The Slav - Sadler

1 d4 d5 2 c4 c6

To my parents.

The Slav

CHESS PRESS OPENING GUIDES Other titles in this series include: 1 90 1 25 9 05 6

Caro-Kann Advance

Byron Jacobs

1 90 1 25 9 06 4

Closed Sicilian

Daniel King

Dutch Leningrad

Neil McDonald

1 90 1 25 9 0 3

X

1 90 1 25 9 1 0 2

French Advance

Tony Kosten

1 90 1 25 9 02 1

Scandinavian

John Emms

1 90 1 25 9 OS 0

Semi-Slav

Matthew Sadler

1 90 1 25 9 0 1 3 1 90 1 25 9 09 9

Sicilian T aimanov

James Plaskett

Trompowsky

Joe Gallagher

For further details for Chess Press titles, please write to The Chess Press c/o Everyman Chess, Gloucester Mansions, 1 40a Shaftesbury Avenue, London WC2H SHD.

Chess Press Opening Guides

The Slav

Matthew Sadler

ir [1Illj

The Chess Press, Brighton

First published 1 997 by The Chess Press, an imprint of First Rank Publishing, 23 Ditchling Rise, Brighton, East Sussex, BN1 4QL, in association with Everyman Books plc Reprinted with corrections 1 999 Copyright © 1 997 Matthew Sadler Distributed by Everyman Chess, Gloucester Mansions, 1 40a Shaftesbury Avenue, London WC2H SHD.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, electrostatic, magnetic tape, mechanical, photocopying, recording or otherwise, without prior permission in writing from the publishers . A CIP c atalogue record f o r t h i s b o o k i s available from the British Library ISBN 1 90 1 259 00 5

Cover design by Ray Shell Dcsign Printed and bound in Great l3ritain by l:3iddles Ltd, Guildford and King's Lynll

CONTENTS 1 d4 d5 2 c4 c6

Bibliography

8

Introduction

9

7

The Old Main Line (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 �fs 6 e3 e6 �xc4 �b4 8 0-0) : Black plays to prevent e4

13

7

The Old Main Line (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 �fS 6 e3 e6 �xc4 �b4 8 0-0) : Black allows e3-e4

27

The New Main Line (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 �fS 6 ttJeS) : Black fights for control of e4

39

The New Main Line (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 �fs 6 ttJeS) : Black counterattacks

53

5

The Smyslov Variation (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 ttJa6)

62

6

The Bronstein Variation (3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 �g4)

72

7

The 4 . . . a6 Slav (3 ttJf3 ttJf6 4 ttJc3 a6) : White plays 5 e3

87

8

The 4 . . . a6 Slav (3 ttJf3 ttJf6 4 ttJc3 a6) : Aggressive options for White

98

9

The Exchange Variation (3 cxds cxdS)

1 08

10

Move-Orders and Transpositions

1 18

11

Odds and Ends

130

2 3 4

Index of Complete Games

142

BIBllOGRAPsH Y

Books

Encyclopaedia o/ Chess Openings vol. D (ECO), Sahovski Informator 1987 Bats/ord Chess Openings 2 (BCO), Kasparov & Keene (Batsford 1989) Winning with the Slav, Schipkov & Markov (Batsford 1994) The Slav for the Tournament Player, Flear (Batsford 1988)

Periodicals

In/ormator ChessBase Magazine New In Chess Yearbook British Chess Magazine Chess Monthly

INfRODUCrlON

1 d4 d 5 2 c4 c6

The skill of preparing an opening is frequently misunderstood: many play­ ers (including some strong grand­ masters) believe that to play an open­ ing well, it is necessary to analyse a great many variations; that no prepa­ ration is complete without at least one queen sacrifice and that to stop before move 20 is akin to criminal negli­ gence. I know from experience that the sheer volume of opening theory can be overwhelming, and this is es­ pecially true for the non-professional player who has little time to keep up with the latest fashions. And yet, having been a professional player now for six years, I know that I have won more games from 'nor-

mal' openings than from any 30-move piece of analysis (and it's not because I haven't done any!) . The brilliant 'I had this position after move 80 on my board at home' games that we see in magazines are the exceptions: beauti­ ful, treasured by every chessplayer, but very, very rare. Chess is a sport and most games are a struggle, and we win games because we fight harder than our opponents, or because we understand the position better. In my opinion, opening prepara­ tion can be successfully reduced to three simple steps: 1. Knowing the main aim of our opemng. 2. Knowing the value of move­ orders. 3 . Understanding typical positions. Therefore, let's apply these ideas to the Slav. Opening Aims

With 2 c4, White challenges the black centre. The natural 2 . . . e6, allowing Black to develop his kingside pieces, has the drawback of blocking the light-squared bishop inside the pawn chain. 2 . . . c6 aims to hold the centre, to develop the light-squared bishop 9

Th e S l a v

outside the pawn chain, and then to play . . . e7-e6 and conclude the black development. However, the course of chess ideas, like love, never runs smoothly! Black must be careful when he develops his light-squared bishop: after 1 d4 ds 2 c4 c6 3 CDf3 CDf6 4 CDc3

that Black can only play a quick . . . �fs if he can successfully defend b7 with his queen. Thus, 1 d4 ds 2 c4 c6 3 CDf3 CDf6 4 e3 �fs

is fine for Black since S cxdS cxdS 6 'iYb3 can easily be met by 6 . . . 'iYc7; however 1 d4 ds 2 c4 c6 3 CDc3 CDf6 4 e3 �fS?! Black would like to play 4 . . . �fS , but he will have great difficulty defending b7 after S cxds cxds 6 'iYb3 !

Now 6 . . . 'iYb6 loses a pawn to 7 CDxds 'iYxb3 8 CDxf6+ exf6 9 axb3 and 6 . . . b6 weakens the queenside light­ squares too much: 7 e4! dxe4 8 CDeS e6 (to stop 'iYxf7+ mate) 9 �bs+ CDfd7 10 g4 �g6 1 1 h4! , intending h4-hS, trap­ ping the bishop. The general rule is 10

S cxds cxds 6 'iYb3 is not good, since 6 . . . 'iYc7 loses a pawn to 7 CDxdS . So how can Black carry out his main idea? Black either has to stop White from playing 'iYb3 , or he has to find a good way to defend b7. This is a typi­ cal opening dilemma: whether to pre­ vent an opponent's threat directly, or whether to arrange the pieces in such a way that the threat is nullified.

In t r o du c tio n

The main line of the Slav runs dS 2 c4 c6 3 tt'lf3 tt'lf6 4 tt'lc3 dxc4.

1

d4

First, Black wins a pawn and threatens . . . b7-b5, making this gain permanent. Second, the b3-square is cunningly taken away from the queen, which means that White can­ not attack b7, and hence that . . . �f5 becomes possible. While White recap­ tures the c4-pawn, Black will develop the light-squared bishop to f5 or g4 and will be looking to complete his kingside development: 5 a4 (surroun­ ding the c-pawn by preventing . . . b7bS) 5 . . . �f5 6 tt'le5 (intending tt'lxc4) or 6 e3 (intending �xc4) 6 . . . e6 are the main lines. So far I have been very enthusiastic about Black's strategy, but now I have to reveal the downside of his play. This sort of schizophrenia is necessary when you play both sides of the Slav, as I do! 4 . . . dxc4 relinquishes control of e4, which makes it easier for White to cramp Black with two central pawns on d4 and e4. But White must be care­ ful that his pawns do not become weaknesses as Black first immobilises, then attacks them. The bottom line is that the player who has the better

understanding of the line will get the best results with either colour. The second idea is to play 1 d4 d5 2 c4 c6 3 tt'lf3 tt'lf6 4 tt'lc3 a6.

The move 4 . . . a6 was first played in Britain by Grandmaster Jon Levitt, but it is Grandmaster Julian Hodgson who has upheld this variation at the highest level, and introduced the most significant ideas. The original idea of 1 d4 d5 2 c4 c6 3 tt'lf3 tt'lf6 4 tt'lc3 a6 was to meet iVb3 with . . . b7-b5, moving the pawn to a protected square. However, ways were found to exploit the drawbacks of . . . b7-b5: the dark-square weak­ nesses on c5, b6, as and the slightly exposed black queenside. Then, in a brilliant piece of unstereotyped think­ ing, Hodgson realised that . . . a7-a6 could allow the rook to defend b7 from a7. No one had dared to play such a strange move before, but Julian did, and this has made some previ­ ously dodgy lines completely viable. However, although avoiding . . . d5xc4 helps to prevent e2-e4, Black's position is less dynamic than in the . . . d5xc4 lines, as it is much harder to break against White's centre with 11

Th e Sla v

. . . c6-cS and . . . e7-eS. My own prefer­ ence as Black is for the 4 . . . dxc4 lines, as they are richer in content and offer a wider range of possibilities to suit many different styles. Move-Orders

Move-orders are a much underrated part of opening preparation. Oppo­ nents don't always play fair! Imagine the scene: you sit down to play, con­ fident that you know your opening at least as well as your opponent, and what happens? He plays the opening in some unusual move-order, and you emerge a bit dazed to find yourself playing a different line to the one you wanted! And unless you work out your move-orders thoroughly, this will continue to happen, time and time again.

12

S o how can this happen in the Slav? If you want to play the 4 . . . a6 Slav, then there is nothing that White can do to muddy the water, which is one of the attractions of this line. 4 . . . dxc4 lines, however, require some care. First, White can try to sidestep them by playing an early e2-e3, protecting c4, e.g. 3 !bc3 !bf6 4 e3 . Black has many reasonable moves here, but none of them fit in with the idea we want to play. The other way for White to play is c4xdS, leading to the Exchange variation; 4 . . . a6 Slav fans should study these positions particu­ larly carefully, since this sort of posi­ tion is very typical of this line, and there are many transpositions. Understanding Typical Positions

Well, for this part, read on . . .

CHAPTER ONE The Old Main Line: Black plays to prevent e3-e4

1 d4 d5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3 dxc4 5 a4 .§i.f5 6 e3 e6 7 .§i.xc4 .§i.b4

8 0-0

This traditional system of devel­ opment for White is especially popu­ lar at club level. White quietly recap­ tures the pawn and puts his king to safety before he starts his plan of e3e4, to obtain a 'perfect' pawn centre. This chapter examines Black's at­ tempts to prevent White from achiev­ ing this goal. First, we need to ask ourselves a few questions in order to establish our approach: Question 1: How will White try and achieve e3-e4? Answer: White has two major ap­ proaches: a) 'i'e2. This is the most dangerous idea, which we shall examine first. b) ctJh4, to remove the bishop on f5, which is helping Black to prevent e3-e4. Question 2: How can Black fight against 'i'e2 and e3-e4? A nswer: Black has three pieces at­ tacking the e4-square: the bishops on b4 and f5 and the knight on f6. When White plays 'i'e2 he is supporting the e3-e4 push with only two pieces: the

queen and the knight on c3. How­ ever, he will achieve the e3-e4 advance with tempo because the e-pawn at­ tacks the bishop on f5. If Black wants to, he can simply pre-empt this by re­ treating the bishop to g6, so that e3-e4 no longer attacks the bishop. Now if White plays e3-e4 regardless, Black can win a pawn by playing . . . �xc3 and . . . ctJxe4. Question 3: What move-order should I play this in? Answer: My own favourite has been to play 8 . . 0-0 9 'i'e2 �g6; 8 . . ctJbd7 9 'i'e2 �g6 is sharper since White can offer a dangerous pawn sacrifice. .

.

Gamel Richardson-Sadler

Islington Open 1995

This was a crucial game for me: I was leading the Islington Open by only half a point and only a win would guarantee first place. However, even more importantly, only a win would be good enough to pip Keith Arkell for the Leigh Grand Prix! 1 d4 d5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3 dxc4 5 a4 .§i.f5 6 e3 e6 7 .§i.xc4 .§i.b4 13

Th e Sla v

S 0-0 0-0 9 'iVe2 jL g 6 1 0 LDe5

10 e4? ! �xc3 ! 1 1 bxc3 tbxe4 wins a safe pawn. Consequently, White re­ verts to 'Plan B ' : he exchanges his knight on f3 for my bishop on g6 and removes an attacker of e4.

1 0 . . . LDbd7 1 1 LDxg6 hxg6 1 2 .lci.d 1

12 e4 tbb6! wins a pawn as 13 e5 (or 13 Mdl �xc3 14 bxc3 tbxc4 15 'i'xc4 tbxe4) fails to 13 . . . 'i'xd4 14 �xe6 fxe6 15 exf6 'i'xf6 with a clear extra pawn. 1 2 ...'iVe7!?

An interesting move, though I imagine that it is not the most accu­ rate - for Kramnik's 12 . . . 'i'a5, see the next game. 1 3 e4

ing t o prevent e3-e4? Well, smce White's aim was to push a pawn within his own territory, it was al­ ways unlikely that we could prevent it for ever. However, by resisting for as long as possible, we have forced White to make a concession, namely that he has had to play tbe5xg6 before being able to play e3-e4. Although White gains the two bishops with this manoeuvre, he exchanges off the piece that would be most affected by the e3e4 advance; on g6, the bishop has little scope if White can maintain his pawn on e4. Moreover, the departure of the knight from f3 means that White loses some control over the central dark squares, d4 and e5 . This last point is seen to great effect in this game. 1 3 . . . e5 1 4 d 5 nacS !

I spent a great deal of time at this stage and realised that I had to force White to release the tension in the centre and play d5xc6. The explana­ tion for this has a lot to do with the central dark squares: without the d­ pawn, I can transfer a knight to e6 (via c5) and exploit the outpost that my pawn on e5 creates on d4. By placing my rook on c8, I was hoping to get my opponent worried about possible threats on the c-file, in order to tempt him into playing d5xc6. 1 5 jL g 5 :!:;lfdS 1 6 dxc6?

Wait a minute! Wasn't Black play14

Here it is! After this mistake, White has to be very careful to avoid a dis­ advantage. The correct plan is ex­ tremely ingenious. Black has two ave­ nues of pressure: he has possible threats along the c-file and he can de­ velop some pressure against e4 by means of . . . tbc5. How can White deal

Th e O ld M a in L in e : B l a c k pla ys to pre v e n t e 3- e 4

with both these threats? With the manoeuvre 16 l':rd3 ! 4.Jc5 17 l':re3 ! ! On e3, the rook covers e4 and defends the knight on c3 along the rank, thus pro­ tecting White's queenside against c-file play. White is slightly better after 16 l':rd3, but the game is still very compli­ cated.

game, I was very impressed with White's attitude: realising that his po­ sition had worsened considerably, White regrouped and concentrated totally on defence. 1 9 iLxe6 �xe6 20 .!:Id 3 gxd 3 2 1 �xd 3 �b3 2 2 .t:!.b 1

l:!:b8 2 3 iL d 2

iLa 5 !

1 6 .� . bxc6 !

My opponent had underestimated this recapture. Although it weakens Black's queens ide pawn structure, Black protects the central light squares, over which he previously had little control due to the exchange of his light-squared bishop. This move is so strong, because White's queenside is so weak: the pawn on a4 gives Black a comfortable slot on b4 for the queen, from which it can attack the a4- and b2-pawns. 1 7 .!:Id 3 ? CLlc5 1 8 .!:Ih 3 ? CLle6 !

White's 17th and 1 8th moves were excessively optimistic as he had no chance of an attack along the h-file. Meanwhile, Black threatens . . . 4.Jd4. With his control of the d-file, and White's weakened queens ide as a clear target, I believe that Black can already be thinking about victory. During the

Protecting the d8-square, so that the rook can use either of the open files on the board. 24 f3 .!:Id8 25 �e2 CLle8 !

Black's knight is the least active of his piec€s. The text prepares to bring it to d4 via c7 and e6. When this hap­ pens, all of Black's troops will be on their optimal squares. 26 iLe 1 CLlc7 27 l:!:c 1 CLle6 28 �c2 �b6+ 29 �f2 CLld4

Intending 3o . . . ihc3 31 �xc3 4.Je2+ winning the exchange, as 32 '>jVxe2 is impossible since the queen is pinned to the king. 30 �h 1 �b3 31

�b 1 l:!:b8 3 2 h3

�c4 3 3 f4 exf4 34 �xf4 ge8 3 5 �d 1

iLxc3 3 6 iLxc3 CLle2 3 7 �f3

gxe4

The first weak pawn falls. 38 a5 a6 39 �d3 CLlxc3 40 �xc4 gxc4 41 bxc3 .!:!.c 5 ! 15

Th e Sla v

what both sides should be aiming for in this variation. ... . .

Game 2

Karpov-Kramnik

Monte Carlo (blindfold) 1995 1 d4 d 5 2 c4 c6 3 CZJf3 CZJf6 4 CZJc3 dxc4 5 a4 jLf5 6 e3 e6 7 jLxc4 3l.b4 8 0-0 0-0 9 �e2 jLg6 1 0 CZJe 5 CZJbd7 1 1 CZJxg 6 hxg6 1 2 .lld 1 �a 5 !

I think that the rook ending is winning now. 4 1 . . .Mxc3 42 Md8+ �h7 43 Ma8 would have regained the a­ pawn, but now after 42 Md8+ �h7 43 Ma8 , Black can simply play 43 . . . Mxa5. 42 .l:i.d8+ 'it> h 7 43 gc8 g 5 44 'it>g 1 'it>g6 45 'it>f2 �f5 46 �c7 f6 47 .l:i.xg 7 .l:i.xa 5 48 �e3 .l:!.e5+ 49 'it>d4 �f4 50 ga7 ge6 !

5 1 �b7

5 1 Mxa6 c5+! wins a rook. 5 1 . . .'it> g 3 5 2 .l:i.b2 .l:i.e 5 5 3 llb6 c5+ 54 'it>c4 f 5 5 5 gxa6 'it>xg 2 5 6 .llg 6 f4 57 h4 f3 58 .llx g 5+ .l:i.xg 5 59 hxg5 f2 60 �xc5 f 1 � 61

c4 �f5+ 6 2

'it>d6 � g 6 + 0 - 1

This was a very important game for me, and an instructive example of 16

This move i s more active, and probably more logical, than 12 ... 'iYe7. While Black is not threatening to win a pawn immediately with 13 . . . �xc3 due to 14 bxc3 'iYxc3 15 �d2 'iYc2 16 �d3 ! 'iYb2 17 Mdb 1 , winning the queen, it does prevent 13 e4, as with White's centre slightly weakened, Black can get away with taking the pawn: 13 . . . �xc3 14 bxc3 'iYxc3 1 5 �d2 'iYxd4 (15 . . . 'iYc2!?) 16 �b4 'iYe5 17 �xf8 Mxf8 when with two pawns for the exchange, Black stands very well. Note that 13 ctJa2 allows 13 . . . 'iYxa4 14 ctJxb4 'iYxa 1 15 ctJa2 (hoping to trap the queen) 15 . . . 'iYb 1 ! , escaping t o f5! 1 3 jLd2

Protecting c3, and intending e3-e4,

Th e O ld M a in L in e : Bla c k pla ys to pre v e n t e 3- e 4

but . .

3 3 �a2 �h2+ 34 \t>f 1 �xh3+ 3 5

1 3 . . . e5 1 4 d 51Iad8

W e 2 tLJ e 5 36 iL e 2 �xg4+ 3 7 W d 2

14 . . . cxds? 15 ctJxds ctJxds 16 �xds 'i'xds 17 �xb4 wins for White.

!!ed8+ 38 We 1 �xd 1 + 3 9 iLxd 1 �g 1

.

40 iLd2 tLJd3+ 4 1 <;t>e2 ldb2+ 0 - 1

1 5 dxe6 bxe6 1 6 iLe1 e4 !

Securing an outpost on d3 for the knight.

S o White's plan of 9 'iVe2 and 10 ctJes seems harmless. Let us take a look at the more direct 9 ctJh4. Game 3 Yusu pov-Kram n ik

Riga 1995 1 d4 d 5 2 e4 e6 3 tLJf3 tLJf6 4 tLJe3 dxe4 5 a4 iLf5 6 e3 e6 7 iLxe4 iL b4 8 0-0 0-0 9 tLJh4!

The most testing idea: White elimi­ nates the bishop on fs without wast­ ing time on 'iVe2. 1 7 iLb3 �e5 1 8 1Le2 iLd6 1 9 g 3 tLlc5 2 0 \t> g 2 �f 5 2 1 lId2 1L e 5 ! 22 gad 1 J::i: b 8 !

The white queenside is looking very weak. 23 iL b 1

iLxe3 24 bxe3 tLJxa4 25

ga2 tLJb2 26 U d 2 tLJe4 2 7 1::i: d 1 tLJe 5 28 h3 tLJf3 29 lIa4 tLJ g 5 30 g4 �e5 3 1 �e2 lIfe8 3 2 !!xa7 tLJf3

9 . . . tLJbd7 ! ?

A typical stratagem: Black's dou­ bled f-pawns will take over the bishop's task of preventing e3-e4. Question 4: What is wrong with 9 . �g6? Answer: White can play 10 ctJxg6 hxg6 1 1 'iVc2! Question 5: Why is it important to put the queen on c2 and not e2? Answer: First, the queen neutralises .

Threatening . . .'iVh2+. Now Black is just winning.

.

17

Th e Sla v

Black's most active plan of . . . 'iYa5, threatening . . . jl,xc3 . Second, with the queen on c2, Black must be careful that when he plays . . . e6-e5 he does not allow 'iYxg6! ( . . . e6-e5 has opened up the a2-g8 diagonal and the f7-pawn is now pinned to the king, so Black cannot recapture on g6) . Of course, we are dealing with subtle nuances rather than big differences, but it is important to understand them none­ theless.

Game 4 Ivanchuk-Bareev

Dortmund 1 995 1 tLlf3 d 5 2 d4 tLlf6 3 c4 c6 4 tLlc3 dxc4 5 a4 jLf5 6 e3 e6 7 jLxc4 jL b4 S 0-0 tLlbd7

1 0 tLlxf5 exf5 1 1 �c2 g 6 1 2 f3 �b6

Preventing e3-e4 by attacking the d4-pawn, which has been weakened by the absence of the white knight from £3 . 1 3 �h 1 �aeS 1 4 �f2 c5 1 5 �h4 .ld:cS ! ? 1 6 jLa2 �fdS 1 7 jLd2 tLlfS 1 S a 5 �a6 1 9 �fd 1 c4

Shutting out White's light-squared bishop. 20 jLe 1 .l:!.eS 2 1 e4 jLxc3 22 jLxc3 fxe4 23 d5 tLlSd7 24 �d4 Y2 - Y2

The draw was agreed in a very murky position. It is now time to consider the other move-order: 8 . . . tLlbd7, intending to meet 9 'iYe2 with 9 . . . jl,g6 as above. 18

Question G: What does Black gain from delaying castling? Answer: 8 . . . tLlbd7 is directed against the plan of an early tLlh4, which we saw in Yusupov-Kramnik. After 9 tLlh4, Black will reply 9 . . . jl,g6, as 10 tLlxg6 hxg6 is extremely dubious for White. Since Black has not castled, his rook is well placed on the semi-open h-file, pointing towards White's king! Black will play . . . 'iYc7 (attacking h2) , castle queenside and then double rooks on the h-file, which is not what White was hoping for when he sensi­ bly (he thought!) took the bishop pair! So what does White do after 8 . . . tLlbd7 9 tLlh4 jl,g6? Give up? Cry? Well, if he's a genius like Ivanchuk, he chooses a third option: he gets sneaky. 9 tLlh4 jLg6 1 0 jLe 2 ! ?

Th e O ld M a in L in e : Bla c k pla ys to pre v e n t e 3-e 4

White wants to take on g6 only once Black has castled; so he plays a useful consolidating move while he waits for Black to commit his king. The text prevents the bishop on g6 from escaping the knight's attentions by 10 . . . ]LhS!? The alternative waiting move, 10 h3 , is considered in the next game.

1 2 . J:lc8 ! ? .

10...0-0

Fans of tactics can investigate 10 . . . ]Lxc3 1 1 l2lxg6 (not 1 1 bxc3 l2lds 12 l2lxg6 l2lxc3! 13 iVc2 l2lxe2+ 14 'iYxe2 hxg6, winning a pawn) 1L.]Lxb2 12 l2lxh8 ]Lxa1 13 ]La3 or 13 iVc2 (unclear - Ivanchuk) and when they have, 1'd be grateful if they could tell me what is going on! How­ ever, more positional players can be happy with Bareev's move. Although White's queen will go straight to c2, the bishop is more passive on e2 than on c4: after . . . e6-eS, Black no longer has to fear iVxg6 (in fact he'd be quite pleased to see it!) as the white bishop is not on the a2-g8 diagonal; and this also means that White cannot reply so easily with d4-ds after . . . e6-eS or . . . c6c5. 11 ct:Jxg6 hxg6 1 2 'iVc2

Since the white queen i s o n the c2, Black tries to inconvenience it by opening the c-file. The immediate 12 . . . cS would be met by 13 l2la2 ! , net­ ting Black's other bishop since 13 . . . ]LaS loses a pawn to 14 dxcS, in­ tending b2-b4. Bareev therefore plays the rook to the c-file in order to facili­ tate . . . c6-cS . The . . . c6-cS break is played less often than . . . e6-eS in the Slav, but it is a typical idea that is well worth remembering. 1 3 e4 ! ?

Since White does not want the c-file to be opened, he prepares to meet 13 ... cS with 14 ds . If only he still had his bishop on c4! This move 19

Th e Sla v

introduces a sharp pawn sacrifice that is probably not quite good enough, so 13 Md1 was suggested by Ivanchuk as an alternative, when he claims a slight advantage for White. 1 3 . . . cS 14 ds exdS 15 ttJxds ttJxds 16 Mxds VlJie7 1 7 b3 ! i s indeed rather better for White. Black's main problems are the weak­ ness of his light squares and his bishop on b4, which is shut off from the rest of Black's pieces by the pawn on cS . The immediate 1 3 . . . VlJie7 is stronger: 14 e4 (also interesting is 14 b3!?, intending 14 ... cS [ 14 ... eS is more sensible] 15 dS iLxc3 16 d6! and VlJixc3 with advantage) 14 . . . cS 1 5 dS exdS 1 6 ttJxds ttJxds 1 7 exdS (17 Mxds loses the e-pawn to 17 . . . ttJf6)

ous play due to his threat of . . . Me2. 1 3 . . . c5 1 4 d 5 exd 5 1 5 exd 5 ! ? l:l:e8 ? !

The start of a series of slight inaccu­ racies that Ivanchuk exploits brutally. Is . . . iLxc3 16 bxc3 ttJxds 17 Md1 ttJ7f6! 1 8 iLf3 (18 c4 ttJb4! unpins) 1 8 . . . VlJie7! is Ivanchuk's recommenda­ tion, as 19 iLxdS ttJxdS 20 MxdS al­ lows mate after 20 . . . VlJiel . 1 6 IId 1 c4

1 7 d6!

A very strong move: the d6-pawn exposes the light-squared weaknesses in the black position by opening the hl-a8 diagonal and freeing dS for the knight on c3 . 1 7 . . . l:l:e6 1 8 i.f4 "i'b6 1 9 l2lb5 i.c5 20 i.g3!

1 7 . . . c4! , freeing the bishop and pre­ venting White from establishing a light-squared blockade of the queen­ side with b2-b3 and iLc4. After 1 8 iLxc4, Black can regain the pawn with 18 . . .tZJb6 19 b3 ttJxc4 (19 . . . ttJxdS 20 iLb2! [20 iLxds Mxc2; 20 Mxds VlJiel+!J leaves White more active due to his two raking bishops) 20 bxc4 VlJih4!, when 2 1 f4 (the only move to save the c-pawn) 2 1 . . .Mfe8 (22 . . . Mel+ is now a threat) 22 g3 VlJig4 gives Black danger20

Th e O ld M a in L in e : Bla c k pla ys to pre v e n t e 3- e 4

Calmly protecting f2. 20 'ctJe4 2 1 .liLg4 f5 22 .liLf3 ! <;t>h7 . .

23 CiJc7 ge5 24 tLJd 5 gxd 5 2 5 !!xd 5 lLlxg 3 2 6 a 5 1 -0

26 hxg3 would also have won. White is just the exchange up with a winning pOSltlOn. Game 5 T opalov-Gelfand

Belgrade 1995

1 d4 d5 2 c4 c6 3 tLJf3 tLJf6 4 tLJc3 dxc4 5 a4 .liLf 5 6 e3 e6 7 .liLxc4 .liLb4

Black's passive play has gIven White a huge space advantage.

8 0-0 tLJbd7 9 tLJh4 .liLg6 1 0 h 3 ! ?

1 9 . . . tLJd6 20 tLJxg 6 hxg6 2 1 .liLf4

Another waiting move. However, Black's bishop has a square!

tLJe8 22 \\Wf3 .liLd 6 23 .liLe3 tLJdf6 24

10 . . . .liLh 5 1 1 g4 tLJd 5 !

26.liLc2? !

Black makes a discovered attack by the queen on the knight on h4. Once the knight retreats, Black will again have g6 for his bishop.

26 dS! eS 27 liJe4 (Topalov) would have led to a clear edge for White.

12 tLJg 2 .liLg6 1 3 tLJa2 .liLe7 ? !

exd 5 3 2 tLJxd 5 tLJxd 5 3 3 .liLb3 <;t>fS

Topalov rightly suggests 1 3 . . . �d6 f3 hS ! as an improvement, striking immediately against the exposed king­ side pawns. 14

g5 tLJh 5 25 !:tfd 1 tLJc7

26 . . .f5! 27 h4 b 5 2S axb6 axb6 29 l:haS \\WxaS 30 \\WxaS J::!. x aS 3 1 d 5 34 .liLxd 5 l:!.a4 35 .liLxb6 l:!.xh4 36 .liLc6 .liLe7 3 7 J::!. a 1 tLJf4 3S .l:!.aS+ <;t>f7 39 .liLeS+ �e6 40 J::!. a 6 �d 5 41 .liLf7+ <;t>e4 42 il..e 3 Yz - Yz

1 4 \\We2 tLJ 5 b 6 1 5 .liLb3 c 5 1 6 tLJc3

0-0 1 7 a 5 cxd4 1 8 exd4 tLJcs 1 9 lLlf4

Instead of 9 liJh4, White has a more testing plan: 9 ik'e2 and 10 e4! 21

Th e Sla v

Game 6 Ivanchuk-Lautier

Linares 1994 1 d4 d 5 2 e4 e6 3 ctJf3 ctJf6 4 ctJe3 dxe4 5 a4 iif5 6 e3 e6 7 iixe4 iib4 8 0-0 ctJ b d 7 9 'iWe2 iig 6 1 0 e4 ! ?

This pawn sacrifice is the problem with this move-order. 1 0 . . . iixe3

The more restrained 10 . . . 0-0 is con­ sidered in the next chapter. 1 1 bxe3 ctJxe4 1 2 iia3 !

c6! and 13 . . . QJexc5 with 14 .ib5 ! 0-0 15 .ixd7 QJxd7 16 .ixf8 . Question 9: How then can Black get his king to safety? Answer: Black can castle queenside instead of kingside. Question 1 0: You don't sound very impressed! Answer: Black's position is horrible! White has a simple and extremely ef­ fective plan: a4-a5-a6, softening up the black queenside, and then, after mov­ ing the bishop on c4, c3-c4-c5 finish­ ing off the job. In reply, Black must seek to play . . . e6-e5 and activate his kings ide pawns. 1 2 . . .. �e7 1 3 afe 1

From el, the rook protects the c3pawn and supports the c3-c4-c5 push. 1 3 . . 0-0-0 .

Question 6: What i s going on? Answer: In return for the sacrificed

centre pawn, White has gained the two bishops and prevented Black from castling kingside. Question 7: Can't Black just grab another pawn with 12 .. .ct:Jxc3? Answer: NO!! 13 'i'b2 (hitting the knight and the pawn on b7) 13 . . . QJxa4 14 'i'b3 ! , threatening both 'i'xa4 and .ixe6, is horrible for Black. Question 8: Can't Black just block the a3-f8 diagonal with 12 . . . c5 and then castle kingside? Answer: This is logical, but 13 dxc5 is awkward, meeting 13 . . . 0-0 with 14 22

,

13 . . . c5 14 QJe5 ! QJxe5 (14 . . . 0-0 15 QJxg6 wins a piece) 15 dxe5 'i'xe5 (or else White plays f2-£3 and h2-h4, trap­ ping the knight) 16 £3 wins a piece, while 13 . . . QJd6 (blocking the a3-f8 diagonal) 14 .ixe6! 0-0 (14 . . . fxe6 15 'i'xe6+ is crushing) 15 .ib3 favours White due to his bishop pair. 1 4 a5

1 4 . . . ctJd6 ! ?

Th e O ld M a in L in e : B l a c k pla ys to pre v e n t e 3- e 4

The young Rumanian player Gab­ riel Schwartzman tried 14 .. J;fhe8 against Razuvaev in Dortmund 1993, but after 1 5 a6! b6 16 ttJh4 ttJd6 17 jLb3 es 1 8 ttJxg6 hxg6 19 iVg4! (preventing Black from activating his kingside pawns with .. .f7-fS by attack­ ing g6) 19 . . -,,�?b 8 20 jLxd6! iVxd6 2 1 jLxf7 �e7 2 2 iVxg6, White stood clearly better. 19 . . . cS cutting out the bishop on a3 , was suggested as an im­ provement, but after 20 dxcS bxcs 2 1 �ab 1 ! (preventing 2 1 . . .�b8 due t o 22 jLxf7+) 2 1 . . .�e7 (defending f7) 22 �cd1! (eyeing the knight on d6) the only positive course of action open to Black is to wring the neck of the man who made this suggestion! 1 5 1l.b3 1l.h5

1 7 . . . b6 1 8 c4 !

Threatening c4-cs. 1 8 . . . c 5 1 9 � a 4 e 5 20 dxc5 b x c 5 2 1 �e3 ge6 22 tLlg 5 .l::rf 6 23 � b 2 h 6 24 tLle4 tLlxe4 25 �xe4 ge6

I would not recommend this posi­ tion to anyone. The game finished:

An idea of Bareev's I believe, trying to inconvenience White by the pin on the knight. Note that 1 s . . . �he8 is met by 16 ttJh4!

gxa6 3 3 �xg7 h5 34 f4 f 5 35 �e5+

1 6 h3

'it>b7 36 gc7+ 'it>b6 3 7 'it>h2 'it>a5 38

16 'iVe3 unpinning, and eyeing the a7-pawn, is also interesting.

4 1 .l::rx a7 'it>xc4 42 1::!. a 8 �c6 43 .l:!.a3

16 . . J!he8

26

ge 1

�g6

27

�a8+

�b8

28

1l.xd7+ gxd 7 29 �xb8+ 'it>xb8 30 gxe5 gdd6 3 1 .l:!.xc5 .l:!.xa6 3 2 gxa6

'it>g3 �e8 39 'it>h4 'it>b4 40 g 3 gg6 �e8 44 ge3 �d 5 45 �f6 � d 7 46 �c3 �e6 47 � b4 �d7 48 �xh 5 gg8 49 �c3 �d6 50 'it>h6 �e6 5 1 �h7 ga8 5 2 � b4+ �d7 5 3 �g7 l::!. a4 54 'it>f6 � d 5 5 5 �c3 �e4 56 g4 fxg4 57 hxg4 �c2 58 .l:!.e7+ �d8 59 �e5 ga6+ 60 l::!. e 6 gxe6+ 6 1 �xe6 �e8 6 2 �d6 �d 1 6 3 g 5 1l.c2 64 f5 1 -0

Game l Ehlvest-Schwartzma n

New York

Open 1996

17 a6!

Softening up the protection around the black king.

1 tLlf3 d5 2 d4 tLlf6 3 c4 c6 4 tLlc3 dxc4 5 a4 �f5 6 e3 e6 7 �xc4 �b4 23

Th e S l a v

8 0-0 ttJ b d 7 9 '¥We2 � g 6 1 0 e4 �xe3 1 1 bxe3 ttJxe4 1 2 �a3 '¥We7 1 3 J:l:fe 1 0-0-0 1 4 a 5 �b8 1 5 �e7 ! ?

prepares to activate the light-squared bishop on the long diagonal. 44 . . . 'i'xf3 loses simply to 45 'i'd6+. 44 . . . a 5 45 iLg2 �e6 46 '¥Wd8+ �a6 47 '¥We7 ttJd7 48 f4 '¥We7 49 �e6+ ttJb6 50 '¥Wxg 6 a4 51 '¥We8 a3 5 2 'i1\Vf8 a 2 53 'i1\Va3+ ttJa4 54 'i1\Vxa 2 'i1\Vd 7 55 �d4 1 -0

If this isn't enough to convince you of the danger in accepting the pawn sacrifice, then try this! Game 8 Hubner-Beliavsky

An interesting manoeuvre, transfer­ ring the bishop to the annoying h2-b8 diagonal. The bishop has already ful­ filled its task on a3 by preventing the black king from castling kingside.

Munich 1994 1 d4 d 5 2 e4 e6 3 ttJe3 ttJf6 4 ttJf3 dxe4 5 a4 �f5 6 e3 e6 7 �xe4 � b4 8 0-0 ttJbd7 9 'i1\Ve2 �g6 1 0 e4 �xe3

1 5 . . . .i:!.de8 1 6 �h4 �a8 1 7 '¥Wb2 f5

11

1 8 ttJd2 f4 1 9 f3 ttJxd2 20 '¥Wxd 2

J::!. f e 1 ! ?

J:l:hf8 2 1 �f 1 e 5 22 d x e 5 J:l:xe 5 2 3

A very aggressive alternative to the old 13 Mfc l . White sacrifices yet an­ other pawn, reasoning that this will merely open more lines for his pieces.

a6

Ehlvest criticises this move, prefer­ ring White after 23 Ma4 Md5 24 'i'a2 ctJc5 25 Md4. This may well be more accurate, but the essential point is that White will always have very good chances because Black's king is weak and White's bishops are strong. 23 . . . b 6 24 :!la4 .i:!.d 5 2 5 :!ld4 :!lxd4 26 '¥Wxd4 ttJ e 5 2 7 .i:!.d 1 :!le8 28 �e4 iLe2 29 J::!. e 1 llxe 1 + 30 1i.xe 1 1i.f5 3 1 1i.d 2 '¥Wd 7 32 'li'xf4 b5 3 3 �f 1 ttJxa6 34 �e3 � b 7 3 5 �f2 �g6 3 6 h4 e 5 3 7 '¥W e 3 �b6 3 8 � h 2 '¥We8 3 9 '¥Wg 5 h 6 40 '¥Wd 2 '¥W e 6 4 1 �g3 �f6 42 �d7 e4 43 �f2+ ttJe5 44 g 3 !

After a time-scramble and a little confusion, White re-establishes con­ trol with this evil little move, which 24

bxe3 ttJxe4 1 2 �a3 'i1\Ve7

13

1 3 . . . ttJxe3

Best and the most critical. 13 . . . 0-0-0 was extremely unpleasant for Black in the game Beliavsky-Akopian, No­ vosibirsk 1993: 14 'i'b2 Mhe8 15 as e5 16 Mab 1 c5 17 �f1 f6 and now 1 8 ctJh4! ctJxc3 19 'i'xc3 exd4 2 0 'i'b3 JiLxb 1 21 Mxb 1 gives White an over­ whelming initiative, as Beliavsky pointed out. Clearly in such lines, the king's rook is much more actively placed on e 1 than on c 1 (as in lines we have seen previously). This is also true of 13 . . . c5, as in Hubner-Hertneck, Munich 1994, when 14 d5! e5 15 �d3! ctJef6 1 6 ctJxe5 0-0-0 17 ctJxd7 Mxd7 1 8

Th e O ld M a in L in e : B l a c k pla ys to pre v e n t e 3- e 4

c4 (Hubner) is the (unpleasant) best that Black can hope for. 14 'i'b2 ctJe4 1 5 a 5 ! ?

now becomes a little random, due to mutual time pressure, but White pulls through in the end.

T o break up the black queenside with a5-a6. 15 I[ac 1 is also interesting. 15 . . . ctJdf6

15 .. .liJd6 (intending . . . 0-0) is met by 16 .!xe6! (16 Vib4 c5 ! [not 16 . . . lZ'lxc4?? 1 7 'ife7 mate] 17 dxc5 lZ'lxc4 1 8 c6

4Jxa3 ! 19 cxd7+ Vixd7 20 Vixa3 Vie7 'ifa4+ Vi d7 22 Vi a3 [preventing kingside castling], which leads to a draw by repetition after 22 . . . Vie7) 22 . . 0-0 17 ltxd7 Vixd7 1 8 lZ'le5 Vic7 19 a6! , breaking up the queenside with an advantage. Instead 1 5 ... a6 (preventing a5-a6) is best, when Hubner suggests 16 I[e3 4Jd6 17 .!xe6 0-0-0 18 ltxd6 Vixd6 19 .!c4 and I[b3 with a dangerous attack.

f 4 Il:.d 5 3 3 .l:i.e4 g 5 3 4 �a4 e 5 3 5

16 tLJe5 a6 1 7 .l:i.ac 1 .l:i.d 8 1 8 jLxa6 ! !

fxe 5 fxe 5 3 6 � a 7 � g 8 3 7 JJ.e7

21

.

1 8 . bxa6 1 9 ctJxc6 ctJg4 20 ctJ e 5 ! . .

22 . . . ctJd6 23 �c7 .l:i.d 7 24 �c3 0-0 25 ctJe5 Il:.dd8 26 �c6? �xc6 27 .l:i.xc6 ctJb5 28 jLxf8 �xf8 29 gxa6 ctJxd4 30 f3 f6 ? ! 31 ctJxg6+ hxg6 3 2

N ow White is winning again.

'i'xa5 2 1 ctJxg4 �b5 2 2 �c2 ? !

37 . . . ctJc6 38 JJ.e6 gc5 39 �g6 �f7

This i s White's first inaccuracy in this fascinating game! Hubner notes that the simple 22 Vixb5 axb5 23 I[c7 I\,a8 24 ltb4 I[a4 25 I[c8+ 'it'd7 26 I\,xh8 I[xb4 27 lZ'le5 'it'c7 28 I[c1 + 'it'b7 29 f3 lZ'lf6 30 lZ'lxg6 would have been clearly better for White. The game

40 �xg 5 g 6 41 h4 �f6 42 �h2 ctJe7 43 �a4 �c6 44 .l:i.g 3 ctJf5 45 gf3 �g7 46 l:te4 �e6 47 h5 ctJe7 48 hxg6 ctJxg6 49 �h3 ctJe7 50 �g3 ctJg6 51 �f5 �e7 5 2 �g4 l:ta7 5 3 � g 5 ctJh8 5 4 � g 4 ctJ f 7 + 5 5 <;t> h 5 + � f 8 56 �g6 � e 8 5 7 � h 4 1 -0 25

Th e S l a v

S u m m a ry

The alert reader will have noticed my profound mistrust of the line 8 . . . 'bbd7 9 'i'e2 �g6 10 e4 �xc3 1 1 bxc3 'bxe4. I honestly cannot understand the at­ traction of these lines for Black. Therefore, if Black wishes to try to prevent e3-e4, then Kramnik's 8 . . . 0-0 9 'i'e2 �g6 is the line for you; Yusupov's plan of a quick 'bh4 is the most testing response. 1 d4 d 5 2 c4 c6 3 ttJt3 CUt6 4 ttJc3 dxc4 5 a4 .ltf5 6 e3 e6 7 .ltxc4 .ltb4 8 0-0 8 . . . 0-0

8 . . . 'bbd7 (D) 9 'bh4 �g6 10 �e2 - game 4 1 0 h3 - game 5 9 'i'e2 �g6 (9 . . . �g4 - see next chapter) 10 e4 �xc3 (10 . . . �g6 see next chapter) 1 1 bxc3 'bxe4 12 �a3 'i'c7 13 .sfe 1 - game 8 1 3 .sfc1 0-0-0 14 as (D) 14 . . . 'bd6 game 6 14 . . . c,t>b8 - game 7 -

9 �e2

9 'bh4 game 3 -

9 . .lt g 6 . .

9 . . . 'bbd7 - see next chapter 1 0 ttJe5 ttJ b d 7 1 1 ttJxg6 hxg6 1 2 .l:td 1 (D)

12 . . . 'i'e7 - game 1 12 . . . 'i'aS game 2 -

B... ttJ b d7

26

14 a5

12 'iJ.d1

CHAPTER TWO The Old Main line: Black allows e3-e4

1 d4 d 5 2 c4 c6 3 l2lf3 l2lf6 4 l2lc3 dxc4 5 a4 i.f5 6 e3 e6 7 i.xc4 i.b4

8 0-0

This chapter deals with lines arising from 8 . . .cubd7 9 'Wie2 0-0 10 e4 iLg6 (or 9 . . . iLg6 10 e4 0-0, turning down the dangerous pawn sacrifice on the way) and 9 . . . iLg4. Question 1 : When White plays e3e4, he gains a commanding central presence with pawns on e4 and d4. Why is Black playing this position? Isn't he just worse? Answer: Central pawns are strong if they are dynamic and able to advance and chase away the opposing pieces. Otherwise, they can present easy tar­ gets for the enemy pieces. In this case, White cannot advance d4-dS, and e4eS leaves a hole on dS for the black pieces. Moreover, Black is threatening to win a pawn with . . . iLxc3 and tiJxe4, now that his king is safely cas­ tled. Therefore, while the d4-e4 centre gives White a definite space advantage, Black has plenty of threats against the white centre, which is the basis of his counterplay. Question 2: What is the difference between playing 9 . . . iLg6 first or 9 . . . 0-0 10 e4 iLg6?

Answer: Good question! Generally, black players play 9 . . iLg6 to pretend that they are willing to take on the pawn sacrifice after 10 e4. Even if they don't intend to take the pawn, the idea is to make White waste a lit­ tle time on the clock thinking about his variations! You never know - a few minutes might be handy later! .

Game 9 G ofshtein-Sadler

Ischia 1996 1 d4 d 5 2 c4 c6 3 l2lf3 l2lf6 4 tLlc3 dxc4 5 a4 i.f5 6 e3 e6 7 i.xc4 i.b4 8 0-0 tLlbd7 9 �e2 0-0 1 0 e4 i.g6 1 1 i.d3

White had to deal with the threat 27

Th e S l a v

of . . . iLxc3 and . . . tbxe4. It is a general rule that the longer you can delay committing your centre, the better, since the later you reveal your hand, the less time your opponent has to adjust to it. 11

. . .

it h 5

the knight on f3 to the queen and thus threatens . . . e6-eS. The alternative 1 1 . . .h6 is considered in Game 12. 1 2 itf4

Trying to avoid e4-eS for a while longer, White brings another piece to bear on eS. Strangely enough, this is probably not the best move. The di­ rect 12 eS is considered in the next game. 12

. . .

'Vj'e7 ! ?

Black threatens 1 3 . . . iLxf3 1 4 'iVxf3 eS! , equalising comfortably. I think that this is a novelty: 12 . . . Me8 had been played before. 1 3 e5

Absolutely necessary. 13

. . .

tZ:l d 5 1 4 tZ:lxd5 cxd 5

Question 3 : What is the pomt of

this move? Answer: The first place to look for

counterplay, is with your pawn breaks. Pawn breaks have two func­ tlons: a) They attack the opponent's pawn structure and force him to react to you. b) They are a 'breakout': they gain space and therefore give more room for your pieces to become active. Black has two pawn breaks in this position: . . . e6-eS and . . . c6-cS . Usually he prefers to aim for . . . e6-eS, since this stops White from playing e4-eS him­ self, inconveniencing the black knight. For example, 1 1 . . .cS 12 eS! tbdS 1 3 tbxds exdS (13 ... iLxd3 14 'iVxd3 exdS 1 5 tbgS ! is unpleasant) 14 iLxg6 hxg6 15 tbgS, with threats of eS-and 'iVg4-h4 and 'iVh7 mate, is nasty for Black. The move in the game pins 28

White has a space advantage due to his pawns on d4 and eS. Normally, with his bishops, knight and queen pointing towards the black king, White could consider launching a kingside attack, but here Black's bishop on hS interferes with this plan: it can exchange itself for the knight on f3 or return to g6 to block any white attack on the b 1-h7 diagonal. Meanwhile Black will challenge for the c-file, exchanging the bishop on

Th e O ld M a in L in e : B la c k a lia ws e 3 - e 4

d3 for the one o n g6 i n order t o free c2 as an entry square for the black major pieces. Black will also transfer his knight to c6 via b8 from where it not only attacks d4, but can invade the white queenside by as-b3 or via b4. So what on earth can White do? Stay calm! White does not want to exchange pieces on the c-file since this would help Black to free his cramped position, so he has two plans. First (my own favourite) , he can concen­ trate on the kingside where White holds most of the trumps: a space ad­ vantage and a large concentration of minor pieces. I would try to push my kingside pawns: 1 5 h3 !!fc8 16 g4 ltg6 17 h4, intending h4-hs. This plan demonstrates the drawback to ltf4, however: White would like to throw the f-pawn forward as well, but the bishop gets in the way. The chances after 15 h3 are, I believe, about equal. White's choice is interesting, but there is always a danger in choosing plans based mainly on tactical points: if there is just a little hole in your calcu­ lations, then you often find that you have just wasted time and must retreat in disarray. On the other hand, such plans are often the most unexpected and the most difficult for the oppo­ nent to deal with!

1 7 . . . Jixd 3 1 8 �xd 3 .l:!.e4 !

Black is now planning . . . b7-bs and 19 b3 is met by 19 . . . Mc7 ! , when 20 'ik'bS does not win a piece as Black has 20 . . . ltc3 , when he stands well. Realis­ ing that his activity on the queenside has come to naught, White goes back to 'Plan A' and expands on the king­ side, but he is several tempi down on what he could have had earlier. 1 9 h4 h 6 20 h5 ctJ e 5 !

1 5 �e3 1:He8 1 6 a 5 ! ? Jig6 1 7 .l:!.a4 ! ?

At first I thought about playing 17 . . Mab8, intending 18 . . . bS to drive away the white rook. Then to my horror I noticed 18 ltbS ! White is threatening to take the knight on d7 and then take my bishop on b4, so 18.. .ctJf8 is natural, but then 19 'ik'b3 ! and my bishop i s trapped! .

Oops! My opponent had missed that one. However, after a big think, he came up with an active defence. 2 1 dxc5 nxf4 22 a6 ! bxa6 23 c6 a 5 2 4 .l:!.e 1

g e 8 2 5 b 3 ge7 26 .l:!. e 2

ne4? !

A rather casual move. 26 . . . Mfs 27 29

Th e S l a v

Mal Mxh5 28 g4 Mh3 29 'It>g2 �h4!! (this lovely move was pointed out to me by Julian Hodgson after the game) was the way to play. 27 '!::!'a 1 jL c 5 ? ! 28 .!::!.x a5 �xc6 ? ?

35 'It>gl Mc1+ leads to immediate mate. So 12 �f4 doesn't seem all that promising for White. What about 12 e5 instead? Game 10 Gelfand-Lautie r

.t-urich 1994

1 d4 d 5 2 c4 c6 3 lLlf3 lLlf6 4 lLlc3 dxc4 5 a4 jLf5 6 e3 e6 7 iLxc4 jL b4 8 0-0 lLlbd7 9 �e2 jLg6 1 0 e4 0-0 1 1 jLd3 iLh5 1 2 e 5

Probably the best move. 29 .!::!. b 5 ? ?

My God! I had missed that White could simply win a piece with 29 �b5 , forking rook and bishop. I had thought that I could play 29 . . . �xf2+ with a discovered attack on the rook on c2, but White just plays 30 Mxf2! Luckily White shared the same blind spot! After 29 Mb5, White is just lost. 29 . . . �c7 30 �a2 iLb6

Trapping the rook. 3 1 <;t>h2 .l::!. c3 3 2 �d2 �c6 33 gxb6

1 2 . . . lLld 5 1 3 lLlxd5

axb6 34 lLld4 .!::!. h 4+ 0- 1

The alternative, 13 liJe4, with in the next game.

1S

dealt

1 3 . . . cxd 5

13 . . . exd5 has been suggested, but since White already has a space advan­ tage on the kingside, I am sceptical about conceding a pawn majority as well in that area. 1 4 �e3 h 6 ? !

A debatable decision. I would pre­ fer 14 . . . �e7, followed by a rook to the c-file and . . . liJb8-c6. 1 5 lLle 1 ! 30

Th e O ld M a in L in e : Bla c k a I/a ws e 3- e 4

pletely dominates the black knight. 29 . . . 'i'xg6 30 'i'e3 � h 7 3 1 3i.f4 'i'f7 3 2 3i.e 5 'i'd7 3 3 h4 \tlg8 34 h 5 !

A very instructive plan: White in­ tends to play f2-f4 and f4-f5, which is particularly dangerous once Black has weakened his kingside with . . . h7-h6. Moreover, the knight will eventually come to d3 with tempo, hitting the bishop on b4. 15 .. .f5 1 6 exf6 'i'xf6 1 7 3i.b5 ttJb8 1 8 tLld 3 ! a6

Fixing the g 7-pawn. 34 . . . 'i'd8 3 5 'i'g3 'i'd7 36 3i.d6 'i'f7 37 'i'e5 'i'f5 38 'i'e2 b4 39 g4 'i'f6 40 3i.e 5 'i'g 5 41 �g2 b3 42 3i.d 6 !

Preventing the knight from becom­ mg active.

18 ... 1td6 19 tLJe5 is not pleasant for Black. 1 9 tLlxb4 axb5 20 a5 ttJa6 2 1 ttJ d 3 ! l:tfe8 22 3i.d 2 g e 2 23 gae 1 g a e 8 24 l:txe2 gxe2 25 ge 1 gxe 1 + 26 3i.xe 1 iLg6 2 7 ttJe5 'i'f5 2 8 h 3 'i'e2 29 lLlxg 6 !

42 . . . �f7 43 'i'd 1 \tlg8 44 3i.g 3 'i'f6 45 3i.e 5 'i'h4 46 3i.g 3 'i'f6 47 'i'd2 'i'e7 48 'i'e3 'i'f6 49 3i.d 6 � h 7 50 'i'x b3 'i'xd4 51 'i'e2+ \tlh8 5 2 'i'e8+ �h7 53 'i'e2+ �h8 54 'i'e8+ �h7 5 5 'i'xe 6 !

Protecting g4. Excellent judgement. In the result­ ing position, White's bishop com-

55 . . . 'i'xb2

56

3i.e 5 'i'c2

57

'i'e7

'i'e4+ 58 �g3 'i'd3+ 1 -0 31

Th e S l a v

After 59 f3 , Black cannot stop mate. His knight has not moved since move 20! Black should be fine after the ex­ change of knights on dS , providing he avoids weakening his kingside. Let us take a look at 1 3 tbe4.

Game 1 1 Xu Jun-Akopian

Moscow Olyrt;lpiad 1994 1 d4 d 5 2 c4 c6 3 CLlf3 CLlf6 4 CLlc3

queen to the kingside. Yes, this is the hacker's option! 1 3 . . . �e7

13 . . . cS is unpleasantly met by 14 ltgS! 'i'aS 15 ltbS! Once Black moves the knight on d7, he will lose the cs­ pawn, and he cannot protect it with a rook due to the bishop on gS. If he protects the knight with l S . . . 'i'c7, then 16 Mac1 is unpleasant. The text prevents ltgS and prepares . . . c6-cS . 1 4 CLlg3

The direct approach. 14 as has also been tried.

dxc4 5 a4 �f5 6 e 3 e 6 7 �xc4 � b4 8 0-0 CLlbd7 9 �e2 �g6 1 0 e4 0-0 1 1 � d 3 �h5 1 2 e 5 CLld 5 1 3 CLle4 ! ?

Question 4: Why does White play

14 as? Many white players d o not enjoy the positions that we have seen in the first two games of this chapter. With the centre closed, and the prospect of exchanges on the c-file, they feel un­ easy about their winning prospects; so recently the plan with 13 tbe4 has come to prominence. Of course, Black keeps his knight outpost on ds and his pawn-break against the centre with . . . c6-cS. However, White's space advantage remains and he retains e4 to transfer first his knight, then his 32

Answer: a4-aS gains queenside space, preventing Black from using the as or b6 squares for his pieces. Question 5: So what's the verdict? Is it a good move? Answer: Absolutely not! These aims are completely irrelevant. 14 . . . cS is logical, striking at the d4-pawn. Neither 1 5 dxcs tbxeS nor 15 tbxcs tbxcs (lS . . . ltxcS!?) 16 dxcS MCS (of course not 16 . . . ltxcs 17 ltxh7+ �xh7 1 S iVc2+ �gS 19 tbgS ltg6 20 iVxcs, winning a pawn) , followed by . . . ltxcs or . . . MxcS, promise White anything.

Th e O ld M a in L in e : B l a c k a //o ws e 3 - e 4

14 . . JLg 6 .

1 5 JLxg6 hxg6

15 ... fxg6 used to be played almost automatically, primarily for defensive reasons: black players were worried that if they recaptured with the h­ pawn, White would play his knight on g3 to gS via e4 and his queen to h4, when Black would have no defence to 'Ih7 mate. By taking with the f-pawn Black retains the option of ... h7-h6 to keep a knight out of gS, and of course, he gains the semi-open f-file for coun­ terplay. And then people realised that White's attack was hardly automatic after 1 5 . . . hxg6, so this move gradually became the main line! 1 6 tLle4 c 5

1 7 tZJc3

This is a perfectly reasonable plan, challenging Black's knight on ds once Black has weakened its support by playing . . . c6-cS, but why did the Chi­ nese player avoid 17 CLlegS? 17 . . . cxd4 18 'iVe4 's'e8 19 'iVh4 CLlf8 seems to be a good reason. The knight on f8 de­ fends against 'iVh7 and if by some miracle White manages to threaten to get a rook on h3, then Black can hit the 'panic button' and chase the knight away with .. .£7-f6. White could, however, try and open up the black kings ide with h2-h4-hS , possibly after 16 'iVe4 cS 17 h4 cxd4 18 hS !? And now it's up to you, the reader! We'll have to wait for practical tests before a conclusion can be reached. 1 7 . . . �b6

17 . . . CLl7b6 18 as CLlxc3 19 bxc3 CLlds 20 c4 CLlb4 also seems reasonable. 1 8 tZJxd 5 exd 5 1 9 dxc5 tZJxc5 20 JLe3 �e6 21 a5 a6 22 .l:!.ac 1 .l:!.ac8 23 .l:!.fd 1 tZJb3 24 .l:!.xc8 .l:!.xc8

Black has an isolated d-pawn, but White's queenside is weak. The posi­ tion is about equal. 25 JLb6 JLd8 ? ! 26 JLxd 8 .l:!.xd8 27 �e 1 ? !

33

Th e S l a v

J::i b 6 J::i d 5 3 6 b4 d 3 3 7 c!;e 1 c!;d7 38

l1.. .'iYaS has a similar idea. After 12 �f4! lIfe8 (12 . . . �xc3 13 bxc3 'iYxc3 loses to 14 �d2 'iYc2 15 �d3 'iYb2 16 lIfb l) 13 h3 lIac8 14 12Ja2!, White had a slight advantage in Beliavsky-Short, Linares 1995, as 14 . . . �f8 (14 . . . 'iYxa4 15 ttJc3! 'iYb3 16 �c4 wins the queen) 15 b4! gains queenside space with tempo: 15 . . . 'iVxa4 16 lIfb l and 12Ja2-c3 traps the queen, while 15 . . . �xb4 16 12Jxb4 'iYxb4 17 lIfb l 'iYaS 18 lIxb7 is unpleasant for Black.

�d2 c!;c7 3 9 J::i d 6 J::i x d6 40 exd6+

1 2 �f4 J::i c 8 1 3 l:Ifd 1 J::i e 8 1 4 h3 a6

�xd6 41 �xd 3 �d5 42 tLJ d 2 tLJc6

1 5 l:Iac 1 !

43 c!;c3 tLJe5 44 tLJb3 tLJc4 45 f4 f6

Black's slightly incautious 25th move allowed White the chance to activate his queen by the lovely 27 'iYe4!, intending 12Jg5 and 'iYh4, as 27 . . . clxe4 allows 2 8 lIxd8+ �h7 29 ttJg5+ winning the queen (analysis by Xu Jun) . The rest is hard-fought, but it was always going to be a draw. 27 . . :Viii e 7 28 'Viii c 3 tLJ c 5 29 'iVb4 �f8 30 �f 1

tLJe6 3 1 'Viiix e7+ c!;xe7 3 2

J::i d 3 d4 3 3 J::i b 3 J::i d 7 34 g 3 tLJd8 3 5

1 d4 d 5 2 c4 c6 3 tLJf3 tLJf6 4 ct:Jc3

After 15 ttJa2 �f8 16 b4 Back has 16 . . . aS! (an excellent idea, breaking White's grip on the c5-square) 17 bxaS 'iYxaS 18 �d2 'iYc7 19 e5 �xd3 20 'iYxd3 12JdS 21 ttJc3 ttJxc3 22 �xc3 ttJb6 23 as ttJdS 24 �el c5 and Black had no problems in LSokolov-Oll, Moscow Olympiad 1994.

dxc4 5 a4 �f5 6 e 3 e6 7 �xc4 Si.b4

1 5 . . . �b6

8 0-0 tLJ b d 7 9 'Viii e 2 Si.g6 1 0 e4 0-0

No better is 15 . . .'iYaS 16 ttJd2! b5 (16 . . . �xc3 17 bxc3 'iYxa4 18 lIal traps the queen) 17 axb5 axb5 18 12Jb3 'iYb6 19 e5 12Jd5 20 12Jxd5 exdS 21 �xg6 fxg6 and Black was probably happy he couldn't see his position in Lautier­ Gelfand, Monaco (blindfold) 1999.

lh - lh

Game 12 Sadler-M iles

British Championship 1 998

1 1 � d 3 h6

1 6 �b 1 !

I really like this development scheme. White consolidates his queen­ side and mobilises all his pieces, ready for any of Black's breaks. 1 6 . . . �h 7 ? !

A slightly risky idea. Black devel­ ops quietly and waits for an opportu­ nity to break with either . . . c6-c5 or ... e6-e5. The problem is that, as in this game, Black can really get sat on! 34

Black should really have taken the plunge with 16 . . . c5, although 17 12Ja2! wins the bishop pair with a nice ad­ vantage for White. 1 7 tLJe5?

This is rather careless though!

Th e O l d M a in L in e : B l a c k a ll o w s e 3 - e 4

17 . . :i'd8 ?

17 ... ctJxeS 1 8 dxeS "iVc7! 19 "iVc4 �f8 is okay for Black.

tremely solid and has no pawn weaknesses.

1 8 ClJc4!

N ow

Black is suffering.

18 . . . ClJb6 1 9 ttJ a 2 ! .ll. f 8 20 b 3 !

Maintaining the knight on c4, as Black will now have to improve White's structure to get rid of it. 20 . J::i. a 8 2 1 �h 1 ttJc8 22 a5 ttJd7 .

.

23 ClJc3 i.b4 24 ttJa4 Wiie 7 25 Wiig 4

itJf6 26 �f3 ttJa7 27 .ll. g 3 ttJd7 28 d5! exd 5 2 9 exd 5 i.xb 1 30 J::i. x b 1

cxd 5 3 1 ttJcb6 ttJxb6 3 2 ttJxb6 gad8 33 ClJxd 5 'i'f8 34 ttJc7 ge7 3 5 J::i. x d8

1 5 . . . a6 1 6 ..\te2 J::i. f d8 1 7 ga4 ..\te7

�xd8 36 ttJ d 5 J:i.e 1 + 3 7 J::i. x e 1 i.xe 1

1 8 f4 CLle8 1 9 .ll. e 3 gac8 20 .ll. f3

38 b4 Wiie 8 ? ? 3 9 ttJc7 ! �c6 40 'iVe3

ttJd6 21 i.e2 g 6 22 J::i. a a 1 1h - Y2

ixb4 41 Wiix a7 .ixa5 42 �b8+ �h7

43 ClJe8 b 5 44 ttJd6 f6 45 Wii a 7 i.b4 46 ClJf5 i.f8 47 �f7 'i'c8 48 CLld4 ia3 49 ttJe6 'i'g8 50 'iVd7 �h8 5 1 ic7 .ib4 5 2 f4 .ic3 5 3 f5 b4 54 id6 1 -0

Here Kasparov shows a more criti­ cal approach for White. Game 14 Kasparov-Bareev

Occasionally 9 . . . .,tg4 is seen instead of 9 .. .,tg6, as in the next two games. .

Novgorod 1994

1 d4 d5 2 c4 c6 3 ttJf3 ttJf6 4 <1Jc3 dxc4 5 a4 i.f5 6 e 3 e6 7 i.xc4 i. b4

Game 13

8 0-0 ttJbd7 9 'iWe2 i.g4 1 0 h 3 !

Khalifman-Kir Georgiev .

Elenite 1994 1 d4 d5 2 ttJf3 ttJf6 3 c4 dxc4 4 itJc3 c6 5 a4 Jl. f5 6 e3 e6 7 i.xc4 ib4 8 0-0 ttJ b d 7 9 �e2 i.g4 ! ?

A rather unusual bore White to tears into a dull ending. 10 J::i. d 1 'i'a 5 !

11

move that aims to by exchanging off e4 'i'h5

1 2 h3

ixf3 1 3 'i'xf3 'i'xf3 1 4 gxf3 0-0 1 5 a5

White has the two bishops and a space advantage, but Black is ex-

The most aggressive idea: now if 1 0 . . . .,thS, White can play for e3-e4 35

Th e S l a v

without fear of the . . . iVaS-hS manoeuvre. 1 0 . . . SLXf3 1 1 jVxf3 0-0 1 2 .i:::l. d 1 jVa5 1 3 e4 e5 1 4 d 5

This bears a distinct similarity to Richardson-Sadler from Chapter 1 . White's queen i s better placed o n f3 than e2, though of course Black's queen is more actively placed as well. 1 4 . . . ttJb6

1 5 �b3 �xc3

1 6 bxc3

cxd 5 1 7 exd5 .i:::l. a c8 1 8 c4! !

Brilliant. 1 8 . . . lLJxc4 19 �gS iVa6 20 d6! lLJxd6 21 �xf6 gxf6 22 iVxf6 lLJc4 23 iVxa6 bxa6 24 Md7 is clearly better for White (Kasparov) .

Which is the better move? A nswer: I can offer no definitive conclusion: in theory, 8 . . . 0-0 and 8 ... lLJbd7 are equally good. However, practical chess is not only about find­ ing the very best move: when making a final decision whether to play 8 ... 0-0 or 8 ... lLJbd7, it is also necessary to take the strength of your opponent mto account. The following game shows why. Game 15 D . Strauss-Lakdawala

USA

1992

1 d4 d5 2 c4 c6 3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 � f5 6 e3 e6 7 �xc4 �b4 8 0-0 ttJbd7 9 'li'b3

1 8 . . . .i:::l. f e8 1 9 � d 2 jVa6 20 d 6 ! ttJbd7 21 �e3 .i:::l. c 6 2 2 a 5 ! .i:::l.x d6 23 �a4 .i:::l. x d H 24 .i:::l. x d 1 .i:::l. d 8 25 c 5 !

Black has been tied up in quite bril­ liant fashion. 25 . . . h 6 26 jVf5 g 6 27 �c2 ! �xa5 28 .i:::l. d 6 <;t>g7 29 jVd 1 �c7 30 h4 �h8 31 h 5 g 5 32 'i¥f3 <;t>g7 33 'i¥f5 b6 34 SLxd 7 ! ttJxd7 3 5 �xg 5 ! 1 -0

Kasparov gives 3 S . . . hxgS 36 iVxgS '.tf8 37 h6 bxcS 38 h7 winning. A really magical game. Question 6: In your own games you have played both 8 . . . 0-0 and 8 . . . lLJbd7. 36

This move forces a draw by repeti­ tion if White desires after 9 . . . a5 (best) 10 lLJa2 (chasing the bishop away) 10 . . . �e7 1 1 iVxb7 Mb8 12 iVa6 (12 iVxc6 Mb6!) 12 ... Ma8 1 3 iVxc6 Mc8 14 iVbS Mb8, as the queen cannot escape from the rook's attack. So if your op­ ponent is much weaker than you, or you desperately need a win, you must play 8 . . . 0-0, as 9 iVb3 iVe7! (protecting b7 and facilitating ... c6-cS) is nothing

Th e O l d M a in L in e : Bla c k a ll o w s e 3 - e 4

for White. Alert readers will observe

1 2 ....txa5

that I chose 8 . . .0-0 against Richardson in just such a must-win situation. The

tiJe4 1 5 tiJxe4! .txe4 1 6 "lif7 ! !

1 3 "lixe6+ '>t>dS

1 4 e5

attempt to avoid the repetition in this game is brutally dealt with.

9 :tlVb6 1 0 e4 ..

A

brilliant

discovery

of

David

Gliksman. The queen moves to allow the e-pawn to advance. 1 6 . . . MfS

10 . . . .tg6

10 .. .ctJxe4 1 1 CLlxe4 �xe4 1 2 �xe6! is clearly better for White. 11 �xe 6 ! fxe6 1 2 a 5 !

The situation is already desperate. 1 6 ... �xf3 1 7 e6! (D.Strauss) 1 7 ... CLlf6 1 8 e7+ �c8 19 �e6+ �c7 ( 1 9 . . . CLld7 20 e8� 20 �f4 + wins, while instead 1 6 . . . �g6 17 �xg7 Me8 1 8 dS ! ! (again D.Strauss) ,

opening

more

central

lines, is crushing as 1 8 . . . cxdS 19 �g5 + �c8 20 Mfc 1 + is appalling for Black. 17

'llIVx g7

Si.d5

1S

e6!

.txe6

19

"lig5+!

Now the point of 1 2 as is revealed: this check picks up the loose bishop on as ! 1 9 .. . '>t>cS 20 Mxa5 �gS 2 1 "lih5 .tg4 22 Mg 5 ! .txh5 23 MXgS+ tiJfS 24 MxfS+ '>t>d7 25 tiJe5+ '>t>c7 26 MxaS

Diverting the bishop to an inferior square.

'i'xd4 27 Me 1 .te2 2S MeS c5 29 �e7+ '>t>cS 30 tiJc6 1 -0

37

Th e S l a v

S u m mary

After 8 . . . LtJbd7 9 "iVe2 0-0 10 e4 �g6 11 �d3 �hS, 12 eS is probably White's best try and after 12 . . . LtJdS, hackers should choose 13 LtJe4 and positional players should favour 13 LtJxds and 14 "iVe3 . These positions are complicated and interesting in all cases. After Kasparov's brutal treatment, 9 . . . �g4 should probably be avoided unless you spot a flaw in 'Gazza's' analysis. If you do, then you're probably Mr Karpov! But remember, if you desperately need a win as Black, 8 . . 0-0 is the only way to play. .

1 d4 d5 2 c4 c6 3 CL'lf3 CL'lf6 4 CL'lc3 dxc4 5 a4 3l.f5 6 e3 e6 7 3l.xc4 3l.b4 8 0-0 8 . . 0-0 .

8 . . . LtJbd7 (D) 9 "iVe2 �g4 10 �dl - game 13 10 h3 - game 14 9 "iVb3 - game 1 5 9 �e2 CL'l b d 7 1 0 e 4 iL g 6 1 1 iL d 3 iLh 5 (D)

11. . .h6 - game 12 1 2 e5

12 �f4 - game 9 1 2 . . . CL'l d 5 (D)

13 LtJxds - game 1 0 1 3 LtJe4 - game 1 1

B

38

. . .

CL'lbd 7

11

. . .

iLh 5

12

. . .

CL'ld5

CHA PTER THREE The New M ain Line: Black fig hts for contro l of e4

1 d 4 d5 2 c4 c6 3 iZJf3 iZJf6 4 iZJc3 dxc4 5 a4 iLf5 6 iZJe5

The sequence 1 d4 d5 2 c4 c6 3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 �f5 6 ttJe5 is the latest fashion at the highest level. 6 ttJe5 truly is a 'modern' move: White does not worry about putting his king into safety; he tries immediately to achieve e2-e4 by moving the knight to e5 and playing f2-£3 . In principle Black should not allow White to achieve e2-e4, since here it is ideally defended by a pawn on f3 , leaving the white pieces free for active operations. In this chapter, Black continues the fight for e4 with an interesting piece sacrifice: 6 e6 7 f3 iL b4 . . .

Pinning the knight on c3 and thus fighting for e4. 8 e4

Question 1 : What is the material

balance? A nswer: Black has four pawns for a

piece, though it is touch and go whether he can hang on to the c4pawn. Even three pawns, however, is good material compensation for the temporarily rather inactive bishop on f1 .

... and after this there is no going back!

Game 16 Lal i c-Sadler

8 .lil.xe4 9 fxe4 iZJxe4 1 0 iLd2 10 'i'f3 leads to a well-known . . .

draw by repetition after 10 . . . 'i'xd4 1 1 'fxf7+ �d8 1 2 �g5+! ttJxg5 1 3 'i'xg7 .ixc3+ 14 bxc3 'i'xc3+ 15 �e2 'i'c2+ 16 �el 'i'c3+, etc. 10

. . .

�xd4 1 1 iZJxe4 �xe4+ 1 2 �e2

iLxd 2 + 1 3 �xd 2 �d 5 +

Hastings

1995/96

1 c4 c6 2 d4 d 5 3 iZJf3 iZJf6 4 iZJc3 dxc4 5 a4 iLf5 6 iZJe5 e6 7 f3 iLb4 8 e4 iLxe4 9 fxe4 iZJxe4 1 0 iLd2 �xd4

11

iZJxe4

�xe4+

12

�e2

iLxd2+ 1 3 �xd2 �d5+ 1 4 �c3 39

Th e Sla v

After 14 �c2 lLJa6, which is consid­ ered in the next game, White always has to worry about . . . lLJa6-b4+, which can be irritating. Therefore white players began to experiment with 14 �c3 , avoiding this sort of counter­ play. The one drawback to the king on c3, however, is that it is just within reach of the black queenside pawns, so Black can play an aggressive con­ tinuation that would not succeed against 14 �c2. 24 �xc2 'i'xb2+ 25 �d3 Mb3+ 26 �e4 f5+ 27 �f4 'i'd4+ 28 �g5 'i'd8+ 29 �h5 'i'e8+ 30 �h4 'i'e7+ 3 1 �h5 g6+ 32 �h6 'i'g7+ 33 �g5 h6+ 34 �f4 g5 mate!

1 4 . . . 0-0 ! 1 5 �e3

This looks incomprehensible - what is wrong with 15 lLJxc4? The problem is 15 . . . b5! 16 lLJe5 and now 16 . . . b4+! The king's exposed placement gives Black a vital extra tempo for the at­ tack. 17 �xb4 is met by 17 . . . lLJa6+! 1 8 �a3 ( 1 8 '»Yxa6 l'hb8+! 19 �c3 '»Yxe5+ winning) 18 . . . Mab8 19 'i'e3 'i'd6+ 20 �a2 lLJb4+ 21 �b 1 'i'd1+ 22 'i'c1 'i'd4! (threatening . . . '»Yxe5 and . . . 'i'e4+) 23 'i'e 1 lLJc2! see

follo wing diagram

Thanks to this powerful knight thrust, Black is now winning by force. Fasten your seat belts, a rather long variation lies ahead! 40

The attempt to turn down the gift with 17 �c2 is met by 17 . . . b3+ 1 8 �c3 lLJa6! 19 'i'e3 ( 1 9 '»Yxa6 '»Yxe5+ 20 �xb3 Mab8+ wins for Black) 19 . . . Mab8 (intending . . . lLJb4) 19 �xa6 'i'a5+ 20 �d3 '»Yxa6+ 2 1 �e4 with a crazy position, where anything could happen (especially to the white king!) . 15 'i'e3 first of all aims to take con­ trol of some dark squares; secondly, White frees the bishop to develop and threatens �xc4. 1 5 . . . b 5 ! 1 6 .i1L.e2

Th e N e w M a in L in e : Bla c k fig h ts fo r c o n t r o l o f e 4

Instead 1 6 axb5 cxb5 1 7 �e2 tiJd7! tLlxd7 'li'xd7 19 �f3 l:I.ac8 20 l:I.xa7 b4+! 2 1 �c2 (Hubner points out that 2 1 �xb4 loses to 2 1 . . .l:I.b8+ 22 �a3 'i'd6+ 23 �a2 l:I.b3 24 'li'f2 l:I.fb8) 2 1 . . .'li'b5 gave White nothing in the stem game Gelfand-Hubner, Munich 1992. Even if White wins both of the black queenside pawns for his b­ pawn, and manages to swap off the queens and both sets of rooks, the resulting ending is likely to be drawn since White has the wrong-coloured rook's pawn for his bishop . Hence, all Black needs to do is to aim for a posi­ tion like 18

and he will draw Slllce the best White can achieve is

which is stalemate! The game con­ tinued 22 l:I.d1 b3+ 23 �b1 g6 24 g4 (taking f5 from the black queen) 24 . . . 'li'b8 25 l:I.a4, when 25 . . . 'li'b5 26 l:I.a7 'li'b8 would have led to a draw by repetition according to Hubner, while 25 . . . 'li'xh2!? 26 l:I.c 1 c3 27 bxc3 l:I.fd8 led to great complications. 1 6 . . . CLld7 1 7 CLlxd 7 'iVxd 7 1 8 'iVc5

A new idea. Piket had earlier played 18 �f3 against Kramnik, but Black's strategy is similar in both cases. 1 8 . a6! 1 9 Uhd 1 Yz - Yz . .

Exciting stuff, this grandmaster chess! 18 . . . a6! was in fact a strong new idea and showed (I think) good under­ standing of the position. Black has 41

Th e Sla v

four pawns for the piece, which is ample. However, his pieces are pas­ sive and he has to find a way to acti­ vate them. White is strong on the light squares (he has a light-squared bishop) , but weak on the dark squares, so I have to put my queen in contact with some dark squares. The c7-square is the obvious spot since from there the queen eyes as, eS , f4 and the h2-pawn. However, I obvi­ ously couldn't play 1 8 . . :iWc7, as 19 axbs would win for White. Therefore 1 8 . . . a6! seemed logical, and after a lit­ tle calculation I saw that it was indeed the best move. For example, after 19 . . . 'i'c7, 20 Md6? Mad8 ! 21 Mxc6 'i'f4! is extremely worrying for White.

The more active alternative, 16 . . .'!iJcS, is considered in Game 22.

1 7 1le2

17 . . . 'i'xg2 wins a pawn, but after 1 8 Mhg1 'i'xh2 19 Mxg7 h e has problems defending his second rank. 1 7 . . . �a8 1 8 g4!

The next game is intended as a cau­ tionary tale for black players, and I hope that my opponent will forgive me for using it in this way. Cynics may point out that I am demonstrat­ ing one of my rare wins from a cata­ strophic British Championship!

Game 1 7 Sadler-Ferguson

British Championship 1996 1 d4 d 5 2 c4 c6 3 lZlf3 lZlf6 4 lZlc3 dxc4 5 a4 llf5 6 lZle5 e6 7 f3 ll b4 8 e4 llxe4 9 fxe4 lZlxe4 1 0 ll d 2 �xd4

11

lZlxe4

�xe4+

12

�e2

llxd 2+ 1 3 �xd 2 �d5+ 1 4 �c2 lZla6

Black cannot hold on to the c4pawn with 14 . . . bS as 15 Md1 'i'cs 1 6 axbS axbs 1 7 'i'f3 ! , attacking f7 and the rook on a8 , is just one way of ex­ ploiting Black's mistake. 1 5 lZlxc4 0-0-0 1 6 �e3 �b8 42

Black has played rather slowly, put­ ting his king to safety in the corner, so White begins to take control. This nice move takes away the annoying check on fS from the black queen. 1 8 . . . �d 7

A novelty. Kramnik-Kir.Georgiev, Moscow Olympiad 1994, had contin­ ued 18 . . .f6 19 Mhd1 'i'g2 20 Mxd8 Mxd8 2 1 'i'xe6 with a clear advantage for White.

Th e N e w M a in L in e : Bla c k fig h ts fo r c o n t r o l o f e 4

1 9 Bad 1 tLJ b4+ 2 0 � b 1 tLJ d 5

Black has got his knight to a central outpost on dS , but it is hard to suggest another active thing for him to do. 2 1 �a3 �c7 22 a5 �b8 23 a 6 !

Softening up the black queenside. 23 . . . b6 24 �f3 .l::!. h e8 2 5 .l::!. h e 1 f6 26 �b3 �a8 2 7 h4! b5 28 tLJe3 �a5

A desperate attempt to break out, but one that is easily refuted.

Question 2: What went wrong?

Why did Black lose without seem­ ingly being able to put up any sort of fight? Answer: Black knew very little about this line and was extremely un­ fortunate that this is one of those variations where knowledge is essen­ tial: the big decisions for Black are strategical - there is almost no chance to calculate your way out of trouble. Black must understand where and when to exchange queens; where to aim to put his knight; and how to arrange his pawns. Without this knowledge, Black has little chance of success. Question 3: How does one acquire this knowledge? Answer: One must study games in this line and draw conclusions from them.

29 tZlxd 5 exd 5 30 nxe8 nxe8 3 1

Game 18

!:!'xd 5 !

K ra m n i k -Lautier

Linares 1994 1 tLJf3 d 5 2 d4 tLJf6 3 c4 c6 4 tLJc3 dxc4 5 a4 �f5 6 tLJe5 e6 7 f3 � b4 8 e4 �xe4 9 fxe4 tLJxe4 1 0 .i1Ld2 �xd4

11

tLJxe4

�xe4+

12

�e2

�xd2+ 1 3 �xd2 �d5+ 1 4 �c2 tLJa6 1 5 tLJxc4 0-0-0

It>c2 �c7 3 5 n d 7 + �b6 36 �xb5

Black has castled queenside, and not kingside. Why? First, castling queen­ side brings a rook immediately to the open d-file; second, if White ex­ changes queens, the black king is well­ placed to protect the queenside pawns from a potential attack by the white knight.

1 -0

1 6 �e5 f6 1 7 �xd 5

31

. . .

�b8

3 1 . . .exdS is met by 32 'iYxds + and mates. 32 nd6 �xa6 33 �xc6 Be 1 + 34

43

Th e Sla v

The more critical 17 'i¥e3 cussed in Games 20 and 2 1 .

lS

dis­

1 7 . . . exd 5

26 . . . g 6 27 �h e 1 �e5 28 a5 tZ:la2 ! 29 �xe5 bxe5 30 �a 1

tZ:lb4 3 1

tZ:le2

tZ:le6 3 2 .lixa6 !;la8 3 3 .li b 5 !;lxa5 34

Here Black has exchanged queens on his own terms: he has forced White to take on ds . We can conclude that the exchange of queens is only acceptable to Black if it improves the black pawn structure. Also, Black should recapture on ds with the c­ pawn: after . . . e6xdS, Black has just a 4-2 majority on the queenside; after . . . c6xdS, Black has a pawn chain of five against just two white kingside pawns on g2 and h2. Black is more likely to be able to create passed pawns and a pawn chain that will re­ strict the white pieces with the latter rather than the former.

!;lxa5 tZ:lxa5 3 5 tZ:l e 1 tZ:lb7 3 6 tZ:lf3 We7 3 7 tZ:lg5 tZ:ld6 3 8 tZ:lxh7 tZ:lf7 !

Just in time! 39 g4 tZ:lxh6 40 gxf5 gxf5 4 1 We3 e5

1 8 tZ:la3 tZ:l b4+ 1 9 W d 2 Wd7 20 !;le 1 !;le8 2 1 !;le3 b6 22 .li b 5+ Wd6 23 .lie2 a6

Preventing lLlbs+.

Black's pawns are now very dan­ gerous and White must play accu­ rately.

24 h4 !;lhf8 2 5 h 5 f5 26 h6

42 .lie6 Wd6 43 .lib7 tZ:lg4+ 44 Wf3

A typical attacking idea for White. Although White runs the risk of los­ ing this pawn, as it is now cut off from the rest of its troops, if White can get a knight to gS or a bishop to g8 . . .

e4+ 45 Wf4 tZ:le5 46 .lia6 e4 47 tZ:lg5

44

tiJd3+ 48 We3 f4+ 49 Wd4 e3 50 tiJf3 e2 51 .lib7 tiJe5 52 tZ:le 1 tZ:ld3 53 tiJf3 tiJe5 54 tiJe 1 f3 5 5 We3 We5 56 Wf2 Wd4 5 7 tiJxf3+ tiJxf3 58 Wxe2 tZ:le5 59 Wd2 tiJd3 60 .lixd5

Th e N e w M a in L in e : B la c k fig h ts fo r c o n tro l o f e 4

Iiix d5 6 1 b 3 Y:l - Y:l

So what if White does not swap queens?

squares, which makes it easy for White to blockade them with his bishop.

Game 19 Karpov-Hjartarson

Tilburg

1988

1 d4 d 5 2 c4 c6 3 t2Jf3 t2Jf6 4 t2Jc3 dxc4 5 a4 .ltf5 6 t2Je5 e6 7 f3 itb4

8 e4 iLxe4 9 fxe4 t2Jxe4 1 0 itd2 �xd4

11

t2Jxe4

�xe4+

12

�e2

ltxd2+ 1 3 'kt>xd2 �d 5+ 1 4 'kt>c2 t2Ja6

1 5 tLlxc4 0-0-0 1 6 �e 5 f6 1 7 �e3 ! ?

At the time, this was a new idea. White is claiming that he has forced his opponent to weaken the pawn structure around his king.

23 . . . �xb3+

24

�x b3

t2Jxb3

25

'kt> x b 3 .l:! d 4 26 h4! .!:!.hd8 2 7 itc4 'kt>c7 28 h5 .!:!.g4 29 h 6 !

1 7 . c5 ! ? .

.

17 . . . �b8 is considered in the next two games. 1 8 'kt>b3 t2Jb4

It's that plan again! Now the black kingside pawns are softened up and White gradually assumes complete control. Black's plan seems very logical: he is aiming to put a knight on d4.

l:!.d6 32 .l:!h6 e4 33 .l:!hxf6 h 5 34

19 gc 1 ! t2Jc6 20 'kt>a3 ! t2Jd4 2 1 t2J a 5 ! e5 2 2 �c3 ! b6 23 t2J b 3

.l:!.xg7 3 7 .l:!.f4 .!:!.g3+ 38 'kt>c2 .!:!.g2+ 3 9

This game is still the model for dealing with . . . c6-c5 and . . . lZJb4. White weakens Black's light squares by forcing all his pawns to dark-

29 . . . l:!.xg2 3 0 hxg7 .l:!.xg7 31

.l:!.cf 1

l:i:6f4 .!:!.d4 3 5 gf7+ .!:!. d 7 36 .l:!.xg7 'kt> c 3 .!:!. g 3 + 40 'kt>d2 It g 4 4 1

l:!.f7+

'kt>d6 42 'kt>e3 a6 1 -0

A really impressive game from Karpov. 45

Th e Sla v

21 �b3 ctJd5 22 �f3 �f4 23 1;;!; f 1

Game 20

�d4 24 �f2 �xf2 25 llxf2 �c8 !

S h i rov- Ba reev

Biel 1991 1 d4 d 5 2 c4 ctJf6 3 ctJc3 c6 4 ctJf3 dxc4 5 a4 lLf5 6 ctJe5 e 6 7 f3 lL b4 8 e4 lLxe4 9 fxe4 ctJxe4 1 0 lLd2 �xd4

11

ctJxe4

�xe4+

12

�e2

lLxd2+ 1 3 �xd2 �d5+ 1 4 �c2 ctJa6 1 5 ctJxc4 0-0-0 1 6 �e5 f6 1 7 �e3 �b8 1 8 1Le2 �xg 2 ! ?

The key move, preparing . . . J:rd7 to contest the second rank. 26 lLg4 f5 27 lL h 5 lld7 28 1;;!; f g2 1;;!; h d8 29 ctJe5 llxg 7 30 llxg 7 ctJf4 31

llxh7 lld 5 !

32 ctJg6 ctJxg6 3 3

lL x g 6 lld 7 !

Seeking t o reduce White's winning chances by exchanging as many pawns as possible. The less greedy 1 8 . . e5 is considered in the next game. .

1 9 1;;!; h g 1 �xh 2 20 1;;!; x g 7 ctJ b4+

Centralising the knight with tempi. The ending is equal according to Bareev. 34 �c4 1;;!; x h 7 35 lLxh 7 � d 7 36 �c5 b6+ 3 7 �c4 a 5 38 lLg6 �d6 39 lLe8 e 5 40 lL h 5 e4 4 1 lLe8 c 5 42 �c3 �e5 43 lLd7 f4 44 lLb5 f3 45 lLa6 Y:z - Y:z

This is probably fine in general for Black, but not very inspiring. I prefer 46

Th e N e w M a in L in e : B la c k fig h ts fo r c o n t r o l o f e 4

the more active plan that Black adopted in the next game.

Came 21 K ra m n i k -Ivanchuk

Linares 1994 1 CLlf3 d 5 2 d4 l2lf6 3 c4 c6 4 l2lc3 dxc4 5 a4 iLf5 6 l2le5 e6 7 f3 iLb4

8 e4 iLxe4 9 fxe4 l2lxe4 1 0 iLd2 �xd4

11

l2lxe4

�xe4+

12

�e2

�xd 2+ 1 3 �xd2 �d5+ 1 4 �c2 l2la6

1 5 4:Jxc4 0-0-0 1 6 �e5 f6 1 7 �e3

29 l2le3 �c5 30 l2lc4 b6 3 1 � b 1

Wb8! 1 8 iLe2 e 5 !

Y:, - Y:,

Neither side has an obvious way to contmue.

This i s the best set-up for Black: pawns on f6 and eS, keeping the queenside pawns where they are, while transferring pieces to the two outposts d4 and dS . Sometimes, Black will transfer the knight to d4 via cS and e6.

It is time to draw a few conclusions: 1. Black should castle queenside. 2. If the queens are to be ex­ changed, Black wants them to be ex­ changed on dS, when he will improve his pawn structure with . . . c6xds . 3 . Black's ideal outpost for his knight is on d4 and not ds . 4. Black's best pawn structure in the middle game is to place his pawns on a7, b7, c6, eS, f6, g7 and h7, since this creates two central outposts, d4 and ds . But this is not the end of the story.

Came 22 K ramni k-Shirov

1 9 93 �e6 20 b3 .nd 5 21 �b2 Rhd8 22 :gad 1 l2l b4 23 �c3 �e7 24 l2le3

Dortmund 1996

Sxd 1 25 Rxd 1 .Q.xd 1 26 iLxd 1 9 6 2 7 �d2 �c7 28 l2lc2 a 5 !

1 l2lf3 d5 2 d4 c6 3 c4 l2lf6 4 l2lc3

An important move, securing the knight on b4. If White could success­ fully play a4-aS, then the knight could become vulnerable and Black would have to be careful.

dxc4 5 a4 iLf5 6 l2le5 e6 7 f3 iL b4 8 e4 iLxe4 9 fxe4 l2lxe4 1 0 iLd2 �xd4

11

tLlxe4

�xe4+

12

�e2

iLxd2+ 1 3 �xd2 �d5+ 1 4 �c2 l2la6 1 5 l2lxc4 0-0-0 1 6 �e3 ! 47

Th e S l a v

�a2 is j ust winning for White: his king is fact perfectly safe. 20 . . . .:gS ! ? and 2 0 . . :i'g2 (threatening . . :i'g6+) are both interesting, but the onus is clearly on Black to find a reasonable continuatIOn. 20 �xd4 �xe2+ 21 ct:Jd2 gd8 22 �c5 I:Ixd 2+ 23 �b3 I:Ixb2+ 24 �a3

With hindsight, very obvious! This is the very latest idea in this line: White avoids giving Black the extra tempo .. .f7-f6, forcing Black to look for another defensive formation. 1 6 . . . eiJc 5

1 7 �e2 �xg2 1 8 ghg 1

�xh2 1 9 I:Ixg7 !

This move was originally thought Amazingly White's king is quite

to be impossible due to 19 . . . .:cI3, but 20 ':h 1 ! is very strong for White after

safe, and now it is j ust a matter of the

20 . . . .:xe3 2 1 ':xh2 or 20 . . :iUxh 1 2 1

material telling in the end.

ilxcl3 .

24 . . . I:Id2 25 I:Ig3 '>iVe4 26 .i::i. b 3 b6 27 �f8+ �b7 28 �xf7 + �a6 29 �f3

1 9 . . . I:Id4? ! ?

�xf3 30 gxf3 �a5 3 1 .i::i. f4 gd3+ 32 �b2 .i::i. h 3 33 ge 1

I:Ih2+ 34 �b3

.i::i. h 3+ 3 5 �c2 I:Ih2+ 36 �d3 I:Ih6 37 I:Ife4 c 5 3 8 I:Ixe6

The black pawns are insufficiently advanced to cause White any real problems. 3 8 . . . I:Ih3+ 39 g 1 e 3 I:Ih 1 40 I:I3e4 I:Ih3+ 4 1 �c4 g g 3 42 gh6 a6 43 I:Ixh7

gg5 44 gb7

gh5

45

ge6

I:Ih4+ 46 �d5 g b4 47 I:Ic6 gd4+ 48 �e6 gb4 49 .i::i. b 8 1 -0

An amazing attempt that j ust falls short. When position,

I

I

started to analyse this

w anted to play 1 9 . . . .:hg8,

White will win easily by attacking the black pawns from the rear with his king.

aiming to meet 20 ':xf7 with 20 . . . .:g2, but 2 1 'iYxcs ':xe2+ 22 �b3 48

':cI3+

23

Postscript: Since the first edition of

Th e N e w M a in L in e : Bla c k fig h t s fo r c o n t r o l o f e 4

this book went t o press, there have been several important games in this variation, which have focused mainly on the previously relatively neglected idea of Black castling kingside (15 . . . 0-0) instead of queenside. This is based on the fine idea of 1 5 . . 0-0 1 6 "Vje5 l:!.ab8 ! .

This wonderful idea has two points: first, Black supports the . . . b7bS advance to open up the queenside against the exposed white king; and second, Black protects the pawn on b7 against the typical White manoeu­ vre 'iWxdS followed by ctJaS or ctJd6. So far, White has even been struggling in this position! 17 .§Le2

17 Md1 ctJb4 + 18 �el ctJa2 + 19 �c2 ctJb4 + was a draw by repetition in P.Cramling-Hector, Malmo 1998. 1 7 . ctJb4+ ! 1 8 Wc3 b 5 1 9 l:!.hd 1 f6 ! . .

20 'iVg 3

�e4

21

ctJe3 ctJ d 5+

22

exercises less influence from f3 than eS, but on the other hand, White threatens the immediate Md1 , chasing the queen from dS. 1 6 . . . �c5

1 8 Wb1

Kramnik's improvement over 19 . . . 'iWgS 20 'iWg4! 'iWcs 2 1 Ma3 Md4 22 'iWhS, when Black had wasted rather a lot of time in Beliavsky-Shirov, Bel­ grade 1997. 20 l.1a3 l.1fd8 21 g4 �g 5

Black is beautifully mobilised. 22 l:!.b3 a5 23 h4 "Vjg6+ 24 Wa 1 ctJc2+ 25 Wa2 ctJ b4+ 26 Wa 1 ctJc2+

% -%

So the idea of castling kingside may, after all, be Black's best course of action. To finish this section, here are two games featuring slightly offbeat at­ tempts by White.

tLlxd 5 cxd 5

Yes, it really is as bad as it looks! I don't know how, in the game Kram­ nik-Van Wely, Tilburg 1998, White managed to hang on!

1 7 Si.e2 ctJ b4+

l:!.ad8 1 9 l:!.c 1 l:!.d4 !

Game 23 Adianto - K ramnik

London (Intel Grand Prix) 1994

23 .§Ld3 l:!.fc8+ 24 Wd2 �b4+ 25

1 d4 d5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3

'iit> e 2 bxa4 26 'it>f 1 l:!.b6 2 7 �e3 �b3

dxc4 5 a4 .§Lf5 6 ctJe5 e6 7 f3 .§Lb4

28 l:!.d2 g 6 2 9 "Vjd4 �b4 3 0 Si.e2

8 Si.g5

'i'xd4 % - %

Fighting for e4 by pinning the black knight, so that 9 e4 is now a threat. Instead 8 ctJxc4 0-0 9 �gS h6 10 �h4 was met by 10 . . . cS! 1 1 dxcS 'iWxd1+ 12 �xd1 (12 Mxd1 �c2! 1 3 Mel �xa4! 14 �xf6 gxf6 1 5 Mal �b3 16 ctJb6 ctJc6 17 ctJxa8 Mxa8 gave Black excellent compensation for the exchange in Beliavsky- Bareev, USSR 1986) 12 . . . Md8+ 13 �el ctJc6! 14 e4 �h7 1 5 �f2 ctJd7! in Akopian-Oll,

A few months after that game, these two players, Kramnik and Van Wely, continued their debate of this variation at Wijk aan Zee 1999 - but this time with colours reversed! Van Wely varied from the standard 16 i"eS with 1 6 �f3

The latest idea. The white queen

49

Th e Sla v

New York Open 1994. 1 5 .. .'�Jd7! intends . . . tLJxc5, highlighting the weakness on b3, and . . .f7-f5 activating the light-squared bishop on h7. After 16 �d3 �xc5! 17 �xc5 tLJxc5 18 �c2 f5! 1 9 exf5 tLJd4! Black stood clearly bet­ ter.

22 lLle3 0-0 23 !:!'a 1 !:!.b3 24 !:!.xa4 SLxe3+ 25 \tg3 z:!.xb2 26 SLf 1 f 5 !

White is in big danger, but some­ how he just hangs on. 27 !:!.xe6 f4+ 28 \th3 �e8 29 !:!.xe8 !:!.xe8 30 �c4+ \th7 3 1 � d 5 �e5 32 !:!.c4 !:!.h5 3 3 g3 �e5 34 !:!.c7 !:!.b6 35 8 . . . h6 9 SLh4 c5 1 0 dxc5 �a 5 ! 1 1 l'i'd4 lLlc6

1 2 lLlxc6

bxc6

1 3 e4

!:!.f7 \tg6 36 !:!.xa7 !:!.b8 3 7 !:!.d7 h5 38 �e7 l:ib2 39 �f8 fxg 3 40 !:!.xg7+

itxc 5 !

\tf6 41

1 3 . . . �g6 14 �f2 i s less good for Black.

lh - lh

1 4 �xc4 �g6

1 5 �a6 �xa6

16

SLxa6 !:!. b 8 !

!:!.f7+ \tg6 42 !:!.g7+ \tf6

Game 24 I .Sokolov-Bareev

Leon 1 995 1 d4 d 5 2 c4 c6 3 lLlc3 lLlf6 4 lLlf3 dxc4 5 a4 SLf5 6 tLle5 e6 7 g 3

White already has trouble with his queenside. 1 7 lLld 1 lLl d 7 ! 1 8 !:!.C 1 � b4+ 1 9 \tf2 lLlc5 20 �e2 lLlxa4 2 1 !:!.xc6 �c5+ 50

Th e N e w M a in L in e : Bla c k fig h ts fo r c o n t ro l o f e 4

White fights for the e4-square in a different way: by putting the bishop on g2. This also helps to dissuade Black from playing the pawn break ... c6-c5, as b7 will be hanging. 7 . �b4 8 -li g 2 �e4! . .

Forcing White to block the long diagonal, which will allow Black to play ... c6-c5 without fear of �xb7.

ctJb5 �xe5 14 �f4 �c5 1 5 ctJc7+ �e7 16 ctJxa8 ctJa6! wins for Black) 1 3 . . . ctJc6 is also good. 1 0 . . . c5 1 1 4:Ja2 �a5 1 2 d x c 5

Or 12 ctJxc4 cxd4 13 ctJxa5 �xa5 14 �xd4 ctJc6 1 5 �c4 �b6+ 16 �h 1 ctJa5!, intending . . . ctJb3 . 1 2 . . . �d 5 1 3 �xd 5 exd 5 1 4 QJxg6 hxg6 1 5 l::!. b 1 4:Jbd7 1 6 �e3 l::!. c 8 1 7 b4 cxb3 1 8 4:J c 1 b2 1 9 4:J b 3 �c3 20 .!::!. f d 1 4:J e 5 21 �d4 -lixd4+ 2 2 '!::!' x d4 b6 23 l::!. x b2

9 f3 -lig6 1 0 0-0

10 e4 c5 11 �e3 cxd4 12 �xd4 'iVxd4 13 �xd4 ctJc6 14 ctJxc6 bxc6 1 5 0-0-0 0-0-0 i s the theoretical recom­ mendation, but 16 �f1 is more pleas­ ant for White due to his superior structure and Black's inactive bishop on g6. 13 . . . ctJfd7!?, instead of 13 ... ctJc6, was my first idea in order to reactivate the bishop on g6 with .. .f7f6 and . . . �f7. However, 14 ctJxc4 f6 15 0-0-0 ctJc6 16 �f2 (intending ... ttJc6) 16 . . . �e7 17 ctJa2! wins the bishop pair, giving White a small ad­ vantage, as 17 . . . �c5 loses to 1 8 lhd7+! �xd7 1 9 �xc5, winning two pieces for a rook. In fact, 1 1 .. .�c7! is stronger: 12 ctJxc4 cxd4 (attacking the knight on c4) 1 3 �xd4 ctJc6 is fine for Black and 1 2 0-0 cxd4 13 �xd4 (13

Better was 23 cxb6 axb6 24 's'xb2 's'c4 with an equal position according to Ivan Sokolov. 2 3 . . . 4:Jc4 24 l::!. b 1 bxc5 25 QJxc5 0-0 26 4:Jd3 4:Ja3 27 l::!. b 7 4:Jc2 28 l::!. f4 l::!. f e8 29 � h 3 l::!. c 3

29 . . . ,S,b8! (I.Sokolov) gave chances for an edge for Black. 30 �f2 g5 3 1 l::!. f 5 g4 3 2 �xg4 4:Je3 3 3 l::!. x f6 4:Jxg4+ 34 fxg4 gxf6 3 5 l::!. x a7 l::!. a 3 36 4:Jf4 l::!. a 2 3 7 a 5 l::!. e 4 38 l::!. a 8+ � h 7

39

l::!. d 8 l::!. x a5 40

4:Jxd 5 l::!. x d 5 41 l::!. x d 5 l::!. x g4 42 �f3 .!::!. a 4 43 g4 �g6 44 h4 '!::!' a 1 45 h 5+ �g7 46 l::!. f 5 l::!. g 1 47 e3 l::!. f 1 + 48 �e4 l::!. g 1 49 l::!. f4 .!::!. g 3 'h - 'h

A tough endgame.

51

Th e Sla v

S u m m a ry

The sidelines do not seem to cause Black any problems, but undoubtedly the most crucial line at the moment is Kramnik's 14 Wc2 ctJa6 15 ctJxc4 0-0-0 16 'iVe3 . In general, such positions are easier to play for White than for Black. 1 d4 d5 2 c4 c 6 3 lLlf3 lLlf6 4 lLlc3 dxc4 5 a4 �f5 6 lLle5 6 . . . e6 7 f3 (0)

7 g3 game 24 -

7 . . . �b4 8 e4

8 �g5 - game 23 8 . . . �xe4 9 fxe4 lLlxe4 1 0 �d2 �xd4 1 1 lLlxe4 �xe4 1 2 �e2 �xd2+ 1 3 'It>xd 2 �d5+ 1 4 'It>c2

14 Wc3 game 1 6 -

1 4 . . . lLl a 6 1 5 lLlxc4 0-0-0 1 6 �e5 (0)

16 'iVe3 16 . . . Wb8 game 1 7 1 6 . . . ctJc5 game 22 -

-

1 6 . . . f6 1 7 �e3

17 'iVxd5 - game 18 1 7 . . . 'It>b8

17 . . . c5 game 1 9 -

1 8 �e2 (0)

1 8 . . . 'iVxg2 - game 20 18 . . . e5 - game 21

7 f3

52

1 6 �e5

18 �e2

CHA PTER FOUR The New Main Line: Black co unterattacks

1 d4 d 5 2 c4 c6 3 ltJf3 ltJf6 4 ltJc3 dxc4 5 a4 i.f5 6 ltJe5

In this chapter we consider lines in which Black does not doggedly fight on for e4, but switches his attention to the d-pawn, arguing that 6 ctJe5 has weakened White's control of d4. Re­ cent attention has concentrated on 6 . e6 7 f3 c5! ? 8 e4 cxd4, which is a specialty of both Boris Gelfand and Alexei Shirov. In the following games you will see the wildly different ways in which they handle this line!

man-Salov, 199 1 , but had not been tried since, as, in his notes, Khalifman had pointed out a continuation that seemed to equalise for Black.

. .

Game 25 Piket-Gelfa n d

Wijk aa n Zee 1996

9

. . .

.§Lg6 1 0 i.b5+ ltJfd 7

10 . . . ctJbd7?? simply loses to 1 1 �g5 . 1 d4 d 5 2 c4 c6 3 ltJf3 ltJf6 4 ltJc3

1 1 'i¥xd4 a 6 !

dxc4 5 a4 i.f5 6 ltJe5 e6

In the 1920s and 1930s, 6 . . . ctJbd7 7 ctJxc4 -&rc7 8 g3 e5 (attacking d4) was popular, but after 9 dxe5 ctJxe5 10 1£4 gd8 11 -&rc 1 �d6 12 ctJxd6+ 'ixd6 13 �g2, White stands better. He has the two bishops and Black cannot activate his queenside pawn maJonty. 7 f3 c5 8 e4 cxd4 9 .§Lxc4 ! ?

A n unusual move that was success­ ful in its first appearance in Khalif53

Th e Sla v

And this was it. If now 12 iLe2, then 1 2 . . . CDxeS 13 'ilxeS CDc6 is very good for Black, so White is forced into exchanges. 1 2 .liLxd 7 + tLlxd 7 1 3 �xd 7 + �xd 7 1 4 tLlxd 7 �xd 7

1 5 � e 2 .liL b 4 1 6

gd 1 + ctJe7 1 7 .liLf4 g h c 8 1 8 gac 1

%-% White obviously has more crucial possibilities. Who better to test the black position than Garry Kasparov?

Game 26 Kasparov-Shi rov

Dos Hermanas 1 996 1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 tLlc3 dxc4 5 a4 .liLf5 6 tLle5 e6 7 f3 c 5 8 e4 cxd4 9 exf5

�e7 13 CDxhS, 13 . . . cxb2 is discovered check, so Black wins. White can try 10 �f2!?, since after 10 . . . dxc3 1 1 'ilxds+ �xdS 1 2 CDxf7+ �e7 13 CDxhS, 13 . . . cxb2 is no longer discov­ ered check, but Black can exploit the other exposed piece in White's posi­ tion: the knight on eS. He can play 10 . . . 'ilc7!, threatening both 1 1 . . .'ilxe5 and 1 1 . . .dxc3 , as White can no longer exchange queens with 'ilxdS+. It seems that White can stop both these threats with 1 1 'ilxd4, but Black has the last laugh after 1 1 . . .iLcs, picking up the queen. Sadly Black is not com­ pletely winning after 1O . . . 'ilc7, as White can play 1 1 CDa2, attacking the bishop on b4, but after 1 1 . . .'ilxeS 12 CDxb4 'ilcS ! Black has powerful com­ pensation for the piece: two pawns and the exposed white king. I think that Black is better here. Garry played the morc natural. . . 1 0 .liLxc4 �d 6 ! ?

An amazing move, adding t o the confusion by attacking another piece. 1 1 .liLb5+ tLlc6 1 2 tLlc4 �c5?

The main line and the only real test of Black's play. 9

. . .

.liL b4 ! ?

Black could not take the knight on c3 , regaining his piece, because after 9 . . . dxc3 10 'ilxdS+ �xdS 1 1 CDxf7+ White wins a rook. However, if we imagine that it is Black's move after 9 . . . iLb4, then 10 . . . dxc3 is possible be­ cause after 1 1 'ilxdS+ �xdS 12 CDxf7+ 54

A serious and, in such a sharp posi­ tion, fatal mistake. Black could simply have retreated with 12 . . . 'ild7 when,

Th e Ne w M a in L in e : Bla c k c o u n t e ra t t a c k s

due to the threat o f . . . d4xc3, White probably has nothing better than to repeat moves with 13 ctJe5 �d6. 1 3 .1Ld2!

Black cannot regain the piece now, as 13 . . . dxc3 14 bxc3 �a5 loses a piece to 15 ctJxa5 . Obviously Shirov did not miss this move; but I believe that he overlooked something extremely cunning later on. 13 0-0 1 4 ctJa2 ..txd2+ 1 5 �xd 2 CiJ e 7 . . .

Black's is threatening ... a7-a6, win­ nin g the bishop, while he can also try ... LZJxf5, intending . . . ctJe3 . White seems to have problems but . . . 1 6 �b4 !

Forcing the exchange of queens as 16.. .'i'xf5 allows 17 �xe7. 16

. . .

�xb4+ 1 7 ctJxb4 a6 1 8 ctJ b 6 !

White does not lose the bishop af­ ter all! The rest is easy for Kasparov. 1 8 axb5 1 9 ctJxa8 �xa8 20 fxe6 bxa4 21 exf7+ �xf7 22 �d2 �e6 23 J:i.hc 1 �d6 24 b3 b 5 25 bxa4 bxa4 26 .8.c4 ctJf5 27 ctJc2 ctJd7 28 gcxa4 .8.xa4 29 .l:1.xa4 ctJb6 30 ctJxd4 1 -0 . . .

A

fine game by Kasparov, but not

one that refutes Black's idea. Let us take another look.

Game 27 Gelfand-Shirov

Dortmund 1996 1 d4 d 5 2 c4 c6 3 ctJc3 ctJf6 4 ctJf3 dxc4 5 a4 ..tf5 6 ctJe5 e6 7 f3 c5 8 e4 cxd4 9 exf5 ..tb4 1 0 ..txc4 �d6

Shirov could not resist trying this idea a second time, but this time he is convincingly mauled. I do wonder why Black has been avoiding 10 . . . dxc3 1 1 �xds+ �xds 12 0-0 (12 ctJxf7+ �e7 1 3 ctJxhs cxb2+) 12 . . . cxb2 13 �xb2 �e7 14 fxe6 fxe6.

White's two bishops, the weak (but extra) pawn on e6 and the slightly exposed king on e7 obviously offer compensation for the pawn, but I don't see an advantage for White. Black's bishop on b4 prevents his op­ ponent from playing Me 1 and ganging up on e6, so 15 ctJd3 is tempting: 15 . . . �d6 is met by 16 Mfe 1 while 1 5 . . . �a5 16 �a3+ is also sub-optimal, as Jon Speelman would say! However, 15 . . . McS ! , attacking the bishop on c4, is the best defence: 16 ctJxb4 Mxc4 is 55

Th e Sla v

good for Black and 1 6 iLxe6 �xe6 17 tDxb4 leaves an equal position. Fi­ nally, 16 iLb3 is met by 16 . . . tDc6, pro­ tecting the bishop, when 17 tDxb4 tDxb4 18 Mfe 1 Mc6! (the point of 1S . . . Mc8) 19 iLa3 as ! , intending . . . �f7, is fine for Black. I feel that the onus is on White to demonstrate more than just sufficient play for the pawn.

Suddenly it is Black's king that is in danger! 1 8 . . . �xh 2 1 9 5lxc6 .l::!. a c8 20 �xf6 �xf4+ 21 �c2 g6 22 .l:!.df 1 �h 2+ 23 .l::!. g 2 �h3 24 fxg6 fxg 6

1 1 5l b 5+ tLlc6 1 2 5lf4 !

25 .i:!xg6+ hxg6 26 �xg 6+ �h8 27 .l::!. h 1 !

Winning the queen. The game is over. 27 . . . .i:!f2+ 28 �b3 .i::!. b 8+ 29 �a3 1 -0

This prevents the capture 1 2 . . . dxc3 due to 13 tDxc6 cxb2+ 14 tDxb4+! , when, thanks t o the great strength of the discovered check, White wins the whole house! 1 2 . . . 0-0 1 3 tLlxc6 �xf4 14 �xd4 !

So it seems as if, despite these two reverses, Shirov's 9 . . . iLb4 may well be just about playable. Let us now take a look at the more restrained continua­ tion 9 . . . ctJc6.

5lxc3+ 1 5 bxc3 �g 5 1 6 f4 �xg 2 1 7

Came 28

O-O-O ! bxc6 1 8 .i:! hg 1 !

l I Iescas-Gelfand

Dos Hermanas 1 996 1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 tLlc3 dxc4 5 a4 5lf5 6 tLle5 e6 7 f3 c 5 8 e4 cxd4 9 exf5 tLlc6 see

follo wing diagram

The older and more solid move. 1 0 tLlxc6 bxc6 5lxc4

56

11

fxe6 fxe6

12

Th e N e w M a in L in e : B la c k c o u n t e ra t t a c k s

gxbS �xbS 20 93 lLa3 2 1 g b 1 .!:!.xb 1 22 <;i>xb 1 lL c 5 23 lLd3 h6 24 c4 lL b4 25 cxd 5 Y2 Y2 -

Game 29 Van der Sterren-Petu rsson

San Bernardino Open

1 992

1 d4 d 5 2 c4 c6 3 ttJc3 ttJf6 4 ttJf3 dxc4 5 a4 lLf5 6 ttJe5 e6 7 f3 c5 S e4 cxd4 9 exf5 ttJc6 1 0 ttJxc6 bxc6 11

1 2 �e2 i s considered i n Game 30, while more adventurous players may like to consider the alternative 12 ltJa2 d3, keeping the piece and hoping to unravel later with g2-g3, �g2 and 0-0, although Black's counterplay is very dangerous! 12

. . .

fxe6 fxe6

1 2 lLxc4 dxc3

13

bxc3

A different move-order that should be met by 13 . . . �xd1+ 14 <;t>xd1 <;t>d7, with the same ideas as in Illescas­ Gelfand above. 1 3 . . . '>il\Va 5 ? 1 4 '>il\Ve2 ! !

dxc3 1 3 '>il\VxdS+ <;i>xdS 1 4 bxc3

This rook sacrifice is a magnificent concept! White is a bit better in this ending, since he has a slightly better pawn structure (fewer pawn islands) and the two bishops, but Black's pieces are active. 14 .':tJd 5 1 5 <;i>d2 lL d 6 1 6 <;i>c2 <;i>d 7 ! . .

The king protects both weak pawns and helps to cover the only open file on the board: the b-file. 1 7 lLd2 .§.hfS 1 S gab 1

gabS

19

1 4 . . . �xc3+ 1 5 <;i>f 1 �xa 1 1 6 '>il\Vxe6+ <;i>dS

16 . . . �e7 loses to 17 �xc6+ <;t>f8 1 8 �xa8 ctJe8 19 <;t>e2 �xa4 20 �d5 ctJd6 2 1 �d3 with a crushing attack (Petursson) . 1 7 <;i>e2 ! !

This quiet move, allowing the rook to j oin in the attack, justifies White's brilliant idea. 57

Th e Sla v

1 7 . . : �xa4 1 8 .l:!.d 1 + �xd 1 + 1 9 �xd 1 � e 5 20 �f7 ge8 2 1 �xg 7 ctJd 7 22 iLf7 .l:!.f8 23 iLe6 ctJf6 24 �b7 .l:!.e8 25 �xa8+ � e 7 26 iLf4+ 1 -0

Came 30

This ending is somewhat more awkward for Black since his king is a little more open. Nonetheless, it is surprising how quickly his position goes downhill. 1 8 . . . .l:!.hd 8 ? !

Topalov-G elfand

Dos Hermanas 1 996

Black must try and activate his knight: ls . .ebds! is stronger, when 19 �xg7 .l:IhgS ! regains the g2-pawn. Af­ ter 1 9 .l:Iac 1 , 19 . . . .l:IheS is best, prevent­ ing 20 �xds cxds 2 1 .l:Ic7 + due to 2 1 . . .�d6+, a discovered check that wins the rook! .

1 d4 d 5 2 e4 e6 3 ctJf3 ctJf6 4 ctJe3 dxe4 5 a4 iLf5 6 ctJe5 e6 7 f3 e 5 8 e4 exd4 9 exf5 ctJe6 1 0 ctJxe6 bxe6 1 1 fxe6 fxe6 1 2 �e2 ! ?

An interesting novelty that aims for a slightly different endgame.

1 9 :l:!hd 1 :l:!ab8 20 g3

This places the g-pawn on a pro­ tected square and thus prevents the black knight from moving. Black now has serious problems which he is unable to overcome. 20 . . . h 5 2 1 gd4 �a5 22 �a3+ �e8 23 :l:!ad 1

.l:!.xd4 24 .l:!.xd4 .l:!.b 1

25

�d3 ge 1 + 26 �f2 �f7 27 �e5 :l:!a 1 28 �e4+ �e8 29 �d3 �f7 30 �xa 7 .l:!.a2+ 3 1 �f 1 :l:!a 1 + 32 �g2 ga2+ 33 �h3

N ow White is just winning. 33 . . . ga3 34 f4 �e 1 35 �c5 .l:!.e3 36 1 2 . . . dxe3 1 3 �xe6+ �e7 1 4 iLxe4

�e4+ �e8 3 7 � b4 Be 1 3 8 �xe 1

�xe6+

:l:!xe 1 39 a5 Ba 1 40 a6 �e7 4 1 gd2

15

�xe6

exb2

16

iL b4+ 1 7 �e2 �e7 1 8 �e4 58

iLxb2

1 -0

Th e N e w M a in L in e : Bla c k c o u n t e ra t t a c k s

Both 6 . . . e 6 7 f3 �b4 and 6 . . . e 6 7 f3 are popular counterattacking sys­ tems at all levels of play. However, Black can also play more slowly, aim­ ing to break out from a cramped posi­ tIOn. c5

Game 3 1 Kramni k-Short

Novgorod 1 994 tiJf3 d 5 2 d4 tlJf6 3 c4 dxc4 4 tZJc3 c6 5 a4 .iLf5 6 tlJe5 tlJbd7 7 tilxc4 tlJb6 8 tlJe5 a 5

Preventing a4-as-a6, breaking up the black queenside. 8 . . . e6 is consid­ ered in the next game, while 8 . . .'�Jbd7, still seeking the exchange of knights, was crushed by 9 iVb3 ! ctJxe5 10 dxes CLlg4 1 1 iVxb7 ctJxes 12 f4 ctJg6 13 e4 _�d7 14 fs ctJes 15 �f4 f6 16 �xes [xeS 17 .s,dl in Kasparov-Timman, Riga 1995 .

The standard plan of development: the queen supports e3-e4 from e2, leaving the rook to cover d4 from d l . 1 4 . . . . tlJfd 7 1 5 tlJd3 ! ?

1 5 ctJxd7 ctJxd7 16 e4 is also slightly better for White. 1 5 . . . �e7 1 6 e4 e 5 1 7 d 5 gfd 8 1 8 .iLe 3? !

A mistake, allowing Black to weaken the white queenside. 18 �d2 would have kept an edge according to Kramnik. 1 8 . . . .iL x c 3 1 9 b x c 3 cxd 5 20 exd 5 tlJc4 ! ?

9 g3!?

A novelty. 9 f3 ctJfd7 10 ctJxd7 CLlxd7 1 1 e4 �g6 12 �e3 e6 13 �c4 �b4 14 0-0 is normal, with a slight advantage for White. 9 . . . e6 1 0 .iLg2 .iLb4 1 1 0-0 0-0 1 2 e3 h6 1 3 �e 2 ! .il.. h 7 1 4 J::i: d 1 !

I think that Short may have missed White's next, but this is actually a good move! 21

.iLxh6 gxh6 22 �g4+ �g 5 23 59

Th e Sla v

�xc4 �ac8 24 � b 5 l::t x c3

1 2 �xf2 ik'xd4+ and . . . ik'xeS.

Black has good counterplay in this murky position.

9 . . . a 5 1 0 e4

31 ttJf4 1Le4 32 �5d2 �c5 3 3 �xc 5

10 g4 ttJfdS! 1 1 h4 (1 1 gxfS ik'h4+ 12 �d2 ik'f4+ wins back the piece) l 1 . . .f6 1 2 gxfS fxeS (Ruzele) is extremely un­ clear.

ttJ x c 5 34 1Lxe4 .l:!.xe4 3 5 ttJ d 5
1 0 . . . 1Lg6 1 1 1Le3 1Lb4 1 2 1Le2 O-O ? !

3 6 ttJxb6 ttJxa4 37 ttJxa4 .6!xa4 38

1 2 ... ttJfd7 is better since 13 ttJxg6 hxg6 gives Black play on the h-file, while 13 ttJxd7 ttJxd7 transposes to S . . . aS 9 f3 .

2 5 ttJe 1 b 6 26 d 6 e4 2 7 � d 5 �f6 28 J:!ad 1 e3 29 fxe3 �xe3 3 0 ttJd3 �c3

J:! d 5

White has a tiny edge in the end­ game but he is unable to make any­ thing of it.

1 3 0-0 ttJfd 7

4 1 J:! a 5 a3 42 g4 'it>g7 43
Now White can take the two bish­ ops.

44 h 3 h4 'it>g6 46 .l:!.a8

1 4 ttJxg 6 !


1Lg 1 �fd8 1 7 �b3

38 . . . l:i.d7 39 h 3 a4

hxg6

1 5
�e7

16

Game 32 Ruzele-Thorsteins

Lyon (European Club

Cup) 1994

1 d4 d 5 2 c4 c6 3 ttJc3 ttJf6 4 ttJf3 dxc4 5 a4 1Lf5 6 ttJe5 ttJ b d 7 7 ttJxc4 ttJb6 8 ttJe5 e6 ! ?

White has effortlessly obtained a wonderful version of the old main line with 6 e3. 1 7 . . .c5

18

ttJ a 2 !

cxd4

19

ttJxb4

�xb4 20 �xb4 axb4 2 1 a5 d 3 22 1Lxd3 ttJe5 23 1Lb5 ttJbc4 24 f4 ttJd6 25 1Le2 ttJc6 26 1Lb6 .l:!.e8 27 e5 ttJc8 28 1Lf3 ttJxb6 29 axb6 ttJe7 30 9 f3

1Lxb7 �ab8 3 1 �a7 ttJc8 3 2 1Lxc8

A sensible reply. 9 as ttJbds (threatening . . . ttJb4 or . . . .\tb4) 10 a6!? is consistent but very risky, while 9 g3 is met by 9 . . . .\tb4 10 .\tg2 ttJe4 when the natural 1 1 .\td2 loses to 1 1 . . .ttJxf2!

�exc8 3 3 �c7 nxc7 34 bxc7 .l:!.c8

60

35 .l:!.c 1
Model strategy by White.

Th e Ne w M a in L in e : Bla c k c o u n t e ra t t a c k s

S u m m a ry

6 . . . e6 7 f3 cS should definitely be studied by Black players. I particularly like Shirov's handling of the line with 8 e4 cxd4 9 exfS �b4. 9 . . .':tJc6 is for the calmer players amongst you who don't mind taking on a slightly worse end­ ing. Probably it is important to choose the right opponent: 9 . . .':tJc6 will be ideal against an impatient attacking player, while 9 . . �b4 would unsettle a more positionally inclined player. If you prefer the more solid 6 . . . ctJbd7 7 ctJxc4 ctJb6, and don't mind the slightly cramped positions that arise from this line, then Thorsteins's 8 . . . e6 looks like a good move-order, since it avoids Kramnik's 9 g3 . .

1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 iLf5 6 ctJe5 6

. . .

e6

6 . . . ctJbd7 7 ctJxc4 ctJb6 8 ctJeS (D) 8 . . . aS - game 3 1 8 ... e6 game 32 -

7 f3 c 5 8 e4 cxd4 9 exf5 (D)

9 �xc4 game 25 -

9

. . .

ctJc6

9 . . . �b4 10 �xc4 'iVd6 1 1 �bs + ctJc6 12 ctJc4 game 26 12 �f4 game 27 -

-

10 ctJxc6 bxc6 1 1 fxe6 fxe6 (D) 1 2 iLxc4

12 'iVe2 game 30 -

12

. . .

dxc3

13 'iVe2 game 28 13 bxc3 game 29 -

-

8 ctJe5

9 exf5

l '

. .

. fxe 6

61

CHA PTER FIVE The S m yslov Variation

1 d4 d 5 2 c4 c6 3 ttJf3 ttJf6 4 ttJc3 dxc4 5 a4 ttJa6

The move 5 . . . ctJa6 was originally an idea of Emanuel Lasker; and it was later taken up and played with success by another World Champion Vassily Smyslov. Recently Ivanchuk and Short have also used it to good effect. Question 1: 5 . . ctJa6 looks a little strange. What does it do? Answer: 5 a4 (preventing . . . b7-b5) has weakened the b4-square. In the main lines with 5 . . . itf5 6 e3, Black puts a bishop on b4 and develops his queen's knight to d7; here, Black leaves his bishop on e7 to keep b4 free for the knight. Black's light-squared bishop will now go to g4 to put pres­ sure on the d4-square. Question 2: So what difference does this make? Answer: In the 6 e3 lines, Black's bishops on f5 and b4 combine to pre­ vent White from easily achieving e3e4. In this line, Black exerts virtually no pressure on e4, and very little on d4, which means that White pretty much has the centre to himself. Question 3: Well that doesn't sound very promising for Black, does it? What am I supposed to do as Black? .

62

Answer: Smyslov's style as Black is perfectly reflected in this system: he is prepared to accept a slight space dis­ advantage and will just place his pieces on good squares where they coordi­ nate well with each other. Since Black's position is very solid, the op­ ponent will not be able to launch a sudden attack and Smyslov will pa­ tiently unravel, gaining space little by little until he frees himself.

Game 33 Ivanchuk -Smyslov

Tallinn (rapidplay) 1 996 1 d4 d 5 2 ttJf3 ttJf6 3 c4 c6 4 ttJc3 dxc4 5 a4 ttJa6 6 e4

Very straightforward play.

Th e S m y s / o v Va ria tio n

6

. . .

i!.. g 4 7 i!.. x c4 i!.. x f3

7 . . . e6 is dealt with in the next game.

A different approach which leads to a complicated middlegame.

8 gxf3 e6 9 i!.. x a6 bxa6

8 i!.. e 3 tiJb4 9 a 5

Black has no problems here. First, he has exchanged two sets of minor pieces, and such exchanges always help the player with less space, since it means that there are fewer pieces in a confined area; second, Black's doubled a-pawns give Black the b-file on which to activate his major pieces and attack the vulnerable white queenside. Black would be much less active if his a6pawn were on b7! Certainly Ivanchuk is happy to exchange queens and es­ cape with a draw.

A typical idea from White, aiming to prevent either . . :iVaS, activating the black queen, or . . . a7-aS, cementing the knight on b4. Black has to be a little careful that this knight, protected only by the bishop on e7, does not get cut off from the rest of his army. 9

. . .

i!.. e 7

9 . . . �xf3 , forcing 10 gxf3 (10 'iVxf3 ctJc2+) , was still possible but Black prefers natural development. 1 0 i!.. e 2 0-0 1 1 0-0 b 5 !

1 0 \We2 a 5 1 1 \Wc4 Mc8 1 2 Mg 1 g6 1 3 r;t>f 1 i!.. g 7 1 4 �c5 �b6 1 5 �xb6 axb6 1 6 i!.. e 3 0-0 1 7 MC 1 tiJ d 7 Y:z - Y:z

If Black delays exchanging on f3 , a more complex situation arises, as we shall see in the next game.

Game 34 N ovikov-Greta rsson

Berlin Open 1995 1 d4 d5 2 c4 c6 3 tiJf3 tiJf6 4 tiJc3 dxc4 5 a4 tiJa6 6 e4 i!.. g 4 7 i!.. x c4 e6

This far from obvious move is the black plan for survival in these middlegames. 63

Th e Sla v

Question 4: What is the pomt of

this move? A nswer: The main idea is that Black

weaker and hence it is easier for Black to achieve the . . . c6-c5 break. 1 6 ttJe 1 1£. x e 2 1 7 ttJxe2 J:t a c 8 1 8

gains just a little more space for his pieces - remember what I said about Smyslov patiently improving his posi­ tion, taking extra territory little by little. The other point is that Black would like to strike at the white cen­ tre with . . . c6-c5, but first he needs a reasonable square for his queen: he can't put it on the c- or d-files, since after . . . c5xd4 these files will be opened and the queen will be in the firing line of white rooks on d1 and c 1 . 1 1 . . .b5 frees b7 for the black queen, where it is absolutely safe. A nice bonus is that after . . . c6-c5, the black queen will join with the knight on f6 in attacking the e4-pawn.

z::!. c 3 c 5 ! 1 9 dxc5 ttJa6 ! !

1 2 �b3 �c7 1 3 Rfc 1 �b7 1 4 � g 5

�d 5 25 ttJc3 �d4 26 �f3 �e 5

R f d 8 1 5 .i1Lxf6 gxf6

Black's control of the dark squares gives him good chances.

A very neat idea. 19 . . . 1lLxc5 would have lost to 20 's'xc5 's'xc5 2 1 iVxb4. 20 ttJd3 ttJxc5 2 1 ttJxc5 axc5 22 axc5 �xc5 23 �c3 �xe4 24 �xf6

27 g3 b 4 2 8 Rd 1 �xd 1 + 2 9 ttJxd 1 �d4 30 �d 3 �g7 3 1 b3 h6 32 a6

%-% This game is a model illustration of Black's middlegame strategy in this vanatlOn. �c5 33 �f 1 �f6

N ow we move on to look at 6 e3.

Game 35 Benz-Gretarsson

A forced recapture as 15 . . . 1lLxf6 loses a piece to 16 iVxb4. The weaken­ ing to Black's kingside is not too seri­ ous, however, since White has no pieces in that area. Moreover, without his dark-squared bishop, White loses a lot of control over the central dark squares, which means that d4 is 64

Oberwart Open 1996 1 d4 d5 2 ttJf3 ttJf6 3 c4 c6 4 ttJc3 dxc4 5 a4 ttJa6 6 e3

The most solid option and proba­ bly the best move. White does not give Black the chance to double his f­ pawns with . . . 1lLxf3 .

Th e S m y s l o v Va ria tio n

A very sharp reply, but I would be intrigued to discover what Helgi had in mind against the ECO recommen­ dation of 14 iLf4! (taking c7 away from the queen) 14 . . . 'iYaS 15 ctJd2!, intending ctJb3 to harass the queen some more. 1s . . . cxd4 16 ctJb3 'iYb6 1 7 a s i s not nice for Black and the reck­ less lS . . . ctJc2 loses a piece to 16 ctJb3 'iYb4 17 ctJa2 'iYxa4 18 ctJc3 ! (18 'iYxc2 iLxe4 causes some problems) 18 . . . 'iYb4 19 Ra4 'iYb6 20 'iYxc2. 6 . . . 1Lg4 7 1Lxc4 e6 8 h 3 1L h 5 9 0-0

1 4 . . . exd 5 ! 1 5 e 5 d4 ! ? 1 6 exf6 1Lxf6

iLlb4 1 0 �e2 !

1 7 1Lf4

A typical manoeuvre in queen's pawn openings: the queen moves to e2, supporting the e4 push, while the rook is played to d 1 , supporting the d4-pawn and discouraging . . . c6-cS due to the opposition of the rook to the black queen on d8 . 1 0 . . . 1Le7 1 1 J:::!. d 1 0-0 1 2 g4 1L g 6 1 3 e4 c5 ! ?

Question 5: What is going on? A nswer: As compensation for the

An unusually active move at this of the Smyslov variation, but Helgi Gretarsson, a fanatic of this variation, has an interesting idea in mind. The more restrained 13 . . .':tJd7 is considered in the next two games.

piece, Black has two pawns, a strong centre and a tempo on the queen with . . . Re8 . Unfortunately, I don't think that this is quite enough; and this is almost entirely due to the bad placing of the bishop on f6, which takes away a brilliant square for the black queen. Perhaps Black could try 1 S . . . Re8 ! ? to meet 16 exf6 (16 iLbs ctJc6) with ei­ ther 16 . . . iLd6 or 16 . . . iLf8 , intending to recapture on f6 with the queen, though I would be the first to admit that it all looks a bit speculative!

14 d 5 ! ?

1 7 . . . J:::!. e 8 1 8 �f 1 a6

stage

65

Th e Sla v

To prevent LLlb5 .

8 h 3 iL h 5 9 0-0 ctJb4 1 0 �e2 iLe7 1 1 tId 1 0-0 1 2 g4 iLg6 1 3 e4 ctJd7 ! ?

1 9 .a.d2? !

White starts to go wrong around here and drifts very quickly into a lost position. 19 LLla2! LLld5 (19 . . . LLlc6 20 �e l ! beginning to exchange pieces) 20 ilg3 , intending �e l , would have given White the better chances.

Black anticipates the threat of llJeS and f2-f4, intending f4-f5 to trap the bishop on g6. 1 4 ctJe5 ! ?

axb5 22 iL x b 5 '>&e4 23 iLxe8 '>&xf4

A very double-edged decision. White allows his central pawns to be doubled, but also frees the f-pawn to advance.

24 .a.e 1

1 4 . . . ctJxe 5 1 5 dxe5 �a5 1 6 f4 .l::!. a d8

1 9 . . . '>&d 7

20 �g 2?

'>&c6

21

ctJb5

h6 25 iL b 5 iLe4 26 .a.xe4

�xe4 2 7 iLc4 �f4 28 b3 .)de8 29

1 7 iLe3 h 6 !

�g 3 '>&xg 3+ 3 0 fxg 3 ctJc6 31 iLd5 ctJa5 3 2 b4 .a.d8 3 3 iLe4 cxb4 34 ctJe 1 b3 3 5 iLd3 iL g 5 36 tIb2 tIc8 3 7 �f 1 g 6 3 8 �e2 iLc 1 39 tIb 1 b2 40 ctJf3 ctJb3 0 - 1

The main line for Black i s consid­ ered in the next game, probably the finest blindfold game ever played. I wish I could play this well in normal chess!

Game 36 K ra m n i k-Ivanchu k

Monte Carlo (blindfold) 1 996

Making an escape square for the bishop. 1 8 �g2 iL h 7 1 9 .a.xd8 .a.xd8 20 .a.d 1 g5!

1 ctJf3 d 5 2 d4 ctJf6 3 c4 c6 4 ctJc3 dxc4 5 a4 ctJa6 6 e3 iLg4 7 iLxc4 e6 66

2 0 . . . a 6 i s considered i n the next game.

Th e S m y s l o v Va ria tio n

2 1 :i::!. x d8+ �xd8 22 �d 2 gxf4 23 ixf4 iL b 6 24 iL b 5 ! !

Black must give perpetual check due to the threat of mate on hS. 24 �xh6 allows 24 . . . iVxeS, so White sacrifices a bishop to keep the queen boxed in on as . 24

. . .

cxb5

25 iLxh6 � c 5

26 �d7

Game 37 Kramni k-Short

Moscow {Intel Grand Prix} 1996

ig6 2 7 �c8+ �h7 28 iLg5

With his threat of �f6 and iVhs mate, White just seems to be winning, but now it is Black's turn to sacrifice a piece, this time to free his queen.

1 tLlf3 d 5 2 d4 c6 3 c4 tLlf6 4 tLlc3 dxc4 5 a4 tLla6 6 e3 iLg4 7 iLxc4 e6 8 h 3 iLh5 9 0-0 tLl b4 1 0 �e2 �e7 1 1 :i::!. d 1 0-0 1 2 g4 iLg6 1 3 e4 tLld7 1 4 tLle5

28

. . .

tLl d 5 ! ! 29 �xc5

Not 29 exdS when 29 . . iVb4! launches a powerful counterattack. .

29

. . .

tLlxc3 30 �xc3 �xa4 3 1 �g3

�xe4 3 2 iLf6 b4 3 3 �c8 �e 1 + 34 �f4 �f2+ 35 � g 5 �d2+ 36 �h4 �h6+ 37 �g3 �e3+ Yz - Yz

Kramnik obviously believes in this continuation for White, but it seems a little hasty to me. Since Black is threatening little in the centre, a sen­ sible move like 14 �f4, taking c7 away from the black queen and 67

Th e Sla v

intending perhaps h3-h4-hS, makes more appeal to me.

'iVxe 5 30 'iVb6 ctJd4 3 1 .1Ld 1 .1L g 8 !

1 4 . . . ctJxe5 1 5 dxe5 �a5 1 6 f4 h 6 1 7 �g2 J::r a d8 1 8 .1Le3 .1L h 7 1 9 J::r x d8 J::r x d8 20 t1.d 1 a 6 !

The new idea, activating Black's queenside majority.

3 2 .i.f3 f6 ! 3 3 ctJ e 2 ctJxe2 34 .i.xe2 �xe4+ 3 5 .i.f3 �c2+ 36 �f2 'iVxf2+ 37 �xf2 a5 38 fxe6 .i.xe6 !

2 1 .1L b 3 �h8 !

So that the bishop on h7 can reacti­ vate itself by means of . . . iLgS , . . .f7-f6 and . . . iLf7! This position would not be to everyone's taste, but Short wins a mce game. 22 J::r d 2 b 5

23 axb5 cxb5 24 f5

ctJc6!

Yes it's free! White will not be able to cope with the two potential outside passed pawns. 39 �e3 b4 40 .1Ld 1 f5 4 1 �f4 fxg4 42 hxg4 �g8 43 �e5 �f7 44 �d6 �f6 45 b3 g5 46 �c5 �e5 4 7 �b5 �d4 48 �xa5 �c3 49 �a4 .1L d 5 0 - 1

Stressing the new weakness on eS. 25 J::r x d8+ � x d 8 26 .i.f4 .i. c 5 2 7 .i.e3 '�b6 28 .1Lxc5 ·�xc 5 29 'iVf2 68

Kramnik's approach with CDeS is rather impatient. A quieter method is demonstrated in the following game, the last of a match between France's 13-year-old star Etienne Bacrot and ex-World Champion Vassily Smyslov,

Th e S m y s / o v Va ria tio n

whose name this variation bears. The match score was a rather crushing 5- 1 , and one player was made t o look vastly inferior in the endgame. But not the player one might have ex­ pected!

Game 38

f4-fs or h3-h4-hs with a clear advan­ tage. Black must play . . . ctJd7 to pre­ vent ctJeS, either before or after . . . b7bS, with a typicaI S . . . ctJa6 position. 1 2 . . J�� c 8 ? ! 1 3 .:iLb3 c 5 ?

This is excessively active from Black at this early stage. 1 4 CL:l b 5 !

Bacrot-S myslov

Albert (sixth match game) 1 996 1 d4 d 5 2 c4 c6 3 CL:lc3 CL:lf6 4 CL:lf3 dxc4 5 a4 CL:la6 6 e3 .:iLg4 7 .:iLxc4 e6

8 0-0 .:iLe7 9 �e2 CL:lb4 1 0 J::i. d 1 0-0 1 1 h3 .:iLh5 1 2 a 5 ! ?

This fine move threatens ctJxa7 and d4xcS followed by ctJd6. 1 4 . . . .:iLxf3 1 5 gxf3 a6 1 6 dxc5 CL:lbd5 1 7 CL:ld6 .:iLxd6 1 8 cxd6 �xd 6 1 9 ga4 .:tIc5 20 f4 �c6 2 1 .:iLd2 �b5 22 �xb 5 l::!. x b5

We have already seen White's idea in Novikov-Gretarsson. Our first thought should be therefore to im­ plement the plan of . . . b7-bS, followed by the transfer of the queen to b7: 12 bs 13 �b3 "Wic7 14 e4 (14 g4 �g6 15 ctJeS! ? , intending a quick h3-h4-hS trapping the bishop) 14 . . . "Wib7. It is obvious that White is better prepared for his opponent's plan than in the above game. The rook covers d4 from d 1 , while the queen on e2 both pro­ tects e4 and attacks bS, making . . . c6-cS more difficult to achieve. After 15 g4 itg6 16 ttJeS White intends either f2. . .

The ending is very nasty for Black, since his knights have no outposts. 23 ga3 CL:le4 24 .:iLe 1 CL:ldf6 25 J::i. c 1 &i.d8 2 6 f 3 CL:ld6 2 7 .:tId 1 CL:lfe8 28 69

Th e Sla v

�a4 gd 5

29 gxd 5 exd 5

30 gd3

37

�f2

CiJxc7

38

gb3

CiJe6

39

CiJc7 31 � b4 CiJ d b 5 3 2 � c 5 f5 3 3

.a.xb7+ �f8 40 .a.b8+ �e7 41 .a.b7+

�b6 gd6 34 e4 !

.a.c7 42 .a.xc7+ CiJxc7 43 �c6 d4 44 b4 1 -0

This allows White to make a mas­ sive stride forwards with his e-pawn as 34 . . . dxe4 loses a piece to 35 Mxd6. 34 . . . �f7 3 5 e5 .a.c6 36 �xc7 .a.c 1 +

70

A fine vlCtOry for the young Frenchman, crowning an amazing match result.

Th e Sm y s / o v Va ria tio n

S u m m a ry

Theoretically, S . . . ctJa6 is doing well for Black and if a system has been played by Smyslov, Ivanchuk and Short then it must have some merit! If you don't mind playing slightly cramped positions, then it could be the system for you. 6 e3 is the most critical test. 1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 ctJa6 6 e4 (D)

6 e3 iLg4 7 iLxc4 e6 8 h3 iLh5 9 0-0 ctJb4 10 'Wie2 iLe7 1 1 Md1 0-0 12 g4 iLg6 13 e4 (D) 1 3 . . . cS game 35 13 . . . ctJd7 14 ctJeS ctJxe5 15 dxeS 'Wia5 16 f4 Mad8 17 iLe3 h6 18 �g2 iLh7 19 Mxd8 Mxd8 20 Md1 20 . . . g5 game 36 20 . . . a6 game 37 12 as game 38 -

-

-

-

6

. . .

ilLg4 7 ilLxc4 (D)

7 iLxf3 - game 33 7 . . . e6 game 34 . . .

-

6 e4

1 3 e4

7 ilLxc4

71

CHA PTER SIX The Bronstein Variation

1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 tLlc3 dxc4 5 a4 jLg4

Like Smyslov's 5 .. .'�Ja6, 5 . . . iLg4 is a variation that has been around for a long time without ever gaining wide­ spread popularity. Question 1 : What is the point of 5 . . . iLg4? A nswer: 5 . . . iLg4 looks to delay e3e4 by putting pressure on the d4pawn. White should avoid 6 e4 as af­ ter 6 . . . e6 7 iLxc4 iLb4 (threatening . . . ctJxe4) , he has problems holding his centre. Therefore 6 ctJe5, gaining a tempo on the bishop, is almost always played. Now after 6 . . . iLh5 White still cannot play 7 e4, as this would allow 7 . iLxd l ! Question 2 : What plans does White have? A nswer: Since White can take the c4-pawn at his leisure, there is no need for him to hurry with ctJxc4. His two most dangerous plans both aim to exploit the slightly precarious posi­ tion of the bishop on h5: a) 7 f3 , which blocks the h5-dl di­ agonal and threatens to achieve e2-e4, while supporting g2-g4 iLg6, h2-h4, intending to trap the bishop. b) 7 h3, intending to gain space on

the kingside with g2-g4 and develop the bishop to g2. This variation is so complicated, however, that a whole book would be needed to explain its ramifications! I will do my best, but I'm afraid you'll only get a brief taster! For the less savage, there is also the quiet 7 g3 , planning a fianchetto.

Game 39 Kram nik-Damljanovic

Moscow Olympiad 1994 1 d4 d 5 2 tLlf3 tLlf6 3 c4 c6 4 tLlc3 dxc4 5 a4 jLg4 6 tLle5 jLh5 7 f3 tLlfd7

. .

72

Question 3: This is a strange­

looking move. What does it do?

Th e B r o n s t e in Va ria tio n

A nswer: Black must challenge the

knight on eS or his bishop will be caught by g2-g4 and h2-h4. Now after 8 g4, 8 . . . �g6 9 h4 4JxeS ! 10 dxe5 'i'xdl+ 1 1 �xdl hs is fine for Black. However, another point is to give Black an unexpected opportunity to continue his pressure against the d4pawn.

knight's natural square, while his bishop is biting granite on hs. 9 4J e4 seeks to exploit this temporary confu­ sion by threatening a devastating check on d6, and also shields the knight on c4 along the fourth rank; so 10 dxeS is now a threat as . . . iVh4+ will no longer achieve anything. 9 . . . SL b4+

10

SLd2 Wife7

11

SLxb4

8 ttJxc4 e 5 !

Wifxb4+ 1 2 Wifd2 Wifxd2+

And this i s it! The knight's move from f6 has freed the h4-d8 diagonal for the queen, giving Black the oppor­ tunity to exploit the slight weakening on the e l-h4 diagonal created by 7 £3 .

Forced, as 12 . . .iVxc4 loses to 1 3 4J d6+.

9 ttJe4

1 3 �xd 2 exd4 1 4 iLled6+ �e 7

The alternative, the enterpnslllg 14 . . . �d8, is considered in the next game. 1 5 iLlf5+ �f6 1 6 iLlxd4 J:;!d8

The old move and a very sensible one. White cannot play 9 dxeS as 9 ... iVh4+! wins the knight on c4, while 9 4JxeS 4Jxe5 10 dxeS 4J d7 1 1 f4 JLb4 is a very risky pawn grab. Black has a substantial lead in devel­ opment and will follow up with . . . 'i'e7 and . . .f7-f6 or . . . g7-gS, opening further lines. 9 g3 is considered in Games 4 1 and 42. Question 4: What does 9 4J e4 do? Answer: Black's pieces are a little strange at the moment: his king's knight is on d7, which is the queen's

We have reached an ending, but the tactical complications continue for some while yet. 1 7 �c3 iLlc5 1 8 e4 ! B.xd4 1 9 �xd4 iLlb3+ 20 �c3 iLlxa 1

21 .iLe2 �e 7

22 J::i. x a 1 iLl d 7 23 b4 !

Question 5: Isn't this just a boring, equal ending? A nswer: Unfortunately for Black, no. White does enjoy a definite edge here, and it all boils down to that wayward bishop on hs. First, Black is going to have to spend a tempo with 73

Th e Sla v

. . . f7-f6 to bring it back into play; and second, if it was still on cS , Black wouldn't have such an annoying weakness on b7! The white knight is excellently placed on c4, as it can at­ tack b7 via d6 (with the help of a rook on dl) or as . Question 6: You mean Black is lost?! A nswer: No, not at all. He only has one real weakness, so he should be able to defend, but it isn't really that much fun.

stopping the knight from activating via eS . Black obviously felt very un­ comfortable round here, since he starts just moving his knight around for no reason.

31 . . . iLlb6 3 2 iLl c 5 iLld7 3 3 iLlb3 96 34 iLld4 iLlb6 3 5 �c4 J:!.b7 3 6 �b3 �d7 3 7 ga5 !

23 . . . f6 2 4 gd 1 iLl b 6

Kramnik suggests that 24 ... Mbs more solid.

IS

2 5 iLl a 5 !

Exchanging knights would greatly simplify Black's defensive task. Now White forces unpleasant weaknesses in the black queenside. 25 . . . iLlxa4+ 26 �b3 iLl b 6 27 iLlxb7

Threatening a breakthrough with e4-eS.

�f7+ 28 �c3

3 7 . . . iLla8? 38 3L a 4 J:!.c 7 39 lIc5 iLlb6

The trade of the a4-pawn for the b7-pawn has been profitable for White, as now he has two targets: a7 and c6.

1 -0

And Black lost on time in this hopeless position: he is just going to lose his c-pawn.

28 . . . gb8 29 �a6 �e8 30 J:!.a 1 iLld7 31 f4 !

Gaining space on the kingside and 74

Let us now take a look at 14 . . . �d8 instead of 14 . . . �e7.

Th e B r o n s t e in Va ria tio n

Game 40 Schandorff-Hellsten

Copenhagen 1996 1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 �g4 6 ctJe5 � h 5 7 f3 IiJfd 7 8 ctJxc4 e 5 9 ctJe4 � b4+ 1 0 �d2 �e7 1 1 � x b4 �x b4+ 1 2 �d 2 �xd2+ 1 3 '.t>xd2 exd4

1 4 ctJed6+

�d8 ! ? 1 9 '.t>c3 �he8 ! 20 ctJac4 .!;l;ab8 2 1 � d 3 ? ctJxd3 2 2 '.t>xd3 ctJc5+ 2 3 '.t>e2 f5! 24 '.t>f 1 f4 2 5 ctJd 1 ctJb3 0 - 1

An amazingly quick defeat for White!

Game 41 Epishin-Pomes

Manresa 1995 1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3

An interesting idea: in the ending above, Black would have loved to have his king on c7 protecting the weak b-pawn! One drawback is that White can take on b7 with check, but at least the position is unbalanced, unlike the safe edge which White eas­ ily obtained in the game Kramnik­ Damljanovic.

dxc4 5 a4 �g4 6 ctJe5 � h 5 7 f3 ctJfd 7 8 ctJxc4 e 5 9 g 3

1 5 ctJxb7+

15 g4!? JlLg6 16 f4 f6 17 f5 JlLeS 1S tLlxb7+ c/£c7 19 tbba5 does not really improve the white position, but in­ stead 15 h4!? (threatening g2-g4 and h4-h5) 1 5 . . .f6 16 g4 JlLeS 17 tbf5!?, at­ tacking g7 and d4, is an interesting attempt.

This has been the most popular choice recently. White prevents . . . �h4+ and threatens d4xe5 .

1 5 . . '.t> c 7

9 . . . � b4

.

16

ctJba5

dxe3+ 1 8 ctJxe3 ctJ b4

ctJa6

17

e4

For 9 . . .f6 see Game 43. 75

Th e Sla v

1 0 dxe5 0-0 1 1 ilL h 3 ! ?

White now threatens e5-e6. 1 1 .. .'�e7 1 2 f4

is ttJh6+ gxh6 19 �xe6 �f7 20 f5 h5 21 0-0 would have led to a murky poslttOn. 1 7 ilL e 3

17 'iVe4!, intending ttJh6+, was bet­ ter. 1 7 . . . LiJa6 1 8 0-0 ilLf7 1 9 LiJe4 It>h8 20 gad 1 �c7 21

b3 ilLf8 22 �b2

LiJec5 23 LiJxf6 ilLxb3 24 !:!.xd8 !:!.xd8 25 LiJg4 ilLd5 26 LiJd4 LiJxa4? ! 27 �a 1 b 5 ?

12 �f4 is considered in the next game. 1 2 . . . .l:!.d8 1 3 �c2 f6

Absolutely necessary in order for Black to free himself. 1 4 e6

14 exf6 ttJxf6 gives Black some counterplay for the pawn. 1 4 . . . LiJc5 1 5 LiJe3

Or 15 f5 Md4 16 ttJe3 ttJba6 17 0-0 MadS with counterplay (Epishin) . 1 5 . . . LiJxe6 1 6 LiJf5

Black is determined to ignore his kingside defences. The punishment is swift. 28 f5 ilL b4 29 LiJe6 ! ilLxe6 3 0 fxe6 ilLc3 3 1 �b 1 �e7 32 LiJ h 6 !

The pawn capture 32 ... gxh6 loses t o 33 Mf7. 32 . . . .l:!.f8 33 LiJf7+ It>g8 34 ilLf5 ':!:!'xf7 35 ilLxh7+ It>h8 36 ':!:!'xf7 �e8 37 �f5 1 -0

Game 42 Parker-Hellsten

Copenhagen 1996 1 d4 d 5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3 1 6 . . . �d 7 ? !

dxc4 5 a4 ilLg4 6 LiJe5 ilLh5 7 f3

A mistake according t o Epishin. The alternative 1 6 . . :�eS 1 7 'iVe4 ttJa6

LiJfd 7 8 LiJxc4 e 5 9 g 3 ilL b4 1 0 dxe5

76

0-0 1 1 ilLh3 �e7 1 2 ilLf4

Th e B r o n s t e in Va ria tio n

1 7 . . . Ile8

17 .. .lbd7 fails to 18 �g5 ! 'MUxe5 (18 .. .f6 19 exf6 gxf6 20 'MUxg6+) 19 �xd8 'MUxg3+ 20 �f1 Mxd8 21 'be4, when Black has insufficient compen­ sation for the exchange. 18

0-0 �c5+

19

<;t>g2 �xc3

20

�xc3 �xc3 21 bxc3 ctJd7

A very interesting idea. 12 . . . g5, to chase the bishop away, is simply met by 13 �d2 (13 �xd7 'bxd7 14 �xg5 'iVxg5 15 'MUxd7 b5 16 f4 is very messy) 13 . . . 'bxe5 14 'bxe5 'MUxe5 15 'be4 Jtxd2+ 16 'MUxd2 g4 17 �g2 with a better position for White. 1 2 . . Rd8 1 3 �c2 �g6 1 4 �f5 .

14 e4 is very risky but not easy to refute. For example, 14 . . . 'MUc5 15 'bd6 fLlxe5 16 'bxb7 (forking queen and rook) 16 . . . 'bxf3+ 17 �f1 'MUc4+ 1 8 'iVe2 seems fine for White, while 14 . . . b5 15 axb5 cxb5 16 'be3 'bxe5 17 0-0 is also difficult to judge.

Black just manages to hold the en­ suing endgame, but the whole line seems extremely uncomfortable for him.

1 4 . . . ctJb6 1 5 �xg6 hxg6 1 6 ctJxb6

ctJ d 7 30 Rbxc4 ctJxe5 3 1 �xe5 Ilxe5

axb6 1 7 h4

22 gfb 1 ga6 23 a 5 gea8 24 gd 1 ctJf8 25 gab 1

b5 26 c4 bxc4 27

gxb7 I!xa 5 28 Il b4 R d 5 29 R c 1 3 2 e4 ga2+ 3 3 <;t>h3 f5 34 gxc6 fxe4 3 5 fxe4 Ilxe4 36 1::!. x g6 <;t>h7 37 gg 5 P.e7 38 Id.h 5+ '/z - '/z

Game 43 Dautov-Ni kolic

Ter Apel 1994 1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 �g4 6 ctJe5 � h 5 7 f3 ctJfd 7 8 ctJxc4 e 5 9 g 3 f6

Preventing . . . g7-g5 .

A very solid continuation, just pro­ tecting the e5-pawn. 77

Th e Sla v

<;t>e5 J::l: e 8+ 41 �f5 ge2 42 b4 �e8 43 gh4 J::l: f 2+ 44 �e6 ge2+ 45 �d6 a5 46 bxa5 <;t>f7 47 gh7+ <;t>xf6 48 gxb7 ne4 49 �xc6 gxa4 50 gb5 1 -0

This is all very sensible. However, White has another rather crazy idea. Game 44 Shirov- N i kolic

Wijk aa n Zee 1993 1 0 dxe5 ctJxe5 1 1 \'i{{x d8+ <;t>xd8 1 2 ctJxe5 fxe5 1 3 iL g 5+ <;t> c 7 1 4 0-0-0 ctJd7 1 5 iL h 3 iLe8 1 6 iLe3 !

1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 iLg4 6 ctJe5 iL h 5 7 f3 ctJfd 7 8 ctJxc4 e 5 9 e4

And this is it! This very natural move was completely ignored until recently. 9 . . . \'i{{ h 4+ 1 0 g 3

The wacky 10 �e2 is considered in the next game. 1 0 . . : 'II¥f 6 !

This very nice move prepares f3-f4, breaking the position open.

Having softened up the kingside with . . . "iVh4+, the queen retreats to f6, where it helps attack f3 with the bishop on h5 , and d4 with the pawn on e5. 1 1 dxe 5 ! �xf3 1 2 ctJd6+ ! <;t> d 8

1 6 . . . ctJc5 1 7 f4 iLd7 1 8 iLxd7 ctJxd7 1 9 f5

19 ctJe4 Me8 20 f5 was even more accurate. The ending is basically very pleasant for White. 1 9 . . . iLe7 20 g4 h 6 21 ctJe4 ctJf6 22 ctJxf6 iLxf6 2 3 h4 iLe 7 24 iL f2 gad8 25 iLg3 iLf6 26 <;t>c2 .!cl.xd 1 27 J::l: x d 1 gg8 28 g 5 h x g 5 29 hxg 5 iLxg 5 30 iLxe5+ <;t>c8 31 gg 1 ge8 3 2 iLxg7 l:;l;.xe2+ 3 3 �d3 l:;l;.d2+ 34 <;t>c3 iLe3 3 5 gg3 iLf4 3 6 gf3 gg2 3 7 gxf4 gxg7 38 f6 g g 8 39 <;t>d4 <;t>d7 40 78

Amazingly, 12 ... �xd6 loses to 1 3

Th e B r o n s t e in Va ria tio n

'iVxd6 i¥xh l 1 4 �,g5 (threatening 'iVe7+ mate) 14 . . .f6 15 exf6 gxf6 16 'iVe6+ WdS (16 . . . WfS 1 7 ilh6+ mate) 17 ilxf6+ lLlxf6 i s i¥xf6+ Wc7 19 'iVe5+! followed by i¥xh5 or i¥xhs with a crushing position, as pointed out by Ivan Sokolov.

a5 25 h4 b6 26 �c2? !

1 3 �xf3 �xf3 1 4 ctJxf7+ �e8 1 5 e6 ! ?

The first new move o f the game! In I.Sokolov-Lautier, Belgrade 199 1 , White had played 1 5 CLlxhS but 15 . . . CLlxe5 ! (Shirov) 16 ilf4 CLlSd7 is nice for Black, as after 17 fIg l , 1 7 . . . ilc5 ! makes sure the rook does not escape. 1 5 . . . ctJ c 5 1 6 � c 4 � x h 1

1 7 ctJxh8

JLxe4 ! 1 8 � g 5

Either 26 Wd2 or 26 g5 would have kept a slight advantage for White ac­ cording to Shirov. 26 . . . ctJca6 27 ctJe4 �a7 28 ctJg 5+ �d5 29 ctJhf7 Ide7 3 0 gh8 ctJd7 3 1 �xh 7 ctJe5 3 2 ctJxe5 M,xh7 3 3 ctJxh 7 �xe5 34 ctJf8 I!>f4 3 5 ctJxg6+ �xg4 36 ctJe5+ �xh4 37 ctJxc6 �g5 Y2 - Y2

Game 45 Nesterov-Imanaliev

Bishkek Zonal 1 993 1 d4 d 5 2 c4 c 6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 �g4 6 ctJe5 � h 5 7 f3 ctJfd 7 8 ctJxc4 e 5 9 e4 �h4+ 1 0 �e 2 ! ?

i s b4 seems to win a piece, but ls . . . ild3 ! (Shirov) 19 e7 il.xc4 (19 . . . ilxe7 20 ilf7+ followed by bxc5 keeps the fun going) 20 exfSi¥ + �xfS 21 bxc5 WgS favours Black. 1 8 . . . �f5

1 9 0-0-0 �e7

20 �xe7

ri;xe7 21 �f 1 g 6

A slight error according t o Shirov. 21 . . . r�e6 22 ilxe6 Wxe6 23 fIfS as is suggested instead, but this also seems quite nice for White. 22 g4 �xe6 23 �xe6 �xe6 24 Idf8 79

Th e Sla v

A magnificent idea, the tactical just­ ification of which lies in my all-time favourite opening trap! 1 0 . . . exd4

Tempting but not the best. Black should react more calmly with 10 . . . jLb4, intending . . . 0-0.

knight on c4. The move is extremely aggressive: White will expand on the kingside with g2-g4 and h2-h4 and try to win Black's light-squared bishop, while rapid queenside castling is also on the agenda.

Game 46

1 1 �xd4 3L c 5 1 2 lZJd6+! �f8

I . Sokolov- Hellsten

12 . . . c,t>e7 loses to 13 ctJfS+ while 12 . . . c,t>ds is met by 13 ctJxb7+.

Malmo 1 995

1 3 �xg 7+ ! ! 1 d4 d5 2 c4 c6 3 lZJc3 lZJf6 4 lZJf3 dxc4 5 a4 3Lg4 6 lZJe5 3L h 5 7 f3 lZJfd 7 8 lZJxc4 e5 9 3Le3 3Lg6

Question 9: This looks odd as well! A nswer: This is a typical idea in this

line. Black realises that the bishop is doing nothing on hS, where it merely bites against the pawn on f3 . There­ fore, he moves it to a more active di­ agonal, delaying the decision of which piece to put on b4: the bishop on f8 or the knight on bS, via a6. 1 3 . . . �xg7 1 4 1ZJf5+

1 0 h4!

Regaining the sacrificed queen and winning a pawn, with a good position to boot! 1 4 . . . �f6 1 5 lZJxh4 lZJa6 1 6 3Lh6 lZJe5 1 7 g4 lZJxg4 1 8 fxg4 3Lxg4+ 1 9 �d2 �ad8+ 20 � c 2 lZJ b4+ 21 �b3 3Le6+ 2 2 3Lc4 3Lxc4+ 23 �xc4 .l:!.d4+ 24 �xc 5 lZJc2 2 5 3Lg7+ 1 -0

Finally, White can try the develop­ ing move 9 jLe3 . Question 8: This looks very strange, doesn't it? A nswer: With 9 jLe3 , White de­ fends the d4-pawn and prepares to meet 9 . . :�h4+ with the simple 10 jLf2 . However, 1 0 dxeS is still not a threat due to 10 . . :iVh4+, winning the 80

This not only aims to harass the bishop on g6 with hS , but also threat­ ens d4xeS by removing Black's re­ source of . . . iVh4+. 1 0 . . . 3Le7 1 1 h 5 ! 3Lf5

1 1 . . .�h4+ is met simply by 12 jLf2

Th e B r o n s t e in Va ria tio n

�xf2+ 1 3 'it'xf2 i,f5 14 liJd6+, win­ nmg a pIece. 1 2 dxe5 0-0 1 3 h 6 ! lLla6 1 4 hxg7 l:!.e8 1 5 g4 XL g 6 1 6 f4 !

Game 47

.

K rasen kov-Sapis

Polish

Championship 1995

1 lLlf3 d 5 2 d4 c6 3 c4 lLlf6 4 lLlc3 dxc4 5 a4 XLg4 6 lLle5 XLh5 7 f3! lLlfd 7 8 lLlxc4 e 5 9 XLe3 XLb4

A more natural developing move than 9 . . . i,g6. 1 0 g4 XLg6 1 1 dxe 5 !

This is possible now, since l 1 . . .�h4+ no longer hits the knight on c4! 1 1 . . . 0 -0 ! ?

Quite amazing! White is only mov­ ing pawns, but Black just seems to be helpless! 1 6 . . . lLlb4

17

Uc 1

lLld5

18

XLg2

Ibxe3 1 9 lLlxe3 XL h4+ 20 �f 1 XLg3 21 f5 �b6 2 2 fxg6 hxg6 23 �d2 �xe 5 24 lLled 5 1 -0

24 . . . cxd5 loses to either 25 Mh8+ \t>xg7 26 �h6+ 'it'f6 27 liJxd5+, fork­ ing king and queen, or instead 25 tt'lxd5 �d4 26 Mh8+ 'it'xg7 27 �h6+ mate! A game of astonishing ferocity, even by Ivan Sokolov's remarkable standards!

This i s a new idea. The alternative 1 1 . . . �e7+ was played in the original game Granda Zuniga-Nikolic, Biel Interzonal 1993, and now 12 liJd6+ i,xd6 13 �xd6 would have given White a safe edge due to his two bish­ ops and the weakness of Black's dark squares. Instead in the game, White went for the crazy complications of 12 f4 �h4+ 13 i,f2 �xg4 14 �b3 liJa6 15 liJd6+ i,xd6 16 �xb7 liJb4 17 �xa8+ �b8 which is just unfathom­ able. 1 2 h4 h6 1 3 h 5 XL h 7 1 4 g 5 ! ? hxg 5 1 5 h6 g 6

81

Th e Sla v

1 9 axb5 cxb5 20 'J}f/e4 !

bxc4 2 1

'J}f/xaS ttJxe5 22 'J}f/xa 7

Question 1 0: Why is White playing so aggressively? He was a pawn up with more space; why did he sacrifice his g-pawn? A nswer: White has played some very strange moves in the opening. First, he put his bishop on e3 in front of the e-pawn, blocking in his light­ squared bishop . Then, instead of de­ veloping his pieces, he advanced his kingside pawns in order to chase Black's light-squared bishop. White has won a pawn, but his pawn ad­ vances have left many weak squares in his position. For example, if Black could get a rook to dS and then play . . . ctJc5, aiming for the weak b3-square, then White's position would become critical. White has raised the stakes with his risky opening play - he must continue in the same aggressive man­ ner or Black will develop and exploit White's weaknesses. 1 6 Rg 1 'J}f/e7 1 7 Rxg 5 :i:!.dS 1 S 'J}f/c2 b5?

A tactical miscalculation. Chekhov recommends lS . . . 'i'e6! 19 'i'e4 ctJa6! (intending . . . ctJdc5) 20 'i'h4 b5!, when I think that Black has good counter­ chances. 82

White is winning now due to his large material advantage. 22 . . . ttJbd7 23 '>jj'd 4 f6 24 Rg3 ctJc5 25 .l'::!. a S ! l:baS 26 '>jj' d 5+ ttJ e 6 27 'J}f/xaS+ ttJfS 2S Wf2 iLd6 29 ttJe4 iLb4 30 iLd4 WhS 3 1 f4 ttJed7 3 2 Re3 '>jj'f 7 33 ttJ g 5 '>jj' g S 3 4 ReS 1 -0

And now things get even more complicated. Let us take a look at 7 h3 .

Game 48 Gelfand - N i kolic

Manila Interzonal 1990 1 d4 d5 2 ttJf3 ttJf6 3 c4 c6 4 ttJc3 dxc4 5 a4 iLg4 6 ttJe5 iL h 5 7 h3 ttJa6 S g4 iLg6 9 iLg2

I find it hard to recommend this line to players of either colour, unless they have six months in which to ana­ lyse the mind-boggling complications! 9 e3 is discussed in the next game. 9 . . . ttJb4 1 0 0-0 iLc2 1 1 'J}f/d2 iLb3

Question 7: Why is Black doing

this? A nswer: If Black were to play nor-

Th e B r o n s t e in Va ria tio n

mally, then White's space advantage would guarantee him a substantial advantage. Black therefore keeps the pawn and challenges White to make something of his lead in development.

iVxg5, which looks like good com­ pensatIon to me.

1 8 �e4 �d8 1 9 f4 �e7 20 e3 0-0 21 �e2 c 5 22 dxc5 �xc 5 23 � d 2 1 2 a5

Black does not really have enough for the exchange.

Threatening a5-a6 to break up the black queenside. 12 tL'le4, attacking the knight on b4, is the other major con­ tinuation, aiming for the attractive trap 12 . . . tL'lxe4 13 iVxb4 tL'ld6 14 'i'xb7! ! , when 14 . . . tL'lxb7 loses to 15 c�oxc6+. 1 2 . . . a6

Stopping a5-a6, but in fact this may not be so dangerous: 12 . . . e6 13 a6 iVc7 14 axb7 iVxb7 15 g5 tL'lfd5 16 e4 tL'lb6 17 d5 looks impressive, but 17 . . . :l':rd8 18 tL'lxc6 tL'lxc6 19 dxc6 iVxc6 20 e5 4Jd5 21 :l':rxa7 �c5, with . . . 0-0 to fol­ low, was nice for Black in the game Moreno-Rogers, Manila Olympiad 1992.

2 3 . . . CLl d 5 2 4 �h 1 � d 7 25 l::i. a c 1 �a4 26 �f3 c3 27 bxc3 �xa 5 28 f5 �c4 29 fxe6 ! liJe7 30 �xh7+ �xh 7 3 1 �e4+ liJg6 3 2 �xc4 fxe6 3 3 �g4 l::i. x f 1 + 34 .ti.xf 1 iVb5 35 c4 �c6+ 36

1 3 liJa4 e6 1 4 g 5 liJ d 7 1 5 liJxd 7

�g 1 �g8 37 �f3 �d 7 3 8 l::i. f 2 a 5 3 9

�xd 7 1 6 tLlb6 �d8 1 7 tLlxa8

�e4 tLle7 4 0 �xa5 tLlf5 4 1 � d 2 1 -0

see

follo wing diagram

1 7 . . . �xa 8 ?

Ftacnik suggests 1 7 ... tL'lc2 1 8 :l':rb 1 'i'xa8 19 e3 �b4 20 iVe2 iVd8 2 1 �d2

If White wishes to duck the critical lines that we saw in the previous game, he can play 9 e3 instead of 9 �g2. 83

Th e Sla v

Came 49

iLlc2!? 15 �xc2 �xd4+ 16 �g2 �xc4 and 17 f5 (Rogers) .

Klarenbeek-Rogers

Dutch Team Championship 1 996 1 d4 e6 2 e4 d 5 3 tZlf3 tZlf6 4 tZle3 dxe4 5 a4 ]Lg4 6 tZle5 ]Lh5 7 h 3 tZl a 6 8 g4 ]L g 6 9 e 3

A quieter attempt, making sure that White regains the pawn. see

follo wing diagram

9 . . . tZlb4 1 0 ]Lxe4 e6

1O . . . 'Llc2+ loses to 11 �xc2 �xc2 12 �xf7 mate, but 1o . . . 'Lld7 1 1 iLlxg6 hxg6 12 �f3, intending �f1-g2, is more normal. White has a small ad­ vantage here due to his slight space advantage and bishop pair.

14

�e2

tZlxb4? !

�xd 7 ]Lxb4

15

tZla2

�e4

17

]Ld2

0-0

16 18

]Lxb4? !

White is just playing for a draw, but he is doing this badly. The text makes the a-pawn very weak. 1 8 . . . axb4 1 9 b3 Ra 5 ! 20 Rad 1 b5 21 axb5 exb5 22 ]Ld3 ]L b 7 !

Black is not going to exchange this bishop, while the as-h 1 diagonal is so tempting! 23 \t>h2 �d 5 24 %:!.b 1 Re8 2 5 Rfd 1 g5!

An unexpected and really strong move. 26 :!:!.f 1 1 1 0-0 a5 1 2 f4 tZld7 1 3 tZlxd 7

Rogers recommends instead 13 e4 �h4 14 �g2 with a slight advantage for White, but with the threat of 15 iLlf3 , intending f4-f5, this looks horri­ ble for Black. 1 3 . . . ]L e 2 !

We have seen this before. Black makes sure that the bishop does not get shut in behind the e4-pawn. 13 . . . �xd7 is strongly met by 14 e4! 84

26 fxg5 loses to 26 . . . �d6 (Rogers), as 27 �g1 �g3+ is terminal. 26 . . . Re3 27 Rb2 gxf4 28 Rxf4 Ra1 29 \t>g3 Rg 1 + 30 \t>h4 h 6 0 - 1

And White lost o n time i n this hopeless position.

Came 50 Leitao-Beliavsky

Erevan Olympiad 1 996 1 d4 d 5 2 e4 e6 3 tZlf3 tZlf6 4 ctJc3

Th e B r o n s t e in Va ria tio n

dxe4 5 a4 �g4 6 tLle5 � h 5 7 g 3

The quiet option. White ignores the bishop on h5 and just develops normally. 7 . . . e6 8 � g 2 � b4 9 tLlxe4 tLl d 5

Attacking c3 . 1 0 'i'b3 0-0 1 1 �d2

11 0-0 fails to 1 1 . . .itxc3 12 bxc3 Jtxe2, winning a pawn. 1 1 . . a 5 1 2 e4 tLl b 6 ! .

An important and typical manoeu­ vre. Black exchanges a pair of knights, relieving his slightly cramped position while attacking d4.

'i'd6 22 �h2 .!:!.ae8 23 .l:!.ee 1 �g6 24

1 3 tLlxb6

'!:!'e2 .l:!.e7 25 tLlb5 'i'e5 26 �f3 g d 7

This rather helps Black. Beliavsky suggests 1 3 ite3 instead.

2 7 �g2 �h8 28 tLle3 .l:!. f d 8 29 gd 1 f6 30 ged 2 �f7 3 1 .!:!.d3 h 5 32 'i'b5

1 3 . . . 'i'x b6 1 4 �e3 e 5 ! 1 5 d5 tLl d 7

e4 3 3 'i'xe4 h4 34 tLle2 � h 7 3 5 gd4

1 6 0-0 tLle5 1 7 h 3 tLlf3+ 1 8 �h 1

ge7 36 'i'd3 '!:!'de8 3 7 gxh4 '!:!'e2 3 8

itJd4

d6 '!:!' x b 2 39 d 7 .l:!.d8 40 �h 1 � h 5 4 1

Clearly Black has now taken over the initiative. The bishop on h5 has suddenly become a major player, sup­ porting the incursion of the black knight into the vulnerable kingside light squares.

1 9 'i'e4 tLle2 20 '!:!'ae 1 tLlxe3 2 1 fxe3

�xh 5 'i'xh 5 42 e5+ f5 43 tLlg 1 � e 5 4 4 .!:!. g 4 .l:!. x d 7 45 'i'xd 7 fxg4 4 6 'i' d 3 + �h6 47 'i'e3 g 3 4 8 'i'xb2 'i'xd 1 49 �g2 'i'xa4 50 'i'xb 7 'i'e2+ 51 �xg3 �xe3 0 - 1

A fine game b y Beliavsky.

85

Th e Sla v

S u m m a ry

I cannot really recommend the 5 . . . itg4 line for Black, not because it is a par­ ticularly bad line, but simply because unless you have loads of time for de­ tailed analysis, you won't be able to feel comfortable playing it. There are many theoretical problems to solve: 7 f3 ttJfd7 8 ttJxc4 e5 9 ttJe4 gives White a safe endgame edge, while 9 g3 is also dangerous. Even the crazy 9 e4 and 9 ite3 pose difficult problems! By contrast 7 h3 gives Black too many counter­ chances, while 7 g3 is a little tame. 1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 dxc4 5 a4 �g4 6 ctJe5 � h 5 7 f3

7 h3 ttJa6 8 g4 �g6 (D) 9 itg2 - game 48 9 e3 game 49 7 g3 - game 50 -

7 . . . ctJfd 7 8 ctJxc4 e 5 (D) 9 ctJe4

9 g3 9 . . . � b4 10 dxe5 0-0 1 1 ith3 Wlic7 12 f4 game 41 12 itf4 - game 42 9 .. .f6 game 43 9 e4 Wlih4 + 10 g3 - game 44 10 �e2 game 45 9 ite3 9 . . . itg6 game 46 9 j�ob4 - game 47 -

-

-

-

..

9 . . . � b4+ 1 0 �d2 Wii e 7 1 1 �xb4 Wiix b4+ 1 2 Wii d 2 Wiix d2 1 3 �xd2 exd4 1 4 ctJed6+ (D)

14 . . . �e7 - game 39 14 . . . �d8 game 40 -

8 86

. . .

�g6

8

. . .

e5

14 ctJed6+

CHA PTER SEVEN The 4

.

.

.

a6 Slav :

White plays 5 e3

1 d4 d 5 2 c4 c6 3 lL:lf3 lL:lf6 4 lL:lc3 a6

In the next two chapters we shall look at the move-order 1 d4 d5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 a6. Question 1 : What is the point of 4 .. a6? Answer: 4 . . . a6 allows the bishop to develop outside the pawn chain to g4 or f5, since after . . . iH5 , 'i"b3 attacking b7 can be met by . . . b7-b5, advancing the b7-pawn to a safe square, or ... J.:!a7!, an ugly looking but brilliant thought of Julian Hodgson's. Question 2: Sounds great. Any drawbacks? A nswer: You had to ask. Black is placing a lot of pawns on light squares , so he can often suffer from weak dark squares. .

Game 5 1 O i l -Anand

Biel Interzonal 1993 1 d4 d 5 2 c4 c6 3 lL:lf3 ttJf6 4 lL:lc3 a6 5 e3

White protects the c4-pawn and prepares to develop his kingside. White's numerous alternatives here are discussed in the next chapter.

5 . . . b5

Black would like to develop his bishop outside the pawn chain by playing 5 . . . �g4, for example, but 6 'i"b3 is strong, as 6 . . . b5 7 cxd5 cxd5 8 a4! breaks up the black queenside. Black must be able to meet a2-a4 with . . .b5-b4; after . . . b5xa4, his a-pawn be­ comes very weak. In fact Black could play an interesting tactical idea here: after 8 . . . �xf3 9 gxf3 he can try 9 . . . b4!? , so that after 10 'i"xb4, 1 0 . . . e5! attacks the white queen. 1 1 'i"b3 exd4 1 2 exd4 ctJc6 13 �e3 �b4 gives com­ pensation for the pawn due to White's weakened pawn structure, but 1 1 'i"b7! gains a tempo by attack­ ing the rook on a8 : 1 1 . . .ctJbd7 1 2 dxe5 ctJxe5 13 �xa6, when Black has 87

Th e Sla v

enormous problems on the light squares. I'm sure that Julian Hodgson would suggest 6 . . . Ma7(!) , but after 7 ctJe5 (threatening 8 ctJxg4 ctJxg4 9 cxd5, winning a pawn) 7 . . . e6 (as usual in the Slav Black does not mind swapping off his light-squared bishop for White's knight) 8 f3 ! 11Lh5 9 g4 11Lg6 10 h4! , White's threat of h4-h5, trapping the bishop, forces Black to play the disastrous 1 O ... h6, when 1 1 ctJxg6 fxg6 1 2 �c2 is just winning for White. The inclusion of 5 . . . b5 6 b3 takes the b3-square away from the white queen, allowing Black to develop his light-squared bishop in greater com­ fort.

For 8 gxf3, see the next game. 8

. . .

e6

This was a novelty at the time of the game, as black players had been experimenting with the violent 8 ... eS, to exploit the absence of the queen from the queenside and the slight weakness of the knight on c3 (it is no longer protected by a pawn on b2) . I prefer Anand's simple move, which carries the same threats but without the risk. 9 :iL.d2 :iL.b4 1 0 �d 1

6 b3

The exchange 6 cxd5 is considered in Game 55. 6

. . .

:iL.g4 7 h 3

A very natural reaction, putting the question to the bishop. 7 11Le2 is dealt with in Games 53 and 54. 7

. . .

:iL.xf3 !

7 . . . 11Lh5 8 g4 11Lg6 9 ctJe5 , intending h2-h4, is rather awkward, as we have seen. 8 �xf3 88

So that 10 . . . �a5 can be met by 1 1 �c2, but this move i s a little meek. A few years ago, I played 10 11Ld3 �aS 1 1 Me l ! ? 11Lxc3 12 MXc3 �xa2 13 �d1 against Jon Levitt, sacrificing a pawn in order to gain the advantage of two bishops against two knights. In fact I won a nice game after 13 . . . 0-0 14 0-0 �a3 15 �a1 ! �xa1 16 Mxa1 Ma7!? 17 cxb5 cxb5 (hoping for 1 8 11Lxb5 ctJe4, swapping off one of my bishops) 18 Mc2 ! , allowing my dark-squared bishop to activate itself via b4. With control of the c-file and Black's pas· sive pieces, I quickly gained a decisive advantage. Black should have played . .. ctJe4 at some point before Mc2 in

Th e 4

.

.

.

a 6 Sla v : Wh i t e p la y s 5 e 3

order to force the exchange of one of White's bishops, but White has rea­ sonable compensation for the pawn. 1 0 . . . 0-0 1 1 �e2 bxc4 ! 1 2 bxc4 c 5 ! This i s a typical freeing manoeuvre for Black. The immediate 1 1 . . .c5 would of course have lost a pawn to 12 cxb5, so Black first exchanges on c4 and then breaks in the centre. 1 3 dxc 5 ?

This is a serious mistake, and after Anand's superb play it almost looks like the losing move. White had to play for equality with 13 cxd5 cxd4 14 exd4 1i.xc3 15 1i.xc3 tLlxd5 . 1 3 . . d4! 1 4 exd4 �xd4 .

17

�a4 �ad8

18

�e1

tDd4!

19

�xb4?

Retreating with 19 1i.d1 was the only (but rather miserable) way to avoid material loss. 1 9 . . . tDxe2+ 20 tDxe2

Or 20 �h l :d3 ! 20 . . :�xa 1 2 1 tDc3 �c 1 22 �a5 �f4 23 �xa6 l:l:a8 24 �d6 �xc4 0 - 1

White had had enough. This game is a really impressive demolition job by Anand. Let us now see what happens if Black recaptures on f3 with the pawn instead of the queen. White has a worse pawn structure and real tactical problems, as Black's pieces quickly become amazingly ac­ tive.

Game 52 Van der Sterren-Shi rov

Biel Interzonal 1993

1 5 �c2

15 iVc 1 (to avoid Black gammg a tempo on the queen after . . . tLlc6-d4) 15 . . J::t d S ! prevents White from cas­ tling. 1 5 . tDc6 . .

15 .. J::r d S is well met by 16 :d1 ! 1 6 0-0 �e 5 !

Freeing the knight.

d4-square

for the

1 d4 d 5 2 c4 c6 3 tDf3 tDf6 4 tDc3 a6 5 e3 b 5 6 b3 �g4 7 h 3 �xf3 8 gxf3 ! ?

The recapture with the queen leaves White vulnerable to a quick . . . 1i.b4. This recapture aims to suffo­ cate Black by playing f3-f4 (preventing Black's . . . e7-e5 break) and c4-c5 (preventing the . . . c6-c5 break) . 89

Th e Sla v

5 liJf3 b 5 6 b3 �g4 7 �e2

8 . . . liJ b d 7 !

Threatening . . . c7-eS . 9 f4 bxc4 1 0 bxc4 dxc4!

By taking the c-pawn, Black pre­ vents c4-cS . Now White cannot stop Black from playing . . . c6-cS himself, and the game soon fizzles out. 1 1 �xc4 e6

Peter Wells and Glenn Flear are well known for their deep knowledge of Slav systems, so this game is espe· cially interesting. 7 . . . e6 8 0-0 ..IiLd6? !

An inaccuracy that has unpleasant consequences. When White plays LiJeS, attacking c6, Black needs to be able to exchange it as quickly as pos· sible for one of his own knights. Therefore Black should either play 8 . . LiJbd7 or 8 . . �e7 (to meet 9 4Je5 with 9 . . . �xe2 10 ¥llix e2 and then 1O . . . LiJfd7!) . In the game, White gets a grip on the dark squares and wins in model fashion. .

.

9 h 3 �h5 1 0 liJe5 �xe2 1 1 liJxe2! 0-0 1 2 liJf4 Wlic7 1 3 liJfd 3 ! liJbd7 1 4 �b2 Wlib7

1 5 '!:!c 1

�ac8

1 6 Re2

1 2 �d2 ..IiL b4 1 3 0-0 0-0 1 4 liJe4 a5

.!:!fd8 1 7 Wlif3 �f8 1 8 J:!fc 1 !

1 5 ..IiLxb4 axb4 1 6 liJxf6+ liJxf6 1 7

White's pieces are coordinating beautifully.

Wlid3 liJ d 5

1 8 �fc l .!:!a5

1 9 J:!ab 1

�d 6 Y:z - Y:z

see

follo wing diagram

Game 53

1 8 . . . bxc4 1 9 bxc4 liJxe5 20 liJxe5

Wells-Flear

�d6 21 liJd3 liJe4 22 �e2 Wlib8 23

Oakham+1994

liJc5 �xc5 24 dxc5 Wlib7 2 5 cxd 5 �xd5 26 ..IiLxg7 �xg7 27 Wlig4+ LtJg5

1 d4 d5 2 c4 c6 3 liJc3 liJf6 4 e3 a6 90

28 e4 J::!. e 5 29 f4 J::!. x e4 30 Wlixg 5+

Th e 4

.

.

.

a 6 Sla v : Wh i t e p la y s 5 e 3

Wf8 3 1 g d 2 �e7 32 �h6+ �g8 3 3

White back.

l:!: c 3

1 7 . . . �b7 1 8 �e2 ga 7 ! !

�h8

34

gg3

gg8

35

gd7

�xe5+ 3 6 � h 2 1 -0

Game 54 K rasenkov-Epishin

Bmo 1 994

I f White now defends the eS-pawn with 19 f4, then 19 . . .tDb S ! 20 l:\,c3 b4! 21 l:\,d3 l:\,cs (Epishin) gains the c-file for Black. 1 9 z:!. e 1 LZlxe 5 ! 20 LZlxe6 �xe6 2 1 LZlxf8 �xe2 2 2 gxe2 LZle4! ! 2 3 bxe4

1 d4 LZlf6 2 e4 e6 3 LZlf3 d 5 4 LZle3

dxe4

a6 5 e3 b 5 6 b3 JiLg4 7 JiLe2 e6 8 h3 Jih5 9 0-0 LZl bd 7 ! 1 0 LZle5 JiLxe2 1 1 CLJxe2 LZlxe5 1 2 dxe5 LZld7

13 exd 5 exd 5 14 JiLb2 JiLe7 1 5 LZld4

The white knight is trapped. Black regains his piece and his queenside pawns prove to be far too much for his opponent to cope with.

�b6 1 6 l:i.e 1 0-0 1 7 ge6

24 JiLd4 ge7 2 5 LZlxh 7 �xh7 26 JiLb6

17 CDc6 �cS is equal according to Epishin. Here White's control of the c-file looks impressive, but with some fine moves, Epishin gradually pushes

ge6 27 JiLa5 JiLa3 ! 28 �f 1

b4 29

�e2 ge5 30 JiLb6 g b 5 31 JiLd4 b3 32 gxe4 b2 3 3 JiLxb2 gxb2+ 34 �f3 JiLb4 0- 1 91

Th e Sla v

Instead of 6 b3, sometimes White plays 6 cxd5 .

�e2 e5 33 h3 e4 V2 - V2

Game 55 Karpov-Short

Dortmund 1 995 1 d4 d5 2 c4 c6 3 tZJc3 tZJf6 4 e3 a6 5 tZJf3 b5 6 cxd 5 cxd 5 7 tZJe5

In the next game we see a tricky al­ ternative move-order from White.

Game 56 Sadler-Hodgson

Hastings 1 995/96 1 d4 d 5 2 c4 c6 3 tZJc3

A different plan: White prevents the black bishop from developing outside the pawn chain and tries to prove that Black's queenside is weak.

Of course playing 3 ttJc3 first gives Black the extra possibility of playing 3 . . . dxc4 (see Chapter 10) . 3 . . . tZJf6 4 e 3 a 6 5 �c2 ! ?

7 . . . e6 8 .1i.d2 .1i.e 7 9 .1i.e2 0-0 1 0 0-0 .1i.b 7 1 1 tZJd3

A typical manoeuvre by White, bringing the knight in contact with the weak c5-square. 1 1 . . . tZJ b d 7 1 2 b4 tZJ b 6 !

White has weak squares too! 1 3 a 4 tZJe4! see

follo wing diagram

1 4 axb5 tZJxc3 1 5 .i,xc3 axb5 1 6 tZJc5 .i,c6 1 7 .1lxa8 �xa8 1 8 .1i.d3 �a2 1 9 � h 5 g 6 20 �e5 �a7 21 ga 1

�b8 22 �x b8 gxb8 23 ga5

.1i.d8 24 .1i.e1 ga8 2 5 !!xa8 tZJxa8 26 g4 tZJb6 2 7 f3 tZJc4 28 .i,f2 .i,g5 29 f4 .i,e7 3 0 � g 2 f6 31 �f3 �f7 3 2 92

5 ttJf3 would o f course simply transpose to the games we have al­ ready seen in this chapter. On c2 the white queen prevents Black from developing his bishop to f5, which

Th e 4

suggests that it would be natural for Black to put his bishop on g4 instead. However, after 5 . . . b5 6 b3 �g4, White's idea is to play 7 tiJge2, and if 7 . . . tiJbd7 then 8 h3 �h5 9 tiJf4, picking up the bishop pair. Question 3: I thought you said that Black wanted to exchange off his light-squared bishop for White's knight! Aren't you contradicting yourself? A nswer: It is a conflict of ideas - as Black you say, ' The bishop on c8 was my problem piece and I'm glad I've exchanged it,' whereas with White you say 'Yes, I've won the bishop pair!' Frankly I would be happy to play either colour! It is clear, how­ ever, that in comparison with the line 5 tiJf3 b5 6 b3 �g4 7 h3 �xf3 , White has gained the two bishops at a much lower cost: he has not had to either weaken his kingside pawn structure, or misplace his queen on the kingside. So basically White has got a good ver­ sion of this typical sort of position. And that is the point of waiting with 5 'lic2. A similar idea for White is 5 �d3 , preventing . . . �f5 . Personally, I would grab this opportunity to transpose into a Queen's Gambit Accepted with 5 . . . dxc4 6 �xc4 e6 7 tiJf3 c5, but I know that not everyone feels the same way! 5 . . . �g4!? 6 'lib3 Ma7 is possible, however, as neither 7 f3 dxc4 (7 �h5 8 cxd5 cxd5 9 g4 �g6 10 �xg6 hxg6 11 g5 tiJh5 12 tiJge2 [12 tLlxd5 tiJg3 !J 1 2 ... e6 1 3 f4 is better for White due to the offside knight on h5) 8 �xc4 �h5, intending . . . tiJbd7 and . . . e7-e5, nor 7 cxd5 cxd5 8 f3 . . .

.

.

.

a 6 Sla v : Wh i t e p la y s 5 e 3

�c8 !?, with . . . tiJc6 and . . . e7-e5 to fol­ low, is advantageous for White. Right, back to the game! 5

. . .

e6 6 iDf3 e 5 !

A very imaginative idea. Black to­ tally changes his plan; he no longer plays for . . . b7-b5, but strikes in the centre with . . . c6-c5 . Question 4: Hasn't Black just wasted a move, since he's played first . . . c7-c6 and now . . . c6-c5? A nswer: This is true of course. White is playing the variation 1 d4 tiJf6 2 c4 e6 3 tiJf3 c5 4 e3 d5 5 tiJc3 a6, with the move 'Iic2 added in for free. Black's contention is that this variation is not advantageous for White normally, and that the move 'lic2 does not make any difference in White's favour. I felt that the best way to try to make use of 'lic2 was to play 7 cxd5, so that after 7 . . . exd5 I could attack a clear target on d5 by bringing my rook to d 1 , exploiting the fact that my queen has already vacated this square. Moreover, my queen could be very useful on c2 to attack a bishop on c5 after d4xc5 �xc5. 7 exd 5 exd 5 8 �e2 iDe6 9 0-0 �e6 ! 93

Th e Sla v

of this defensive course, White plays for the initiative. 1 3 a 3 ! exf3 1 4 iLxf3 ctJc6 !

Julian did not like the look of 14 . . . ttJdS 15 dxcs ttJc7 16 iLxb7 with three pawns and an initiative for the piece. 1 5 d5 ctJe5 1 6 dxe6 ctJxf3+ 1 7 gxf3 fxe6 1 8 b3 �xd 2 1 9 iLxd2

The start of a superb tactical plan. Normal development could have eas­ ily ended in disaster. For example, if 9 . . . iLe7 10 Md1 0-0 then 1 1 dxcS iLxcs 12 ttJxds wins a pawn, since the bishop on cS hangs. Therefore Black begins a complicated tactical manoeu­ vre that seeks to exploit the exposed position of the queen on c2 . 1 0 gd 1 ctJ b4 ! 1 1 �d 2

1 1 'iVb 1 'iV cS , threatening . . . iLfS, could be embarrassing. 1 1 . . . ctJe4! 1 2 ctJxe4 dxe4

Black's queenside pawn maJonty even gives him a slight pull, but after a few adventures the game was eventu­ ally drawn. 1 9 . . . �d 7 !

20 iLc3+ �c6

21

ga2

iLe7 22 gad2 gad8 23 gxd8 gxd8 24 gxd8 iLxd8 2 5 iLxg7 c4!

Sacrificing a pawn to activate the king. 26 bxc4 �c5 2 7 f4 �xc4 28 e4 iLc7 29 f5 exf5 30 exf5 � d 5 3 1 a4 b 5 32 axb5 axb5 33 iLc3 iLe5 34 iLd2 �e4 35 �f 1 �xf5 36 h3 iLf4 37 iLc3 �e4 38 �e2 iLd6 39 iLd2 �d4 40 iLe3+ �c4 41 iLd2 iLb4 42 iLxb4 !

White's pieces are not coordinating well, and if White were to continue routinely with 13 ttJeS, then 13 . . . cxd4 14 exd4 MCS , threatening . . . ttJc2, would be extremely annoying. Instead 94

I spent a while just checking that the pawn ending was drawn. Remem­ ber that 4 1 . . .b4 just leads to a draw after 42 iLxb4 as Black has the wrong­ coloured bishop for the rook's pawn!

Th e 4

.

.

.

a 6 Sla v : Wh i t e p la y s 5 e 3

42 . . . <;t>xb4 43 <;t>d3 <;t>a3 44 <;t>c2 b4

7 . . . �d 7 S ttJf3 e 6 9 ttJe5 �cS 1 0 f3

45 <;t>b 1 <;t>b3 46 h4 <;t>c3 47 h5 � - �

ttJfd7 1 1 ttJxd 7 ttJxd7

Since 47 . . . h6 48 f4 �d4 49 �b2 �e4 50 �b3 �xf4 51 �xb4 �g4 52 cj;c3 �xh5 53 �d2 �g4 54 �e1 �f3 55 �f1 draws Gust) for White.

Also possible was 1 1 . . .'i'xd7! ?

Another tricky move for White is 5 a4 .

Gam? 57 Atali k-Miles

Hastings 1995/96 1 d4 d 5 2 c4 c6 3 ttJc3 ttJf6 4 e3 a6 5 a4

An unusual move in this position when White has already committed himself to e2-e3, shutting in his dark­ squared bishop. 5 . . . �f5 6 �b3 ga7 7 a 5 !

Suddenly, t o his horror, Miles real­ ised that after the natural 7 . . . e6, 8 '{!Vb6! is extremely strong, as after the forced 8 . . . 'i'xb6 9 axb6 �a8 10 c5 , intending b2-b4 and b4-b5 breaking through, White has a magnificent end­ ing. Tony, practical as ever, just played a few necessary defensive moves and got on with the game!

1 2 cxd 5 cxd 5 1 3 e4 dxe4 1 4 fxe4 �g6 1 5 e 5 ! §Le7 1 6 §Le2 0-0 1 7 0-0 gaS 1 S §Le3 .l:!.bS 1 9 �a4 �dS 20 .l:!.ad 1 MCS 2 1 d5 �c5 22 §Lf2 �xf2+ 2 3 gxf2 ttJc5 24 �b4 �g 5 25 dxe6 fxe6 2 6 gxfS+ <;t>xfS 27 �d4 <;t>gS 2S �c4 Wie7 2 9 b4 ttJ d 7 30 §Lb3 ttJfS 31 �d6 �eS 3 2 §La4 �f7 3 3 Mf 1 �f5 3 4 ttJe4 �g 6 3 5 ttJ c 5 h 6 3 6 �d 1 �e4 3 7 B f 2 §Ld 5 3S §L c 2 Wig 5 39 h3 �e3 40 �e7 <;t>hS 4 1 �f7 ttJh7 42 �f4 �e 1 + 43 <;t> h 2 .l:!.gS 44 �g3 �xb4 45 ttJd3 �a3 46 ttJf4 �xg 3+ 47 <;t>xg3 gdS 4S ttJg6+ <;t>gS 49 ttJe 7 + <;t>hS 50 ttJg6+ <;t>gS 5 1 95

Th e Sla v

%-% After 5 a4 Black can simply play 5 . . . e6, aiming to put a bishop on the hole on b4, but this leads more to a Semi-Slav type of position, so for Slav devotees, I offer a few other ideas: a) 5 . . . g6 is interesting, leading to a sort of Schlecter Slav (see Chapter 1 1) . ClJe7+ �h8 52 ClJg6+

96

b) 5 . . . 'ii c 7 is not as stupid as it seems: Black defends b7 in advance. If 6 ct'Jf3 then 6 . . . �g4 7 'ii b 3 �xf3 8 gxf3 e6! c) 5 . . . Ma7!? with the same idea. If 6 as then maybe 6 . . . �e6!? 7 ct'Jf3 (7 'ii b 3 dxc4!) 7 . . . dxc4 8 ct'Jg5 �g4 9 f3 �h5 with a completely unclear position.

Th e 4

.

.

.

a 6 Sla v : Wh i t e p la y s 5 e 3

S u m m ary

In general Black is doing fine in these lines, but since the 4 . . . a6 Slav is such a recent development, there is still scope for improvements for both colours. 4 Qjf3 a6 5 e3 bS 6 b3 oltg4 7 h3 �xf3 8 'i'xf3 e6 9 �d3 �b4 10 �d2, as in Sadler-Levitt, is worth further tests, and the game Sadler-Hodgson is certainly crazy enough to be worth analysing! 1 d4 d5 2 c4 c6 3 tUf3 tUf6 4 tUc3

4 e3 a6 (D) 5 'i'c2 game 56 5 a4 game 57 -

-

4 . . . a6 5 e3 b 5 6 b3

6 cxdS game 55 -

6 . . . iLg4 (D) 7 h3

7 �e2 e6 8 0-0 - game 53 8 h3 game 54 -

7 . . . iLxf3 (D)

'i'xf3 game 5 1 8 gxf3 game 52

8

-

-

4

. . .

a6

6

. . .

iLg4

7

. . .

iLxf3

97

The 4

.

.

.

a6 Slav:

Aggressive o ptions for White

1 d4 d 5 2 c4 c6 3 lLlf3 lLlf6 4 lLlc3 a6

In this chapter we deal with the more attacking systems against the 4 . . . a6 Slav, in which White delays e2e3 in order to develop the bishop on c 1 to an active square outside the pawn chain. We shall first examine S cS, which has virtually become the main line. Question 1 : S cS looks like a begin­ ner's move. Isn't it bad to release the tension in the centre so early? A nswer: White is aiming for a 'big clamp' on the centre: S cS stops Black from breaking with . . . c6-cS, and �f4 will prevent Black from achieving . . . e7-eS . White will then either launch a queenside offensive with b2-b4, a2a4 and b4-bS, or he will organise a central break with e2-e4. Overall, this is a very ambitious plan. The queenside break . . . b7-b6 is not great for Black, since cSxb6, . . . ik'xb6 leaves him with weak dark squares on cS, b6 and as, and a backward c-pawn. Question 2: Black can't play on the queenside, and he can't get in any of his breaks. What can he do? A nswer: . . . c6-cS is not on the agenda, but . . . e7-eS should not be im98

possible to achieve if Black can organ­ ise his pieces properly. White has closed the centre, so Black has more time to achieve his plan, as he never has to worry about a sudden central breakthrough. Moreover, Black can spare a lot more pieces than usual to prepare . . . e7-eS, since he doesn't need any to defend his own centre. Game 58 K ra m n i k-Shirov

Vienna 1 996 1 lLlf3 d 5 2 d4 c6 3 c4 lLlf6 4 lLlc3 a6 5 c 5 .iLf5 6 �b3 .l:!.a7 !

I couldn't help laughing when Jul­ ian Hodgson told me that he played . . . a7-a6 in order to facilitate . . . .l::!. a 7,

Th e 4

.

.

.

a 6 Sla v : A g g r e s s i v e o p t io n s fo r Wh i t e

but the idea is a good one. It really is not clear whether the queen on b3 is any better placed than the rook on a7: on b3 the queen attacks very little and gets in the way of one of White's main plans, the queenside pawn storm with b4, a4 and bs. If the queen moves away from b3, then the rook can simply return to as ! The more conventional 6 . :�cS 1S discussed in the following game.

White positionally. White had to an­ ticipate . . . e7-eS and either prevent it with 17 eS or play 17 iYc4, intending 17 . . . eS 1S dS . 1 7 . . . e 5 ! 1 8 �c4 �f6 1 9 dxe5 lLlxe5

.

7 .ltf4 lLl b d 7 8 h 3 h6 9 e3 g 5 !

The pos1tlOn is now horrible for White: Black can just gang up on his weak pawns. 20 �d4 gaa8 !

Having done its job, the rook comes back into play. 2 1 � e 3 J::!. a d8 2 2 .lte2 � e 7 23 J:Ihe 1

This move shows a good under­ standing of the position. Black leaves his weakened queenside alone and grabs space on the kingside, where White has little going for him. The move . . . g7-gS also allows the bishop to come to g7, supporting the . . . e7-eS break.

lLld7 24 �d4+ lLlf6 25 �e3 �e5 26 .ltf3 h 5 !

Preparing . . . gS-g4, driving away a defender of e4. 2 7 a3 gfe8 28 nxd8 nxd8 29 .ltd 1 J::!. d 4 30 .ltc2 J::!. c4 3 1 Wb 1

1 0 .lth2 .lt g 7 1 1 lLle5 0-0 1 2 f3

Unwilling to let Black have things his own way, White prepares the e2e4 break. This raises the stakes, as ... e7-eS will become doubly effective against an expanded white centre. 1 2 . . . lLlxe5 1 3 .ltxe5 lLl d 7 1 4 .ltxg7 �xg 7 1 5 e4 dxe4 1 6 fxe4 .ltg6 1 7 O-O-O? A

careless move that condemns 99

Th e Sla v

S . . . e6 9 ctJe5 3Le7 1 0 g4 3Lg6 1 1 e3

3 1 . . . 1:!.xc5

The first pawn falls . . . 3 2 �d2 1:!.c4 3 3 �dS ctJxe4 3 4 3Lxe4

3LdS 1 2 3Le2 ctJxe5 1 3 3Lxe5 3Lc7 14 ctJa4

3Lxe4+ 35 ctJxe4 .!:!xe4

. . . and now the second. 0- 1

A really good exposition of the ideas behind the 4 . . . a6 Slav.

Game 59 I . S okolov-Shirov

Erevan Olympiad 1 996 1 d4 d 5 2 c4 c6 3 ctJf3 ctJf6 4 ctJc3 a6 5 c 5 3Lf5 6 �b3 �cS ! ?

This is a normal plan, trying to gain the two bishops by exchanging the knight for the dark-squared bishop on ds . However, there is a tactical problem on this occasion. 1 4 . . . 3La5+! 1 5 ctJc3

Or 15 Wfl CDe4 16 Wg2 f6, intending . . . e6-e5 . 1 5 . . 0-0 1 6 h4 h6 1 7 0-0-0 ctJd7 1 8 .

3Ld6 geS 1 9 ghg 1 3L c 7 2 0 h 5 3Lh7 21 g 5 .1Lxd6 22 cxd6 hS

I'm not sure why Shirov did not want to repeat his previous game. Perhaps it was nothing objective, just the desire not to be too predictable. The idea behind this move is ex­ tremely neat: Black will develop normally with . . . e7-e6, . . . CDbd7, . . . :ie7 and . . 0-0, and will then prepare . . . e6e5 by means of . . . MeS and . . . :idS-c7, making use of the fact that the queen has vacated dS ! .

7 3Lf4 ctJ b d 7 S h 3 ! ?

Black i s going t o castle kingside, so White wants to have his pawn storm ready on that side of the board. 100

I think that Black IS Just better here: White is having to make all sort of positional concessions and his at­ tack just isn't getting anywhere.

Th e 4

.

.

.

a 6 Sla v : A g g r e s s i v e o p tio n s fo r Wh i t e

2 3 gxh6 g x h 6 2 4 e 4 dxe4 2 5 tLia4

� b 5 �e2+ 60 �c3 f4 61 gd8 �f3

�d8 26 tLic5 tLixc5 2 7 dxc5 �f6 28

6 2 .l:!.d2+ �e 1 0 - 1

�c3 �xc3+ 29 bxc3 Uab8 30 gd4

Obviously Black is rather happier than White in this line at the mo­ ment!

a 5 31 �d 1 b6 3 2 cxb6 gxb6 33 jLc2 .!:!.d8 34 d7 .!:!.b5 3 5 .!:!.gd 1 '!:!'xh 5

Another very popular idea for White is to play 5 a4. Question 4: What is the point of this move? A nswer: 5 a4 merely aims to pre­ vent Black's idea of . . . b7-b5, and force him to find another plan. Though it severely weakens the b4-square, this move does give White the possibility of a4-a5, cramping the black queen­ side. Black wins yet another pawn. Though Sokolov fights hard, the re­ sult is never in doubt.

Pushkov-Epishin

36 c4 �g7 37 �b2 �f8 3 8 .ita4

Russian Championship 1 995

Game 60

nc5 3 9 \t>c3 �e7 40 .!:!.d6 e3 4 1 fxe3 �e4 42 .!:!.6d4 .!:!. e 5 4 3 gf 1 �g2

1 tLif3 tLif6 2 c4 c6 3 tLic3 d 5 4 d4

44 gg 1 gxe3+ 45 gd3 .!:!.xd3+ 46

a6 5 a4 e6

\ii x d3 '!:!'xd 7 + 47 �e3 �h3 48 �xc6 nc7 49 � b 5 e 5 50 Uh 1

.ite6 5 1

\ii d 3

\ii f4 54 .!:!.h4+ �g4 55 gh8 f5 56

The most natural and overwhelm­ ingly the most popular choice. The game loses a little of its Slav character I'm afraid, but Black does get to oc­ cupy that weak b4-square.

ne8 g c 5 5 7 a4 �f3 58 �a6 �f2 59

6 �g5

51 . . . �f6

52 .!:!.xh6+ �g 5

53 gh8

101

Th e Sla v

The fianchetto 6 g3 is dealt with in the following game.

�g7 38 �ba3 .i:txb2 39 �xa6 .i:tf4 40

6 . . . tiJ b d 7 7 e3 iLe7 8 iLd3 0-0 9 0-0

.i:td3 43 na5 Wf6 44 na6+ �f5 45

dxc4 1 0 iLxc4 c 5 1 1 a 5 cxd4 1 2

ga7 �e6 46 .l:!a6+ .i:td6 47 ga5 f5

exd4 b 5 1 3 axb6 tiJxb6

�6a4 �xa4 4 1 nxa4 nd2 42 �g 3

48 �a8 Wd5 49 �f2 nb6 50 l:IdS+ �e6 51 �g3 nb1

52 h 5 gxh 5 53

l:Ih8 h4+ 54 �xh4 �d5 5 5 XlfS f4 56 .i:td8+ �e6 57 ne8+ �f6 58 .l:!fS+ �e7 59 �h8 gg 1 60 nh7+ �f6 6 1 l:Ixh6+ � g 7 62 n e 6 .l::i. x g2 6 3 gxe5 �f6 64 J::!. a 5 .l::!. g 1 % - %

Let us see what happens if White opts to fianchetto . Game 61 Cifuentes-Dreev

Black has already equalised.

Wijk aan Zee 1 995

1 4 iLe2 iL b 7 1 5 tiJe5 h 6 1 6 iLh4 tiJfd 7 1 7 iLxe7 �xe7 1 8 tiJxd 7 �xd 7

1 c4 c6 2 d4 d 5 3 tiJc3 tiJf6 4 CUf3

1 9 tiJ a 4 tiJxa4 20 l::l x a4 l::l f d8 2 1

a6 5 a4 e6 6 g3 tiJbd7 7 iLg2 iLb4 8

�a 1

0-0 0-0 9 �b3 a5 1 0 gd 1 b6

�c6 2 2 iLf3 �b6 2 3 iLxb7

�xb 7 24 �c 1 �d6 2 5 h 3 gad8 26 gcc4 e 5 !

Black wins a pawn, but can't quite convert it into a win.

gg6 30 .i:tg4 l:I d d 6 3 1 gb8+ �h7 3 2

White will find it hard to achieve e2-e4, since Black has his bishop en­ trenched on b4 and can always play . . . �xc3 , removing a defender of e4. Without this idea, however, White may struggle to find a plan.

� b 3 �xb3 3 3 g x b 3 gd 1 + 34 � h 2

1 1 tiJe5 tiJxe 5 1 2 dxe5 tiJd7 1 3 cxd 5

g f 6 3 5 f3 g d 2 3 6 h4 g 6 3 7 ga4

exd 5 1 4 iLf4 �e7

27 gab4 �e4 28 d5 �xd 5 29 �a3

1 02

1 5 e4 d4!

16

Th e 4

.

.

.

a 6 Sla v : A g g r e s s i v e o p tio n s fo r Wh i t e

gxd4 lZJxe 5 1 7 J£. x e 5 \'ixe5

swift attack. Game 62 Ward-Levitt

British Championship 1 995 1 d4 d 5 2 e4 e6 3 lZJe3 lZJf6 4 lZJf3 a6 5 J£.g5 dxe4

Black's two bishops and queenside pawn majority give him a clear advan­ tage. 1 8 .l:!.d 2 J£.e6 1 9 \'i c2 gad8 20 gad 1 gxd2 2 1 gxd2 b 5 22 h 3 h 5 23 h4 bxa4 24 \'ixa4 J£.xc3 2 5 bxe3 \'ixe3 26 ge2 \'ib4 2 7 \'ixe6 a4 28 \'ie3 Vi1Vb3 29 �f 1 ge8 3 0 \'ix b3 axb3 3 1

6 a4

gd2 g e 2 3 2 gd8+ � h 7 3 3 e 5 J£.f5

Also possible is the immediate 6 e4!?

0-1

6 . . . h 6 7 J£. h 4 lZJd5 8 e 4 lZJ x e 3 9

This does all seem fine for Black but if you cannot live happily with­ out developing your bishop outside the pawn chain, then S . . . �fs 6 iVb3 Ma7 seems interesting. After 7 as, threatening 8 iVb6, 7 . . . dxc4 8 iVxc4 CLlbd7 is not so stupid. For example, 9 �gS h6 1 0 i.. xf6 (10 i.. h 4 gS!? 1 1 i.. g 3 i.. g7 seems fine for Black) 10 . . . exf6 (Please don't even think of 10 . . . CLlxf6, allowing 1 1 iV cS ! , winning a piece) 1 1 e4 i.. g4 (1 1 . . .i.. h 7!?; 1 1 . . .i.. e 6!?) with a very murky position. S i.. g S is a sharp continuation that is only occasionally seen. White de­ velops h is bishop to its most aggres­ sive square and is willing to sacrifice the pawn on c4 for the chance of a

bxc3 b 5

Black has gone about this in a strange way. S . . . Lbe4 6 i.. h4 (6 h4 has been played but I don't believe it! Even 6 . . . h6 is not stupid, when I don't see the value of h2-h4!) 6 . . . Lbxc3 7 bxc3 dxc4 8 e4 bs is more natural if Black wants this type of position, since the omission of a2-a4 and h7-h6 must help Black a little. After S . . . dxc4 6 a4 another idea is to play 6 . . . �fS (a more natural Slav move) , aiming for a little trap: 7 e3 Lbe4!? 8 Lbxe4 i.. x e4 9 i.. x c4 i.. xf3 ! 1 0 iVxf3 �aS+! 1 1 �e2 �xgS 12 �xf7+ �d8 , when White does not have sufficient compensation for the piece. 1 0 lZJe5 \'ie7 1 1 J£.g3 \'ib7 1 2 g b 1 1 03

Th e Sla v

Wii a 7 1 3 Wii f 3 e6 1 4 3l.e2 g6 1 5 Wiif 6

The opening has not been a success for Black, but somehow he hangs on and almost turns the tables com­ pletely. 1 5 . . . g h 7 1 6 �f3 3l. b 7 1 7 0-0 tiJd7 1 8 tiJxd7 �xd7

1 9 �f4 �e8

After this freeing break, Black has no problems. 11

tiJxd 7 tiJxd7

1 2 3l.e3 3l.e 7

13

dxc5 tiJxc5 1 4 Wiix d8+ gxd8 1 5 gd 1 0-0 1 6 a3 gxd 1 + 1 7 �xd 1 M.c8 1 8 3l.e2 �f8 1 9 �c2 tiJa4 20 3l.d2 3l.d6

20

21 h 3 �e7 22 Bd 1 tiJxc3 23 3l.xc3

axb5 axb5 21 ga 1 ! Wii b 6 22 :!::i. x a8+

b4 24 axb4 3l.xb4 25 gd3 a5 26

3l.xa8 23 g a 1 Wii b 7 24 d5 cxd 5 25

�b3 3l.xc3 2 7 Bxc3 .l:i.xc3+ 28 �xc3

exd 5 3l.e7 26 �e3 3l. g 5 2 7 f4 3l.e7

�d6

28 3l.f3 Wii c 8 29 d 6 3l.xf3 30 Wiix f3 �c5+ 31 3l.f2 Wiix d6 3 2 .!:!.a8+ 3l.d8 3 3 ga7 Wii d 5 34 Wiix d 5 exd 5 35 g b 7 h 5 36 �xb5 f5 3 7 M.xd 5 .!:!. d 7 38 ge5+ �f7 3 9 g 3 3l. f6 40 M.c5 V2 - V2

Next, the adventurous 5 CLJeS . I don't like this move at all for White and in the next game Shirov neutral­ ises it very easily with simple devel­ opment.

Game 63

tiJxc4 b5 8 tiJe5 3l.b 7 9 e4 e6 1 0 f3

You might have expected the play­ ers to have called it a day here, espe­ cially when you consider that this was the last round of a gruelling Olym­ piad for both players: Beliavsky was playing his 14th game on board 1 and Shirov his 1 3th!

c5!

29 �d4 e 5+ 30 �c3 �c5 31 3l.c4 f6

Beliavsky-Shirov

Erevan Olympiad 1 996 1 d4 d5 2 c4 c6 3 tiJf3 tiJf6 4 tiJc3 a6

1 04

5 tiJe5 tiJ b d 7 6 3l.f4 dxc4!

7

Th e 4

.

.

.

a 6 Sla v : A g g r e s s i v e o p tio n s fo r Wh i t e

32 h4 �c6 3 3 �g8 h 6 3 4 � c 4 � d 7 3 5 h 5 f5 3 6 e x f 5 � x f 5 3 7 �d3 �e6 38 �e4 �f7 3 9 g4 �c4 40 �b7 � b 5 41 �a8 �c4 42 �e4 �d5 43 �xd 5 �xd 5 44 b3 e4 45 f4 e3 46 �d3 e 2 4 7 �xe2 �e4 48 g5 �xf4 49 gxh6 gxh6 50 � d 3 �g4 5 1 �c4 �xh 5 5 2 �b5 �g4 5 3 �xa 5 h5 54

b4 h4 55 b5 h3 56 b6 h2 57 b7 h 1 � 58 b8� �a 1 + 59 �b6 �b 1 + 60 rtJc7 �x b8+ 6 1 �xb8 V2 - V2

In the end even these two battlers cannot continue the game! A really impressive display of fighting spirit from both players. The final idea to be considered is 5 'i'b3 , preventing the development of the light-squared bishop by attacking b7. The advantage of this move is that Black cannot counter in normal Slav fashion, but the Semi-Slav approach is easy to understand, and nice for Black.

6 1Lg 5 ? !

6 cxd5 cxd5 7 iLg5 iLe7 S e 3 h 6 9 iLxf6 iLxf6 10 iLd3 ctJc6 1 1 0-0 0-0 12 Mac 1 was played in Piket-Shirov, Aruba 1995, and now Shirov suggests 12 . . . iLd7, as 13 �xb7 ctJa5 14 �b4 �e7 snares the queen. 6 . . . dxc4 7 �xc4 b 5 8 �d 3 c 5 9 a4 cxd4! 1 0 LiJxd4 b4 1 1 LiJe4 1Lb7 1 2 �xf6 gxf6 1 3 ld.d 1 1Le7 1 4 �f3 b 3 !

Game 64 Lautier-Ba reev

Linares 1994 1 d4 d 5 2 c4 c6 3 LiJf3 LiJf6 4 LiJc3 a6 5 �b3 e6

The most popular move for Black. 6 �xc4 iLf5 7 g3 is the 4 �b3 dxc4 5 �xc4 iLf5 line with the extra moves 4 ctJc3 a6, which favour White more than Black, while 5 . . . Ma7 6 iLf4! is awkward. I briefly considered 5 . . . b5 6 cxd5 cxd5 7 a4 b4!? S �xb4 ctJc6 9 'i'b3 Mbs 10 �dl iLf5, intending . . . ctJc6-b4, but 1 1 ctJh4 chases away the bishop and just leaves White a pawn up. 5 . . . dxc4

To meet 1 5 ctJxe6 with 1 5 ... �a5+! 1 5 e3 �a5+ 1 6 �e2

Watch this king! 1 6 . . . rtJf8 1 7 �f4 �e5 1 8 f3 �xf4 1 9 exf4 �g7 20 f5 LiJc6 21

fxe6

LiJxd4+ 22 .i;l;,xd4 ghd8 23 gd7 gxd 7 24 exd 7 J:i.d8 25 rtJe3 f5 26 LiJg3 iLc5+ 27 �f4 ! 1 05

Th e Sla v

g4xfS-f6-f7-f8-g8 ! !

27 . . . 1Ld6+ 28 �xf5 gxd 7 29 ctJe4 1Lc7 30 1Lc4 g e 7 3 1 ctJc3 ge5+ 32

40 . . . 1Lb4 4 1

�g4 � h 6 33 f4 ge3 34 1L d 5 f5+ 3 5

gxc3 43 ge 1 .l:i.c2 44 1Ld5 .l:i.d2 45

�xf5 1Lc8+ 3 6 �f6 1Lxf4 3 7 1Lxb3

1Le4 1Lh5 46 gf 1 1Lg6 47 1Lb7 gd4

1Lg5+ 38 �f7 1Lg4 3 9 �f8 ! ! 1Le7+

48 a 5 na4 49 gf6 gxa5 50 nxa6

40 �g8 ! !

g h 5 5 1 h3 ge5 52 ga8 � g 5 53 1Lf3

The white king completes a remarkable J ourney: �e2-e3-f4xfs-

h 5 12 - 12

1 06

1Lf7 1Lxc3 4 2 bxc3

A brilliant game!

Th e 4

.

.

.

a 6 Sla v : A g g r e s s i v e o p t io n s fo r Wh i t e

S u m m a ry

If you had asked me a few months ago, I would have told you that Black had a few problems in the 4 . . . a6 Slav, but now I am not so sure! 5 cS should be avoided for the time being, while 5 a4, 5 ctJeS and 5 'iYb3 don't really seem to promise a great deal, though 5 a4 and 5 'iYb3 can be good weapons if you know that your opponent does not like to play systems with . . . e7-e6. 5 �gS is aggressive and deserves further tests. 1 d4 d 5 2 c4 c6 3 CLJf3 CLJf6 4 CLJc3 a6 5 c 5 (D)

5 a4 e6 (D)

6 �gS - game 60 6 g3 game 61 -

5 �gS game 62 5 ctJeS game 63 5 'iYb3 game 64 -

-

-

5

. . .

jLf5 6 �b3 (D)

6 . . . .l:"!.a7 game 58 6 . . . 'iYc8 game 59 -

-

5 c5

5

. . .

e6

6 �b 3

107

CHA PTER NINE The Exchange Variation

1 d4 d 5 2 c4 c6 3 cxd 5 cxd 5

The Exchange variation is often dismissed as just a dead draw, but sev­ eral top players, Yusupov and Vaiser in particular, play it to win. What do they see in this variation? Although the pawn structure is symmetrical, there is the open c-file to play with and we know from experi­ ence the annoying pressure that White can exert on b7. White also has the advantage of moving first in this position. Indeed, if Black does not know what he is doing, his position can very easily become highly un­ pleasant. Question 1 : What plans has Black tried in this position? A nswer: Black has tried two ap­ proaches: a) To put the bishop on c8 outside the pawn chain, on either f5 or g4. b) To unbalance the pawn structure with . . . ctJe4xc3 . Question 2: And what does White do? A nswer: White also has two ideas: He can put his king's knight on either f3 or e2 (after �d3) . Question 3: What is the difference between them? 1 08

A nswer:

a) ctJf3 allows White to increase his pressure on the c-file by playing ctJe5, attacking a black knight on c6. If Black exchanges knights with . . . ctJxe5, then White has removed the main barrier to invasion on the c-file. b) With �d3 and ctJge2 White aims, not to attack on the c-file, but rather to keep Black passive by preventing the light-squared bishop from devel­ oping outside the pawn chain. The white bishop on d3 controls f5, while the absence of a knight from f3 means that . . . �g4 can be met by f2-f3 .

Game 65 M i l ov-Sadler

Isle ofMan 1994 1 d4 d 5 2 c4 c6 3 cxd 5 cxd 5 4 iLf4

White must be wary of move­ orders here as 4 ctJc3 e5 !? 5 dxe5 d4 6 ctJe4 ctJc6 7 ctJf3 �f5 8 ctJg3 �g6 9 a3 �c5 10 "i'b3 ctJge7, as in Tozer-Levitt, London (Lloyds Bank) 1993, gives Black good counterplay for the pawn. 4 . . . lLlc6 5 e3 lLlf6 6 lLlc3

White could try 6 �d3 ! ? here to prevent 6 . . . �f5. Black's best is 6 . . . i,g4 7 ctJe2 �h5 ! , intending . . . �g6 to swap

Th e Ex c h a n g e Va ria tio n

off bishops. 7 �b3 ctJaS ! 8 �bs+ �d7 also poses few problems. 6 . . �f5 .

The old main line. The modern 6 . . a6 is considered in the next game. .

the main line, which does not seem bad for Black, but at the board I got a little carried away. The positional idea is correct: since White has released the tension in the centre very early, Black feels much more able to take action on the wings. Indeed this idea was seen to great effect in Kramnik-Shirov from Chapter 8, but in this case White has the open c-file to help his queenside play. 1 1 �g3 h5 1 2 h3 g4 1 3 hxg4 hxg4 1 4 tZJd2 �h4 1 5 tZJe2 �a5 ! ?

7 ]L b 5 e6 8 tZJf3

4 ctJf3 ctJf6 S ctJc3 ctJc6 6 �f4 �fS 7 e3 e6 8 �bs is the normal way to reach this position. White is threaten­ ing ctJeS, ganging up on the knight on c6.

S o far so good. With my last move, attacking the knight on d2, I was hop­ ing for the reply 16 'sfd 1 , when 16 . . . �c2! 17 �xc2 �xb5 is reasonable for Black. 1 6 e4 ! !

8 . . . tZJd7

Breaking the pin on the knight and thus dealing with the annoying threat of ctJeS . 9 0-0 �e7 1 0 �b3 g 5 ! ?

After a great deal of thought, my opponent found a very strong idea, blowing the centre open. 1 6 . . . �xd 2 1 7 exf5 ]Lxg3 1 8 tZJxg3 �xd4 1 9 t!.fe 1 �d8 20 fxe6 fxe6 2 1 l:!xe6

Bold but probably not the best. 1O . . . ,Sc8 1 1 �xc6 bxc6 12 �b7 as is

Well, this is not great for Black, but at least I'm still alive! My opponent was already in serious time-trouble, and after the game I was really upset that I had made things so easy for him 1 09

Th e Sla v

in the rest of the game.

a6 5 cxd5 cxd5 6 .tf4 ttJc6.

!!xd4

Question 4: Why is . . . a7-a6 useful in

4:Jxb3 24 �xd 5 4:Jbd4 2 5 �e4 4:Jxb5

this position? A nswer: 6 . . . a6 is a constructive waiting move: Black keeps the white pieces from occupying the b5 square, which means he no longer has to worry about .tb5, pinning his knight on c6. Question 5: What if White just plays 7 e3? A nswer: Then Black plays 7 . . . �g4 8 h3 .txf3 9 iVxf3 e6. This is another example of Black giving up his light­ squared bishop for White's king's knight in the Slav. In this case, since White has played cxd5 so early, Black has been able to play his knight to c6, its most natural and best square. The manoeuvre . . . .tg4xf3 also removes the attacking idea ttJe5 . Question 6: So what does White do? A nswer: The only way that White can go for advantage is to avoid play­ ing e2-e3 too early and thus sidestep . . . .txf3 .

21 . . . �c7

22

�d 1

4:Jc5

23

26 �x b 5 �ad8 2 7 .l::i. e 1 � d 2 28 4:Je4 .l::i. e 8 29 �e3 �c2 30 a4 .l::i. e 7 3 1 4:Jf6 I!.xe3 32 4:Jd5+ �c8 3 3 4:Jxe3 :!:!.d2 34 4:Jx94 �d4

3 5 4:Je5 :!:!.xa4 36

4:Jxc6 bxc6 3 7 �b3 �c7 38 f3 c 5 39 �f2 �c6 40 94 c4 41 �b8 a 5 42 �e3 �c7 43 �b5 �c6 44 :!:!.b8 Wc7 45 .l:!a8 I!.a2 46 95 �b7 47 96! c3 48 9 7 c 2

Everything else is also hopeless, e.g. 4S ... cxb2 49 .l:IbS+ �xbS 50 gSiV + �a7 5 1 'i¥xa2 and wins. 49 �d2 1 -0

Now let us take a look at the fash­ ionable 6 . . . a6.

Game 66 H odgson-Sadler

Ischia 1 996 1 c4 c6 2 4:Jf3 d5 3 cxd 5 cxd 5 4 d4 4:Jf6 5 4:Jc3 4:Jc6 6 iL.f4 a6

The modern main line, which will be of particular interest to 4 . . . a6 Slav players, as this line can occur by transposition after 3 ttJf3 ttJf6 4 ttJc3 1 10

7 4:Je 5 ! ? e6

Black does not fear S ttJxc6 bxc6, as White is not sufficiently active to pre­ vent Black from freeing himself with

Th e Ex c h a n g e Va ria tio n

. . . c6-cS . 8 e3 CLlxe5 9 iLxe5 iLd7 !

7 . . . iLf5

An important new idea. 9 . . . bs was also possible, but that does give White a bit of a target to attack on the queenside. 9 . . . �d7 develops the bish­ op to the as-h 1 diagonal (preventing White from ever achieving e3-e4) , while still allowing Black to cover the cS-square with the advance . . . b7-b6 if necessary.

7 . . . �g4 S ctJ eS ! (the point) is annoy­ ing, but now S ctJ eS is met by s . . . ncs . 8 e3 e6 9 LDe5 LDxe5 1 0 iLxe5 CLld7 1 1 iLg3 .!:!.c8 1 2 iLd3 iLxd3 1 3 �xd 3 iL e 7 1 4 0-0 0-0 1 5 '!::!' c 2 �a 5 Y:z - Y:z

Despairing o f making anything against this plan of . . . �g4xf3 , white players turned to 'Plan B ' : �d3 and ctJe2.

1 0 iLd3 iLc6 1 1 �f3 CLld7 1 2 iLg3

Game 68

Y:z - Y:z

A real no-holds-barred classic! In mitigation, England were playing Switzerland that day in 'Euro 96' so we did have other things on our minds. (Although after watching the match for half an hour we began to think that even our game might have been more exciting!)

Yusupov-Sh i rov

Zurich 19941%

1 d4 d 5 2 c4 c6 3 CLlc3 CLlf6 4 cxd 5 cxd 5 5 iLf4 CLlc6 6 e3 iLg4

Game 67 Andersson-Epishin

Ter Apel 1995 1 CLlf3 af6 2 c4 c6 3 d4 d5 4 cxd 5 cxd 5 5 LDc3 CLlc6 6 iLf4 a6 7 .1:Ic 1

Here we see another attempt to de­ lay e2-e3 . 1 1 1

Th e Sla v

Question 7: Why not 6 . . . a6 here? A nswer: After 6 . . . a6 White can play

7 i,d3 i,g4 8 tLJge2 (but not 8 f3 �hs and Black can exchange bishops with . . . i,g6) . Question 8: What's the problem? Can't I just go 8 . . i,xe2? A nswer: You certainly can, but in comparison with the line 6 tLJf3 a6 7 e3 i,g4 8 h3 i,xf3 9 'iVxf3 , we see that Black has an inferior version: White has played i,d3 instead of h2-h3 , and his queen is better placed on e2 than on f3 . This position is perfectly play­ able for Black, but there just seems no point in going in for a worse version of something if there is no real need to. Question 9: What is the point of the 6 . . . i,g4 7 f3 i,d7 manoeuvre? A nswer: By forcing f2-f3 , Black weakens a dark square on e3 in White's position. Moreover, Black has more chance of achieving the . . . e7-eS break, since the white knight can no longer go to f3 . Finally, the bishop is not too badly placed on d7, as it sup­ ports Black's counterplay with . . . b7bS. Question 1 0: Wait a minute, if I get this via a transposition from the . . . a7a6 Slav, then I won't be able to play it in this way will I? A nswer: Don't worry! White can't use this subtle move-order if he trans­ poses to the Exchange variation via the 4 . . . a6 Slav, since he will either have shut in his dark-squared bishop (3 tLJc3 tLJf6 4 e3 a6) or will have al­ ready committed his knight to f3 (3 tLJf3 tLJf6 4 tLJc3 a6) . .

7 f3 Si. d 7 8 Si. d 3 e6 1 12

Also playable is 8 . . . g6! ? , intending 9 . . i,g7 to support the . . . e7-eS break. .

9 Si.g3

The aggressive 9 g4 is considered in the next game, while 9 tLJge2 allows Black to win the bishop pair with 9 . . . tLJhS . 9

. . .

Si.e 7 1 0 tLJge2 0-0 1 1 0-0 a6

Black's main counterplay is to ex­ pand on the queenside, placing his knight on the c4 outpost and advanc­ ing the a- and b-pawns. 1 2 tLJc 1 ! ? tLJa5 1 3 tLJ b 3 !

An imaginative idea: on b3, the knight protects d4, allowing White to expand in the centre with e3-e4, while eyeing the cS-square, which will be weakened by . . . b7-bs. The one

Th e Ex c h a n g e Va ria tio n

problem is that the knight will be in the line of fire when Black plays . . . a6as-a4. 1 3 . . . lLle4 1 4 �e2 J:i.e8 1 5 e4 b5 1 6 e5 lLlh 5 ! 1 7 f4 lLlxg3 1 8 hxg3 a 5 1 9 2Lxh7+ �xh7 20 �h5+ �g8 2 1 �f2 f5 Y, - y,

After 22 �h l White will force per­ petual with iVhs+ and iVhs+. Game 69 Vaiser-Nalbandian

Erevan Open 1 996 1 d4 d 5 2 e4 e6 3 lLlc3 lLlf6 4 exd 5

Although White's position is a little loose, his extra pawn will count in the end.

exd 5 5 .itf4 lLle6 6 e3 .itg4 7 f3 2Ld7

20 . . . �b6 21 lLld2 0-0 2 2 g5 h5 23

8 .itd3 e 6 9 g4

�e3 a 5 24 a 3 2Lb7 2 5 .l:!.b1 �a6 26 ge7 2Le6 27 .l:!.e1

lhe7 28 .itxe7

.it b 5 29 lLlg3 ge8 3 0 �b2 g 6 3 1 lLlge4 a4 3 2 l:!e3 2Le6 3 3 .itg3 .it d 5 3 4 .l:!.xe8+ �xe8 3 5 lLle3 .ite6 3 6 e4 �a6

This aggressive thrust is V aiser' s trademark: White advances on the kingside, while Black counter-attacks on the queenside. 9 . . . a6

9 . . . i, b4 is interesting, so that after 10 a3 i,aS 1 1 b4, the bishop returns to c7 and supports the central break . . . e6-eS . 1 0 g e 1 h 6 1 1 h4 ge8 1 2 lLlge2 lLla5

37 d5 .ite5+ 38 �g2 exd 5 39 exd5 2Lxd 5 40 lLlde4 .itxe4 4 1 lLlxe4 iU8 42 �d4 �e2+ 43 .itf2 f5 44 gxf6 �b5 45 lLlg 5 lLle5 46 �e4 .ith6 47 f7+ 1 -0

1 3 �f2 lLl e4 1 4 ge2 b 5 1 5 lLlb 1 2Le6 1 6 �e 1 .ite7 1 7 b3 lLld7 1 8 bxe4 bxe4 1 9 2Lxe4 dxe4 20 .l:!.xe4

Finally, we take a look at the most recent attempt from Black, 6 .. .cDe4. 1 13

Th e Sla v

This game brought the whole idea to popular attention.

Game 70 Portisch-Kra m n i k

Biel lnterzonal 1993 1 d4 d 5 2 c4 c6 3 cxd 5 cx d 5 4 ct:Jf3 ct:Jf6 5 ct:Jc3 ct:Jc6 6 �f4 ct:Je4 ! ?

The key to this concept. Black an­ ticipates that the centre will open with c3-c4 or e3-e4, and places his bishop on an influential diagonal. 9 .lte2

An idea o f Iosif Dorfman. Question 1 1 : What is the point of the . . . lZJe4xc3 manoeuvre? A nswer: With . . . lZJe4xc3 , Black closes the c-file, blocking one of White's major assets. Moreover, he unbalances the pawn structure, isolat­ ing the white a-pawn but giving White the chance to play a pawn to c4 for the second time! Question 12: What if White had played one of the subtle move-orders, such as 3 cxd5 cxd5 4 �f4 lZJc6 5 e3 lZJf6 6 lZJc3? A nswer: Well, 6 . . . lZJe4 is still possi­ ble and leads to unclear play after 7 lZJxe4 (or 7 lZJge2! ? , intending f2-f3 to recapture on c3 with a knight) 7 . . . dxe4, intending a quick . . "e7-e5 and . . . �b4+. 7 e3 ct:Jxc3 8 bxc3 g 6 ! 1 14

9 c4 �g7 10 cxd5 'i'xd5 is a posi­ tion from the 4 �g5 Griinfeld (1 d4 lZJf6 2 c4 g6 3 lZJc3 d5 4 �g5 lZJe4 5 �f4 lZJxc3 6 bxc3 �g7 7 e3 c5 8 cxds cxd4 9 cxd4 'i'xd5 , etc.) , which is fine for Black, while Kramnik shows that 9 �d3 �g7 10 0-0 0-0 1 1 e4 �g4 12 h3 �xf3 1 3 'i'xf3 e5! 14 dxe5 (14 exdS lZJxd4! 1 5 cxd4 exf4 equalises) 14 . . . d4!, intending . . . lZJxe5, is also fine. Finally, 9 lZJe5 is dealt with in the next game. 9 . . . �g7 1 0 0-0 0-0 1 1 c4 dxc4 1 2 �xc4 �f5

Th e Ex c h a n g e Va ria tio n

Kramnik gives 12 . . . �g4 13 h3 �xf3 14 �xf3 es 1 5 dxes �xes 16 �xes �xes as equal: 17 �xb7 �xh2+ 1 8 �xh2 �h4+ 19 �gl 'iVxc4 keeps the balance. He also suggests 12 . . . a6!?, intending ... b7-bs. 1 3 Rc 1 l;l c 8 1 4 � e 2 a6 1 5 h 3 ?

1 5 ds bs 16 �b3 leads to a slight advantage for White according to Kramnik. 1 5 . . . ttJa5 1 6 Ji.d3 Ji.xd3 1 7 �xd3 �d 7 1 8 .l::!. c 3 b 5 1 9 Rfc 1 ttJc4

A very impressive demonstration of Black's strategy. Game 71 Cifuentes-Leyva

Cien/uegos 1 996 1 d4 c5 2 c3 cxd4 3 cxd4 d 5 4 ttJf3 ttJc6 5 ttJc3 ttJf6 6 Ji.f4 ttJe4

After a weird transposition, we are back to the main position. 7 e3 ttJxc3 8 bxc3 g 6 9 ttJe5 Ji. g 7 ? 20 �e2 ttJb6 21 'f1.c7 �e6 22 Ji. g 5 4Jd 5 ! 23 'f1. 7 c 5 h 6 2 4 Ji.h4 b4 ! 25 �b2 ttJc3 !

I like the way in which Black has moved his outpost from c4 to c3 !

A careless error. As Lalic has shown, Black can equalise with 9 . . . 'iVas! 10 �b3 �g7 1 1 �bs �xes! 12 �xes 0-0. 1 0 ttJxc6 bxc6 1 1 �a4 !

26 .l::!. x c8 Rxc8 27 \t>h 1

27 'iVxb4 ctJe2+ wins. 27 . . . \t> h 7 ! 28 'f1.a 1

Or 28 'iVxb4 ctJxa2! 28 . . . a5 29 �b3 �xb3 30 axb3 g 5 3 1 Ji. g 3 a4 see

follo wing diagram

32 ttJd2 a3 33 l;l c 1 e5 34 d 5 a2 3 5 'f1.a 1 e 4 3 6 d 6 'f1. a 8 3 7 ttJc4 ttJb5 3 8 .Jt e 5 ttJxd6 39 l:!xa2 .l::!. x a2 4 0 Ji.xd6 gxf2 41 0-1

Ji.xb4 l:!f 1 + 42 \t>h2 l;lb 1

A very awkward move: Black 1 15

Th e Sla v

suddenly finds himself in big trouble. This is a good illustration of what can happen to Black if he does not take enough care.

�e8 22 VJlib3 �b8 23 VJlia2 �h8 24 .!:!:fd 1 VJlia8 25 h4 h 5 26 '!:!: d e 1 exd4 27 exd4 .!:!:e4 28 �e2 � h 7 29 J::t d 1 �a7 3 0 iLe7 .!:!:a6 3 1 a 5 .!:!:xd4

1 1 . . . SL d 7 1 2 iLa6 ! e5 1 3 �a3 exd4 1 4 exd4 0-0 1 5 0-0 iLe8

3 2 SLb6

Black has to shed a pawn to meet the threat of iLb7, winning the ex­ change.

Now the inevitable advance of the white a-pawn swiftly proves to be decisive. 32 . . . .!:!:xd 1 + 33 .!:!:xd 1 VJlib7 34 �c5

1 6 iLxe8 .!:!:xe8 1 7 VJlixa7 f6 1 8 a4 e 5

�a8 3 5 �xd 5 VJlixd 5 36 �xd 5 f5 37

1 9 SL g 3 � a 8 20 VJlie5 �a5 2 1 VJlie3

�d8 .!:!:a6 38 �d7 �g8 39 �a7 1 -0

1 16

Th e Ex c h a n g e Va ria tio n

S u m m ary

All three major continuations of the Exchange variation seem to be doing fine for Black. Personally I would recommend either 6 . . . a6 or 6 . . . .iiJ5 , as 6 . . .liJe4 leads to positions that are more typical of the Griinfeld than the Slav. 1 d4 d5 2 c4 c6 3 cxd 5 cxd 5 4 11. f4 CLlc6 5 e3

5 liJf3 liJf6 6 liJc3 6 . . . a6 (D) 7 liJe5 game 66 7 �c 1 game 67 6 . . . liJe4 7 e3 liJxc3 8 bxc3 g6 (D) 9 �e2 game 70 9 liJe5 game 71 -

-

-

-

5 . . . CLlf6 6 CLlc3 1Lg4

6 . . . �f5 game 65 -

7 f3 lL d 7 8 lL d 3 e6 (D)

9 �g3 game 68 9 g4 game 69 -

-

6. . . a 6

B

. . .

g6

B

. . .

e6

1 17

CHA PTER TEN Move-Orders an d T ranspos itions

1 d4 d5 2 c4 c6

This chapter is extremely impor­ tant both for white players, and for black players who wish to play lines with . . . d5xc4. Until recently, it was thought that White had no need to be too accurate with his move-order. However, due to the efforts of Ivan Sokolov, this is no longer true. We shall first consider 3 e3 . With this move White aims to remove the force from . . . d5xc4 (White can simply recapture with the bishop and has no need to play a2-a4, preventing . . . b7b5) , and thus to prevent Black from entering the Slav. Usually, the game continues 3 . . .cbf6 4 e3 e6, leading to the Semi-Slav, a very interesting open­ ing but not the one we want to play. The following game shows the way to meet 3 e3 .

Game 72 K rasen kov-I . S o kolov

Malmo

1 995

1 d4 d 5 2 c4 c6 3 e3 jLf5 !

I n the introduction, I said that . . . �f5 can only be good for Black if he can defend the b-pawn with his queen. Sokolov noticed that after 4 1 18

cxd5 cxd5 5 iYb3 , 5 . . :Vllic 7 is possible as 6 iYxd5 loses to 6 . . . iYxc 1+. 3 . . . ilfs is not the end of the world for White, but he has only very small chances of gaining an advantage once Black has developed his queen's bishop outside the pawn chain, while White has shut his inside.

4 �b3 �c7 5 cxd 5 cxd 5 6 CiJc3

6 ,�b5+ is considered in the next game. 6 . . . e6 7 jLd2 CiJc6 8 CiJf3 CiJf6 9 jLe2 a6 1 0 0-0 jLd6 1 1 gfc 1 0-0

Black has developed very naturally, and has equalised comfortably. 1 2 a3 �e7 1 3 jLe 1 h6

Ivan Sokolov suggests that 13 . . . MadS, aiming for . . . e6-e5, was most accurate.

M o v e - O r d e rs a n d Tr a n sp o s i tio n s

1 4 ctJa2 gac8 1 5 ctJb4 ctJa5 1 6 �a2 iLc7 1 7 ctJe5 ctJe4 1 8 ctJbd3

attack continues despite Black's deficit of a rook. However, White had to try this as 25 iiLd3 ? , trying to deflect the bishop on e4 from its protection of the d5-pawn, simply loses a piece. Sokolov finishes very efficiently. 2 5 . . :�xd8 26 dxe5 ctJdc4 27 iLxe4 dxe4 28 .l:!.c 1 r;t>h8 29 �b 1 ctJxe5 30 .l:!.xc8 �xc8 3 1 �xe4 ctJac6 32 h 3 ctJxb4 33 axb4 ctJ c 6 3 4 h4 � d 8 3 5 h 5 �e7 36 �f 5 �xb4 3 7 �c8+ r;t> h 7 38 ,, 5 hxg5 39 �f 5+ r;t>g8 40 �e6+ r;t>f8 41 �f5+ r;t>e8 42 �g6+ r;t>d8 43 �xg 7 �g4+ 44 r;t>f 1 �xh 5 4 5 �x b7

Somewhere around here, White of­ fered a draw, but Black was having none of it! 1 8 . . .f6 1 9 iL b4 ctJd6 20 ctJf4 gfd 8 ! 21

g4 iL e 4

22 .l:!. x c 7 ?

�xc7

23

ctJxe6 � e 7 2 4 ctJ x d 8 fxe 5

�h 1 + 0-1

46 �e2 t2Jd4+ wins the queen on a8 . A very important game: Sokolov won his last four games in the tour­ nament to pip Krasenkov by half a point!

Game 73 N ogueiras-I . Sokolov

Erevan Olympiad 1 996 1 d4 d 5 2 c4 c6 3 e3 2Lf5 4 �b3 �c7 5 cxd 5 cxd 5 6 2L b 5+ ctJd 7 ! ?

2 5 iL d 3 ?

After the natural 25 iiLxa5, winning a piece, 25 . . . nc2 26 iiLdl �h4 gives Black a very dangerous attack: 27 iiLxc2 �xg4+ 28 �f1 �g2+! (28 . . . iiLxc2 29 �xd5+!) 29 �e2 �g4+ 30 �e l �gl+ leads to a draw by per­ petual, while Ivan Sokolov also men­ tions 29 . . . iiLxc2!? when, with threats of . . . t2Je4 or . . . t2Jc4 and . . . �g4+, the

7 ctJc3 e 6 8 2Ld2 ctJe7 ! ?

An interesting development o f the knight. 1 19

Th e Sla v

9 J:rc 1 ctJc6 1 0 ctJa4 J::t c 8 1 1 ctJe2

square, away from the centre.

iLe 7 1 2 0-0 0-0 1 3 ctJf4 ctJdb8 ! ?

4 e4

There is no way that White is going to get through on c6! 1 4 iL d 3 iLxd 3 1 5 ctJxd3 iLd6 1 6 93 �e7 1 7 ctJ c 3 a 6 1 8 ctJa4 �c7 1 9 Mc2 Y2 - Y2

Exchanges will follow on the open c-file. The next move that we shall exam­ ine is 3 tLlc3, which has exactly the same idea as 3 e3: after 3 ... CLJf6, 4 e3 would prevent . . . dSxc4 and lead to the highly theoretical pastures of the Semi-Slav after 4 ... e6. This was the main move-order for several years until that man Ivan Sokolov again got cracking.

Game 74 · LautieH .So kolov · Groningen 1995

The most natural continuation: White gets in his desired e2-e4 as soon as possible. If White plays 4 a4 in 0[­ der to prevent . . . b7-b5, then Black can exploit the fact that White has not played CLJf3 by playing 4 ... e5! 5 e3 (5 dxe5 �xd1 + is very nice for Black) 5 . . . exd4 6 exd4 �e6! Instead 4 e3 b5 5 a4 b4 6 CLJe4!? is the critical test of this idea, when 6 ... �dS (not 6 . . . �a6 7 CLJc5!; but 6 . . . CLJf6!? 7 CLJxf6 + exf6 8 �xc4 �d6 9 'Wic2 0-0 1 0 �d3 g6 led to a quick draw in Epishin-Sakaev, 5t Petersburg 1997) 7 CLJg3 CLJf6 (7 ... e 5!?) 8 CLJf3 �a6 (8 ... h5!?) is very unclear. 4 . . . b5 5 a4 b4 6 ctJa2 ctJf6 7 f3

Advancing with 7 e5 is possible, al­ though 7 ... CLJdS 8 �xc4 e6 9 CLJf3 ile7 10 �d2 as 1 1 CLJc1 CLJdl 12 CLJb3 ilb7 13 'Wie2 c5! was absolutely fine for Black in Yakovich-Sadler, European Team Championship, Pula 1997.

1 d4 d 5 2 c4 c6 3 ctJc3 dxc4!

7 . . . e 5 8 iLxc4?

Once White has already committed his knight to c3, the sequence . . . b7-b5, a2-a4 (to regain the pawn) ... b5-b4 gains a tempo on the knight on c3, which must then move to an inferior

White sacrifices a pawn, but he must have missed something, as he never gets anything for it. The correct move is 8 dxe5, which is considered in the next game.

1 20

M o v e - O r d e rs a n d Tra n sp o s i t io n s

ing path for White. Game 75 Hjartarson-Gulko

Reykjavik 1 996 1 d4 d5 2 c4 c6 3 liJc3 dxc4 4 e4 b5 5 a4 b4 6 liJa2 liJf6 7 f3 e5 8 dxe 5 ! �xd 1 + 9 �xd 1 liJfd 7 1 0 e6 ! ?

8 . . .'�Vxd4 9 �c2

9 'i'b3 , attacking f7, is strongly met by 9 . . . iilcS ! 10 iilxf7+ cJ;; e 7, when Black's threat of . . . 'i'f2+ gives him an overwhelming position. 9 . . . it c 5 1 0 it g 5 ita6 ! 1 1 itb3 ite7 1 2 liJe2 �b6 1 3 liJg3 h 6 1 4 itd2 g6 1 5 � c 1 liJfd 7 1 6 liJf 1 liJ c 5 !

Hjartarson suggests that 10 f4 iila6 1 1 iile3 f6 is unclear, but 10 . . . CLJcS may be stronger, meeting 1 1 iilxc4 by 1 1 . . .iila6! 12 iilxa6 CLJbxa6 with ideas of . . . CLJb3 and . . . CLJxa4 as well as . . . CLJxe4. 1 0 . . . fxe6 1 1 itxc4

1 7 a 5 �b7 1 8 liJxb4 liJxb3 1 9 �xb3

White has the more attractive pawn structure and hence the better long­ term prospects, but Black's piece ac­ tivity and the slightly open position of the white king should give him equally good chances.

c5!

1 1 . . . ita6

This wins a piece due to the pin on the knight.

1 2 itxa6 liJxa6 1 3 ite3 it c 5 1 4 �e2

White's position is truly horrible.

20 �d 5 cxb4 2 1 �xe 5 f6 0 - 1

White players will obviously not want to repeat this experience. A re­ cent game has shown a more interest-

Also possible was l 1 . . .CLJeS! ? see

follo wing diagram

1 4 . . 0-0-0? .

Hjartarson says that Black should have taken this opportunity to swap 12 1

Th e Sla v

off bishops by 14 . . . ii,xe3 1 5 �xe3 0-0-0 16 ctJh3 ctJe5 17 Mac 1 with mu­ tual chances. In the game, Black fails to make the most of his chances and falls into an unpleasant ending.

4 1 e5 �c6 42 �f2 �e4 43 ne2 ng4 44 e6 1 -0

Nice technique from Hjartarson, and a critical new idea for White. In view of Ivan Sokolov's success with 3 . . . dxc4 against 3 ctJc3 , you may be wondering whether Black can play 3 . . . dxc4 against 3 ctJf3 but, amongst others, the following game has put the line under a cloud. It is so unbalanced, however, that I would not be at all surprised if there is a sneaky resource for Black!

Game 76 Miles-Hodgson

Hastings 1 995/96

1 5 .l1i. g 5 ! ctJf6 1 6 b3 .l1i.d4 1 7 n c 1 �b7 1 8 ctJ h 3 e 5 1 9 �hd 1 c5 20 .l1i.e3 .l1i.xe3 21 nxd8 �xd8 22 �xe3 h 6 23

1 ctJf3 d 5 2 d4 c6 3 c4 dxc4

ctJf2 gd6 24 gc2 ctJd7 25 ctJc 1 h5

3 ctJf3 ii,f5 4 cxd5 cxd5 5 'i'b3 'i'c7 is tactically possible as 6 'i'xd5 allows mate after 6 . . . 'i'xc 1+. The crucial dif­ ference with the 3 e3 line is that White has not blocked in his dark­ squared bishop, which means that after 6 ctJc3 e6, 7 ii,f4! is extremely nasty: 7 . . . 'i'xf4 loses the rook on as after 8 'i'xb7, while 7 . . . 'i'b6 8 'i'xb6 axb6 9 e3 , intending ii,b5+, �e2 and then Mhc 1 to invade on the c-file, gives Black a very depressing ending to defend.

26 ctJfd3 g5 27 ctJ b 2 g4 28 ctJcd 3 ! gxf3 29 gxf3 g g 6 30 ctJc4!

4 e3 b 5 5 a4!

White's pieces are ideally placed, and Black can no longer hold his pawn weaknesses. 30 . . . ng 1

31

nb2 ctJc7 3 2 ctJcxe5

ctJxe5 3 3 ctJxe5 ne 1 + 34 �f2 gh 1 3 5 �g2 nc 1 36 f4 ctJe6 37 f5 ctJg5 38 f6 �c7 3 9 f7 ctJ h 7 40 ctJg6 .!;i.e 1 1 22

see

fol/o wing diagram

The and 3 vance tempo White them.

main difference between 3 ctJc3 ctJf3 is that Black cannot ad­ his queenside pawns with against 3 ctJf3 . This gives plenty of time to undermine

M o v e - O r d e rs a n d Tra n sp o s i tio n s

1 5 0-0 "¥IlIc7 16 �f3 ! (16 ttJb3? ttJe4!, aiming for c3 , is to be avoided at all costs) 16 . . . �a7 17 cS is very nice for White. 1 2 . . . CLlfd7 1 3 c 5 ! 0-0 1 4 CLlec4!

5 . . . e6 6 axb5 cx b 5 7 b3 iL b4+ 8 iLd2 iLxd 2+ 9 CLlbxd 2 a5 1 0 bxc4 b4

A very confusing situation: Black has two passed queenside pawns while White has more central control. The essential conflict is whether Black can get his pawns moving or whether White can blockade them so that they will become weak.

ttJc6 20 0-0 Wifc7 21 Bfc 1 iLxc4 22

1 1 CLle 5 !

CLlxc4 CLlce7 23 �b3 Ba7 24 g 3 !

Exchanges help Black free himself. 1 4 . . . CLlf6 1 5 iLe2 CLl d 5 1 6 iLf3 iLd7 1 7 �c2 iLb5 1 8 CLld6 iLa6 1 9 CLl2c4

R b 8 2 5 �g 2 !

Very instructive: Black's pawns are going nowhere, so White quietly im­ proves his position, removing the possibility of back-rank mates and making sure that if Black does queen, it won't be with check. 25 . . . h6 26 Rc2 CL'lc3?

Threatening 12 ttJxf7! 'it>xf7 1 3 "Wf3+, winning the rook o n a8 . 1 1 . . . CLlf6 1 2 �a4+ !

This is the key idea, forcing Black's pieces into a nasty tangle. 12 . . . ttJbd7 loses a pawn to 13 ttJc6, while 12 . . . �d7 1 3 ttJxd7 is exactly what White wants: 13 . . . ttJbxd7 14 �e2 0-0 1 23

Th e Sla v

Allowing a combination, but Black was struggling to find anything to do.

b3 0-0 1 2 e4?

27 nxa 5 ! �xa 5 28 tLlxa5 �xa 5 29 �xc3 bxc3 3 0 �xb8+ \t>h7 3 1 �b7 f5 3 2 �xe 7 c 2 3 3 �xe6 c 1 � 34 �xf5+ g 6 35 �f7+ \t>h8 3 6 �f8+ \t>h7 37 .li d 5 !

Mate cannot be averted. 3 7 . . . �h 1 + ! ? 3 8 �xh 1 1 -0

The next line is 3 tLlf3 tLlf6 4 e3. I am surprised at how popular this variation is, as it really doesn't prom­ ise White very much at all, and his results have not really been that good.

Game 77 Kozu l - l il escas

Erevan Olympiad 1 996

1 2 . . . tLlc5 ! ! 1 3 dxc5 dxe4 ! 1 4 �xd8 �fxd8

The point: Black regains his piece with a vastly superior position, as the knight on f3 is tied to the defence of the bishop on d2. 1 5 tLla4 exf3 1 6 �fd 1 �d3 1 7 itc3

1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 e3

:ad8 1 8 ne1

.lif 5 !

20 tLlb2 �f3 21 \t>xg2 z:i.f5 2 2 t2Ja4

I don't think that this needs any comment!

�d3 23 .l:!.cd 1 .l:i:df3 24 �e2 h 5 ! 25

:8d7 1 9 :ac 1 fxg2

h3 g 5 !

This kingside advance finishes off the game. 26 b4 axb4 27 axb4 g4 28 hxg4 tLlxg4

5 .lid 3

5 cxd5 cxd5 6 iVb3 is the other way to play, leading to positions very simi­ lar to Game 72. 5 . . . .lixd3 6 �xd3 e 6 7 0-0 tLlbd7 8 tLlc3 1I.. b 4 9 1I.. d 2 a5 1 0 a3 .lie7 1 1 1 24

Now f2 must fall. 29 .l:i:h 1

:xf2+ 30 nxf2 tLlxf2 3 1

M o v e - O rd e rs a n d Tra n sp o s i tio n s

J:!.a 1 h4 3 2 tZ:lb6 h3+ 33 �h2 J::!. g 5 34 J:!.g 1 .l:!.xg 1

3 5 �xg 1 tZ:lg4 36 tZ:lc8

�f8 0 - 1

And finally, 4 'iVb3 and 4 'iVc2. In both case White's queen protects c4 and prevents the light-squared bishop from developing safely: 4 'iVb3 attacks b7, while 4 'iVc2 covers the f5-square. However, these moves do nothing to further White's development, while exposing the white queen to attack by Black's minor pieces. Although nei­ ther line promises much, they are both popular with positional players who wish to avoid any sharp options. Recently, black players have been trying a Semi-Slav approach, 4 . . . e6 5 g3 dxc4 6 'iVxc4 b5 7 'iVc2 j,b7, with . . .tDbd7 and . . . c6-c5 to follow, but as befits a book on the Slav, I will con­ centrate on our beloved plan of de­ veloping the light-squared bishop out­ side the pawn chain.

1 1 'iVxb4 attacks b7 and d4) 1 1 'iVxb4 e6 12 'iVc3 l:!.c8 ! , winning the queen, as 13 'iVd3 allows 1 3 . . . l:!.xc1+ mate! Instead of 6 . . . cxd5, 6 . . . CtJxd5! ? (hitting the bishop on f4) 7 j,g3 e6!?, intend­ ing a quick . . . c6-c5, is also interesting. Note that 5 CtJc3 transposes to 4 CtJf3 a6 5 'iVb3 . 5 �xc4 .itf5 6 tZ:lc3 tZ:lbd7 7 g 3 e6 8 .itg2 .ite7 9 0-0 0-0 1 0 e 3

Be warned! The 'natural' 10 l:!.d1 loses the exchange to 10 ... j,c2!, as 1 1 l:!.d2 allows 1 1 . . .CtJb6! , trapping the queen!

Game 78 Akopian-Shirov

Wijk aan Zee Open 1993 1 d4 d 5 2 tZ:lf3 tZ:lf6 3 c4 c6 4 �b3 dxc4

The most solid and reliable option. Black diverts the queen from its attack on b7, thereby allowing the light­ squared bishop to develop to f5. I suppose that 4 . . . a6 is possible here, as 5 j,f4 (preventing . . . l:!.a7!) 5 . . . b5 6 cxd5 cxd5 7 a4 b4! 8 CtJbd2 (8 'iVxb4 e5! wins a piece) 8 . . . CtJc6 9 l:!.c 1 j,b7 10 CtJe5 (to remove the knight on c6, which both guards b4 and blocks the c-file) is met by 10 . . . CtJa5! (10 . . . CtJxd4

1 0 . . . tZ:le4 1 1 �e2

A novelty. 1 1 CtJd2 is the old move, when theory recommends 1 1 . . .CtJxd2 12 j,xd2 e5! 13 e4 cxd4 14 exf5 dxc3 15 j,xc3 j,f6! 16 l:!.ad 1 ? ! j,xc3 1 7 bxc3 'iVc7 with equality. White must consider 16 j,b4!?, retaining the bishop pair. 1 1 . . . tZ:lxc3 1 2 bxc3 .ite4 !

A typical manoeuvre, preventing e3-e4 and neutralising White's bishop on g2. 1 3 c4 c5 1 4 ld:.d 1 wtic7 see

fol/o wing diagram

1 5 .itb2 tZ:l b6 ! ? 1 25

Th e Sla v

White i s material up, after all! Game 79 Razuvaev-Sturua

Erevan Open 1 996 1 d4 ltJf6 2 c4 c6 3 ltJf3 d 5 4 �c2 g6

The start of an eccentric plan from Shirov. 1 5 . . . MfdS , intending 16 . . . MacS and then . . . ct:Jd7-bs-c6 to pressurise c4 and d4, was also interesting. 1 6 gac 1 jfLf6 1 7 jfLf 1 !

To chase the bishop from e4 with­ out allowing the exchange of bishops. 1 7 . . . ltJa4 ! ?

1 8 jfLa 1

.l::i. f d8 1 9 ltJd2

A very sensible idea. Black allows his bishop to develop to g7 whilst supporting . . . �cS-f5 , attacking the queen. 4 . . . a6 5 �f4 b5?! 6 cxb5 cxb5 7 ct:Jbd2! (intending ct:Jb3) is best avoided, as White's pieces are well placed to exploit the queenside dark­ square weaknesses, so Black should try 5 . . . dxc4 6 'iVxc4 e6 7 e3 b5 S 'iVc2 �b7, intending . . . ct:Jbs-d7 and . . . c6-c5. Finally, 4 . . . dxc4 5 'iVxc4 transposes to the previous game.

jfLc6 20 ltJ b 3 b 6 21 d5

5 �f4 �g7 21 . . . jfLxa 1

2 2 dxc6 jfLf6

23 jfLg2

ltJc3 24 .l::i. x d8+ gxd8 2 5 �c2 gd6 26 .!de 1

gxc6 27 jfLxc6 �xc6 28

ltJd2 a6 29 ltJ b 1 ltJe4 30 h4 b 5 3 1 cx b 5 axb5 3 2 ltJ d 2 ltJc3 3 3 ltJ b 1

%-% I am a little surprised that White accepted the draw here. Although his opponent has good counterplay,

ltJe4 3 4 ltJ d 2 ltJc3

126

5 . . . ct:Ja6!? is considered in the next game and 5 . . . dxc4 is also good: 6 'iVxc4 �g7 7 e3 (7 ct:Jc3 0-0 S e4 b5 leads to a typical Griinfeld position which, though reasonable for Black, may not appeal to pure Slav players) 7 . . . 0-0 S �e2 �e6 9 iVc1 ct:Jbd7 10 0-0 c5! 1 1 ct:Jc3 ct:Jd5 ! 12 Mdl MCS 13 ct:Jxd5 �xd5 14 dxc5 ct:Jxc5 15 �c4 ct:Jd3 ! 16

M o v e - O r d e rs a n d Tra n sp o s i tio n s

g,xd3 g,xc4 1 7 'iVd2 e6 gave Black no problems in Goldin-Yusupov, Tilburg 1992. However, s . . . iiJs 6 'iVb3 'iVb6 7 cS 'iVxb3 S axb3 is a touch better for White, and not very exciting for Black.

axb3 are improved for Black, as the white knight is misplaced on gS, which makes . . . e7-eS easier to achieve.

6 e3 0-0 7 tLlc3 ilLe6

And finally, an amazing sacrificial line in this most solid of openings!

1 1 . . . ilLg4 1 2 cxd 5 tLlxd 5 1 3 tLlxd 5 cxd5 1 4 �xb6 tLlxb6 Y2 - Y2

Game 80 Alburt-S habalov

USA

Championship 1996

1 d4 d 5 2 c4 c6 3 tLlf3 tLlf6 4 �c2 g 6 5 ilLf4 tLla6 ! ? 6 e3 ilLf5 7 �b3 tLlb4 ! !

S tLlg 5

S 'iVb3 dxc4 9 �xc4 (9 'iVxb7 ctJdS ! ? , intending to trap the queen i n the corner after 10 'iVxaS 'iVb6, is ex­ tremely murky but not worse for Black) 9 . . . �xc4 10 'Vi'xc4 ctJds 1 1 �g3 ctJa6, intending . . . ctJdS-b6 and . . . c6-cS is about equal. S . . . ilLf5 9 �b3 �b6 1 0 ilLe2 tLlbd7 1 1 tLlf3

A typically inventive idea of Ku­ preichik. S �xb4 e 5 9 �x b7

9 cS cxf4 10 exf4 b6 1 1 ctJeS bxcs 12 'iVb7 �d7 13 ctJxd7 ctJxd7 14 'iVxc6 g,cS 15 'iVxds cxd4 16 �bs �b4+ (16 . . . g,c1+ 17 �d2 g,xh 1 1S 'Vi'eS+! 'iV e7 19 'Vi'xhS+ 'Vi'fs 20 'Vi' eS+ leads to a draw by repetition) gave Black a powerful lmtlatlve in Epishin­ Kupreichik, Russia 1989. 9 . g b S 1 0 �xc6+ ilLd7 . .

11

�xf 6 !

�xf6 1 2 ilL x e 5 � b 6 1 3 b3 ilL b4+ 1 4

The endings with 1 1 cS 'iVxb3 12

tLlbd2 0-0 1 5 ilLxbS J::!. x bS 127

Th e Sla v

wmmng. 1 6 . . . 'ii' a 5 1 7 �c4 �g4 1 8 0-0 �xd2

White has some pawns, but Black has the big guys! 1 9 CLJe5 �f5 20 a3 'ii' c 3 21 d 6 �g7 22 CLJxf7 .l::!. b 6 23 e4 �c8 24 .l:i:fd 1 �f4 25 e 5

1 6 cxd 5

A strange move to play since Shabalov had already won a convinc­ ing game in this line. 16 i,d3 is the theoretical recommendation, meeting 16 . . . i,g4 with 17 �e2 ik'a5 (threatening . . . i,b4xd2) 18 �hd l , which i s a bit of a mess. 17 O-O!? seems more natural, to counter 17 . . . ik'a5 with 18 h3 ! i,h5 (18 . . . i,xd2 19 hxg4 i,c3 20 �ac 1 i,b2 2 1 �c2 ik'xa2 22 4::l d2! [intending �b l] 22 . . . �xb3 23 4::lx b3 ik'xb3 24 �b l ik'xd3 25 �cxb2 dxc4 26 �c 1 , intending �bc2 with an advantage) 19 a3 ! i,xd2 20 b4! i,xb4 21 axb4 ik'xb4 22 �ab l ik'f8 23 g4! ,

1 28

25 . . Jbb3 26 � x b 3 'ii'x b3 27 CLJd8 'ii' b 6 28 g 3 'ii' x d8 29 g xf4 'ii' h 4 30 f3 'ii'x f4 3 1 �f2 �b7 3 2 .l:i:d3 �a6 3 3 d7 'ii'x h2+ 34 �e3 'ii' h 6+ 35 �e2 'ii' h 4 3 6 �d2 �f7 37 d5 �e7 38 e6 'ii'f 2+ 39 �c3 'ii' c 5+ 40 �d2 �xd3 41 .l::!. c 1 'ii' d 4 42 .l::!. c 8 � b 5+ 43 �c2 �xd7 0- 1

A fascinating game.

M o v e - O rd e rs a n d Tra n sp o s i tio n s

S u m mary

3 e3 JLfs equalises for Black; 3 CDc3 dxc4 is a very interesting line and only Hjartarson-Gulko (Game 7S) is a possible attempt to play with White; 3 CDf3 dxc4 is very unbalanced but seems to be good for White; 3 CDf3 CDf6 4 e3 JLfs is nothing for White; while 3 CDf3 CDf6 4 'iVc2 and 4 'iVb3 are also nothing spe­ cial. Hence 3 CDf3 CDf6 4 CDc3 dxc4 is the most accurate order for both sides. 1 d4 d 5 2 c4 c6 3 e3

3 CDc3 dxc4 4 e4 bs S a4 b4 6 CDa2 CDf6 7 f3 eS (D) 8 JLxc4 game 74 8 dxeS game 75 3 CDf3 3 . . . dxc4 game 76 3 . . . CDf6 4 e3 JLfs game 77 4 'iVb3 game 78 4 'iVc2 g6 S JLf4 (D) S . . . JLg7 game 79 S . . . CDa6 game 80 -

-

-

-

-

-

-

3 . . . i1Lf5 4 �b3 �c7 5 cxd 5 cxd 5 (D)

6 CDc3 game 72 6 JLbS + game 73 -

-

7

. . .

e5

5 �f4

5 . . . cxd5

1 29

CHA PTER ELEVEN Odd s and Ends

1 d4 d5 2 c4 c6

In this chapter, we take a brief look at all the sidelines that I couldn't fit into the main chapters! A)

T h e Winawer Cou nter­ G a m b i t : 3 CL'lc3 e5

This gambit was all the rage four or five years ago, but the following game somewhat dampened the ardour of the black players.

Game 81 . Kasparov-N ikolic

Manila Olympiad 1 992

7 . . . \'l;Vd 6

Black has also tried 7 . . . cS, but after 8 CL'lf3 Vic7 9 e3 dxe3 1 0 �xe3 CL'lf6 1 1 �d3 White i s slightly better due t o his lead in development.

1 d4 d 5 2 c4 c 6 3 ltJc3 e 5 4 dxe5

8 ltJf3 ltJf6 9 \'l;Vc2 �e7 1 0 O-O-O!

d4 5 ltJe4 \'l;Va5+ 6 � d 2 !

0-0

This natural move had been practi­ cally ignored prior to this game (6 CL'ld2 had been the main line) . 6 . . . . \'l;Vxe 5 7 ltJg 3 !

This move looks obvious, but it was Kasparov's new idea. Of course, the genius is not in the move itself, but in the astonishing attacking plan that flows from it. It is not easy to guess that White is gearing up for kings ide attack! 1 30

O d ds a n d En ds

1 1 e3

Azmaiparashvili, a long-time ana­ lyst for Kasparov, played 1 1 jLc3 agamst Eslon in Seville 1994. 1 1 . . .dxc3 !? 12 �xd6 cxb2+ is given as unclear by ' Gazza' , but 'Azmai' claims an advantage after 1 3 'iVxb2 jLxd6 14 e4 (threatening e4-e5) 14 . . . jLf4+ 15 'It>b 1 . This assessment is objectively correct, as White's mate­ rial advantage, coupled with the threat of e4-e5, driving away the knight on f6 and exposing the bishop on f4 to attack by White's pieces ('iVb2-d4, ctJg3-h5) , should tell in the end. How­ ever, in a practical game White's dark­ square weaknesses and exposed king give Black definite counter-chances. Eslon played 1 1 . . .'iVf4+, unpinning, and White sacrificed a pawn with 12 e3 dxe3 13 fxe3 (13 �d4!? looks very strong, meeting 13 . . . 'iVh6 with 14 �h4! 'iVg6 1 5 jLd3 ! , trapping the queen, and 13 . . . 'iVc7 with 14 jLd3 ! ? or just 14 fxe3 with very dangerous at­ tacking play) 1 3 . . . 'iVxe3+ 14 'It>b 1 ctJa6 15 ctJd4 (aiming to put a knight on f5 and threatening �de 1 , winning the bishop on e7) .

17 ctJxf5 'iVe6 18 jLd3 White had a superb attacking position for the pawn. This looks very smooth and, in his annotations, Azmaiparashvili comments that if Black tries to im­ prove with 15 . . . ctJb4 (instead of . . . jLb4) then 16 'iVb3 c5 17 �e 1 wins after 17 . . . 'iVgS 18 �xe7 cxd4 19 jLxb4, but in fact 16 . . . jLcS! is extremely good for Black, as 17 jLxb4 'iVxb3 18 ctJxb3 jLxb4 and 17 �e 1 'iVf4! (threatening . . .jLxd4) 18 jLxb4 'iVxd4 both leave Black simply a pawn up . The position is extremely risky for Black, of course, and I would not recommend this sort of position against a good attacking player, but Black's tactical chances must not be underestimated. 1 1 . . . dxe3 1 2 fxe 3 !

This is the key to the whole idea as 12 jLxe3 'iVc7 promises nothing for White. The text allows the bishop on d2 to move with tempo to the attack­ ing diagonal a1-h8. 1 2 . . .'�c7

12 . . . ctJa6 is considered in the next game. 1 3 SLc3 SLg4? !

1 4 SLd3 tLlbd7 1 5

SLf5 !

Now after 1 5 . . . jLb4 16 ctJdf5 jLxf5

Naturally, White wants to occupy 131

Th e Sla v

the f5-square with his knight. 1 5 . . . 1l.xf5 1 6 CLlxf5 ZUeS 1 7 CLlxg7 ! ! �xg7 1 S �f5 CLlfS 1 9 h4!

If Black wishes to play this line, he needs to find an improvement on 12 . . . �c7.

Game 82 Rogozenko-Bets

Moldovan Championship 1 994 1 d4 d 5 2 c4 c6 3 CLlc3 e 5 4 dxe5 d4 5 CLle4 �a5+ 6 1l.d2 �xe 5 7 CLlg3 CLlf6 S CLlf3 �d 6 9 �c2 1l.e7

10

0-0-0 0-0 1 1 e3 dxe3 1 2 fxe3

Following Kasparov's example. In order to chase away the knight on fs if it should surface on g6. 1 9 . . . h 6 20 g4? !

20 �g4! iDg6 2 1 h5 was even stronger according to Kasparov. 20 . . :�cS 21 �xcS .!:!:axcS 22 g 5 !

1 2 . . . CLla6

12 . . . .sdS is possible, preventing 13 Jtc3 as 13 . . . �xd1+ 14 �xd1 .sxd1+ 15 'ii;>x d1 exchanges queens, destroying White's attacking possibilities. 1 3 1l.c3 �c7

The ending i s still very unpleasant for Black, and Kasparov powers through with his customary energy.

Since this turns out so badly, Black must consider 1 3 . . . �e6, keeping the queen close to the kingside to help with defence, while threatening . . . iDb4 and . . . iDg4 as well as . . . �xe3+.

22 . . . CLlSd7 23 e4 �cdS 24 � d f 1 �fS

14 a 3 !

25 gxf6 1l.xf6 26 e5 1l. g 7 27 .i:i.hg 1 c5 2S �c2 .!:!:e6 29 .!:!:g4 1l.hS 30 b4

see

follo wing diagram

.!:!:b2 1l. g 7 34 �b7 .i:i.xa2+ 3 5 � b 3

Preventing 14 . . . iDb4, activating the knight.

1;!a6 3 6 e6 .l::!. x e6 3 7 .!:!: x g 7 1 -0

1 4 . . . CLlg4 ! ? 1 5 '!:!:e 1 1l.f6

b6 3 1 bxc 5 bxc5 32 .!:!:b 1 �a6 3 3

1 32

O dds a n d En ds

A nswer: This is another Smyslov favourite. Black accepts a slight space disadvantage and develops his king's bishop on g7, avoiding . . .e7-e6 in or­ der to retain the option of developing his bishop on c8 outside the pawn chain if he wishes.

Game 83 G u l ko-Salov

Reykjavik (World Cup) 1991 This manoeuvre, exchanging the dangerous bishop on c3, briefly gave black players hope in this line. 1 6 h 3 ! jLxc3 1 7 hxg4

Obviously 17 . . . .txe l loses to 1 8 iYxh7+ mate.

1 l2Jf3 d 5 2 c4 c6 3 e3 l2Jf6 4 l2Jc3 g 6 5 d4 jLg7 6 jLe2 0-0 7 0-0 b6

A solid move, developing the bishop to b7 to support Black's cen­ tre. The more active 7 . . . dxc4 8 .txc4 .tg4 is dealt with in the next game.

1 7 . . . h 6 1 8 �xc3 �xg3 1 9 g 5 ! hxg 5 20 jL d 3 �d8 2 1 l:!.h7 f6 22 l:Ieh 1 �c7 23 l2Jxg 5 �f8 24 c 5 ! jLg4 25 �xf6+! 1 -0

8 cxd 5 cxd 5 9 l2Je 5 !

25 . . . gxf6 26 �h8+ cJi;e7 27 �lh7+ is mate! B)

The S c h leeter S l a v : 3 CD c 3 CD f 6 4 e3 9 6

Question 1 : What is the idea behind

this system?

When the central pawn structure becomes fixed, the first side to gain a central space advantage will be able to claim the initiative. 9 . . . jLb7 1 0 jLd2 l2Jfd 7 ? !

10 . . . CDc6 was stronger, meeting 1 1 f4 with 1 1 . . .CDe8 ! , intending to de­ velop the knight to d6, when Black only stands a little worse. 1 1 f4 ! f6 1 2 l2Jf3 ! ?

1 2 CDd3 i s also possible. Black has 1 33

Th e Sla v

not developed his pieces harmoni­ ously: the king's knight stops the queen's knight from developing to d7 and has no moves of its own.

iLxe7+

CL:Jxe7

34

W'h6+

CL:Jg6

35

W'xg6+ � e 7 36 W' h 7 + �d6 3 7 CL:Jf7+ rt;e7 3 8 iLf5 1 -0

Game 84

1 2 . . JH7 1 3 iLd3 CL:Jf8

D reev-Piket

Dortmund 1994 1 d4 CL:Jf6 2 e4 e6 3 CL:Jf3 d 5 4 e3 g6 5 CL:Je3 iLg7 6 iLe2

White can try 6 �d3 (preventing . . . �f5) 6 . . 0-0 7 h3 (preventing . . . il,g4) if he wants to prevent the plan in the game, although 7 . . . c5 is an interesting reply. The game transposes to a quiet variation of the Griinfeld, in which White has played the useful, though hardly earth-shattering, extra move h2-h3 . .

1 4 f5? !

A little hasty according to Gulko, who prefers 14 g4! e6 15 f5 . 1 4 . . . gxf5 1 5 iLxf5 e6 1 6 iLd3 CL:Je6 1 7 CL:Je2 �d6 1 8 CL:Jg3 CL:Jg6 1 9 CL:Jh5 iLh8 20 W'e2 gaf8 21 a 3 e 5 22 dxe5

6 . . . 0-0 7 0-0 dxe4 ! ? 8 jLxe4 iLg4

This is a much more active idea: Black will follow up with . . . 4Jbd7 and a quick . . . e7-e5.

fxe 5 23 CL:Jg5 gxf 1 + 24 .i:l:.xf 1 iLe8? 2 5 .i:l:.xf8+ W'xf8 26 iLe4!

9 h 3 iLxf3 1 0 W'xf3 CL:Jbd7 1 1 iLb3

A very nice move: 26 . . . dxe4 loses to 27 "ik'c4+. 26 . . . CL:Jee7 2 7 iLb4!

iLb7 28 W'g4

\�i"c8 29 W'f3 �f8 3 0 �h3 �e8 3 1 CL:Jf6+ 1 34

�g7

32

W'xh 7+

�xf6

33

The critical line is 1 1 gd1 e5 12 d5 e4!? 13 4Jxe4 4Jxe4 14 "ik'xe4 4Jb6 15 �b3 (threatening d5xc6; Bareev sug­ gests 15 gb 1 !?, protecting b2 and seek­ ing to avoid the time-loss with �c4b3xd5 as in the game) 15 . . . cxd5 16

O dds a n d En ds

,�xds CiJxds 17 Mxds 'i'b6.

ity and gains space. 1 8 '¥fie2 tLl b 6 ! 1 9 gxd8 gxd8 20 ge 1 tLlc4 !

Imperceptibly, White has drifted into big trouble: Black controls the central dark squares and White's queenside is an easy target.

Question 2: What is going on here? A nswer: Black has sacrificed a pawn

for a lead in development and pres­ sure against the b2-pawn. However, this play is sufficient to regain the pawn, but not to achieve complete equality: i s 'i'd3 MadS 19 e4! itd4 20 Wh l itxf2 2 1 ith6 Mxds 22 exdS Mds 23 Mdl 'i'd6 24 'i'c3 , intending 'i'g7+ mate, when White's passed d-pawn and Black's weak kingside gave White a slight advantage in Bareev-Kramnik, Novgorod 1994.

2 1 jLc 1 tLle6 22 '¥fif 1 '¥fic5 23 jLb3 tLld2 24 jLxd2 gxd2 2 5 jLxe6 fxe6 26 e 5 gxb2 27 tLle4 '¥fixe 5 28 '¥fid3 IWd 5 29 IWg3 jLe5 3 0 f4 jLd4+ 3 1 �h2 c 5 3 2 \\!,Vh4 gxa2 3 3 \\!'ve7 h 5 !

1 1 . . . e 5 1 2 gd 1 '¥fi e 7 1 3 e4? ! exd4 1 4 gxd4 gad8 1 5 jLe3 tLlc5 1 6 jLc2 tLlfd7 1 7 gdd 1 b 5 !

3 4 \\!'ve8+ � g 7 3 5 \\!'ve7+ � h 6 3 6 h4 IWf5 3 7 tLlg3 \\!'vxf4 38 gxe6 g a 1

Here White lost on time. Black activates his queenside major-

What I like about the Schleeter Slav is its flexibility: on move 7, Black has 1 35

Th e Sla v

a huge range of plans. We have seen: a) 7 . . . b6 reinforcing the centre by fianchettoing the other bishop; and b) 7 . . . dxc4 8 i1Lxc4 i1Lg4 to break quickly in the centre by means of . . . tLlbd7, . . . i1Lxf3 and . . . e7-e5; but Black can also try c) 7 . . . a6 to expand on the queenside with . . . b7-b5, before or after . . . d5xc4; and d) 7 . . . tLle4!? to unbalance the pawn structure with . . . tLlxc3 . Bates-Sadler, British Championship 1995, contin­ ued 8 'iVb3 b6 9 tLlxe4 (9 cxd5 tLlxc3 ! 10 bxc3 cxd5 intending . . . tLlc6-a5, hit­ ting the queen) 9 . . . dxe4 10 tLld2 f5 1 1 f3 e5 1 2 dxe5 exf3 ! 1 3 tLlxf3 tLld7, re­ gaining the pawn with a better pawn structure; . . . and the list continues! There really is huge scope for personal ideas. C)

3 tLlf3 tLlf6 4 tLlc3 dxc4 5 e3

5 . . . b5 6 a4 b4 7 CLJb 1

7 tLla2, attacking the b4-pawn, re­ gains the pawn by force, but seriously misplaces the knight. The text is more ambitious. 7 . . . 1La6 8 1Le2

8 tLlbd2 regains the pawn, but after 8 . . . c3 9 bxc3 i1Lxf1 10 tLlxfl bxc3 Black will break with . . . c6-c5 and equalise. 8 . . . c 5 ! 9 0-0 CLJc6 1 0 dxc5 CLJ a 5 ! 1 1 CLJbd2 e6 1 2 b3 1Lxc5 1 3 1Lb2 c3 1 4 1Lxa6 cxb2 1 5 1L b 5+ cI;e7

This was one of my favourites when I was little, but I'm not quite sure why I decided to play it against Bareev! It is actually not a very promlSlng con­ tinuation.

Game 85 Sadler-Bareev

Hastings 1992/93 1 d4 d 5 2 c4 c6 3 CLJf3 CLJf6 4 CLJc3 dxc4 5 e3 see

fol/o wing diagram

Trying to manage without the standard 5 a4, which prevents the ad­ vance . . . b7-b5. 1 36

The king is very safe in the centre. Bareev rapidly outplays me, but around the time control begins to ask too much of his slight advantage. 1 6 .i:t b 1 a6 1 7 1Le2 �b6 1 8 .i:txb2

O d ds a n d En ds

l2Jd 5 1 9 l2Jc4 l2Jxc4 20 .ltxc4 l2Jc3 21 �c2 �c7 2 2 �c 1 Ithd8 23 .l::i. d 2 a 5 2 4 � c 2 g 6 2 5 .l::i. c 1 .ltd6 26 g 3 .ite5 27 .itf 1 h 5 28 .it c 4 �c 5 29 h 4 .lt f 6 30 .itf 1 l:! a c 8 3 1 .l::i. x d8 l:!xd8 3 2 l2Jd2 .l::i. c 8 3 3 l2J c 4 g 5 34 hxg 5 �xg 5

As far as I am concerned, 5 e4 just loses a pawn, but some die-hards just keep on playing it! Foremost amongst them is the attacking Swedish player Tiger Hillarp-Persson. So here is one of his typical efforts.

3 5 � h 7 � h 8 3 6 � d 3 .i:!. d 8 3 7 �h7

Game

.l::i. h 8 3 8 �d3 �c5 39 J:!c2 l:!d8 40

86

�h7 h4 41 gxh4 Rh8 42 �d3 lad8

Hillarp-Persson - Acs

43 �h 7 .l::i. h 8 44 �d 3 .i:!.g8+ 45 .ltg2

Budapest 1996

.l::i. d 8 46 �h7 .l::i. h 8 47 �d3 :ad8 48 �h7 l2Jxa4?

1 c4 c6 2 l2Jf3 l2Jf6 3 d4 d 5 4 l2Jc3

Too risky. After the tactics, Black only just manages to hold on to the draw.

l2Jg 5 ! ?

dxc4 5 e4 b 5 6 e 5 l2Jd 5 7 a4 e6 8

The modern line. 8 axb5 tL'lxc3 9 bxc3 cxb5 10 tL'lg5 .i,b7 1 1 'i'h5 g6 1 2 'i'g4 .i,e7 i s the old continuation, when Black will follow up with . . . tL'ld7-b6-d5, while White plays for tricks!

49 .l::i. a 2 ! l2J b 2 50 l2Jxa5 �c 1 + 5 1 c;t>h2 lah8 5 2 l2Jc6+ �xc6 5 3 l:!a7+ c;t>d6 54 �xh 8 .itxh8 5 5 .ltxc6 c;t>xc6 5 6 .l::i. x f7 l2Jd3 5 7 h5 l2Jc5 58 h6 l2Jxb3 5 9 Itf8 .ite5+ 60 f4 .itb2 6 1 e4 l2Jc5 6 2 e 5 l2Jd7 6 3 h 7 l2Jxe5 64

8 . . . .lte7 9 h4 ! ? h6 1 0 l2Jge4 b4

fxe 5 .itxe5+ 65

10 . . . .i,a6 immediately is interesting, to avoid weakening the queenside too early. The text leads to absolute chaos, though I think that Black is fine.

c;t>g2 b3 66 h 8�

.itxh8 67 J:!xh8 c;t>c5 68 c;t>f2 c;t>c4 69 c;t>e2 b 2 7 0 lab8 c;t>c3 71

c;t>d 1

e5

YZ - YZ D)

T h e S l a v G a m bit : 3 tL'lf3 tL'lf6 4 tL'lc3 dxc4 5 e4 b 5 6 e 5 tL'ld5 7 a4 e6

1 1 l2J b 1 .lta6 1 2 �g4 g6 1 3 l2Jbd2 c3 14 l2Jc4 .itxc4 1 5 .itxc4 a 5 1 6 h 5 g 5 1 7 0-0 l2Jd7 1 8 bxc3 l2J7b6 1 9 .ltb3 l2Jxc3 20 l2Jxc3 bxc3 2 1 �e4 137

Th e Sla v

tiJ d 5 22 f4 gb8 23 ll.a2 gb4 24 fxg 5 �b6 ! 25 ll.xd 5 exd 5 26 �f3 IWxd4+ 27 ll.e3 �xe 5 28 IWxf7+

Here 6 . . . �c8 !? 7 tLl£3 takes us back to square one, but 7 e3 e5 8 ilxc4 exd4 9 exd4 ile7 10 0-0 0-0 1 1 h3 tLla6!? 12 ct'lf3 ct'lc7!, intending ... ile6, leads to an unclear position. In the next game we see Black allowing the exchange on f5.

Game 87 Savchenko-N inov

Cappelle La Grande Open 1 994 1 d4 d 5 2 e4 e6 3 tiJf3 tiJf6 4 tiJe3 dxe4 5 a4 .liLf5 6 tiJh4 e6 7 tiJxf5 exf5 8 e3 .liLd6 28 . . . �d8

29 gae 1

l:!e4 30

�b6+

� d 7 3 1 l:!xe4 �xe4 3 2 ll.e5 J:!. h 7 33 �f2 ll.xg 5 34 J:!.e1 �f 5 3 5 �e2 e5 3 6 �b5+ �e8 3 7 �e6+ ge7 38 IWa8+ � d 7 39 IWxd 5+ �e8 40 �a8+ �d7 41 gd 1 + ll.d2 42 �d5+ �e8 43 gf 1 .liLf4 44 �g8+ �d7 4 5 �d5+

1h - 1h

An amazing game! E)

3 tLl f 3 tLl f 6 4 tLl c 3 dxc4 5 a4 �f5 6 tLlh4 ! ?

This i s an interesting little idea that has even been tried by Kasparov. 1 38

This cannot be bad, but I wonder whether Black cannot develop more effectively. White's basic idea is to play iVf3 , h2-h3 and g2-g4, removing the f5-pawn and thus undermining Black's control of e4. In order to pre­ vent this plan, I would therefore sug­ gest protecting the f5-pawn with . . . iVd7, putting the bishop sensibly on e7 and then developing the queen's knight to b4 via a6. This line is of course very similar to Yusupov­ Kramnik from the 6 e3 main lines, but since White has taken on £5 so early, allowing Black to prevent e3-e4

O d ds a n d En ds

with a pawn on fS , rather than his pieces, Black has more flexibility with his piece placement: S . . . aS 9 iLxc4 CUa6 10 'i'f3 'i'd7 1 1 h3 CUb4 12 0-0 ,YLe7 is fine for Black.

ga7 40 gb8 ge7 4 1

9 1Lxe4 0-0 1 0 0-0 CL.lbd7 1 1 a 5 a6

gxb5 ga7 43 Ub4+ We5 44 £a4

28 . . . Wf6 29 gb4 gd7 30 gb6+ We5 31 Wf2 ge7 3 2 b4 we4 3 3 Ud6 h4 34 We2 g 5 3 5 Wd2 ge7 36 e6 bxe6 3 7 gxe6 ga7 38 gd6 gb7 39 gb6 b 5 axb5 42

1 2 h 3 h5 1 3 �f3 g 6 1 4 e4! CL.lxe4

Wd6 45 a6 We5 46 We3 f 5 47 ga5+

1 5 CL.lxe4 fxe4 1 6 �xe4

Wb6 48 gxd 5 Wxa6 49 gxf5 £g7 50 Wd3 Wb6 51 We2 g4 5 2 gh 5 gxh3 5 3 gxh3 £a7 54 Wd3 We6 5 5 gxh4 Wd5 56 Uh5+ We6 5 7 We4 ga2 58 gh6+ Wf7 59 Wf4 1 -0

3 iLf4

F)

With 3 iLf4 White intends e2-e3, when he will have safely developed his queen's bishop outside the pawn chain. The move . . . h7-hs has rather weak­ ened the black kingside and now 'i'xg6+ is threatened.

Game 88 Psakhis-Sadler

Megeve (peA rapidplay) 1 994

1 6 . . . wg7 1 7 1Ld2 CL.lf6 1 8 �f3 CL.l d 5 1 9 gfe 1 1L e 7 20 1Lxd 5 exd 5 2 1 1L b4 1L d 6 22 � e 5 �xe5 23 dxe5 �d7 24 gad 1 gad8 25 gd4 gfe8 26 Uxe8 �xe8 27 �e3 �xe3 28 fxe3

1 d4 d 5 2 e4 e6 3 �f4 dxe4!

Basically winning a pawn. 4 CL.le3 e6 5 e3 b5 6 �f3 �a5 7 g4 b4 8 CL.le4 b3+ 9 CL.le3 1Lb4 1 0 CL.lge2 �xa2 !

White's space advantage gives him a very pleasant rook ending.

Now after 11 Mxa2 bxa2 the pawn 1 39

Th e Sla v

queens!

ttJa4 26 �e 1 �xc3+ 27 � O .:gxg7

1 1 g c 1 �xb 2 1 2 e4 ttJf6 1 3 g 5 e 5

28 �xe6+ fxe6 29 �e4 exd 5 30

1 4 g x f 6 exf4 1 5 �xf4 � a 3 1 6 �e3

�e6+ �b8 3 1

b2 1 7 .:g b 1 �a2

�e4+ !

ttJg3 �c2 3 2 �g2

Threatening 1 8 . . . 'iVxb l+. 1 8 �d 1 �b3+ 1 9 �d2 ttJa6 20 fxg 7 gg8 2 1 �h6 �e6 2 2 � h 3 ttJ c 5 !

Making things safe! 33 �xe4 dxe4 34 .:ghd 1 gxd 1 gxd 1 �e5 0 - 1 23 �xh 7 0-0-0 2 4 d 5 cxd 5 2 5 exd 5

1 40

Not a good advert for 3 �f4!

35

O d ds a n d En ds

S u m m a ry

The Schleeter variation is a reasonable alternative to 3 . . . dxc4 after 3 CDc3 . However, note that Black can only play this line after e2-e3 by White, as 3 CDf3 CDf6 4 CDc3 g6 5 cxds cxdS 6 itf4! leads to a superior version of the Ex­ change variation, which is rather depressing for Black. White players looking for an offbeat alternative might care to examine 6 CDh4!? 1 d4 d 5 2 c 4 c 6 3 i2J c 3

3 CDf3 CDf6 4 CDc3 dxc4 (D) 5 e3 game 85 5 e4 game 86 5 a4 itfS 6 CDh4 game 87 3 itf4 dxc4 - game 88 -

-

-

3 . . .e5

3 . . . CDf6 4 e3 g6 5 CDc3 itg7 6 ite2 0-0 7 0-0 (D) 7 . . . b6 - game 83 7 . . . dxc4 game 84 -

4 dxe5 d4 5 i2Je4 'iVa5 + 6 3Ld2 'iVxe 5 7 i2Jg3 i2Jf6 8 i2Jf3 'iVd6 9 'iVc2 3Le7 1 0 0-0-0 0-0 1 1 e 3 dxe3 1 2 fxe3 (D)

1 2 . . . iVc7 - game 81 1 2 . . . CDa6 - game 82

4 . . . dxc4

7 0-0

1 2 fxe 3

1 41

INDEX OF COMPLETE GAMES

Adianto-Kramnik,

London (Intel Grand Prix) 1994 Wijk aan Zee Open 1993 Alburt-Shabalov, USA Championship 1996 Andersson-Epishin, Ter Apel 1 995 Atalik-Miles, Hastings 1 995/96 Bacrot-Smyslov, A lbert {sixth match game} 1 996 Beliavsky-Shirov, Erevan Olympiad 1996 Benz-Gretarsson, Oberwart Open 1 996 Cifuentes-Dreev, Wijk aan Zee 1 995 Cifuentes-Leyva, Cienfuegos 1 996 Dautov-Nikolic, Ter Apel 1 994 Dreev-Piket, Dortmund 1 994 Ehlvest-Schwartzman, New York Open 1996 Epishin-Pomes, Manresa 1 995 Gelfand-Lautier, Zurich 1 994 Gelfand-Nikolic, Manila Interzonal 1 990 Gelfand-Shirov, Dortmund 1 996 Gofshtein-Sadler, Ischia 1 996 Gulko-Salov, Reykjavik (World Cup) 1991 Hillarp-Persson - Acs, Budapest 1996 Hjartarson-Gulko, Reykjavik 1 996 Hodgson-Sadler, Ischia 1 996 Hiibner-Beliavsky, Munich 1 994 Illescas-Gelfand, Dos Hermanas 1 996 Ivanchuk-Bareev, Dortmund 1 995 Ivanchuk-Lautier, Linares 1994 Ivanchuk-Smyslov, Tallinn (rapidplay) 1996 Karpov-Hj artarson, Tilburg 1988 Karpov-Kramnik, Monte Carlo (blindfold) 1995 Akopian-Shirov,

1 42

49 125 127 111 95 69 1 04 64 1 02 115 77 134 23 75 30 82 55 27 133 137 121 110 24 56 18 22 62 45 16

In dex o f C o mp l e t e G a m e s

Karpov-Short,

Dortmund 1995 Novgorod 1994 Kasparov-Nikolic, Manila Olympiad 1992 Kasparov-Shirov, Dos Hermanas 1996 Khalifman-Georgiev.Kir, Elenite 1994 Klarenbeek-Rogers, Dutch Team Championship 1996 Kozul-Illescas, Erevan Olympiad 1996 Kramnik-Damljanovic, Moscow Olympiad 1994 Kramnik-Ivanchuk, Linares 1994 Kramnik-Ivanchuk, Monte Carlo (blindfold) 1996 Kramnik-Lautier, Linares 1994 Kramnik-Shirov, Dortmund 1996 Kramnik-Shirov, Vienna 1996 Kramnik-Short, Moscow (Intel Grand Prix) 1996 Kramnik-Short, Novgorod 1994 Krasenkov-Epishin, Bmo 1994 Krasenkov-Sapis, Polish Championship 1995 Krasenkov-Sokolov.I, Malmo 1995 Lalic-Sadler, Hastings 1995/96 Lautier-Bareev, Linares 1994 Lautier-Sokolov.l, Groningen 1995 Leitao-Beliavsky, Erevan Olympiad 1996 Miles-Hodgson, Hastings 1995/96 Milov-Sadler, Isle ofMan 1994 Nesterov-Imanaliev, Bishkek Zonal 1993 Nogueiras-Sokolov.l, Erevan Olympiad 1996 Novikov-Gretarsson, Berlin Open 1995 Oll-Anand, Biel Interzonal 1993 Parker-Hellsten, Copenhagen 1996 Piket-Ge1fand, Wijk aan Zee 1996 Portisch-Kramnik, Biel Interzonal 1993 Psakhis-Sadler, Megeve (PCA rapidplay) 1994 Pushkov-Epishin, Russian Championship 1995 Razuvaev-Sturua, Erevan Open 1996 Richardson-Sadler, Islington Open 1995 Rogozenko-Bets, Moldovan Championship 1994 Ruze1e-Thorsteins, Lyon (European Club Cup) 1994 Sadler-Bareev, Hastings 1992/93 Sadler-Ferguson, British Championship 1996 Sadler-Hodgson, Hastings 1995/96 Sadler-Miles, British Championship 1998 Savchenko-Ninov, Cappelle la Grande Open 1994

Kasparov-Bareev,

92 35 130 54 35 84 124 72

47 66 43 47 98 67 59 91 81 118 39 105 120 84 1 22 108 79 119 63 87 76 53 1 14 139 101 1 26 13 132 60 136 42 92 34 138 1 43

Th e Sla v

Schandorff-Hellsten,

Copenhagen 1996 Biel 1 991 Shirov-Nikolic, Wijk aan Zee 1 993 Sokolov.I-Bareev, Leon 1 995 Sokolov.I-Hellsten, Malmo 1 995 Sokolov.I-Shirov, Erevan Olympiad 1996 Strauss.D-Lakdawala, USA 1992 Topalov-Gelfand, Belgrade 1 995 Topalov-Gelfand, Dos Hermanas 1996 Vaiser-Nalbandian, Erevan Open 1996 Van der Sterren-Petursson, San Bernardino Open 1992 Van der Sterren-Shirov, Biel lnterzonal 1993 Ward-Levitt, British Championship 1995 Wells-Flear, Oakham 1 994 Xu Jun-Akopian, Moscow Olympiad 1994 Yusupov-Kramnik, Riga 1 995 Yusupov-Shirov, Zurich 1 994 Shirov-Bareev,

1 44

75 46 78 50 80 1 00 36 21 58 1 13 57 89 103 90 32 17 111