2 Ad5 29 �g3 ! tDe1 ! 30 'tWxe1 'tWf3+ 31 �h2 'tWh1 + with perpetual check.

• h4 23 .txe4 fxe4 24 .te3

24 E!a2 e3 ! (24 . . . 'tWx h3+ 25 �g1 E!cd8 26 E!d2 =) 25 E!xe3 'tWc4+ + , or 24 tDf4 E!xf4 25 Axf4 'tWx h3+ 26 �g1 e3 -+.

0) 22 �e3 'tWf4 ! (22 . . . 'tWh4 ? 23 E!e2 'tWx h3+ 24 �g1 �g5 25 f4 and wins) 23 E!e2 'tWx h2 24 �e1 , and now :

02) 24 . . . E!cd8 25 ttlf1 'tWg1 (25 . . . 'tWh1 26 E!e3) 26 h4 (preventi ng . . . �g5) 26 . . . E!d7 (26 . . . g5 27 'tWc2 E!c8 28 'tWb1 g x h4 29 Axe4 fxe4 30 'tWxe4 h3 31 f3 (0) 27 'tWc2 E!fd8 28 E!d1 and it is hard to assess this position ;

. . •

27 . . . 'tWf3 28 �g1 (28 e6 gc2 29 �g1 ).

207

28 �g1 .g4+ 29 �1 Draw agreed

***


White is threatening 9 tDe5.

B o r i s G e l fa n d - L a z a ro B r u z o n O l y m p i ad , B l ed 2002 Ca talan Opening [E04]

B . . . Ad6

Every player has his own way of establishing an open ing repertoire and prepari ng for an event. Some prepare concretely for tourna ments and l i ke to switch from one variation to another without paying m uch attention to subtleties. I, however, bel ieve that you should study open ings deeply and not be in a h urry to reap the benefits. As you may have noticed , on a n u m ber of occasions I have used novelties that were invented many years earlier. Th is game broke all the records, as it contains a novelty that I analysed back in 1 985 . I had to wait half of my l ife to use it ! My opponent was then four years old and I doubt whether he yet knew the moves. Lazaro Bruzon was the world under-20 champion in 2000, and together with Le n ier Dominguez he is probably the biggest hope of Cuba since the time of Jose- Rau l Capablanca.

1 d4 dS 2 c4 e6 3 �f3 �f6 4 g3 d x c4 S Ag2 a6 Black tries to hold on to the pawn.

6 0-0 �c6 7 e3 Ad7 After 7 . . J �b8 8 tDfd2 8 . . . �d7 9 tDxc4 b5 10 tDcd2 White regains the pawn with a sl ight advantage (Gelfand -V. M i lov, Bie1 2000) .

Thus 8 . . . b5 ? is sim ply a blu nder: 9 ttJe5 ttJxe5 10 dxe5 ttJd5 11 ttJxd5 exd5 1 2 �xd5 ± (Gelfand -Van Wely, Melody Amber, Monaco rapid 2001 ). However, interesting is 8 . . . Ab4 9 ttJe5 ttJxe5 10 dxe5 Axc3 11 bxc3 ttJd5 12 Aa3 (12 �g4 0-0 13 e4 f5 14 exf6 tDxf6 15 �e2 �e7 00) 12 . . . �g5 (12 . . . Ac6 13 �g4) 13 �d4!? (White should keep the i n itiative, not regain mate rial ; 13 Axd5?! exd5 14 �xd5 0-0-0, or 13 f4 �g6 00 ) 13 . . . 0-0-0 14 �a7 (14 f4 !?) 14 . . . Ac6 15 .!3.fd1 �xe5 16 .!3.d4 with full compensation for the two pawns, but not more (Gelfand Adams, Engh ien-Ies- Bains 2003).

9 .e2 bS 10 e4 Ae7 Dangerous is 10 . . . e5 11 dxe5 ttJxe5 1 2 ttJxe5 AxeS 1 3 f4 Axc3 (13 . . . Ad4+ 14 Ae3 Ac6 1 5 e5 Axg2 1 6 �xg2 ±) 14 bxc3 c6 15 f5 and White has a strong i n itiative (Tkachiev Bel iavsky, Enghien-Ies-Bains 1 999).

1 1 dS �b4 In an unpubl ished game from my youth Black tried 11 . . . exd5 12 e5 ttJe4 13 ttJxd5, but after 1 3 . . . tDc5 14 tDf6+ ! his king got stuck i n the centre and his .!3.h8 was out of play: 14 . . . �f8 1 5 tDxd7+ �x d7 1 6 .!3.d1 tDd3 1 7 tDe1 .!3.d8 18 tDxd3 cxd3 1 9 �e4 tDd4 20 Ad2 ttJe6 21 .!3.ac1 h5 22 Aa5 ± (Gelfand - I . Botvinnik, Belarussian Championsh ip 1 985) .

B �c3 a

12 �eS exdS b

c

d

...

e

a

8

8 7

.1

b

c

d

e 8

7 6

6

5

5

5

5

4

4

4

4

3

3

3

3

2

2

a

208

b

c

d

e

'If

Game 48

Gelfand - Bruzon, Olympiad , Bled 2002

13 a3!? N 8

I found this interesting idea with my coach Al bert Kapengut immed iately after the game with l Iya Botvi nnik. A few years ago I tried this l i ne with Black against Anatoly Karpov and obtai ned an equal position after the harmless 13 exd5 0-0 14 a3 �d3 15 �xd3 Ag4 16 YWe3 cxd3 17 YWxd3 YWd7 (Karpov Gelfand, Dos Hermanas 1 999) .

8

7 6 5 4

1 3 . . . �d3 14 ttlxd3 cxd3 1 5 trxd3 The point of White's idea is that he wants to take on d5 with a piece after e4-e5.

15

• . .

dxe4

15 . . . c6 1 6 exd5.

1 6 ttlxe4 ttlxe4 16 . . . Ae6 1 7 �xf6+ Axf6 1 8 Ac6+ ! �f8 19 YWc2 with fine compensation for the pawn .

17 trxe4 17 Axe4 Ah3 1 8 YWf3 Axf1 1 9 Ac6+ �f8 20 Axa8 Ac4 and I am u nable to find any thing concrete here, but it is obvious that White has fu ll com pensation for the pawn.

19

. • .

£d6?

Bei ng already short of time (thanks to F I D E w e were forced t o play with the sense less new time control), Bruzon panicked . If 1 9 . . . Af6?! 20 YWc2 , but the cool-headed 19 . . . ge8 ! 20 YWc2 Ad6 21 Axd6 cxd6 22 gxd6 gb6 23 gd4 ;t would have left Wh ite with only a slight edge. Another sugges tion of Lukacs and Hazai may also be good : 19 . . . gb6, and if 20 Axc7 YWxc7 21 YWxe7 ge6 22 YWxd7 ? ge1 +.

20 £xd6 cxd6 21 gxd6 gb6? This is simply a blunder (21 . . . YWc8 22 YWf4 i) .

22 gd2

17 . . . lib8

a

b

c

8

Black could bail out by 1 7 . . . 0-0 1 8 YWxa8 YWxa8 19 Axa8 gxa8 ;t with good d rawing chances. White would like to exchange rooks or bishops, but I don't see how he can do this. A good idea may be 20 gd1 Ae6 21 Af4 Af6 22 gd2 c5 23 ge1 h5 24 Ae5, as rec ommended by Lukacs and Hazai i n New in Chess Yearbook 68.

d

•

e 8

7 6 5

5

4

4

3

3

2

2

18 £f4 White is a pawn down , but the pressu re of his bishops is strong . 18 ge1 g b6 ! 19 Af4 was tempting, but after 1 9 . . . ge6 20 YWb7 0-0 21 Axc7 (21 gxe6 Axe6 22 YWxa6 Ac4) 21 . . . YWc8 Black holds on .

18

. . .

Black cannot avoid loss of material due to the pin on the d -fi le.

22 . . . ge8 22 . . . YWc8 23 YWd4.

0-0 19 gfd1

23 trd5 ge7 24 Ah3 Black resigns

209


a

B o r i s G e l fa n d - Peter A e s P o l a n d S u perc u p , 2 0 0 3 Sicilian Defence [86 7]

b

c

d

e 8

7 6

This was the first time I played for the Polo nia Warsaw club. The event was the Poland Supercup between the two strongest teams in the national league. Our opponents repre sented the city of Grodzisk Mazowiecki , the birthplace of the legendary M iguel Najdorf, a person adored by almost all chess players in the world for his pure love of the game. Obvi ously I wanted to play my best in this city. My opponent, the 2001 world j u n ior champion, is a pupil of the famous Hungarian theoreti cians Laszlo Hazai and Peter Lu kacs. It is no wonder, therefore, that he is excellently prepared in the openings and last year he scored a number of memorable wins against some of the world 's best players. I n the first round of the event he was extremely close to adding my name to the l ist of players he had beaten , so I arrived for this game fu l ly concentrated .

1 e4 Th is move could have been a l ittle surprise for my opponent, as 1 d4 has been my main weapon for more than a decade. However, I had been working hard to widen my reper toire and to play 1 e4 as wel l .

1 . . . e 5 2 �f3 d6 3 d 4 exd4 4 �xd4 �f6 5 �e3 �e6 6 Ag5 e6 7 .d2 a6

5

5

4

4

3

3 2

�------��-

�

1 e4 was my main weapon (about 20 years ago) that this plan contains more venom than it appears.

9 . . . Axe6 9 . . . bxc6 is bad due to 1 0 e5 ! d xe5 1 1 Axf6 g xf6 1 2 �e4 Ae7 13 �h6 f5 14 � d6+ Axd6 15 gxd6, when Black is on the verge of los ing, as his king is stuck in the centre and he has numerous weaknesses.

10 .e1 .a5 After 1 0 . . . Ae7 1 1 e5 � d5 1 2 Axe7 �xe7 White has a pleasant choice between 13 �e4 d x e5 14 c4 �f6 1 5 �d6+ �f8 1 6 �xe5, and 13 � xd5 Axd5 14 c4 gc8 (14 . . . Axc4!?) 15 �b1 Axc4 1 6 gc1 d5 1 7 b3 b5 1 8 bxc4 bxc4 1 9 �a1 , when the two pawns do not fully com pensate for the piece (Short Mednis, Brighton 1 983) .

1 1 f4 Ae7 a

The Richter- Rauzer Variation was extremely popular in the 1 990s, but now the N aj dorf and the Sveshnikov are Black's main weapons in the Sicilian.

b

c

d

e

--"-1'-"""

S 0-0-0 J.d7

6 5

(see next diagram)

4

4

3

9 �xe6 This is a real sidel ine of theory. 9 f3 or 9 f4 is played in nearly all of the top-level games. However, I remembered from the time when

210

�------�--�

�

Game 49

Gelfand -Acs, Poland Supercup, 2003

1 2 .th4!? This idea of Gennady Kuzm in brought a lot of success to him and his pupil Maya Chi bur dan idze (then the women 's world champion) in the early 1 980s. Earl ier that year (2003) I checked this line again and I d iscovered that, for no apparent reason , since then no body had tried it. So, I decided to do so myself! 1 2 Ad3 h6 13 Ah4 g5 gives Black good counterplay.

12 . . . Hc8 It was obvious that Peter was not fam i l iar with White's plan , as he spent a lot of time in the open ing stage. 12 . . . '�c7 1 3 Ad3 0-0-0 14 .E!f1 .E!d7 15 �b1 �b8 16 h3 e5 17 Ag3 exf4 18 Axf4 ± gave White the advantage in Kuzmin - Liptay, (European Cup 1 983) . 1 2 . . . b5 1 3 e5 d xe5 14 fxe5 and if 14 . . . coe4 ? 1 5 Axe7 coxc3 1 6 Ad6 ! +-. Bad is 12 . . . coxe4? 13 Axe7 coxc3 14 Axd6 coxa2+ 15 �b1 't'fxe1 16 .E!xe1 +-, or 12 . . . O-O? 13 e5 .

1 3 .td3 White's idea is that the only natural develop ment of Black's i n itiative i nvolves . . . b7-b5. However, with his queen on a5, this is un favourable on both the 1 2 t h and 13t h moves due to e4-e5.

13

. . .

14 . . . O-O !? 1 5 .E!f1 b 5 1 6 e 5 d x e5 1 7 fxe5 ttld5 1 8 ttlxd5 Axd5 1 9 Af6 allows White a strong attack (if 19 . . . Ac4 20 Ax h7+) , but 15 . . . .E!fe8 ! ? , as suggested by Lev Psakh is, is interesting. Th is move looks i l logical , as Black moves his rook off the f-fi le, which is l i kely to become open . H owever, here, as i n many other l ines of the Sicil ian , this is the best place for the rook, and now 16 e5 (16 't'fe2 !?) is not so dangerous, in view of 1 6 . . . d xe5 1 7 fxe5 ttld5 1 8 ttlxd5 Axd5 19 Af6 Ac4 20 Ax h7+? �xh7 21 't'fh4+ �g8 22 't'fg5 Af8 and White can resig n .

1 5 Hf1 !? The most am bitious move. 15 Axe7 't'fxe7 1 6 .E!f1 cof6 ;t; would leave the i n itiative with White, but not more.

15 . . . �xf4! Black is looki ng for cou nter-chances.

16 Hxf4 Harm less is 16 Axe7 ttlxd3 1 7 't'fg3 't'fxe7 1 8 't'fx g7 't'ff8 (18 . . . COx b2 1 9 't'fx h8+ �d7 20 .E!xd6+ �xd6 21 't'fxc8) 19 't'fxf8+ .E!xf8 =.

1 6 . . . g5 a

b

c

d

e

•c7

Dangerous is 13 . . . b5 14 e5 d x e5 15 fxe5 cod5 16 Axe7 coxc3 17 Ad6.

14 �b1 14 .E!f1 !? was perhaps more subtle, when 14 . . . coh5 would have much less point.

14 . . . �h5!?

9

Now 14 . . . b5 is met by 15 Axf6 Axf6 (15 . . . gxf6 leads to a theoretical position that favou rs White) 1 6 ttld5 Axd5 1 7 exd5 e5 18 g4 with an undisputed advantage for White. A simi lar idea one move earl ier was premature, as after 14 Axf6 Axf6 15 ttld5 Axd5 16 exd5 Black can gain strong cou n terplay by 1 6 . . . '�b6 ! '

h

11

17 .txg5 It was very tem pting to play flamboyantly and sacrifice some more material with 17 cod5 !?, but I was u nable to find anything concrete after either 17 . . . Axd5 18 exd5 gxf4 19 dxe6 O-O ! 20 't'fe4 f5 21 't'fxf4 d5 00 , or 1 7 . . . exd5 18 exd5 g xf4 19 dxc6 bxc6 20 .E!d2 !? (20 Af5

211


� b8 21 �c3 0-0 22 Af6 �fe8 ! 00 ) 20 . . . �f8 21 �e2 , although White still has the initiative.

21 . . . d5 22 �e2 would leave Black helpless.

22 ttie4 Af4

17 . . . Axg5 18 Ilf3 If B lack were able to transfer his bishop to e5, he wou l d be out of trouble, but I hoped that the pressure on the f-file would prevent him from doing this.

18

. . .

Black could have put up more of a resistance by 22 . . . �g8 23 �f6 (23 �d6 gc7 24 Ax h7 gg7) 23 . . . Axf6 24 gxf6 �g7 , when despite White's advantage the game stil l goes on.

23 g3 Ah6 24 ttif6

•e7?

Too slow. Active play was cal led for: 1 8 . . . �a5 ! 1 9 �f2 �f8 (or i m med iately 19 . . . �c5 20 �xf7 �xf2 21 gxf2 �e7 22 gdf1 gcf8 ;t) 20 gf1 �c5 ! (weaker is 20 . . . gc7 21 Ac4 (21 �a7 b5, or 21 �d5 exd5 22 gf5 �d2 23 exd5 Ad7) 21 . . . �e5 22 Ad5 ! (22 �b6 Ad8) a n d Wh ite mai ntains his at tack) 21 gxf7 �xf2 22 g7xf2 ;t gxf2 23 gxf2 �e7 . At the cost of a pawn Black has re pelled the opponent's attack and , thanks to his bishop pair, he has good chances of sav ing the game.

24 �d6 gc7 .

24 . . . Ilc7 The o n ly defence. If 24 . . . Ac6 25 �d7+ +-, or 24 . . . �c5 25 �xc5+ gxc5 26 �xe8 �xe8 27 gxf7 +-. a

c

d

e

8

8 7

6

19 .f2 Threaten ing ge1 followed by �d5.

1 9 . . . @f8 20 Ilf1 Ae8

8

b

5

5

4

4

3

3 2

a

8

b

c

f

9

h

'if

7

25 .b6!

6 5

5

4

4

3

3

After quite a long think I found a precise way to win . 25 �x h7+ gxh7 26 Ax h7 f5 was not so clear.

25 . . . tllg7

2

a

b

c

'It

Black is hoping to bring hi s roo k i nto the game by . . . gc5-e5, but . . .

21 e5! ± A typical Sicil ian idea. White vacates the e4 square for his knight.

21 . . . dxe5

After 25 . . . �c5 26 �xc5+ �xc5 27 �xe8 �xe8 28 �xf7 or 25 . . . �c6 26 �a5 gc5 27 �d7+ ! White wins. 25 . . . Ac6 was more stu b born : 26 �3f2 ! (26 �d5 ? exd5 27 �xf7+ �xf7 28 �xc7 would allow my opponent to escape after 28 . . . Af4 ! 29 ffb6 �f6 30 gxf4 e4 with coun terplay) 26 . . . e4 (26 . . . Ag7 27 �x h7+ gxh7 28 Axh7 f5 29 g4 +-)

212

Game 50

Shabalov - Gelfand , Bermuda, 2004

Black is helpless. Despite the relative mate rial balance, his pieces are extremely bad ly placed , his king may come under attack and his pawns are weak.

29 . . . l1ee7 30 £e4 £g5 31 h4 11f1 +

and o nly now 27 &DdS exdS 28 gxf7+ 'tlVxf7 29 'tlVxc7 'tlVxf1+ 30 Axf1 +-. 26 ctsxe8+ I1xe8 27 I1xf7+ "xf7 28 I1xf7+ 11xf7 29 a4 +a 8

b

c

d

e

9

Peter doesn 't want to suffer a slow death and he loses in a few moves. After 31 . . . Af6 32 as Black can only sit and wait, while White chooses between advancing his queenside pawns, or playin g his queen to hS fol l owed by the advance of his pawn to gS. However, this wou ld require some precision .

h

32 �a2 £16 33 £xb7 a5 34 "xa5!

8

I---�

The last finesse.

7

34 . . . l1xb7

35 "a6

It is impossible to save both rooks. 3

35 . . . l1xb2+ 36 �xb2 e4+ 37 c3 11f3 38 "xe6 11xg3 39 "c6 I1g2+ 40 �b3 h5 41 a5 I1g1 42 "xe4 Black resigns

2

a

b

c

9

h

*** Game 50

A l e x a n d e r S h a b a l ov - B o r i s G e lfa n d Berm u d a , 2 0 0 4 Sicilian Defence [896J For already the 1 2 t h year i n a row the island of Berm uda hosted a chess festival in Jan uary. This event is made possible by two people - the sponsor N ick Faul ks and the organ iser N igel Freeman . Both of them are 100% devoted to chess, and it is no won der that everyone who has played i n their tou rnaments wants to return the next year! Nick and N igel managed to stage one of the highest-rated tournaments i n the history of the Americas ! I really enjoyed playi ng there and I played a number of entertaining games.

1 e4 c5 2 ctsf3 d6 3 d4 cxd4 4 ctsxd4 cts16 5 ctsc3 a6 6 £g5 On my first evening in Bermuda I had a d is cussion over d i nner with N ick Faul ks about fashion in chess. He asked , among other things, why everyone is going for the 6 Ae3 line against the Najdorf, whereas 6 AgS has been as though abandoned for no reason . I was amazed by his intu ition . The next morn ing Vladimir Kramn i k played 6 AgS agai nst Peter Svidler i n Wij k aan Zee and two more games were played i n the Corus B and C tournaments. H owever, you cannot accuse Alex Shabalov of being a copycat player. Th is line was the main weapon of Latvian players (led by M i khail Tal and Alvis Vitolins) back i n

213


the 1 970s and I remem ber that Alex played it against me in a j u n ior tournament in 1 979 and achieved a winning position , but missed the win.

6 . . . e6 7 f4 �bd7 7 . . . �b6 is mostly played , which I see too as being a q uestion of fashion . No one has come even close to refuting 7 . . . tDbd7 as wel l as other li nes.

8 .f3 .c7 9 0-0-0 b5 10 Ad3 During the last few years my opponents have tried to gain an advantage after 10 Ax b5 or 10 e5 .

10 . . . Ab7 1 1 ghe1 .b6 I experienced d ifficu lties after 11 . . . 0-0-0 12 a3 Ae7 13 �e2 ! h6 (13 . . . tDc5 14 Ax b5 !? with the i n itiative) 14 Ah4 �c5 1 5 Af2 (15 Ax b5 ? tDfxe4 1 6 tDxe4 Ax h4 -+) 15 . . . d5 (15 . . . �b8 !? ;I;; ) 16 e5 tDfe4 ?! 17 Axe4 dxe4 in the first game of my candidates match with Michael Adams (Wij k aan Zee 1994) and now strong was 18 �e3 ! f5 (18 . . . �a4 1 9 �xe4) 19 exf6 g xf6 20 �b1 ± . a

b

c

d

e

�-......, ,-,,-J-i"'''''---I

6 5 4

12

. . .

•xd4

1 2 . . . exd5 is bad because of the intermed i ate 1 3 tDc6 ! Axc6 14 exd5+ Ae7 1 5 dxc6 �c5 1 6 Axf6 g xf6 1 7 Af5 (Chiburdan idze Dvoirys, USSR Championshi p Sem i-Final, Tal linn 1 980) .

13 Axf6 gxf6 14 Axb5 .c5 14 . . . '�x d 1 + 1 5 �xd1 ax b5 1 6 tDc7+ is also possible, but I think it is in Wh ite's favour to eliminate one pair of rooks from the board .

1 5 b4 1 5 �xf6+ followed by the capture on d7 is more chal lenging and I think it is worth test ing in practice.

15

. . .

•xb5 16 �c7+ �e7

1 6 . . . �d8 1 7 �x b5 ax b5 1 8 �h5, aiming at the weakness on f7 .

17 �xb5 axb5 18 .h5!? N Alex plays i n ful l accordance with the ideas of the Riga chess school - activity for the pieces first and foremost ! He had already spent most of his allotted time and now it was my turn to try and d iscover the sub tleties of this position. Theory considered it to be i n Black's favour on the basis of 18 �d3 ga6 19 �x b5 gb6 + (Mnatsakanian Kr. Georgiev, Yerevan 1 982) .

18 . . . gxa2

3 2

�------�--� �

12 �d5 After some thought my opponent goes for ward ! After 1 2 tDxe6? fxe6 1 3 �h3 e5 ! 14 �d5 Axd5 15 exd5 0-0-0 Black won in the famous game Geller- Polugayevsky, (Portoroz Interzonal 1 973). 1 2 tDb3 is a sol id move, which I faced i n a game with Jaan Ehlvest (Linares 1 991 ).

As soon as I found my 1 9t h move, it be came clear to me that this was the right way ! If 18 . . . e5 ?! 19 �b1 , while after 18 . . . Ac6 1 9 gd3 !? the initiative passes to Wh ite. I also rejected 1 8 . . . Ag7 1 9 gxd6 ! �xd6 20 �xf7, when than ks to the threat of gd1 + Wh ite re gains the piece and maintains his attack.

19 .xb5 I saw that 1 9 �b1 ga6 20 e5 d5 was harm less for Black (20 . . . fxe5 21 fxe5 �xe5! 22 gxe5 Ag7 is even stronger).

214

Game 50


20 �x b7 is less accurate than 20 � b1 , as it g ives Black the add itional interesting op tion 20 . . . Axf4+ 21 � b1 .E!aa8 !? (21 . . . .E!ha8) 22 �b5 Ae5 .

20

19 .th6!! The idea beh ind this move is clear. For the modest price of just one piece, Black bri ngs two pieces into the game and begins an at tack on the wh ite king. It was not so easy to venture this, as Black will be left with just two pieces for the queen and nothing can be proved with concrete variations, but intu ition told me that Black's com pensation should be sufficient for at least a draw. I liked much less the alternative 1 9 . . J �a1 + 20 �d2 (or 20 �b2 .E!xd1 21 .E!xd1 Axe4 22 .E!a1 Ag7 23 .E!a7 f5+ 24 �b3 .E!d8 (0) 20 . . . .E!xd 1+ 21 �xd1 Aa8 00 , while i f 1 9 . . . .E!a7 ? 20 .E!xd6 ! �xd6 21 .E!d1 + Ad5 22 exd5 ± . 20 e5?! Both sides are obviously paying more at tention to the initiative than to material fac tors. White wants to prevent the opponen t's bishop or knight from reaching the e5 square. 20 g3 allows Black a choice between forcing a draw by 20 . . . .E!c8 21 �x b7 .E!cxc2+, or a good game after 20 . . . .E!b8!? 21 .E!xd6 .E!xc2+ 22 �d1 (22 �xc2 Axe4+) 22 . . . .E!c7 + . The most critical line was 20 � b1 !? .E! ha8 21 �x b7 Axf4 22 c3. Here Black has a num ber of possibilities. I was i ntending to play 22 . . . Ae5 (22 . . . .E!2a7 23 �b5 �e5 fol lowed by perpetual check on the a-file is the com puter's suggestion ; 22 . . . .E!8a3 !? is i nterest ing) 23 �c6 .E!8a3 24 �c4 .E!xg2 25 .E!e2 .E!xc3 with full com pensation for the queen . • . .

• • .

fxe5?!

If 20 . . . Axf4+ 21 � b1 .E!ha8 22 exd6+ �d8 23 �x b7 .E!a1 + 24 �b2 .E!1 a7 25 �xa8+ .E!xa8 26 c4 ± . But both players missed the strong 20 . . . .E!c8 ! , recom mended by M ikhail Gol ubev in the daily internet newsletter Chess Today. After 21 exd6+ (21 .E!xd6 .E!cxc2+ 22 �d1 .E!c7, or 21 exf6+ �xf6 22 .E!e2 Ae4 23 .E!xe4 .E!cxc2+ 24 � b1 .E!cb2+ 25 �c1 �c5) 21 . . . �d8 22 �d3 (22 �x b7 .E!cxc2+ 23 �b1 .E!cb2+ 24 �c1 Axf4+) 22 . . . .E!cxc2+ 23 �xc2 Axf4+ Black is close to a win.

21 .xb7 .txf4+ 22 cc!?b1 Ilha8 23 93 My opponent accom pan ied this move with a draw offer. I was expecting 23 .E!xd6 .E!a1 + 24 �b2 .E!8a2+ (of course, Black can force a d raw by perpetual with 24 . . . .E!1 a2+, as bad is 25 � b3 .E!8a3+ 26 �c4 .E!xc2+ 27 � b5 �xd6) 25 �b3 �xd6 26 .E!xa1 .E!xa1 27 c4 .E!c1 28 g3 00 (28 c5+ ? .E!xc5), but Alexander thought that Black would give up his bishop or knight for the pair of passed pawns, after which Black wou ld have the advantage i n the end ing.

23

. • .

Il2a7

I rejected the draw offer for several reasons. First of al l , I believed and sti l l bel ieve that Black has fu ll com pensation . And on top of that, it would be a pity to abort the game in such an entertaining position .

24 .c6 After 24 �xa7 (24 �e4 d5) 24 . . . .E!xa7 25 gxf4 exf4 I have more trust in Black's pawns than in White's b-pawn .

24

• • •

Ila6

Black could also have forced a draw here by 24 . . . Ad2 25 .E!xd2 .E!a1 + 26 �b2 .E!8a2+ 27 �b3 .E!a3+ 28 �b2 = .

215


a

b

c

d

29 gxh7 29 gg4 Ae3. 29 . . . �f6 I didn't want to al low White counterplay after 29 . . . ttlb6 30 gf1 . 30 h4?

e 8

7 6 5

5

4

4

3

3

2

2

a 8

K

b

c

d

...

e 8

7 6

�------�--� �

25 "c3?! I was more worried that Wh ite would keep an eye on the black rooks, thus l i m iting their activity: 25 �b7 !? Ah6 (25 . . . Ag5 26 c3 Af6) 26 c3 Ag7 (26 . . . d5 27 gxd5 exd5 28 gxe5+, or 27 . . . ga1 + 28 � b2 g8a2+ =), and if 27 gxd6!? ga1 + 28 � b2 g8a2+ 29 � b3 �xd6 30 gxa1 gxa1 31 c4 + with a position simi lar to that d iscussed in the note to White's 23rd move.

25 . . . £h6 25 . . . Ag5 !? It was hard to decide which was the more precise route for the bishop onto the long diagonal .

26 ge4? Now the rook becomes stranded in the cor ner of the board . H owever, it is very hard to suggest (especial ly i n time trouble) how White could prevent the simple plan of . . . d6d5, . . . Ag7 and . . . ttld7-b6-c4 etc. (if 26 gd3 d5).

26 . . . d5 27 gh4 d4! 28 "b2 The fianchetto is not the best option for the most powerful piece.

28 . . . £g5 There is no reason to play 28 . . . ttlb6, al lowing 29 gd xd4 ! exd4 (29 . . . ttla4 30 �a3 exd4 31 b5+ gd6 32 gxd4) 30 gxh6, even though after 30 . . . gd8, as indicated by Gol u bev, Black is doing fine.

5

5

4

4

3

3

2

2

I was concerned about 30 b5 ! , when Black would have to choose between 30 . . . ga1 + 31 �xa1 gxa1 + 32 �xa1 ttlxh7 , wh ich should be good enough to win , and the at tempt to win by a mating attack : 30 . . . ga4 31 b6 ! ttld5 (31 . . . ttl x h7 32 b7 gb8 33 �b5 00 ) , and now : A) 32 b7 ? gb8 33 gd3 e4 34 ga3 (34 gxd4 ttlc3+ 35 �xc3 gx b7+) 34 . . . gb4 35 gb3 e3 ! 36 gx b4 e2 -+ ; B) 32 gd3 ! (a strong defence) 32 . . . e4 33 gxd4 ! Af6 (33 . . . ga1 + 34 �xa1 gxa1 + 35 �xa1 Af6 36 b7 Axd4+ 37 � b1 +-) 34 gxa4 (34 b7 !?) 34 . . . gxa4 35 c3 ttlxc3+ 36 �c2 and it is White who is i n the driving seat. 30 . . . �d5? I was surprised by the last move and was somehow fixed with the idea of the knight going to c3 via d5. 30 . . . ttle4 31 b5 ga4 32 gd3 (32 h xg5 ttlc3+) 32 . . . Ad2 was win ning on the spot. 31 gd3 £d2?! Again played u n der the i l l usion of a q uick mate. The win wou ld have been m issed after 31 . . . e4 32 h x g 5 exd3 33 �xd4 ga1 +

216

Game 50


34 � b2 gSa2+ 35 � b3 = , but the cool headed 31 . . . Af6 32 b5 (32 gf3 e4) 32 . . . ga4 was the strongest, when White cannot stop the march of the black pawns : 33 b6 e4 34 b7 gbS 35 gb3 (35 ga3 d3) 35 . . . d3 36 c3 e3.

32 Dxd2 �c3+ 33 «i&c1 Only here did I discover that I was not mating my opponent.

40 Dc7+ More practical chances were offered by 40 gde7 , when after 40 . . . �d5 (40 . . . Ci!>d6! 41 gd7+ Ci!>e5 is better) 41 gh6! �e4 42 ghxe6+ �f3 43 gd7 White manages to prevent the king's advance.

40 . . . �d5 41 Dcd7+ 41 ghd7+ Ci!>e4 42 gc6 e5 .

33. . . Da1+ 34 Oxa1 Dxa1 + 35 �b2 Db1 + 36 �a3 8

8

7 6

6

5

5

4

4

3

3

2

2

a

b

c

9

The time control was reached and I immedi ately spotted a beautifu l king manoeuvre.

h

36 . . . e4!

41 . . . �e4!

The right decision ! Black has to use his last tru m p - the passed e - pawn ! 36 . . . gf1 37 Ci!>b3 00 would have let the king out of the mating net.

37 Df2 Wh ite's only chance is to try to win the f7 pawn and coord i nate his rooks. After 37 gxd4 e3 he has to g ive up his rook for just the e-pawn .

37 . . . e3 38 Dfxf7+ �d6 39 Dd7+ «i&c6 39 . . . Ci!>e5 was more accurate, but i n time trouble I avoided taki ng a committing deci sio n . 40 gh5+ �e4 41 gd6 e5 42 ge6 Ci!>f3 would have been simi lar to the game.

The king is aiming for c2 , where it will be an integral part of the mating construction .

42 Dde7 «i&f3 43 Dxe6 �e2! 44 Dd7 �d2 45 Dxe3 White has to give up his rook to avoid mate. If 45 gxd4+ Ci!>xc2 followed by . . . gb3#. 45 . . . �xe3

After 45 . . . Ci!>xc2 46 gxc3+ White would be able to com plicate matters. 46

Df7 �d5 47 b5

47 h5 Ci!>d2 .

* * *

217

47 . . . �d2 White resigns


12 . . . Had8

B o r i s G e l fa n d - Al exa n d e r G r i s ch u k R u s s i a n Leag u e , Dagomys 2004 Tarrasch Defence [D34J After some setbacks i n recent years, chess is regaining its place among the most pres tigious sports in Russia. A n u m ber of high level events took place i n 2004 , and i n ad d ition the team championsh i p was trans formed into a professional league, gaining in status and strength. This year it was splen d id ly organ ised i n Dagomys, a suburb of the Black Sea resort of Soch i . I represented the Termosteps team from Samara, which is owned by Igor Burshtein , a big fan of chess and a close friend of mine. Our team was al lowed to play in the top league when another team failed to appear at the open ing cere mony. As a result, I managed to join the event o nly in the second round . It took me some days to warm up, but I was very pleased by my standard of chess i n the second half of the event.

1 d4 d5 2 c4 e6 3 �c3 c5 The open ing choice of my opponent came as something of a surprise to me. The Tar rasch Defence was at the core of Alexander's repertoire until 2002 . In the last few years he has switched to the N i mzo- In d ian and vari ous forms of the Slav.

4 cxd5 exd5 5 �f3 �c6 6 g3 �f6 7 £g2 £e7 8 0-0 0-0 9 .lg5 c4 Deviating from 9 . . . cxd4 10 �xd4 h6 11 Ae3 ge8 as we played in Biel 2001 .

10 �e5 .le6 1 1 b3 "a5 12 "d2! A strong sem i-novelty (mean ing that it was never at the centre of theoretical discussions and only a few games have been played with it) , that throws the entire line with 9 . . . c4 into question . Generally speaking, I don 't believe that it is possible to refute any open i ngs or variations. However, i n this case it seems to me that i n all li nes Black is doomed to passive defence.

8

8

7 6

6

5

5

4

4

3

3 2

c

e

9

'If

1 2 . . . Ab4 1 3 Axf6 g xf6 14 �xc6 bxc6 15 gfc1 occurred in a couple of games. I think that Black has no compensation for the serious weaknesses in the vicinity of his king . Wh ite is also better after 1 2 . . . �a6 1 3 bxc4 dxc4 14 �xc6 bxc6 15 �c2 ;t . And the other rook move is also bad , as after 1 2 . . . gfd8 1 3 �xc6 bxc6 14 bxc4 d xc4 1 5 Axc6 the rook on a8 is hanging. 1 3 bxc4! Ab4 The point of Wh ite's idea can be seen from the followi ng line: 13 . . . �xe5 14 �xd5 ! �xd2 (White is clearly better after 14 . . . gxd5 15 cxd5 �xd2 1 6 Axd2 (an amusing case of two d-pawns attacking two pieces on the e -fi le) 1 6 . . . �c4 17 d xe6 �xd2 18 gfd1 ±) 15 �xe7+ �h8 16 Axd2 �xc4 17 Ab4 gfe8 18 Ax b7 ! (White spoiled his advantage with 18 d5 �xd5 1 9 �xd5 Axd5 20 gfd1 Ae6 O'Cinneide- Halkias, European Team Cham pionsh ip, Leon 2001 ) 18 . . . gd7 (or 18 . . . gb8 19 Ac6 gx b4 20 Axe8 �xe8 21 gab1 ) 1 9 Ac6 gexe7 20 Axe7 gxe7 21 gfc1 and the rook and two pawns are clearly stronger than the two kn ights. However, the line holds ! A few months after the game was played , the German 1 M Blauert fou n d a beautifu l combi nation : 13 . . . �xd4!! 14 �xd4 d x c4 1 5 �e3 Ac5 1 6 �f4 Ad4 or 1 6 �f3 Ad4 , regain in g the piece with comfort. H is opponent, St. Pedersen , took the wise decision to make a d raw after

218

Game 51

Gelfand - G rischuk, Russian League, Dagomys 2004

14 ttlxdS 'tWxd2 1 S ttlxe7+ �h8 1 6 Axd2 ttlxe2+ 1 7 �h1 .E!xd2. It is rather White who has to equalise here.

14 �xc6 bxc6 1 5 l1fc1 Here I spent a lot of time, trying to d is cover whether Wh ite had anything better than the game conti nuation, but I wasn 't i m pressed by 1S .E!ac1 d x c4 16 Axc6 'tWb6 17 dS (17 Axf6 gxf6 18 dS? 'tWxc6) 17 . . . .E!xdS (17 . . . Axc3 18 'tWxc3 ttlxdS 19 'tWxc4 ttle7) 1 8 AxdS ttlxdS 1 9 ttlxdS Axd2 20 ttlx b6 Axc1 .

1 5 . . . dxc4 1 6 Axf6 16 Axc6 'tWb6 17 dS ttle4 is fine for Black.

1 6 . . . gxf6 17 e3 c5 18 d5 It is i mportant to exchange the light-square bishops, as the e6 bishop is the key piece that holds Black's position together.

Alexander rejected 21 . . . .E!fd8, as 22 .E!xc4 'tWd2 23 .E!xcS leaves White with an extra pawn. 22 11xc4?! A superficial move that allows Black to ex change queens. H owever, I was very op tim istic about the double -rook end i n g . I n m y first encounter with Anatoly Karpov i n Linares 1 991 I had a simi lar ending, in which with Blac k I was defend i ng a better pawn structure (my pawn was on g6, not on f6) and where I was confident that we wou ld soon agree a d raw. But the 1 2t h world champion taught me a memorable lesson, allowing me no chances in such a harmless-looking pos ition . My hope of exchanging my a- and c pawns for the a-pawn proved t o b e a mere illusion .

Anatoly Karpov - Boris Gelfand Linares 1 991 a

18 . . . Axc3 1 9 l1xc3 Obviously, Wh ite is not i n a hurry to ex change queen s : 19 'tWxc3 'tWxc3 20 .E!xc3 AxdS 21 AxdS .E!xdS 22 .E!xc4 .E! b8 23 .E!c2 .E!b6.

d

e

8

8

7

7

6

6

4

4

•

3 2

8

7

a 6 5 4

2

Th is position has been reached almost by force after Black's 13t h move. Black faces a difficult defence.

b

c

11

29 g4 If 29 .E!cS .E!a6!? intending . . . .E!a4. 29 hxg4 30 hxg4 � It was better to gain some activity by giving u p the c - pawn : 30 . . . cS !? 31 .E!xcS (31 �g3 .E!a6 32 a4 �f6) 31 . . . .E!xcS 32 .E!xcS .E!a6 33 .E!c2 .E!a4 with good drawing chances. 31 �g3 �e6 32 a4! Threatening 33 as. 32 . . . �d7 I failed to sense the danger and defended too passively. 32 . . . �d6 33 .E!d4+ �e6 34 .E!cS ! . . •

3

21 . . . l1c8

c

5

19 . . . Axd5 20 Axd5 l1xd5 21 1Ic2

8

b

219


23

gd7 was stronger, and if 35 ge5+, as sug gested by Karpov, then 35 . . . �xe5 36 gxd7 �f6 37 gxa7 g b4 fol lowed by the advance of the c-pawn .

. . .

•xe2 24 111 xe2

8

8

33 g 5 lla6 34 Ild4+ �e8 34 . . . �e6 ! ? , and if 35 gc5 gd7 36 a5 gd5 ! (Karpov) .

35 !le5 !lb6 36 em4 !ld7?! 37 !lxd7 �xd7 38 �e5 ± �e7 39 f4 Ilb4 40 lla5 Ilb7 41 e4 a

b

c

d

...

e

8

8

7

7

6

6

5

5 4

3

3

2

2

White has achieved complete domi nation .

41 . . . !le7 41 . . . gd7 42 gc5 gd6 43 f5 gxf5 44 exf5 f6+ 45 gxf6+ gxf6 46 ga5 and wins (Karpov) .

42 lle5 lle8 43 !le3 lle8 44 !le4 !le8 45 Ilb4 lle7 46 a5! �d7 47 Ilb3! +- �e7 48 a6 �d7 49 em6 �e8 50 Ilh3 Ild7 51 f5 g xf5 52 exf5 e5 53 lle3 lle7 54 g6 fxg6 55 fxg6 �d7 56 g7 lle8 57 Ilg3 1-0 Now let us return to my game with Grischuk, where it is obvious that if White can keep the queens on in favourable circumstances, it should be to his advantage. For this rea son 22 gc1 ! was stronger, when 22 . . . gd3 ? is bad because of 23 gxc4 1!fd2 24 gxc5 ! . An other idea that deserved consideration was 22 gb1 I?�

22

. . .

•d2 23 lle1

If 23 VWe4, then 23 . . . gcd8, threatening . . . 1!fe2 and . . . gd1 +.

So, a double-rook end ing has been reached . For the sake of clarity, let me rem i nd you that the rook end ing without the queen side pawns is theoretically drawn , whereas if White were able to win the c5 and f6 pawns in return for the a-pawn , he would win in similar fashion to the game Romanish i n Polugayevsky from t h e 1 974 USSR Cham pionship. I also think that exchanging one pair of rooks wi ll favour Black, as Wh ite has this option at any point. Moreover, if White should manage to activate his second rook (as in the game) , all Black's pawns will be come targets.

24 . . . !le6 A strong move, preventi ng White from oc cupying the b -fi le. If 24 . . . �g7, then 25 gb2 gc6 26 gb7.

25 �g2? A pOintless move, played automatically. I think I overestimated the size of White's edge. 25 g b2 gb6 was harmless, but a stronger option was 25 ga4 ! gc7 26 �g2 f5 27 �f3, when White would at least be able to exchange a pair of rooks on the d - line (�e2 followed by gd2 or gc3-d3) i n favou rable circumstances.

25 . . . lla6! The rook has reached an active position, from where it ties down one of its counter parts to the defence of the a-pawn .

220

Gelfand - Grischuk, Russian League, Dagomys 2004

Game 51

26 a4 After 26 �f3 ga3 I don 't see how White can make any progress.

26 . . . h6 Short of time, my opponent fails to fi nd the best way to defend . I think that a better idea was to put the pawn on f5 by 26 . . . �g7 27 �f3 �g6 28 �e2 (28 gg4+ gg5) 28 . . .f5 . Now i t wou ld seem that White has noth ing better than to exchange a pair of rooks by 29 gd2 gad6 (29 . . . ge5 30 gd8 ; 29 . . . gxd2+ 30 �xd2) 30 gxd5 gxd5 31 gc2 �f6 32 gb2 when he keeps good winning chances. After the i m med iate 26 . . . f5 ? 27 gxc5 gxc5 28 gxc5 gxa4 29 gxf5 Black loses, for ex ample: 29 . . . a5 30 �f3 ga1 31 h4 a4 32 ga5 a3 33 h5 h6 34 g4 �g7 35 e4 ga2 36 �e3 ga1 37 ga6 ga2 (37 . . . a2 38 �f4) 38 e5 ga1 39 �f4 gf1 40 f3 ga1 41 �f5 a2 42 f4.

27 �3 �8 28 �e2 a

b

c

d

e

ga1 34 h4 a5 35 �f3 a4 36 ga5 a3 37 g4 ! (37 �f4 ? ga2 38 f3 ga1 39 h5 a2 40 g4 �h7 41 e4 �g7 leads to a draw) 37 . . . gh1 (37 . . . a2 38 h5) 38 h5 ga1 39 ga6, or 31 . . . ga1 32 gxh6 a5 33 h4 a4 34 gh8 a3 35 ga8. I th i n k that the best defence was 28 . . . gh5, keeping the king and the other rook i n their best positions : 29 h4 gd5 30 g2c3 ! (30 gb2 gb6 31 gb5 gxb5 32 a x b5 �g7 33 �f3 gd6 34 gxc5 g b6 allows Black to liqu idate into a drawn endi ng) 30 . . . �g7 31 gd3 gxd3 32 �xd3, although Wh ite retains excellent winning chances.

29 J:lb2! J:ld6 30 J:lb8+ �e7 31 J:lh4 Black is unable to protect all his weaknesses. If 31 gc8 gda6 32 gc7+ �e6.

31 . . . J:lb6 Or 31 . . . gc6 32 ge4+ ge6 33 gb7+ ! �d6 (33 . . . �f8 34 gh4 gc6 35 gxh6 �g7 36 gh4 c4 37 gb4 c3 38 ghc4) 34 gf4 .

...

f

32 Dc8 �d7 33 J:lh8 J:lb4 34 J:l4xh6 ...

8

8

7

7

8

6

7

1...1.-.-'

I...,...../--'

5

6

4

5

3

4 3

3

2 f

28

•. .

9

h

Da5?

Th is allows the rook at c2 to come i nto the action . I g uess that Grischuk was afraid of losi ng the c5 pawn at some stage after e3e4. This is the main problem in such pos itions - even if it is possible to defend them d u ri n g home analysis, when you are actu ally playing it is extremely d ifficult to decide which threats are real and which are illusory.

34

...

J:lb6

34 . . . gaxa4 35 gxf6 gb2+ 36 �f3 gaa2 37 �g2 �e7 38 gf4 is equally hopeless for Black. 35

J:l6h7 Dxa4 36 Dxf7+ �c6

If 36 . . . �e6, then 37 ghf8.

28 . . . �e7 loses a pawn after 29 gxc5 gx c5 30 gxc5 gxa4 31 gh5 and White is winning, for example: 31 . . . �f8 32 gxh6 �g7 33 gh5

221

37 Dxf6+ �b5 38 Dxb6+ a x b6 39 h4 J:la2+ 40 �3 c4 41 h5 J:la7 42 J:lcB J:lf7+ 43 �e2 �b4 44 g4 J:lh7 45 14 Black resigns

Com b inations I include here a number of game extracts, in which there was either some interesting tactical play, or else a com binational solution proved possible. For train i ng purposes the reader may wish to study the diagram positions and decide what he would have played , before checking how the the author continued .

Va l e r y Ya k u b e n y a - B o r i s G e lfa n d M i n s k 1 9 80 a

b

c

d

e

..

f

19 . . . �xb2! I had no problem in finding this combination , as I made a simi lar one at the age of 1 2 !

20 .id3

8

The point of the sacrifice is 20 �x b2 Aa3+ 21 �xa3 �c3+ 22 �b3 a4 23 Aa7 axb3 24 cxb3 �c5 ! 25 Ax b5+ �f8 -+. Or 20 Ax b5 gb8 ! 21 �x b2 gxb5+ 22 �b3 a4.

7 6 5

5

4

4

3

3

2

2

20 . . . �xd3 21 cxd3 �f6 22 �e4 �xd5 Black is simply two pawns up.

23 flhg1 b4 24 fldf1 .ixh4 25 "f3 25 �x h4 �xe3.

25 . . . �xe3 26 "xe3 b3 27 a x b3 flb8 28 d4 0-0 29 g5 hxg5 0-1

1 6 . . . �xb2! 17 �xb2 17 �xd5 Axd5 18 �x b2 Ac4. 17 . . . .ixa3+ ! 18 �b3 18 �xa3 �xc3+ 19 �a2 �x h3 -+ . 18 . . . "a5 19 flb1 �c5+ 0-1

B o r i s G e l fa n d - An d rey P a v l ov B e l o r u s s i a n J u n i o r Team C h a m p i o n s h i p , P i n s k 1 9 84

M i c h a e l A d a m s - B o r i s G e l fa n d E u ropean J u n i o r C h am p i o n s h i p , Arn h e m 1 9 8 8 a

b

c

d

e

a

7

6

6

5

5

4

4

3

3

d

e

f

h

7

8 7

c

8

..

f

b

6

6

5

5

4

4

3

3 2

L...-

....::....

_______

_--I

lf

2

Here I calculated a lengthy combination .

1 5 cxd5 cxd5 1 6 �xd5! �xf2

222

Combinations

Or 16 . . . gc8 17 � c3. Better was 16 . . . £xd5 17 £ xe4, when Black is 'only' a pawn down.

17 Axh7+ cc!1hS 1S IIxe6! E l i m i nating the central defender of Black's position.

1S . . . fxe6

43

Uxd7! Oxd7 44 �x16 OdS

The point of Wh ite's combination is 44 . . . ffxd6 45 Wlx h6+ ! g x h6 46 � xe8+. 45

Oh5 IIfS 46 �e8! ct>gS

46 . . . ffxe8 47 ffx h6+ � g8 48 Wlxg7#.

47 0g6 1-0

18 . . . gxe6 1 9 ffxf2 � x h7 20 �g5+.

47 . . . gf7 48 �f6+ � f8 (48 . . . gxf6 49 A xf6 Wlxf6 50 ffxf6 g xf6 51 d7) 49 ff h7 gxf6 50 ff h8+.

19 �g5 UfS

S e r g ey M ov s e s i a n - B o r i s G e l fa n d S i g e m a n & C o . , M a l m o 1 9 99 a

20 £gS!!

b

c

d

..

e

8

8

7

7

6

6

5

5

4

4

3

The key move of the combination.

2

20 . . . lIf5 21 �f7+ cc!1xgS 21 . . . gxf7 22 ff h7#. 22 �xdS IIxdS 23 �e3 1116 24 �c4 Ilea 25 £g5 IIf5 26 Oe2 IIxg5 27 ct>xf2 Axh2 2S Oxe6+ ct>h7 29 Oxd7 1-0

2

36 . . . lIxd5! This queen sacrifice leads to a forced mate.

37 11xdS+ B o r i s G e l fa n d - An g u s D u n n i n g t o n O akh a m 1 9 88

37 gxg6+ � h8.

37 . . . lIxdS 38 Oxc4 IIxd1+ 39 ct>h2 £f4+ 40 g3 hxg3+ 41 ct>g2 IId2+ 0-1

223

Combinations

B o r i s G e l fa n d - Peter Syi d l e r Koop Tj u c h e m , G ro n i n g e n 1 9 9 6 a

b

c

d

With this typical sacrifice Blac k seizes the i nitiative.

22 cxb4 �xb4 23 "b1 �xd3 24 ne2 e4

e

The strong kn ight at d3 makes White's pos ition hard to defend .

8

25 b3 Ad4 26 �f1 a6 27 £e3 b5 28 a x b5 a x b5 29 �cd2 £f6 30 "c2 c4 31 bxc4 bxc4 32 �b1 nb8 33 �c3 nb2 34 "xb2 �xb2 35 flxb2 £xc3 36 nc2 Ag7 37 £d4 nd8 38 ncd2 nxd4 39 nxd4 £xd4 40 flxd4 c3 41 fld1 Ab3 42 nc1 c2 43 �e3 "c3 44 �xc2 Axc2 45 £f1 e3 46 h5 "d2 0-1

6 5

5

4

4

3

3

�------�--�

�

23 £e6!! This intermediate move completely destroys the coordination of Peter's pieces. If 23 Axf7 f5 24 �e6 gxc1 25 gxc1 tlVd2 26 g b1 gc8 with counterplay.

Lj u b o m i r Lj u b oj e Y i c - B o r i s G e lfa nd M e l o d y A m b e r, M o n aco b l i n d 1 9 99 a

b

c

...

d 8

7

23 . . . flb8

6

6

5

5

4

4

24 £xf7 1-0

3

3

24 . . . f5 25 �e6.

2

2

23 . . . gce8 24 �xe8 gxe8 25 tlVxf7, or 23 . . .fxe6 24 �xe6 Ah6 25 �xf8 gxf8 26 gc7.

a

Va l e ry S a l oy - B o r i s G e l fa n d H oo g o v e n s B l itz, Wij k aan Zee 1 9 9 8 a

b

c

d

e

...

f

b

c

9

h

White has carelessly 'forgotten ' to castle and Black exploits this.

14 . . . �xf2! 1 5 cc!1xf2 �g4+ 16 cc!1g1

8

Or 16 i'e1 �xe3 17 tlVb1 Af5.

16 . . . �xe3

6

Black has two pawns for the sacrificed piece and h is knight at e3 is a 'central stri ker' in front of goal .

5

3

17 "d2 2

2

a

b

21 . . . �db4

c

d

e

f

9

1 7 tlVb3 tlVb6 1 8 �bd4 Ah3 ! .

17 . . . Af4

h

Black should never regain the exchange on d1 , unless he has a forced win of material .

224

Combinations

18 £d3 18 Ac5 Ag4.

18 . . . £g4! More and more pieces are joining the attack. Wh ite cannot prevent mate without substan tial loss of material.

19 �bd4 19 !!f1 'tWb6 20 tDfd4 Ah6 21 !!f2 tDd1 22 !!f4 a6 23 tDd6 !!e1 + 24 'tWxe1 Axf4.

19 . . . £xf3 20 �xf3 "b6 21 £as 21 tDd4 tDxd1 22 'tWxf4 !!e4.

21 . . . �xd1+ 22 £xb6 £xd2 0-1

The final touch ! Too many black pieces are attacked . 23 Axc5 ? �c4 24 �xd7 �xa3 25 �x b8 Axc5+ . 23 . . . �cxe4 24 fxe4 £e8 24 . . . tDxe4 25 �xd7. 2S "f3 £e7 26 �xa8 "xa8 27 £cS White remains the exchange up for no com pensation . No further commentary is re quired . 27 . . . "b8 28 as "c7 29 flac1 "d7 30 £d3 �g7 31 "e3 £f7 32 £b6 Dxc1 33 Dxc1 "a4 34 £cS gS 35 "f2 "d7 36 "c2 h4 37 h3 £g6 38 .tb6 �h6 39 "c7 "e8 40 £cS "f8 41 flf1 1-0

A l exa n d e r M otylev - B o r i s G e l fa n d R u s s i a v . Rest o f t h e Wo rl d , M oscow rap i d 2 002

B o r i s G e l fa n d - Al exey S h i rov O l y m p i ad , I stan b u l 2 0 0 0 a

b

c

d

e

f

a 8

b

c

d

e

f

..

h 8

7

7

6

6

6

5

5 4

a

b

c

d

e

f

9

h

3

2

2

� 1

�

21 �bxd6! Not only most spectacu lar, but also the most efficient way to put an end to Black's resis tance. 21 �c3 b5 22 ax b5 Axb5 (22 . . . ax b5 23 'tWxa8 'tWxa8 24 !!xa8 !!xa8 25 �b6) 23 tDxb5 'tWx b5 24 !!fe1 ± .

21 . . . �xd6 22 �b6!! 22 Axc5 �xc4 23 Axc4 b6 (23 . . . Axa4 24 d6+ �h7 25 Ae6) 24 Ad6 ± and Wh ite is 'only' a pawn up.

22 . . . £f8 23 b4

3

Wh ite thought that he had eliminated Black's counterplay by playing a2-a4. 1 6 . . . �a3+ ! 17 �a2 Unpromising is 17 bxa3 'tWxc3 18 Ab2 'tWc7 =t . 1 7 . . . bS!! The point. Black beg i ns a d i rect assault on the king. 18 gS 1 8 �xb5 'tWxc2 1 9 'tWxc2 �xc2 , 1 8 Axb5 tDx b5 1 9 tDxb5 'tWxc2 (1 9 . . . 'tWc4 !?), 18 bxa3 'tWxc3 , or 18 �xa3 b4+ 1 9 �a2 bxc3. 18 . . . �d7 I don't see how White can save the game.

225

Combinations

19 �dS

28 �e7+

19 Ax b5 �x b5 20 �x b5 �c4 (20 . . . �xc2) 21 �c3 �c5, or 19 �x b5 �xc2 20 �xc2 �xc2 21 Ah3 �b4+ 22 �a3 d5.

19 . . . AxdS 20 nxdS bxa4 21 c!>xa3 21 �a1 �c4 ! and the knight at a1 is l ittle help to White, whereas 21 . . . �xc2 22 �xc2 �x c2 23 �xc2 !!xc2 24 Ae3 leaves h i m some hope.

21 . . . axb3 22 c!>xb3 a4+ 23 c!>a2 a3!

28 bxc3 !!xa2 , or 28 �xc3 �xc3 29 bxc3 !!xa2 .

28 ... c!>h8 29 �xe8 nxa2! 30 bxe3 Black draws despite bei ng the exchange down after 30 !!a7 !!x b2 31 !!d1 �d2 32 .§.a3 �b3+ 33 !!x b3 !!x b3 34 �c2 !!a3 35 .§.d3 Ab4, or 30 !!hd1 Ax b2+ 31 �c2 Af6+ 32 � b3 !! b2+ 33 �a3 !!b8 34 !!c7 Ab2+ 35 �a4 �c3+.

Black has only one goal - the white king !

30 Ila1 + 31 c!>e2 Ilxh1 32 �d6! nxh2 33 �xe4 fxe4 . • .

24 bxa3 �b6 2S nbS 25 !!d3 d5!? 26 exd5 �a4.

a

2S . . . �a4 26 nb3 dS

8

The bishop joins the attack

7

27 exdS �e3+ 28 c!>a1 neb8 29 d6 ,txd6 0-1 White is helpless against 30 . . . !!x b3 31 cxb3 Axa3 .

b

c

d

1--"--

6

6

5

5

4

4

3

3

2

Laj o s P o r t i seh - B o r i s G e l fa n d Li n ares 1 9 9 0 a

b

c

d

e

......;;....

L..-_______

..

f

8

Black sti ll has to play accurately to hold the draw.

8

7

34

6

6

5

5

4

4

3

3

----' 'l1

_

Ild2 nh4! 35 nf2

35 c4 !!xf4 36 �c3 (36 c5 e3) 36 . . . !!f5.

35 . . . c!>g7 36 c!>d2 36 c4 g5 ! 37 fxg5 e3 =.

36

.. .

2

ct>ts 37 c!>e3

37 c4 !? a

b

c

d

e

f

9

37

h

It looks like as though White is in the driving seat, as he is two pawns up and �e7+ is threatened , but . . .

...

c!>fs 38 e4

38 !!a2 ? !!xf4 39 !!f2 g5.

38 Ilg4 39 Ile2 Ilg3+ 40 c!>f2 c!>xf4 41 eS nd3 . •.

41 . . . e3+ ? 42 �e1 ! .

27 . . . Axe3!! 27 . . . !!xa2 28 � b1 ! �xc3+ 29 �xc3 !!a6 30 !!c1 ± leaves Wh ite with an extra pawn and serious winning chances.

226

42 e6 nd8 43 e7 ne8 44 neS h S 45 g3+ c!>g4 46 Ile4 c!>fs 47 c!>e3 c!>g4 48 c!>f2 c!>fs 49 c!>e3 Y2-Y2

Combinations

B o r i s G e lfa n d - G i o r g i G i o rg a dze U S S R Yo u n g M asters , U z h g o rod 1 9 87 a

b

c

d

M issi ng the spectacu lar fin ish 31 .E!xd4+ cxd4 32 e5+ .E!xe5 33 .E!c6#.

31 . . . �e7 32 097+ �d6 33 Oh6+ �e7 34 gxc5 1-0

e

6

B o r i s G e l fa n d - Al exa n d e r S h a b a l ov U S S R Yo u n g M asters , Vi l n i u s 1 9 88

5

5

4

4

3

3

8

2

7

a

--1 11

_ :. L...-_______----:_

If you notice the idea of the combination on move 23, you shouldn't hesitate to go for it !

b

.

c

d

e

9

h 8

6

6

5

5

4

4

3

3

18 �xh6+ �h7

2

18 . . . �f8 19 �g6! and if 19 . . . fxg6 20 tDxg6#. a

b

A L...-______________��__� v

19 �f5 19 �h5 gxh6 20 �xf7+ £g7 and Black hangs on.

19 . . . g6 20 Oh3+ �g8 21 �h6+ �g7 It may seem that Wh ite's q ueen and his knight at h6 are awkward ly placed , but . . .

22 f5! £xe5 23 �xf7 ! Destroying the cover of the opponent's king .

23 . . . �xf7 24 Oh7+ £g7 25 Oxg6+ �e7 The king tries to run away from the dan ger zone, as otherwise the game will not last long . 25 . . . �f8 26 f6 Ah8 27 .E!f4 ! .E!xe3 28 .E!af1 and , despite his two extra pieces, Black has no defence agai nst the threat of VWh7 , or 25 . . . �g8 26 f6 .E!e7 27 f7+ �f8 28 �h7 .E!xf7 29 .E!xf7+ �xf7 30 .E!f1 + �e7 31 �xg7+ �d6 32 e4 .

26 Oxg7+ �d6 27 dxc5+ bxc5 28 e4! The attack goes on ! The extra piece is not felt.

28 . . . Aa6 29 Dfd1 d4 30 Dac1 g b8 31 0h6+

Black has seized the i n itiative, whereas I have failed to break through on the kingside, and my opponent's strong bishop, knight at c4 and a3 pawn dominate on the other side. Now . . . �d7, attacking the kn ight at g6, is a threat. For this reason Wh ite must com plicate matters at all costs. Note that, for the moment, the knight at g6 is cutting off the rook at h7 and is u ntouchable i n view of �xg6+ and �x h7 .

26 �xa3!! At the cost of a piece Wh ite beg ins an attack on the black ki n g . Th is is the best practical chance, even though Black should be able to win with best play.

26 . . . £xa3 Or 26 . . . tDxa3 27 .E!1 x b2 �xc3 28 .E!b8+ gx b8 29 .E!x b8+ �d7 30 tDf8+ �c6 (30 . . . �e7 31 h3) 31 h3 .E!h8 32 �xg7 and Black hardly has more than perpetual check by 32 . . . �c1 + 33 �h2 �f4+.

227

27 Db8+ gxb8 28 gxb8+ �d7 29 Of3

Combinations

Or 29 �h4 �a4 ! 30 �d8+ �c6 31 h3 fxg6 32 gc8 �a7 33 �e8+ �b7 34 gxc7+ �xc7 35 �f7+ �b8! (35 . . . � b6 36 �xe6+ Ad6 37 �xd5 00) 36 �g8+ �b7 37 �x h7 (37 �f7+ �a6 38 �xe6+ �b6 39 �g8 �c7) 37 . . . �c6 and Black keeps his extra piece.

38 gb1 Ag5.

38 •e1 + 39 �g2 .e4+ 40 �g1 .e1 + 41 �g2 .e4+ 42 �g1 �d4 . . .

42 . . . �c4 43 gb1 (43 �xc7+ ? Ac5) with nu merous threats. 43

29 . . . fxg6 30 .f7+ £e7

a

The weakness of Wh ite's back ran k forces him to waste a tempo.

8

7

31 �g8 �b6 ! , or 31 Ab4 �d6 ! '

31 . . . .tixe5 32 dxe5 .xc3 33 .g8

��-r��

...

b

8

31 g3

8

.xc7

8

6

6

5

5

4

4

3

3 2

7 6 5

5

4

4

3

3

Regaining one of the pieces. Here the game was adjourned and I naively expected that, in view of the vulnerable position of his king, Black would have to take a draw. But Alexan der d isplayed his famous fighting spirit and played on.

2

33

. . .

43

Th is careless move allows me to achieve sufficient cou nterplay, despite being two pieces down . Stronger was 33 . . . �c6 34 �c8 (34 �x h7 Ac5) 34 . . . gh8 ! ! (d iverting the q ueen from the attack) 35 �b7+ (35 �x h8 Ac5 36 gb1 Axf2+, or 36 �c8 �e1 + 37 �g2 �xf2+ 38 �h3 �f1 +) 35 . . . �d7 36 gx h8 Ac5 37 �b1 �xe5 + and Black repels the attack, while keeping a sufficient material advan tage. 34

�c6 35 .b7+ �d7 36 .c8+ �d6 37 Db6+ �c5 38 .b7 . . .

. . .

46 . . . Ac5 47 g b5 ! �c1 48 g b4+ �d3 49 gb3+.

47 .a7 ! I f Wh ite were t o regain t h e piece with 47 �f4+ �e4+ 48 �xe4+ �xe4 49 gx b4+ d4 + he would find h imself in a dubious rook end ing.

47 . . . •e4+ 48 �g1 .e1 + 49 �g2 .e4+ 50 �g1 �d3 51 .a4 £d2

.c8+ !

White doesn 't pay any attention to the rook on h7 , which is not taki ng part i n the game, and just concentrates on the opponent's monarch. 34

•e1 + 44 �g2 .e4+ 45 �g1 .e1 + 46 �g2 £b4

•xe5?

51 . . . Ac3 ? 52 �d1 + +-.

52 .b5+ .c4 53 .b1 + .c2 54 .b5+ .c4 55 .b1 + �d4 Black is trying to protect his king with his pieces, but as his rook at h7 is out of play, his chances are slim.

228

Combinations

a

b

c

d

e

f

9

8 7

And here Shabalov abandoned all attem pts to win, since after 69 . . . �d6 (69 . . . �c6 70 VWe8+ J;d7 71 VWc8+ J;c7 72 VWa6+ �d7 73 J;xd5+ +-) 70 J;b6+ �c7 71 VWa6 (71 J;e6 !?) it is rather Wh ite who has chances: 71 . . . J;d7 72 VWb7+ �d8 73 VWb8+ �c8 74 VWe5 ! Ac3 75 �e6.

h 8

r-'r==o

6

6

5

5

4

4

M i kh a i l G u re v i ch - B o r i s G e l fa n d S KA- M e p h i sto , M u n i c h 1 9 92 8

56 0xg6 0d3

7

56 . . J �h8 57 �xg7+ �c5 58 �c7+.

6

57 Oxe6 Oe4

5

Other moves are simi larly u nsuccessfu l : 57 . . . J;h8 5 8 J;b3 Ac3 5 9 �b6+, o r 57 . . . h5 58 J;b3.

4 3 2

58

Od7 Oe1 + 59 �g2 Oe4+ 60 �g1 g5 61 Oa4+ �d3 62 Ob5+ Oc4 63 Ob1 + Oc2 64 Ob5+ �e4 65 Oe2+ �d4 66 Db4+ �c5

8

8 7 6

��

6

5

5

4

4

3

67 11b5+ Wh ite could have regai ned the piece by 67 �b5+ �d6 68 J;d4, but after 68 . . . �c5 ! 69 �xc5+ �xc5 70 J;xd2 d4 he is i n trou ble, despite his extra pawn, as the d - pawn is extremely strong.

67 . . . �d4 68 llb4+ �c5 69 llb5+ �d4 Y2-Y2

1.

L-��

��

�____

__ __ __ __ __

The black queen is trapped and it seems as though the game is over. But one shou ldn 't be desperate and should keep on seeking chances. I know well from my own expe rience that a huge n u m ber of games have been saved by stubborn defence. 30 . . . 0h3 31 llh2 Oxh2+ ! This queen sacrifice is Black's best chance. Bad is 31 . . . VWf5 ? 32 �g6+ VWxg6 (32 . . . �g8 33 �e7+) 33 VWx g6 Af5 34 VWg5 Ax b1 35 �x b1 �fe4 36 VWe3 . 32 �xh2 Af5 The coord ination of the white pieces is very poor and this gives Black good hopes. 33 Ob3?! Stronger was 33 VWd1 ! , keeping an eye on the bishop at e2 . After 33 . . . �fe4 ?! 34 Ab2 J;xa3 !? 35 Axa3 �c3 (35 . . . Axe5 36 Ab2 �c3 37 Axc3 Axc3 38 Ad3) 36 VWe1 Axe5 37 Ad3 ! �de4 38 Axe4 �xe4 39 Ab2 White wins. H owever, my opponent was not sure about 33 . . . Ax b1 34 �x b1 �fe4 35 �d3 Eta2

229

Combinations

with some cou nterplay, accord i ng to Gure vich's annotations i n Informator.

33 . . . Iiae8! Pinning the wh ite pieces on the e -fi le. Also possi ble was the simple 33 . . . Ax b1 34 ttlx b1 (34 �x b1 gae8 35 Af4 ttlfe4) 34 . . . gae8 (34 . . . ttlfe4 35 ttlf3 gae8 36 �c2) 35 Af4 ttlg8 36 �c3 gf5 37 �d2 gexe5 38 Axe5 gxe5 with a probable draw.

34 Af4 34 Ad3?! gxe5 35 Af4 ttlg4+ 36 �g2 Axd3 37 �xd3 gh5 with attacking chances.

34 . �d7 ! 35 Ah5

39 . . . �e3+ Being short of time, I chose the simplest and safest optio n . All three resu lts would have been possible after 39 . . . Ac5 !? (threatening . . . ttlf2) 40 gb3, and now : A) 40 . . . �g8 41 �d2 Ad7 (41 . . . Axa3?! 42 gxa3 ttlxc4 43 �c3 �xa3 44 �xa3) 42 Axd6 ttle3+ 43 gxe3 Axe3 (43 . . . Ah3+ 44 �f2 Axe3+ 45 �xe3 gxe3 46 �xe3 cxd6 47 ttlb5 +-) 44 �h2 cxd6, and it is sufficient to exchange the lig ht-sq uare bishop for the kn ight to secure a draw;

. .

35 Ad3 �xe5 36 Axf5 gxf5 and the rook and knight fully compensate for the missing queen .

35 . . . �xe5 36 Axe8 36 gf1 ttlg4+ 37 Axg4 Axg4 with the initia tive, or 36 Axe5 gxe5.

36 . . . �g4+ 37 �g1 Wh ite's last chance to fight for a win was 37 �g2 gxe8. Despite Black's big material deficit, I think that he has sufficient cou nter chances, as all his pieces are taki ng part i n the attack.

37 . . . Ad4+ 38 �1 Iixe8 Suddenly Wh ite is faced with dangerous threats against his king ( . . . ttlf2 and . . . Ah3+).

39 1td1

B) 40 . . . ttlf2 !? 41 �a1 + �g8 42 g4 Ax g4 with fu ll com pensation .

40 Axe3 Axe3 Threatening . . . Ah3+.

41 �g2 41 gb2 ? Ah3+ 42 �e1 Ac5+ 43 ge2 Ab4+ 44 �f2 gf8+ 45 �e3 (45 �g1 gf1 + 46 �xf1 Axf1 ) 45 . . . Ax a3 , and if 46 �a1 + �g8 47 �xa3 ttlxc4+ .

41 . . . Ae4+ 42 �h3 �f7 !? Black could also have made a draw 'from the weaker side' by 42 . . . Ax b1 43 �x b1 Ac5 44 �b5 �xb5 45 �x b5 ge7 with a fortress.

39 gd1 '!? �h2+ 40 �g2 ge2+ 41 �h1 Ae4+ -+. a 8

b

c

d

...

e 8

7 6

6

5

5

4

4

3

3

2

2

43 Iib3 Forcing Black to give perpetual check. 43 g4 �g5+ 44 �g3 (44 �h4 ttlf3+ 45 �g3 gf8) 44 . . . gf8 45 gb3 Af4+ 46 �f2 Ae5+ 47 �e3 Af3 00 and the white king is vulnerable. 43

•. .

�g5+ 44 �h4

44 �h2 ? gfB 45 gxe3 ? gf2+ 46 �g1 ttlh3#.

44

. . .

�f3+

44 . . . Af3 ? 45 gxe3 +-.

45 �h3 �g5+ 46 �h4 �f3+ %-%

230

Combinations

l Iya S m i r i n - B o r i s G e l fa n d B elo russsian C h a m p i o n s h i p , M i n s k 1 9 84 a

b

c

d e

G e n n a d y S a g a l ch i k - B o r i s G e l fa n d U S S R J u n ior Championsh i p , Yu r m a l a 1 9 8 5

.

a

b

c

d e

9

•

h

8

8

7

7

8

8

6

6

7

7

5

5

6

6

4

4

5

5

3

3

4

4

2

3

3 2

a

Al l three fi les are open i n front of the white ki ng. But White is going to cover all his weaknesses by Aa4-b3, so Black has to act quickly.

h

Bad is 32 Axd3 �gd7, winning material.

25 ti)xe4 .b4

32 . . . d2!

Rei nstating the threat of . . . Ag5 in view of the pin on the 4t h ran k.

26 Ilhf1 26 f6 �x b2+ 27 �d2 �xf6 would not en able White to escape, as there is noth ing to protect his king.

•xb2+

A far-advanced pawn is always a strong tru m p ! Moreover, Black creates a danger ous threat, which White fails to notice.

33 .e2? 33 Ae2 is stronger, but after 33 . . . Ae3 the strong pawn on d2 ensures Black's advan tage. Now he begins a com bination .

Of course, not 26 . . . �xa4 ? 27 �f6+.

33 . . . Ilxg4 34

.xg4 Aa6 35 ti)xd2 Ilxd2

36 Ild1

27 �d2 Ag5! Exchanging the wh ite knight, the only re maining defender.

28 ti)xg5 hxg5 29 .e4 Ilb4 30 .c6 .d4+ 31 �e2 .e4+ 32 �2 Ilxa4 Black has won a piece and the game is over.

33 Ilxd6 Ile8 34 .xe4 Ilexe4 0-1

9

29 . . . d4 30 ti)e4 d3 31 cxd3 cxd3 32 Axg4

Removing the knight from the centre, but creating the deadly threat of . . . Ag5.

. . .

c

Black's kingside pawns are weak, so it time to start an attack in the centre.

24 . . . ti)h7 !

26

b

A last attem pt to resist.

36 . . . Ilxb2+ ! 0-1 Wh ite resigned , antici pating 37 �x b2 (37 �a1 Ad4) 37 . . . Aa3+ 38 �xa3 (38 �b1 Ad3+ 39 �xd3 �c1 #) 38 . . . �c3+ 39 �a4 b5#.

231

Combinations

J o e l L a u t i e r - B o r i s G e lfa n d F I D E Wo r l d C h a m p i o n s h i p rap i d , G ro n i n g e n 1 9 97

Paul Van der Sterren - Boris Gelfand S KA , M u n i c h 1 9 94

a

b

c

d

e

..

f

8

8

7 6

6

5

5 4

3

3 2

In a typical King 's I n d ian game I went for a d i rect attack, but failed to g ive mate. White has all the tru m ps on the q ueenside, and so Black has to look for a way to save the game.

I n an i m portant tie-break game from this world championsh ip knock-out event, I have managed to seize the initiative. An exchange sacrifice is the best way to develop it.

18 . . . l1xd2! 19 l1xd2 dxe3 20 fxe3

31 . . . • h4! Too slow is 31 . . . �g5 32 b6 Ah3 33 b x c7 �xf1 34 c8� Axc8 35 E!.xf1 . Also insuffi cient is 31 . . . Ah3 32 b6 ! (or 32 �xc7 �g5 33 E!.c2 �xf1 34 chxf1 Axg2+ 35 che2 Axf3+ 36 �d3) 32 . . . c x b6 33 E!.x b6 �xf1 34 �xf1 E!.xg2 35 �xg2 Axg2+ 36 chxg2 and Wh ite retains an advantage.

32 .xc7 ! There is no time for 32 �b2 E!.g5 , and if 33 E!.xc7 ? (Wh ite can sti l l d raw by 33 E!.c2 �h1 + 34 chf2 �h4 35 chg1 ) 33 . . . �h1 + 34 �f2 �xf1 35 E!.xc8+ �g7 36 �e1 �e3+ 37 chd2 �xg2+ 38 chd3 �xf3 -+ with a mat ing attack. Or 32 b6 ? �h1 + 33 chf2 �xf1 .

32 . . . �e2+ ! 33 £xe2 l1xg2+ ! 34 �xg2 £h3+ Black is two rooks dow n, but this does n ot change the outcome of the game.

20 E!.e2 exf2+ 21 E!.xf2 �g4 -+ .

20 . . .•e5! Black wins a second pawn and with it the game, than ks to the poor placement of White's pieces, especially his kn ight at a4.

21 .f1 21 E!.e1 Ab4 -+ . No better is 21 E!.ad1 Ag4 22 E!.f1 (22 �xa7 Axd1 23 E!.xd1 �g4 !) 22 . . . �xe3+ 23 E!.ff2 (23 E!.df2 Af5) 23 . . . Ab4 24 E!.d3 �c1 + 25 E!.f1 �c2 .

21 . . .•xe3+ 22 .f2 .a3! 23 .xa7 23 �f4 E!.c8.

23 . . . �e4 24 £xe4 The only defence against . . . Ac5+. After 24 E!.c2 Ac5+ ! the game concl udes with a smothered mate : 25 E!.xc5 �e3+ 26 chh1 �f2+ 27 chg1 �h3+ 28 chh1 �g1 + 29 E!.xg1 �f2#.

24 . . . £xe4 25 l1e2

35 �h1

25 E!.f1 Ac6 -+ .

The only move.

25 . . . l1a8 26 .e3 .xe3+ 27 l1xe3 l1xa4

35 . . . £g4+ 36 �g2 .g3+ 37 �h1 .h3+ Y2-Y2

and White resigned a few moves later (0-1 ).

232

Combinations

B o r i s G e lfa n d - J e r o e n P i ket M e l o d y Amber, M o n aco b l i n d 1 99 9 a

b

c

d

e 8

6 5 4 3 2

L...-_______--=-__.....

{f

Wh ite concludes the game with a simple but pleasing com bination .

32 �xe6+ ! fxe6 33 llxd7 .xd7 34 �f5+ �xf5 35 .xd7+ @h6 36 h xg6 h xg6 37 .xe6 @h7 38 Ad3 1-0

31 . . . �b8 32 ge8+ Af8 33 f4 (or 33 gc5 gd6 ! , but not 3 3 . . . gcxd8? 3 4 Axd8 �f7 3 5 gxd5 �xe8 36 gf5 !) 33 . . . gd6 ;!;; he could have re tained chances of a successful defence. 32 11ed1 ! Surprisingly, this rook temporarily leaves the e -fi le. 32 gcd1 would allow 32 . . . gd5. 32 . . . Ae5 Or 32 . . . gxd1 33 gxd1 , winning a piece. 33 llxd4 Axd4 34 lld1 c5 Black was counting on this tactical shot, but after 35 bxc5 �xc5 36 llxd4 �b3 37 lle4! he had to resign in view of 37 . . . �xa5 38 ge8#.

B o r i s G e l fa n d - A l exa n d e r B e l i a v s ky Reg g i o E m i l i a 1 9 91 a

b

c

d

e 8

8

B o r i s G e l fa n d - C h r i sto p h e r L u tz C an d i d ates To u rn a m e n t , D o rt m u n d 2 0 0 2

7 6

8

5

5

4

4

3 2

6 5

a

4 _�""'---l

3 2

'--_______--=-__.....

�

{f

31 �d8! Attacking the c6 pawn and cutting the rook at c8 out of the game.

31 . . . 11xd4? Black is reluctant to defend passively, but the heavy concentration of pieces on the d file leads to immediate loss of material . With

2

b

c

d

e

9

h

{f

It may seem that Black has gained counter chances, but Wh ite fi nds a precise way to fin ish the game. 23 Af7! Axd2 Black also wou ldn 't have been saved by 23 . . . �xf5 24 Axe8 Axd2 25 Ag6 ! (25 Axd2 �g3 allows some counter chances) 25 . . . �xg6 26 �xg6 hxg6 27 Axd2 with a techn ically won end ing, or 23 . . . gxf7 24 gxd4 (after 24 �xf7 gfB 25 �xf8+ Axf8 26 Ae3 Ac5 27 gfd1 �f3+ 28 g xf3 White would sti ll req uire some techn ique to win) 24 . . . �xd4 25 �xf7 gf8 26 Ae3.

233

Combinations

24 Ag6! +The key intermed iate move. 24 Axd2? ge5 ! and bad is 25 Ag6 77 �x g6 ! -+ .

24 . . . �e2+ 25 �h1 h6 26 Axd2 Black cannot prevent Ax h6.

26 . . . Ele7 26 . . . �xf2 27 gxf2 Ctlg3+ 28 �h2 Ctlxh5 29 Ax h5.

27 Axh6 1-0 I n view of 27 . . . gxh6 28 �x h6+ �g8 29 Ah7+ gxh7 30 �x b6.

8 7 6

,....--""""'-1

Kiri l does not find the correct response and he allows a mating attack on his exposed King. The best defence was to give u p t h e q ueen for three pieces after 1 6 . . . Ctlxf4 ! 1 7 Ctlxd8 gxd8 1 8 �c2 (18 gxe7 ? gxd1+ 1 9 gxd1 �f8) 18 . . . Ctlfe6 19 �xc4 gd4 ! 00 . It is hard to assess this position . Black's pieces are awkward ly placed , but if he can man age to develop them properly, thei r attacking potential may tel l .

16 . . . �b6 17 Axd5 Ae6 18 �h5 m8 19 gxe6 ! �xe6 20 ge1 gxf7 21 gxe6 +- was indicated by Maxim Notkin .

5 �"""""'---1

16 . . . �xf7?

The computer shows another interesting de fence : 16 . . . Ctle3 !? 17 �xd8 Axd8. However, Wh ite can keep an edge by 1 8 Ctld6 ! Ac7 1 9 gxe3 ! (19 Ctlxe8 Axf4) 1 9 . . . gxe3 20 fxe3 Ae6 (20 . . . Axd6 21 Ax d6 Ctld3 22 Ad5+) 21 Ctlx b7 Axf4 22 Ctlxc5 Axe3+ 23 �h1 Axc5 24 Axa8 ±, or 18 . . . Ctld3 19 Ad5+ �f8 20 gxe3 gxe3 21 Axe3 ± .

B o r i s G e l fa n d - K i r i l G e o rg i ev O l y m p i ad , C a l v i a 2004

6

guess that my opponent had reckoned only on 1 6 Axd5 Ae6, when Black is doing fi ne.

4 3

1 7 Axd5+

2

1 5 . . . �xd5? Black played this quite quickly. He definitely didn't pred ict White's next move. I was wor ried about 15 . . . Ae6 ! 1 6 Ctlxe7+ gxe7, when after 1 7 �c2 (or 17 �x d8+ gxd8 1 8 Ctlxc4 Ctld3 19 ge2 Ctlxf4 20 gxf4 ged7 21 Ctlxa5 b6 22 Ctlc6 gd1 + 23 ge1 gxe1 + 24 gxe1 gd2) 17 . . . gc8 !? (17 . . . Ctld3 18 ged1 Ad5 19 Ctlxd3 Axg2 20 �xg2 �d5+ 21 f3 cxd3 22 gxd3 =) 1 8 gad1 �e8 it is Wh ite who has to prove that he has equal chances.

16 �xf7 ! I had spotted the weakness of Black's f7 sq uare (his rook has moved to e8) , which allows White to start a dangerous attack. I

During the game I thought that 17 �h5+ �g8 18 Axd5+ �h8 19 Af7 was also strong . However, Black can hold by 1 9 . . . Ae6 ! (after 1 9 . . . gf8 20 Ag6 h6 21 Ax h6 Ag4 ! 22 Axg7+ �xg7 23 �xg4 Ag5 24 Ac2 Wh ite would seem to win) 20 Axe6 (20 gxe6 Ctlxe6 21 Axe6 �b6, or 20 Ag6 Ag8) 20 . . . Ctlxe6 21 gxe6 Af6 =.

17 . . . �g6 The king is obliged to move out into the open field. Bad is 1 7 . . . Ctle6 18 �h5+ �g8 1 9 gxe6 �h8 ( 1 9 . . . Axe6 20 Axe6+ �h8 21 Af5) 20 gh6 ! g x h6 21 Ae5+ Af6 22 �x e8+ ! (22 Axf6+ �xf6 23 �xe8+) 22 . . . �xe8 23 Axf6 with a spectacular mate, or 17 . . . Ae6 1 8 �h5+ �g8 1 9 gxe6 Ctlxe6 20 Axe6+ �h8 21 Af5 h6 22 �g6 �g8 23 Ae6+ �h8 24 Ax h6.

234

18 Ele5!

Combi nations

Th reaten ing mate in one and preventing . . . Af5 .

1B . . . .tf5 The game can not be saved by 1 8 . . . h6 19 � h5+ � h7 20 Af7 f!a6 (20 . . . f!f8 21 �g6+ � h8 22 A xh6 f!xf7 23 �xf7 �f8 24 A xg7+ �xg7 25 f!h5+) 21 Axe8 with a decisive ma terial advantage.

19 Ilxf5! �xf5 20 .h5+ The black king has gone too far from base and is going to be mated .

20 . . . .tg5 21 .xh7+ �6 Or 2 1 . . . g6 22 � h3+ � f6 23 A xg5+ � xg5 24 � h4+ �f5 25 �f4 with another nice mate.

22 .txg5+ �xg5 23 .tf7 I had seen this move from afar and was q u ite proud of it. However, as the com puter shows, Wh ite could have g iven mate in 9 moves : 23 f4+ � f6 24 � h4+ � f5 (24 . . . � g6 25 f5+ ! � xf5 26 �f4+ � g6 27 Af7+) 25 � h5+ g5 26 � h7+ � g4 27 h3+ � xg3 28 �f5 ! This qu iet move creates the u nstoppable th reat of �g4. Nevertheless, my decision is strong enough.

23

. . .

•d6

23 . . . � f6 24 �g6+ � e7 25 f!e1 + � d7 2 6 f!xe8, or 23 . . . f!e6 24 � h4+ �f5 25 �f4# .

24 Oxg7+ �5 25 .txeB 1-0 I n view of 25 . . . f!xe8 26 � f7+ .

* * *

Monaco 200 1 : Watched by a number of colleagues and journalists, Boris Gelfand analyses his game with Vishy Anand.

235

Endings I ncluded here are extracts from games, i n which matters were decided i n the endgame. Generally speaking, the positions are arranged according to material, from m i nor piece endings to endings where heavy pieces play the main role. Again , the positions may be used by the reader for training purposes.

Le o n i d M i l ov - B o r i s G e lfa n d USSR J u n ior Championsh i p , K i rovabad 1 9 84

B o r i s G e l fa n d - Ve se l i n To p a l ov H oog oven s , Wij k aan Zee 1 9 96 a 8

b

c

d

1_________-

e

f

9

h 8

8

8

7

7

6

6

6

5

5

5

4

r-�

"""""-1

4

7

4

3

3 2

7

r-___-.-==

2

3 2

2 �------�--�

�

41 �e2 Forced , as Black wins the pawn end ing after 41 f3 Axc3 42 bxc3 e5 43 a5 b5 44 axb6 c x b6 45 d6 xd6 46 xf4 a5 47 e3 a4 48 d3 e5.

63 . . . �xb4!! The bishop cannot stop both pawns.

64 �xe4 92 65 Ah2 �c4 66 �e3 Or 66 f3 b4 67 xg2 d5 ! (67 . . . b3 68 Ae5).

66 . . . b4 0-1 If Wh ite tries to rearrange his pieces to keep an eye on both pawns by 67 f2 (67 d2 d4 68 c2 e3 69 b3 f2) , then 67 . . . g1 � +! (67 . . . d5 68 e3 b3 69 d3 b2 70 c2 e4 71 x b2 d3 72 b3 e2 73 c4 f2 74 d4 g1 � 75 Axg 1 + xg1 76 e4 g 2 7 7 f5 =) 6 8 xg1 (68 Axg1 b3) 68 . . . d5 and the b - pawn queens.

41 . . . Axb2 42 �xf4 b5 43 �d3 Ae5 43 . . . Ad4 was another try: 44 ax b5 a5 (here too 44 . . . a x b5 45 f3 does not create any threats) 45 f3 a4 46 e2 a3 47 �c1 (pre mature is 47 b6 ? a2 48 b7 Aa7 -+) 47 . . . Ab2 (47 . . . c5 48 �b3+ c4 49 �xd4 xd4 50 b6 and the pawns queen simultaneously) 48 b6 c x b6 (48 . . . Axc1 49 b7 a2 50 b8� a1 � 51 �d8+ c5 52 �xc7+ and Wh ite draws from a position of strength) 49 �a2 c5 50 d3 b5, and now : A) Black should win after 5 1 c2 ? c4 52 d6 b4 53 d7 (53 �c1 Axc1 54 xc1 c3 55 b1 b3 56 d7 a2+ -+) 53 . . . b3+ 54 b1 Af6 55 e5 Ah4 56 �c1 Ad8 57 �e2 a2+ 58 b2 Ac7 59 d8� Axe5+ 60 c1 a1 �+;

236

Endings

B) 51 d6 ! (it is time to d ivert the opponent's king , even at the cost of two passed pawns) 51 . . . b4 (or 51 . . . �xd6 52 �c2 and after Wh ite's king reaches b3 he has a fortress) 52 d7 Af6 53 �c2 �c4 54 �c1 and Black cannot break through .

B o r i s G e l fa n d - Pa b l o S a n Seg u n d o M a d r i d 1 996

Bad was 43 . . . bxa4? 44 �x b2 a3 4 5 �c4+.

44 axb5 a5! 45 cc!n3 a4 46 �e2 46 �g4 may wel l also have been possible, but I saw a clear d raw after the move i n the game.

46

a3

. • .

a 8

b

c

d

e

I-..---=�-

8

7

7

6

6

5

5

4

4

3

3

2

2

48. . . l1h5? Black fal ls i nto the trap and captures the bishop, but loses the game. I was more afraid of 48 . . . gh7 ! 49 c6 (49 f7 �f8 50 gf6 gg7 , or 49 gxa5 gh5) 49 . . . �e5 50 Axf4 (50 gxa5+ �e4 with counterplay) 50 . . . �xc6 51 �g3 ;t and White is only slightly better.

49 f7 ! l1)(g5+ 50 cc!n2 �------�--�

�

Despite his extra piece, Black cannot pre vent one of the pawns from queening.

47 b6! By luring the black pawn onto the b -fi le, White creates the outl ine of a fortress. After 47 �c1 �c5 48 �d3 �x b5 49 �a2 �c5 Black has excel lent winning chances.

47 . . . cxb6 48 �c1 �c5 49 �d3 Af4 49 . . . b5 50 �c2 (50 d6 ! =) 50 . . . �c4 51 �b1 Af4 52 d6 Axd6 53 �a2 was simi lar to the game.

50 �c2 �c4 51 �b1 b5 52 d6! By diverting the opponent's king, Wh ite gains an im portant tempo.

52 . . . Axd6 53 �a2 �d4 54 �b3+ White has achieved his dream and has built a fortress.

54 �)(e4 55 �c1 �d4 56 �b3+ �c3 57 �c1 �b4 58 �b3 �a4 Y2-Y2

50

• . .

�f8

50 . . . �xc5 51 gxg6.

51 c6 Also good enough was 51 ga8 �h7 (51 . . . �e6 52 c6 +-) 52 gh8 gg6 53 gxh7 gf6 54 c6 �xc6 55 gh6 gxh6 56 f8tlf.

51

. • .

f3

51 . . . gh5 52 ga8 gh2+ 53 �g1 gc2 54 gxf8 �e6 55 gh8 �xf7 56 c7 gxc7 57 gh7+ , or 51 . . . �d6 52 ga8 �h7 (52 . . . �g6 53 gg8) 53 gh8.

52 11a8 52 c7 gg2+ 53 �xf3 gc2 .

. . .

237

52

. . •

�g6 53 c7 I1g2+ 54 cc!n1 1-0

Endings

B o r i s G e l fa n d - V i c t o r Ko rch n o i Dos H e r m a n as 1 9 99 8

8

7

7

�f6 58 Ad8+ �e6 59 �g8 �d7 60 Ag5 �e6, and if 61 h6 tjj x h6+ 62 Ax h6 �d5 63 �f7 �c4 64 �e6 � b3 65 Ag7 a5 66 �d5 a4, exchanging the last pawn.

8

6

7

5 4

4

3

3

6 5 4

2

3 a

b

c

d

e

f

9

2

h

A �______________�____� v

The bishop is stronger than the knight, White's king is i ncom parably more active, and the e6 pawn is strong. However, he has to play precisely to avoid missing the win.

47 h3! After 47 h4 �e7 Wh ite does n ot have any usefu l move, as 48 Ac5+ �e8 49 �f6 tjj c7 al lows the knight to 'stick' to the e6 pawn.

47

. • .

�e7 48 h4

Now it is Black to move and he is forced to step back.

48 . . . �e8 49 c!>f6 �d6 50 �g6 �c4 Black has a lost pawn ending after 50 . . . �e7 51 Ac5 ! �xe6 52 Axd6 �xd6 53 �x h5 �e7 54 �g6 �f8 (54 . . . a5 55 �g7) 55 �f6 a5 56 �e5 b5 (56 . . . �g7 57 �d5 �g6 58 �c5 �h5 59 �b5) 57 �d5 a4 58 �c5 when he is one tempo too late.

a

b

c

d

e

f

9

h

51 . . . b6 51 . . . tjj x b2 52 �xh5 was the alternative: A) 52 . . . b6 53 Ad4 (or the computer's 53 �g6 bxc5 54 h 5 ! and the pawn is unstoppable) 53 . . . tjjc 4 54 �g6 �e7 55 h5 �xe6 56 h6 tjjd 6 57 h7 ttlf7 58 Ax b6 ; B) 52 . . . tjjd 3 53 Ad6 b5 54 �g5 (but not 54 �g6 ? b4 55 h5 b3 56 h6 b2 57 h7 ttlf4+ ! 58 �g7 tjj g 6) 54 . . . b4 55 h5 and the pawn queens with mate.

52 Ad4 �e7 53 �xh5 �xe6 54 �g6 �d6 55 Axb6 55 h5 tjjf5 56 h 6 ? tjj x h 6 57 �xh6 �d5 58 Ax b6 �c4 was Black's last chance.

55 . . . �f5 56 h5 �e7+ 57 �g7 �f5+ 58 �g6 �h4+ 59 �g7 �f5+ 60 �8 Now White only has to win the knight for the h-pawn , which is quite simple.

51 Ac5! (see next diagram)

60 . . . 't!n6

A very precise move. White ind uces the pawn to move to b6, where it will be cap tured . The straightforward 51 Ad4 �e7 52 �xh5 �xe6 53 �g6 tjjd 6 54 h5 tjjf5 wou ld have allowed Black to escape after 55 Ab6 (or 55 h6 tjj x h 6 56 �xh6 �d5 57 Ac3 �c4) 55 . . . ttle7+ 56 �g7 ttlf5+ 57 �f8

61 Ad8+ �e6 62 �g8 �h6+ 63 �g7 �f7 63 . . . tjjf 5+ 64 �g6.

238

64 Ac7 1-0

Endings

B o r i s G e l fa n d - S e r g ey M ov s e s i a n E u ropean Team C h a m p i o n s h i p , Bat u m i 1 9 99 8

8

7

7

6

6

90 .td5! �g4 90 . . . ttlfS 91 �h3. a

b

c

8

8

7

7

6

6

5

5 4 3

4

4

3

3

2

2

2

2

L...-.....;."..

--:.

_ _ _ _ _ _

L...-

--:.

_ _ _ _ _ _ _

-' 'It

_ _

-' 'It

_ _

91 .tc6!! �xh4

78 h4

91 . . . ttlfS 92 Ad7.

I saw that White could hold the d raw after 78 �f3 h4 79 g x h4 g x h4 80 �e3 ttldS+ 81 �f3 ttlf4 82 Af1 ttle6 83 �e3 ttlgS 84 Ag2 . This is a position of m utual zugzwang . White must not allow the oppo nent's king to go to f4. Thus if he has to move he loses after 1 Af1 f4 2 �f2 �e4 followed by . . . f4-f3, . . . �f4 etc. But he cannot be g iven the move. However, d u ring the game I had some doubts about this and I chose a more risky path.

92 .td7 Again a m utual zugzwang position saves Wh ite. If it were h i m to move, he wou ld be in trouble after 1 �f3 ttle8 ! , and if 2 Axe8 ? �h3.

92 . . . �g5 Y2-Y2

To m a sz M a r kows k i - B o r i s G e l fa n d A k i b a R u b i n ste i n M e m o r i a l To u rn a m e n t , P o l a n i c a Zd roj 2 0 0 0

78. . . g4 79 ct>t2 �e4+ 80 �g2 �d6 81 .td3 �d4 82 .tb1 �e3 83 .tc2 �e4 84 .tb3 f4 85 g xf4 �xf4 86 .tc2

a

Black is trying to win the h4 pawn and White has to play extremely carefully to draw.

b

c

d

..

e

8

8

7

7

6

6

5

86 . . . �f6 87 .td3 �e8 87 . . . g3 88 Ae2 ttlg4 89 �h3 �e3 !? 90 Axg4 �f2 91 Af3 ! �xf3 and White is saved by stalemate.

4

4

3

3

2

2

88 .tg6 �g7 89 .t17 g3 So, Black has found a way to win the h4 pawn . However, the awkward position of the knight at g7 allowed me to find a study like draw. If 89 . . . �eS 90 �g3 �fS (90 . . . �f6 91 Ax hS) 91 Ab3 �f6 92 Ad1 .

51

•. .

Dxh5

I was not sure that I wou ld be able to save the game by preventing the b-pawn from queen i n g . H owever, this is Black's only chance,

239

Endings

as normally rook and bishop endings with opposite colour bishops are very hard for the defend i ng side, as I proved i n the same tournament against Peter Svidler (see be low).

52 Dxh5+ �xh5 53 �c4 �g5 54 b4 f5 55 Ad5 � The king is run n i ng to b8 to block White's potential passed pawn.

56 b5

And my opponent trusted that I would find 63 . . . Axa7 64 �xa7 �g6 65 Af3 (65 �b6 �h5 66 Af3+ �h4 67 �xb5 �g3) 65 . . . b4 66 �b6 (66 Ad1 b3 67 Ax b3 �h5 68 Ad1 + �h4) 6 6 . . . b 3 6 7 Ad1 b 2 6 8 Ac2 �h5 69 Axf5 �h4 70 Ab1 f5 !! 71 Axf5 �g3.

Peter Sv i d l e r - B o r i s G e l fa n d A k i b a R u b i n st e i n M e m o r i a l To u rnament, P o l a n i c a Zd roj 2 0 0 0

56 �b5 Ac3.

56 . . . �e7 57 b6 57 �c5 �d7 58 b6 ax b6+ 59 �x b6 (59 axb6 �c8) 59 . . . Ac7+ 60 �b5 �c8 61 a6 �b8.

6

57 . . . axb6 58 a6

5

58 ax b6 �d7. 58

. . .

3

Ab8 59 c!>b5 Aa7 a

b

c

d

2

e

8

8

L.....-

_______...._ .::... ----I

7

68

Df7+ �e5 69 Dh7 Da2 70 fth3 c!>e4 71 �g1 Ae3+ 72 �1 g5 73 Dg3 ftf2+ 74 �g1 Db2+ 75 �1 �4 76 Dg2 Db1+ 77 c!>e2 ftb2+ 78 �1 ftb1 + 79 c!>e2 c!>e4 80 Ag8 ftb2+ 81 �1 ftb1+ 82 c!>e2 ftb2+ 83 �1 Db3 84 c!>e2 fta3 85 Af7

6 5

5

4

4

3 2

L.....-

_______...._ .::... ----I

Better defensive chances were offered by 85 Ae6 Af4 86 �f2 gf3+ 87 �g1 .

lf

Now Black w ill be forced to g ive u p his bishop for the a-pawn , but it turns out that Wh ite cannot keep his last pawn on f3 . 59 . . . �d6? 60 �x b6 �xd5 61 �b7 .

60 c!>c6 Or 60 f4 f6 61 �c6 b5 62 �b7 Ae3 63 Ab3 �d6 64 Ac2 �e6 65 a7 Axa7 66 �xa7 �d5 67 Axf5 �d4 68 Ab1 f5 ! ! 69 Axf5 �e3.

60 . . . b5 61 c!>b7 Ae3 62 f4 62 a7 Axa7 63 �xa7 �d6 64 Axf7 �e5 65 �b6 �f4 66 Ad5 b4 67 �c5 b3.

62

. • .

lf

� 63 a7 Y2-Y2

240

85

. . .

.lf4 86 .le6 a

b

c

...

d

8

8

7

7

6

6

5

5

4

4

3

3

2

2

tt:J

Endings

86 . . . Ag3!

Pete r L e ko - B o r i s G e l fa n d D o rt m u n d 1 99 6

Shutting the rook out of the game.

87 Af7 g4 88 Ah5 De3+ 89 �d2 Ae1 + 90 �d1 g3 0-1 Black is threatening . . . Af2 , thus winning at least the d5 pawn (91 ge2 Af2).

B o r i s G e l fa n d - J o e l L a u t i e r B i e l 2001 8

8

7

7

7

7

6

6

5

5

4

4

3

3

2

2

"if

69 �b4? Th is mistake leads to a very instructive end ing. After 69 �f4 �xf4 70 Axf4 �xf4 71 �xe6 e3 Black wins, but Leko, tired i n the seventh h o u r o f play after a long d e fence, does not notice the simple draw after 69 �e1 e3 70 Ac1 .

3

69 . . . e3!

2

---"'

L...-_______

.....

__

An im portant intermediate move.

"if

70 Ae1

White is a pawn down and his chances i n a long fight are slim. But he can save the game by perpetual check.

70 �xd5 exd2 71 �c3 Ac4 72 c6 �f4 73 �c5 Aa6 74 �d4 �f3 (74 . . . f5 ? 75 c7 �f3 76 c8� ! Axc8 77 �d3 =) 75 c7 �f2 76 �e4 Ac8 77 �d3 �e1 -+.

70 . . . �xb4 71 Axb4 Ac4

44 Af5! �g7 44 . . . g xf5 45 �f7+ Ag7 46 �xf5+ �g8 47 �d5+ is similar.

71 . . . Ac8 !?

72 Ac3 Necessary, i n order to prevent 72 . . . �e4 fol lowed by the advance of the f-pawn. If 72 �e7 �g5 73 c6 f5 74 c7 Aa6 75 �e6 �f4 (75 . . . f4 ?! 76 �e5 �g4 77 �d4) 76 �d5 (76 Ac5 �e4 77 Axe3 Ac8+) 76 . . . �f3 77 �e5 Ac8 78 �d4 �f2 , or 72 c6 �e4 73 c7 Aa6.

Axg6!

I nsisting on the capture of the bishop. 45 . . . �xg6 46

8

L..._ .______...._ .::;... ---I

4

45

8

"e6+ A16 47 "g8+ Ag7

47 . . . �h6 48 �f8+ Ag7 49 �f4+.

72 . . . �g5

48 "e6+ �h7 49 "f5+ �g8 50 "d5+ �8 51 "d8+ � 52 "d7+ 1/2-1/2

Also i nteresting was the suggestion of Roberto Cifuentes : 72 . . . e2 (threatening

241

Endings

73 . . . 't!?e4) 73 't!?e7 (73 c6 't!?e4) 73 . . . 't!?g6 74 't!?d6!? (74 c6 f5 75 c7 Aa6 76 't!?e6 't!?g5 77 't!?e5 't!?g4) , and now 74 . . . 't!?g5 ! (74 . . . Aa6 75 't!?d5 't!?f5 76 c6, or 74 . . . f5 75 't!?e5 't!?g5 76 't!?d4 =) 75 c6 't!?f4.

73 c6 Wh ite has nothing better. H is king is tem porari ly cut off.

73

8

8

7

7

6

6

5

5

4

4

3

3

2

2

. • •

83 Ae1 + 't!?g2 84 't!?e4 f3 85 't!?e3 Ab7 86 Ah4 't!?f1 87 c8� Axc8 88 't!?xf3) 82 Ac5+ 't!?d3 (82 . . . 't!?d2 83 't!?xf4) 83 Af2 f3 84 't!?f4 = ; B) 77 . . . f4 78 't!?e5 Ab5 79 c7 Aa6 80 't!?f5 (or 80 Aa3 Ab7 81 Ab4 e2 82 't!?f5 , simi lar to the note to Wh ite's 75th move) 80 . . . Ac8+ 81 't!?e5 't!?g3 82 't!?e4 Ab7+ 83 't!?f5 't!?f3 84 't!?e5 't!?g4 85 't!?d4 e2 86 Ad2 't!?g3 87 Ae1 + 't!?g2 88 Ah4 't!?f1 89 c8� Axc8 90 't!?e4, drawing.

74

f5?

When I checked my oid commentary I found out that the end ing was more d ifficult than I thought i n 1 996. Black could have won by 73 . . . Aa6 74 't!?d5 f5 75 't!?d4 't!?f4 ! 76 Ab4 't!?f3.

75

A) 7 7 . . . e 2 78 Ad2 A h 3 79 c7 f4 80 't!?e5 Ac8 81 Ab4 't!?e3 (81 . . . 't!?g4 82 't!?d4 't!?g3

.ta6 75 .td4

According to analysis by Roberto Cifuentes, 75 't!?d4 't!?f4 76 Ab4 't!?f3 77 c7 e2 (77 . . . f4 78 't!?e5 =, or 77 . . . 't!?e2 78 't!?e5 Ac8 79 't!?f4 =) 78 Ae1 't!?g2 79 't!?e3 't!?f1 80 't!?d2 leads to a draw. However, as Alexander Rustemov no ticed in our jOint analysis, Black can reach this position with Wh ite to move and win this ending : 80 . . . Ab7 ! ! (80 . . . f4 81 Ah4 leads to a mutual zugzwang position which will con tin ually turn up at the end of most of our l i nes ; 81 . . . Ac8 82 't!?d3 Ab7 83 c8� Axc8 84 't!?e4 =) 81 Ah4 (81 Ag3 f4 82 Ah4 Aa6, or 82 Ae1 f3) 81 . . . Ac8 ! 82 Ag3 f4 83 Ah4 Aa6 (zugzwang) 84 Ae1 (84 c8� Axc8 85 't!?d3 Ab7) 84 . . . f3 .

74 c!>e5? Return ing the favou r. Stronger was 74 't!?c5 ! (hitting the bishop) 74 . . . Af1 (74 . . . Aa6 al lows the study-like 75 't!?b6 Ac8 76 't!?c7 Ae6 77 't!?d6 Ac4 78 't!?c5 , return i ng to the same position) 75 't!?d4 (Wh ite provokes Black's pawn i nto advancing to f4, where it cann ot be protected by the bishop, and because of his poorly placed bishop on f1 , Black can not prevent this) 75 . . . 't!?f4 (an instructive l i ne is 75 . . . f4 76 c7 Aa6 77 't!?e4 't!?g4 78 Ab4 Ab7+ 79 't!?d4 't!?h3 80 Ae1 't!?g2 81 't!?d3 't!?f1 82 Ah4 e2 83 c8� ! Axc8 84 't!?e4) 76 Ab2 !! (76 c7 ? Aa6 77 Ab2 't!?f3 78 't!?e5 Ac8 -+) 76 . . . 't!?f3 77 Ac1 , and now :

. . •

• • .

e2 76 .tc3 c!>g4 77 .te1

77 c7 Ac8 ! 78 't!?d4 't!?f3 79 Ae1 transposes i nto the game.

77

. . •

.tc8 78 c!>d4 f3 79 c7

After 79 't!?d3 B lack wins i n simi lar fash ion to the note to Wh ite's 75th move : 79 . . . Aa6+ 80 't!?d2 't!?g2 81 c7 't!?f1 82 Ag3 Ab7 ! (82 . . . f4 ? 83 Ah4) 83 Ah4 (83 Ae1 f4 84 Ah4 Aa6) 83 . . . Ac8 (83 . . . f4 ? 84 't!?d3 e1 � 85 Axe1 't!?xe1 86 c8� Axc8 87 't!?e4) 84 Ae1 f4 85 Ah4 Aa6.

79

. . •

c!>g2 80 c!>e3 f1 81 .th4

81 't!?d2 f4 82 Ah4 Aa6 with zugzwang .

242

81

• . •

e1 1t+ 0-1

Endings

B o r i s G e l fa n d - Ko n st a n t i n L e r n e r U S S R C h a m p i o n s h i p Se m i - F i n al , N o r i l s k 1 9 87

Pave l M a r t i n ov - B o r i s G e lfa n d USSR J u n ior Championsh ip, Yu r m a l a 1 9 85

8

1--.....""'-

7 6 5

t-'-=-o

4

2

I n this im portant game from the USSR Junior Championship I had to seal my next move.

47 . . . g5! It was tem pting to go for the pawn end ing with an equal n u m ber of pawns, but after 47 . . J:!a7 ? 48 gxa7 ct1xa7 49 ct1f2 ct1b6 50 ct1f3 ct1c6 51 ct1f4 ct1d6 52 ct1f5 ct1e7 53 ct1g6 ct1f8 54 ct1h7 ! (Wh ite prevents the opponent's king from hiding i n the corner; after 54 g4 ct1g8 55 h4 ct1h8 56 ct1f7 ct1h7 57 g5 h xg5 58 h xg5 ct1h8 he has nothing better than to g ive stalemate with 59 g6) 54 . . . ct1f7 55 g4 ct1f6 56 h4 h5 57 g5+ ct1f7 58 ct1h8! Wh ite wins the g7 pawn and the game.

43 . . . gb7?

Following Tarrasch's advice regarding plac ing the rook beh ind a passed pawn i n rook endings, but this position is an instructive exception . 43 . . . gf6+ ! 44 ct1e5 gf2 45 gxg7 b3 46 g4 b2 47 gb7 ct1c8.

44 gg4!! +Thanks to the strong position of his king and d6 pawn , Wh ite fi nds a way for his rook to reach the 8t h ran k. 44 gb3 ct1e8 =.

44 . . �e8 .

44 . . . b3 45 ga4 gb8 46 gf4. 45

45 . . . ct1d8 46 gc5 b3 47 ga5 gb8 48 gf5.

48 gf6+ �c7 49 gxh6 �d7 50 gf6 50 gh5 ga5 51 h4 ga1 + 52 ct1f2 g x h4 53 gxh4 ct1e6 54 gf4 is an easy d raw, ac cord ing to theory.

46 gc7 b3 47 gxg7 �8 48 gf7+ �g8 49 d7! Securing the good position of the king. 49 gf1 ge8+ ! 50 ct1d5 ge2 .

50 . . . ga5 51 �2 �e7 52 gf3 �e6 53 g4 ga2+ 54 �g3 ga1 55 h3 gb1 56 ga3 56 gf5 is met by 56 . . . g b3+ 57 ct1g2 g b2+ 58 ct1f1 gb3.

56

••.

gc4 gb8

�

And White's further attempts to win were pointless (y2-Y2).

243

49 . . . b2 50 gf1 �g7 (see next diagram)

Endings

a

b

c

d

e

f

9

ctJxc1 67 h7 ctJd2 68 ctJg7 and Black has no defence.

h

8

8

7

7

6

6

In home analysis I managed to find a pretty way to save a half point. 61 . . . ctJd4 ? 62 gf2 ! transposes i nto the previous note.

5 4

4

3

3

2

2

L...-___________----'

62 Df2 62 g4 gxe2 63 h8ff c2 = and White is unable to prevent the manoeuvre of the rook to d2 and d1 .

1f

62 . . . ctJb31

51 Db1 1 Avoiding the last trap : 51 ctJe7 ?? b1 ff 52 gxb1 gx b1 53 d8ff ge1 + = .

The right way ! Black h ides his king from the checks.

51 .. J;;l b6+ 52 ctJe7 gb7 53 ctJea 1-0

63 g4 Black is just in time after 63 ctJg7 c2 64 gxc2 ctJxc2 65 g4 ctJd3 66 g5 ctJe4 67 g6 ctJf5.

M a n u e l R i va s - B o r i s G e lfa n d Dos H e r m a n as 1 994

63

• •.

gxf2 64 haW a

8

8

7

7

6

6

5

5

4

4

3

3

2

2

b

c

...

d

8

8

7

7

6

6

5

5

4

4

3

3

2

2

� 1f

L...-______________�____

Here White had to seal his move. Confident of an easy win , he quickly sealed

61 h7? thus throwing away the win. He could have achieved it with the subtle 61 gf2 !! , for ex ample: A) 61 . . . ctJe4 62 h7 ctJd3 63 gf3+ ctJd2 64 gh3 gxg2+ (64 . . . gxh3 65 g x h3 c2 66 h8ff c1 ff 67 ffh6+) 65 ctJh5 c2 66 h8ff c1 ff 67 ffh6+ ; B) 61 . . . ctJc4 62 gf4+ ctJd3 (62 . . . ctJb3 63 g4) 63 g4 c2 64 gf1 ctJd2 65 g5 c1 ff 66 gxc1

64 . . . gb21 With the threat of . . . c3-c2 and . . . gb1 , which White is unable to prevent. Black did not want to suffer in the infamous ending with queen and b-pawn v. queen after 64 . . . c2 65 ffb8+ ctJa2 (65 . . . ctJc4 66 ffc7+ ctJd3 67 ffd6+ ctJe2 68 'fWc5 +-) 66 'fWa7+ ctJb1 67 ffb6+ ctJa1 68 ffxf2 c1 ff . I n a practical game it is easy to lose such an ending, even though it is theoretically drawn.

244

Endings

65 "h3 Or 65 �b8+
65 . . . �h1 66 ge2 UeS 67 �g4 �g1 68 m3 ggS! Cutting off the white king from the h-pawn .

65. . . �b4 66 "hS �b3 67 "gS+ �c2

69 ge1 + �h2 70 ge2+ �h1 7 1 m4

Paradoxical ly, Black blocks the advance of his c-pawn, but he has time to regroup.

71 ge1 + gg1 72 gxe3 gg3+ .

71

68 �h5

. • .

gg3 72 gb2 �g1 73 ga2 m1 0-1

68 �c4 gb3 69 g5
68 . . . �b1 69 g5 c2 70 "h7 %-%

Boris Gelfand - Adrian Mikhalchishin O l y m p i ad , I stan b u l 2 0 0 0 a

J o e l L a u t i e r - B o r i s G e l fa n d Dos H er m a n as 1 9 9 5 8

'-----

7

b

c

d

e

f

9

h

8

8

7

7

6

6

5

5 4

6 5

3

3

4

2

2

3 a 2

b

c

d

e

f

2

9

h

1f

White is a pawn up, but it is far from clear if he can win this position .

49 gf4

This ending is of theoretical importance.

63

• • •

ge4!

Black would have thrown away the win by 63 . . .
64 m3 Black's task is simple after 64 gg8 gg4 65 ga8
64

. . .

An attem pt to cut off the king. 49
49 . . . gh3 Black has to waste a few tempi to create counterplay. If 49 . . . h5 50 g5 gd3 51
50 a5 ga3

�h2!

This forced but strong intermed iate move is the point of the whole ending.

Bad is 50 . . . gxh4 51
51 Uf5 �h7 52 gb5 ga4

65 gg2+ 65 gg8 e2 .

52 . . . �g8 !? 53 h5.

245

Endings

53 cc!?t3 �g8 54 h5 tMl? Allowi ng White to i m prove position of his rook, which turns out to be decisive. Black should have got rid of his weak g7 pawn by 54 . . . g5 ! , as suggested by M i khalchish i n : A) 5 5 �b8+ �f7 5 6 �b7+ (56 �a8 �f6 57 a6 �g7) 56 . . . �f6 57 � b6+ �e5 58 a6 �a3+ 59 �e2 �f4 60 �xh6 �xg4 61 �d2 �h4 = ; B ) 55 � b 6 �xa5 5 6 � x h 6 �f7 5 7 � d 6 �e7 (57 . . . �e5 !?) 58 �d4 �f6 ;!;; with good drawing chances. For a long time I thought that 54 . . . g6?! was even stronger, but on checking the analysis for this book I discovered a study-l i ke wi n :

C) 55 �b8+ ! (cutting off the king on the 8th ran k) 55 . . . �g7 56 �b7+ , and now : C 1 1 ) 56 . . . �g8 57 �a7 g x h5 58 g x h5 �f8 59 a6 �e8, when : C 1 1 1 ) it is premature to li berate the oppo nent's king with 60 �a8+ ? �f7 61 �e3 �g7 (61 . . . �f6 62 �g8, or 61 . . . �a5 62 �d4 �xh5 63 a7 �a5 64 �h8 �xa7 65 �h7+) 62 �d3 �a5 63 �c4 �xh5 64 �b8 (64 �b4 �h1 65 �c8 � b1 + 66 �a5 �a1 + 67 �b6 �b1 + 68 �a7 �g6) 64 . . . �a5 65 � b7+ �g6 66 � b6+ �g7 67 �b4 �a1 68 � b5 h5 69 �c6 �g6 70 � b7+ �g5 71 a7 �xa7+ 72 �xa7 h4 = ; C 1 1 2) 6 0 �e3 ! and White is winning : C 1 1 2 1 ) 60 . . . �f8 61 �d3 �e8 62 �a8+ (62 �c3 !? �d8 63 �b3 �a1 64 �c4 �c8 65 �a8+ �c7 66 a7) 62 . . . �f7 (62 . . . �d7 63 a7 �c7 64 �h8 �xa7 65 �h7+ �b6 66 �xa7 �xa7 67 �e4 and Black is just too late to reach f8 with his ki ng) 63 �c3 �g7 64 �b3 �a1 65 �b4 +-;

A) 5 5 �e3 leads only to a draw : A1 ) 55 . . . �xg4? 56 �b8+ �g7 57 � b7+ �f6 (57 . . . �g8 58 a6 �a4 59 a7) 58 � b6+ �e5 59 �xg6 ; A2) 55 . . . g5 is simi lar to 54 . . . g5; A3) 55 . . . g x h5 56 g x h5 �f7 57 �d3 �e6 58 �c3 (58 � b6+ �f5) 58 . . . �d6 59 � b3 �a1 60 � b4 � b1 + 61 �c4 �c1 + (61 . . . �a1 ? 62 �b6+ �c7 63 �b5) 62 �d4 �a1 63 �b6+ �c7 64 �xh6 �xa5 65 �h8 � b7 66 h6 �a6 ! = . An i m portant ru le - Black can save the game if he manages to attack the pawn from the side. On the other hand, if the black rook attacks the pawn from behind, this end ing is lost; B) 55 h xg6 �g7 56 �g3 �xg6 57 �h4 �a1 58 �b6+ �g7 59 �a6 �h1 + 60 �g3 h5 61 g x h5 �x h5;

C 1 1 22) 60 . . . �d8 61 �a8+ ! (at j ust the right time; 61 �d3 �c8 62 �c3 �a5 63 �a8+ �c7 64 a7 � b7 65 �h8 �xa7 66 �xh6 is a draw, as we already know) 61 . . . �c7 (the king is too late returning to the kingside after 61 . . . �e7 62 a7 �f7 63 �h8 �xa7 64 �h7+) 62 �f3 ! (improving the position of the king) 62 . . . �a5 (62 . . . �h4 63 �h8 � b6 64 �x h6+, or 62 . . . �b6 63 �h8 �xa6 64 �a8+ �b5 65 �xa4 �xa4 66 �e4) 63 �g4 �g5+ 64 �h4 �a5 65 a7 �b7 66 �h8 �a4+ 67 �g3 �xa7 68 �xh6 �b7 69 �g6 +-; C 1 2) 56 . . . �f8 !? (the king hastens to at tack the opponent's rook) 57 �a7 (57 g5?! g x h5 58 g x h6 �g8 59 �a7 h4 =) 57 . . . g x h5 58 g x h5 �e8 59 a6 �d8 60 �a8+ �c7 61 a7 (61 �g3 �a5 62 a7 �b7 63 �h8 �xa7 64 �xh6 � b5 =) 61 . . . �b7 62 �h8 �xa7 63 �xh6 � b7 64 �h7+ ! (a strong i ntermed iate move; Black has gai ned a tempo com pared with the previ ous l i ne, and now 64 �g6 �h4 ! 65 h6 �c7

246

Endings

66 �g3 gh1 67 �f4 �d7 = allows him to es cape, or 64 gh8 gc4 ! and he succeeds i n ach ieving t h e desired position o f his rook) 64 . . . �c6 (64 . . . �b8 65 �g3 followed by h5h6 and march of the king to g8) 65 h6 (the king prevents Black from attacking the pawn from the side) 65 . . . gh4 66 gh8.

55 gb7+ 55 g5 h xg5 56 gxg5 �f6 57 g b5 g5 (57 . . . �e6 58 �e3 �d6 59 gg5 �c6 60 �d3 ga1 61 gg6+ ±) 58 h6 (58 g b6+ �f5 59 a6 g4+) 58 . . . �g6 59 g b6+ �h7 60 a6 gf4+ 61 �g3 ga4 62 gd6 gf4 leads on ly to a draw.

55 . . . �6 56 gb6+ �e5 56 . . . �f7 57 a6 and after gg6 the king is free to march across to support the a-pawn .

and I was not sure if this ending was won or drawn. However, there is a study-like win by 71 �d4 ! (71 gh8 �c6 72 h6 �b7 73 �d4 gg6 ! =) 71 . . . �a5 72 �e4 ! (72 gh8 �b6 73 h6 � b7 74 gfB gh5 75 gf6 �c7 76 �e4 �d7 77 �f4 �e7) 72 . . . gb5 73 gh8, and now : A) 73 . . . �a6 74 h6 and Black is u nable to achieve the desired position : 74 . . . gb4+ (74 . . . gb6 75 h7 gh6 76 ga8+) 75 �d5 gb5+ 76 �c4 gh5 77 �d4 ; B) 73 . . . g b4+ 74 �d5 gb5+ 75 �c6 g b6+ 76 c!lc7 and although Black's rook is cor rectly placed , his king is not, and he cannot save this ending.

63 �b2 ga5 64 gxg7 gxa6 65 gg6 Winning a second pawn.

65 . . . ga4 66 gxh6 �d5 67 gg6 1-0

57 a6 lla3+ 58 �e2 �d4 59 Ild6+ �e4 60 �d2 �e5 61 gg6 �d5 62 �c2 a

b

c

d

e

f

9

•

h

8

8

7

7

B o r i s G e l fa n d - V l a d i m i r K ra m n i k Can d i d ates Q u arter- F i n a l (6) , Sang h i N aga r 1 9 94 a

5 4

4

3

3

2

2

b

c

d

e

8

8

7

7 6

5 4 3 2

The king is aiming to break across to the pawn , but the opponent's pieces try to stop it.

'---_______...._ .;;.... --I

62 . . . �c4

'lJ

61 �h3

During the game I was worried about 62 . . . �c5 ! 63 �b2 ga4 64 �b3 ga1 65 gxg7 gxa6 66 gg6 gd6 ! 67 �c3 (White cannot win the pawn ending after 67 gxd6 �xd6 68 g5 �e6 69 g6 �f6) 67 . . . gd4 ! (67 . . . �d5 68 g5, and if 68 . . . h xg5 69 gxd6+ �xd6 70 h6) 68 g5 (here too 68 gc6+ �xc6 69 �xd4 �d6 = is insufficient) 68 . . . gd5 69 ga6 � b5 70 gxh6 gxg5. I calcu lated the line up to here

Here the game was adjourned and I sealed this move. During home analysis together with my seconds Alexander Huzman and Valery Atlas, our first i m pression was that Wh ite's position was completely won . But gradually this was replaced by a more balanced assessment, according to which B lack maintains good practical chances of saving the game by sacrificing his central

247

Endings

pawn and trying to exploit the exposed pos ition of the wh ite king. The final verdict, that White is winning, kept eluding me both be fore and after the resumption.

61 . . .•f6! The only move that preserves Black's chances of defending successful ly. Other contin uations lose : A) 61 . . . �h7 62 �b6 ! (transposing i nto the rook end ing is Wh ite's main idea) 62 . . . �x b6 (62 . . . �g8 63 �xd6 gxd6 64 gc7 gf6 65 g4 g6 66 h xg6 fxg6 67 ge6 gdf8 68 gee7 +-) 63 axb6 gb7 64 gc7 gd7 (64 . . . gxc7 65 bxc7 gc8 66 ge7 f6 67 gd7 a5 68 gxd5 gxc7 69 gxa5) 65 gc8 gx b6 66 gee8 g6 67 gh8+ �g7 68 gcg8+ �f6 69 fxg6 +-; B) 61 . . . �f8 62 �c3 �f6 (or 62 . . . �b8 63 gc6 �b1 64 gc8) 63 gc6 gd6 (63 . . . �g5 64 f6) 64 �c5 �g8 65 gxd6 �xd6 66 ge8+ �h7 67 �xd6 gxd6 68 gb8 +-.

62 gcxd5 After the game Kramnik adm itted that in his opinion 62 f4 !? would have posed Black the hardest problems, lead ing to a queen end ing after 62 . . . gd6 63 gc xd5 gxd5 64 gxd5 gxd5 65 �xd5 ± .

During our h o m e analysis w e were i n itially inclined to go for this option, but I was afraid that, despite the two extra pawns, there wou ld be no easy way for White to solve 'the problem of his king ' , which constantly has the sword of Damocles hoveri ng over

its head due to the existing pawn configu ratio n . Also, earl ier that year I had fai led to win superior queen endings against M ichael Adams and Alexander Bel iavsky, and for psycholog ical reasons I decided to avoid the queen end ing now. Later I overcame this problem after winning a queen ending in a game with Jul ian Hodgson (see p. 250). Instead of the text move, 62 gexd5 could have been tried . After 62 . . . �g5 63 g4 Black has a choice : A) 63 . . . �f4 ? ! 64 gxd7 gxd7 65 gc8+ �h7 66 �c3 ! (even stronger than 66 gc4, provoking 66 . . . ge7 (66 . . . gd8 !? ; 66 . . . gxd4 67 gxd4 �xd4 68 �xf7) 67 d5 �d2 68 �c2 ! +-) 66 . . . gxd4 (66 . . . �xd4 67 �xd4 gxd4 68 gc6) 67 �c6 and Black can not de fend against both �g6+ ! and �a8 ; B) 63 . . . gxd5 64 gxd5 gxd5 65 �xd5 �e3 (65 . . . �c1 66 �a8+ �h7 67 �e4) 66 �a8+ �h7 67 �e4 �f2 68 f6+ , or 64 . . . gf8 ± ob tain ing a simi lar position to th e game.

62 . . .•c6! An unpleasant surprise ! In our home analysis we had paid most attention to the tem pting 62 . . . �g5 , after which we came up with a convi ncing way to convert Wh ite's advan tag e : 63 gxd7 gxd7 (63 . . . �x h5+ ? 64 �g2 gxd7 65 g4 �g5 66 ge8+ �h7 67 �b8 g6 68 �e5 +-) 64 f6 ! (opening lines for the wh ite pieces) 64 . . . �xf6 (64 . . . �c1 65 ge8+ �h7 66 �d3+ g6 67 h xg6+ fxg6 68 �g2 ! �g5 69 �b3) 65 �b8+ � h7 (65 . . . gd8 66 ge8+ gxe8 67 �xe8+ �h7 68 �e4+ �g8 69 d5 �d8 70 �e5 �xa5 71 �e8+ �h7 72 �xf7 �d2 73 �e6 �d1 74 �g4 ! +-) 66 �b1 + g6 (a big concession ; now the black king will be exposed) 67 �e4 �g7 68 h xg6 �xg6 69 gd5 ! gxd5 (69 . . . �xe4 70 fxe4 gxd5 71 exd5 and the pawn end ing is hopeless) 70 �xd5 �d3 71 �g4 and Wh ite should be able to win this queen end ing. We also thorough ly examined the continu ation 62 . . . gxd5 63 gxd5 ge8 64 ge5 gd8

248

Endings

which makes real ising the advantage consid erably more difficult. Here Wh ite can choose between 65 'tlfe3 'tlfc6 66 'tlfe4 'tlfa4 67 gd5 ge8 68 'tlff4 'tlfd1 69 f6, weakening the oppo nent's ki ng, and the subtle 65 'tlfa4 !? �f8 (65 . . . � h7 66 'tlfc4) 66 'tlfc2 ! ? , provoking 66 . . . 'tlfg5 ? 67 f6 ? 'tlfxf6 68 'tlfh7 g5 69 h xg6 and wins.

63 Ilxd7 .xd7 64 .b6! �h7 65 g4 An undesirable weakening of the king's pos itio n , but I cou l d n 't find any other way of freeing the rook from havi ng to defend the f5 pawn.

65

. . .

f6

65 . . . 'tlfxd4 66 'tlfxd4 gxd4 67 ge7 leads to a lost rook end ing.

Probably the best practical chance was the original 67 d5!? By sacrificing a pawn , Wh ite takes control o f th e g1 -a7 diagonal (the queen is going to protect his king) and launches an attack: A) 67 . . . gxd5 68 ge7 , when White wins after 68 . . . gxf5 ? 69 g xf5 'tlfxf5+ 70 �g3 'tlfg5+ 71 �f2 'tlfh4+ 72 �e2 , 68 . . . 'tlfc1 69 gxg7+ �xg7 70 'tlfb7+ �f8 71 'tlfxd5 'tlfh1 + 72 �g3 'tlfg1 + 73 �f4 'tlfc1 + 74 �e4 'tlfe1 + 75 �d4 'tlfd2+ 76 �c5 'tlfxa5+ 77 �c6, 68 . . . 'tlfc4 69 'tlfa7 �h8 70 �g3, or 68 . . . gd3 69 'tlfxf6 gxf3+ 70 �h4, and only 68 . . . 'tlff8 ! 69 g b7 gd3 leaves Black with counterplay; B) 67 . . . 'tlfc3 ! (it is best to avoid the ' G reek gift') 68 ge3 'tlfc1 69 ge7 'tlff4 = .

66 Ile4 .ca!

67

The best defence, found by Kram nik after a forty m i n ute think. After 66 . . . 'tlfd5 67 'tlfxa6 ga8 68 'tlfe6 'tlfxa5 69 'tlff7 'tlfa1 (69 . . . 'tlfb5 !? 70 �h2 'tlfb2+ 71 �g3 'tlfb8+ 72 f4 'tlfb5, keeping the rooks on, is probably better) 70 'tlfg6+ �h8 71 ge8+ Wh ite has every chance of winning.

67 Ile7 a

b

c

..

d

8

8

7

7 6

5 4

4

3

3

2

2

•c1 !!

A splend id resource, which enables Kram nik to save the half point. 67 . . . 'tlfc3 68 �g3 gxd4 (68 . . . 'tlfc1 69 d5!) 69 'tlfb7 �h8 70 ge8+ �h7 71 'tlfb8 +-.

68 d5! Making it harder for Black to d raw. After 68 'tlfxd8 'tlfh1 + 69 �g3 'tlfxf3+ ! 70 �h4 'tlfxg4+ the game ends with an elegant stale mate, w h i l e after 6 8 gxg7+ �xg7 6 9 'tlfxd8 'tlfh1 + 70 �g3 'tlfg1 + 71 �f4 'tlfc1 + 72 �e4 'tlfc6+ ! Wh ite is forced to block the exit for his king with 73 'tlfd5 'tlfc2+ and it all ends in perpetual check. 68

The natural continuatio n , creating the i l l u s io n o f an easy win for Wh ite. T h e post mortem analysis showed that 67 'tlfc5 !? would not have posed Black many problems after 67 . . . 'tlfb7 (pin n i ng the rook) 68 'tlfe7 'tlfb3 69 ge3 'tlfd5.

. . .

. . .

•14

After 68 . . . gxd5 69 gxg7+ ! (even stronger is 69 'tlfa7 ! 'tlfg5 70 'tlfa8 gd8 71 gxg7+ �xg7 72 'tlfxd8, but not 71 'tlfxd8? 'tlfxg4+ 72 fxg4, again with stalemate) 69 . . . �xg7 70 'tlfb7+ �h8 71 'tlfxd5 +- the white king avoids the pursuit.

69 �g2 The outcome of the game is not changed by 69 'tlfb3 gb8 70 'tlfe3 'tlfd6 71 'tlfa7 gg8 72 'tlfb7 'tlff4 73 d6 'tlfxd6 74 ge6 'tlfd1 =.

249

Endings ...

66 . . . lWh3+ (66 . . . h3 67 lWf8+ c;!?h5 68 lWh8+ c;!?g4 69 lWc8+ c;!?xf4 70 lWc7+ leads to per petual check; if 66 . . . c;!?h5 67 lWd1 + lWg4 68 lWd5+ lWf5 69 lWf3+ c;!?h6 70 c;!?g2) 67 c;!?f2 c;!?h5 68 lWe5+ lWf5 69 lWxf5+ g xf5 70 c;!?f3 and White saves the pawn ending.

4

66 . . .• h3+

3

It is necessary first to 'unstalemate' the op ponent's kin g . If 66 . . . c;!?h5 67 lWe5+ c;!?g4 68 lWe6+ c;!?xf4 69 lWe5+ with a draw.

2

c

d

e

f

9

h

67 �g1 �h5 68 .d4

69 . . . flc8! The final blow ! By sacrificing a third pawn, Kramnik gets his pieces to coord inate won derfu lly, and despite his material advantage Wh ite has to be satisfied with a d raw. If 69 . . J �b8 70 lWxf6.

70 .xa6 flc3 71 .e2 .c1 72 .f2 .d1 73 fle1 .xd5 74 fla1 74 a6 ,!;.a3 75 a7 lWa5 also leads to a draw.

74 . . . fld3 75 fla2 flxf3 76 .xf3 Y2-¥2 After 76 . . . lWxa2+ 77 c;!?h3 lWxa5 Black re gains all his sacrificed material .

J u l i a n H o d g s o n - B o r i s G e l fa n d Koop Tj u c h e m , G ro n i n g e n 1 9 9 6 ... 8

L-�

6 5 4 3

2

2

A) 70 . . . lWe3+ 71 c;!?f1 (71 c;!?h2 lWg3+ 72 c;!?h1 lWh3+ 73 c;!?g1 lWf5 74 c;!?f2 c;!?g4 75 c;!?e3 c;!?g3 -+) 71 . . . h3 (71 . . . lWe4 72 lWd1+ c;!?h6 73 lWg4 =) 72 lWd5+ c;!?h6 (72 . . . c;!?g4 73 lWf5+ !! again with stalemate) 73 lWg5+ c;!?h7 (73 . . . c;!?g7 74 lWh4 lWf3+ 75 c;!?e1 lWh1 + 76 c;!?e2 lWg2+ 77 c;!?e3 c;!?f7 78 lWh7+ with a d raw) 74 lWg4 ! (not 74 lWh4+ c;!?g7, when White loses because of zugzwang) and Black cannot make progress ; B) 70 . . . lWe2 ! (Black needs to play sub tly) 71 lWf6 lWe1 + (71 . . . lWg4+ 72 c;!?f2 lWf5 73 lWxf5+ gxf5 74 c;!?f3 is the same draw that we have already seen) 72 c;!?g2 lWg3+ 73 c;!?f1 lWd3+ 74 c;!?f2 lWd5 + , and I don't see how White can save the f4 pawn and the game.

7

3

After this White loses his last pawn. Even one tem po down, he should have tried to activate his king : 68 c;!?f2 ! (68 lWd2 lWf5) 68 . . . lWg3+ 69 c;!?f1 (69 c;!?e2 g5) 69 . . . lWf3+ (it is not yet the time for 69 . . . g5 - 70 lWd1 +) 70 c;!?g1 (70 c;!?e1 lWe4+ 71 c;!?f2 c;!?g4) , and now :

68 . . .•f5 -+

Now there can be no doubt about the out come of the game.

65

. . .

•g3+ 66 �h1 ?

The king should try to get out of the corner. After 66 c;!?f1 ! Black cannot active his king :

250

69 �g2 �g4 70 .d1 + �xf4 71 .d2+ �g4 72 .d1 + �g5 73 .d8+ �h5 74 .h8+ �g5 75 .d8+ .16 76 .d2+

Endings

�h5 77 Oe2+ �h6 78 �h3 Of4 79 �g2 Og3+ 80 �h1 Oh3+ 81 �g1 Of5 82 �h2 g5 83 Oe8 016 84 Oc8 Of4+ 85 �h1 �h5 86 Oe8+ �g4 87 Oe6+ Of5 88 Oc4+ �g3 89 Oc7+ Of4 0-1

White would be able to save the game after 55 . . . .l;d1 + 56 c,!>h2 c,!>f2 57 .l;xe3 c,!>xe3 58 a7 .l;d8 59 a8tW .l;xa8 60 eDxa8 f4 61 eDc7 ! = .

56 a7 gd3 57 a80+ �g3 a

b

c

d

e

f

9

h 8

8

P re d r a g N i ko l i c - B o r i s G e lfa n d C an d i d ates M atc h , S araj evo 1 991 a

b

c

d

e

f

9

6

h

5 8

4

6 5

5

4

4

3

2

a

r-'-=-.

2

c

d

e

f

9

h

As White has no checks, he has to g ive u p his q ueen for t h e rook, a n d then t h e knight ending is lost.

2

L...-_______-:-._----I

b

A L...-_______-:-._----I V

'lt

58

Wh ite's only try is to d ivert the opponent's rook.

53 gb3 �g2+ First i m proving position of the knight (53 . . . .l;xb3 54 a7) .

Oa1 Ild1 + 59 Oxd1 �xd1 60 �xe6 g6 61 b4 f4! 62 �c5

62 eDd4 eDc3 63 b5 f3 64 �f1 eDd5 65 eDc2 g5 -+ , or 62 eDg5 eDc3 63 �f1 h6 64 eDf7 (64 eDe6 eDd5 -+) 64 . . . eDd5 65 eDx h6 eDe3+ 66 �e2 f3+ 67 �xe3 f2 -+.

54 �1 �e3+ 55 �g1 gxb3!!

***

251

62 . . . �c3 63 �d3 g5 64 �f2 h5 65 �1 �h2 0-1

Ap p end ix I n d ex of P l aye rs A

Acs, Peter 211 , 213 Adams, M ichael . 1 3, 14, 59, 60, 62 , 98, 1 21 , 1 25, 204, 248 Agrest, Evgeny 1 25 Akopian , Vladimir 69 Aleksandrov, Alexey 67 Alterman , Boris 50 Anand, Aruna 185 Anand, Viswanathan 1 2 , 13, 15, 27, 42 , 62 , 101 , 110, 1 1 2 , 113, 155, 185, 187, 188, 1 91 , 1 97, 1 98 Atlas, Valery 67, 1 22, 133, 247 Averbakh , Yuri 10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

D Damljanovic, Branko 98, 99 Dolmatov, Sergey 119 Dominguez, Len ier 208 Dorfman , lossif 17 Dreev, Alexey 13, 1 7, 5 0, 5 9, 69 , 177 . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. .

F Faulks, N ick Flohr, Salo Freeman , N igel Ftacn ik, Lubomir

213 12 213 21 , 24

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

G Gelfand, Abram Gelfand , Boris Gelfand , Nella Geller, Yefim Georgiev, Kiri l Glek, Igor Golovey, Marietta Golovey, Tamara Golu bev, M i khail Grischuk, Alexander Gross, Stefan Gulko, Boris Gurevich , M i khail

8, 10-12 7-16, 19, 75 10 11 , 12 88, 92 , 234 70, 159 12 11 215, 216 206, 21 8-221 17 30, 55 17, 36, 200, 230

. . . . . . . . . . . . . . . . . . . . . . . . .

B Bareev, Evgeny Beliavsky, Alexander Benjam in, Joel Beyl i n , M i khail Blauert, Jorg Bohm , Hans Boleslavsky, Isaac Bologan , Viorel Botvinnik, l Iya Botvinnik, M i khail Bronstei n , David Bruzon, Lazaro Burshtein, Igor

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . .

16, 50, 59, 148 11 , 73 , 86, 248 1 93 10 218 147 11 3 6 , 37, 69 209 94 10 208, 209 218

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

H

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Hazai , Laszlo H ickl , Jorg Hodgson, Jul ian Hubner, Robert H uzman , Alexander 152, 181 , 247

209 , 210 164 248 121 13, 16, 47, 1 22, 133,

. . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

C Capablanca, Jose Chandler, M urray Chebanenko, Vyacheslav Chern i n , Alexander Chiburdan idze, Maya Cifuentes, Roberto

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 83, 208 81 113 1 7, 18, 26, 39, 42 211 241 , 242

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

Iliescas, M iguel

252

. . . . . . . . . . . . . . . . . . . . . . . . . . .

36, 155

Appendix

Index of Players

•

Ivanchuk, Vasily 1 2 , 14, 1 7, 26, 27, 30, 31 , 50, 51 , 63, 67, 69 , 86 , 98, 121 , 1 72 . . . .

K Kamsky, Gata . . . . 1 2 , 14, 27, 34, 37, 151 Kapengut, Al bert . 11 , 1 2 , 1 9, 20, 64, 67, 209 Karpov, Anatoly . 1 2-14, 16, 38, 59, 70, 71 , 80, 97, 1 51-153, 1 72, 1 95, 209 , 219, 220 9, 1 2-14, 16, 17, 30, 70, Kasparov, Garry 73-77, 79 , 80, 101 , 113, 139, 1 72 Khalifman , Alexander . . . 13, 15, 1 7, 50, 51 Khen kin, Igor . . . . . . . . . . . . . 50 Kholmov, Ratmir . . . . . . . . . . 52 , 54, 55 Kindermann , Stefan . . . . . . . . . . . . . . . . . 34 Kogan , Mark . . . . . . . . . . . . . 133 Kok, Bessel . . . . . . . . . . . . . . . 94 Korch noi, Victor . . . . . 14, 153, 154, 1 93, 203 Kram nik, Vladimir 7, 13, 14, 16, 27, 31 , 41 , 98, 133-135, 137, 148, 155, 1 63, 164, 1 72 , 187, 1 97, 2 1 3 , 248-250 Kuzm i n , Gennady . . . . . . . . 211 . . . . . . .

Novikov, Igor . . . . . . . . . . . . . Nu n n , John . .. ...... . . .

.

. .

. . . . . . . . . . . . . .

011,

.

. 21 . 1 2 , 74

. . . . . . . . . . . . .

.

. .

. .

. . .

. . .

.

.

o

Lembit Olms, Manfred Olpiner, Zohara

. .............. . 50 .. . ....... .. 154 . . . . . . . . . 207

. . . . . . . . . . . . . . . . .

. . .

.

.

. .

.

.

. . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

.

.

.

. .

. . . . .

. . . .

.

. . . . . . . . . . . . . .

.

.

. .

. . .

.

.

. . .

.

.

.

. .

.

.

.

.

. . . . . . . . .

.

.

.

. .

.

. .

.

.

.

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . .

. .

. . . .

. . . . . . . . . . . . .

. . .

.

. . .

.

.

. . .

.

.

.

. .

. .

. . . . .

. . . . .

. . . . . . .

. . .

.

.

. . . .

.

.

.

.

. . . . . . . . . . . . . . . .

.

. . .

Pedersen , Steffen . . . . . . . . . . . . . 218 Pein , Malcolm . . . . . . . ....... .. . 19 Pelletier, Yannick . . . . . . . . . . 151 , 1 77 Petrosian , Tigran . . . . . . . 1 1 , 1 29 Piket, Jeroen . 14, 80, 81 , 145, 147, 159, 204 Pillsbury, Harry Nelson . . . . . . . . . . . . . . 205 Poldauf, Dirk .... ........ . .. .. . 16 Polgar, Judit .. . ... . . ... . . 14, 35 Polugayevsky, Lev . . . . . . . . . 101 Portisch , Lajos . . . . . . . . . . . 70, 181 Postovsky, Boris . . . . . . . . 16 Psakhis, Lev . . . . . . . . . . . . . . . . . . . . . 90, 211 . . . . . . .

. . .

.

.

. . . .

. .

. . . .

.

.

.

. .

.

. . .

. . . . . . .

M Makarychev, Sergey . . . . . . . . . . . . . . . . 89 M i khalchishi n , Adrian . . . . . . . . . . . 246 M i les, Anthony . . . . . . . . . . . . 10 Morozevich , Alexander . . 16, 200, 204, 205 Movsesian , Sergey . . . . . 1 76 . .

. . . .

.

. . . . . . . . . .

. . . . . . . . . . . . . . . . .

.

. .

.

.

.

.

.

.

.

.

. . . . . . . . . . . . . . .

. . . .

.

. .

.

.

. . . .

.

.

. . .

. . .

.

.

.

.

.

. . .

. . .

.

. .

.

.

.

.

. .

.

.

.

. . . . . . . . .

.

.

.

. .

. . .

. .

. .

N Najdorf, Miguel . . . . . . . . 210 Neat, Ken . . . . . . . . . . . . 1 95 N i kolic, Predrag . . . . . . . . . . . . . . . . . . . 1 2 , 14 N imzowitsch, Aaron . . . . 65, 152 Noteboom , Dan iel . . . . . . . . . 145 Notki n , Maxim . . . . . . . . 234 .

.

.

.

. . .

. .

.

.

. . . . . . . . . . . . . . . . . . .

. .

.

. . . . . . . . .

.

.

. .

. . . . . . . . . . . . . . .

.

. .

.

.

.

. . . .

. . . . . . . .

. . .

. .

. . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . .

.

. . . . . .

. . . . . .

. . . . .

. . . . . .

. . . . . . . .

.

.

.

.

.

.

. .

.

.

R Razuvaev, Yuri . . . . . . . . . 11 , 15, 85 , 106, 155 Rentero, Luis . . . . . . . . . . . . . . . . 73 , 93 Romero Holmes, Alfonso . . . . . . 68 Rozentalis, Eduardas . . . . . . . . . . . . . . . . 94, 95 Rubi nstein , Akiba . 15, 46, 85 , 110, 1 69, 1 80, 188 Rublevsky, Sergey . . . . . . . . . . 46, 1 63, 167 Rustemov, Alexander .. . . . 242 .

.

.

. . . . . . .

. . .

.

.

. .

. . . . . . . . . . .

.

.

.

.

. .

.

.

.

. . .

.

. .

.

. . .

. . . .

.

S Sagalch ik, Gennady . . . . . . . . . . . . 67 Sakaev, Konstantin . . . . . . . . . . . . . . . . . . 34 Salov, Valery . . . . . . . . 1 2-14, 73 , 150, 200 Savchen ko, Stan islav . . 67 Semen iuk, Alexander . . . . . . . . . . . . . . . 31 Serper, Grigory . . . . . . . 50, 1 72 Shabalov, Alexander 50, 139, 213-215, 228, 229 Shi pov, Sergey . . . . . . . . 1 73, 183, 188, 189 Shir�Y, Alexey 14, 15, 41 , 42 , 44, 45, 50, 69, 98, 103, 109, 113, 114, 1 1 7, 139, 141 , 142 , 165, 1 69 Short, N igel . . . . . . . . . . . . . . . . 1 2 , 83 , 94, 97 .

. .

. . . .

. . .

. .

.

. .

253

. . .

. .

. .

.

.

.

.

.

. .

.

. .

.

. . .

.

. . . . . . . . . . . . . . . . . . . .

.

. . . . .

. .

. . . . . . . . . . . . . . . . .

.

. .

.

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

.

.

. . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . .

.

. . . . . . . .

.

L Lanka, Zigurds . . . . . . . . . . . . . . . . . . . 117 Lautier, Joel . . . 9 8 , 1 72 , 1 73, 1 75, 200 Leko, Peter . 13, 14, 16, 38, 39, 98, 241 Lj ubojevic, Ljubomir . . . . . . . . 14 Lobron, Eric . . . . . . . . . . . . . . 55, 57, 58 Lukacs, Peter . . . . . . . . . . . 1 95, 209, 210 .

p

. .

.

.

.

. . . . .

. .

.

. . .

. .

. .

Appendix

Shulman, Yuri Smiri n , l Iya Sokolov, Ivan Sosonko, Genna Spassky, Boris Stein itz, Wilhelm Stoh l , Igor Sutovsky, Emil Svidler, Peter

•

Index of Players

39, 145 1 7, 50, 63, 67, 84 42 1 92, 1 97 95 74 21 159, 1 60 13, 15, 213, 224, 240

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

Topalov, Vesel i n Tseitl i n , Mark

. . .

14, 16, 1 29, 131 , 1 53, 159 26

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

V Van Wely, Loek Vitolins, Alvis

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 95 213

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . .

y

Ye J iangchuan Yudasin, Leonid Yusupov, Artur

181 139, 140 1 2 , 89, 90, 103

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

T Tal , M i khai l Tarrasch , Siegbert Ti mman , Jan Timoshchenko, Gennady

. . . . . . . . . . . . . . . . . . . . . . . . . . .

11 , 139, 213 243 1 2 , 14, 97 186

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . .

Z

. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

Zelkind, Edward

254

. . . . . . . . . . . . . . . . . . . . . . . . . . . .

10, 11

Appendix

•

Index of Games

I n d ex of G a m e s A page number in bold indicates that the first-named player had the black pieces. Underlining indicates a memorable game. A page number in brackets means a particular game was referred to on that page. A page number ital icised means a game fragment (a combination or an endgame position). A

Acs, Peter- Gelfand 21 0 Adams, Michael - Gelfand 59, (1 21 ), 1 21 , 1 25, 1 25, 208, 214, 222 Alexi kov, Alexander-Shariyazdanov 1 70 Anand, Viswanathan - Fritz 203 - Gelfand 63, 1 1 0, 1 56, 1 85, 1 87 - i l iescas 42 - Kramnik 27 - Krasen kow 1 98 Andersson , U lf- Gelfand 64 Andrianov, N i kolay-Vaisser 1 9 Aseev, Konstantin - Gelfand 1 86 Averbakh , Yuri - Spassky 95

B Balashov, Yuri - Gelfand 148, (1 84), 1 85 Bareev, Evgeny - Gelfand 1 1 1 - Kramnik 111 Basin, Leonid - Gelfand 1 7, (1 52) , 1 54 Beliavsky, Alexander- Gelfand 70 , 87, 233 - Kallai 114 - Shirov 114 -Tkachiev 208 Benjam i n , Joel - Gelfand 1 93 Blauert, J6rg - Pedersen 21 9 Botvinnik, l Iya- Gelfand 208 Bren ninkmeijer, Joris- Gelfand 70 Bruzon, Lazaro - Gelfand 208 C Chandler, Murray- Gelfand 80 Chern in, Alexander- Horvath 42 Chiburdanidze, Maya- Dvoirys 214 o

Damljanovic, Bran ko - Gelfand 98 Daniliuk, Sergey- Solozhen kin 36 Dautov, Rustem - Gelfand 64

Deep Blue- Kasparov 59 Delchev, Alexander-Gelfand 1 97 Dim itrov, Vladimir- Komarov 1 70 Dizdar, Goran - Schandorff 114 Dlugy, Maxim - Gelfand 70 Dorfman , lossif- Gelfand 1 7 Dunnington, Angus- Gelfand 223 Dvoirys, Semen - Chiburdan idze 214 - Rozentalis 94

E Ehlvest, Jaan - Gelfand 71 , 214 - Kasparov 204 F Fressinet, Laurent - Gelfand 204 Fritz-Anand 203 Ftacnik, Lubom ir-Gelfand 21 - Gel ler 74 - N unn 74 - Sakaev 21 G Gavri kov, Viktor- Gurevich 1 7 Gelfand, Boris-Acs 210 -Adams 59, (121 ), 121 , 1 25, 1 25, 208, 214, 222 -Anand 63, 110, 156, 185, 187 -Andersson 64 -Aseev 186 - Balashov 148, (1 84) , 185 - Bareev 111 - Basin 1 7, (1 52), 154 - Beliavsky 70, 87, 233 - Benjamin 1 93 - Botvi nnik 208 - Brenninkmeijer 70 - Bruzon 208 - Chandler 80 - Damljanovic 98 -

255

--

--

Appendix

Index of Games

•

Gelfand, Boris- Dautov 64 - Delchev 1 97 - Dl ugy 70 - Dorfman 1 7 - Dunnington 223 - Ehlvest 71 , 214 - Fressinet 204 - Ftacnik 21 - Georgiev 86, 1 76, 234 - Gheorghiu 70, (1 63) - G iorgadze 22 7 - G reenfeld 1 92 -Grischuk 203, 218, 218 - Gunawan 70 - G u revich (1 72) , 200, 229 - Hodgson (248) , 250 - I l iescas 155 - Ionov 155 - Istratescu 1 81 - Itkis 1 7 - Ivanchuk (26), 26, 2 7, 3 0 , 31 , (35) , 36, 47, 67 , (74) - Kamsky 34, 34, 35 - Karpov 1 51 , 209 - Kasparov 73 , 74, 83 - Ki nderman n 34 - Korchnoi 1 54, 238 - Kramn i k 133, (1 63), 1 64, 1 65, 1 70, 1 87, 24 7 - Lautier 1 72, 1 72, 1 73, (20 0) , 232, 24 1 , 245 - Leko 38, 24 1 - Lerner 243 - Lev 50 - Lj ubojevic 224 - Lobron 55 - Log inov 47 - Lutz 74, 233 - Macieja 28, 156 - Malaniuk 73 - Malisauskas 62 - Markowski 1 69, 239 - Martinov 243 - M ikhalch ish in 245 - M ilov 1 92, 208, 236 - Morozevich 203 -

--

Gelfand, Boris - Motylev 225 - Movsesian 1 77, 223, 239 - N ijboer 74 - N ikolic 251 - Pavlov 222 - Pelletier 151 - Pig usov 64 - Piket 81 , 145, 192, 233 - Polgar 35 - Pol ugayevsky 1 01 - Portisch 226 - Rivas 244 - Romero Hol mes 67 - Rublevsky 1 63 - Sagalch ik 231 - Salov 148, 224 - San Segundo 23 7 - Savchenko (67), 68 - Schebler 1 82 - Semen iuk 30 - Shabalov 21 3, 22 7 -Shirov 41 , 46, 1 03, 1 05, 1 1 1 , 113, 1 1 7, 1 1 7, 1 1 8, 1 39, 1 64 , 225 - Short 83, 94, (203) - Smirin 63, 85, 231 - Stoh l 24 - Sutovsky 159 - Svidler 224 , 240 -Timoshchen ko 1 86 -Tkachiev 1 64 -Topalov 56, 64, 74 , 1 29, 159, 1 60, 204, 236 - U libin (134), 138 -Vaisser 146 -Van der Sterren 232 -Van Wely 1 95, 208 -Yakubenya 222 -Ye Jiangchuan 180 -Yudasin 1 1 7, 1 39 -Yusu pov (88) , 92 , 1 03, (154) Geller, Yefim - Ftacnik 74 - Polugayevsky 214 - Spassky 99 Georgiev, Kiri l - Gelfand 86, 1 76, 234 - Sakaev 1 77 Georgiev, Krum - M natsakanian 214

256

-

-

-

--

Appendix

•

Index

Gheorghiu, Flori n - Gelfand 70, (1 63) Giorgadze, Giorg i - Gelfand 227 Gipslis, Aivars - Petrosian 1 04 Gonzalez, Juan - Perez 20 Greenfeld , Alon - Gelfand 1 92 Grischuk, Alexander- Gelfand 203 , 218, 21 8 Gufeld , Eduard - Kochiev 205 Gunawan , Ronny - Gelfand 70 Gurevich , M i khai l - Gavri kov 1 7 - Gelfand (172) , 200, 229 H

Hal kias, Stel ios - O'Cinneide 21 8 Hodgson, Julian - Gelfand (248), 250 - Sadler 1 25 Horvath, Jozsef- Chern in 42 Howell, James- Mack 83 H ubner, Robert - Pieper-Emden 114

Iliescas, M iguel -Anand 42 - Gelfand 1 55 lonov, Sergey - Gelfand 1 55 Istratescu , Andrei - Gelfand 1 81 Isu pov, Vladimir- Kiriakov 1 61 Itkis, Boris - Gelfand 1 7 Ivanch uk, Vasily- Gelfand (26) , 26, 27, 30, 31 , (35), 36, 47, 67, (74) - Rokhmanov 1 21 -Topalov 1 72 Ivanov, Sergey- Svensson 18 Ivanov, Victor- Shulman 1 45

K Kal lai , Gabor- Beliavsky 1 1 4 Kamsky, Gata- Gelfand 34, 34, 35 Karpov, Anatoly- Gelfand 1 51 , 209 - Kasparov 73 , 1 21 - Kramnik 200 Kasparov, Garry- Deep Blue 59 - Ehlvest 204 - Gelfand 73, 74, 83 - Karpov 73, 1 2 1 - Kramnik 27 - Polugayevsky 1 01

of

Games

Kasparov, Garry-The World 164 -Vaganian 92 Kau la, Roman - Krasenkow 1 70 Keres, Pau l - Korchnoi 130, 1 98 Khal ifman , Alexander- Leko 36 - Sveshn i kov 106 Kindermann , Stefan - Gelfand 34 Kiriakov, Petr- Isupov 161 Kirov, Nino-Tosh kov 1 8 Koch iev, Alexander- Gufeld 205 Komarov, Dimitri - Dim itrov 170 Korchnoi , Victor- Gelfand 1 54, 238 - Keres 1 30, 1 98 - Shirov 114 Kotronias, Vasil ios - Rozentalis 95 Kouatly, Bachar-Touzane 1 9 Kram nik, Vladimir-Anand 27 - Bareev 1 1 1 - Gelfand 1 33, (1 63) , 164, 165, 1 70, 1 87, 247 - Karpov 200 - Kasparov 27 -Topalov 1 77 Krasenkow, M ichal -Anand 1 98 - Kaula 170 Kuzm i n , Gennady - Liptay 211

L Lautier, Joel - Gelfand 1 72, 1 72, 1 73, (200) , 232, 24 1 , 245 Lautier, Joel - Leko 38 - Sokolov 42 Leko, Peter- Gelfand 38, 24 1 - Khal ifman 36 - Lautier 38 -Vaisser 38 Lerner, Konstanti n - Gelfand 243 - Sorokin 149 Lev, Ronen - Gelfand 50 Li ptay, Laszlo - Kuzm in 21 1 Ljubojevic, Lj ubomir- Gelfand 224 Lobron, Eric - Gelfand 55 Log inov, Valery- Gelfand 47 Lputian , 5m bat - Polgar 1 87 Lutz, Christopher-Gelfand 74, 233

257

Appendix

•

M Macieja, Bartlomiej - Gelfand 28, 1 56 Mack, Andrew- Howel l 83 Malani uk, Vladimir- Gelfand 73 Malisauskas, Vidmantas- Gelfand 62 Markowski , Tomasz - Gelfand 1 69, 239 Martinov, Pavel - Gelfand 243 Mednis, Edmar-Short 21 0 M ichaelsen , Nils- Pein 1 9 M i khalchish i n , Adrian - Gelfand 245 M i lov, Leon id - Gelfand 236 M i lov, Vadi m - Gelfand 1 92, 208 Mnatsakan ian , Eduard - Georgiev 214 Morozevich , Alexander- Gelfand 203 Motylev, Alexander-Gelfand 225 Movsesian , Sergey- Gelfand 1 77, 223, 239 N Nadanian , Ashot - Strelbin 28 N ij boer, Friso - Gelfand 74 N i kolic, Predrag - Gelfand 25 1 Novi kov, Igor-Tu kmakov 21 N u n n , Joh n - Ftacnik 74 o

O'Cinneide, Mel - Halkias 218 Olafsson , Fridrik- Sax 83 Oral , Tomas - Shu lman 40 p

Pavlov, Andrey - Gelfand 222 Pedersen, Steffen - Blauert 21 9 Peek, Marcel -Timmer 1 9 Pei n , Malcolm - M ichaelsen 1 9 -Vaisser 1 8 Pelletier, Yannick- Gelfand 1 51 Perez, Juan - Gonzalez 20 Petrosian , Tigran - Gipslis 1 04 Pieper-Emden , Carsten � H Obner 1 1 4 Pigusov, Evgeny- Gelfand 64 Piket, Jeroen - Gelfand 81 , 1 45, 1 92, 233 Polgar, Judit- Gelfand 35 - Lputian 1 87 Polugayevsky, Lev - Gelfand 1 01 - Geller 214 -

Index of Games

Polugayevsky, Lev - Kasparov 101 - Roman ishi n 220 Portisch , Lajos - Gelfand 226 -Xie Jun 1 81

R Rivas, Manuel - Gelfand 244 Rokhmanov, Sergey- Ivanchu k 1 21 Romanish i n , Oleg - Polugayevsky 220 Romero Holmes, Alfonso - Gelfand 67 Rozentalis, Eduardas - Ovoirys 94 - Kotronias 95 -Wah ls 95 Rublevsky, Sergey- Gelfand 163 S Sadler, Matthew- Hodgson 1 25 Sagalchik, Gennady- Gelfand 23 1 Sakaev, Konstantin - Ftacnik 21 - Georgiev 1 77 Salov, Valery - Gelfand 1 48, 224 San Seg undo, Pablo- Gelfand 237 - Shulman 1 45, (1 46) Savchen ko, Stan islav - Gelfand (67) , 68 Sax, Gyula- Olafsson 83 Schandorff, Lars- Oizdar 1 1 4 Schebler, Gerhard - Gelfand 1 82 Semeniuk, Alexander - Gelfand 30 Shabalov, Alexander- Gelfand 213, 227 Shariyazdanov, Andrey -Alexi kov 1 70 Shi pov, Sergey- Sokolov 42 Shirov, Alexey- Bel iavsky 1 1 4 - Gelfand 41 , 46, 1 03, 1 05, 1 1 1 , 1 1 3, 1 1 7, 1 1 7 , 1 1 8 , 139, 164, 225 - Korchnoi 1 1 4 Short, N igel - Gelfand 83 , 94, (203) - Mednis 210 Shulman, Yuri - Ivanov 145 - Oral 40 - San Seg undo 145, (146) Smiri n , l Iya - Gelfand 63, 85, 23 1 Sokolov, Ivan - Lautier 42 - Sh ipov 42 Solozhen ki n , Evgen iy- Oaniliuk 36 Soroki n , Maxim - Lerner 1 49 -

258

-

Appendix

Sosonko, Genna-Vandevoort 1 93 Spassky, Boris-Averbakh 95 - Geller 99 Stohl, Igor- Gelfand 24 Strelbi n - Nadan ian 28 Sutovsky, Em i l - Gelfand 1 59 Svensson , Bengt- Ivanov 1 8 Sveshni kov, Evgeny- Khal ifman 1 06 Svid ler, Peter- Gelfand 224, 240

T The World - Kasparov 1 64 Ti mmer, Robert - Peek 1 9 Timoshchen ko, Gennady - Gelfand 1 86 Tkachiev, Vladislav- Beliavsky 208 - Gelfand 164 Topalov, Veselin - Gelfand 56, 64, 74, 1 29, 1 59, 1 60, 204, 236 - Ivanchuk 172 - Kram nik 1 77 Tosh kov, Ti homir- Kirov 1 8 Touzane, Ol ivier- Kouatly 1 9 Tu kmakov, Vladimir- Novi kov 21 U Ulibin, M i khai l - Gelfand (1 34), 1 38

•

I ndex of Games

V Vagan ian , Rafael - Kasparov 92 Vaisser, Anatoly-Andrianov 1 9 - Gelfand 1 46 - Leko 38 - Pein 1 8 Van der Sterren , Pau l - Gelfand 232 Van Wely, Loek-Gelfand 1 95, 208 Vandevoort, Pascal - Soson ko 1 93

W Wah ls, Matth ias- Rozentalis 95 X

Xie J u n - Portisch 1 81 Xu J u n - Yermolinsky 1 76 y

Yakubenya, Valery- Gelfand 222 Ye Jiangchuan - Gelfand 1 80 Yermolinsky, Alex-Xu Jun 1 76 Yudasin, Leon id - Gelfand 1 1 7, 139 Yusu pov, Artur- Gelfand (88) , 92, 1 03, (1 54)

259

Appendix

•

Index of Opponents

I n d ex of O p p o n e n ts A

Acs, Peter 210 Adams, Michael 59, 1 21 , 1 25 Anand, Viswanathan 110, 185

B Bruzon, Lazaro 208 C Chandler, M urray 80

L Lautier, Joel 172 Leko, Peter 38 Lev, Ronen 50 Lobron, Eric 55 Log inov, Valery 47 M Malisauskas, Vidmantas 62 Markowski, Tomasz 169 p

o

Damljanovic, Branko 98 Delchev, Alexander 197 Dorfman , lossif 17

F Ftacnik, Lubom ir 21

Piket, Jeroen 145, 192 Polugayevsky, Lev 101

R Rublevsky, Sergey 163

G Georgiev, Kiril 86, 176 Gheorghiu, Florin 70 Grischuk, Alexander 203, 218 Gurevich , M i khail 200

S Salov, Valery 148 Shabalov, Alexander 213 Shirov, Alexey 41 , 105, 1 1 3, 117, 139 Short, N igel 83, 94 Sutovsky, Emil 1 59

Ivanchuk, Vasily 26, 30, 67

T Topalov, Vesel i n 1 29

K Kamsky, Gata 34 Karpov, Anatoly 1 51 Kasparov, Garry 73 Korchnoi, Victor 1 54 Kram nik, Vladimir 133

V Van Wely, Loek 195 y

Ye Jiangchuan 180 Yusupov, Artur 103

260

Appendix ..

•

I ndex of Open ings

_..

I n d ex of O p e n i n g s Benoni Defence A43 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Bogo-Indian Defence E 1 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103, 154

Queen's Gambit 037 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 040 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 041 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 97

Caro-Kann Defence 8 1 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

Queen's Gambit Accepted 020 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 21 028 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148

Catalan Opening E04 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105, 208 E05 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 85

Queen's Indian Defence E 1 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 E 1 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

Dutch Defence A89 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 29 English Opening A07 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Grunfeld Defence 002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 079 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 95 085 . . . . . . . . . . . . . . . 17, 21 , 26, 30, 34, 38, 41 King's Indian Defence E81 70 E84 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 E94 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67, 159, 1 69 E97 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 80 E99 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Petroff Defence C42 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 , 203 Pirc-Ufimtsev Defence 807 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 25

Queen's Pawn Opening A41 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Ragozin Defence 038 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 92 Semi-Slav Defence 045 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 046 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 72 047 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Sicilian Defence 836 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 852 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 867 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210 890 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 892 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 896 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139, 213 Slav Defence 015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113, 1 76 031 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Tarrasch Defence 034 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218

261

My Most Memorable Games (gnv64).pdf

Recommend Documents