Attacking With 1.d4 - Dunnington

attackin�

with

by Angus Dunnington

EVERYMAN CHESS

First published in 2001 by Everyman Publishers pk, formerly Cadogan Books pk, Gloucester Mansions, 140A Shaftesbury A venue, London WC2H 8HD Copyright © 2001 Angus Dunnington

The right of Angus Dunnington to be identified as the author of this work has been asserted in accordance with the Copyrights, Designs and Patents Act

1988.

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, photocopying, recording or otherwise, without prior permission of the publisher.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library. ISBN 1 857 44 258 X

Distributed in North America by The Globe Pequot Press, P.O Box 480, 246 Goose Lane, Guilford, CT 06437-0480.

All other sales enquiries should be directed to Everyman Chess, Gloucester Mansions, 140A Shaftesbury Avenue, London WC2H 8HD

tel: 020 7539 7600 fax: 020 7379 4060 email: [email protected] website: www.everyman.uk.com

The Everyman Chess Opening Guides were designed and developed by First Rank Publishing.

EVERYMAN CHESS SERIES (formerly Cadogan Chess) Chief advisor: Garry Kasparov

Commissioning editor: Byron Jacobs

Typeset and edited by First Rank Publishing, Brighton. Production by Book Production Services.

Printed and bound in Great Britain by The Cromwell Press Ltd., Trowbridge, Wiltshire.

CONTENTS

I

Bibliography 4

Preface 5

1 Queen's Gambit Accepted with 3 e4 7 2 Queen's Gambit Declined and Slav Defences 41

3 1 d4 d5 2 c4: Black's Second Move Alternatives 67

4 King's Indian and Benoni: The Four Pawns Attack 80 5 Griinfeld Defence with 4 il.f4 101 6 Nimzo-Indian Defence with 4 f3 113

7 Dutch Defence 130

8 Other Defences 139

I

BIBLIOGRAPHY

I

Books

The Modern Benoni, Kinsman (Everyman 2000)

The Queen's Gambit Accepted, Ward (Batsford 1999)

Beating the King's Indian and Benoni, V aiser (Batsford 1997)

Queen's Gambit Declined, Sadler (Everyman 2000)

Modern Defence, Speelman and McDonald (Everyman 2000)

English Defence, King (Everyman 1999)

Periodicals Informator

ChessBase Magazine The Week in Chess Chess

PREFACE

I

It is significant that a definition of 'attack' is as follows: 'to take the initiative in a game'. I had this in mind when selecting the opening systems recommended in these pages, be­ cause the very nature of 1 d4 openings is such that the kingside is by no means the only sector of the board in which either side is able to engineer attacking prospects. It is a common misconception - usually with young players - that to attack is to pressure only the opponent's king (or the king's de­ fences), but a weakness is a weakness, and by concentrating on this or that vulnerable square or pawn in the enemy camp it is pos­ sible to win a game.

Nevertheless, you will notice anyway that throughout the book the lines tend to be quite aggressive, mostly with an emphasis on space, fluid development, inducing weak­ nesses in the enemy camp, generating an initiative etc. Above all, however, it is impor­ tant that a system is sound and that specific plans and features are relatively easy to

un-derstand.

In order to facilitate familiarisation with major characteristics (structures, areas of attack etc.) several of the systems involve similar plans from White, and I have deliber­ ately chosen variations that require certain moves (and move orders) to be employed. Additionally, for the sake of convenience, and to avoid confusion, I have altered the move orders of several games.

I must stress that this is not intended to be a watertight repertoire book, rather a tool with which to help the reader open 1 d4 with confidence.

For their thoughts and advice I would like to thank fellow Yorkshireman Richard Pal­

liser and IM Andrew Kinsman (an honorary

Yorkshireman since winning the 2nd White Rose Masters in 2000).

IM Angus Dunnington,

Castleford, June 2001

I

CHAPTER ONE

I

Queen's Gambit

Accepted with 3 e4

1 d4 d5 2 c4 dxc4 3 e4

Natural and popular, the QGA can be a very solid system to face. Black simply ac­ cepts the c4-pawn and reacts accordingly to White's chosen set up in the centre. Rather than allow Black restriction tactics after 3 tt:Jf3 lt:Jf6 4 e3 cS, for example, when White can easily see his attacking chances fade away

only to be left with an isolated d-pawn, we

will concentrate exclusively on the no­ nonsense 3 e4. This immediately erects a potentially troublesome pawn centre that Black must contest early or drift into passiv­ ity. Games 1-2 see Black use knights to chal­ lenge d4 and e4 respectively, while Games 3-4 feature two contrasting approaches begin­ ning with 3 ... cS. In Games 5-6 Black adopts the traditional 3 ... eS.

4 ll:lf3

The knee-jerk 4 dS is also possible, when 4 ... tt:JeS S ..tf4 lt:Jg6 6 ..tg3 eS 7 ..txc4 tt:Jf6 8 lt:Jc3 ..td6 9 .tbs+ ..t>f8 is not unlike the main game. 4 ..te3 is natural, when 4 ... lt:Jf6 S f3 eS

---. 6 dS lt:Jd4!? 7 ..txd4 exd4 8 'ifxd4 c6 9 lt:Jc3

Game 1

Van Wely-V . Milov 0-0 has been assessed as unclear. The

varia-French League 1999

._ ______________ ..td7 9 'ifxaS a6 pays very close attention to

1 d4 d5 2 c4 dxc4 3 e4 ll:lc6

Obstructing the c-pawn can be justified here because White no longer has e2-e3 to protect the d4-pawn. The immediate pres­ sure is designed to induce a concession from White, while a further challenge in the centre with ... e7-e5 is planned.

White's queen. Illescas-Sadler, Linares Z 1995 did much for the popularity of 3 ... lt:Jc6 when Black won after 10 tt:Jbl? lt:Jxe4 1 1 ..t>dl c3!, preventing the queen's escape. In­ stead of 10 tt:Jbl White has 10 lt:Ja4 lt:Jxe4 1 1 ..txc4 bS with chances for both sides. With 4 tt:Jf3 White avoids these complications in

favour of simple development, hoping to be the one who benefits most from Black's plan.

4 . . . i.g4 5 d5

This forcing advance seems to hold more promise of an advantage than the alternative S ii.xc4.

5 . . . ll:\e5 6 i.f4

Again White selects the most direct con­ tinuation. 6 'if d4 tl:ixf3+ l gxf3 ii.xf3 8 ii.xc4 eS ! is another line that typifies the attraction to 3 ... tl:ic6. 6 tl:ic3 tends to transpose to the main line after 6 ... tl:if6, but it also invites Black to challenge the dS-pawn. After 6 ... e6 White can escape the pin and create one of his own with l 'if a4+ tl:idl 8 tiles, while Kas­ parov-Deep Thought, New York 1989 went 6 ... c6 l �f 4 tl:ig6 8 �e3 cxdS 9 exdS tiles (9 ... a6 has been suggested) 10 'ifd4! tl:ixf3+ 1 1 gxf3 �xf3 12 ii.xc4 with the threat of �bS+. Now 12 ... tl:if6 and 12 ... a6 make sense, rather than Deep Thought's 12. .. 'if d6, which met with 13 tl:ibs 'iff6 14 'ii'cS 'ifb6 1S 'ii'a3 e6 16 tilcl+! etc.

6 . . . ll:\g6

The most sensible move, putting the ques­ tion to the bishop. 6 ... tl:id3+ l ii.xd3 cxd3 8 'ii'xd3 and 6 ... tl:ixf3+ l gxf3 �dl 8 ii.xc4 simply speed up White's development. After 6 ... �xf3 White should avoid l 'ifa4+? bS! 8 'ifxbS+ c6 9 'ii'a4 ii.xe4 with advantage to Black and instead follow the course of Z.Varga-Fochtler, Agria 1992, where White was happy to trade: l gxf3 tl:id3+ 8 �xd3 cxd3 9 'ii' xd3 c6 10 tl:ic3 tl:if6 1 1 0-0-0 cxdS 12 'ifbS+!? 'ildl 13 tl:ixdS! tl:ixdS (13 .. J�c8+ 14 tl:icl+ .l:txcl+ lS �xcl 'ifxbS? 16 .l:td8 mate, or 13 ... 'ifxbS? 14 tl:icl mate) 14 'ifxd7+ �xdl lS .l:txdS+ with a clear lead.

7 i.e3

Also seen is l �g3 , after which Black can push his e-pawn either one or two squares: a) l ... tl:if6 8 tl:ic3 e6 9 �xc4 exdS was the course of XuJun-N.Nikolic, Belgrade 1988. After 10 ii.xdS!? c6 (10 ... tilxdS 1 1 tilxdS retains White's pull) 1 1 'if a4! �xf3 12 gxf3 �el 13 ii.b3 0-0 14 eS tl:idl lS l:.dl 'if c8

White essayed 16 ii.xfl+.

The game continued 16 ... .l:txfl ll e6 .l:txf3 18 exdl 'if f8 19 'ifb3+ 'if fl 20 0-0 and Black had to worry about the dl-pawn, although this does seem preferable to the bizarre 16 ... �xfl ll e6+ �xe6 18 'if c4+ (18 'ifg4+ �fl 19 .l:txdl slightly favours White) 18 ... �f6 19 tl:ie4+ �fs 20 tild6+ ii.xd6 2 1 .l:txd6, e.g. 21...tilf6 2 2 h4.

b) l ... eS is a more natural approach, hop­ ing to leave the g3-bishop closed out of the game. 8 �xc4 and now:

b 1) 8 ... a6 defends the often useful (for White) bS-square but presents White with the opportunity to activate his other bishop with 9 d6!?, e.g. 9 ... cxd6 (to defend the bl­ pawn next move) 10 'ifb3 'if el (or 10 ... 'ii'dl 1 1 tl:igS llih6 12 f3 ..ths 13 tl:ic3 with good play for the pawn) 1 1 tilc3.

b2) 8 ... ii.d6 keeps the position closed. 9 ..tbs+ ii.dl (Costa suggests preserving the light-squared bishop with 9 ... �f8!?) 10

ii.xdl+ 'ifxdl 11 tilc3 tl:if6 12 0-0 was played

in Timman-Lautier, WijkaanZee 1994. Now the aggressive 12 ... hS can be comfortably handled with 13 tilh4, so the game continued 12 ... 0-0 13 .l:tcl bS 14 'if c2 .l:tfb8 1S tildl ! b4 16 tile3 'ifbS ll tilc4 with an advantage to White.

With l ii.e3 White is able to monitor the f4-square as well as the queenside, and the bishop is free to change location.

After 7 ... tt:'if6 8 tt:'ic3 e6 9 ..txc4 exdS 10 exdS ..lld6 11 ..llbS+ ..lld7 Black is doing fine. Instead 9 'ti'a4+ 'ii'd7 10 'ii'xd7+ offers White good chances:

a) 10 ... tt:'ixd7 takes Black's eye off e4 and dS. 1 1 tt:'id4 exdS 12 h3 cS 13 tt:'idbS d4 14 hxg4 0-0-0 has been evaluated as unclear but the simple 14 tt:'ic7 + looks good for White. Smyslov-Fontaine, Cannes 1996 went 1 1 ..txc4 exdS 1 2 tt:'ixdS ..lld6 1 3 tt:'id4 c6 14 tt:'ic3 tt:'ideS 15 ..llfl ..td7 16 h3 f6 17 0-0-0 i.cS and White was ready to get his kingside pawns rolling, leaving him considerably bet­ ter after 18 f4 tt:'if7 19 g3 hS 20 tt:'ia4 ..td6 21 fS tt:'igeS 22 tt:'ie6 <lite7 23 tt:'iacS ..llxe6 24 tt:'ixe6 b6 25 ..te2 l:.h7 26 a3 h4 27 g4 cS 28 gS.

b) 10 ... <litxd7 aims to connect the rooks quickly. Now 1 1 ..llxc4 exdS 12 0-0-0!? ..llxf3 13 gxf3 tt:'ieS 14 tt:'ixdS! tt:'ixc4 15 tt:'ib6+ <litc6 16 tt:'ixc4 gives White a clear advantage, but 12 ... <litc8 13 exdS ..lld6 limits White to an edge. Note that recapturing with the king leaves the f7-pawn unprotected, which is why Karpov-Lautier, Melody Amber (rapid) 1997 went 1 1 tt:'igS exdS 12 tt:'ixf7 l:.g8 13 f3 i.e6 14 tt:'igS c6 15 0-0-0 with pressure against dS and on the light squares in general.

8 ..ixc4

Van W ely deliberately holds back his knight rather than play the automatic 8 tt:'ic3 tt:'if6 9 ..txc4, when 9 ... a6 10 ..lle2 ..lld6 1 1 tt:'id2 offers an exchange of Black's more mobile bishop. Play might continue 1 1...txe2 12 'ii'xe2 0-0 13 0-0 'ti' d7 14 f3 l:.fc8 15 l:.fdl tt:'ie8 16 tt:'ib3 h6 17 l:.acl <lith7 18 l:.c2 l:.cb8 19 l:.dcl bS as in Azmai­

parashvili-Piket, Wijk aan Zee 1993. Then 20

tt:'iaS eyes the c6-square, securing White a slight advantage.

8 ... ..id6

Some players prefer to keep White out of bS with 8 ... a6 here. Since the kings have yet to castle White now has a couple of plans available.

a) In Alterman-N.Nikolic, Zagreb Z 1993,

Q u e e n 's G a m b i t A c c epted with 3 e 4

White decided the enemy minor pieces on the kingside made for attractive targets: 9 h3 ..lld7 10 h4!? ..lld6 11 hS tt:'i6e7 12 h6 gxh6 and it was time for the kings to head to the queenside after 13 tt:'ic3 tt:'ig6 14 'ti'd2 'ii'e7 15 0-0-0 0-0-0. It is true that the pawn sacrifice has damaged Black's structure, but Black's influence on the dark squares is significant. Consequently White turned his attention to the other wing: 16 <litbl tt:'if6 17 'ii'e2 (now 17 ..llxh6 is okay) 17 ... l:.dg8 18 tt:'id2 tt:'ig4 19 ..llxa6?! bxa6 20 'ii'xa6+ <litd8 21 'ii'a8+ ..llc8 22 tt:'ic4 tt:'ixe3 23 fxe3 'ii'd7 with an odd situation in that, for the moment, Black can do little with his extra piece. The game was eventually drawn after White was able to make inroads down the c-file, but it is inter­ esting that the combination of the closed centre (instigated by Black) and White's space advantage afford the first player such flexibility.

b) The more circumspect 9 0-0 will appeal

to most players. Then 9 ... tt:'if6 10 ..te2 ..td6 11 tt:'if d2 leads to similar play to that in the note to White's 8th move should Black trade bishops. Instead Karpov-V.Milov, Switzer­ land 1997, continued 11...td7 12 tt:'ic3 0-0

13 l:.cl 'ti' e7 14 a3 bS 15 tt:'ib3

Black's queenside expansion has left both cS and c6 potentially weak but there is the usual dark-square grip in compensation. At­ tempting to alleviate the pressure with 15 ... c6 looks sensible but is, in fact, what White is

hoping for in this type of position, since after 16 dxc6 .i.xc6 17 lbaS! Black learns that opening the position can leave him too vul­ nerable on the light squares, e.g. 17 ... .i.xe4 18 lbxe4 lbxe4 19 .i.f3 f 5 20 'W d5+ 'it>h8 2 1 -txe4 fxe4, and now 22 l:tfdl poses serious problems. No doubt aware of such an even­ tuality Black stuck with the strategy of con­ centrating on the dark squares, exploiting the closed centre to launch a kingside offensive with 15 ... lbf4 16 -tf3 'it>h8 (16 ... g5!?) 17 lba2 g5

This no-nonsense thrust is consistent with the general plan but creates another light square weakness on f5. White, meanwhile, has yet to make a concession on the kingside, permitting him to further remind his oppo­ nent of the holes on the other flank: 18 tLlc5 l:tg8 (18 ... g4 19 -te2) 19 lbb4 l:tg6 20 'Wc2 g4 2 1 .i.e2 l:tag8 22 l:tfdl (22 lbbxa6 .i.c8 23 lbb4 ltJ6h5 24 g3) 22 ... lb6h5 23 g3 .i.c8 24 lbc6! (there is no need to give Black what he wants with 24 gxf4? g3 25 hxg3 lbxg3 26 fxg3 l:txg3+) 24 ... 'Wg5 25 ..tfl l:th6 26 'Wc3! ltJf6 27 lbd3 ! (again 27 gxf4?! 'Wh4 28 -tg2 exf 4 29 -td4 f3 helps only Black) 27 ... 'Wh5 28 h4! gxh3 29 lbdxe5 l:tg7 30 .i.xf 4 lLixe4 3 1 'We3 �f5 32 -txh6 h2+ 33 'it>xh2 tLlxf2 34 .i.xg7+ 'it>xg7 35 l:td4 1-0. An excellent in­ structive display from Karpov, highlighting the power of the traditionally under-rated queenside attack. White's knights on c5 and c6 certainly make their presence felt!

9 ..ib5+

In Smyslov-Semkov, Rome 1990, the for­ mer world champion was in fighting mood. After 9 h3 .i.d7 10 h4 h6 1 1 h5 lbf8 12 lbh4 lbe7 13 g4 White was in danger of clamping down on both sides of the board. Conse­ quently Black hit out with 13 ... b5 14 .i.b3 c6

15 tLlc3 b4 but 16 dxc6 .i.xc6 17 tLld5 tLie6 18 l:tc 1 -txd5 19 exd5 tLif 8 20 -ta4+ tLld7 21 l:tc6! highlighted yet again how Black needs to be careful on the light squares. The text leads either to an exchange of light-squared bishops that is favourable to White or a misplacement of Black's king.

9 .. . �f8

By now we are aware of the problems Black can face on the light squares and, with the centre closed and a number of black pieces posted on the kingside, tucking the king away on f8 is not too inconvenient.

1 0 0-0

Sensibly opting for quick and easy devel­ opment. With Black's rooks still waiting to get acquainted the queenside will not be easy to defend. However, White successfully left his king in the centre in readiness for the ending in I.Nikolaidis-V alkesalmi, Moscow

0 lympiad 1994, at the same time addressing

the possibility of any future counterplay on the kingside: 10 .i.e2 lbf6 1 1 lbfd2 -td7 12 g3 h5 13 h4 tLlg4 14 .i.xg4!? .i.xg4 15 'Wb3 b6 16 tLlc3 a6 17 f3 -td7 18 tLlc4 �e7 19 tLlxd6 �xd6 20 a4! 'it>g8 21 'Wa3 l:tb8 22 'Wxd6 cxd6 23 'it>e2 f6 24 l:thcl 'it>f7 25 b4

l:thc8 26 b5 aS, and now 27 tLldl! followed

by sending the knight to c4 to attack both b6 and d6 kept Black busy, giving White time to switch to the kingside.

1 o ... eits 1 1 ttJbd2

I like this idea of keeping the knight on b 1 for a while. If White does not intend to offer an exchange of bishops with -te2 and tLlf3-d2, then posting the queen's knight on d2 supports the f3-knight while still protecting e4, thus freeing White's queen. If Black re­ treats his bishop after an inevitable h2-h3 the

knight already has access to c4 while, in the event of ... b7-b5, there is tt:'id2-b3-a5-c6 etc.

1 1 ... h5

A logical follow-up to the play thus far, the closed centre setting the scene for flank attacks. However, White need not worry about the coming offensive since he has enough pieces either on or near the kingside. The same cannot be said of Black's defensive resources on the other wing.

1 2 -.Wc2 a6 1 3 ..ltd3 h4 14 h3 ..ltxf3 1 5 lLixf3 lLih5 1 6 lUc1 ! tLlgf4 1 7 .llf1

This is why White sent his king's rook to the c-file, vacating f1 for the bishop in order to provide the g2-pawn with necessary extra protection. There is a temptation in this kind of situation to leave a rook near the king for defensive purposes, but here the f2-pawn is not a target.

1 1 ... :ha 1 s °iWb3 b6 1 9 :ca �gs 20

l:.ac 1 -.We7 21 a4 a5 22 �h 1 !

A prophylactic measure worth remember­ ing. An unwelcome major piece is sure to come to the g-file soon, when the h3-pawn will come under attack, so White takes time out to unpin his g-pawn in advance. Over on the queenside Black is powerless on the light squares and, as we will soon see, rather pre­ cariously placed on the dark squares.

22 ... :ga 23 tLld2 �h7 24 lLic4

With his kingside quite safe White has been able to conduct his queenside assault with little difficulty, culminating in

consider-Q u e en's G a m b i t A c c ep t e d with 3 e 4

able pressure against b6, c7 and d6. It is in­ teresting that White's rooks, queen, knight and dark-squared bishop combine to attack these weaknesses, while the remaining bishop plays an equally vital role in defending the kingside.

24 ... ..ltb4 25 d6!

Ironically it is the blockade of this pawn upon which Black's strategy is based in this line. Now the c7-b6-a5 pawn chain is about to be seriously undermined, and the opening of the a2-g8 diagonal for the white queen causes Black further trouble.

25 . . . cxd6

Black's queenside also falls apart after 25 ... .txd6 26 tL!xd6 cxd6 27 .txb6.

26 lLixb6 :ds 27 l:c7 -.We6 28 -.Wxe6?!

There is no doubt that White stands very well after the queen exchange, but 28 .tc4 looks much stronger, e.g. 28 ... dS 29 tLlxd5 tLlxg2 30 tL!f4. Nevertheless with an attack no longer a realistic prospect Black's pieces -particularly his rook - look out of place on the kingside.

28 •.. fxe6 29 :7c4

29 tLld7 followed by �a7 and .tb6 is not easy to meet, but White has another plan in mind.

29 . . . l:f6

The rook rushes to re-enter the game.

30 :xb4!?

Despite this exchange sacrifice working out well for White it would have been

sim-pler to try and exploit the existing advantage.

30 ••. axb4 31 a5

Each one of White's pieces points to the queenside, while three of Black's - two of them notoriously slow knights - are far away.

31 . . . d5 32 a6 :t1 33 J:l.a1 J:l.a7 34 lLixd5 J:l.xa6 35 :xa6 exd5 36 exd5 .l:xd5

36 ... lbxd5 37 i.g5.

37 .Jlc4 J:[d1+ 38 �h2

The smoke has cleared and White's bish­ ops are enough to give him a potentially de­ cisive lead.

38 . . . lbt6 39 l:.a5 lLig6 40 .l:b5 e4 41 :xb4 lLie5 42 .Jlb3 J:l.d8 43 .Jlg5?!

White should be winning without looking for a second pawn. A more incisive way to use the bishops and passed pawn is 43 i..d4 l:te8 44 l:tb7 �g6 45 i..a4 l:te6 (45 ... l:ta8 46 i..c2) 46 b4 etc.

43 . . . �g6 44 .Jlxh4 .l:d2 45 J:l.b5 lLid3 46 .Jle6 lLixt2 47 .Jlt5+ '.t>h6 48 .l:b8 g6 49 .l:b6?!

49 i..xg6 parts with the other bishop and looks preferable, e.g. 49 ... lb6g4+ 50 hxg4 �xg6 5 1 l:tb6+ �g7 52 �g3 e3 53 Wf3 lbd1 54 b4 and White has passed pawns on both flanks.

49 . . . '.t>h5 50 J:l.xt6 '.t>xh4 51 J:l.xg6 e3 52 .Jlg4?!

It is time for 52 b4 since the e-pawn is go­ ing no further after 52 ... e2 53 l:te6.

52 . . . lLid3 53 .l:h6+ �g5 54 .l:e6 �t4 55 J:[t6+ �e5 56 J:l.t8 lLie 1 57 J:l.e8+ �t4 58

:ta+ �e5 5 9 :ts+ �e4 6 0 :n J:xg2+ 61 �h 1 J:l.xg4! 62 hxg4 lLit3 63 '.t>g2 e2 64 .l:h1 e 1 � 65 J:be1+ lLixe 1 + 66 �g3 Yz-Yz

Game 2

Continental Open 2000

This time Black hits the e4-pawn, practi­ cally forcing its advance as 4 lbc3 e5 prom­ ises White little.

4 e5 lLid5 5 .Jlxc4 lLib6

The most obvious and the most popular, putting the question to the bishop. It is also possible to insert 5 ... lbc6 6 lbc3 and now:

a) 6 ... i..e6 has the nasty threat of 7 ... lbxc3 8 bxc3 i..xc4. Hauchard-Collas, Cappelle la Grande 1997 continued 7 i..b3 lbxc3 8 bxc3 i..d5 when, instead of the natural 9 lbf3 White tried the provocative 9 lbe2 with a view to an early eviction of Black's bishop. After 9 ... i..xg2 10 l:tgl i..d5 1 1 c4 i..e6 12 dS lbxe5 13 dxe6 lbf3+ 14 Wfl 'ii'xdl+ 15 i..xdl lbxgl 16 i..a4+ c6 17 Wxgl fxe6 White had two good pieces for a rook but was running short of pawns. Obviously such lines require careful handling from both sides, although I'm not convinced 6 ... i..e6 is worth the trouble in any case.

b) 6 ... lbb6 7 i..b5 i..d7 8 lbf3 with a fur­ ther branch:

bl) 8 ... a6 9 .id3 .ig4 10 .ie4 e6 and Black had developed his light-squared bishop outside his pawn centre in Glek-Kozlov, Frunze 1988. However, White's bishop stands very well on e4, monitoring squares that its counterpart can no longer defend. After 1 1 0-0 .ie7 12 .ie3 .ih5 White's con­ trol of the centre gave him an edge, which remained in place after 13 d5 ll'ixe5 14 dxe6 'iixdl 15 l:tfxdl ll'ixf3+ 16 gxf3 fxe6 17 i.xb7 l:tb8 18 .ic6+ �f7 19 b3 thanks to his more active pieces and superior pawn struc­ ture.

b2) 8 ... e6 9 0-0 lt:Jb4 (9 ... lt:Je7 10 .id3 .ic6 1 1 lt:Jg5 h6 12 'ifh5 g6 13 ll'ige4! .ig7 14 'ii'g4 gave White something to aim at in Beliavsky-Portisch, Thessaloniki 1984) 10 .ie2 .ic6 11 a3 lt:J4d5 12 lt:Je4

There is no denying Black has a firm grip on the d5-square! How useful this will prove is another matter, particularly if White is able to lessen the influence of a central knight by playing 'around' it and exploiting his extra territory. Korneev-Arakelian, Kstovo 1994 continued 12 ... "iid7 13 lt:Jc5 .ixc5 14 dxc5 ll'ia4 15 "iic2 0-0-0 16 .ie3!? (White wants to prove that the knight on the rim is indeed 'dim) 16 ... lt:Jxe3 17 fxe3 "iid5 18 l:tacl 'ife4 19 'ifxe4 .ixe4 20 b3 ll'ib2 21 ll'id4 with a clear advantage to White. More testing is 12 ... lt:Jf4 13 .ixf4 .ixe4, removing White's annoying knight. After 14 ll'ig5 .ic6!? 15 .th5! g6 16 .tf3 .txf3 17 'ifxf3 .ie7 18

Q u e e n 's G a m b i t A c c ep t e d with 3 e4

lt:Je4, or 16 ... ll'id5 17 .td2 .ie7 18 ll'ie4, the holes in Black's kingside will be a constant worry. This leaves the alternative retreat 14 ... .ig6 15 .if3, when 15 ... lt:Jd5 16 .td2 .ie7 17 lt:Je4 0-0 18 'ifb3 l:tb8 19 l:tacl is a shade better for White according to Huzman, while Korchnoi-Sadler, Arnhem 1999, went 15 ... l:tb8 16 'if cl .ie7 (16 ... c6!? limits White to the usual space advantage) 17 .ixb7!. Now 17 ... l:txb7 18 'ifc6+ ll'id7 19 "iixb7 .ixg5 20 .ixg5 'ifxg5 21 "iixa7 .ie4 22 f3 is poor for Black after either 22 ... .idS 23 'if xc7 or 22 ... .id3 23 l:tfcl 'ife3+ 24 �hl 0-0 25 l:txc7 etc. Instead the game went 17 ... 0-0 18 .ic6 'ifxd4 19 ll'if3 and White stood better.

6 i..d3

I prefer this to the alternative 6 .ib3 be­ cause without the f5-square available Black must still solve the problem of where to de­ velop his light-squared bishop. Nonetheless dropping back to b3 is a viable option. After 6 ... ll'ic6 play might develop as follows: 7 .ie3 i.f5 8 ll'ic3 e6 (8 ... lt:Jb4 9 'if f3! lt:Jd3+ 10 �fl rebounds on Black according to Atalik) 9 ll'ige2 and now:

a) 9 ... .ie7 10 0-0 "iid7 1 1 ll'ig3 .ig6 12 f4

Gormally-Law, British League 1997. Launching the f-pawn looks like an effective plan in this line, reminding Black who has more space and endeavouring to open the position for the light-squared bishop with d4-d5 or f4-f5. Consequently Black chose to

14 'ifxb3! i.d3 (taking on dS invites f4-f5, trapping the bishop) 15 .l:.fdl i.c4 16 dxe6 "iixe6 17 "iic2 0-0 18 'ii'e4! and White is do­ ing well.

b) Black immediately went for the bishop in Giorgadze-Narciso Dublan, Linares 1999. After 9 ... tLlaS 10 i.c2 i.xc2 1 1 "iixc2 i.e7 12 0-0 0-0 13 .l:.adl .l:.e8 White again pushed his f-pawn: 14 f4 ltJdS 15 tLlxdS 'ifxd5 16 tLlc3 'iic6 17 fS tLlc4 18 i.cl exfS 19 'iixfS "iie6 20 'iif3 c6 21 tLle4 with a slight pull.

If these lines are to your liking it is worth taking a closer look at 6 i.b3, but make sure Black's control of the d5-square does not become a long-term plus. To be avoided is 6 e6? tLlxc4 '7 "iia4+ tLlc6 8 exf7+ �xf7 9 'iixc4+ i.e6 etc.

Returning to the popular 6 i.d3, Black's task of achieving smooth development is more problematic.

6 . . . lllc6

Don't expect too many opponents to oblige with 6 ... 'ifxd4?? 7 i.bS+ and the queen is lost.

7 ..te3

It is a little unusual to see bishops finding posts before knights, but here the idea is to deny the c8-bishop an outing. The helpful 7 lLle2, for example, permits 7 ... i.g4, whereas now both fS and g4 remain unavailable, hence the suggestion in the note to Black's next move.

7 . . . lllb4

This natural move is seen the most often. Black aims to punish White's uncompromis­ ing bishop. Two other strategies have also

been tried: ·

a) 7 ... i.e6 rather stubbornly addresses the problem of the queen's bishop. After 8 tLlc3 Black has played a number of moves:

al) 8 ... 'if d7 9 tLlf3 i.g4 10 h3 i.xf3 1 1 "iixf3 e6 12 .l:.dl 0-0-0 13 0-0 tLlb4 1 4 i.e4 was clearly better for White in Korchnoi­ Hiibner, Tilburg 1987. Later, in Karpov­ lvanchuk, Reggio Emilia 1991, Black im­

proved with 9 ... 0-0-0 10 h3 tLlb4 1 1 i.e2 fS

12 0-0 h6 13 a3 ltJ4d5, and now 14 lLlel (heading for c5) gave White an edge.

a2) 8 ... tLlb4 9 i.e4 tLl4d5 10 tLlge2 fS 1 1 exf6 exf6 1 2 tLlf4 w as seen in Shaked­ Ibragimov, Berlin 1997. White has an iso­ lated d-pawn that does seem well blockaded but 10 .. .fS has led to a positional concession involving a weakening of the light squares in Black's camp. After 12 ... tLlxf4 13 i.xf4 Black's attempt to reduce his opponent's piece activity with 13 ... i.d6 14 i.xd6 'ifxd6 15 i.xb7 .l:.d8 met with 16 'ifhS+ i.f7 17 'ii'b5+ tLld7 18 "iie2+ and White was a safe pawn up.

a3) 8 ... tLlc4 9 i.e4 i.c8 is the odd course chosen by Black in Furman-Noskov, Mos­ cow 1991. White is happy to allow the cap­ ture on e3 in these positions because fxe3 both supports the d4-pawn and opens the f­ file, so Black has achieved less than nothing. In fact White struck immediately with 10 e6, which looks like a nice alternative to the comfortable advantage that results from sim­ ple development (e.g. 10 tLlge2).

a4) 8 ... i.dS is a consistent follow-up to 7 ... i.e6. Then 9 tLlf3 e6 10 0-0 slightly fa­ voured White in Salov-Hiibner, European

T earn Ch. 1989. Again White can 'ignore' the

central bishop until such time that an ex­ change on d5 can be carried out in favour­ able circumstances. With Black so involved with the d5-square he has less influence in other areas and less room in which to ma­ noeuvre.

b) With 7 ... g6 Black waits to determine the future of his queen's bishop. However, I don't believe the fianchetto is appropriate here. After 8 tLlc3 i.g7 (8 ... tLlb4 9 i.e4 i.fS 10 i.xfS gxfS is an interesting way to use 7 ... g6, although I doubt the damage to Black's kingside structure is worth the dS­ square) 9 lLlge2 0-0 White has two avenues to explore.

bl) 10 h4 is a blunt yet effective means with which to demonstrate that ... g7-g6 is simply too risky. Faced with h4-h5 Black

must seek counterplay in the centre: 10 ... tt:Jb4

1 1 .te4 when Portisch-Spraggett, Wijk aan

Zee 1985 continued 1 1...tZ::i4d5? 12 hS .i.e6 (12 ... tt:Jxe3 13 fxe3 serves only to strengthen White's centre, while 12 ... c5 13 hxg6 hxg6 14 .th6! .txh6 15 l:!.xh6 'it>g7 16 nh2 maintains White's initiative) 13 'i'c1 tZ::ixe3 (13 ... c5 14 hxg6 fxg6 15 .th6 is dangerous) 14 'i'xe3 (14 fxe3!?) 14 ... c5 (14 ... tt:Jd5 15 .txd5 .txd5 16 hxg6 hxg6 17 tt:Jf4 e6 is uncomfortable for Black but at least avoids the following) 15 d5!

Now 15 ... .txdS runs into 16 hxg6 hxg6 17 'ii'h3 ne8 18 tZ::ixd5 tZ::ixd5 19 e6!, e.g.

19 ... tt:Jf6 20 exf7+ �xf7 21 .i.xg6+! �xg6 22 tt:Jf4+, or the lesser evil 19 ... 'i'd6 20 l:!.dl fxe6 21 .txg6. Instead the game went 15 ... tt:Jxd5 16 .i.xd5 .i.xd5 17 0-0-0 e6 18 tt:Jf4 l:!.e8 (18 ... 'i'c7 19 hxg6 hxg6 20 tt:Jcxd5 exd5 2 1 tZ::ixd5 and White dominates) and now 1 9 h6! .i.h8 20 tt:JfxdS exd5 21 l:!.xd5 'i'e7 22 f4 would have given White a clear lead, e.g. 22 .. .f6 23 f5! etc.

Returning to Black's 1 1th move, 1 1...c5 has been suggested as an improvement. After 12 d5 .tf5 Black seems to be doing fine, so best is 12 dxc5. Then 12 ... 'i'xdl+ 13 tZ::ixdl tt:Jc4 14 f4 .tf5 15 .txf5 gxfs 16 �f2 tt:Jd3+ 17 �f3 favours White, so this leaves 12 ... tt:Jc4 13 'ii'b3 tZ::ixe3 14 'i'xb4 a5 when both 15 'i'd4 tt:Jf5 16 'i'xd8 l:!.xd8 17 tt:Jd5 (e.g. 17 ... .txe5 18 g4) and 15 'i'a4 tt:Jg4 16 .i.xb7 .i.xb7 17 'i'xg4 .i.xe5 look promising

Queen's Gambit Accepted with 3 e4

for White.

b2) 10 .te4 expresses an interest in the

centre rather than a kingside attack. In Tim­

man-Korchnoi, European Team Ch. 1997, 10 ... tt:Jb4 1 1 'i'b3 a5 12 tt:Jf4! placed a fourth white piece within range of the crucial d5-square and consequently left Black struggling. After 12 ... e6 13 a3 tt:J4d5 (13 ... tt:Jc6 14 .l:!.dl is very pleasant for White) 14 tt:Jfxd5 exdS 15 .txdS (15 tt:Jxd5?? .i.e6) 15 ... a4 16 'i'a2 l:!.a5 (16 ... tt:Jxd5 17 'i'xd5) 17 .i.c4 tZ::ixc4 18 'i'xc4 Black did not have enough for a pawn.

8 ..li..e4 f5

This is the thrust upon which Black's opening strategy is based in this particular line. Black will not allow the bishop to re­ main in such a commanding position in the middle of the board and, in order to fight for this outpost, White must give up his ad­ vanced e5-pawn. 8 ... tt:J4d5 9 tZ::ic3 c6 10 'i'f3!? provides White with a menacing set up, while after 8 ... c6 the game Sadler-Shaw, Isle of Man 1994, continued 9 tZ::id2!? .i.e6 10 tt:Jgf3 'ii'd7 1 1 tt:Jb3 .i.xb3 12 'i'xb3 e6 13 0-0 .i.e7 14 l:!.fdl tt:J4d5 15 .td2 a5 16 a3 and Black was missing his bishop.

9 exf6

Of course White refuses to give way, and taking the pawn retains the advantage of the move.

9 ... exf6 1 0 a3

The beginning of a complex and practi­ cally forced sequence that Miton treats with

some skill. The chief alternative is 10 lDc3 fS when the second pawn to land on fS puts the question to the bishop.

a) Karpov-Short, Linares match (3) 1992,

went 11 i.f3 (keeping an eye on d5 and tying

the bishop down to the b7-pawn) 1 l...iD4dS 12 i.d2 i.e6 13 lDge2 ii'd7 14 0-0 0-0-0 lS .l:!.el with an edge for White. Surprisingly Black's blockade on dS is not as secure as it first appears, as the logical 1S ... .l:!.g8 16 Si.gs .l:!.e8 17 lDf4 lDxf4 18 i.xf4 gS 19 Si.es i.g7 meets with Curt Hansen's 20 i.xg7 'ii'xg7 21 dS and White stands better. The solid 1 l...c6 merits attention. In Zaja-Ganguly, Istanbul Olympiad 2000 White managed to remove his opponent's light-squared bishop after 12 iDh3 i.e6 13 0-0 i.d6 14 .l:!.el 0-0 lS lLJgS i.c4 16 b3 i.f7 17 lDxf7 .l:!.xf7, but then 18 a3 lD4dS 19 lDxdS lDxdS 20 i.xdS cxdS 21 'ii'f3 'ii'h4 was fine for Black.

b) Dropping back to b 1 is more popular. In fact 1 1 i.b 1 only temporarily ignores the dS-square, as a later a3 will offer the a2-square as an active long-range post for the bishop. After 1 1...lD4dS 12 lDf3 it is Black's turn to decide where is best for his king's bishop.

b l) Timman-Salov, FIDE Candidates match, Sanghi Nagar 1994, continued 12 ... i.d6 13 Si.gs 'ii'd7 14 'ii'e2+ 'ii'e6 lS lDeS, highlighting one of the potential prob­ lems for Black caused by an early push of the f-pawn(s) - the weakness of the eS-square. After 15 ... 0-0 1 6 0-0 lDxc3 17 bxc3 i.xeS 18 dxeS 'ii' c6 19 i.d3 i.e6 Black could do with a bit of help on the dark squares but must only be a shade worse.

b2) 12 ... i.b4 13 i.d2 Ji.e7 is designed to reduce White's support of the d4-pawn and is better than 13 ... 0-0 14 iDxdS i.xd2+ 15 'ii'xd2 lDxdS 16 0-0 when the hole on eS is significant. The we have 14 0-0 0-0 lS .l::.el i.f6 16 a3! �h8 17 i.a2 f4 18 lDeS Hiibner­ Sulskis, European Team Ch, Pula 1997. Black's occupation of dS is looking less se­ cure and the eS-square is a genuine problem.

Consequently he now sought t o generate complications with a further advance: 18 .. .f3, the point being that 19 iDxf3 i.g4 20 lDe4 ii.xf3 2 1 'ii'xf3 i.xd4 22 'ii'd3 lDf6 is about even. Instead the game went 19 lDxdS iDxdS 20 lDf7 + .l::.xf7 2 1 i.xdS .l::.f8 22 i.xf3 i.xd4 (22 ... 'ii'xd4 23 i.b4 cS 24 'ii'xd4 i.xd4 25 i.c3 and e7 beckons) 23 i.c3! i.xc3 24 'iixd8 .l::.xd8 25 bxc3 .l:!.b8 26 .l::.e7 and White's lead was still intact.

Incidentally the immediate 12 ... i.e7 was agreed drawn in Hebden-Drasko, Vrnjacka Banja 1991. Of course there is still everything to play for. Note that with the bishop still on e3 it is tempting to hit it with .. .f5-f4 at some point, but then White's other bishop comes to life on the b l-h7 diagonal.

1 0 ... fS

Another one! In fact Black's uncompro­ mising response is called for as 10 ... lLJ4d5?! 1 1 'ii'hS+ g6 12 i.xg6+ hxg6 13 ._.xh8 nets White too much material for insufficient compensation, while 1 l...'iti>e7 is hardly an attractive move.

11 axb4

11 . . . fxe4

Huzman gives this move a '!' and it does look better than the tempting 1 1...i.xb4+ 12 iDc3 fxe4 13 'ii'hS+ when Black can spend so much time checking out 13 ... g6 14 'ii'eS+that he overlooks 14 'iibS+ picking up the bishop. After 13 ... 'iii>f8 l 4 lDge2 Black tends to bring his queen to e8 soon.

a) 14 ... i.e6 lS 0-0 and now:

al) 1S ... 'iti>g8 16 tt:\xe4 'ii'e8 17 'ii'xe8+! l:!.xe8 18 l:!.fcl (18 l:!.xa7 i.c4 19 tt:\4g3 i.a6 looks risky but might be good for White) 18 ... aS 19 l:!.xc7 tt:\dS 20 l:!.ccl ! leaves White with an extra pawn, e.g. 20 ... i.fS 21 tt:\4g3 i.g4 22 h3 i.xe2 23 tt:\xe2 tt:\xe3 24 fxe3 i.d2 (24 ... l:!.xe3? 2S l:!.c8+ .if8 26 l:!.fl) 2S l:!.c7 i.xe3+ 26 'iti>fl b6 27 l:tb7. Dreev­ Svidler, Elista 1997, went instead 18 ... i.c4 19 tt:\2g3 i.dS (19 ... aS 20 tt:\cS i.dS 21 i.f 4 and Black's queenside is still under pressure) 20 i.d2! i.xd2 (20 ... aS 21 i.xb4 axb4 22 tt:\cS limits White's advantage) 2 1 tt:\xd2 iLc6 22 l:txa7 'iti>f7 23 tt:\f3! tt:\d7 24 dS! i.xdS 2S l:!.xc7 with excellent winning chances for White.

a2) 1S ... 'i'e8 intends to recapture on e8

with the king, so this time White does best to

decline the offer with 16 'ii'h4, as played in A.Ziegler-Brynnel, Sweden 1998. Then 16 ... i.c4 17 f3?! i.xc3 18 fxe4+ 'iti>g8 19 lLixc3 i.xfl 20 l:txfl 'iid7 21 dS h6 saw Black wriggle out, so rather than sacrifice the exchange White should have played 17 l:f.fcl i.d3 18 tt:\g3, when Black's problem king is compensation for the pawn.

b) 14 ... We8 lS 'ii'h4 'iti>g8! improves on lS ... iLfS 16 0-0 i.xc3? 17 tt:\xc3 tt:\c4 18 'i'f4 tt:\d6 19 l:taS, Tregubov-Stajcic, Harkany 1992. After 16 0-0 i.fS White can try 17 dS i.g6 18 tt:\f4 i.d6 19 i.xb6 iLxf4 20 Wxf4 cxb6 21 l:tfel 'if f7 22 'ii'd2 h6 23 tt:\xe4, which offered decent prospects of an advan­ tage in Bezgodov-Nikitin, Tomsk 1998, or 17 tt:\g3 i.g6 18 tt:\gxe4 h6 19 dS i.xe4 20 tt:\xe4 tt:\xdS 2 1 i.cS iLxcS 22 tt:\xcS b6 23 'ii'd4 which has been assessed as slightly bet­ ter for White despite the pawn deficit. Tregubov-Nikitin, St Petersburg 199S illus­ trates what White is looking for: 23 ... 'i'fl 24 tt:\d3 cS 2S We4 l:te8 26 tt:\eS We6 27 l:tael! hS 28 'ii'b l ! and Black was struggling. Per­ haps 23 ... tt:\f6 is better, although White's compensation is obvious.

12 ltlc3!

Queen's Gambit Accepted with 3 e4

Here White should refrain from 12 'ii'hS+ g6 13 'i'eS+ 'iti>f7, first because 14 'ilfxh8?? loses to 14 ... iLb4+, but really because 14 tt:\c3 i.xb4 lS 'ii'xe4 l:!.e8 16 'i'f3+ 'iti>g8 17 tt:\ge2 'ifdS and lS i.gS i.d6! 16 i.xd8 iLxeS 17 dxeS l:txd8 18 tt:\xe4 i.fS are both preferable for Black.

12 . . . i.e6

12 ... i.xb4 13 'i'hS+ 'iti>f8 14 tt:\ge2 leads us

back to the note to Black's eleventh move.

13 ltlh3 i.d6 14 ltlgS i.c4

Part of the grand plan.

15ilg4

lS tt:\gxe4, on the other hand, inconven­ iences White more than Black, e.g. lS ... 0-0 16 b3 .ids 17 bS i.xe4! 18 tt:\xe4 i.b4+ 19 tt:\d2 'ii'dS 20 0-0 i.xd2 2 1 'ii'xd2 'ii'xbS and this time the knight will be fantastic on dS, Ibragimov-Y akovich, St. Petersburg 1998.

15 . . . i.xb4 16 ltlxh7!

The point. This is the position both play­ ers have been aiming for since 8 .. .fS. Neither king sits comfortably in the centre, but the e4-pawn is a sitting duck, and White hopes to emerge from the complications with a mate­ rial lead that will provide realistic winning chances as the game moves into the ending.

16 . . . ild7 17 'ilfg6+ �dS 1 8 ltlg5 :es 19 ltlgxe4 ltld5 20 c,£.ld2!?

This remarkable move escapes the pins on the e-file yet calmly retains the pin on the aS­ e 1 diagonal, not forgetting walking into po­ tential hazards on the cl-file! However, young

Miton's contribution to opening theory is in fact quite logical, as the king is heading for the relatively safe haven on cl. Black's king is no better, and Black is a pawn down. Before 20 Ji.d2 had been played, when 20 ... Ji.xc3 21 bxc3 'if c6! highlighted the problem on the e­ file and saw White happy to make the draw in Ehlvest-Volzhin, Koszalin 1998.

20 . . . c5

With White's centre in danger of becom­ ing sufficiently solid Black strikes before White has time to bring his king's rook to the middle of the board.

21 Wc 1

As per plan. 2 1 lLid6? cxd4 is going a little too far and backfires on White.

2 1 . . . lllxe3 22 fxe3 cxd4 23 exd4

23 . . . i.d3

Introducing another pin. 23 ... 'ii'xd4 is critical - should Black be afraid of 24 l::tdl Ji.d3 or will he hang on to see his bishops devour the knights? Well, after 25 'if g3 Ji.xc3 26 lLixc3! l::te3 White has 27 l::ta4.

24 'i'g5+ i.e7

24 ... 'i&tc8 25 lLic5 does not help Black.

25 'i'a5+?

It is difficult to criticise White since his whole strategy has been wonderful thus far, but the simple 25 'ii'xg7 might well leave Black with nothing to show for a two pawn deficit. In such a complex position, with un­ comfortable kings, awkward pins and the presence of queens, White can be forgiven

for 'playing safe'.

25 . . . �cS

25 ... b6 26 'ii'xb6+! axb6 27 l::txa8+ 'i&tc7 28 l::ta7+ picks up a second pawn.

26 :d 1 b6

Again the cl-pawn is safe: 26 ... 'i'xd4? 27 'ii'f5+ 'i&tb8 28 'ii'f4+ 'i&tc8 29 'if g4+ 'i&tb8 30 'ii'g3+.

27 'i'a4

No doubt forcing the exchange of queens is what White had in mind when turning down the g7-pawn earlier and, under the circumstances, this is understandable. How­ ever, Black now proceeds to defend the end­ ing very well indeed, a situation that could have been avoided with 25 'ii'xg7.

27 . . . i.xe4 28 'i'xd7+ Wxd7 29 lllxe4 i.b4 30 lllc3 a5 3 1 '&td2

White has an extra pawn and decent pros­ pects of converting it, but with the clock ticking and a strong opponent the task can be rather difficult. The game continued:

31 . . . i.d6 32 g3

32 h3 i.f4+ 33 'i&td3 l::te3+ 34 'i&tc4 is noth­ ing for White to be afraid of.

32 . . JlhS 33 l:.h 1 %ilh3 34 �af1 ?!

34 lLid5 'i&tc6 3 5 lLie3.

34 . . . g5 35 llle4 i.e7 36 �d3 a4 37 :a1 b5 38 lllf2 :ha 39 lllg4 :ha6! 40 :hc1 b4 41 llle5+ Wes 42 %ilc6

42 l::tc7 is more ambitious, e.g. 42 ... a3 43

l::tel! axb2 44 l::tbl l::ta3+ 45 'i&tc2, or 43 ... a2

.l:!.xal 47 .l:!.b7) 46 ll\g5 al"ii 47 ll\h7+ �g8 48 .l:!.xal .l:!.xal 49 ll\f6+ �f8 50 .l:!.d7.

42 . . . a3 43 :x a6 :x a6 44 bx a3 bx a3 45 llJ c4 'it>d7 46 'it>c3 Y. - Y.

Black's long-range bishop carries out a game-saving dual-purpose role.

Game]

A talik-Gyimesi

Yugoslav Team Ch. 1998

1 d4 d5 2 c4 dx c4 3 e4 c5

While it is quite logical to exploit Black's early queenside pawn majority immediately, 3 ... c5 invites White to continue with his terri­ torial claim in the centre.

4 d5

4 ll\f3 cxd4 5 "iixd4 is a risk-free varia­ tion, but we are concerned with the less compromising push of the d-pawn. Quite simply White hopes to regain his sacrificed pawn with an advanced centre, leaving his opponent cramped.

4 . . . e6

Black sets about dealing with the impor­ tant d5-square head on. The next main game features the wild 4 ... ll\f6 5 ll\c3 b5.

5 lDc3

5 .ixc4 tends to transpose after 5 ... ltJf6 6 ll\c3 exd5 7 ll\xd5 ll\xd5 as 7 ... ll\xe4? walks into 8 "iie2. Alternatively 5 ... exdS?! 6 .ixdS is already difficult for Black, since 6 ... ll\f6? permits 7 .ixf7+!, and blocking the d-file with 6 ... .id6 runs into 7 e5!, e.g. 7 ... .ixe5? 8 .ixf7+ etc.

5 . . . ex d5 6 llJx d5

Maintaining a piece on d5 is a key part of our strategy here, for the square ceases to be a weakness for Black once White captures with his pawn (unless this gives him a power­ ful passed pawn, of course).

6 . . . llJe7

Black does best to avoid 6 ... ll\f6 7 .ixc4 i.e6 8 'ii'b3.

7 .i.x c4 llJx d5 8 .i. x d5

Q u e en 's Gambit A c c ep t e d with 3 e 4

8 . .. .i.e7

By far the most popular response to White's threatened 9 .ixf7+. Petrosian­ Radulov, Plovdiv 1983 went 8 ... .id6 9 "iih5 'iff6 10 ll\f3 'ii'g6 and the exchange of queens did nothing to diminish White's pres­ ence in the centre. After 8 ... ll\c6 9 .ixc6+ bxc6 Black's poor queenside pawns are worth parting with the bishop.

9 lDf3

It is a matter of choice whether White se­ lects f3 or e2 for his knight. The difference between Atalik's choice and 9 ll\e2 is that the latter seeks to maintain maximum control over the d5-square. Using e2 also rules out a pin after ... .ig4 since now White has f2-f3. Let us look at a few examples of how play can go after 9 ll\e2.

a) 9 ... ll\d7 10 0-0 ll\b6 1 1 ll\c3 0-0 12 .tf 4

This is typical. White is ready to meet the challenge to his bishop with the supporting knight. Black has tried a number of moves in the diagram position.

al) 12 ... i..g5 harasses the other bishop but presents white with the cheeky response 13 i..d6, the point being that the bishop is safe in view of 13 ... 'ii'xd6? 14 i..xf7+ etc. How­ ever, Black can persevere with 13 ... i..e7 when White has 14 e5! with the prospect of having at least one rather large bishop right in the heart of Black's position. After 14 ... i..xd6 15 exd6 the new passed pawn can prove equally unwelcome, the bishop over­ sees both halves of the board and 15 ... l2Jxd5 16 'ii'xd5 merely adds to Black's problems.

a2) 12 ... 'it>h8 rules out any funny business with i..xf7+ and frees the f-pawn in case Black elects to challenge the remaining centre pawn. I like 13 Wf3 here, intending to meet 13 ... l2Jxd5 with 14 .U.adl to stay in charge of the d5-square. Comas Fabrego-Cifuentes Parada, Platja d'Aro Barcino 1994, went 13 'it'h5 l2Jxd5 14 l2Jxd5 i..d6 15 e5 i..b8 16 .U.adl and Black was being pushed back. Af­ ter 16 ... 'i'e8 17 'i'h4 'iie6 18 .l:l:fel .l:l:e8 19 h3 i.d7 White struck with 20 lLif6!? gxf6 21 .U.xd7 i..xe5 (2 1...'iixd7 22 'ifxf6+ 'it>g8 23 i..h6) 22 i..xe5 fxe5 23 .U.edl, but after 23 ... Wf5 24 .l:l:xb7 .l:l:g8 walked into 25 .U.c7?? 'iff3!. This is a pity because with by far the better pawn structure White is clearly better, e.g. 25 .U.dd7 .U.g7 26 .U.xa7 .U.ag8 27 g3.

a3) 12 ... i..f6 13 i.d6 i..e7 14 e5 trans­ poses to 'al', while White also has 13 eS!?. Then 13 ... l2Jxd5 14 lLixdS i..g5 (14 ... i..e7 15 e6!) 15 i.xg5 'ii'xgS 16 f4! is precisely what White is looking for. In Khalifman-Vulfson, St Petersburg 1995, Black retreated: 13 ... i..e7 14 i..e4! Wxdl 15 .l:l:fxdl and White's author­ ity in the centre was still present after the exchange of queens. After 15 ... g5 (undermin­ ing the support of the eS-pawn) 16 i..e3 .l:l:b8 (16 ... l2Jc4? 17 lZJdS) 17 b3 i..e6 18 lLibS Black's queenside was starting to creek.

a4) 12 ... lLixdS 13 lLixdS i..d6 14 eS i..c7

15 .U.cl b6 16 Wf3 i..e6 17 .U.fdl i..xd5 was agreed drawn in Rausis-Lev, Eupen 1997. White went for more in Kacheishvili­ Spangenberg, World U26 Team Ch., Roque Saenz Pena 1997, finding himself on the way to a modest material lead after 15 e6!? i..xf4 16 e7 i..xh2+ 17 'it>h 1 'iid6 18 exf8'i' + 'it>xf8 19 g3. There followed 19 ... Wh6 20 'ii'cl! 'i'c6 2 1 WgS i..e6 22 .U.adl h6 23 Wes .l:l:d8 24 'it>xh2 .U.xd5 25 .U.xd5 'ifxd5 26 'i'b8+ 'it>e7 27 'i'xa7 'it>f6 28 Wb6 �g6, and now 29 .U.cl would have caused Black the most problems.

b) 9 ... 0-0 and now:

b l) 10 0-0 l2Jd7 1 1 lLic3 i..f6 12 i..e3 i..xc3 13 bxc3 'ifc7 14 c4 l2Jf6

Zontakh-Mirkovic, Novak Nikolic Mem. 1998. White's bishop is too strong to be al­ lowed to survive in this kind of position, but after 15 'ii'd2 lLixdS 16 cxd5 i..d7 17 i.f4 'i'd8 18 .U.fel .U.e8 19 .U.acl .U.c8 20 .U.c3 White was ready to swing his rook over to the kingside, leaving Black with the standard dilemma of being unable to contest the dark squares.

b2) 10 i..e3 'ii'b6 1 1 0-0 l2Jc6 12 'ii'd2 i..e6 13 lLif4 lLie5 14 b4!? 'if xb4 (14 ... l2Jc4 15 i..xc4 i..xc4 16 .U.fcl) 15 'if xb4 cxb4 16 lLixe6 fxe6 17 i..xe6+ 'it>h8 18 f4 lLid3 19 .U.adl .l:l:ad8 20 i..dS and the bishop pair fa­ voured White in Granda Zuniga-Gendler, Yerevan Olympiad 1996.

and now 12 ... ll\xd5 13 ll\xd5 transposes to 'a4', while in Hertneck-Kallai, European Cup, Strasbourg 1994, the knight travelled further with 12 ... ll\e6. Then 13 i..e3 .l:.b8 14 a4 a6 15

a5 b5 16 axb6 'iixb6 17 b4! �c7 18 bxc5

i.xc5 19 �d2 was enough to provide White with something to build on in the form of his supremacy in the centre. For the moment the

a-pawn is irrelevant, while White's grip on d5

and his kingside pawn majority are relevant indeed.

Incidentally 9 �h5 looks ambitious but soon peters out to equality. W ard-J .Howell, British League 1996, continued 9 ... 0-0 10 ll\f3 ll\d7 1 1 ll\g5 i..xg5 12 i..xg5 ll\f6 13 i.xf6 (Ward gives 13 'iih4 � aS+! 14 i..d2 'ifa4) 13 ... 'iixf6 14 0-0, and now instead of 14 ... �xb2 15 .l:.abl 'iif6 16 .l:.fcl Black chose the safer 14 ... .l:.b8! 15 .l:.acl b6.

With 9 ll\f 3 White intends to let the d5-square look after itself, in some cases drop­ ping the bishop back to c4 or b3, or trading on e6 after ... i..e6. Meanwhile White's knight protects d4, monitors e5 and is ready to transfer to c4 (via d2 or e5) should an oppos­ ing piece need to be removed (or attacked) or the cl-pawn - after a recapture on d5 - ad­ vanced.

9 ... 0-0 1 0 0-0

1 0 ... llla6

Occasionally starting on the edge of the board can be the most flexible way to de­ velop a piece, and here the knight is able to

Q u e en 's G a m b i t A c c ep t e d with 3 e 4

drop back to c7 to contest the d5-square or continue to e6 (-d4), or advance to b4. Of course the b4-square can be reached via c6, but in 'a', below, White puts a stop to this after 10 ... ll\c6. Here are examples of alterna­ tives to Gyimesi's speciality.

a) 10 ... ll\c6 looks fine until one considers that White might well part with his excellent bishop if it means causing serious and per­ manent damage to Black's queenside pawns. Romero Holmes-Estremera Panos, Sala­ manca 1998, is typical of how White should patiently go about addressing his opponent's new weaknesses after 1 1 i..xc6 bxc6. The game continued 12 i..e3 i..e6 13 'iic2 'iiaS 14 .l:.fcl .l:.fd8 and now White avoided 15 i..xc5? i..xc5 16 �xc5? .l:.dl+ and exploited the position of his knight with 15 ll\e5!

Notice that since doubling Black's pawns White has directed his forces against c4, c5 and, now, c6!. Additionally White is also ready to mobilise his kingside pawn majority with f2-f4.

b) 10 ... ll\d7 1 1 �e2 �c7 met with 12 e5! in Beim-Wagman, Aosta 1990. With his knight on f3 this centre/kingside expansion is available to White. After 12 ... ll\b6 13 i..e4 i..e6 (13 ... i..g4 14 i..xh7+ �xh7 15 'iie4+) 14 .l:.el Black might consider placing his rooks on d8 and e8, although White's pieces point to the kingside and the e-pawn provides White with much room for manoeuvre. Black played 14 ... .l:.ae8 15 ll\g5 i..xg5 16

i.xg5 i.d5 in order to further alleviate his defensive task with another trade of minor pieces. White's response, however, serves to remind us how a space advantage can soon grow to decisive proportions after ostensibly logical moves from the defender: 17 i.xh7+! <t>xh7 18 'Yi'h5+ <t>g8 19 i.f6!

A fitting culmination of White's strategy ­ Black has taken over the d5-square but g7 is the new focus of attention. After 19 ... gxf6 20 'it'g4+ <t>h7 21 exf6 <t>h6 22 f4 Black will soon be mated.

c) 10 ... 'it'b6 1 1 i.e3!

We have already seen that Black's queen can be embarrassed when it shares the a7-gl diagonal with White's bishop. Here the b2-pawn is safe because 1 1...'ii'xb2 12 l:r.bl and

13 i.xb7 opens the queenside in White's favour. Now Yusupov-Ehlvest, Belfort 1988 saw the automatic l 1...i.e6? allow White's intended 12 b4! with a clear advantage to White, since 12 ... 'ii'xb4 13l:r.b 1 is even worse for Black than before. Meanwhile the c5-pawn is pinned, and White threatens to open the b-file anyway. Black came up with a per­ fectly sound continuation in Ibragimov­ Estrada Nieto, Ubeda 1997: 1 1...lZ'lc6 12l:r.cl i.g4 13 h3 i.e6 14 i.xe6 fxe6 15 'ifc2 and now Chris Ward justifiably recommends 15 ... lZ'lb4, when Black must be only slightly worse. Instead there followed 15 ... 'ii'a6? 16 i.xc5 i.xc5 17 'ii'xc5 'ii'xa2 18 'ii'b5! and White threatened both 19 l:r.al and 19 'ii'xb7.

In fact the game was over after only ten more moves: 18 ... a6 19 'ifxb7 lZ'la5 20 'ifc7 l:r.f7 21 'ife5 lZ'lb7 22 lZ'lg5 l:r.e7 23 l:r.c7 l:r.xc7 24 'ifxc7 'ifxb2 25 e5 'ii'b3 26 'ii'f7+ <t>h8 27 lZ'lxe6 l:r.g8 28 lZ'ld8! 1-0.

Finally if Black maintains the pin with 13 ... i.h5 White has 14 g4! i.g6 15 lZ'ld2, heading for the attractive c4-square and toy­ ing with the idea of trapping Black's bishop with f2-f 4-f5 etc. Again this is a key differ­ ence between posting the knight on f3 and e2.

1 1 i.b3!?

As if expecting a future ... lZ'lc7 /b4 to be too inconvenient White elects to retreat his bishop anyway. Otherwise why not just get on with normal development? After 1 1 i.f 4 'ifb6, 12 lZ'le5 again hopes to justify 9 lZ'lf3 by teaming up with the queen's bishop to cover the d6-square after landing on c4. Now 12 ... 'ifxb2? 13 i.xf7+! is final, e.g. 13 ... l:r.xf7 (or 13 ... <t>h8 14 'ifh5 etc.) 14 'it'd5. Alterna­ tively 12 ... i.e6 13 lZ'lc4 'ii'd8 14 i.xe6 leaves White with both superior pieces and pawn structure.

The logical continuation is 1 1...lZ'lc7

From cl the knight hits d5 and supports

... i.e6 without voluntarily accepting an iso­ lated pawn on e6 after the bishop trade we just considered. Now 12 i.xc7 'ii'xc7 13 h3 'ii'b6 14 'ii'c2 i.e6 15 l:r.adl l:r.ad8 16 b3 i.xd5 17 exd5 i.f6 is approximately equal as it is not clear how White can successfully

infiltrate with his knight. However, in Z.Varga-Afek, Budapest 199 1, White pro­ vided us with an interesting plan which in­ volves a 'positive' retreat, namely 12 .i.c4 i.e6 13 .te2!?. The idea is to eventually ex­ ploit White's kingside pawn majority and extra space to generate play in the centre and on the kingside. Should Black seek to avoid an unpleasant offensive with 13 ... 'ii'xdl 14 .l:r.fxd l, White will already have the d-file and Black's knight will be misplaced on c7. The game went 13 ... lZ'le8 14 'ii'c2 h6 15 l:!.fdl 'i'b6 16 lZ'leS! lZ'ld6 17 i.e3 l:!.ac8 18 f4 with the by now familiar plan of lZ'lf3-e5 and f2-f4. White's bishops perform a dual-purpose task in that they support the thematic ad­ vance of the kingside pawns while simulta­ neously monitoring the queenside in order to slow down Black's counterplay. It is surpris­ ing how quickly White's kingside pawns can trouble Black's minor pieces and kingside in general, and awareness of this fact makes White's game so much easier to play. Here

the tempo of the game soon changed:

18 ... 'ii'c7 19 .tf2 g6 20 g4!? .tf6 21 lZ'lxg6! fxg6 22 eS and Black's third rank was begin­ ning to look a little suspect.

1 1 . . .lbc7

Keeping an eye on dS and preparing to challenge the bishop anyway by bringing his own to e6. Against 1 1...lZ'lb4 it has been sug­ gested that White follow up 12 'ii'xd8 l:!.xd8 with 13 i.gS!?, which looks enough for a tiny

Q u e en 's G a m b i t A c c ep t e d with 3 e4

pull. Black experiences some discomfort but with queens off he has reasonable chances to secure a level game. 1 1...'ii'b6 again meets with 12 lZ'leS when, as well as the usual out­ post on c4, White has the f7-pawn in his sights. After 12 ... i.f6 13 lZ'lc4 'ii'c7 14 'ii'f3 the prospect of the tempo-gaining .tf 4 aug­ ments White's lead. White should respond to

12 ... i.e6 with 13 lZ'ld7 .txd7 14 'ii'xd7, e.g. 14 ... .tf6 15 'ii'fS g6 16 'ii'f3 .tg7 when both 17 l:!.dl l:!.ad8 18 i.gS and 17 i.gS!? favour White thanks mainly to the poor knight.

1 2 'ilx dS

A queen trade does not necessarily mean an end to uncompromising or aggressive play. In this case White judges that his devel­ opment advantage is sufficiently significant to permit him to keep the momentum going, and there is an important factor to consider here in that the 'natural' recapture on d8 is the inferior choice. Moreover White's king­ side pawn majority is no less mobile without a queen on the board, and while other pieces remain the plan of a kingside offensive will be effective.

1 2 .. . � x dS

Gyimesi proposes the ostensibly less ac­ tive 12 ... .txd8 as an improvement. The rea­ soning behind this is that in the game Black's knight proves awkwardly placed on e6, whereas on c7 at least the dS-square is pro­ tected. Therefore by recapturing with the bishop Black defends the knight and earns time to get his queenside in order, no longer having to worry about 13 .tf4. Consequently 13 i.e3 b6 14 l:!.adl .tf6 15 eS .te7 16 .tgS!? has been suggested as White's best try, and it is true that the onus is still on Black to keep his opponent's advantage to a mini­ mum. However, this looks preferable to the greedy 14 ... i.b7 15 lZ'leS i.xe4 16 l:!.d7 .tg6 17 lZ'lxg6 hxg6 18 l:!.fdl, e.g. 18 ... lZ'le6 19 i.dS, or 18 ... l:!.c8 19 .txf7+ etc. Perhaps Black might consider 14 ... i.e6 15 .i.xe6 lZ'lxe6 with the intention of evicting White's rook after 16 nd7 l:!.e8 17 l:!.fdl lZ'lf8.

1 3 i.f4

Now White is happy to lure the knight to e6, where it will be a target - eventually - for the f-pawn. Forcing it from c7 also hands the d5-square back to White.

1 3 .. . ll:ie6

Note that 13 ... i.d6? walks into a pin after 14 .l:!.adt, when both 14 ... ltJb5 15 ltJg5 (ex­ ploiting another pin) and 14 ... ltJe8 15 i.g5 .l:!.d7 16 i.a4 practically win for White.

1 4 i.g3!

White has a development lead, the better pieces and the d5-square holds more promise than does the d4-square for Black. Add to this his potentially more dangerous pawn majority, and the exchange of queens be­ comes irrelevant.

14 . . . i.f6 ! ?

Black provokes the e5-pawn. Others: 14 ... i.d7 15 i.d5 i.c6 16 .l:!.adt i.f6 17 i.xc6 bxc6 18 i.e5 i.xe5 19 lLlxe5 ltJd4 20 f3 .l:!.ab8 (20 .. .f6 21 ltJc4) 21 b3 is typical, with Black's queenside weaknesses too big a price to pay for the outpost on d4.

The immediate 14 ... ltJd4 15 lLlxd4 favours White after both 15 ... .l:!.xd4 16 i.d5 and 15 ... cxd4 16 i.dS.

Finally the attempt to pretend normal de­ velopment is enough with 14 ... b6?! 15 .l:!.fdl! i.b7 invites White to cement his grip on d5 with 16 i.d5!, when the bishop is wonder­ fully positioned on g3 to facilitate the ad­ vance of the d-pawn in the event of an

ex-change on dS. Worse for Black is 15 ... i.a6? 16 lLle5 i.f6 17 i.dS.

1 5 ll:ie5! �8

At first glance an odd looking move, but now Black can land his knight on d4 without i.xf7 being check. After the hasty 15 ... ltJd4?, for example, White has 16 i.xf7+ �f8 17 i.dS and now 17 ... ltJe2+ 18 �hl ltJxg3+ 19 fxg3 pins the bishop! 15 ... .l:!.d2!? needs to be investigated. 16 lLlc4 16 ... .l:!.e2 17 e5 i.e7 18 i.dl .l:!.e4 19 ltJd6 leaves the rook punching air, and 16 .l:!.ab1!? also looks good for White. First 16 ... ltJd4?! 17 i.xf7+ �f8 18 i.d5 lLle2+ 19 �h t lLlxg3+ 20 fxg3 again backfires, while 16 ... i.xe5 17 i.xe5 ltJd4 18 i.dS reminds Black that half of his queenside is still at home.

16 f4!

After seeing the black king move away from the a2-g8 diagonal and on to the f-file White is more than willing to permit the doubling of his pawns if this means clearing the way for his rook.

1 6 .. J�d2?!

After this act of aggression White's pawns dominate. Black should try testing his oppo­ nent's plan regardless of the subsequent at­ tention to his f-pawn: 16 ... i.xe5 17 fxe5 .l:!.d2 (17 ... .l:!.d4 18 i.dS) 18 .l:!.f3! lLlg5 (18 ... .l:!.xb2 19 .l:!.aft ltJg5 20 .l:!.xf7+ ltJxf7 21 e6) 19 .l:!.xf7+! lLlxf7 20 .l:!.fl

If White can keep Black under pressure after taking on f7 the sacrifice will have been

worthwhile. After 20 ... 'ot>e8 21 i.xf7+ (21 .l:r.xf7? b5!) 21...'ot>d8 22 i.h4+ '.t>c7 23 i.d5!? .l:r.xb2 24 l:tf7+ 'ot>b6 25 e6 Black's rook and bishop are still shut out of the game, White's e-pawn is close to glory and Black's kingside pawns are doomed. Black does have the c­ pawn, but it lacks support. The other try is 20 ... b5 21 i.xf7, when 21...l:tdl 22 l:txdl

'ot>xf7 23 l:td5 is good for White, as is

21...l:tb8 22 i.f4!? 'ot>xf7 (22 ... l:te2 23 e6 l:tb6 24 i.h5; 22 ... l:txb2 23 e6 l:tb6 24 i.g5) 23 i.xd2+ 'ot>e6 (23 ... 'ot>e7 24 i.e3) 24 l:tf8, e.g. 24 ... l:ta8 (24 ... 'ot>xe5 25 i.f4+) 25 l:te8+ 'ot>f7 26 l:td8 'ot>e6 27 i.f 4 c4 28 'ot>f2 b4 29 'ot>e3 c3 30 bxc3 bxc3 3 1 'ot>d4 i.b7 (3 1...c2 32 .l:r.d6+ 'ot>e7 33 l:tc6) 32 l:td6+ 'ot>e7 33 i.g5+ �e8 34 l:te6+ 'ot>f8 35 l:te7 l:td8+ 36 'ot>xc3 i.xe4 37 l:txa7 etc.

1 7 lDf3!

Better than 17 lZJc4 i.d4+ 18 'ot> h 1 l:te2 19 i.dl l:txe4 20 i.f3 l:txf4 21 i.xf4 lZJxf4 which might give White an edge.

1 7 .. . �xb2 1 8 e5 i.dS 1 9 f5 lDg5

19 ... lZJd4 20 lZJxd4 cxd4 21 e6 fxe6 22 fxe6+ i.f6 23 i.d6+ 'ot>e8 24 l:tae1 and White brings his final piece into play.

20 e6

20 . . . lDxf3+

Or 20 ... lZJe4 21 i.e5 l:te2 22 l:tael! l:txel 23 l:txel lZJf6 24 exf7 i.e7 25 g4 b5 (25 ... i.d7 26 lZJg5; 25 ... lZJxg4 26 i.d6!) 26 g5 c4 (26 ... i.b7 27 gxf6 gxf6 28 i.f4 c4 29 i.d 1 !) 27 i.c2!, and now 27 ... lZJg4 28 i.xg7 +

Q u e e n 's G a m b i t A c c ep t e d with 3 e4

'ot>xf7 29 g6+ hxg6 30 fxg6+ 'ot>xg7 31 l:txe7 + 'ot>f6 (3 1...'ot>g8 32 i.e4 l:tb8 33 i.d5+) 32 l:tf7+ 'ot>e6 33 l:tf8 and 27 ... i.b7 28 lZJd4 lZJdS 29 g6 hxg6 30 fxg6 i.c5 31 'ot>fl win for White.

21 �xf3 �d2 22 i.f4 l:td4 23 i.e5 i.f6

Black's rook is in danger of running out of steam, e.g. 23 ... l:td2 24 i.c3 l:te2 25 l:tdl.

24 i.xf6 gxf6 25 �e 1 fxe6

Not 25 ... c4? 26 e7+ '.t>e8 27 i.a4+ i.d7 28 l:tg3 i.xa4 29 l:tg8+ 'ot>d7 30 l:txa8 etc.

26 fxe6 �e7

The king is not an ideal blockader.

27 �g3 ! ?

27 l:th3 is more logical, e.g. 27 ... 'ot>d6 (27 ... c4 28 l:txh7+ 'ot>d6 29 l:th8 cxb3 30 el) 28 l:txh7 c4 29 l:th8! 'ot>e7 30 i.c2.

27 . . . b5 28 l:tg7+ Wd6 29 �gs �e7

White wins after both 29 ... c4 30 e7 and 29 ... i.b7 30 l:txa8 i.xa8 3 1 e7 i.c6 32 l:te6+.

30 i.c2! l:td5 31 l:tg7+??

Whoops. White can get another passed pawn rolling after 3 1 i.xh7 l:te5 32 l:txe5 fxe5 33 i.e4 l:tb8 34 h4, e.g. 34 ... c4 35 h5 b4 36 h6 b3 37 axb3 cxb3 38 h7, or 34 ... 'ot>xe6 35 h5 'ot>f7 36 i.dS+ 'ot>e7 37 h6.

31 . . . �d6 32 e7 i.e6 33 i.e4 �eS ! 34 i.xd5 i.xd5 35 .:te2?

Again White ignores the h7-pawn: 35 l:txh7! i.xa2 36 l:tdl+ 'ot>e6 37 l:tel+! draws. Now Black is in the driving seat ...

35 ... c4 36 .:txh7 c3 ! 37 >ii>f2 b4?

ea-ture in the rest of the game - understandable in such a complex ending. Correct is 37 ... i..c4! ! 38 l:te3 b4 39 l:th6 c2 40 l:txf6+ �d7 41 l:tel i.xa2 42 l:ta6 b3 43 l:txa7+ Wc6 44 .i::tcl Wd6.

38 �e3 !

Excellent. White faces facts and is pre­ pared to part with his once mighty e-pawn.

38 . . . Wc5?!

38 ... l:txe7+ 39 l:txe7 Wxe7 40 Wd4+ Wd6 41 h4 a5 42 h5 i.xa2! 43 l:txa2 b3 44 Wxc3 bxa2 45 Wb2 We7 46 g4 Wf7 47 �xa2 Wg7 48 Wb3 Wh6 49 Wa4 f5 with a draw.

39 l:f2 .i.xa2!

A good try. Instead 39 ... aS 40 l:txf6 i.xa2 41 l:tf8 wins for White.

40 l:xa2 b3 41 l:th5+?

Wrong rook! 41 l:taS+ wins: 41...Wd6 (4 1...�b4 42 l:txa7 c2 43 Wd2 l:tc8 44 l:tb7+ Wa3 45 Wet; 41...�b6 42 l:ta3 b2 43 l:tb3+

WaS 44 Wd3; 41...Wc6 42 l:ta3 c2 43 Wd2; 41...Wc4 42 l:th4 mate) 42 l:ta6+ Wd7 43 l:txa7+ �e6 44 l:tb7 b2 45 �d3 l:tc8 46 e8'ii' + !. 4 1 ... Wd6 42 �a6+? 42 l:tal! b2 43 l:tdl+ Wxe7 44 l:tc5 Wf8+ 45 �f4 l:tb8 46 l:tbl l:tb3 47 �e4 l:ta3 48 �d3 . 4 2 .. . �xe7 43 l:th7+ Wf8+??

The final blunder. White still has work to

do after 43 ... �d8+! 44 �d3 c2 since here after 45 l:taxa7 Black can queen his pawn. I have a feeling that White should have a win somewhere, but there is always the simplify­ ing 45 l:td6+ �c8 46 l:tc6+ �b8 47 l:tcc7 l:tel! 48 l:tb7+ �c8 49 l:txa7 �b8 50 l:tab7+ Wc8 5 1 l:txb3 cl'i' 52 l:tc3+ 'ii'xc3+ 53 Wxc3.

44 Wd3 c2 45 .:taxa 7 1-0

After 45 ... �g8 46 l:tag7+ Wf8 47 l:tb7 Wg8 48 l:thc7 and 45 ... l:te7 46 l:ta8+ l:te8 47 l:txe8+ Wxe8 48 l:tc7 (or 48 �d2) are fairly simple, while the other try 45 ... l:td8+ is met by 46 We4 l:te8+ (46 ... �g8 47 l:tag7+ Wf8 48 l:tb7 Wg8 49 l:thc7) 47 Wf5 l:te7 48 l:ta8+ l:te8 49 �xf6! �g8 50 l:tg7+ �f8 51 l:tf7+ mates. Game 4 Sakaev-lbragimov Russian Ch. ₁₉₉₉ 1 d4 d5 2 c4 dxc4 3 e4 c5 4 d5 l'Lif6 5 l'Lic3 b5

Bizarre but entirely logical! Black protects his extra pawn and threatens to undermine the defence of White's e4-pawn by attacking the knight with ... b4. Of course the b5-pawn itself is not def ended, but White should refrain from taking it, e.g. 6 lZJxb5?! 'ii'aS+ 7 lZJc3 lZJxe4 8 i..d2 lZJxd2 9 'ii'xd2 .ia6 10 lZJf3 lZJd7 11 i.e2 g6 and Black was better, Korchnoi-Lindinger, Wichern 1997.

6 .i.f4

6 e5 has the right look but is less effective than Sakaev's choice. After 6 ... b4 7 exf6 bxc3 8 bxc3 Black should play 8 ... lZJd7! when 9 'ii'a4 (9 fxe7 i..xe7 10 .ixc4 lZJb6 l l i.b5+ .id7 is fine for Black) 9 ... exf6 10 i.f4 'i'b6 1 1 i.xc4 i.d6 is equal. Instead 8 ... exf6?! 9 i.xc4 i.d6 10 'ii'e2+ invites Black into an ending in which White's superior pawn struc­ ture counts for something, while Bacrot­ Peric, Corsica (rapid) 1997, continued 8 ... 'i'aS 9 'i'd2 gxf6 10 .ixc4 lZJd7 1 1 lZJf3 lZJb6 12 i.e2 c4 13 0-0 i.b7 14 l:tdl l:td8 15 'i'f4 with a development lead for White worth a pawn (or two). Black chose 15 ... lZJxd5? rather than the better 15 ... i.xdS, and soon paid the price: 16 'i'xc4 lZJxc3 17

:xd8+ 'it>xd8 18 i.fl i.xf3 19 i.d.2! 'iic7 20 'ifd3+ tt:Jds 2 1 llct 'ii'b6 22 llbt!? 'ii'c7 23 :bs 'it>e8 24 llxdS i.xdS 25 'ii'xdS.

With 6 i.f4 White simply brings a piece to

a good square and prepares to add weight to lt:JbS in the event of ... b5-b4 by hitting the cl-square. Consequently Black has three ways of defending bS before attacking the knight.

6 . . . i.a6

As a6 may well be the best square for the bishop - especially as from here the c4-pawn is given added protection as well as the bS­ pawn - this relatively recent idea is a promis­ ing alternative to the outwardly more aggres­ sive pin (see below).

Others:

a) 6 ... 'ii'aS is popular.

al) 7 a4 is better than its reputation, al­ though after 7 ... tt:Jxe4 White should ignore 8 axbS lt:Jxc3 9 llxaS lt:Jxdl 10 'it>xdl lt:Jd7 1 1 i.xc4 lt:Jb6 1 2 b3 i.b7 1 3 lt:Jf3 i.xdS in favour of 8 lt:Jge2 lt:Jd6 9 axbS 'ii'b6 10 i.xd6, e.g. 10 ... exd6 11 lt:Jg3 i.e7 12 i.xc4 0-0 13 0-0 i.f6 14 'i'c2, Beliavsky-Kamsky, Linares 1993, or 10 ... 'ii'xd6 1 1 lt:Jg3 g6 12 .i.xc4 i.g7 13 0-0 0-0 14 lt:Jge4, with a slight edge for White in both cases despite the ab­ sence of his dark-squared bishop.

a2) Again 7 eS cries out to be played but should be avoided. After 7 ... lt:Je4 8 lt:Jge2 lt:Ja6 9 f3 lt:Jb4! 10 fxe4 lt:Jd3+ 1 1 'it>d.2 g6!! Black was having all the fun in Gelfand­ Anand, Linares 1993.

see follo wing diagram

There followed 12 b3 i.g7 13 bxc4 lt:Jxf4 14 lt:Jxf4 i.xeS 15 lt:Jfe2 b4 16 'i'a4+ (16 lt:Ja4 i.d7) 16 .. .'ii'xa4 17 lt:Jxa4 i.xal 18 lt:JxcS 0-0 and by now Black was winning. Later, in Beliavsky-Comp Fritz 4, Slovenia (rapid) 1996, White tried 12 e6 lt:Jf2 13 'i'el lt:Jxhl 14 a4 fxe6 15 axbS 'i'b4 16 'it-cl. This does look like an improvement and is worth further investigation, but White has to be at least as careful as Black in this line, so it

Q u e en 's G a m b i t A c c ep t e d with 3 e 4

makes more sense to settle for 'a3' or 'a4', both of which seem superior anyway.

a3) 7 f3 sensibly bolsters the centre and denies Black use of the g4-square. Now 7 ... gS?! has been seen occasionally but the pawn is too much to pay for Black's subse­ quent Benko-style activity after 8 i.xgS. Over on the queenside 7 ... b4 is well met by 8 'ii' a4+. The main line is 7 ... lt:JhS 8 i.d.2 tt:Jd7 9 f4

Vyzmanavin-Azmaiparashvili, Burgas 1994. After the forced 9 ... g6 (9 ... lt:Jhf6? 10 eS) Vyzmanavin proposes 10 eS i.h6 1 1 lt:Jge2 with the threat of snaring a piece by launching the g-pawn. Black's best is the obvious 1 1 ...b4, e.g. 12 lt:Je4?! i.b7, so Ward suggests 12 'i'a4 with an evaluation of un­ clear. Instead the game went 10 i.e2 i.g7 (10 ... b4 1 1 'Wa4! is a reply given by Ftacnik) 1 1 es b4

and now White preferred the thematic 12 'i'a4 to 12 lZ'le4 i.b7 13 i.xc4 lZ'lb6 14 'ii'b3 in view of 14 .. J1d8. Then the retreat 12 ... 'ii'b6 works out well for White after 13 lZ'le4 i.a6? 14 e6! fxe6 15 dxe6 'ii'xe6 16 lZ'lxc5, or the lesser evil 13 ... i.b7 14 i.xc4. However, Black might consider 12 ... 'i'd8 13 lZ'le4 0-0 when the consistent 14 g4 means parting with a couple of pawns for the piece after either 14 ... lZ'lxf4! 15 i.xf4 lZ'lb6 or 15 ... lZ'lxe5. Nevertheless in these complex positions the extra piece tends to be more useful than pawns. If this is not to Black's liking Ward's clever 12 ... 'i'a6!? is a realistic possibility, forcing the exchange of queens on Black's terms. The game itself continued 12 ... 'fi'xa4 13 lZ'lxa4 i.a6 14 g4 lZ'lxf4 15 i.xf4 i.b5 16 b3 i.xa4 (16 ... lZ'lxe5) 17 bxa4 lZ'lxe5 18 0-0-0 c3 19 i.b5+ 'itf8 20 h3 l:!.d8 21 l:!.h2 and both sides had chances of mak­ ing something of their respective material 'gain'.

a4) 7 i.d2 is most successful in the line 7 ... b4 8 e5 bxc3 9 i.xc3, e.g. 9 ... 'i'a6 10 exf6 exf6 1 1 b3 i.e7 12 i.xc4 'i'd6 13 lZ'le2 0-0 14 0-0 with a pleasant position for White in Shirov-Kramnik, Linares 1993. However, 8 ... lZ'lg4!? 9 e6 lZ'lf6 is stronger for Black than it first appears and will be just as inconven­ ient to White as e5-e6 has been for Black. Another possibility to consider is 7 ... e5 8 f4 lZ'lbd7 9 lZ'lf3 i.d6 with a view to establishing a grip on the dark squares. All in all I prefer 7

f3.

b) 6 ... a6 guards b5 with a lowly pawn rather than bringing out a piece. After 7 e5 b4 8 exf6 bxc3 9 bxc3 Black has tried three moves:

bl) 9 ... 'ii'aS 10 l:!.cl gxf6 1 1 i.xc4 h5 12 lZ'lf3 lZ'ld7 13 0-0 lZ'lb6 14 i.c7

Legky-Peric, Martinez 2000. The diagram position is hardly any better for Perie than his game against Bacrot, above. The fact that

an IM can find himself in trouble so early

illustrates how Black's game-plan can so eas­ ily fall apart in the 5 ... b5 variation. In fact after 14 ... l:!.a7 15 d6 exd6 16 l:!.el+ i.e6 17 i.xb6 'i'xb6 18 l:!.bl 'ii'd8 19 i.xe6 fxe6 20 l:!.xe6+ 'itf7 21 'ii'dS 'itg7 22 lZ'lh4 Black's situation was embarrassing enough to resign at once.

b2) 9 ... lZ'ld7 10 'i'a4 'i'b6 1 1 fxe7 i.xe7 12 i.xc4 'i'b2 13 l:!.cl i.d6 14 i.e3 l:!.b8 15 tt:Jf3 0-0 16 i.d3 f5 17 0-0 and Black - faced with the prospect of lZ'lf3-d2-c4 - was much worse in Markeluk-Juarez, Buenos Aires 1989.

b3) The same players reached the same position in a tournament in Acasusso 199 1. Then Black tried 9 ... gxf6 10 i.xc4 lZ'ld7 11 'i'a4 i.g7 and White endeavoured to justify his early queen sortie with 12 'i'c6!? l:!.a7 13 l:!.bl 0-0 14 d6!? exd6 15 i.xd6 lZ'le5 16 i.xe5 l:!.e8 17 lZ'le2 fxe5. White's light-square control is enough to leave him comfortably