Yelena Dembo

EVERYMAN CHESS Gloucester Publishers pic www.everymanchess.com

First published in 2007 by Gloucester Publishers plc (formerly Everyman Publishers plc), Northburgh House, 10 Northburgh Street, London EC1V OAT

Copyright © 2007 Y elena Dembo

The right of Yelena Dembo 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.

ISBN13: 978 1 85744 521 3

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, Northburgh House, 10 Northburgh Street, London EC1 V OAT; tel : 020 7253 7887; fax: 020 7 490 3708; email: info@everymanchess.com; website: www.everymanchess.com

Everyman is the registered trade mark of Random House Inc. and is used in this work under licence from Random House Inc.

EVERYMAN CHESS SERIES

Chief advisor: Byron Jacobs Commissioning editor: John Emms Assistant editor: Richard Palliser

Typeset and edited by First Rank Publishing, Brighton. Cover design by Horatio Monteverde. Production by Navigator Guides. Printed and bound in the US by Versa Press.

C o n t e n t s I

Bibliography Introduction

Part One -The Exchange Variation (1 d4 tt:lf6 2 c4 g6 3 tt:lc3 d5 4 cxd5 tt:lxd5 5 e4 tt:lxc3 6 bxc3 .tg7)

4 7

1 Introducing 7 tt:lf3 c5 12

2 The Critical8 �b1 18 3 .te3 Systems 35 4 The Fashionable 7 .tc4 51 5 White's 7th Move Alternatives 71

6 4 cxd5 tt:lxd5 without 5 e4 81

Part Two - 3 tt:lc3 Without The Exchange

7 The Russian System 8 � a4+ Systems 9 .tf4 Systems

10 .tg5 Systems 11 The Solid 4 e3 12 Offbeat 4th Move Alternatives

Part Three - White Postpones tt:lc3

13 The Fianchetto System 14 Angling for a Samisch: 3 f3

Index of Variations

92 102 106 127 149 159

164 184

189

Bibl i o g ra p h y I

Griinfeld Sou rces

Winning With The Grunfeld, Andras Adorjan & Jeno Dory (Macmillan 1987) A revolutionary and early book on the Griinfeld, now outdated in some places but still very relevant in others. This is really where it all started!

The Complete Grunfeld, Alexei Suetin (Batsford 1991) Large reference work, covering all the major systems.

Beating the Grunfeld, Anatoly Karpov (Batsford 1992) Essentially a collection of Karpov's best games in the Griinfeld, containing deep and instructive material on his favourite systems.

Fianchetto Grunfeld, Adrian Mikhalchishin & Alexander Beliavsky (Cadogan 1998) Significant reference work on the g3 systems by two renowned analysts.

Understanding the Grunfeld, Jonathan Rowson (Gambit 1999) A must-have for all Griinfeld players. This book explains both the basics of the opening as well as more intricate concepts.

The Grunfeld, Nigel Davies (Everyman 2002) Repertoire book for Black featuring plenty of interesting ideas.

King's Indian & Griinfeld: Fianchetto Lines, Lasha Janjgava (Gambit 2003) Large reference work on the g3 systems; very comprehensive and well researched.

Bib l iog raphy

Starting Out: The Griinfeld, Jacob Aagaard (Everyman 2004) Everyman's own introductory guide to the Griinfeld; a very good alternative to Rowson's book.

Challenging the Griinfeld, Edward Dearing (Quality Chess 2005) A book dedicated to the 7 tt:lf3 c5 8 .l:.b1 Exchange Variation; definitely the bible for adherents of this system.

An Expert's Guide to the 7 ii.c4 Gruenfeld, Konstantin Sakaev (2nd edition, Chess Stars 2006) Not just a repertoire book, but perhaps the most comprehensive and remarkable opening work ever! Covers absolutely everything under the sun on the 7 ii.c4 Ex­change Variation for both sides and includes an incredible amount of original analysis.

Other Sou rces

Beating the Fianchetto Defences, Efstratios Grivas (Gambit 2006) Repertoire book for White, covering the 4 tt:lf3 Ji.g7 5 i.g5 tt:le4 6 Ji.h4 variation, on which Grivas is the world's leading expert.

Beating the Indian Defences, Graham Burgess & Steffen Pedersen (Batsford 1997) Repertoire work for White, covering the 7 Ae3 Exchange Variation.

Beating the King's Indian & Griinfeld, Timothy Taylor (Everyman 2007) Repertoire book for White, covering two offbeat lines against the Griinfeld.

Black is still OK!, Andras Adorjan (Batsford 2004) Collection of articles about various aspects of chess, including original ideas of the author in several Griinfeld lines.

Dynamics of Chess Strategy, Vlastimil Jansa (Batsford 2003) A strategic chess tutorial by a legendary grandmaster, with a large and very in­structive section on several Griinfeld systems.

Play 1 d4!, Richard Palliser (Batsford 2003) Repertoire book for White, covering 4 tt:lf3 Ji.g7 5 Ji.g5 tt:le4 6 cxd5.

The Soviet Chess Conveyor, Mikhail Shereshevsky (Semko 1994) An instructional book and not an opening manual, but with a few paragraphs on a

5

Play t h e G r ii nfe /d

specific line of the 4 .if4 system.

Data bases, Period ica ls and Websites

Chess Informant, ChessPublishing.com (Griinfeld coverage chiefly supplied by Tisdall, Davies and Flear), Mega Database 2007 (ChessBase), New In Chess Magazine,

New In Chess Yearbook, The Week In Chess ( 1 -639) and UltraCorr (Chess Mail) .

6

I n t ro d uct i o n I

Welcome to the fascinating world of the Griinfeld! The book you are holding is a repertoire work for players who wish to employ

the Griinfeld with Black. It is aimed at players of intermediate level up to that of Grandmaster (hopefully!) and aspires to provide everything you need to know about this dynamic defence which begins 1 d4 li:Jf6 2 c4 g6 3 lt:lc3 d5.

A History of the Gri.infeld

The Griinfeld Defence (I believe that 'counterattack' would be a more appropriate term!) was introduced into the higher echelons of competitive chess in the heyday of the hypermodern movement, namely in the 1920s and 1930s. It bears the name of the Austrian master Ernst Griinfeld who was its first devoted adherent. Never­theless, like several opening set-ups conceived in those days, it was forced to re-

7

Play t h e G r ii nfe ld

main rather a theoretical backwater for some time, presumably because i t ap­peared too avant-garde to the more classically-minded spirit that prevailed back then.

The evolution of the Soviet Union as a chess power provided the necessary boost to catapult the Gri.infeld to the status of a respectable opening. The more dynamic, less stereotyped way of thinking that characterized the best Soviet play­ers of that time found a ferti le field of expression in the Griinfeld. Indeed, the powerful World Champion Mikhail Botvinnik regularly employed the Gri.infeld in his games and his example was soon followed by other Soviet stars, most nota­bly his successor to the throne, Vasily Smyslov. The Gri.infeld even made its ,,·c.:. to the World Championship matches between the two, thus establishing itself a,:::: sound and viable opening.

The Sixties were a decade crucial to the further development of the Griinte::. with the opening being adopted by Bobby Fischer in some high-profile games, ::­particular two against Boris Spassky. Both those games had a dismal outcome for the American, although the opening itself was hardly to blame. Then the end of the Sixties saw the birth of the two great schools of the Griinfeld. The Hungarian school, led by the innovator Andras Adorjan and Zoltan Ribli, quickly established itself as the leading Gri.infeld ' family' in the world, followed by the Czechoslovak school, whose leading protagonists were Vlastimil Jansa, Jan Smejkal and Lubomir Ftacnik. In fact, under the influence of those outstanding players, the two countries developed a whole host of other players who greatly contributed to Griinfeld theory; Peter Leko being the most obvious but by no means only exam­ple.

Andras Adorjan was the author (along with his compatriot Jeno Dory) of the first really influential book on the Griinfeld in 1987, but even more importantly he introduced the opening to the young and ambitious Garry Kasparov at the time of his first World Championship match with Anatoly Karpov. It was not until their third match in 1986 that Kasparov actually employed the Gri.infeld, but when he did so it came with remarkable success; Karpov proved completely unprepared, tried several systems and was generally unable to trouble Kasparov. Overall, Kas­parov's opening choice was deemed a success and the Gri.infeld was back at the top; a position it retained throughout the legendary pair's further matches in 1987 and 1990.

At the end of the Eighties some young Soviet masters, mainly from St Peters­burg, introduced a system against the Griinfeld which would dominate tourna­ment praxis in the Nineties, namely the 7 lt:lf3 c5 8 l:tb1 Exchange Variation. The system was analysed heavily in every corner of the world and at first kept scoring points for White, especially thanks to the efforts of Alexander Khalifman, Boris Gelfand and Vladimir Kramnik. In turn several top players, especially Vishy An-

8

I n t ro d u ct io n

and, Alexei Shirov and Peter Leko, began to defend the Griinfeld and managed to find some solutions to the 8 .l:i.b1 problem. Indeed, the variation is currently declin­ing in popularity, whereas 7 .ic4, a popular favourite in the Sixties, has returned to the forefront of fashion.

The Nineties also saw a massive emigration of Soviet chessplayers, with one of the main destinations being Israel . There the third great Griinfeld school was born, largely thanks to the efforts of the legendary Mark Tseitlin. Nowadays there are a lot of Israeli grandmasters who employ the Griinfeld, most notably Emil Sutovsky, Boris A vrukh, Viktor Mikhalevski and Alexander Huzman, and the Israeli school currently represents the most active laboratory of opening ideas in the Griinfeld.

Finally, one cannot sum up the current status of the Griinfeld without mentioning the Russian grandmaster Peter Svidler, at this moment the world's strongest Griinfeld practitioner. Whereas Kasparov, Anand and Leko eventually dropped the Griinfeld from their repertoires, opting for more solid openings that are less vulnerable to computer-assisted research, Svidler has persistently em­ployed the beloved opening of his childhood and still does, suffering a few pain­ful defeats along the way but also scoring several victories. It is also encouraging to see that, after a brief intermission, the Griinfeld is again appearing regularly at the highest level, in the hands of young and ambitious players such as Alexander Grischuk, Shakhriyar Mamedyarov and David Navara. Indeed, the scene may well be set for a full-scale revival of this great opening!

My Gri.infeld Qua l ifications

As for myself, my relationship with the Griinfeld began while I was living in Is­rael in the 1990s. I was introduced to the opening there, playing and analysing it extensively for a number of years; something which continued when I later moved to Hungary and felt the influence of the first great Griinfeld school. Of late, I have also become a little infatuated with the King's Indian, but have continued to stay fully up-to-date with the latest Griinfeld developments and to search for new re­sources for Black. Perhaps having now written this work, it is time for me once again to make the Griinfeld my exclusive defence to 1 d4!

About This Book

Rather than focus chiefly on the general themes of the opening, I have aimed first and foremost to supply a complete repertoire for Black with the Griinfeld. En route we will discuss typical tactical motifs and strategic ideas as we come across them, but there is especially plenty of theoretical coverage. Indeed, I can assure

9

Play th e G r ii nfe ld

you that, should you study this book comprehensively, your preparation would be on a par with that of almost any IM in the world! Every single significant op­tion at White's disposal has been covered in considerable depth where necessary, but by pointing out the main concepts and plans for both sides, I believe that this work is also very suitable for the ordinary club player.

Due to the Griinfeld' s very concrete and often quite tactical nature as an open­ing, I have opted for a variation tree approach within this book. Indeed, most of the time the Griinfeld is all about piece activity and being the first to create threats. Concepts such as fixed pawn structures, slow manoeuvring and long-term planning are fairly rare in the Griinfeld, in contrast to more strategic openings such as the Ruy Lopez. Moreover, both sides usually have several choices at the many theoretical crossroads which we will come across. There is plenty of theory to be discussed, although I will not be advocating any lines which require exces­sive rote-learning - I'm assuming that the reader isn't preparing for a World Championship match and so doesn't want to learn 30+ moves just to force a draw! - and so a variation tree approach, rather than a complete games format, again makes the best sense.

So what can you expect to find in this book? The answer is simple : it contains coverage from Black's perspective of all White's reasonable (and some not-so­reasonable) tries against the Griinfeld, offering at least one way to meet each of them successfully. At certain critical junctures I have opted to offer more than one option, primarily in those systems which are very popular and have developed a large body of theory. That should help to ensure that this repertoire will remain viable for many years. I have also decided not to enter the more complex and theoretically-heavy main lines, in favour of more positional schemes and without making any sacrifices in the quality department. Indeed, I believe that the sug­gested repertoire is the optimal combination of opening ambition, preparatory study and efficiency for the non-professional player!

The Griinfeld contains a number of theoretical lines, but even here understand­ing the key plans for both sides is very important. Thus our coverage begins with the most natural form of the Exchange Variation, with White playing 7 CDf3 and then trying to develop normally. It is essential to understand why such an ap­proach does not work out well, so as to better understand the motivation behind the critical 8 �bl of Chapter Two.

By no means all white players want to take on the Griinfeld with an especially theoretical approach. Unfortunately for us, some of the available quiet set-ups can be rather dull, such as 4 CDf3 i.g7 5 i.g5 or 4 e3; to use a cliche, watching grass grow can sometimes be more fun! Please note that when a white system falls into this category, I have almost always opted to recommend the particular line that fully neutralizes it. Just learning that accurate response for Black should mean that

10

I n tro d u c t i o n

you never need worry ever again in your Grunfeld career about the white system in question.

Of late, one can also observe white players turning towards some very new and ambitious systems, especially the 4 cxd5 Ci:Jxd5 5 ctJa4 and 5 �d2 variations. Nowadays, the former is almost extinct, for reasons we will see in Chapter Six, but the latter continues to figure in the repertoire of several grandmasters. It chiefly aims to steer play into relatively uncharted territory; an approach rather character­istic of our times, especially in the Grunfeld. All such systems deserve some re­spect and I have certainly examined them closely, even suggesting some entirely novel ideas when the existing options failed to satisfy me. In Chapter Fourteen, for example, one will find 3 f3 ! ? - a popular move, but also one wrongly neglected by many previous Grunfeld works - met with the dynamic 3 . . . e5 !? ; a move fully in keeping with our active black approach.

Finally, I would like to inform you, dear reader, that every single line in this book has been checked with at least one analysis engine. However, the Grunfeld is very dynamic and often leads to rather unbalanced positions. That makes it a fas­cinating opening to play, albeit also a little daunting at times to evaluate . Please note that many lines are assessed as giving Black 'the initiative' . That does not necessarily mean that Black is better, but rather that he is calling the shots. In prac­tice, though, having to defend accurately against a Grunfeld initiative often proves too much for White, so do strive to gain the initiative!

Acknowledgements

I am very much obliged to my parents, Vladimir Dembo and Nadezhda Fokina, as well as to my husband Sotiris Logothetis, for their immense help in the creation of this book.

My gratitude is also due to John Emms for the help, suggestions and useful material he provided throughout, as well as to Richard Palliser, who successfully turned a complex manuscript into the book you are now holding!

Yelena Dembo, Athens,

March 2007

1 1

Ch a pte r On e I Exc h a n ge Va riat io n : Int rod u c i n g 7 ttJf3 cs

1 d4 Ci:Jf6 2 c4 g6 3 Ci:Jc3 ds 4 cxds ct:Jxds

5 e4 ct:Jxc3 6 bxc3 Ji.g7 7 Ci:Jf3 cs

Apart from the highly-theoretical 8 .l:!.b1, the subject of our next chapter, White has tried a few other ideas.

A: 8 Ji.e2 B: 8 h3

c: 8 Ji.bs+

The coverage of Line A shows the motivation behind both Lines B and C,

namely to prevent Black from applying pressure against d4 with . . . Ci:Jc6 and . . . .i.g4. So let's see first what happens if White tries to develop naturally:

A) 8 Ji.e2 ct:Jc6

White must already tread carefully. To see why consider:

a) 9 e5? ! is too committal and after 9 . . . 0-0 10 0-0 cxd4 1 1 cxd4 Ji.e6 12 .i.e3 .i.d5 13 Ci:Jg5 (H.Grooten-P.Boersma, Amsterdam 1982) 13 . . . £6!, as suggested by Krnic, Black has an excellent game.

Exc h a n g e Var ia t io n : I n t roduc ing 7 l:iJf3 c5

b) 9 .ltb2?! cxd4 10 cxd4 'iYa5+ 1 1 �fl .ltg4 1 2 'iYd3 .l:td8 13 .ltc3 'iYa3 was again quite good for Black in H.Konar­kovska-M.Chiburdanidze, Palma de Mallorca 1989.

c) 9 .l:tb1 ? ! cxd4 10 cxd4 tt:lxd4 1 1 tt:lxd4 'iYxd4 12 .ltb5+ �f8 13 'iYe2 a 6 14 .id3 b5 15 0-0 .lte6 simply leaves Black with an extra pawn.

d) 9 0-0? ! cxd4 10 cxd4 tt:lxd4 1 1 tt:lxd4 'iYxd4 12 .ltb5+ �f8 1 3 .lte3 'iYxd1 14 .l:taxd1 �e6 15 �xa7 f6 again sees White fail to justify his material in­vestment.

Therefore White's choice is re­stricted to the risky 9 d5? ! and to shor­ing up the defence of d4.

A1: 9 ds A2: 9 .lte3

A1) 9 dS?! A rather premature exchange sacri­

fice . 9 ... i.xc3 + 10 .ltd2 .ltxa1 11 'iYxa1 tt:ld4 12 tt:lxd4 cxd4 13 'iYxd4 0-0

rial, at the cost of seriously exposing his king, or continue the game an ex­change down, with only minimal prac­tical chances of an attack. 14 0-0

White plays for compensation. He can do so too with 14 .ltc3 f6 15 h4, but 15 . . . 'iYb6 16 'iYd3 .ltd7 17 0-0 .l:tac8 18 i.d4 'iYa5 is very good for Black. In­stead 14 i.h6 recovers the exchange, but at too high a price: 14 . . . 'iVa5+ 15 �fl f6 16 .ltxf8 �xf8

17 f3 ( 17 g4 iYxa2 18 e5 �g8 is better for Black) 17 . . . .ltd7 18 �f2 .l:tc8 19 .l:tb1 .l:tc2! 20 .l:tb2 .l:txb2 21 'iYxb2 'iYc5+ 22 �g3 (22 �fl b5 gives Black a clear ad­vantage) 22 . . . b5 23 a3 was R.Hernan­dez-M.Sisniega, Matanzas 1992, and Black could now have won in the fol­lowing way, as pointed out by Sis­niega: 23 . . . g5 ! 24 'iYb4 (or 24 h3 'iYg1 25 d6 h5 26 dxe7+ �xe7 27 'iYb4+ �f7 and wins) 24 . . . 'iYe3 25 .ltfl h5 26 h3 'iYg1 with a winning position. 14 . . . 'iYb6 15 'iYa1

Or 15 'iYc3 f6 16 .ih6 .l:tf7 17 i.e3 (S.Mohr-S.Lputian, Altensteig 1989)

Now White can win back the mate- 17 . . . 'iYd6 18 f4 e6 with an obvious plus.

1 3

·· fe ld t h e Grun P lay

i,h6 16 18 � d7 16 l:i:b1 Wic7 �!, 3 l:i:c8

15........ Wid4 b5 20 lOfe Ji.xfS �xf8 19

British League . J Rowson, In M.Pem- . er hand. 2000, Black had the upp

l forcing a Immediate y concession from White .

0 l:i:b1 Ji.xf3 10 es

bette• afte' I • d4

Black was d4 12 cxd4 .ax f3 (or 11 Ji.xf3 ex

. th an extra l l gx

14 0 0 b6 w> 7 13 i.xd4 Ci:Jxd4 -d4 Ji.xd4 13 l:i:xb ) 1 1 cxd4 12 ex

1. p Svidler, pawn . . .

in T.Geza wn- . -0 14 0-0 e6 0 990 Leningrad 1 .

Wia5+ 12 Ji.d2 d4 11 cxd4 10 ... cx

1 4

A Khasin, V Utemov- . The 12 Wid2 of . f Black after · good or USSR 1986, IS

Ji.e6. 12 Wixd2+ 13 Ci:Jxd2 4 Wia4 Wid7 15

... l:i: 1l:i:d8 1 12 ... Wic7 13 c

Ji.e3 o-o

nly increa-1 ms have 0 White's prob e Alushta 2006 . . hin y Zinchenko, sed, D.Tis - .

Exch a n g e Var ia t io n : I n t roduc ing 7 li:Jf3 c5

8 ... 0-0 9 i.e2 Instead 9 ii.e3 'ifa5 10 'ifd2 (10 i.d2

Ci:Jc6 11 d5 Ci:Je5 12 ct:Jxe5 i.xe5 13 'ifb3 .Ub8 14 f4 i.g7 15 e5 b5 is quite good for Black) 10 . . . Ci:Jc6 11 .Ud1 was tried in F .Gomez-A.Zapata, Havana 2001, and now 1 1 . . .i.e6! would have set White some problems. 9 ... cxd4 10 cxd4 ct:Jc6 11 ii.e3

Now Black cannot further increase the pressure on d4 and White is set to castle. It is thus the moment to exploit White's small loss of time. u ... fs !

12 exfs White can also play the similar 12

'ifb3+ (instead 12 ii.c4+ 'it>h8 13 e5 b5 14 ii.e2 f4 15 i.cl .Ub8 is better for Black) 12 . . . 'it>h8 13 exf5 (13 .Ud1 fxe4 14 ct:Jg5 Ci:Jxd4 15 'ifc4 - 15 Ci:Jf7+ .Uxf7 16 'ifxf7 ii.e6 17 'ifxe6 ct:Jxe6 18 .Uxd8+ .Uxd8 does not help White either - 15 . . . 'ifa5+ 16 .Ud2 ct:Jxe2 17 Ci:Jf7+ .Uxf7 18 'ifxf7 i.d7 was a disaster for White in W.Arencibia-P.Eljanov, Ubeda 2001) 13 . . . gxf5 14 .Ud1 f4 15 i.cl ctJa5 16 'ifa3 ii.e6 17 d5 ii.xd5 18 ii.b2 ( 18 Ci:Jg5 .Uf5 19 Ci:Je6 'ifd6 20 'ifxd6 exd6 21 ct:Jxg7 'it>xg7

22 i.xf4 ii.xa2 23 i.xd6 .Ue8 also leaves White in trouble) 18 . . . e6 19 ii.xg7+ 'it>xg7, which occurred in Y.Kruppa-K.Sakaev, Neum 2000, but Black has already taken the upper hand; for example 20 Ci:Jd4 .Uf6 21 0-0 Ci:Jc6. 12 ... 'ifas+!

This disruptive check highlights the downside to 8 h3 in the best possible way. 13 'it>f1

The alternatives are not even suffi­cient for equality :

a) 13 'ii'd2 'ifxd2+ 14 'it>xd2 ii.xf5 15 ii.c4+ 'it>h8 16 .Uhd1 (B.Grachev-V.Belov, Vladimir 2002) 16 . . . ii.xh3 17 gxh3 .Uxf3 with an edge for Black.

b) 13 ii.d2 'ifxf5 14 ii.c3 (or 14 0-0 Ci:Jxd4 15 Ci:Jxd4 ii.xd4 16 i.h6 as in T.Halay-A.Van Weersel, Vlissingen 2005, and now 16 . . . .Ud8 again leaves White worse) 14 . . . i.e6 15 0-0 .Uad8 16 .Ub1 ( 16 'ifa4 i.d5 17 .Uad1 a6 18 .Ud2 .Ud6 is also quite good for Black) was played in A.Korobov-A.Goloshchapov, Ordzhonikidze 2001 . Now simple and good was 16 . . . ii.xa2 17 .Uxb7 ii.d5 in­tending . . . e5, with an edge.

1 5

Play t h e Griinfe ld

13 • • . i.xfs 14 'ii'b3 + 'it>h8 15 l:td1 e6 16

lbgs 'ii'b4 17 'ii'xb4 lbxb4 18 g4 ..tc2 19 l:tc1 :l.ae8

Black has at least equalized, N.Kor­niushin-V.Yandemirov, Tomsk 2001 .

C) 8 i.bS+

Unlike after the 7 .i.bS+ of Chapter Five, Black cannot play . . . c6 and so should opt to block with his knight. Instead an exchange of bishops on d7 promises good equalizing chances, but also robs Black's position of much of its dynamism. 8 ... lbd7 9 0-0 0-0

10 i.gs

1 6

White has tried several alternatives, without coming close to gaining an advantage:

a) 10 .l:!.bl 'i!Vc7 1 1 'iWb3 (or 1 1 :tel, as in L.Oll-I.Sokolov, Wijk aan Zee (rapid) 1993, and now Black is fine after l l . . . .l:!.d8 12 .i.a3 b6) l l . . .lbb6 12 .i.a3 .i.e6 13 dS .i.g4 14 lbd2 (R.Damaso­L.Ftacnik, Lisbon 2000) 14 . . . .l:!.ad8 IS h3 .i.c8 intends . . . e6 gives with good play.

b) 10 a4 'i!Vc7 11 .i.gS (or 11 'i!Ve2 lDf6 12 h3 b6 13 dS, P.Van der Sterren­L.Gutman, Wijk aan Zee 1987, and now 13 . . . .i.b7 prepares the standard idea of . . . :ad8 and . . . e6) 11 . . . lbf6 12 :el .l:!.d8 13 'i!Vb3 i.e6 14 dS i.g4 IS lbd2 h6 16 i.h4 lDhS 17 f3 i.c8 18 i.f2 lbf4 19 .i.fl b6 20 aS l:tb8 21 axb6 axb6 22 .l:!.abl .i.d7 gives Black the initiative, A.Shirov­G.Kamsky, Manila Olympiad 1992.

c) 10 .i.d3 cxd4 11 cxd4 lbcS 12 i.c2 i.g4 13 i.e3 (S.Mohr-J .Van Mil, Buda­pest 1991 ) 13 . . . .i.xf3 14 'ii'xf3 .i.xd4 IS :!.adl eS with advantage to Black.

d) 10 .i.a3 b6 11 .i.c6 l:tb8 12 eS 'i!Vc7 13 'i!Va4 a6 14 i.e4 was seen in Buna­kov-V.Timofeev, Kaluga 2000, and now 14 . . . bS IS 'i!Vc2 aS is a logical way for Black to pursue his initiative.

e) 10 i.e3 lbf6 1 1 eS lbdS 12 l:tcl .i.g4 again leaves Black on top. 10 ... h6!

Forcing the bishop to make a deci­sion. 11 ..te3

Instead 1 1 .i.h4 lbf6 12 .l:!.el lbhS! 13 :tel (L.Alburt-L.Ftacnik, Hastings 1980/81 ) 13 . . . i.g4 14 dS gS IS .i.g3 lbxg3 16 hxg3 e6 is equal according to Ftacnik, while 1 1 i.f4 was tried in

Exch a n g e Va r iat io n : I n tro d u c ing 7 ti:Jj3 c5

A.Huzman-E.Sutovsky, Pula 2000, and now 1 1 . . .tt:lf6 12 l::te1 i.g4 gives Black the initiative. 11 ... tt:lf6 12 ii.d3 tt:lg4

Now White must allow the ex­change of his important dark-squared bishop, although the exchange does strengthen his centre. 13 es tt:lxe3 14 fxe3 i.e6 15 i.e4 i.ds 16 i.xds

Instead 16 �d3 cxd4 17 exd4 (infe­rior was 17 cxd4 �d7 18 l::tfcl l::tfc8 19

i.xd5 �xd5 with an edge for Black in A.Graf-H.Odeev, Beirut 2000) 17 . . . l::tc8 18 i.xd5 �xd5 19 l::tab1 l::tc7 is at least equal. After the text, 16 . . . �xd5 17 �3 �e4 leads to a complicated position in which Black's long-term chances are preferable; he has the superior minor piece and the more flexible structure.

Conclusion

8 i.e2 is rightly considered inferior; after all, if this line was good for White, the entire Griinfeld would be in trou­ble! Both 8 h3 and 8 i.b5+ contain a drop of poison and should be met in an active and accurate way if Black wishes to fight for an opening advantage: 8 h3 represents a slight loss of time, which is best exploited with a timely . . .f5 strike, while 8 i.b5+ is best countered by keeping all the pieces on and aiming to take advantage later of the exposed position of the bishop.

1 7

C h a pte r Two I Exc h a n ge Va riat io n : The Crit ica l 8 Mb1

1 d4 tLlf6 2 c4 g6 3 tLlc3 ds 4 cxds tLlxds 5 e4 tLlxc3 6 bxc3 .ltg7 7 CLlf3 cs 8 l:!.b1

This is one of the most critical tests of the entire Griinfeld. White removes his rook from the long diagonal, thereby preparing to meet the pressure against his centre with the d4-d5 ad­vance, regardless of the fact that the c3-pawn may be left en prise. This line was introduced in the Eighties, most notably by Viacheslav Eingorn, but became really popular towards the end of that decade, thanks to the efforts of Boris Gelfand and Alexander Khalif­man, who have both scored numerous victories with this system. Later on, the white side found another prominent supporter in Vladimir Kramnik.

This whole system has been deeply researched for many years and the the­ory has advanced to alarming propor­tions. Indeed, the black player is well advised to prepare for it meticulously. To help a little, I have suggested that Black avoids the risky and heavily

1 8

theoretical main lines . Instead we will explore a system that is both position­ally fully sound and has scored well in practice. This does not mean that Black mustn't prepare well for 8 J:!.b1, but rather that Black's survival depends much less on long forced lines and much more on active positional play. 8 . . . 0-o 9 .lte2

Instead 9 .lte3?! is inconsistent, as White wants to push d4-d5, not protect d4: 9 . . . .ltg4 10 'i¥d2 (10 l:!.xb7 .ltxf3 1 1 gxf3 CLlc6 gives Black pressure) 10 . . . cxd4 1 1 cxd4 CLlc6 12 d5 .ltxf3 13 gxf3 CLle5 14 .lte2 'i:Vc8 gives Black the initiative, N.Rashkovsky-M.Ghinda, Lvov 1981.

After 9 .lte2, 9 . . . cxd4 10 cxd4 'i:Va5+ 1 1 .id2 'i:Vxa2 12 0-0 leads to the main tabiya of the 8 J:!.b 1 system, which has been debated for many years and in countless games. Black has tried nu­merous schemes, without ever really managing to extinguish White's initia­tive. Indeed, the positional aspects of White's compensation (a strong centre,

Exch a n g e Va r ia t io n : T h e Cr it ica l 8 I:.b 1

a lead in development, the possibility success when they result in a perpetual of creating a powerful central passed check or repetition. pawn, the lack of good squares avail­able to Black's minor pieces and the exposed black queen) guarantee lasting pressure and force Black to tread very carefully.

Moreover, the immense amount of theory involved in these lines, as well as the forcing nature of most of the en­suing variations, makes the whole ven­ture of capturing the a2-pawn quite unappetizing for Black. I have a healthy respect for material in general, but in this particular case I believe that, from a practical viewpoint at least, the pawn just isn't worth taking!

Another natural option is 9 . . . tt:Jc6, but this falls in with White's plans. Af­ter the standard sequence 10 d5 tt:Je5 (10 . . . i.xc3+ 1 1 �d2 �xd2+ 12 'iYxd2 is a very risky pawn-grabbing operation) 1 1 tDxe5 i.xe5 12 'iYd2! e6 13 f4 i.c7 14 0-0 exd5 15 exd5 i.a5 it seems to me that Black is skirting the precipice .

White has many dangerous con­tinuations, while Black's defences are often considered to be crowned with

Therefore, 9 . . . CDc6 is not a good choice for the practical player!

After 9 �e2 there is, however, a third option: 9 ... b6!

The reader should note that this is not a rare sideline. Indeed, it has actu­ally become the main choice of not just Peter Svidler, but also several very strong grandmasters who regular em­ploy the Grunfeld, such as David Navara and Pavel Eljanov. Further­more, Garry Kasparov has tried it on occasion in the past, and so has Boris

1 9

Play t h e G r u nfe ld

A vrukh, the Israeli GM who has con­tributed much to the theory of 8 .Ub1 from both sides of the board. Further­more, I am fully convinced that this system will continue to figure promi­nently in modern high-level praxis and, due to its solid and sound nature, will survive any attempts by White to refute it!

Note too that with 9 . . . b6, Black opts for natural development: the light­squared bishop will usually go to b7 or a6, the knight to d7, the rooks to the centre and Black may perhaps play . . . e6. In this way he generally succeeds in keeping White's centre at bay, aim­ing either to attack it later with his pieces or to break it up with . . . f5, taking advantage of his good light square control . The resulting struggle is more positional than tactical and also fits in nicely with the Griinfeld' s general spirit. 100-0

By far White's main move. Instead: a) 10 .l¥.e3? ! again does not fit in

with White's strategy. Black continues with 10 . . . .1¥.b7 and now:

a1) 11 e5 (this advance surrenders the central light squares; a typical prob­lem for White in this variation) 1 l . . .cxd4 12 cxd4 'Lla6! has the simple plan of transferring the knight to d5 via c7.

Now 13 'iVd2 (or 13 h4 'Llc7 - I don't like the weakening 13 . . . h5 and also see no reason to waste time on it - 14 h5 'Lld5 15 hxg6 hxg6! - maintaining struc­tural integrity; the attack down the h­file is hardly dangerous - 16 .l¥.h6 .l¥.xh6

20

17 .Uxh6 Wg7 18 'iVd2 'Llc3 and White's position is on the verge of collapse) 13 . . . 'Llc7 14 h4 'Lld5 15 h5 'Llxe3 16 fxe3 (16 'iVxe3? ! 'iVd5 gives Black a clear edge, due to White's loose pawns on a2 and d4) 16 . . . e6 17 .l¥.d3 .Uc8 left Black better in I.Nemet-V.Korchnoi, Switzer­land 1985; White's h4-h5 plan has turned out to be a waste of time.

a2) 1 1 'iVd3 .l¥.a6 12 'iVd2 'iVc8 13 0-0 (preparing to relieve the some of the pressure with .l¥.h6) 13 . . . .1¥.xe2 14 'iVxe2 'iVa6

15 'iVxa6 (or 15 'iVd2 .Ud8 16 .l¥.h6 -16 .l¥.g5 'Lld7 17 .l¥.xe7 .Ue8 18 .l¥.d6 .Uxe4 19 dxc5 'Llxc5 is equal - 16 . . . .1¥.h8 17 'iVf4 'iVxa2 which is unclear according to Golod; this is an arguably a better ver­sion of the standard pawn sacrifice for Black since a pair of minor pieces has been exchanged and White's central initiative is still to get rolling) 15 . . . 'Llxa6 16 .Ufcl .Uac8 (16 . . . cxd4?! 1 7 cxd4 .Ufc8 18 d5 gives White a slight edge, as 'Lld4-c6 is coming) 17 d5 (intending to strengthen the centre with c4) 17 . . . f5 ! and Black has good counterplay, M.Roiz-V.Golod, Beer Sheva 2001 .

b) 10 h4 is an aggressive thrust, but again rather inconsistent with 8 .Ubl . Black can now prevent h5 with the natural 10 . . . i.g4, leading to another divide:

b1) 1 1 Wfl cxd4 ( 1 l . .:�c7 is another option) 12 cxd4 tt:lc6 (Rowson) is a simple way to a good position; for ex­ample, 13 d5 tt:la5 14 h5 i.xh5 15 tt:lg5 i.xe2+ 16 'iVxe2 h6 17 tt:lf3 h5 with an edge for Black.

b2) 11 i.e3 tt:lc6 (Black has switched back to the standard Grunfeld scheme of development and White is under pressure) 12 e5 cxd4 13 cxd4 .Uc8 ! ? (also good is 13 . . . 'iVd7 14 h5 i.xh5 - but not 14 . . . i.xf3? ! 15 i.xf3 .Uad8 16 hxg6 hxg6 17 .Uh4! and White consolidates his centre, A.Vaisser-S.Mohr, San Bernar­dino 1989 - 15 d5 tt:la5 with an edge; White's central pawn duo is weak) 14 h5 i.xh5 and I find it rather unlikely that White can fully justify his opening play.

c3) 11 tt:lg1 i.xe2 12 tt:lxe2 sees White insist on his attacking plan, but one can only take so many liberties in the opening phase.

Exc h a n g e Va r iat io n : T h e Cr it i c a l 8 '!:.b 1

S.Krivoshey-F.Vallejo Pons, German League 2005, continued 12 . . . 'iVd7! (a strong and accurate move, threatening . . . 'iVg4, whereas both 12 . . . tt:lc6 13 d5 tt:le5 14 f4 and 12 . . . cxd4 13 cxd4 'iVd7 14 d5 give White a slight plus) 13 i.e3 (13 d5?! 'iVg4! is problematic for White, but he might try 13 f3 when 13 . . . tt:lc6 14 dxc5 - 14 d5 tt:le5 15 f4 tt:lc4 16 'iVd3 'iVa4 appears OK for Black - 14 . . . 'iVc8 is un­clear according to Vallejo, but White's position has certainly been severely weakened) 13 . . . tt:lc6 14 'iVa4 (or 14 h5 .Uad8 15 hxg6 hxg6 16 d5 'iVg4 17 'iVc2 tt:le5 18 'lt>fl f5 ! and White suddenly finds himself facing a powerful attack) 14 . . . .Ufd8 15 d5 tt:la5 16 'iVxd7 .Uxd7 when Black was better, and he can also consider 15 . . . tt:le5 !? 16 'iVxd7 .Uxd7 with an edge.

Returning to 10 0-0: 10 ... i.b7

The main move, but there is one worthwhile alternative, namely 10 . . . 'iVc7! ? .

The idea of this move order is to avoid the dangerous pawn sacrifice 10 . . . i.b7 1 1 d5 and to transpose to Line

21

Play t h e G r iJ nfe ld

B1 with 1 1 'i&'d3 .l¥.b7. White's only really independent option after 10 . . . 'i&'c7 is 1 1 .l¥.g5 .l¥.b7 12 'i&'d3 h6! (it is important to kick the bishop from its active post, so as to play . . . tt:ld7) and now:

a) 13 .l¥.e3 leads to very similar posi­tions to Line B1; for example, 13 . . . .l:i.d8 14 d5 e6 (14 . . . e5? is a clear misunder­standing: 15 tt:ld2! tt:lc6 16 tt:lc4 .l¥.a6 17 'i&'d2 .l¥.xc4 18 .l¥.xc4 tt:la5 19 .l¥.a6! �h7 20 f4 and White was much better in E.Fomichenko-S.Beshukov, Novoros­sijsk 1995; Black should be playing for . . . e6 instead of blocking the position which here usually allows White to seize the initiative) 15 c4 tt:ld7 with typically complex play.

b) 13 .l¥.h4 is more persistent, but Black has a strong counter with 13 . . . cxd4 (and not 13 . . . tt:lc6? when 14 .l¥.g3! causes a lot of problems) 14 cxd4 'i&'f4!, highlighting a lack of coordina­tion in White' s camp. P.Haba-G.Feher, Zalakaros 2000, continued 15 d5 .l¥.a6 16 'i&'d1 'i&'xe4 17 .l¥.xa6 tt:lxa6 18 .l:i.e1 'i&'f5 19 .l¥.xe7 (now the main question is: how strong is the white d-pawn?)

2 2

1 9 . . . .l:i.fc8 (as Haba points out, 1 9 . . . .l:i.fe8 20 d6 tt:lc5 21 tt:ld4 'i&'d7 22 tt:lb5 tt:le6 is also possible; it is very hard for White to break the blockade on d7 and his pieces are about to be pushed back) 20 d6 (20 'i&'a4 tt:lc5 21 .l¥.xc5 .l:i.xc5 is fine for Black) 20 . . . tt:lc5 21 tt:ld4 and now 2l . . .'i&'d7? ! 22 h4! h5 23 tt:l£3 'i&'f5 24 .l:i.e3 ! gave White a dangerous initiative. Black should prefer 21 . . . .1¥.xd4! (eliminating the knight and thereby ensuring the viability of the blockade on d7) 22 'i&'xd4 tt:ld7 23 .l:i.bcl .l:i.c5 (Haba) when he has everything under control.

Returning to 10 . . . .1¥.b7:

Now White has two options:

A: 11 ds B: 11 'i&'d3

There are also a couple of rare al­ternatives :

a) 11 ii.d3 ! ? (White sacrifices a cen­tral pawn for some dark square play, but it seems insufficient to me) 1 l . . .cxd4 12 cxd4 .l¥.xd4 13 tt:lxd4 'i&'xd4 14 .l¥.b2 'i&'d6! (the correct retreat as

Black must play . . . e5 at some point to blunt the b2-bishop; White will then aim to remove the e5-pawn with f4 and so Black needs to secure as much dark square control as possible) 15 'i¥d2 tt:Jc6 and soon . . . e5 will follow, while Black can even play . . . tt:Jd4 later. White has some compensation, but I don't feel it is especially threatening.

b) 1 1 e5? ! is, as usual in this system, mistaken: l l . . . cxd4 12 cxd4 �d5! (im­mediately seizing control of some criti­cal squares, while preparing . . . tt:Jc6 without blocking the active bishop)

13 'i¥a4 tt:Jc6 14 �e3 'i¥d7 15 'i¥a3 (Black has full central control and is now ready to strike at the advanced e5-pawn) 15 . . . f6! 16 exf6 exf6 17 .Ufd1 .Uad8 and Black was obviously for preference in G.Michelakis-J.Rowson, Yerevan Olympiad 1996. It' s worth continuing with this game for a while, as Black's play was very instructive: 1 8 .Ubcl 'i¥d6! 19 'i¥a4 .Uf7! 2 0 h4 tt:Je7 2 1 h5 CLJf5 22 hxg6 hxg6 23 .Uc3 �f8 ! 24 �c4 tt:Je7! 25 'i¥c2 .Uh7 26 'i¥e4 �xc4 27 .Uxc4 'i¥d5! 28 'i¥xd5+ tt:Jxd5 and Black's ad­vantage had continued to grow.

Exc h a n g e Va r ia t io n : T h e Cr it ica l 8 'l:.b 1

A) 11 d5! ? This aggressive pawn sacrifice is

fully in the spirit of the 8 .Ub1 system and it is surprising that it has only re­cently attracted serious attention. Black should meet it with great accuracy and care. The best plan is to accept the pawn and then to immediately retreat the bishop back to g7. After that, Black should aim to exchange the light­squared bishops with . . . 'i¥c8 and . . . �a6. The expediency of this plan cannot be emphasized enough; any delay may result in White breaking through with d6 or e5-e6, creating serious problems both in the centre and for the black king. 11...�XC3

Now 12 �g5 �g7 13 'iVb3 'i¥d6 is fine for Black, who can continue . . . CLJd7, and so White usually chooses from:

A1: 12 'i¥b3 A2: 12 �c4

A1) 12 'i¥b3

Dangerous, but not objectively

2 3

Play t h e G riinfe / d

promising. Black should now proceed with the aforementioned plan. 12 ... �g7 13 �f4 'i¥c8!

14 .l:i.fe1 Black is better after 14 e5 e6 15 d6 (or

15 dxe6 'i¥xe6! 16 �c4 'i¥f5 17 'i¥e3 CLJc6 18 e6 'it>h8 with advantage - A vrukh) 15 . . . ct:Jd7 16 .l:i.fe1 a6. Note that such a blockade of White's central pawn duo is a common feature of this line. Should it prove successful, Black will gain the time necessary to consolidate and start exploiting his queenside majority. 14 ... �a6 15 es

V.Dobrov-P.Eljanov, Internet (blitz) 2003, deviated with 15 �xa6 'i¥xa6 16 e5, but after 16 . . . CLJd7 17 �g5 (both 17 .l:i.e4 h6 and 17 CLJd2 'iVb7 18 ..l:i.bd1 .l:i.ac8 are no improvement for White) 17 . . . ..l:i.ae8 18 .l:i.bd1 (or 18 ..l:i.e4 h6, solv­ing Black's problems and leaving him with the advantage) 18 . . . c4 19 'i¥c2 CLJc5 20 d6 exd6 21 exd6 .l:i.xe1+ 22 CLJxe1 ..l:i.e8, Black clearly had the upper hand. 15 ... �xe2 16 .l:i.xe2 e6!

The final link in Black's plan: he in­vites d5-d6, after which White cannot create any serious threats.

2 4

17 d6 17 ct:Jg5? doesn't work out well:

17 . . . h6 18 dxe6 (or 18 ct:Je4 exd5 19 'i¥xd5 ct:Jc6 20 CLJd6 'i¥d7 with a clear advantage for Black, as Avrukh has analysed) 18 . . . hxg5 19 exf7+ ..l:i.xf7 20 e6 ..l:i.e7 21 �xg5 ct:Jc6 22 �xe7 CLJxe7 and Black was much better in V.Belov­P.Eljanov, Internet (blitz) 2003. 17 ... CLJd7

As Avrukh notes, Black is better.

A2) 12 �c4 �g7! A prudent retreat as Black avoids

any tempo-gaining moves by the white queen.

13 'iVd3 White hampers . . . i.a6, but he might

prefer: a) 13 'iVe2 'iVc8! 14 i.g5 (Black is fine

too after 14 i.b2 i.xb2 15 .Uxb2 tt::ld7, intending . . . 'iVc7 and . . . tt::lf6) 14 . . . f6 ! (time is of primary importance ! ) and:

a1) 15 d6+? Wh8 (also possible is 15 . . . e6!? 16 i.f4 tt::lc6 17 .Ufd1 'iVd7 -Sutovsky) 16 dxe7 .Ue8 17 i.h4 .Uxe7 18 e5 tt::lc6 (Sutovsky) leaves Black better.

a2) 15 i.f4 i.a6 16 .Ufd1 (16 d6+ i.xc4 17 'iVxc4+ e6 18 e5 tt::lc6 is nothing for White) 16 . . . i.xc4 17 'iVxc4 Wh8 18 h3 tt::ld7 seems quite comfortable for Black.

a3) 15 i.h4 i.a6 16 .Ufe1 i.xc4 17 'iVxc4 Wh8 18 a4 tt::la6 and the knight will settle on b4, while White's initia­tive is dying; Black is for preference.

b) 13 i.f4 i.a6 (of course there's no need for . . . 'iVc8 here) 14 'iVe2 (14 i.xa6 tt::lxa6 15 'iVe2 tt::lc7 16 .Ubd1 a6 is fine for Black) 14 . . . 'iVc8! sees Black correctly retain the tension between the bishops so as to restrict the movements of the white queen. Now:

b1) 15 e5 i.xc4 16 'iVxc4 (B.Gelfand-A.Mikhalchishin, Portoroz 2001)

Exc h a n g e Va r ia t io n : The Cr i t ica l 8 �bl

16 . . . 'iVa6 17 .Ub5 (Black gains the advan­tage after 17 'iVe4 tt::ld7 18 e6 tt::lf6 19 exf7+ .Uxf7 20 'iVe6 'iVc8! - Mikhal­chishin) 17 . . . tt::ld7 seems OK for Black.

b2) 15 i.g5 ! ? i.xc4 16 'iVxc4 .Ue8 is also fine for Black.

b3) 15 .Ubd1 (a suggestion of Mik­halchishin's) 15 . . . i.xc4 16 'iVxc4 'iVa6 and Black is for preference; he can play . . . b5 and . . . c4, as well as . . . tt::ld7. 13 ... 'iVc8!

14 ..tgs Instead 14 e5 i.a6 15 i.g5 i.xc4 16

'iVxc4 .Ue8 17 .Ufe1 'iVf5 ! ? i s good for Black, as his queen is very active (but not 17 . . . 'iVa6 18 'iVh4 when White can attack), and similarly 14 i.b2 i.xb2 15 .Uxb2 i.a6 16 i.xa6 (or 16 tt::le5 e6 17 tt::lg4 exd5) 16 . . . 'iVxa6 does not seem too dangerous. 14 ... .Ue8 15 es

White has managed to advance his central pawns side-by-side to the 5th rank and has decent compensation for the sacrificed pawn. Black's position looks uncomfortable, but appearances can be deceptive! There is actually no easy way for White to break through

2 5

Play t h e G r iJ nfe /d

Black's defences, while Black will force a useful exchange with . . . j_a6. 15 ... j_a6 16 �bd1

16 �fe1 j_xc4 17 'ifxc4 'iff5 should be OK for Black. 16 ... j_xc4 17 'ifxc4 'ifa6

Probably best, whereas White re­tains good compensation after the al­ternatives 17 . . . 'iff5 18 �fe1 'Lld7 19 e6 and 17 . . . 'Lld7 18 �fe1 e6 19 d6 h6 (19 . . . 'ifc6 20 'iff4) 20 j_e7 'ifc6 21 h4 b5 22 'iff4 c4 23 h5 (Haba). 18 'ifh4 'ifb7

Black is wise to avoid the greedy 18 . . . 'ifxa2 19 j_xe7 'Lld7 20 �fel . 19 j_h6!

Facilitating White's main idea in these positions which is to launch an attack after an e6-advance. Black must then defend accurately, but I believe that he can. 19 ... 'Lld7 20 e6 fxe6

21 'Llgs! 'Llfs! Haba shows with some nice varia­

tions how Black loses after 21 . . .j_f6? 22 dxe6 'Lle5 23 f4 'Llc4 24 j_g7! ! j_xg5 25 fxg5 'Lle3 26 �f2 'ifc6 (or 26 . . . 'Llxd1 27 'iff4! �xg7 28 'ife5+ 'it>g8 29 �f7) 27

26

j_a1 ! �ed8 28 �e1 'ifxe6 29 'ifh6 'Llf5 30 �xe6 'Llxh6 31 gxh6. 22 dxe6 j_xh6 23 'ifxh6 �ed8

Black has both defended well on the kingside and retained his extra pawn. Indeed, White's attacking wave has been brought to a temporary halt. This position was reached in P.Haba­J.Banas, Austrian League 1997, and it' s worth noting that Haba himself feels that Black's resources are more than adequate, as illustrated by one of his variations: 24 f4 'ifc7 25 f5 �xd1 26 �xd1 'iff4 27 fxg6 hxg6 28 �fl 'ifd4+ 29 �h1 'ifh8!, exchanging queens and leaving Black clearly better.

B) 11 'ifd3

With this move White protects both the e4- and the c3-pawns, thereby pre­paring to push d4-d5 at the first oppor­tunity. We will now consider two op­tions for Black.

81: 11 ... 'ifc7 82: 11 ... j_a6

81) 11 .. .'iVc7!? Aiming for a complicated game,

rather than exchange pieces as Black does in B2. 12 ds

White's best try, whereas 12 .1i.g5 h6 transposes to the notes to Black's lOth move (with lO . . .'iVc7) and shouldn't be too troubling. 12 ... tt::ld7

13 .Jtgs Another option is 13 'iVc2 tt::le5 14

tt::lxe5 'iVxe5 15 c4 e6 16 �b2 'iVc7 17 .1i.xg7 Wxg7 which is approximately level; White has established a protected passed pawn on d5, but Black can eas­ily blockade it on d6 and undermine its support with . . . b5. I feel that Black has no problems here, as was shown by a recent high-level example: 18 a4 exd5 19 exd5 .l:i.fe8 20 .1i.d3 .1i.c8 21 a5 .l:i.b8 22 axb6 axb6 23 'iVc3+ 'iVe5 24 'iVxe5+ .l:i.xe5 25 .l:i.b3 .1i.f5 with complete equality in E.Bacrot-D.Navara, European Team Ch., Gothenburg 2005. 13 ... e6

Black should avoid 13 . . . e5? 14 tt::ld2!, but 13 . . . .l:i.fe8 !? 14 .l:i.bd1 tt::le5 is possible;

Exc h a n g e Va r ia t io n : T h e Cr i t ica l 8 J:.b 1

Black was then better after 15 .1i.f4?! tt::lxd3 16 .ixc7 c4 17 tt::ld2 tt::lb2 18 .l:i.cl .l:i.ac8 19 �g3 b5 in F.Handke­A.Timofeev, Halkidiki 2000 . 14 .l:i.bd1

Instructively 14 c4 should be met by 14 . . . e5! (Turov) . Note the essential dif­ference with an earlier . . . e5, namely that the white knight no longer has access to the c4-square, from where it controls both d6 and e5. After 14 c4 e5, Black can maintain a dark square blockade before slowly preparing . . . f5 with good prospects, especially since White's knight lacks a good role. 14 ... tt:Jes

15 'iVd2 Instead 15 tt::lxe5? 'iVxe5 16 �e7 .l:i.fb8

simply costs White a pawn. 1S ... tt::lxf3+

Another interesting idea is 15 . . . 'iVd6!?, increasing the pressure against d5. Then, for example, 16 .1i.h6 .1i.xh6 17 'iVxh6 tt::lxf3+ 18 �xf3 exd5 19 exd5 .l:i.fe8 leads to equality. 16 .1i.xf3 'iVes 17 �e7 exds!

The simplest path to equality. In­stead 17 . . . .l:i.fe8 18 d6 'iVxc3 19 'iVg5 h6

2 7

P lay t h e G riinfe ld

20 'Yi'h4 g5 21 'Yi'hS occurred in A.Lesiege-M.Turov, Montreal 2000, and now Black should play 21 . . .'i¥e5! ? with an unclear position, but I don' t see any reason to take this risk.

After the text, Turov has analysed 18 i.xf8 'it>xf8! (incorrect is 18 .. J:hf8 19 exd5 'i¥xc3 20 'i¥f4 'i¥e5 since 21 'i¥a4! leaves Black facing some serious prob­lems) 19 exd5 'i¥xc3 20 'iVf4 'iVeS 21 'i¥h4 (21 'i¥a4 'i¥d6 leaves Black in con­trol) 21 . . .h5 22 d6! (White has to really prevent the blockading and strong . . . 'i¥d6) 22 . . . �xf3 23 gxf3 �f6 24 'i¥a4 .l:i.d8.

Here Black has promising compen­sation; he enjoys some positional ad­vantage and the d6-pawn should be rounded up in the near future. White may capture on a7, but his weakened kingside means that he is the only one in any real danger of losing the game.

82) 11 ... �a6 This has always been considered

the main line. Black relieves some of the congestion in his camp with the exchange of bishops, although on the

28

other hand he does surrender some dynamism. White now lacks a com­fortable square for his queen and must choose between:

821: 12 'iVc2 822: 12 'i¥e3

821) 12 'iVc2 This retreat has attracted some re­

cent attention. White gets his queen out of the way of his dark-squared bishop.

12 ... cxd4! It makes sense to open the c-file in

order later to gain a useful tempo on

the white queen. 13 .i.xa6

White must interpose this exchange, because 13 cxd4 'ii'c8! emphasizes the loose state of his position and at least equalizes for Black. 13 ... ltJxa6 14 cxd4 1i'd7

Black is very comfortable here. He can centralize his rooks, perhaps bring his knight to e6 and play down the c­file, while White's centre cannot ad­vance easily. 15 .i.a3

Instead 15 .i.g5 l:tac8 16 1i'e2 ltJc7 17 l:lfd1 lLle6 18 .i.e3 'ii'a4 19 d5 l:tc2 20 l:ld2 l:lxd2 21 1i'xd2 ltJc5 was fully equal in J.Lautier-P.Svidler, Cap d' Agde (rapid) 2003. 15 ... l:tfe8!

Freeing the queen by protecting the e7 -pawn is the correct choice, whereas 15 . . . l:tac8?! 16 'ii'e2 lLlc7 17 d5 f5 1 8 l:tfd1 (Krasenkow) gives White an edge. 16l:tfd1

Or 16 1i'c4 Vi'b7 17 1i'a4 (J.Lautier­S.Shipov, Internet blitz 2004), and now 17 . . . l:tab8 intends . . . b5 and allows Black to take the upper hand.

Exc h a n g e Va ria t i o n : T h e Cr it ic a l 8 I:.b1

16 ... l:tac8 17 'ii'b3 e6 18 h4 'ii'b7

Black again has everything under control and can look to the future with confidence. In B.Lalic-V.Tomescu, Porto San Giorgio 2004, White pro­ceeded with his intended central ad­vance: 19 d5 exd5 20 l:txd5 and now Black can play 20 . . . l:tc3 with the initia­tive (instead of the game's 20 . . . l:r.cd8 21 l:lbd1 l:txd5 22 exd5 h6 23 d6 'ii'd7 24 1i'd5 l:le6 25 g3, although even here 25 . . . .i.f6 would have kept the position unclear).

822) 12 'ii'e3

Keeping the queen centralized and

2 9

P lay t h e G riinfe ld

away from trouble down an open c-file, but this does hamper the development of the cl-bishop. At this juncture Black again has a choice:

8221: 12 . . . e6 8222: 12 . . .'t/Vd7

8221) 12 . . . e6 This is a generally useful move,

preventing the critical d4-d5 advance. On the other hand, Black is not ready here to play . . . 'i¥a4, which means that White is justified in exchanging the light-squared bishops. 13 .txa6!

As White has no useful waiting move, he should embark on this ex­change in order to regroup his pieces. Instead 13 dxcS il.xe2 14 'i¥xe2 'i¥c7 is a typical pawn sacrifice. A.Khalifman­S.Kudrin, New York Open 1998, con­tinued 15 cxb6 axb6 16 il.e3 lLld7 17 il.d4 llfc8 when Black was ganging up against White's queenside pawns, while White lacked a constructive plan. 13 . . . tt:Jxa6 14 'i¥e2

3 0

Forcing the knight to retreat and clearing a path for the cl-bishop. White must play this immediately since oth­erwise Black can actively protect the knight with . . . 'i¥d7-a4; for example, 14

lld1 ? ! 'i¥d7! 15 'i¥e2 'i¥a4 16 il.gS cxd4 17 cxd4 llfc8 18 lld2 llc4 with an obvious initiative. 14 . . . lLlb8

Prudent, whereas 14 . . . 'i¥c8 is rather too passive and Black would prefer to have his queen' s rook on c8. 15 lld1 'i¥c7

Black now intends to develop ac­tively with . . . lL:lc6 and he appears to have a reasonable game, whereas the less justified 15 . . . cxd4 16 cxd4 'i¥d7 17 h4! 'i¥a4 18 il.g5lLlc6 19 'i¥d2llac8 20 hS favoured White in C.Barus-R.Dineley, Bled Olympiad 2002.

8222) 12 . . . 'i¥d7!

Probably Black's strongest option, and this purposeful move is certainly critical . Black waits for White to ex­change on a6 himself, while preparing both . . . lld8 and . . . 'i¥a4. White now faces an important decision between imme-

diately advancing with 13 d5, which involves a pawn sacrifice, and exchanging on a6.

82221: 13 �xa6 82222: 13 dS

Alternatively: a) 131ld1 'ifa4 14l:1d2 �xe2 151:1xe2

l:1c8 16 d5 CDd7 17 'ifd3 c4 18 'ifd2 lDc5 19 l:1b4 'ifa5 (preparing . . . CDd3) 20 l:1xc4 'ifa6 favours Black; the c3-pawn will fall after which Black's pieces are quick to invade the white position.

b) 13 h4 �xe2 14 'ifxe2 'ifg4!? high­lights the weakening aspect to White's 13th move.

c) 13 l:1b2 looks a little strange and doesn't contribute much to White's position:

13 . . . e6 14 h4 l:1d8 15 l:1d1 cxd4 16 cxd4 'ifa4 (Black already stands quite well) 17 l:1bd2 �xe2 ( 17 ... CDc6 18 �xa6 'ifxa6 19 h5 LDa5 !? is another idea, in­tending . . . LDc4) 18 'ifxe2 LDc6 with equality.

d) 13 dxc5 �xe2 (13 . . . bxc5? is a seri-

Exc h a n g e Variatio n : Th e Critica l 8 'i:Ib1

ous mistake: 14 l:1xb8! l:1axb8 15 �xa6 and White was winning in B.Gelfand­I.Sokolov, Oakham 1988) 14 'ifxe2 bxc5 15 'ifc4 'ifc6

1 6 e5 (or 1 6 l:1d1 ltJd7 17 �g5 lDb6 18 'ifd31:1fe8 19 'ife3 'ifa4 20 'ifxc5 'ifxe4 21 l:1e1 'ifc2 22 l:lbcl 'ifxa2 23 l:1xe7 h6 24 l:1xe8+ l:1xe8 25 �e3 l:1c8 and Black was better in F .Elsness-B.Avrukh, Is­tanbul Olympiad 2000) 16 . . . LDd7 1 71:1e1 e6! (more accurate than 17 . . . lDb6 18 'ifh4 'ifa4 19 'ifxe7 'ifxa2 20 l:1b5 with an edge for White in L.Van Wely­B.Avrukh, Neum 2000; Black is now ready to kick the white queen with . . . lDb6 and then to invade on the queenside) 18 �g5 ( 18 'ifh4 is now met by 18 . . . 1:1fb8! avoiding the exchange of dark-squared bishops with J.h6; then 1911a1 'ifc7 20 �h6lDxe5 21lDg5 �xh6 22 'ifxh6 £6 23 LDxe6 'ife7 24 ltJf4 'if£7 25

l:1ad1 l:1d8 left Black's pieces much more relevantly placed in P .Murdzia­A.Kovchan, Cappelle la Grande 2003) 18 . . . 11fb8 (18 . . . h6 19 �e7 l:1fc8 20 �d6

lDb6 also gave Black an edge in S.lvanov-K.Urban, Polish Team Ch. 2001 ) 19 l:1bd1lDb6 20 'ife2 'ifa4 211:1d6

3 1

P lay t h e G r u nfe ld

was seen in J.Markos-A.Greenfeld, European Team Ch., Leon 2001, and now Black can obtain the advantage with Krasenkow' s suggestion of 21 . . .lt:Jc4 !? 22 \i'e4 \i'xa2 23 .l:td7 lt:Jb6 24 .l:tc7 \i'a5; White's compensation is in­sufficient.

82221) 13 �xa6 lt:Jxa6

This exchange doesn't really bother Black, since he can now play . . . \i'a4. 14 \i'e2

Alternatively: a) 14 d5 \i'a4 transposes to a posi­

tion we will consider after 13 d5 \i'a4 14 Axa6 lt:Jxa6, but Black should probably prefer here 14 . . . lt:Jc7!? (but not 14 . . . c4 due to 15 Cbd4lt:Jc5 16 lbc6 with advantage) 15 c4 e5, intending to bring the knight to d6 and later play the move . . . f5.

b) 14 "li'd3 \i'a4 15 �g5 .l:tfe8 16 d5 c4 17 \i'e3 lt:Jc5 18 lt:Jd4 e6 allowed Black to seize the initiative in P.Hummel­D.Gormally, Oakham 2000; . . . c4 is not always a bad move, it must just be well timed!

c) 14 h4 '1i'a4 15 .l:tb3 (or 15 a3 cxd4

3 2

16 cxd4 .l:tfd8 17 �b2 .l:tac8 18 h5 .l:tc2 which gives Black good counterplay) 15 . . . Cbc7! (15 . . . \i'xa2? is a mistake in view of 16 .l:ta3 \i'c4 17 lt:Jd2) 16 dxc5 (16 d5? \i'xa2 17 c4 \i'c2 ! highlights the lack of coordination among White's pieces; after 18 lt:Jd2 �d4 19 \i'f4 lt:Je8 20 .l:th3 lt:Jg7 21 g4 f6 22 \i'h6 J:lf7 Black was much better in R.Lev-V.Golod, Israeli League 2002) 16 . . . \i'xa2 17 .l:tb2 \i'c4 18 cxb6 axb6 19 .l:txb6 \i'xc3 (or 19 . . . lt:Je6 !? with decent compensation) 20 \i'xc3 �xc3 is equal according to Golod. 14 ... lt:Jc7!?

I feel that this is Black's most accu­rate move order. He strives to improve the scope of his knight and will trans­fer it to either e6 or b5 . By keeping his queen on d7 for a move, he discourages both 15 �g5 (which would now be met by 15 . . . lt:Je6 with tempo) and 15 d5 (now met by an immediate 15 . . . e6). 15 .l:td1 \i'a4!

The consistent approach. Another interesting concept is the dynamic 15 . . . cxd4 16 cxd4 f5 ! ? which also seems quite acceptable; for example, 17 \i'c4+ ( 17 lt:Je5 \i'e6) 17 . . . e6! 18 �a3 .l:tfd8 19

.l:i.e1 fxe4 20 .l:i.xe4 tt:ld5 with a complex situation, but one in which Black stands fairly well . 16 Ji..gs .l:i.fe8

White's position suddenly appears quite loose. Indeed, 17 �d2?! (Black is also more than fine after 17 d5 Si.xc3 18 e5 f6, while 17 �e3 tt:lb5! 18 e5 .l:i.ad8 leaves White's centre under serious pressure) 17 . . . .l:i.ad8 18 Ji..h6 e5 ! is pow­erful; for example, 19 Ji..xg7 Wxg7 20 .l:i.e1 (20 �g5? exd4 21 e5 - 21 cxd4 tt:lb5 - 21 . . .tt:le6! 22 �f6+ �g8 23 �4 �xa2 24 cxd4 �d5 was excellent for Black in Y.Ulko-B.Avrukh, Moscow 2002) 20 . . . exd4 21 cxd4 cxd4 22 .l:i.b4 �aS 23 .l:i.xd4 (23 tt:Jxd4 tt:lb5! is strong) 23 . . . �xd2 24 .l:i.xd2 .l:i.xd2 25 tt:lxd2 tt:le6 (Krasenkow) when Black has an edge in the endgame.

82222) 13 dS This is of course the critical test, in­

volving yet another pawn sacrifice . 13 ... Ji..xe2!

Black must accept the offered pawn, as otherwise White will consolidate his centre. Thus 13 . . . �a4 is inferior: 14

Exch a n g e Va r iat io n : Th e Cr i t ic a l 8 I:. b 1

Si.xa6 tbxa6 (or 1 4 . . . �xa6 1 5 e5 ! ? �xa2 16 �e4 with the initiative Krasenkow) 15 �e2! tt:lc7 (15 . . . Ji..xc3? doesn't work in view of 16 .l:i.b3 Ji..b4 17 tt:le5 l:tac8 1 8 l:th3 with a crushing at­tack, E.Najer-S.Yuferov, St Petersburg 2000, but 15 . . . c4 ! ? may be possible) 16 c4 e5 17 .l:i.b3 tt:le8 18 Ji..b2 f6 (V.Diu­Shomoev, A.Krasnodar 2002) 19 �e3 prepares f4 and gives White an edge. 14 �xe2 .!txc3

Again both consistent and accurate, whereas 14 . . . e6? ! 15 c4 l:te8 16 Ji..b2 Ji..xb2 17 �xb2 exd5 18 exd5 was good for White in R.Vera-M.Ripari, Malaga 2001 . 15 Ji..h6

Alternatively, 15 .l:i.d1 Ji..g7 16 Ji..b2 (or 16 e5 e6 17 d6 tt:Jc6 intending . . . .l:i.ad8 and . . . f6, when Black will de­stroy White's central pawn chain) 16 . . . kxb2 17 l:txb2 e6! (breaking up White's centre just in time) 18 dxe6 �xe6 19 e5 tt:lc6 20 l:td6 �e8 21 �e4 .l:i.d8 22 .l:i.bd2 .l:i.xd6 23 .l:i.xd6 tt:ld8 and with the consolidating . . . tt:le6 and the active . . . �5 available, Black is better. 1S . . . Ji..g7

3 3

P lay t h e G r u nfe ld

A prudent retreat since 15 . . J1e8?! 16 l::tb3! allows White to switch all o f his pieces to a dangerous kingside attack. 16 ii.xg7 Wxg7 17 i::tfd1

White' s compensation has a sound positional foundation and Black must now continue actively. 17 .. .'iVa4! 18 l::tbc1 'Lld7 19 es l::tad8 20 l::tc4

White has also tried 20 e6 CDf6 21 exf7 (now care is required, as White threatens 'Llg5-e6) 2l . . .'i!Ve4! 22 'li'xe4 'Llxe4 23 l::te1 'Llf6 24 'Llg5 i::txd5 25 h4 l::txf7 26 'Llxf7 Wxf7, but at the end of the day Black was better in B .Gelfand­P.Svidler, Monaco (rapid) 2005. 20 • • . \i'a6

We've followed V.Mikhalevski­S.Kudrin, Minneapolis 2005. Black is

3 4

ready to respond to any further aggres­sion from White and may even play . . . £6 to clarify the situation in the centre. The position remains complicated, but I see no reason for Black to complain; he has an extra pawn and no immedi­ate danger is apparent.

Concl usion

The aggressive 8 l::tb1 i s one of the stern­est tests of the Gri.infeld. By adopting the 9 . . . b6 variation, Black obtains a solid and flexible position which offers good prospects of counterplay. Play is chiefly based on standard Grunfeld positional and tactical motifs, and many strong players employ the suggested reper­toire, which speaks volumes for its in­herent soundness.

Particular attention should be paid to the various versions of the d5-advance, sacrificing the pawn on c3. Black should definitely accept the offer and then play accurately to neutralize White's initiative. Do note that in these lines, general principles will not suf­fice; good preparation is necessary. However, this is much less so than in the extremely complex main lines of the 8 l::tb1 system.

C h a pte r T h ree I Exc h a n ge Va r iat io n : il.e3 Syste m s

1 d4 tt:Jf6 2 c4 g6 3 tLlc3 ds 4 cxds tt:Jxds 5 e4 tt:Jxc3 6 bxc3 kg7

There are two different versions of the ke3 system, depending on whether or not White plays ctJf3:

A: 7 ctJf3 cs 8 ite3 B: 7 ite3

A) 7 ctJf3 cs 8 ke3

With this system, White intends to develop his queenside first with Vi'd2

and .l:Ibl or .l:Icl, preparing to meet . . . ctJc6 with d5. In this way he manages to diffuse Black's standard scheme of applying pressure against his centre. Anatoly Karpov made very efficient use of this system (and especially the more flexible move order 7 ke3 - see Line B) in his time, while Vladimir Kramnik successfully flirted with it for a while.

The drawback to White's set-up is the fact that his king remains in the centre for a long time. Black's best way of exploiting this is with . . . Vi'a5, a very annoying move that often forces an exchange of queens on d2, after which White usually has to misplace a piece as he recaptures. This, allied to White's rather slow development, can leave Black with a dangerous initiative should White be just a little inaccurate. 8 ... Vi'a s! 9 Vi'd2

The other interpositions are less ef­fective:

a) 9 kd2 0-0 10 ke2 (10 kd3 ctJc6 1 1

3 5

Play th e G r u nfe ld

d5 tt:le5 12 tt:lxe5 �xe5 13 'iib3 'iic7 gives Black the initiative, as Krasenkow has analysed, and 10 'iib3 tt:lc6 1 1 �d1 'iic7 is similar) 10 . . . �g4 11 0-0 e6 (Black' s set-up is flexible and quite comfortable) 12 a4 (or 12 �b1 when 12 . . . 'W/xa2 is sound but allows a repeti­tion, and so Black might prefer 12 . . . �xf3 13 �xf3 tt:lc6 with pressure against d4) 12 . . . tt:lc6 13 �b1 (neither does 13 d5 exd5 14 exd5 tt:le7 solve White's problems) 13 . . . 'W/c7 14 d5 exd5 15 exd5 tt:le7 and White's position looks very uncoordinated, while 15 . . . tt:la5 was also fine for Black in A.Beliavsky­A.Adorjan, Baden 1980.

b) 9 tt:ld2 cxd4! 10 tt:lc4 (10 cxd4?! loses control of the position: 10 . . . tt:lc6 11 d5 tt:ld4! 12 �b1 �d7 13 �d3 �a4 14 'iicl 0-0 15 0-0 �ac8 and Black has a strong initiative) 10 . . . dxe3 ! ? (a promis­ing queen sacrifice) 11 tt:lxa5 �xc3+ 12 '<t>e2 �xa5 13 'W/a4+ (or 13 Wxe3 �b6+ 14 W£3 tt:lc6 with excellent compensa­tion - Adorjan) 13 . . . tt:lc6 14 fxe3 0-0 15 '<t>f2 f5 and Black's initiative is very dangerous. 9 . . . 0-0

3 6

The usual choice here has actually been 9 . . . tt:lc6, leading to some very complicated positions. Then Black's pressure against d4 prevents White, after an exchange on d4, from recaptur­ing on d2 with his knight, unlike after the text move. Doing so is certainly more comfortable for White than tak­ing with the king, but the resulting po­sitions are still fully acceptable for Black. I have analysed them in detail and feel that Black' s resources are fully adequate for equality and perhaps even more, especially should White play inaccurately. Two other positive features of 9 . . . 0-0 are that it enables a quick . . . �d8, a very significant gain in case White does recapture with his king on d2, and that the advance d4-d5 does not come with tempo, unlike after 9 . . . tt:lc6.

White must now decide where to place his rook:

A1: 10 �b1 A2: 10 �c1

A1) 10 �b1 This causes some problems after

9 . . . tt:lc6, as then 10 �b1 b6 is met by the annoying 1 1 �b5 . Here, however, it is absolutely harmless - another good reason to prefer 9 . . . 0-0 . 10 ... b6 11 �d3

Alternatively: a) 11 �b5 is always comfortably met

by 1 1 . . .'W/a4, when the queen is actively placed; for example, 12 dxc5 (or 12 �b3 �g4 13 �b5 'W/a5 14 tt:lg5 a6 15 �d3

'Llc6 16 h3 .id7 which favoured Black in V.lkonnikov-P.Carbonnel, Paris 1992; and both 12 .ie2 .ib7 and 12 .ih6 .ixh6 13 \i'xh6 .ig4 are no better) 12 . . . bxc5 13 �xeS .ia6 14 .id4 'Lld7 when Black has excellent compensation for the pawn in view of his very active pieces and White's undeveloped king­side.

b) 11 �cl 'Lld7 12 dS occurred in T.Hillarp Persson-E.Liss, Copenhagen 1996, and now I think that Black has very good play after 12 . . . \i'a4!? 13 \i'c2 \i'xc2 14 �xc2 fS 15 exfS gxfS . 11 ... 'Llc6 12 �bs

As usual, 12 e5? ! is too committal : 12 . . . �d8 13 �bS (13 0-0? cxd4 14 cxd4 \i'xd2 wins a pawn) 13 . . . \i'a4 14 �b2 (White collapses after both 14 0-0 .ia6 15 �b2 .ixd3 16 \i'xd3 cxd4 17 cxd4 'Llxe5 18 'LlxeS .ixeS and 14 \i'c2 \i'xc2 15 .ixc2 ka6 16 �b3 'LlaS 17 �a3 .ib7) 14 . . . .ia6 and Black enjoys the initiative, M.Rivas Pastor-V.Epishin, Dos Her­manas 1994. 12 ... \i'a4 13 �b2 .ia6 14 0-0 .ixd 3 15 \i'xd 3

Exch a n g e Va r i a t io n : i. e 3 Sys t e m s

I .Ben Menachem-E.Liss, Ramat Ha­sharon 1992, and now the simple 15 . . . �ad8 further increases the pressure on d4 and favours Black.

A2) 10 �c1 cxd4 11 cxd4 \i'xd2+ White now faces an important deci­

sion: recapturing with the king is fairly common, while doing so with the knight is an attempt to exploit the omission of . . . 'Llc6.

A21: 12 'it>xd2 A22: 12 'Llxd2

A21) 12 'it>xd2 �d8!

Already one of the aforementioned advantages of 9 . . . 0-0 over 9 . . . 'Llc6 comes to the fore. 13 �c7

The most aggressive, but White has several other options:

a) 13 d5?! fS doesn't help White at all .

b) 13 .ic4 'Llc6 14 dS?! is also prema­ture and, for example, 14 . . . 'Lla5 15

We have been following the game �hd1 'Llxc4+ 16 �xc4 fS ! 17 'it>e1 fxe4 18

3 7

P lay th e G r ii nfe ld

l:txe4 i..c3+ 19 �f1 i..f6 20 g4 b6 left White under great pressure in C.Hoi­K.Helmers, Esbjerg 1982.

c) 13 h3 is a bit slow: 13 ... l2Jc6 14 i..c4 (14 d5? e6 15 i..b5 exd5 1 6 i..xc6 dxe4+ 17 'iii>e2 bxc6 was winning for Black in H.Barrios-A.Ruiz, Bogota 1991 ) 14 . . . l2Jxd4 15 i..xd4 i.xd4 16 l2Jxd4 l:ixd4+ 17 'iii>e3 e5 and i t i s White who has to seek ways to equalize.

d) 13 i..d3 l2Jc6 raises the question of how White will defend d4.

K.Sakaev-R.Ruck, Panormo (rapid) 2002, continued 14 d5 (or 14 l::tc4 f5 15 exf5, B.Damljanovic-D.Antic, Subotica 2000, and now 15 . . . gxf5 16 ltc5 e6 17 i..c4 �f8 gives Black comfortable play) 14 ... l2Jb4 15 i..c5 (it's easy for White to err here; he was quickly struggling af­ter both 15 i..b1 ? e6 16 a3 exd5 ! 17 axb4 dxe4+ 1 8 'it>e2 exf3+ 19 �xf3 i..e6, in P.Kekki-J .Aijala, Tampere 2003, and 15 lthd1? e6 16 �e1 exd5 17 exd5 l2Jxa2 in T.Steiner-G.Trammell, Tulsa 2006), and now Black could have secured an edge with the simple 15 . . . i..h6+ 1 6 i..e3 i..xe3+ 17 �xe3 e6.

e) 13 i..b5 i.d7!? 14 i..xd7 (14 i..d3? !

3 8

l2Jc6 15 d5 l2Jb4! 1 6 i..c4 b5 1 7 a3 bxc4 18 axb4 e6 leaves White in trouble, a s the position is opening up for the black bishops) 14 . . . l2Jxd7 15 l:ic7 (or 15 'it>e2 l2Jf6 16 l2Jd2 l2Jg4 17 h3 l2Jxe3 18 fxe3, S.Trofimov-R.Swinkels, Belfort 2005, and now 18 ... e5 19 d5 f5 when Black has the initiative and White's central pawns are weak) 15 . . . l2Je5! (with this neat tactic, Black at least equalizes)

16 l2Jxe5 (or 16 ltxb7 l2Jxf3+ 17 gxf3 i..xd4 18 i..xd4 l:ixd4+ 19 �e3 lta4 with easy equality for Black in S.Pedersen­J.Rowson, Oxford 1 998) 16 ... .i.xe5 and now White must be accurate with 17 l:txb7 i..xd4 18 .l:r.hb1, a s he was in O.Ruiz Mata-F.Vera Pons, Santa Co­loma 2000, after which 18 . . . e5 is equal. Instead 17 l::txe7? ! i..xd4 18 l:!.xb7 l:iac8! 19 i..xd4 ltxd4+ 20 'iii>e3 l:ia4 (the black rooks are more active and the white king somewhat exposed) 21 lia1 (or 21 ltd1 ltc3+ 22 'iii>f4 ltxa2 23 lidS+ �g7 with an edge for Black in E.Stephan­V.Petzold, Bayern 2003) 21 . . .l:ic3+ 22 'it>f4 ltc2 gives Black the initiative; a very important factor in a double rook ending.

f) 13 �e1 escapes the pin, but fails otherwise to impress: 13 . . . e6 14 �c4 (or 14 �g5 f6 15 �e3 tt:lc6, intending . . . f5, which is fine for Black, as is 14 d5 exd5 15 exd5 ctJd7 and the d5-pawn will surely prove weak, especially with White's king's rook shut out of play) 14 . . . tt:lc6 15 �g5 (15 �d1 b6 16 �e2 �b7 created very unpleasant threats in T.Roussel Roozmon-S.Glinert, Rich­mond 2002) 15 . . . �d7 and White's prob­lems with the defence of d4 continue to persist. 13 . . . tt:lc6

14 d5 White hopes to create counter­

threats before Black manages to fully develop his pieces, but this strategy doesn't quite work out. The alterna­tives are less ambitious, but not really any better:

a) 14 �b5 tt:lxd4 15 tt:lxd4 �xd4 16 �xd4 �xd4+ 17 �e3 �b4 (17 . . . e5 18 �c4 equalizes immediately, but there's no reason to allow White to equalize so easily! ) 18 �c4 �g7 19 �d1 �g4 20 f3 �c8 and White is again seeking ways to equalize, T.Carvalho-L.Tavares da

Exc h a n g e Var ia t io n : .fi. e 3 Sys te m s

Silva, Brasilia 1985. b) 14 �d3 tt:lxd4 15 tt:lxd4 �xd4 16

�xd4 (White should prefer 16 �xe7 �f6 17 �c7 �e6 18 a4 �e5 with equal­ity, N .Stewart-D .Drewelius, Bonn 1998) 16 . . . �xd4 17 �xe7 �e6 18 �e3 �adS and Black had a powerful initiative in M.Kuhn-S.Grimm, German League 1991 . 14 .. . e6 15 �gs

There is no turning back for White now, especially since 15 �cl ? exd5 16 exd5 tt:lb4 loses a pawn. Instead, the other aggressive option, 15 tt:lg5, rather backfires after 15 . . . exd5 16 tt:lxf7 �d7! 17 �xd7 �xd7 when White is not too well coordinated and Black remains the more active even after 18 exd5 (or 18 tt:ld6 �e6! 19 tt:lxb7 dxe4 with the initiative) 1 8 ... �xf7 19 dxc6 �xc6 20 �c4+ �e8. 15 .. .f6 16 �c1

Instead 16 �h4 exd5 17 e5 �h6+ wins, as . . . g5-g4 and then .. . fxe5 fol­lows. 16 ... exds 17 exds tt:lb4 18 �c4 bs 19 �b3

White had to avoid losing with both

3 9

Play t h e G r u nfe ld

19 .i.d2 lt::la6 and 19 .i.xb5 fxg5, but the text, as seen in G.Kotlyar-L.Ftacnik, Reno 1991, still leaves Black with a clear advantage after the continuation 19 .. .<it>h8 20 lLle5 Iif8 21 .i.d2 fxe5 22 .i.xb4 .l::txf2.

A22) 12 lt::lxd2

An important tabiya for both the .i.e3 system in general and especially our preference for 9 . . . 0-0. The position may look too simple, but it's actually far from that! Indeed, both sides have good chances to outplay the opponent: White by seizing more space in the cen­tre and invading down the c-file; Black by attacking the a-pawn or by forcing White to advance prematurely in the centre . 12 . . . e6

The most accurate, discouraging d4-d5 which can give White useful control of the c6-square in the event of a queenside fianchetto. Indeed, Black wants to play . . . b6, both to prevent White's knight from leaping forwards from b3 and to develop the c8-bishop actively.

4 0

13 lLlb3 This is the most flexible. Alternatively: a) 13 .l::tc7? ! is too ambitious, and

after 13 . . . lt::lc6 14 e5 f6 Black has the ini­tiative.

b) 13 e5?! is again too adventurous. After 13 . . . .i.d7 14 lt::le4 .i.c6 Black has the initiative and the d4-pawn will be­come a target.

c) 13 .i.b5 .i.d7! 14 .i.xd7 lt::lxd7 is an attempt by White to speed up his de­velopment, but it also relieves Black's congestion: 15 �e2 (15 .l::tc7? is strongly met by 15 . . . Iifc8 ! ) 15 . . . .l::tfc8 16 lt::lc4 (or 16 lt::lb3 .i.f8 17 .l::txc8 Iixc8 18 Iicl .l::txcl 19 .i.xcl b6 20 �d3 .i.d6 21 g3 h5 and Black equalized in J.Schulz-J.Plachetka, Stare Mesto 2004) 16 . . . .i.f8 17 'it>d3 (17 .i.f4 Iic6 was also fine for Black in R.Hiibner-A.Adorjan, Bad Lauterberg 1980) 17 . . . b5 1 8 lt::la5 .i.a3 19 .l::txc8+ Iixc8 20 lib1 a6 21 .l::tb3 .i.d6 22 h3 f5 23 f3 �f7 was again equal in R.Akesson­F .Elsness, Gausdal 2001; note the cen­tral strike . . . f5, which is a typical idea for Black after the exchange of light­squared bishops.

d) 13 ..ic4 supports a d5-advance: 13 . . . tt:lc6 14 tt:lf3 (14 tt:lb3 �d8 15 d5? ! exd5 16 ..ixd5 tt:lb4 is just weakening and leaves White too exposed) 14 . . . ..id7 15 d5 (instead 15 0-0 tt:la5 16 ..ie2 �fc8 gave Black enough time to prepare . . . b5 and . . . tt:lc4 in C.Fernandez-K.Garrido, Euskadi 2000; White now took the bull by the horns with 17 d5 exd5 18 exd5, but it didn't serve him well and after 18 . . . b5 19 �xc8+ �xc8 20 ..ixa7 �c2 21 ..id3 �xa2 Black was on top) 15 . . . exd5 16 ..ixd5 �ac8 17 0-0 b6 18 �fd1 tt:lb4 and in B.McNally-J.Stevenson, Edin­burgh 1997, Black had comfortably equalized. 13 ... b6 14 ..id3

The alternative 14 �c7 is again pre­mature, this time in view of 14 . . . a5 ! which highlights the lack of squares for the b3-knight, while 14 ..ib5 ..ib7 (14 . . . ..ia6?! 15 a4! is good for White) 15 f3 a6 kicks the bishop away and pre­vents White from invading on c6. 14 ... ..ia6!

Pretty much forcing an exchange of bishops and thereby solving all Black's problems.

Exc h a ng e Va r ia t io n : j_ e 3 Sys te m s

15 ..ixa6 In R.Stone-V.Ivanchuk, New York

1988, White preferred 15 �e2 ..ixd3+ 16 �xd3. The presence of the king in the centre is a mixed blessing for White; it may prove useful if the rooks are ex­changed, but it can also become a tar­get. The game continued 16 . . . tt:la6 17 a3 (17 a4 tt:lb4+ 18 �d2 �ac8 is equal) 17 . . . �fd8 18 �c4 ..if8 19 a4 �ab8 (pre­paring to open files with . . . b5) 20 �e2 f6 ! 21 �a1 ..id6 22 h3 �f7 when Black was for preference as White had run out of ideas. 1S ... tt:Jxa6 16 �e2

White's best plan is to exchange all the rooks, and the text move prepares this. 16 ... �fc8 17 �xc8+ �xc8 18 �c1 �xc1 19 ..ixc1 �f8 20 ..ia3+ �e8

With equality, M.Skliba-V.Bures, Brno 2005 .

This endgame, quite typical of the ..ie3 variation, is of course equal and should be drawn with best play from both sides, but the structural imbalance gives both players reason to try for more.

4 1

Play t h e G r ii nfe ld

B) 7 .lte3

White foregoes ctJf3 in order to use the tempo in a better way, namely to move the rook away from a1 one move earlier. This move order is more flexi­ble than 7 ctJf3 c5 8 .lte3; a variation to which White can always transpose should he wish. White's newly ac­quired options deserve respect, but Black shouldn't have anything to fear and should follow the same plan as in Line A. 1 ... cs 8 'i/Vd2 'i/Vas !

Once again White must make a de­cision as to where to place his queen's rook:

4 2

81: 9 .l:!.c1 82: 9 .l:!.b1

81) 9 .l:!.c1 This leads to a very similar

queenless middlegame to the ones dis­cussed above. 9 . . . cxd4 10 cxd4 'i/Vxd2+ 11 'it>xd2

Instead 1 1 .ltxd2 is rather unnatural and Black is fine after 1 1 . . .0-0:

a) 12 ctJf3 e6 13 .ltb5 tt'Jc6! 14 .ltxc6 bxc6 15 tt'Je5 (15 .l:!.xc6? .ltb7 is, of course, horrible for White) 15 . . . .ltb7 16 .ltb4 (and not 1 6 .l:!.b1 .lta6 1 7 tbxc6? .ltd3) 16 . . . .l:!.fd8 17 .ltc5 f6 18 ctJc4 .lta6 (V.Lazarev-T.Ernst, Oberwart 1991) and Black will occupy the b-file, with at least equality.

b) 12 .ltc3 e6 13 ctJf3 .i:!.d8 14 .ltc4 tt'Jc6 15 d5 .ltxc3+ 16 .l:!.xc3 exd5 17 .ltxd5 ctJb4 gives Black easy equality, G .Burgess-B.Ostenstad, Gausdal 1997.

c) 12 d5 e6 13 .ltb5 (13 .ltb4?! is a misguided attempt to maintain a passed d-pawn: 13 . . . .l:!.e8 14 d6 tbc6 15 .lta3 e5 16 .ltc4 .lte6 1 7 tt'Je2 .ltf8 18 .ltd5 .ltxd5 was much better for Black in A.Vaisser-W.Schmidt, Trnava 1983) 13 . . . .1td7 14 .ltxd7 tt'Jxd7 15 dxe6 fxe6 16 tt'Jf3 .l:!.fc8! 1 7 'it>e2 tbc5 18 tt'Jg5 h6 19 .lte3 hxg5 20 .l:!.xc5 .l:!.xc5 21 .ltxc5 g4! and an endgame with equal chances arose in W.Schmidt-V.Jansa, Vrnjacka Banja 1983. 11 . . . 0-0

Now the move ctJf3 at any stage will transpose to Line A21, but White can also continue in independent vein.

12 i..bs Alternatively: a) 12 d5 e6 13 i..c4 exd5 14 i..xd5 (or

14 exd5 tt'Jd7 15 tt'Je2 .i:!.d8 16 .l:!.hd1, S.Bender-T.Pielmeier, German League 2001, and now 16 . . . ctJe5 with the initia­tive) 14 . . . tt'Jd7 15 f3 tt'Jf6 !? 16 i..b3 a5 and Black is on top.

b) 12 .l:!.c7 .i:!.d8 13 i..b5 (A.Al Khateeb-B.Kouatly, Doha 1993) 13 . . . a6 14 i..c4 tt'Jc6 is fine for Black

c) 12 i..d3 tt'Jc6 13 tt'Je2 .i:!.d8 again raises awkward problems to do with White's d-pawn. L.Van Wely-E.Sutov­sky, Pamplona 1999, continued 14 .l:!.c4 (14 d5 tt'Je5 15 i..b1 e6 was better for Black in S.Tennant-L.Cohen, USA 1994) 14 .. .f5 15 .l:!.b1 (instead 15 e5 i..e6 is good for Black, as is 15 .l:!.hcl i..e6 16 .l:!.a4 i..d7!?, while after 15 f3 fxe4 16 fxe4 i..g4 17 .l:!.b1 .l:!.d7 18 d5 tt'Je5 19 .l:!.cb4, V.Kostic-S.Dvoirys, Oberwart 1999, and now 19 . . . e6 20 .l:!.xb7 exd5 21 .l:!.xd7 i..xd7 Black has the initiative) 15 .. .fxe4 16 i..xe4 i..f5 17 i..xf5 gxf5 when Black al­ready enjoyed a promising position, especially after 18 'it>e1 .i:!.d7 19 f3 .l:!.ad8 20 'it>f2 'it>f7 21 .l:!.b5 e6.

Exch a n g e Va r ia t io n : .1i. e3 Sys t e m s

12 . . . .l:!.d8 13 tt'Je2 tt'Jc6 ! An important concept; Black puts ac­

tivity above structural considerations. 14 i..xc6 bxc6

15 .l:!.hd1 Instead: a) 15 .l:!.c5? e5 ! creates problems for

White: 16 i..g5 f6 17 i..e3 exd4 18 i..xd4 i..a6 and White's position is collapsing, J .Hjartarson-I .Stohl, Copenhagen 1982.

b) 15 .l:!.xc6 i..b7 16 .l:!.c7 i..xe4 17 f3 i..f5 18 .i:!.d1 (18 .l:!.xe7 .l:!.ab8 gives Black too much activity) 18 . . . 'it>f8 saw the two bishops and superior structure guaran­tee Black a lasting plus in N.Kelecevic­J .Smejkal, Sarajevo 1982.

c) 15 f3 e5 16 .l:!.xc6 exd4 17 i..g5 .l:!.e8 18 tt'Jf4 was seen in A.Graf-R.Ruck, Bu­dapest 2004, and now 18 . . . f5 ! gives Black the edge. 1S . . . .l:!.b8 16 .l:!.xc6

Or 16 .l:!.c2 .l:!.b4 17 'it>e1 .l:!.a4 18 f3 i..a6 19 .i:!.dd2 i..c4 20 tt'Jc3 .l:!.a3 21 ctJd1 i..b5 with good counterplay in L.Persson­B .Kyhle, Stockholm 1994, while 16 f3 i..a6 1 7 'it>e1 18 .i:!.d2 only allows White to maintain equality. 16 .. . i..b7 17 .l:!.c7 i..xe4

4 3

Play t h e G r ii nfe ld

A critical situation. Black has a couple of pawns en prise, but White's king is exposed and his pawns loose. K.An­dreev-A.Gubanov, St.Petersburg 1997, continued 18 f3! (as Ftacnik notes, 18 l::i.xe7? l::i.b2+ 19 'it>e1 i.xg2 20 l::i.xa7 i.f3 21 l::i.d2 l::i.b1+ 22 l::i.d1 l::i.xd1+ 23 'it>xd1 i.xd4 is decisive; White should also avoid 18 l::i.xa7 l::i.b2+ 19 'it>e1 i.xg2 20 l::i.xe7 i.f8 21 .\tel l::i.xa2) 18 . . . i.d5 19 l::i.xa7 (19 l::i.xe7 l::i.b2+ 20 'it>e1 l::i.xa2 21 tt:lc3 'it>f8! is good -Ftacnik) 19 . . . i.c4 20 'it>e1 e5! (maximum activity!) 21 dxe5 l::i.xd1 + 22 'it>xd1 i.xe5 23 f4 i.f6 and Black had excellent play.

82} 9 l::i.b1 b6

4 4

Our habitual way o f meeting l::i.b1 and one which normally tempts White into exploiting his temporary control of b5 .

821: 10 l::i.bs 822: 10 i.bS+

Practice has also seen: a) 10 i.c4 i.b7 11 f3 0-0 12 tt:le2 tt:lc6

13 l::i.d1 l::i.ad8 saw White again coming under pressure in the centre in A.Pugachov-V.Tseshkovsky, Leeuwar­den 1994. The game continued 14 0-0 tt:le5 15 i.b3 i.a6 16 l::i.fe1 tt:ld3 17 l::i.fl tt:le5 18 l::i.fe1 tt:lc4 19 i.xc4 i.xc4 and White was struggling to hold d4.

b) 10 l::i.cl looks strange, but White argues that . . . b6 is a serious weakness should Black now enter an endgame .

However, Black can exploit the ex­tra tempo: 10 . . . 0-0 11 tt:lf3 ( 1 1 d5 e6 in­tends . . . l::i.d8 and seems promising for Black) 1 l . . .i.b7 12 d5 tt:ld7 13 .ltd3 (or 13 c4 'Yi'xd2+ 14 .ltxd2 e6 15 .ltd3 f5 16 exf5 gxf5 17 dxe6 l::i.ae8 and White re­grets his premature central advance)

13 . . . c4! when 14 i.xc4? .l:!.ac8 supplies excellent compensation, while 14 i.b1 tt:lc5 15 i.h6 e6 16 i.xg7 'it>xg7 17 Vi'd4+ f6 18 0-0 (L.Portisch-S.Sarno, Reggio Emilia 1992/93) 18 . . . exd5 19 exd5 .l:!.ad8 20 .l:!.fe1 i.xd5 ! 21 .l:!.e7+ .l:!.f7 22 .l:!.xf7+ 'it>xf7 leaves Black clearly better.

821) 10 .l:!.bs 'ifa4

11 .l:!.b3 Black is also fine after the others: a) 11 .l:!.b2 i.a6! 12 i.xa6 (or 12 f3

i.xfl 13 'it>xfl tt:lc6 with the initiative) 12 . . . tt:lxa6 13 f3 (13 tt:le2?! e5 ! highlights the weakness of both white central pawns: 14 0-0 .l:!.d8 15 'ifd1 Vi'c4! 16 .l:!.d2 0-0 17 f3 exd4 18 cxd4 cxd4 19 tt:lxd4 tt:lb4! and Black was better in Y.Dok­hoian-S.Dvoirys, Helsinki 1992) 13 . . . 0-0 with an edge for Black.

b) 11 dxc5?! wins a pawn, but at too high a price in terms of structure. All of a2, c3 and e4 now become rather vul­nerable; for example, 1 1 . . .0-0 12 .l:!.b4 Vi'a5 13 i.c4 (13 cxb6 axb6 is much bet­ter for Black - Adorjan and Feher) 13 . . . tt:lc6! 14 .l:!.b3 (or 14 .l:!.b5 Vi'a4 15 i.e2 .l:!.d8 16 Vi'cl i.a6 and as Aagaard notes,

Exc h a n g e Va r ia t io n : j_ e3 Sys t e m s

Black i s much better) 14 . . . .l:!.d8 15 'ifcl bxc5 (but not 15 . . . tt:le5? due to the nasty 16 .l:!.a3 ! ) 16 .l:!.a3 'ifc7 17 f4 i.b7 18 i.e2, as in I .Ibragimov-C.Toth, Dortmund 1992, and now 18 . . . tt:la5 favours Black since . . . f5 is imminent and e4 weak. 11 . . . 0-o 12 i.bs Vias

13 ctJf3 White should avoid 13 tt:le2? ! since

13 . . . a6! 14 i.d3 (14 i.c4 b5 15 i.d5 .l:!.a7 16 dxc5 e6 17 c6 .l:!.c7 18 i.f4 exd5 19 i.xc7 Vi'xc7 20 exd5 .l:!.d8 21 'iff4 i.e5 22 Vi'f3 Vi'd6 wins everything, as analysed by Adorjan and Feher) 14 . . . tt:lc6 15 d5 tt:le5 16 0-0 f5 ! left Black much better in L.Portisch-A.Adorjan, Hungarian Ch., Budapest 1991 . 13 . . . i.d7 !

Quickly mobilizing Black's remain­ing pieces and now White must be careful just to maintain the balance. 14 i.xd7

Instead 14 i.e2 .l:!.c8 15 0-0 (15 d5? c4 16 .l:!.b4 tt:la6 wins material) 15 . . . cxd4 16 cxd4 Vi'xd2 17 tt:lxd2 .l:!.c2 18 d5 .l:!.xa2 19 i.c4 .l:!.a4 (Adorjan and Feher) is much better for Black. 14 . . . tt:lxd7

4 5

Play t h e G r ii nfe /d

15 0-0 If 15 d5? c4 16 M.b4 tt:Jf6 17 Vi'c2 tZ:lxe4

18 M.xc4 tZ:lxc3 19 iLd2 M.ac8 winning. 1S ... cxd4 16 cxd4 Vi'xd2 17 iLxd2 M.fc8

Now 18 M.c3 M.xc3 19 iLxc3 M.c8 gives Black an edge, since 20 .l:i.cl? iLh6 21 M.c2 tLlf6 wins material, but even the superior 18 .l:i.fb1 M.c2 19 .l:i.3b2 M.ac8 20 <t>f1 e6 (Adorjan and Feher) leaves Black at least equal.

822) 10 iLbS+ The difference with 7 tLlf3 c5 8 iLe3

is that White has this check, but it still doesn't trouble Black. 10 ... iLd7

4 6

The two most logical retreats are now:

8221: 11 i..d3 8222 : 11 i..e2

Retreating the bishop to e2 is White's most common choice, arguing that the black bishop is misplaced on d7, but things are not so simple, as we will see.

White has also tried: a) 11 i..xd7+ tZ:lxd7 is too simplistic

to cause any problems: for example, 12 tZ:le2 (12 tZ:lf3 tLlf6 attacks e4 in an awk­ward way) 12 . . . 0-0 13 0-0 (13 d5? ! Vi'a4 14 f3 f5 once again leaves White over­extended) 13 . . . M.fd8 (due to the pin down the d-file, the d7-knight acquires further prospects) 14 iLg5 cxd4 15 cxd4 (or 15 iLxe7 dxc3 16 'ifc2 M.e8 17 iLb4 VieS 18 f3 M.ac8 and Black is better) 15 . . . Vi'xd2 16 iLxd2 tLlc5 17 e5 tZ:le6 18 iLe3 .l:i.d7 was better for Black in Wang Yaoyao-S.Atalik, Beijing 1996.

b) 11 iLc4 leaves the bishop rather loose: 1 l . . .tZ:lc6 12 tLle2 0-0 13 M.d1 (both 13 iLb3 cxd4 14 cxd4 Vi'xd2+ 15 ct>xd2 tZ:la5 and 13 dxc5 M.ad8 14 cxb6 iLh3 15 Vi'c2 tLle5 ! are promising for Black) 13 . . . b5 14 iLd5 e6 15 iLxc6 iLxc6 16 d5 (White is also struggling after 16 £3 M.fd8 17 0-0 b4 18 cxb4 ifxb4 - Ftacnik) 16 . . . exd5 17 exd5 M.fd8 18 c4 'ifa4 and Black was a lot better in I .Levitina­M.Chiburdanidze, Shanghai Candi­dates 1992.

8221) 11 iLd3

Seemingly the most logical retreat. However, while this supports e4, it also interferes with the queen's protection of d4.

Exch a n g e Va r i a t i o n : i.. e 3 Sys t e m s

for Black in P.Taylor-Jo.Hodgson, Hampstead 1998) 15 . . .'ifxa2 16 M.b2 'ifa3 17 M.xb6 M.fd8 18 '2Jd4 (Black is fine, too, after 18 i.d4 tt::lxd3 19 'ifxd3 ..ltg4) 18 . . . M.ac8 19 .ita6 M.a8 20 i.e2 M.ac8 (Rowson) and Black isn't worse.

b) 13 M.b2 M.ad8 ! ? (or 13 . . . M.fd8 14 0-0 cxd4 15 cxd4 'ifxd2 16 M.xd2 '2Jb4 1 7 i.b1 .itb5 18 a3 '2Jc6 and Black equal­ized comfortably in N.Babu-A.Gok­hale, Mumbai 2003) 14 0-0 i.e6 15 d5 ctJe5 with a complicated position where Black's chances are certainly not worse. 13 ... .l:i.fd8!

Gaining central counterplay is the correct response.

11 ... 0-0 12 tt::le2 14 hS cxd4 15 cxd4 M.ac8! 16 'ifxas Instead 12 tt::lf3 '2Jc6 13 M.cl (or 13 .:t:Jxa s

M.b5 'ifa4 14 M.b2, D.Adla-M.Sion Cas-tro, Zaragoza 1995, and now 14 . . . M.ac8 intends . . . .itg4, with pressure) 13 . . . .itg4 14 d5 M.ad8 15 'ifc2 (15 dxc6?! .itxf3 16 gxf3 c4 i s disastrous for White, but perhaps preferable is 15 0-0 i.xf3 16 gxf3 e6 ! 1 7 c4 'ifxd2 18 .itxd2 '2Jd4 with an approximately equal position -Krasenkow) 15 . . . .itxf3 16 gxf3 '2Jd4 17 'ifd1 was seen in H.Olafsson­A.Huzman, Moscow 2004, and now 17 . . . e6 18 0-0 tt::lc6 leaves Black better. 12 ... tt::lc6 13 h4

An attacking gesture that is not fully justified by the position. Alterna­tively:

a) 13 dxc5? ! tt::le5! ? (or 13 . . . bxc5 14 M.b5 'ifa4 15 M.xc5 tt::le5 16 0-0 M.fc8 ! ? which gives Black good compensation according to Rowson) 14 cxb6 axb6 15 0-0 (15 i.xb6 ifxa2 16 .:t:Jcl '2Jxd3+ 17 'ifxd3 i.xc3+ 18 'it>fl 'ifa4 was excellent

Black's accurate play has left White's h4-h5 advance rather out of place in the position. R.Akesson­J .Timman, Malmo 1997, continued 17 i.g5 (17 M.cl was better) 1 7 . . . f6! 18 i.d2 '2Jc4 19 .itb4 e5 ! 20 hxg6 hxg6 21 dxe5 (or 21 d5 aS 22 .itd2 b5 with the initia­tive) 21 . . .'2Jxe5 22 i.a6 M.c2 23 f4 '2Jc6 24 .itc3 i.g4 25 i.c4+ 'it>f8 and Black had a clear advantage.

4 7

Play t h e G r ii nfe ld

8222) 11 ..te2

11 .. . ..tc6! Black nicely regroups his pieces

with this move. 12 ..td3

White is now wasting time moving his bishop back and forth, but he has no real choice here. Indeed, 12 f3? ! tt'Jd7 leaves White in trouble after both 13 Mel (S.Volkov-N.Kalantarian, Cappelle la Grande 1999) 13 . . . 0-0 14 ctJh3 Mad8, threatening . . . ..tb5, and 13 ctJh3 0-0 14 ctJf2 (R.Sherbakov-R.Akesson, Port Erin 2000)14 . . . Mfd8, preparing . . . cxd4. 12 ... ctJd7

13 tt'Je2

4 8

The most natural, developing White's last piece and intending to support e4 further with f3, but White has also tried:

a) 13 ctJf3 ctJf6 14 d5 was seen in R.Sherbakov-E.Vorobiov, Moscow 1999, and now 14 . . . tt'Jxe4 15 ..txe4 ..txc3 16 dxc6 ..txd2+ 17 ..txd2 (17 tt'Jxd2 0-0-0 is also great for Black) 17 . . ."i¥a4 intends . . . 0-0-0 with some advantage.

b) 13 h4 is a bit too rushed: 13 . . . Md8 14 h5 tt'Je5 ! (a recurring motif in this line) 15 dxe5 c4 16 ..td4 cxd3 17 ctJf3 ..txe4 (White is positionally destroyed, so it's no surprise that his kingside ag­gression comes to naught) 18 h6 ..tf8 19 tt'Jg5! ..txg2 20 Mh2 (20 Mh4?! e6 stops the attack) 20 .. ."i¥d5 21 "iYxd3 (intending f3 to trap the bishop on g2) 2l . . . f6 (the only move) 22 Mb5 (22 exf6? "iYxg5 wins for Black) 22 . . . "iYc6 23 exf6 exf6 24 f3 (or 24 tt'Jxh7 Mxh7 25 "iYxg6+ Mf7 26 h7 Mxd4! 27 Me5+ 'it>d8 28 Me3 Me7 and Black wins - Timman) 24 . . . fxg5 25 Mxg2 ..txh6! and Black was much better in J .Timman-A.Shirov, Wijk aan Zee 1999.

c) 13 Mel e6 and now: el ) 14 ctJe2 0-0 15 0-0 (or 15 h4, as in

S.Ovseevitch-V.Belov, Alushta 2000, and now 15 . . . Mad8 16 h5 ctJe5 17 dxe5 ..txe4 18 ..txe4 Mxd2 19 ..txd2 ..txe5 is much better for Black) 15 . . . Mfd8 16 Mfd1 ..ta4 17 ..tc2 (so far R.Akesson­M.Roiz, Cappelle la Grande 2006) and now 17 . . . ..tb5 favours Black.

c2) 14 ctJf3 ctJf6 (logical, now that White has renounced the option of f3) 15 ctJe5 ..tb5! 16 c4 "iYxd2+ 17 'it>xd2 (or 17 ..txd2 ..ta6 with the initiative)

17 . . . cxd4 18 cxb5 dxe3+ 19 'it>xe3 tt:lh5 20 f4 tt:lxf4 21 'it>xf4 g5+ 22 'it>xg5 i.xe5 and Black had an obvious advantage in view of White's horribly-exposed king in R.Sherbakov-E .Vorobiov, Maikop 1998. 13 .. . lld8!

The most accurate. Note that Black ignores castling in favour of more pressing matters, such as creating im­mediate threats. 14 f3

Other moves have been also tried, albeit without much success:

a) 14 d5 tt:le5 and Black has the ini­tiative.

b) 14 0-0 cxd4 15 tt:lxd4 (or 15 cxd4?! 'ifxd2 16 i.xd2 tt:lc5 ! 1 7 dxc5 llxd3 18 i.e3 bxc5 19 i.xc5 and Black is better, as Aagaard has pointed out) 15 . . . i.b7 16 llb5 Vi'a3 1 7 llb3 Vi'd6 with advan­tage for Black.

c) 14 h4 cxd4 15 cxd4 tt:lc5 16 'ifxa5 bxa5 17 i.b5 i.xb5 18 llxb5 tt:lxe4 is at least equal for Black, V.Epishin­L.Ftacnik, German League 1998. 14 . . . 0-0

Black now has nothing better than

Exc h a n g e Va r ia t io n : .il. e3 Sys t e m s

to castle, but thanks to his accurate play he has managed to obtain suffi­cient counterplay.

15 h4 This aggressive move is White's

only real plan. The way his pieces are developed leaves him without any constructive central options; for exam­ple, 15 llcl i.b7! 16 0-0 tt:le5 !? (another example of this standard tactic; the fol­lowing sequence is now quite forced) 17 dxe5 c4 18 tt:ld4 cxd3 19 e6 i.xd4 20 cxd4 Vi'xd2 21 i.xd2 llxd4 22 llc7 i.c8 23 exf7+ 'it>xf7 (Aagaard) leaves Black better. 1s ... hs !

I t makes sense to prevent h4-h5 al­together. White now struggles to find something else to do. 16 i.gs

Instead Kasparov analyses both 16 'it>f2 cxd4 17 cxd4 tt:le5 ! 18 i.b5 'ifxd2 19 i.xd2 i.xb5 20 llxb5 tt:lc4, with the ad­vantage, and 16 llb2! ? i.a4! 17 'it>f2 tt:le5 ! (again!) 1 8 dxe5 c4 19 tt:ld4 cxd3 20 e6 i.c2 21 exf7+ 'it>xf7 22 tt:lc6 Vi'xc3 23 tt:lxd8+ llxd8, with a powerful initia­tive, while 16 g4? is powerfully met by

4 9

Play t h e G r ii nfe /d

Aagaard's 16 . . . c4! 1 7 �c2 (17 �xc4 tt:ie5 wins material) 17 . . . hxg4 18 h5 gxh5 19 �h6 tt:ie5 ! . 16 . . . .l:!.fe8 17 .l:!.c1 i.b7

It' s notable how Black slowly im­proves his position, while White is un­able to undertake anything positive.

18 0-0 In the stem game, White preferred

the unsuccessful 18 d5? tt:ie5 and suf­fered badly: 19 �b1 ?! tt:lc4 20 'iVf4?? (but even the superior 20 'iVd3 i.a6 21 0-0 tt:ia3 22 'iVd2 tt:ixb1 23 .l:!.xb1 i.xe2 24 'iVxe2 'iVxc3 - Ftacnik - leaves Black

5 0

with a clear extra pawn) 20 . . . i.e5 and White's queen was embarrassingly trapped in T.Shaked-G.Kasparov, Til­burg 1997. 18 ... tt:ies !

Yes, this move again! 19 dxes c4 20 tt:ld4 cxd3 21 e6 �a6!

Black has a clear advantage, accord­ing to Ftacnik.

Concl usion

The i.e3 systems clearly deserve re­spect, but they have lost much of their appeal. In both variations Black should play an immediate . . . 'iVa5, intending to exchange on d4 and the queens on d2 after White plays .l:i.cl . The resulting endgames are fine for Black, who re­tains enough asymmetry to fight for the full point. Instead .l:!.b1 should be met with . . . b6, when �b5+ does not disturb Black as much as White would hope. Indeed, what was once the main line now looks decidedly good for Black!

Ch a pt e r F o u r I Exc h a n ge Va r iat io n : The Fa s h io n a b l e 7 iLc4

1 d 4 'Llf6 2 c 4 g6 3 'Llc3 d s 4 cxds 'Llxds 5 e4 'Llxc3 6 bxc3 .tg7 7 .tc4

This move introduces the most popular and critical system against the Gri.infeld nowadays. Its immense popularity at the highest levels is slowly being transferred downwards, mostly thanks to a monumental refer­ence work by Konstantin Sakaev. The book in question is the most deep, de­tailed and comprehensive opening book I have ever seen! The second edi­tion covers the entire body of theory after 7 .tc4 from both sides and will be the point of reference for many years to come. In its massive 450+ pages it con­tains an incredible amount of original analysis and new ideas, reviving for­gotten lines and burying others . Un­derstandably, the material in that book forms the core of this chapter, particu­larly in Line A. Indeed, there I have opted to indicate where I have discov­ered improvements on Sakaev's analy­sis; the rest of the suggested improve-

ments can be attributed to the Russian grandmaster.

One should not ignore both the theoretical danger posed by Sakaev' s book and the popularity of this system in practice . I have taken this into ac­count when writing this chapter, as you will understand when we move on to the main line of 10 0-0. Naturally this chapter is theoretically the heaviest in this book - there simply was no other choice !

The message is clear: this system is very dangerous. White can effortlessly support his centre with 'Lle2, .te3 and f3, while he retains the option of mount­ing a rapid kingside attack should the opportunity arise. Essentially, Black has only one real target on which to base his counterplay, namely the somewhat loose c4-bishop, but that is enough to give him good queenside counterplay. 7 ... cs 8 'Lle2 'Llc6 9 .te3 o-o

After some logical moves, we reach a major theoretical divide:

5 1

Play t h e G r u nfe ld

A: 10 .U.c1 B: 10 0-0

Castling is White's usual approach nowadays, but he can also initiate an immediate kingside attack with a quick h4. Line A is a prophylactic way of be­ginning that aggressive plan, whereas the direct 10 h4? proved to be prema­ture after 10 . . . cxd4 1 1 cxd4 'Yi'd6 12 l:!.cl l:!.d8 13 d5 ct:Jes 14 �3 .td7 in R.Naranja-L.Portisch, Siegen Olympiad 1970; Black is already much better.

More circumspect is 10 l:!.b1 when Black can play in a similar vein to how he does in Line B :

10 . . . ct:Ja5 1 1 .td3 e5! ? ( 1 l . . .b6 is also playable, just as after 10 0-0 ct:Ja5 1 1 .td3, but I see n o reason t o refrain from the immediate 1 l . . .e5 as White hasn't castled yet) 12 d5 (12 dxe5 is interesting, leading to a complicated position after 12 . . . 'Yi'c7 13 ct:Jf4 'Yi'xe5; the white knight has at least found a role) 12 . . . b6 (12 . . . c4! ? is another possi­bility) 13 c4 (13 'Yi'cl ?! f5 14 .tg5 .tf6

5 2

15 .th6 .tg7 1 6 .tgs 'Yi'd6 17 f3 c4 18 .tc2 f4 was quite good for Black in A.Dreev-K.Thorsteins, Reykjavik 1990) 13 . . . f5 14 ct:Jc3 f4 15 .td2 'Yi'g5 16 'iilfl ctJb7! (bringing the knight to the block­ading d6-square) 17 f3 'Yi'e7 18 ct:Jb5 g5 19 h3 h5 20 .te2 .td7 21 a4 ct:Jd6 was agreed drawn in A.Dreev-G.Kamsky, Reykjavik 1990; the position has ac­quired obvious King's Indian charac­teristics and, of course, there is still a lot of play left in it.

A) 10 .U.c1

This is the main alternative to 10 0-0 . White does not wish to castle and hopes to use his h-pawn as a battering ram against the castled black king. This aggressive line was very popular in the late Eighties and early Nineties, but has nowadays faded into semi-oblivion, as Black has found ways to highlight the drawbacks to not castling. 10 . . . cxd4

In this particular position, this ex­change and the ensuing check on aS create certain coordination difficulties for White.

Exch a n g e Va r ia t io n : Th e Fas h i o n a b l e 7 i.. c4

11 cxd4 'ifas+

Now White must displace his king should he wish to launch an attack. 12 �f1

Instead 12 .td2 is not especially challenging: 12 . . . 'ifh5 13 .tc3 (13 d5 t:be5 14 t:bg3 'ifxd1 + 15 �xd1 t:bxc4 16 .S.xc4 e6 was good for Black in K.Braekkan-B.Ostenstad, Gausdal 1995) 13 . . . 'ifg4 14 t:bg3 'ifxd1+ 15 �xd1 �d8 16 e5 (Black is fine, too, after 16 t:be2 .td7 17 0-0 �ac8) 16 . . . .td7 17 .tb3 .l:Iac8 and Black has easily equalized.

Also possible is the rather timid and inconsistent 12 'ifd2. After 12 . . . 'ifxd2+ 13 �xd2 .l:td8,

with his king in the centre of the board, White finds it hard to keep his centre protected:

a) 14 d5 e6 15 .l:Ihd1 (15 .tg5 f6 16 .tf4 t:ba5 17 .tb5 .td7 is fine for Black) 15 . . . exd5 16 exd5 t:ba5 17 �e1 t:bxc4 18 .l:Ixc4 b6 gives Black a promising posi­tion; his bishops might have their say later on.

b) 14 �c3 t:ba5 15 �b2 t:bxc4+ 16 .l:Ixc4 b6 17 �d1 f5 18 f3 fxe4 19 fxe4 .tb7 was at least equal for Black in N.Grotnes-F.Elsness, Norwegian Team Ch. 2000.

c) 14 .l:Ihd1 t:bxd4 15 t:bxd4 .txd4 16 .txd4 �xd4+ 17 �e3 �xd1 18 �xd1 .tg4 19 f3 �c8 20 .td5 (J.Juptner-E.Mikuev, Czech League 2000) allows White, after some exertions, to regain his pawn and equalize!

d) 14 .td5 t:bxd4 (also fine and per­haps even preferable is 14 . . . e5 15 .tg5 �d6 16 dxe5 t:bxe5) 15 .txd4 (after 15 t:bxd4 e6 16 t:Llb5 exd5 17 t:bc7 dxe4+ 18 �c2 .td7 19 t:bxa8 �xa8 Black had the advantage in S.Lopez Abad-D.Podesta, Buenos Aires 1993; White's king is ex­posed and Black has two active bish­ops) 15 . . . e6 16 .txg7 �xg7 17 t:bd4 exd5 18 e5 .l:Id7 19 .l:Ic3 .l:Ie7 (but not 19 . . . b6? ! 20 .l:Ihcl .ta6 21 .l:Ic7 .l:Iad8 22 e6 with a clear edge for White) 20 �hcl .tf5 with equal chances.

Returning to White's usual choice, the critical 12 �fl : 12 . . . 'ifa3 !

A very strong and important move. The black queen, having enticed White to abandon his castling rights, can now return to d6 if necessary, while its con-

5 3

P lay t h e G r u nfe ld

tact with the cl-rook creates certain tactical threats on d4. In this way Black develops immediate counterplay and doesn't give White enough time to get his attack going.

White has tried several moves here:

A1: 13 'i!Vd2 A2: 13 h4 A3: 13 �c3 A4: 13 'i¥b3

Less critical are: a) 13 dS li:laS 14 ..id3 .i.d7 intending

. . . l:tac8 and Black has the initiative. b) 13 eS iLg4 14 'iWb3 'ti'xb3 1S axb3

lt:laS 16 f3 it.d7 17 �f2 i:.ac8 18 .i.d2 lt:lxc4 19 bxc4 f6 forces the break-up of White's centre, A.Abdulla-P.Konguvel, Calcutta 2002.

c) 13 it.b3 eS 14 dS t'Dd4 1S li:lxd4 exd4 16 iLf4 .i.d7 17 £3 .l:i.ac8 18 'it>f2 aS 19 'ti'd3 bS 20 d6 l:tc3 was horrible for White in V.Krutti-A.Keatinge Clay, Harkany 1999.

A1) 13 'ilfd2 l:tdS !

5 4

Introducing the aforementioned threat against d4. 14 d5

Alternatively: a) 14 h4? ! ignores the threat at

White's peril : 14 . . . lt:lxd4 1S ..ixd4 (or 1S l:tc3 'iVd6 16 .l:td3 eS with a clear plus) 1S . . . it.xd4 16 hS 'i!Vb2 17 'ii'f4 e6 and White's compensation seems insuffi­cient.

b) 14 l:Id1? ! (this doesn't actually prevent . . . li:lxd4) 14 . . . li:lxd4! (also prom­ising for Black is 14 . . . lt:laS !? 1S .i.d3 ..id7 16 h4 hS, with the idea of . . . l:tac8 and . . . lt:lc4) 1S ..ixd4 eS 16 .i.dS exd4 17 li:lxd4 �h8 18 lt:lc2 \!i'cS and Black was better in M.Laframboise-S.Gravel, Montreal 2003.

c) 14 f3? ! is similarly irresponsible: 14 . . . li:lxd4 1S .U.c3 'iVaS! (1S . . . 'ilfd6?! 16 li:lxd4 .i.xd4 17 .i.xd4 'i!t'xd4 18 1Wxd4 lhd4 19 �e2 it.d7 20 iLdS allows White to regain the pawn) 16 li:lxd4 .i.xd4 17 .i.xd4 eS 18 .idS exd4 19 If.d3 'ifxd2 20 .U.xd2 ..ie6 21 l:txd4 ..ixdS 22 exdS .U.ac8 and in view of White's weak back rank and the passive h1 -rook, Black is better . 14 . . . lt:les 15 ..ibs

Exch a n g e Va r ia t io n : Th e Fas h i o n a b l e 7 il. c4

In the case of 15 .tb3 b6 (threaten­ing . . . .ta6 and . . . tLld3) 16 f4 tLlg4 17 .td4 e5 (17 . . . .txd4 18 'i¥xd4 .ta6 19 �d1 'tlfd6 20 e5 'tlfc5 is also excellent) 18 .tb2 'tlfe7 19 h3 lLlf6 Black gains a strong ini­tiative due to White's exposed king. 15 . . . b6

Preparing . . . .ta6 and a later inva­sion on d3 and/or down the c-file. White must now play very precisely to stay in the game. 16 'i¥c3

Instead 16 .td4 .ta6 17 .txa6 't!Vxa6 gives Black the upper hand, while both 16 �c3 'i¥a5 17 a4 .ta6 18 �b3 'i¥xd2 19 .txd2 .txb5 20 axb5 �ac8 and 16 l2Jd4 .ta6 17 .txa6 'i¥xa6+ 18 'i¥e2 'i¥a3 are plain bad. Finally, the overambitious 16 f4? ! tLlg4 17 .td4 was met by the strong 17 . . . e5! in L.Ftacnik-I.Gurevich, Biel Interzonal 1993. 16 .. . 'i¥xc3 17 l2Jxc3 ts 18 .tgs �f8 19 f3 a6 20 .te2 bs 21 �f2 .tb7

Black now has some obvious moves, such as . . . �ac8 and . . . l2Jc4,

This reveals another point behind 12 . . . 'i¥a3 ! ; White cannot play f3 as then the e3-bishop hangs. 14 hs

Instead 14 'i¥d2 �fd8 15 f3 lLlxd4 16 fxg4 l2Jf3 17 'i¥c2 'i¥xe3 18 gxf3 'tlfxf3+ 19 �g1 't!Ve3+ 20 �fl �d2 should result in a miniature, while 14 .tb3 .txe2+ 15 'it'xe2 �ac8 (but not 15 . . . 't�Vb2+? ! 16 �c2 tLlxd4+ 17 .txd4 'i¥xd4 18 'i¥xd4 .txd4 19 �c7 and White is better) 16 d5 (both 16 h5 't�Vb2+ and 16 �c2 �fd8 are no better) 16 . . . tLle5 gives Black the initiative.

which promise him good play. 14 . . . .txhs ! It's as simple as that! Black's point

A2) 13 h4 .tg4! is that White cannot play 15 �xh5 gxh5

5 5

Play t h e G r ii nfe /d

16 ltJg3 since 1 6 . . . i.xd4 ( 16 . . . 'ii'd6 also suffices) 17 i.xd4 ltJxd4 18 ltJxh5 'ii'd6 defends and wins.

A3) 13 l:.c3

This solves the immediate tactical problems but leaves the rook mis­placed. 13 . . . 'ii'd6 14 h4

The alternatives are no better: a) 14 f4? ! is premature due to 14 .. . e5! . b ) 14 l:.d3 ltJa5 15 �d5 (15 i.b5 i.d7

is also good for Black) 15 . . . e6 16 �f4 'ii'a6 17 �b3 �d7 18 h4 .ib5 19 l:.e3 ltJxb3 20 axb3 �xd4 and Black was winning in C.De Wolf-V.Mikhalevski, Vlissingen 1999.

c) 14 'ii'd2 l:td8 15 l:.d3 b6 16 f3 ( 16 i.f4 �4) 1 6 . . . �b7 17 �f4 'iVb4 18 �d5 l:.ac8 19 'it>f2 e6 20 i.xc6 'ii'xd2 21 nxd2 l:.xc6 was fine for Black in M.Franic­M.Rade, Mravinci 1993.

d) 14 f3 ltJa5 (a simple way to equal­ize is 14 .. .'�Jxd4 15 ltJxd4 - 15 e5 ltJxe2 16 exd6 ltJxc3 17 'ii'd2 exd6 offers excel­lent compensation for the queen -15 . . . e5 1 6 l:.d3 exd4 17 �xd4 i.e6) 15 �d3 i.e6 16 i.f4 'iVb4 17 �d2 b6 18 l:.c7

5 6

'ii'd6 1 9 i.f4 'ii'd8 20 d5 �d7 2 1 i.a6 e6 22 dxe6 .ixe6 23 'ii'c2 i.c8 24 �b5 a6 with an edge for Black, W.Browne­S.Kudrin, Philadelphia 1992. 14 ... h 5

White now lacks a sensible plan. 15 'ii'd3

Also very good for Black was 15 'ii'd2 e5 16 d5 ltJa5 17 �d3 b6 18 i.h6 �d7 19 g3 ltJb7 20 Wg2 ltJc5 21 �e3 ltJxd3 22 l:.xd3 f5 in F.Elsness­B.Predojevic, Pula 2005. Note how White's exposed king is a recurring problem for him in this variation. 1s ... ttJas 16 �f4

1 6 i.b5 a6 17 �a4 b5 18 i.c2 ltJc4 is also advantageous for Black. 16 .. . es 17 dxes 'ii'xd3 18 i.xd3 ttJc6

White is still to get his king' s rook into the game and Black is better after both 19 l:tc5 l:.d8 20 �c2 �e6 and 19 i.c4 l:.d8 20 i.d5 i.e6 21 nc5 ltJxe5.

A4) 13 "ii'b3 A change of plan, but not a bad

move because a queen exchange would leave White with the preferable posi­tion due to his strong centre.

Exch a n g e Va r ia t io n : Th e Fas h i o n a b l e 7 .i.. c4

13 . . .'iVd6! Black logically avoids the exchange

and will now gain further time with . . .'�Ja5. 14 i.ds

The only sensible way to cover the d4-pawn, since 14 'ilfc3 is well met by 14 . . . i.e6! and 14 e5 is too committal: 14 . . .'iVd8 15 'ilfa3 i.f5, intending . . . .l::!.c8 and . . . tt:Ja5, gives Black the initiative. 14 . . . tt:Jas ! 15 'ii'bs

Instead 15 'i!t'a4 'ilfd8 1 6 .id2 tt:Jc6 17 i..xc6 bxc6 leaves White badly lacking a good follow-up since grabbing the c6-pawn is rather suicidal . 1S . . . 'ii'd8 !

The journey of the black queen ap­pears ridiculous, but just consider the moves White has made in the mean­time. As a result, Black is now firmly in control of the initiative. 16 llcs .id7! 17 'i!t'xas b6 18 'ii'd2 bxcs 19 i..xa8 cxd4! 20 i.xd4 i..xd4 21 'ii'xd4 'i!t'xa8

This forced sequence has led to some simplification of the position, but White's kingside pieces remain mis­placed. Significantly, White cannot play 22 'i!t'xd7? .l:.d8 23 'ii'a4 'i!Vxe4 24 'iVb3 .l:.d2 25 l2Jg3 in view of the deci­sive 25 . . . 'ii'd4!, as pointed out by Atalik, but even after 22 f3 i.b5 23 �f2 .Ud8 Black is obviously better.

B) 10 0-0 The absolute main line of this sys­

tem now runs 10 . . . i..g4 1 1 f3lDa5

when White has a choice between sacrificing an exchange for good attacking chances (12 i.d3 cxd4 13 cxd4 i..e6 14 d5), sacrificing a pawn for promising positional compensation (12 i.d3 cxd4 13 cxd4 i.e6 14 .l:.cl), and actually winning one while simplifying

5 7

Play t h e G r ii nfe ld

the position (12 �xf7+) . All three lines are extremely complicated and im­mensely theoretical . Sakaev's mam­moth work covers them all extensively and it would be a foolish task to use up half of this book trying to prove that Black is OK against all of these danger­ous plans. Moreover, in several of the ensuing positions, a draw is the most that Black can achieve.

Instead by covering an immediate 10 . . . 'L'la5, I have opted for a rare but by no means inferior set-up. Indeed, this has acquired some recent popularity and been played by both Svidler and lvanchuk. It both offers fertile ground for research and, in the spirit of this book, allows Black to fight on equal terms without having to memorize masses of variations. In addition, I have also opted to cover what is now Black's most popular alternative to the main line, namely the 10 . . . i.d7 system. Thus, should any reader feel a little uncomfortable employing a rare and fairly novel system, they can instead prefer a tried and tested variation.

Our coverage now divides into:

81: 10 . .. 'L'las 82: 10 . . . �d7

81) 1o . . . 'L'las ! ? The strategical basis behind

10 . . . 'L'la5 is clear: since White's set-up is designed to absorb central pressure, Black instead focuses on the c4-square. In a sense, this move is also quite logi­cal as the c4-bishop is White's only

5 8

loose piece in his otherwise harmoni­ous set-up.

11 �d3 b6 Black offers the c5-pawn some sup­

port, for reasons which will become clearer after his next move. The inten­tion is not really to develop the light­squared bishop on b7, but rather to play . . . e5, trying to force White to clar­ify the central tension.

The d4-d5 advance is then White's most obvious response, but this rather plays into Black's hands; he can bring the aS-knight to the blockading square d6 and prepare . . . f5. The resulting posi­tions seem perfectly satisfactory for

Exch a n g e Va r ia t io n : Th e Fas h i o n a b l e 7 fi. c4

Black, which is why White should probably opt for dxc5 instead. That is generally an undesirable move, but also one which can prove strong under the right circumstances. 12 .l:!:c1

This is, I believe, the only real way to trouble Black, enabling the advance of White's central pawns .

Instead 12 f3 e5 13 dxc5 (or 13 d5 f5) 13 . . . .1ie6 14 .l:!:b1 lt:Jc4 15 i.xc4 i.xc4 16 �b2 (16 cxb6? ! axb6 only serves to en­hance Black's pressure) 16 . . . b5 leaves Black clearly the only side with any positive prospects .

More logical, but also not especially accurate is 12 'i¥d2. Black again re­sponds with 12 . . . e5 and now:

a) 13 dxc5 .lie6 14 .l:!:fd1 'i¥c7 again offers full compensation for the pawn, as Black is in control of the c4-square. Note that in the positions resulting from dxc5, White's e2-knight is always a problem piece for him since it lacks a good square to move to.

b) 13 d5 is not really the best ap­proach to the position; then one in­structive example was the 13 .. . f5 14 .lig5 'i¥d6 15 .lih6 f4 16 .lixg7 'it>xg7 of P .Nielsen-V .I vanchuk, Monaco (blind­fold) 2006. Black is already very com­fortable, as the continuation of the game showed: 17 f3 c4 18 .lic2 g5 19 �h1 i.d7 20 g3 .l:!:ae8 21 gxf4 gxf4 22 .:i.g1+ 'iilh8 with a small but obvious advantage for Black.

c) 13 dxe5 .lixe5 14 .l:!:ad1 i.b7 leaves Black comfortably placed and his pieces with good scope. 12 . . . es 1

Black strikes back in the centre and White must now deal with the threat to d4. 13 dxc5 !

I believe that this is White's only way to fight for an advantage. Notably it was Topalov's choice when faced with this line in a game against Svidler.

Alternatively: a) 13 dxe5? ! aims for a quick king­

side pawn advance, but Black needn't be worried about this at all : 13 . . . i.xe5 (Black might even try 13 . . . .1ie6 ! ? 14 f4 i.xa2 - but not 14 . . . lt:Jc4? 15 .lixc4 .lixc4 16 'i¥e1 ! - 15 lt:Jg3 with complicated play, although White's kingside play could prove dangerous) 14 f4 i.g7 15 e5 (now the g7-bishop i s blocked, but that has come at too high a price for White on the light squares) 15 . . . i.e6 16 'i¥c2 lt:Jc4 sees Black's pieces easily find their way to good squares and the ini­tiative is on his side.

b) 13 d5?! fails to impress after 13 . . . f5 (or even 13 . . . c4 14 i.c2 .lid7 15 .l:!:b1 CLJb7 16 .lia4 .lixa4 17 'i¥xa4 CLJd6 18 f3 f5, with good counterplay for Black in V.Seifert-J .Zezulkin, Bayern 1999) 14

5 9

P lay t h e G r ii nfe ld

f3 c4 15 .ib1 f4 16 i.f2 g5 when Black is better; the future of White's minor pieces, especially the b1-bishop and the e2-knight, seems rather bleak, while Black has an obvious and good plan. 13 . . . i.e6!

As usual, this eyes the c4-square. Note that while the e5-pawn does block the beloved Gri.infeld bishop, it both deprives the white knight of a good future and offers Black's position a high degree of stability; a factor which ensures that he can develop his queenside counterplay at leisure.

Instead 13 . . .'i¥c7 looks like a reason­able move, but it is not incisive enough: 14 cxb6 (14 c4! ? is also promising) 14 . . . axb6 15 c4 i.e6 16 c5 bxc5 17 .l:Ixc5 '1i'd7 18 tt:lcl ! ? and White retains a plus.

White must now decide how to proceed on the queenside:

811: 14 cxb6 812: 14 C4

811) 14 cxb6 This capture is, of course, very sig-

6 0

nificant since Black must now b e able to demonstrate sufficient compensa­tion. Due to both his strong central con­trol and White's weakened queenside, he is fully able to. 14 . . . axb6

14 . . . tt:lc4? ! is not so good in view of 15 b7 .l:Ib8 16 i.xc4 i.xc4 17 i.c5 .l:Ie8 1 8 .l:Ib1 and White is clearly better. 15 C4

Naturally, White must prevent . . . tt:lc4 at all costs, as otherwise Black's play unfolds effortlessly and with great vigour. 1s . . . fs !

A dynamic and strong advance, whereas 15 . . . tt:lb7? is simply too slow: 16 tt:lc3 tt:lc5 17 i.e2 and White accom­plishes his main positional task, namely that of bringing his knight to a good square. 16 exfs

16 f3 is more solid, but Black can immediately equalize with 16 . . . fxe4 17 i.xe4 (or 17 fxe4 .l:Ix£1 + 18 Wxfl .l:Ic8 and the c4-pawn falls, with at least equality for Black) 17 . . . tt:lxc4!, if not actually hope for more.

Exch a n g e Va r iat io n : Th e Fas h i o n a b l e 7 Ji.c4

16 . . . gxfs 17 f3 This is pretty much compulsory;

White must restrain Black's advancing pawn duo. 17 ... tt:'lb7 !

Only now does Black improve the placement of his pieces and with tempo. Instead 17 . . . e4? is a big mistake in view of 18 fxe4 fxe4 19 .:xf8+ 'iVxf8 20 i.xe4, but 17 . . . .:c8 is playable, albeit slightly better for White after 18 cS bxcS 19 .:xeS (and not 19 i.xcS? .:xeS 20 .:xeS 'iVh6) 19 . . . .:xcS 20 i.xcS .:£7 21 'iVh1! tt:'lb7 22 i.b6 (22 i.f2 i.xa2! equal­izes) 22 . . . 'iVd6 23 a4 tt:lcS 24 i.xcS 'iVxcS+ 2S �hl .

18 tt:'lc3 .:a3 Black's initiative is growing rapidly:

the advance . . . e4 is now threatened, the knight is coming to cS and so on. White is rather thrown on to the defensive and there is no question that Black has adequate compensation for the sacri­ficed pawn.

812) 14 C4 !? This was introduced by the always

well-prepared Veselin Topalov in his

aforementioned game with Svidler. In this ambitious way, White prevents the occupation of c4 and prepares to bring his knight to dS via c3. Accomplishing that manoeuvre would guarantee White a lasting positional advantage and so Black must urgently sharpen the struggle.

14 ... bxcs ! 15 i.xcs i.h6!? Svidler's suggested improvement in

Informant 96 (and the following analy­sis is based on his notes) over the stem game which saw 1S . . . .:e8?! , after which the opening battle concluded in White's favour: 16 i.e3 .:e7 (16 . . . .:c8 17 cS allows White to safeguard his extra pawn) 17 'iVc2 .:d7 18 .:fd1 tt:'lc6 19 a3 'iVaS (19 . . . .:b8 20 tt:'lc3 tt:'ld4 21 i.xd4 exd4 22 tt:'ldS gives White a safe edge, as Golubev notes) 20 tt:'lc3! (20 'iVc3 .:ad8 allows Black counterplay) 20 . . . 'iVxa3 (20 . . . tt:'ld4 21 'iVh2 is also good for White) 21 tt:'ldS and White had the upper hand in V.Topalov-P.Svidler, Linares 2006.

The text move seems much stronger to me and now the position becomes very complicated.

6 1

Play t h e G r ii nfe ld

16 .l:!.c3 This protects the d3-bishop, thereby

freeing the white queen. Alternatively: a) 16 Sl.xf8 is obvious, but offers

White nothing after 16 . . . SI.xcl 17 tbxcl Wxf8 18 4:'lb3 ( 18 'i¥a4 .l:!.c8 19 .l:i.d1 'i¥c7 is similarly fine for Black as c4 drops) 18 . . . SI.xc4 19 4:'lxa5 Axd3 20 4:'lc6 'i¥c7 21 'i¥xd3 'i¥xc6 with equality.

b) 16 .l:!.c2 .l:!.e8 intends . . . .l:!.c8 and is also fine for Black: 17 £4 (17 4:'lc3 4:'lxc4 1 8 4Jd5 Sl.xd5 19 exd5 4Jd6 is compli­cated but seemingly fully acceptable, as Black has a firm blockade on d6 and is about to set the e- and £-pawns rolling) 17 . . . .l:!.c8 18 Sl.xa7 .l:!.e7 19 Sl.e3 .l:!.d7. White must now return the material to relieve the pressure and after 20 .l:i.d2 4:'lxc4 21 Sl.xc4 .l:!.xd2 22 'i¥xd2 'i¥xd2 23 Sl.xd2 .l:!.xc4 Black has at least equalized.

c) 16 £4 .l:!.e8 again sets White the problem of how to proceed:

cl ) 17 .l:!.c2 transposes to 1 6 .l:!.c2 .l:!.e8 17 f4, above.

c2) 17 Sl.e3 .l:!.e7 (17 . . . .l:!.b8! ? can also be considered) 18 'i¥c2 .l:i.d7 19 .l:!.cd1 .l:!.c8 20 c5 Sl.xa2 21 'i¥xa2 .l:!.xd3 22 .l:!.xd3

6 2

'i¥xd3 sees Black regaining his material while retaining some pressure.

c3) 17 .l:!.c3 .l:!.c8 18 Sl.xa7 4:'lxc4 19 'i¥c2 'i¥a5 20 Sl.xc4 (alternatively, 20 Si.f2 tba3 21 .l:!.xc8 .l:!.xc8 offers Black excellent compensation in view of his very active pieces) 20 . . . 'i¥xa7+ 21 Wh1 Sl.xc4 22 .l:!.xc4 'i¥a6 23 .l:!.xc8 .l:!.xc8 forces 24 'i¥d1 exf4 25 4Jxf4 Sl.xf4 26 .l:!.xf4 'i¥xa2 fizzles out to full equality. 16 .. . .l:!.e8 17 Sl.a3

Preparing to meet . . . .l:!.c8 with c4-c5. 17 .. . tbc6!

The knight heads for the d4-square and frees a5 for the queen.

18 'i¥a4!? Passive play allows Black's initia­

tive to develop unhindered; for exam­ple, 1 8 c5 'iVaS! with excellent play for Black. Therefore I think that White' s best option i s to embark on the active text move and the resulting exchange sacrifice. 18 . . . i.d2 19 .l:!.d1 i.xc3 20 4Jxc3

This position is the result of some critical and ambitious play from both sides. Black has a slight material advan­tage, but he is missing his dark-squared

Exch a n g e Va r ia t io n : Th e Fas h io n a b l e 7 .i. c4

bishop. This is not so much a problem on the kingside, as the eS-pawn keeps lines closed there, but rather in the cen­tre and on the queenside. It is hard to give a definite evaluation of this posi­tion and certainly both sides have their chances, as well as good outposts for their knights. However, if pushed I would say that Black probably stands fairly well after 20 . . . .l::!.c8 .

82) 10 .. . .Jid7

This quiet move represents a very solid and positional way of playing. Black simply prepares to play on the queenside and he has a very clear plan of obtaining control of the key c4-square, with moves such as . . . �c7, . . . .l::!.c8, . . . bS (or . . . .JibS) and, of course, . . . lt:JaS-c4. Achieving that plan usually guarantees him good play. White, in turn, must try to disturb Black' s afore­mentioned plan. He has two ways of doing so: the dS-advance and the cap­ture on cS. Both options have advan­tages and drawbacks, but overall I feel that only d4-dS can really trouble Black. Against that, Black should re-

spond by closing the centre with . . . eS, just as in Line B 1 .

This 10 . . . .Jid7 system, just like 10 . . . lt:JaS, was introduced into high­level practice by Peter Svidler and re­mains the regular choice of the Israeli Griinfeld aficionados - a good recom­mendation indeed!

Now White usually elects to remove his queen' s rook from the long diago­nal:

821: 11 .l::i.b1 822 : 11 .l::!.c1

Many other moves have been tried here, but they are generally rather ir­relevant and offer Black excellent play:

a) 11 dS is premature: 1 1 . . .lt:JeS (note that the knight is actually quite well placed here, since booting it with f4 is not an option in view of . . . lt:Jg4) 12 .Jib3 �aS 13 Wh1 (13 f4? ! lt:Jg4 14 .Jid2? c4! drops material, and also promising for Black is 13 c4 bS! 14 cxbS .itxbS accord­ing to Krasenkow) 13 . . . .itbS 14 f4 �a6 1S .l::!.e1 lt:Jd3 and Black was better in P.Verdier-O.Houhou, Paris 200S.

b) 1 1 f4 is an aggressive but loosen­ing move. The most efficient reply is 1 1 .. .cxd4! 12 cxd4 .itg4 13 dS lt:JaS 14 .Jid3 e6, breaking up White' s centre .

c) 1 1 a4 cxd4 12 cxd4 .l::!.c8 13 .l::i.cl lt:JaS 14 .Jid3 .l::!.xcl 1S .itxcl (V.Dobrov­P.Jaracz, Pardubice 2001) 1S . . . �6 (Krasenkow suggests the interesting 1S . . . �c7! ? 16 .ita3 .l::!.e8 with counter­play) 16 .ite3 .l::!.c8 sees Black taking over the initiative.

6 3

Play th e G r ii nfe ld

d) 1 1 dxc5 is a capture which usu­ally restricts White' s middlegame po­tential, since if Black recovers the pawn, White is left with a miserable position. Black's best approach is again to fight for the c4-square, this time usually with . . . tt:Ja5 and . . . i..e6/b5. Al­ternatively, he may prefer to play . . . b6 at an appropriate moment, gaining typical compensation after an exchange on b6.

Play might continue l l . . :�c7! ? (1 1 . . .4Ja5 1 2 it.d3 'iVc7 is similar, but probably less accurate in view of 13 4Jd4 .l:tfd8 14 .l:tbl b6 15 cxb6 axb6; Krasenkow believes that Black has good compensation here, but I think that after 16 'iVe2, intending a nasty 4Jb5, White is definitely better) 12 f4 tt:Ja5 13 it.d3 .l:tfd8 with good compen­sation; we will see more of this type of sacrifice in note 'a' to White's 12th move in Line B21 .

e) 1 1 'iVd2 is not a threatening move and actually offers Black a pleasant choice:

el) l l . . .b5 ! ? is a simple equalizing device;

6 4

e.g. 12 i..xb5 4Jxd4 13 i..xd7 tt:Jxe2+ 14 'iVxe2 'iVxd7 and Black is fine.

e2) l l . . .'iVa5 12 �2 (12 .l:tacl .l:tac8 transposes to Line B22) 12 . . . .l:tab8 (again 12 . . . b5 ! ? 13 'iVxb5 'iVxb5 14 i..xb5 4Jxd4 equalizes) 13 i..b3 'iVa6 14 .l:tfdl cxd4 (14 . . . 4Ja5 is also playable: 15 dxc5 i..c6 16 �c2 it.xe4 17 �xe4 �xe2 18 �xe7 4Jxb3 19 axb3 i..xc3 is equal, while here 16 4Jg3 tt:Jc4 17 �cl tt:Jxe3 18 'iVxe3 'iVa5 gave Black the initiative in S.Williams­B.Macieja, European Championship., Warsaw 2005) 15 cxd4 .l:tbc8 and Black has the upper hand.

821) 11 .l:tb1

Exc h a n g e Va r ia t io n : Th e Fas h io n able 7 i.. c4

This has emerged as the most threatening continuation, applying pressure down the b-file. While an immediate capture on b7 is impossible due to a . . .'�Ja5 fork, White does pre­vent the desirable .. .'iVa5 (as the pawn could then be captured) and thus forces Black into a less active deploy­ment of his forces. 11 . . .'iVc7!

Simply protecting the b7-pawn with a useful move. Black is now indirectly attacking the c4-bishop, a factor which forces White's hand in the centre.

There is another consistent move in this position that I would like to draw your attention to, namely l l . . .:i::i.c8! ? .

Surprisingly, this has been quite rare in practice but, in the few exam­ples available to me, the opening phase has concluded quite successfully for Black: 12 .\td3 (12 dxc5 Wilc7 is very similar to positions we will examine in the notes to White's 12th move in our main line) 12 . . . e5 13 d5 tt:la5 and now:

a) 14 Wi/cl Wi/e7 15 �hl f5 16 exf5 gxf5 17 c4 b6 18 .ltg5 Wi/f7 19 tt:'lg3 tt:'lb7 20 it.e2 Wi/g6 21 f4 e4 22 tt:'lh5 .ltd4 23

l:.b3 e3 was very unclear in P.Cramling-A.Groszpeter, Biel 1990.

b) 14 Wiid2 (more common) 14 . . . £5 15 .ig5 (15 f3 f4 16 .if2 c4 17 it.c2 b6 18 �b4 �f7 was quite satisfactory for Black in M.Van der Werf-F.Nijboer, Leeuwarden 2001) 15 . . . Wi/e8 1 6 ex£5 (or 16 c4? ! b6 17 tt:'lc3 tt:'lb7 18 �bel f4 19 f3 h6 20 .lth4 g5 21 .ltf2 �f6 22 �e2 h5 23 �bl �g6 and Black's kingside attack was in full swing in J.Planas-J.De la Villa, Palma de Mallorca 1991) 16 . . . gxf5 17 tt:'lg3 Wilg6 18 f4! e4 19 .lte2 Wild6 (much worse was 19 . . . kf6? 20 tt:'lh5 .idS 21 c4 b6 22 �bel with an attack for White in A.Shariyazdanov-E.Vorobiov, Tomsk 2004) 20 tt:'lh5 kh8 21 Wi/e3 h6 22 .ih4 Wi/xd5 and Black is firmly in charge of the initiative.

Overall, I feel that l l . . .�c8 ! ? is a fruitful area for research, but do note that Black should avoid the once popu­lar l l . . .a6? ! . The problem is that this creates significant weaknesses on the queenside; a factor best highlighted by 12 dxc5! after which Black has strug­gled. 12 kf4

6 5

P lay t h e G r ii nfe ld

Harassing the queen i s White's best option here, but he has tried a number of alternatives:

a) 12 dxcS 'LlaS 13 it.bS (or 13 .id3 .:tfd8 14 �c2, J.Horvath-D.Boros, Hun­garian League 2004, and now 14 . . . 'Llc6 ! ? intends . . . 'LleS and . . . �aS, thereby giv­ing Black his fair share of the play) 13 . . . it.xbS 14 .:txbS 'Llc4 1S it.d4 eS 16 �d3 a6 with good compensation for Black, Kir .Georgiev-B .Predojevic, Zlati­bor 2006.

b) 12 �d2 a6 13 .:tfd1 (13 dxcS .:tad8 supplies typical compensation) 13 . . . bS 14 .idS e6 1S it.f4 �aS gives Black the initiative, P.Skatchkov-A.Timofeev, Sochi 200S.

c) 12 'Llf4 .:tad8 13 'LldS �c8 14 .ibS (S.Citak-I .Kurnosov, Istanbul 200S) 14 . . . <;io>h8 avoids White's threat and leaves his pieces misplaced.

d) 12 .id3 removes the bishop from its exposed position on c4. However, Black can increase the pressure against White's centre with 12 . . . .:tfd8 and now:

d1) 13 dS 'LleS 14 c4 e6 1S 'Llf4 b6 16 it.e2 h6 17 �cl <;io>h7 18 'Lld3 exdS 19 cxdS fS gave Black sufficient counter-

6 6

play in J .Lahner-M.Meszaros, Frydek Mistek 200S.

d2) 13 �c1 .:tac8 14 .:td1 prepares �a3, both to prevent . . . �aS ideas and force a release of the central tension. However, it does appear somewhat artificial; for example, 14 . . . a6 1S 'iVa3 (1S dS? 'LleS 1 6 c4 bS 17 cxbS 'Llxd3 1 8 .:txd3 it.xbS 19 .:td2 'iVaS leaves Black on top according to Berkes) and now in F .Berkes-E.Sutovsky, Paks 200S, Black should have played 1S . . . cxd4! 16 cxd4 'iVaS 17 �xaS 'LlxaS, when he appears to have secured dynamic equality. Berkes offers 18 dS! 'Llc4! 19 .:txb7 'Llb2 20 .:td2 (or 20 .:txb2 it.xb2 21 .ib6 .:te8 with counterplay) 20 . . . 'Llxd3 21 .:txd3 .ibS 22 .:td2 e6 23 dxe6 fxe6! 24 .:tc7! .:tb8 with good compensation for the pawn in view of Black's activity.

d3) 13 f4 ..ltg4 14 d5 e6! (Black's play is based on this important tactical de­tail; instead 14 . . . 'LlaS? 15 c4 would, of course, be a positional disaster) 1S dxc6 c4 (regaining the piece) 16 cxb7 .:tabS (Berkes) and Black already has the ini­tiative. Indeed, after the natural 17 .id4 it.xd4+ 1 8 cxd4 .:txd4 19 'ii'c2 .:txd3, Black's counterplay is rather awkward.

Returning to 12 it.f4: 12 . . . 'iVc8!

This might look like a waste of time, but the truth is that the white bishop is misplaced on f4. It is both a target for . . . eS and prevents more aggressive plans (f4-f5). The text also sees Black retain tac­tical threats against the c4-bishop, whereas 12 . . . eS is somewhat premature; after 13 it.g3! the central tension cannot be resolved in Black's favour.

Exch a ng e Va r ia t io n : Th e Fas h io n a b l e 7 i. c4

13 ds Instead 13 'i¥d3 .I:i.d8 14 .I:i.fcl tt:'laS 1S

�bS a6 16 �xd7 'ifxd7 17 'i¥g3 .I:!.ac8 equalized very comfortably in J .Kolly­B.A vrukh, Biel 2002, while 13 .I:i.cl .I:i.d8 (13 . . . a6 ! ? 14 a4 tt:'laS 1S �d3 bS looks to me like a promising way to grab the initiative) 14 dS tt:'leS 1S .txeS �xeS 16 f4 �g7 17 'ifb3 .I:i.b8 18 a4 b6 19 eS �e8 20 tt:'lg3 a6 21 .I:!.cd1 bS 22 axbS axbS 23 i.e2 'ifb7 was seen in M.Rodshtein­A.Huzman, Ashdod 2004; although White's central presence was quite good, Black had decent counterplay. 13 . . . tt:Jas 14 �d3 es

Black proceeds with the standard method. Just as in the 10 . . . tt:'laS varia­tion, the plan is to bring the aS-knight to the strong blockading square d6 and to advance on the kingside with . . . fS, in King's Indian style. White must now make a crucial decision: by retreating his bishop to g3 he clearly intends to challenge the stability of Black's centre with a quick £4, while the retreat to e3 signifies a more patient approach; White will support his centre with f3 and aim for queenside play.

15 �e3 This approach is endorsed by Sa­

kaev. Instead 1S �d2 c4 16 .tc2 b6 17 f4 f6 18 tt:'lg3 iYcS+ 19 �h1 exf4 20 �xf4 tt:'lb7 21 a4 .I:!.ae8 was a surprisingly toothless approach from White in A.Shirov-A.Areshchenko, Foros 2006; Black is clearly fine here.

The major alternative is 1S �g3 and now:

a) 1S . . . fS 16 f4 (of course White must not allow . . . f4)

16 . . . fxe4 (16 . . . exf4! ? is a viable alter­native, after which 17 .I:!.xf4 c4 18 �c2 'ifcS+ 19 .if2 'i¥d6 reaches a complicated position; Black has surrendered some critical squares, such as d4, but White's centre is under attack) 17 i.xe4 tt:'lc4 18 'i¥d3 (18 d6!? i s an interesting idea; then 18 . . . tt:'le3 19 i.xb7 tt:'lxd1 20 .txc8 .I:!.axc8 21 .I:i.bxd1 .tbS sees the forcing sequence leave Black with good compensation in view of his powerful bishops on the open board) 18 . . . tt:'ld6 19 fxeS tt:'lxe4 20 'ifxe4 �fS 21 'i¥c4 .txb1 22 d6+ �h8 23 .I:i.xb1 'iffS was the course of M.Carlsen­D.Navara, Wijk aan Zee 2007. Here White's central pawn duo looks fright-

6 7

Play th e G r ii nfe /d

ening, but Black has his trumps as well, most notably the extra exchange; the position is dynamically balanced.

b) Also possible is 15 . . . c4 when Black must make sure that he can maintain his pawn on e5, preventing the white knight from acquiring the excellent d4-square: 16 .ic2 b6 (16 . . . f5 ! ? i s another option) 17 iYd2 tLlb7 18 f4 f6 ! (essential) 19 'lt>h1 tt:Jd6 20 tLlg1 iYd8 with complicated play in K.Sakaev­B.A vrukh, Izmir 2004. 1s . . .fs 16 exfs

Allowing . . . f4 will result in a smooth kingside attack for Black, so this is the only reasonable move. 16 ... gxfs 11 f3

17 it.g5 just leaves the bishop ex­posed after the obvious 17 . . . iYe8.

Sakaev reaches this position in his book, suggesting that White has a slight but stable edge. Indeed, any ad­vance of Black's e- and f-pawns will lose control of some key central squares, while White can gradually bring his forces over to the kingside. Positionally, White definitely seems to have things going his way, even if only

6 8

slightly, but Black's position contains some very dynamic resources which I feel Sakaev has underestimated. Play might continue: 17 . . . b6 18 iVe1

18 .ig5 c4 19 .ic2 iYcS+ 20 'lt>h1 tt:Jb7 sees Black nicely regrouping his pieces; the knight is coming to d6 and White's position is not very harmonious. 18 .. . f4!

This advance looks rather ugly, abandoning the e4-square and weaken­ing the b1-h7 diagonal, but it does cru­cially interfere with White's plans. In­deed, White's bishop now really has to retreat to f2, but that prevents the op­tion of iYh4. 19 it.f2 c4 20 it.c2

20 it.e4 is very similar. 20 . . . tLlb7

The black knight is en route to ei­ther d6 or c5 (the more aggressive op­tion), after which Black can set his queenside majority in motion. Alterna­tively, he can even play . . . .if5, trying to wrest control of the e4-square with the intention of later advancing with . . . e4 himself. Meanwhile, White's pieces are

Exc h a n g e Va r ia t io n : Th e Fas h io n a b l e 7 .i. c4

distinctly misplaced, particularly the e2-knight, and it' s hard to find a good plan for him. Therefore this position seems quite promising for Black.

822) 11 .l:!.c1 .l:!.c8 Instead 1 1 . . .'Lla5?! 12 it.d3 Wic7 13

'Llf4 it.c6 14 e5 .l:!.ad8 15 W/g4 was prom­ising for White in J.Tarjan-O.Roma­nishin, Indonesia 1983; a good illustra­tion of White's more aggressive ideas.

12 Wid2 There is one significant but defi­

nitely inferior alternative, namely the mi.stimed 12 d5? ! . A.Onischuk­P.Svidler, Halkidiki 2002, continued 12 . . . 'Lle5 13 .ib3 W/a5 14 h3 (14 f4? ! is, of course, met by 14 . . . 'Llg4) 14 . . . W/a6 (this whole plan of fighting for the c4-square was introduced by Griinfeld guru Peter Svidler; Black intends . . . .ib5, gaining complete control of the light squares in White's camp) 15 a4 (both 15 f4 .ib5! and 15 c4 b5 are excellent for Black -Svidler) 1 5 . . . 'Llc4 16 .ig5 and now, as Svidler notes, Black's best continuation is 16 . . . f6 ! 17 it.f4 ( 17 .ih4 .ih6 18 .l:!.c2 'Lld6! with the threat of . . . c4 is very

good for Black) 17 . . . 'Llb2 18 W/d2 'Lld3 19 .l:!.cd1 'Llxf4 20 'Llxf4 c4! 21 .ic2 .ih6 with advantage. 12 ... W/as

The most consistent and combative move, pursuing Black's standard plan, although there is also 12 . . . b5! ? 13 it.xb5 'Llxd4, equalizing. 13 Wib2

Another illustration of the afore­mentioned Svidler plan occurs with 13 d5 'Lle5 14 .ib3 W/a6! and now:

a) 15 .ig5 e6 16 dxe6 .ixe6 17 it.d5 .ixd5 18 exd5 .l:!.fe8 19 'Llg3 'Lld3 20 .l:!.b1 c4 with a clear plus for Black in V.Georgiev-L.Ftacnik, Bled Olympiad 2002.

b) 15 .l:!.fd1 it.b5 16 'Llg3 'Llc4 17 .ixc4 it.xc4 18 .l:!.c2 .l:!.fd8 19 W/e1 W/a4 20 .l:!.ccl b5 and again Black is much better, R.Eynullaeva-E.Vorobiov, Moscow 2002.

c) 15 c4 b5! 16 it.h6 (16 f4?! failed to convince after 16 . . . bxc4 17 fxe5 cxb3 18 axb3 .ixe5 19 .l:!.xc5 Wid6 20 .l:!.xc8 it.xh2+ 21 'lt>h1 .l:!.xc8 in K.Murugan-B.Sivanan­dan, Chennai 2004) 16 . . . it.xh6 (16 . . . bxc4 17 .ixg7 'lt>xg7 18 kc2 'Lld3 19 .l:!.b1 .l:!.b8 also worked out well in D.Komljenovic-

6 9

Play th e G r il nfe ld

E.Vorobiov, Cappelle la Grande 2003) 17 �xh6 bxc4 18 �c2 l:tb8 and Black's position is preferable.

White also has 13 l:tfd1 when Black can maintain the balance with Ep­ishin's 13 . . . e6 or completely equalize with the forcing sequence 13 . . . cxd4 14 cxd4 �xd2 15 .txd2 lbxd4 16 lbxd4 .ixd4 17 �h6 .tb2 18 �xf8 .ixcl 19 .txe7 lhc4 20 l:txd7 l:txe4 21 �fl, when White regains his pawn and the result­ing ending should be drawn. 13 ... b6 14 f4 e6

A necessary precaution against the aggressive f4-f5 . 1S l:tfd1

Instead, 15 lbg3 cxd4 16 cxd4 was tried in A.Kaminik-V.Mikhalevski, Givatayim 2004, and now after 16 . . . l:tfd8!? 1 7 �f2 �a3, intending . . . lba5, Black has the initiative. 1s ... l:!.fd8 16 dxcs .tf8!

Activating the bishop and hoping to weaken White's dark squares.

17 cxb6 17 �5? ! �xeS 18 .ixc5 bxc5 19 l:tb1

.te8 20 �xa5 (or 20 �f2 l:txd1 21 l:txd1

7 0

lib8 with the initiative) 20 . . . lixd1 + 2 1 l:txd1 lbxa5 was seen in M.Prusikin­P.Svidler, German League 2002; Black is better in view of his potentially more active pieces. 17 ... �cs 18 �d2 axb6 19 lbd4 �a8!?

Krasenkow considers Black to have good compensation. Indeed, White has several weaknesses and his centre, par­ticularly the e4-pawn, seems rather loose, while Black may play, for instance, . . . lba5 and . . . .ia4. All in all, it looks like White's pieces are in a bit of tangle.

Con cl usion

I cannot emphasize enough the neces­sity of knowing the lines in this chapter well! Black can easily meet Line A by remembering the important manoeuvre . . .�a5-a3-d6 and its main ideas. Overall, White's position after 10 l:!.cl seems generally unsound to me. The main line with 10 0-0 is a totally different story. Then 10 . . . lba5 has only recently become popular, but seems to offer Black good prospects, at least on the basis of the available material. I have spent a con­siderable amount of time examining the resulting positions and feel satisfied with Black's play, but there is no doubt that the theoretical debate will continue - keep an eye on current tournament practice! Meanwhile, 10 . . . .td7 has sur­vived several tests at the highest level and is a reliable option. The very fact that it is regularly employed by both Svidler and Sutovsky suggests that it is fully sound.

·�.�.��� �·

c h a pte r F ive I ::::;:z: � ��· '�

��W//� � � m r� � rs 11. � � ff� 11. p-.,!_1;

Exc h a n ge Va riat io n : § ����t�Jt1§ Wh ite's 7th Move Alte rnat ives

1 d4 CLJf6 2 c4 g6 3 CLJc3 ds 4 cxds lbxds 5 e4 lbxc3 6 bxc3 i..g7

We shall now examine four at­tempts by White to stray from the beaten path:

A: 7 i..a 3 B: 7 'iWa4+ c: 1 i..bs+ o: 1 i..gs

While these variations have not ac­quired mainline status, they do merit attention as they represent logical and strategically sound systems of devel­opment. The common thread running through them is the intention to dis­rupt Black's natural development scheme of . . . c5, . . . lbc6 and . . . i..g4. Achieving this aim enables White to develop without concerns about his pawn on d4, after which his central superiority should ensure a slight but steady plus. Naturally, Black is not

helpless against these ideas and, once again, accurate and purposeful play is the key to success.

A) 7 .ia3

By making . . . c5 harder to achieve, White hopes to buy enough time to develop comfortably and support his centre. 7 . . • CLJd7

Black in turn argues that although he will be unable to pressurize the cen­tre with . . . lbc6 and . . . i..g4, the a3-

7 1

Play t h e G r ii nfe ld

bishop is badly placed. As Black's posi­tion is very flexible, he can seek other means of troubling White. 8 tt:lf3

Instead 8 l::.cl c5 9 d5 'iUaS 10 'iUb3 0-0 1 1 tt:lf3 was seen in G.Garcia Gonza­les-J .Lechtynsky, Cienfuegos 1985, and now 1 1 . . .tt:lf6 12 .id3 tt:lh5 gives Black the initiative . s . . . cs

White usually now chooses between two aggressive-looking tries:

A1: 9 �c4 A2: 9 'iUb3

Less good than these is 9 'iUd2?! since after 9 . . . 'iUa5 10 .tb2 tt:lb6 11 �e2 ( 1 1 l::.d 1? is erroneous: 1 1 . . . .tg4 12 a3 0-0 13 �e2 l::.ac8 14 h3 .txf3 15 gxf3 tt:la4 with a clear advantage for Black) 1 l . . . .tg4 White faces some pressure.

A1) 9 .ic4 After this move Black must respond

accurately. 9 . . . cxd4! 10 'iUb3

7 2

Instead 10 0-0? dxc3! 1 1 ii.xf7+ ( 1 1 tt:lg5 0-0 with two extra pawns) 1 1 . . .Wxf7 12 tt:lg5+ We8 13 tt:le6 (or 13 'iUb3 tt:le5) 13 . . . 'iUb6 14 tt:lxg7+ Wf7 just wins for Black and 10 .txf7+? Wxf7 1 1 tt:lg5+ We8 12 tt:le6 'iUaS 13 tt:lxg7+ Wf7 is also very good for Black. 10 .. . 0-0 11 cxd4 tt:lb6

Now White has to decide whether or not to allow the exchange of his bishop. 12 .ie2

The alternative is 12 0-0 tt:lxc4 13 'iUxc4 b6 14 l::.acl (14 l::.ad1 .tg4 is equal) 14 . . . .ig4 15 tt:le5 .ixe5 16 dxe5 l::.e8!? when Black has an equal share of the chances in this unbalanced posi­tion. He similarly had no problems after 16 . . . l::.c8 17 'iUdS .te6 18 'iUxd8 l::.fxd8 19 �xe7 l::.xcl 20 l::.xcl l::.e8 21 l::.c7 (or 21 .t£6 �xa2 22 f4 aS with the initiative) 21 . . .l::.c8 22 l::.xc8+ �xc8 in B.Lalic-D.Gormally, British Champion­ship, Torquay 2002. 12 .. . �e6 13 'iUb4 l::.c8 14 0-0

Black is at least equal after 14 'iUxe7 'iUxe7 15 .ixe7 l::.fe8 16 .ih4 (16 ii.d6 .tc4 17 e5 f6 is equal, while Black is

Exc h a n g e Va r ia t io n : Wh ite 's 7th M o ve A l tern a t ives

better after both 16 i..g5 i..c4 1 7 e5 f6 and 1 6 i..a3 �c4 1 7 e5 f6) 1 6 . . . i..c4 1 7 e5 f6. 14 . . . �c4

15 �fe1 After 15 �xc4 'Llxc4 16 �acl b5 !

(Black has achieved a good position) 1 7 'iVxb5 ( 17 'iVb3 'iVd7 18 i..c5 a 6 seems fine for Black) 17 . . . 'Llxa3 18 'iVa4 e5 ! (the only move, but a good one! ) 19 �xc8 'iVxc8 20 'iVxa3 exd4 21 'iVxa7 d3 22 'iVe3 the position should eventually be drawn, although there is some play left. 15 ... i..xe2 16 �xe2 'Llc4 17 �c1

Instead 17 �d1 'iVd6 18 'iVxd6 ( 18 'iVxc4 �xc4 19 i..xd6 exd6 is also better for Black) 18 . . . exd6 19 i..cl �fe8 20 i..f4 b5 favoured Black in V.Milov­P.Svidler, FIDE World Ch., Moscow 2001 . With the text, though, White should be able to maintain the balance; for example, 17 . . . 'Llxa3 18 �xc8 'iVxc8 19 'iVxa3 �d8 20 e5 a6 21 'iVxe7 (or 21 h3 e6) 2l . . .�xd4! 22 �e1 'iVd7 is equal.

A2) 9 'iVb3 0-0

10 �e2 Other tries fail to impress: a) 10 i..d3 'iVc7 11 0-0 �b8! 12 .ib5

b6 13 �ad1 a6 14 .id3 b5 15 .ib1 .ib7 was good for Black in L.Evans­V.Korchnoi, Buenos Aires 1960.

b) 10 i..c4 �b8 11 e5 ( 1 1 0-0 b5 12 i..e2 i..b7 13 'iVc2 b4 14 cxb4 cxd4 15 �ad1 �c8 also looks quite good for Black) 1 1 . . . b5 12 �xf7+ �xf7 13 e6 �xf3 14 exd7+ c4 15 dxc8'iV 'iVxc8 16 'iVc2 'iVe6+ 1 7 �fl �f7 and White was in trouble in W.Heidenfeld-A.Tsvetkov, Marianske Lazne 1951 .

c ) 10 �d1 cxd4 1 1 cxd4 'Llf6 12 .id3 (12 e5? ! .ie6 13 'iVxb7 'Lld5 14 'iVa6 i..c8 is again good for Black) 12 . . . i..g4 13 'iVxb7 'iVa5+ 14 i..b4 (Black is better, too, after 14 'iVb4 'iVh5 15 0-0 .ixf3 16 gxf3 'iVxf3) 14 . . . 'iVh5 15 �xe7 �fb8 16 'iVc7 �xf3 1 7 gxf3 'iVxf3 18 0-0 'Llh5 and White's king is a constant source for concern. 10 . . . iVc7

Another idea is 10 . . . �b8! ? intending . . . b5; for example 1 1 0-0 b5 12 �ad1 a6. 11 o-o �b8 12 es e6 13 'Lld2

73

P lay t h e G r u nfe /d

This position was reached in I.Papaioannou-M.Krasenkow, European Team Ch., Leon 2001, and now Black should play 13 . . . b5 ! 14 lLle4 b4 15 cxb4 i.b7 16 lLlxc5 (or 1 6 lLld6 cxd4 17 l::r.fcl '1ifh6 18 b5 �d5 19 i.c4 i.xe5 20 �xd5 �xd6 21 :c6 'i!Va5 22 �xd6 l:txb5 23 'i!Vh3 lhd5 24 �xf8 lLlxf8 with good prospects) 16 . . . lLlxc5 17 dxc5 (after 17 bxc5 i.xg2 18 �e3 �xfl 19 lhfl .l::!.fe8 Black will break with .. . £6) 17 . . . i.xe5 18 i.b2 �xh2+ 19 'itrh1 �f4 20 l:tfd1 �e7, when he intends . . . l:tfd8 and . . . e5 with good play.

B) 7 �a4+

74

This disruptive check has recently acquired a certain popularity, chiefly due to the efforts of Vadim Milov and Mikhail Gurevich. It is actually a quite dangerous move - dangerous in a sneaky way! White's aim is to entice Black to misplace his pieces, so that he will later be unable to apply enough pressure on White's centre . Black needs to react accurately, but should he do so, I believe he is able to highlight the main drawback of the check, namely that the white queen is no longer sup­porting d4. 7 . . . �d71

In my mind, clearly best, whereas other moves allow White to develop comfortably.

Now White must choose between:

81: 8 'i!Vxd7+ BZ: 8 'i!Vb3

Instead 8 �a3 is similar to 8 'i!Vb3 but, I feel, less accurate . Black should then immediately prepare . . . c5: 8 . . . b6 9 lLlf3 i.b7 10 �d3 c5 1 1 i.e3 0-0 12 0-0 (instead, the 12 l:id1 of E.Sevillano­A.Yermolinsky, Stratton Mountain 1999, should be met with 12 . . . e6 13 dxc5 'i!Vc7 14 0-0 lLld7 with equality -Yermolinsky) 12 . . . cxd4 13 cxd4 e6 14 l::r.fd1 lLlc6 (A.Szeberenyi-D.Howell, Budapest 2004) when Black has devel­oped comfortably and stands fully equal.

81) 8 �xd7+ The queen exchange is, of course,

Exch a n g e V a ri a t i o n : Wh i t e ' s 7 th M o ve A l te rn a t i ve s

fundamentally critical. Let's not forget that a strong centre can be an asset, not only in a complicated middlegame but also in a more simplified endgame. Black is able to strike back and secure good play, but he must play consis­tently. 8 • • • tt:Jxd7!

This is the best recapture. Black will not be able to pressurize d4 in the standard Gri.infeld manner, but he can pursue other plans, particularly the attack on e4, but also the occupation of c4 with . . . tLlb6 and .. . i.e6. Another idea is . . . b6 and . . . i..b7 when White has ei­ther to advance his e-pawn, thereby handing Black the d5-square, or protect it with i!.d3, leaving the bishop ex­posed to tactics. 9 tLlf3

Definitely not the only move: a) 9 i.e3 intends to meet . . . i!.b7 with

f3, firmly supporting e4. However, this plan somewhat clogs up White's king­side development: 9 . . . 0-0 10 i.c4 c5 1 1 tLle2 tLlb6! ? (I very much like this idea, gaining time to develop counterplay, whereas 1 l . . .b6?! is inferior due to 12

e5! intending e6) 12 i.b3 cxd4 13 cxd4 i.d7 and I feel that Black already has the initiative.

b) 9 i.a3 c5 10 tLlf3 b6 leads to equal­ity, as the a3-bishop is shut out of play.

c) 9 i.d3 b6 10 i.e3 i.b7 11 tLle2 0-0 12 0-0 �fd8 (now Black has tactical ideas based on the loose state of the d3-bishop) 13 �ad1 (13 �fd1 c5 has the idea of 14 a4 cxd4 15 cxd4 tLlc5 with advantage, as suggested by Huzman and Vainerman) 13 . . . c5 14 e5 �ac8 (as Huzman and Vainerman point out, no good is 14 . . . i!.d5?! 15 tLlf4 i.xa2 16 �a1 i.b3 17 �fb1 c4 18 ii.e4 �ac8 19 �xa7) 15 f4 cxd4 ( 15 . . . e6 16 g4 was agreed drawn here in Y.Piskov-A.Huzman; Huzman suggests that Black has good play after 16 . . . i!.h6, but I think that 17 tLlg3, intending tLle4, promises White an edge) 16 cxd4 i.d5 with comfortable equality. 9 . . . b6

10 i!.c4 Instead 10 i!.d3 tt:Jc5 1 1 i.c2 i.a6 12

i!.e3 tLld3+ 13 '.t>d2 tLlb2 14 '.t>cl tLlc4 15 i!.a4+ b5 (better than 15 . . . '.t>f8 16 i.c6 with an edge, although actually agreed

75

Play t h e G r u nfe ld

drawn here in I .Khenkin-P.Jaracz, An­dorra 2001 ) 16 ..ib3 lt:Jxe3 17 fxe3 gives a complicated position with chances for both sides. 10 . . . �b7 11 0-0 0-0

Now Black threatens . . . ..ixe4. 12 es e6 13 tt:Jgs �fd8 14 f3 cs

Black has fully equalized, S.Volkov­B .Macieja, Batumi 2002.

82) 8 �b3

White's main move. He hopes to benefit from Black's 7 . . . �d7, which rules out . . . �g4 and practically forces the light-squared bishop to be devel­oped to b7, to maintain his centre. As

7 6

we shall see, however, the queen can become a target on b3. 8 . . . 0-0

Black should now opt for a queen­side fianchetto, followed by a harassing . . . lt:Jc6-a5 . 9 �e3

9 lt:Jf3 b6 10 i..e3 transposes, but less good is 10 i..b5 c6 11 i..c4 c5 12 d5 ..ia6 13 a4? ! �g4 14 0-0 �xe4 15 ..ixa6 lt:Jxa6 16 �e1 �d3 with an edge for Black, S.Volkov-M.Turov, Sochi 2004. 9 . . . b6 10 tDf3

Playing instead for f3 fails to con­vince: 10 �d1 i..b7 1 1 f3 lt:Jc6 12 �a4 �fd8 13 ttJh3 lt:Ja5 14 �xd7 �xd7 15 �f2 e5 16 i..e2 �ad8 and Black was better in V.Milov-P.Eljanov, Geneva 2004.

10 . . . lt:Jc6!? A rare idea that I have investigated

and which looks promising to me! In­stead 10 . . . .ib7 is more common, but White can then achieve a harmonious set-up with 1 1 .id3 lt:Jc6 12 �d1 lt:Ja5 13 �1 c5 14 0-0 . In my opinion, 10 . . . lt:Jc6 is the most annoying move order for White, as he would like to meet . . . l2Ja5 with �1, but without blocking the

Exch a n g e Va ri a t i o n : Wh i te 's 7 th M o ve A l t e rn a t i ve s

�al . Furthermore, Black does not commit his bishop to b7; as we shall see, it may find greener pastures! 11 �C1

Securing the retreat of the queen to bl . Instead after 11 .id3 tLla5 12 'iVc2 c5 13 �d1 .ib7 14 0-0 �ac8, the threat of . . . cxd4 forces White to move his queen again.

White might also try 1 1 �d1, hoping for a transposition to the 10 . . . .ib7 variation. However, Black can then remain in independent waters with 1 1 . . .4Ja5 12 'iVb1 (or 12 'iVc2 c5 13 .td3 cxd4 14 cxd4 'iVd6 with good play; note that Black has retained the options of . . . �g4 and . . . .te6) 12 . . . c5 13 .td3 'iVc7 (again with the idea of . . . .ig4) 14 0-0 (or 14 h3 .ie6, aiming for the c4-square) 14 . . . .tg4 15 �fe1 .txf3 16 gxf3 �ad8 and Black stands very well, in view of White's kingside weaknesses and lack of a good plan. 11 ... .ib7

The immediate 11 . . . 4Ja5 !? is also en­ticing: 12 'iVc2 (12 'iVb1 c5 13 dxc5 bxc5 14 .ixc5 'iVc7 offers the standard com­pensation) 12 . . . c5 13 .td3 .tb7 and Black, with ideas of . . . c4 and . . . f5, has assumed the initiative. 12 .id3

12 'iVb1 retreats without being prompted, in order to meet . . . tLla5 with .tb5. However, this further loss of time enables Black, who has practically com­pleted his development, to strike in the centre: 12 . . . e5 13 d5 tLle7 14 c4 c6 15 �d3 f5 and Black is in the driving seat. 12 ... tt:Jas 13 'iVb1 cs

Also promising is 13 . . . 'iVg4!? 14 0-0

f5 15 4Jd2 e5 with the initiative. 14 o-o �adS

This position is rather complex, al­though the white rook does seem a lit­tle misplaced on cl .

C) 7 .ibS+ Another disruptive check, but yet

again Black is not unduly worried, es­pecially since this move, in comparison to 7 'iVa4+, prematurely commits the bishop, while the check can be met in a natural way. 7 . . . c6

Black does not renounce . . . c5 at all, but simply delays it until after he has castled. He will then try to argue that the white bishop is misplaced on a4. 8 .ia4 0-0 9 tt:Je2

Of course, 4Jf3 would be strongly met by . . . .ig4 at any point. g . . . cs

So Black has achieved . . . c5 after all! The question now is: how well is the bishop placed on a4? 10 0-0

Unsurprisingly, 10 d5?! is prema­ture: 10 . . . e6 ! 11 .te3 tLla6 (now the

7 7

P lay th e G r u nfe /d

pawn on d5 will come under heavy pressure) 12 �b3 exd5 13 .ixd5 tDc7 14 .ltc4 (14 .ixc5 .l:i.e8 15 0-0 tt:lxd5 16 exd5 b6 17 .lte3 �a6 gives Black excellent compensation as the white pieces are unpleasantly pinned) 14 . . . 't!Ve7 15 f3 b5 16 i.b3 c4 17 i.c2 b4 (White is already struggling to hold the balance) 18 cxb4 (18 .l:i.b1 ?! bxc3 19 0-0 .l:i.d8 20 't!Vcl �a6 21 a4 tt:le6 22 f4 tt:ld4 was disastrous for White in M.Ulibin-S.Dvoirys, Geneva 2001 ) 18 . . . .l:i.d8 19 't!Vcl i.xa1 20 't!Vxa1 't!Vxb4+ 21 �f2 'ti'a5 and Black is better; he will cover the kingside dark square weaknesses with . . . tt:le8, while the passed c4-pawn will become a strong asset after . . . .lte6 and . . . .l:i.ac8. 10 ... tt:lc6 11 .lte3 tt:Ja s

The position is almost identical to a very important one examined in the 7 i.c4 system, except that here the c4-square is open to occupation. Now White has two main options:

C1: 12 dxcs (2: 12 .l:i.b1

78

C1) 12 dxcs tt:lc4 12 . . . i.g4!? 13 f3 i.e6 intending

. . . tt:lc4 is a way to play for more than equality; Black has the standard com­pensation. 13 'ti'xd8 .l:i.xd8 14 .tgs .lid7 !

Black invites exchanges, a s the weak white queenside pawns will pro­vide good endgame targets. 15 i.b3 tt:Jas 16 i.xe7

Worse is 16 .l:i.fd1 ? ! .ltc6 17 f3 ..if8 18 e5 .l:i.xd1+ 19 .l:i.xd1 e6 20 .lte3 .l:i.c8 21 tt:l£4 .ie7 22 c4 i.e8 23 ctJd3 tt:lc6 24 f4 ctJb8! with the advantage for Black, B.Avrukh-I.Sokolov, Amsterdam 1999; after . . . tt:la6 the cS-pawn will drop. 16 .. . .l:i.e8 17 i.d6 .l:i.xe4 18 tt:ld4 tt:lxb3 19 axb3 .ltxd4

Here a draw was agreed in V.Topalov-G.Kasparov, Linares 1998.

C2) 12 .l:i.b1!? More ambitious . White tries to

hamper Black's development. 12 .. . b6 !

Black replies in tune! This move amounts to a pawn sacrifice, but one which is very promising in my opinion.

Exch a n g e Va ri a t i o n : Wh i te 's 7 t h M o ve A l t e rn a t i ves

Also playable and fully acceptable is 12 .. .<:t:Jc4 again: for example, 13 .ltg5 h6 14 .lth4 (V.Topalov-P.Svidler, Linares 1998) 14 . . . a6 15 .ltb3 tLla3 16 M.a1 cxd4 17 cxd4 tLlb5 with equality.

13 dxcs Accepting the challenge. Instead 13

.ltb5 tries to bring the bishop back to d3 and is rather an admission of failure . B .Avrukh-J .Timman, Amsterdam 1999, continued 13 . . . .\te6 14 d5 i.d7 15 .\td3 e5 16 f4 and now Black should opt for 16 . . . exf4! (and not Timman's risky 16 . . . £5? ! 17 tLlg3 c4 18 i.c2 g5 19 fxe5 £4 since A vrukh could now have played the strong 20 i.d4! fxg3 21 M.xf8+ 'iiVxf8 22 hxg3 with excellent compensation) 17 .ltxf4 .ltg4 with chances for both sides, or even 17 . . .'�e7! ? . 13 . . . i.a6 14 .tbs .ltxbs 15 M.xbs l2Jc4

Black has occupied the target square and stands well; for example, 16 i.g5 (E.Gleizerov-T.Oral, Stockholm 1999) 16 .. .'�c7 17 'iiVa4 a6! (or Timman's 17 . . . 4Ja5 18 'iiVa3 'iiVe5 !, with the initia­tive) 18 M.bb1 b5 19 'ilVb4 M.fc8 with an obvious advantage, as White cannot save the c5-pawn.

o) 1 .tgs!?

An interesting sideline, introduced by Yuri Kruppa. 7 . . . cs !

As White has forfeited the option of supporting d4, this natural move is also the most consistent one. 8 MC1 !?

White prepares to meet . . . 4Jc6 with d5 since he cannot support d4. 8 . . . o-o 9 ds

Kruppa and Komarov have sug­gested 9 tLl£3, but White cannot main­tain the pawn on d4 in any case : 9 . . . i.g4 10 d5 (or 10 'ilVb3 'ilVc7 1 1 'iVa3 4Jd7 12 i.e3 i.x£3 13 gxf3 e5 with both

79

Play th e G r iJ nfe ld

the initiative and certain positional pluses) 10 .. .<�:Jd7 when Black is all set to break with . . . f5 and White is forced on to the defensive . 9 . . Ji'd6!?

I like this move, unpinning the e­pawn. Black intends to exploit his lead in development to break up White' s centre with . . . e6 and/or . . . f5 . 10 tiJf3 e6

10 . . . tDd7!? is possible, intending .. . e6. 11 i.c4 es

Not really a change of heart, but an attempt to prove White's minor pieces misplaced; . . . f5 is coming next. 12 tiJd2

Instead 12 0-0 h6 13 .lte3 (13 il.h4 tt:Jd7 14 tiJd2 tiJb6 is good for Black, who plans . . . il.d7, . . . .:ae8 and .. . f5) 13 . . . £5 14 .ltd3 f4 15 .ltd2 g5 16 h3 Vig6 leads to a massive King's Indian-style pawn-storm on the kingside. Thus the text was pre­ferred in B .Taborov-D.Dzupin, Kiev 2006, and now 12 . . . h6 13 .lte3 (13 i.h4 tt:Jd7 - not 13 . . . £5? 14 il.e7! - 14 0-0 tiJb6 15 i.g3 £5 also gives Black the initiative) 13 . . . £5 14 exf5 gxf5 15 0-0 tiJd7 has the obvious intention of . . . e4 and . . . tt:Je5.

8 0

Black has active and good play and stands at least equal.

Conclusion

The lines in this chapter are by no means harmless. Simplistic natural play by Black will not suffice for equal chances -concrete knowledge is required, but Black can equalize with correct play without too much effort. After 7 i.a3, Black should insist on playing . . . c5, after which the opening debate revolves around the functionality of the bishop on a3; Black usually emerges successful.

The two checks on a4 and b5 are quite different from each other. 7 .ltb5+ is, in my opinion, an overrated line, as Black can obtain a good, normal Gri.in­feld position with standard counterplay. However, 7 Via4+ is the most critical line in this chapter. The queen exchange on d7 should not present Black with any problems, but 8 'i¥h3 is really danger­ous. Hopefully, my suggestion against it will stand the test of time and solve the slight problems Black has recently been facing in this line!

Ch a pte r S ix I 4 cxd 5 ct:Jxd 5 wit h o ut 5 e4

1 d4 'Llf6 2 c 4 g6 3 'Llc3 d5 4 cxd5 'Llxd5 Apart from the standard 5 e4, White

has a few other ideas at his disposal. These chiefly aim to strengthen the im­pact of the coming e4 by preventing or discouraging a knight exchange on c3.

A: 5 'Lla4 B: 5 j,d2 C: 5 'Llf3

A) 5 'Lla4

This extraordinary idea was intro­duced by the Armenian grandmaster, Ashot Nadanian. White wants to play e4 without allowing an exchange on c3, while also preventing . . . cS. The move is not entirely harmless, but it can hardly prove effective should Black respond dynamically. 5 ... e5 ! 6 dxe5

Instead, 6 a3 e4 is not what White wanted, while after 6 e4 'Llf6 he comes under some pressure in the centre:

a) 7 il..gS exd4! 8 eS (8 \ib3 h6 is also very good for Black) 8 . . . il..b4+ 9 'Llc3 (as Rowson observes, 9 'lt>e2? Vi'dS! is strong) 9 . . . 'iad7! 10 ..txf6 dxc3 (Rowson) and Black wins.

b) 7 dxeS 'iaxd 1 + 8 'lt>xd 1 'Llxe4 9 .i.e3 .i.d7 10 'Llc3 'Llxc3+ 1 1 bxc3 il..a4+ and Black's positional advantage, due to his better pawn structure, is undeni­able. 6 ... 'Llc6!

I believe Black does not need to do anything drastic yet. Simple, active

8 1

P lay t h e G r iJ nfe ld

development should do the job, as any attempt to protect the e5-pawn is bound to be weakening, while White's only developed piece is the knight on a4.

At this point White has two princi­pal continuations:

Al: 7 a3 A2 : 7 tiJf3

Instead 7 .id2 should be met by 7 . . . tt:Jxe5! ? 8 e4 (8 �3 .ie6) 8 . . . tt:Jb4 9 .ic3 tt:Jed3+ 10 .ixd3 tt:Jxd3+ 1 1 'it>fl �g8 intending . . . b5, with advantage for Black, and 7 e4 .ib4+ (7 . . . tt:Jdb4 can also be played) 8 .id2 tt:Je3 ! ? also looks promising for Black, with the idea 9 fxe3?! 'ifu4+ 10 g3 'ifxe4 1 1 tZ:lf3 .ig4 12 .ig2 .ixf3 13 .ixf3 .ixd2+ 14 'ifxd2 'ifxf3 15 �fl tt:Jxe5 ! 16 'ifd4 'ii'h5 and wins.

Al) 7 a3 This is really taking things a bit too

far. I just cannot accept that White can play the opening like this!

8 2

7 ... .\tfs ! 7 . . . tt:Jxe5 is not in our interests; de­

velopment is first and foremost the main goal. 8 tt:Jf3 Vi\Vd7

It already looks like White is head­ing for a miniature! 9 e3

Not forced, but in any case Black has fantastic compensation for the pawn and a very strong initiative:

a) 9 .ig5 h6 10 .ih4 (or 10 i.f6?! tt:Jxf6 11 exf6 Vi\Vxd1+ 12 '>t>xd1 0-0-0+) 10 . . . g5 1 1 .ig3 0-0-0 (Kantsler) gives Black promising play.

b) 9 g3 0-0-0 is good for Black. c) 9 .id2 0-0-0 10 �cl Vi\Ve8 was also

promising for Black in J .Rukavina­A.Mikhalchishin, Nova Gorica 1999.

d) 9 tZ:lg5 h6 10 e4 hxg5 11 exd5 tt:Jxe5 ( 1 1 . . .0-0-0 ! ? is suggested by Ein­gorn; after the logical 12 .ixg5 .ie7 13 dxc6 Vi\Vxc6 14 'ifcl .ixg5 Black has the initiative) 12 .ixg5 .ie7 13 VilVd4 i.xg5 14 'ifxe5+ 'ife7 15 .ib5+ �f8 16 Vi\Vxe7+ 'it>xe7 with excellent compensation for Black. 9 ... 0-0-0 10 .ie2

Or 10 i.b5 a6 (Eingorn's 10 . . . 'ib'e6 is also good; for example, 1 1 tt:ld4 tt:lxd4 12 exd4 f6 with the initiative) 11 i.xc6 'ib'xc6 12 ctJd4 'ib'e8 13 f4 (13 0-0 'ib'xe5 is much better for Black) 13 . . . f6 and I would be surprised if White can sur­vive the early middlegame. 10 ... 'ib'e7 11 iVb3

Note that 1 1 tt:ld4? tt:lf4! is a power­ful shot, while 1 1 i.d2 tt:lxe5 12 'iVb3 i.g4 is quite good for Black. 11 ... i.g7

Black has a powerful initiative, B.Kantsler-B.Avrukh, Israel 1999.

A2) 7 ctJf3 i.g4!? Instead 7 . . . tt:ldb4?! was played in the

fairly well-known game S.Lputian­A.Shirov, Montecatini Terme 2000, but as Eingorn points out, it is not good. After 8 a3! 'ib'xd 1 + 9 Wxd 1 tt:la6 Black can only fight for a draw, and in par­ticular 10 b4 i.e6 1 1 Wc2 i.f5+ 12 Wb2 i.g7 13 ii.f4 seems very bad for Black to me.

The text move is an interesting sug­gestion of Eingorn' s, while another of his ideas is 7 . . . ii.f5 8 tt:lc3 (8 a3 'ib'd7

4 cxds ti:Jxds with o u t 5 e4

leads us back to Line A1) 8 . . . tt:lxc3 9 'ib'xd8+ l:txd8 10 bxc3 i.g7, which he assesses as equal . In fact, after a rea­sonable continuation such as 11 i.g5 l:td5 12 i.f6 0-0 13 e3 l:tc5 Black has the initiative . This is reason enough to con­demn White's opening play, but 7 . . . �g4 appeals to me even more!

8 a3 'ib'd7 9 lLlc3 Black is also rather comfortable af­

ter 9 h3 i.xf3 10 gxf3 0-0-0; for exam­ple, 1 1 e4 tt:lf6 12 'ib'xd7+ tt:lxd7 13 i.g5 (13 f4 ctJd4 is bad for White) 13 . . . l:te8 regaining the pawn, while White is left with an inferior pawn structure. 9 ... tt:Jxc3 10 'ib'xd7+ �xd7 11 bxc3 i.g7

Black has excellent positional com­pensation for the pawn and promising long-term chances.

B) 5 ii.d2 Yet another system which has been

popular of late. White intends to recap­ture on c3 with the bishop, thereby hoping to counterbalance the activity of the Griinfeld bishop on the long di­agonal . However, this move is some­what unnatural and that prompts Black

83

Play t h e G r il nfe ld

to avoid the exchange on c3, after which White will have to spend a tempo moving the bishop again. s ... i.g7 6 e4

Note that 6 tt:lf3 is covered below in Line Cl . 6 ... tt:lb6 7 �e3 o-o

Black has two principal ideas to at­tack the white centre: he can pressurize the d4-pawn with . . . tt:lc6 and . . . �g4, or strike against e4 with . . . f5 . The proper selection of plan depends on White's choice at this important crossroads:

81: 8 h3 82: 8 f4 83 : 8 1i.bs 84: 8 �e2

Before we move on to White's main options, let' s take a look at some rare alternatives:

a) 8 tt:lf3 �g4 9 �e2 tt:lc6 transposes to Line B42.

b) 8 a4? ! weakens White's position for no reason: 8 . . . a5 ! 9 �e2 tt:lc6 10 d5 tt:lb4 11 tt:lf3 ( 1 1 l:tcl f5 12 tt:lb5? c6 13

8 4

dxc6 'l!Uxd1 + 14 �xd1 tt:lxa4 was a disas­ter for White in L.Alburt-J.Benjamin, New York 1993) 1 l . . . f5 with an edge.

c) 8 l:tcl c6 9 i.e2 tt:la6 10 tt:lf3 �g4 1 1 0-0 tt:lc7 12 'l!Ud2 i.xf3 13 gxf3 'l!Ud7 left the white king under attack in P.Iotti­N.Davies, Porto San Giorgio 1999.

d) 8 'l!Ud2 intends to further support d4, but is inefficient: 8 . . . tt:lc6 9 l:td1 (or 9 0-0-0 e5 10 d5 tt:ld4 1 1 tt:lf3 �g4 12 i.e2 �xf3 13 gxf3, M.Mason-T.Mirabile, Philadelphia 2006, and now 13 . . . c6 with the initiative) 9 . . . e5 10 d5 tt:ld4

1 1 f4 ( 1 1 �xd4? exd4 12 tt:lb5 :S.e8 was disastrous for White in Z.Gorecki­F .Windekilde, Copenhagen 2001, as was 1 1 �d3 f5 12 �g5 'l!Ud6 13 f3 fxe4 14 fxe4 c6 in D.Kljako-I.Balinov, Bled 1998) 1 1 . . .1i.g4 12 fxe5 �xd1 13 �xd4 �h5 and Black is better, as in the long run White will be unable to support his e5-pawn.

e) Finally, the exotic 8 h4 didn't turn out well in F .Vallejo Pons-P.Svidler, Morelia 2006: 8 . . . tt:lc6 9 e5 �f5 (9 . . . h5! ? i s another option) 10 i.e2 ( 10 g4 �e6 1 1 h5 tt:ld5 is no improvement) 10 . . . tt:lb4 1 1 l:tcl ( 1 1 g4 tt:lc2+ 12 �fl tt:lxe3+ 13 fxe3

..lie6 was the game, with a clear plus for Black) 1 l . . .cS! and White's centre is col­lapsing.

81) 8 h3? ! This prevents . . . ..lig4, but again costs

White valuable time. s . . . fs !

9 exfs Instead 9 l2Jf3 (9 'i¥d2 fxe4 10 l2Jxe4

i.e6 and . . . .idS is good for Black) 9 .. .f4 sees Black change direction and focus on d4 once more: 10 ..lid2 (10 ..licl l2Jc6 11 dS ltJeS 12 .i.e2 e6! was good for Black in I.Bitansky-R.Romon Paves, Aviles 2000) 10 . . . l2Jc6 11 dS (11 'ikb3+ 'iii>h8 12 l2Je2 aS 13 i.c3 l2Jb4!? 14 i.xb4 axb4 1S �xb4 eS is promising for Black) 1l . . .l2Je5 12 ltJxeS (or 12 �3 aS 13 a4 l2Jxf3+ 14 gxf3 e6 1S i.c4 fie7 with a clear plus, Y.Zim­merman-C.Pedersen, Budapest 2002) 12 . . . ..lixeS 13 'ifb3 (Y.Zimmerman­S.Atalik, Bled 1999) 13 . . . e6 14 dxe6 'i¥e8 1S e7+ .l:!.f7 16 0-0-0 �xe7 and with . . . i.e6 coming, Black is better. g ... .txfs! 10 l2Jf3

10 'ifb3+ 'iii>h8 1 1 g4 i.d7 12 i.g2 l2Jc6 13 l2Jge2 ltJaS 14 'ii'd1 l2Jbc4 was

4 cxd5 CiJxd5 with o u t 5 e4

very good for Black in S.Peric-L.Milesi, Cannes 1997. 10 .. . l2Jc6 11 �b3+

White also comes under pressure after 1 1 .i.e2 'ii'd6 (or simply 1 1 . . .eS ! ? 12 dxeS 'ii'xd1+ 13 l:!.xd1 ltJxeS with equal­ity, V.Ruban-P.Leko, Tilburg 1993) 1 2 0-0 .l:!.ad8 13 .l:!.cl (S.Wehmeier­A.Greenfeld, Lippstadt 2004) and now 13 . . . 'iib4 prepares . . . eS . 11 . . .'ii?h8

12 0-0-0 Instead 12 dS ltJeS 13 ltJxeS i.xeS 14

.l:!.d1 aS 1S g4 (both 1S a3? ! a4! and 1S i.d3 a4! 16 'iib4 a3 17 i.d4 �d6! give Black an edge) was seen in B.Predojevic­M.Rade, Neum 200S, when I think that Black can gain the upper hand with 1S . . . ..lid7 16 i.g2 a4 17 'iib4 a3. Likewise, 12 .:!.d1 aS! 13 ltJgS (or 13 a3 a4 14 �a2 ..lic2 1S .l:!.d2 ..lib3 16 �1 ltJdS with some advantage) 13 . . . l2Jb4 with the initiative. 12 .. . a s !

Harassing the white queen. Now White must avoid 13 ltJeS a4 14 l2Jf7+? l:.xf7 1S �xf7 �d6 16 g4 l!£8 17 �x£8+ ..lix£8 18 gxfS a3, but even 14 �S a3 1S b3 'i¥d6 leaves him in trouble.

8 5

Play t h e G r u nfe ld

82) 8 f4 Aggressive, but too loosening.

8 ... lbc6! 9 ds White should avoid 9 lbf3? ..ltg4 10

d5 (or 10 e5 lbb4 with full control over d5 and an edge) 10 . . . lba5 11 ..ltd4 ..ltxf3 12 gxf3 e6 when his centre comes un­der pressure. 9 ... lbb8 10 a4

10 lbf3 c6 11 'i!Vb3 cxd5 12 lbxd5 lbxd5 13 exd5 lbd7 14 ..lte2 'i!Va5+ 15 ..ltd2 'i!Vb6 16 ..ltc3 ..ltxc3+ 17 bxc3 'i!Ve3! was better

aS lb6d7 12 lbf3 'i!Ve7! (Black now threatens . . . exf4) 13 i.d3 exf4 14 .ltxf4 lbe5 15 lbxe5 .ltxe5 16 i.xe5 'i!Vxe5 17 0-0 lbd7 18 i::tcl 'i!Vd4+ 19 Wh1 lbe5 left White on the defensive in H.Pilnik­S.Reshevsky, New York 1942. 11 .. . gxfs 12 a s

S o far w e have followed a sugges­tion of Davies' and now I feel that Black has good play after 12 . . . lb6d7 13 exf5 c6.

for Black in M.Euwe-S.Reshevsky, Hol- 83) 8 .ltbS!? land 1938, while Black enjoyed very good compensation after 10 i.d4 ..lth6 11 'i!Vf3 c6 12 h4 cxd5 13 exd5 e5 14 ..ltxe5 ctJ8d7 15 ..ltd4 lb£6 16 0-0-0 'i!Vd6 in V.Korchnoi-L.McShane, Igualada 2005. 1o ... es !

Correctly fighting for the central dark squares . 11 fs !?

Alternatively, 11 fxe5 ..ltxe5 12 lb£3 was F .Gheorghiu-L.Ftacnik, Palma de Mallorca 1989, and now strong is 12 . . . ..\txc3+! 13 bxc3 l:te8 14 'i!Vd4 f5 ! 15 ..ltd3 (or 15 ..ltg5 'i!Vxd5! - Ftacnik) 15 . . . fxe4 16 .lk.xe4 c5 (Ftacnik), while 1 1

8 6

A recent trend which discourages . . . lbc6 . I must say, though, that in my view this is taking the concept of pro­phylaxis a little too far, even if several strong grandmasters have tried it. 8 ... 'i!Vd6

A good response, preparing . . . lbc6 as well as . . . l:td8 . 9 lbge2

9 lbf3 a6 10 ..lte2 lbc6 threatens . . . .ltg4 and is unpleasant for White. 9 . . . a6 10 ik.d3 l:td8 11 0-0 lbc6 12 dS lbes 13 ..ltc2 lbbc4!?

The logical 13 . . . e6 14 dxe6 ..ltxe6 was attractive for Black in A.Aleksandrov-

A.Grischuk, Internet (blitz) 2004. After the text, 14 i.cl i.d7 gives Black a promising initiative.

84) 8 i.e2

White contents himself with simple development. 8 ... tt:Jc6

Now a decision must be made re­garding the d4-pawn.

841: 9 d s 842: 9 lt:Jf3

841) 9 ds tt:Jes 10 i.d4

4 cxds tlJxds with o u t 5 e4

White is threatening f4, so Black must react promptly. 1o .. . cs !

The only move according to Svidler, but also a very strong one; Black buys some time to create threats against b2. 11 i.xcs tt:Jec4 12 i.xc4

Black is much better after 12 i.xb6?! "i¥xb6 13 i.xc4 "i¥xb2, and 12 "i¥b3 lt:Jxb2 13 l:tcl (13 "i¥xb2? lt:Ja4 wins the house) 13 . . . i.xc3+ 14 "i¥xc3 lt:J2a4 is also good for Black, D.Palo-A.Jerez Perez, An­dorra la Vella 2004. 12 ... tt:Jxc4 13 "i¥b3

Note that 13 "i¥e2 lt:Jxb2 will trans­pose after 14 "i¥xb2. 13 ... tt:Jxb2 ! 14 "i¥xb2 "i¥c7 15 i.b4

Instead 15 "i¥b4 a5 1 6 "i¥a3 b6 17 lt:Jge2 bxc5 18 0-0 i.a6 left Black much better in I .Shliperman-B.Avrukh, Cala Galdana 1996, while 15 "i¥a3 b6 16 i.b4 a5 17 l:tcl axb4! 18 "i¥xb4 (18 "i¥xa8? i.xc3+) 18 . . . i.a6! 19 lt:Jge2 "i¥d7 supplies excellent compensation, due to the two bishops and the hanging state of the white knights; Black will further open up the position with . . . e6. 1S ... as 16 tt:Jge2 axb4 17 "i¥xb4 i.g4!

8 7

Play th e G r iJ nfe ld

This occurred in G.Kacheishvili­P.Svidler, Szeged 1994. Black is better here, in view of White's obvious coor­dination difficulties.

842) g lt:lf3 .lil.g4 10 ds .lil.xf3

White now faces another important decision: 11 gxf3

Instead 1 1 .ixf3 avoids structural damage at some cost in terms of time. Then 1 1 . . .lt:le5 12 .ie2 lt:lec4 13 .tel c6 14 dxc6 bxc6 gives Black good counter­play; for example, 15 'ti'c2 (15 0-0 'ti'xd1 16 lhd1 .l:tfd8 leaves Black with a healthy lead in development) 15 . . . 'iVd4! (a strong centralizing move) 16 0-0 .l:tfd8 17 .l:tbl lt:la3! 18 bxa3 'iVxc3 19 'iVxc3 (Black is also better after 19 'iVb3 c5 20 ..ie3 'ti'a5! ?, intending 21 'i/Ub5 'iVxb5 22 .l:txb5 .id4) 19 . . . .ixc3 20 .ia6 c5 21 .if4 .id2! and White faced seri­ous problems in I.Sokolov-E.Sutovsky, Hastings 2000 . 11 ... LtJes

White is about to set his central pawn mass rolling, but he must first cover the c4-square and that gives

8 8

Black just enough time to develop counterplay. 12 'iVb3 !

The critical move, taking c4 away from the black knights and preparing f4. Alternatively:

a) 12 i.d4 e6! 13 f4 lt:led7 14 i.xg7 Wxg7

15 'iVd4+ (15 dxe6 fxe6 is very good for Black as the f4-pawn is extremely weak) 15 . . . 'iVf6 16 0-0-0 (Rowson' s 16 e5 'i/Uh4 again highlights the weakness of f4, while 16 'iVxf6+ lt:lxf6 17 .if3 exd5 18 e5 lt:lfd7 19 lt:lxd5 lt:lxd5 20 .ixd5 c6 21 .ib3 lt:lc5 leaves Black better) 16 . . . exd5 17 lt:lxd5 (or 17 exd5 lt:lc8 18 lt:le4 'i!Nxd4 19 .l:txd4 lt:le7 20 Wb1 lt:lb6 which was pretty favourable for Black in G.Rojo Huerta-M.Turov, Linares 2000) 17 . . . lt:lxd5 18 exd5 .l:tad8 (so far M.Tyrtania-A.Lagunow, Berlin 1994) 19 .if3 lt:lb6! and Black can organize strong pressure against d5 after the exchange of queens, followed by . . . .l:td6 and . . . .l:tfd8.

b) 12 0-0 lt:lec4 13 �g5 h6 14 .ih4 was tried in R.Sheldon-C.Beaumont, Newport 1999, and now the simple

1-L .fS is promising. 12 . . . c6 13 f4!

Instead 13 0-0-0 cxdS 14 iLlxdS (14 �xb6 'ii'xb6 1S "ib'xb6 axb6 16 iLlxdS e6 1 7 iLlxb6 .l:txa2 is quite promising, too, :'or Black) occurred in V.Korchnoi­G .Kasparov, Wijk aan Zee 2000, and :1ow Black could obtain the edge, ac­cording to Lagunov, after 14 . . . "ib'c8+ 1S �b1 iLlxdS 16 .l:txdS e6 17 .l:tcS "ib'd7 18 :d1 "ib'e7! 19 .l:f.dcl iLlc6. 13 . . . iLled7 14 dxc6

Black gains good pressure down the .:-file after 14 .l:td1 .lk.xc3+ 1S bxc3 cxdS 16 exdS 'ii'c7. 14 . . . bxc6

15 0-0-0 Various other moves have been

tried here, but they don't promise any­thing; for example, 1S .l:td1 eS 16 fxeS "ii'h4 17 e6 .lk.xc3+ 18 "ib'xc3 "ib'xe4 19 ex£7+ .l:txf7, with a good game for Black, or 1S 0-0 .lk.h6 !?, followed by . . . 'ii'c7 with pressure against f4. 1S . . . e6!?

Black is now better in every line; for example, 16 'ii'a3 �4 (K.Miton­R.Wojtaszek, Goa 2002), or 16 .l:td6? !

4 cxd5 liJxd5 w i th o u t 5 e4

"ib'c7!, or 16 "ib'c2 "ib'e7 17 h4 .l:f.fb8 18 hS aS 19 hxg6 hxg6 20 iVd2 a4 21 ik.d4 eS 22 fxeS cS ! (R.Sheldon-J .Rowson, Southend 1999) .

C) 5 iLlf3 .lk.g7 White can now opt for one of the

main lines with 6 e4 iLlxc3 7 bxc3 cS 8 .l:tb1 (see Chapter Two) or prefer to de­velop his dark-squared bishop.

(1: 6 ik.d2

c2: 6 .ig5

C1) 6 .lk.d2

This is a fairly harmless version of Line B, above. 6 . . . 0-0

7 .l:tcl The most common, but White might

also try: a) 7 "iVb3 iLlb6 8 .l:td1 (or 8 e3 iLl8d7 9

iLle4 aS 10 .l:f.cl eS 1 1 .ic3 "ib'e7 12 dxeS iLlxeS 13 iLlxeS .ixeS 14 .ixeS 'ii'xe5 1S iLlc3 .lk.e6 16 "ib'c2 .l:f.fd8 and Black was better in K.Sabri-T.L.Petrosian, Tehran 200S) 8 . . . .ie6 9 'ii'c2 iLlc6 10 i.f4 oc-

8 9

P lay th e G r u nfe ld

curred in G.Lorscheid-F .Holzke, Cap­pelle la Grande 2002, and now 10 . . . . :t:JdS gives Black a nice initiative.

b) 7 e4 i2lb6 8 �e3 �g4 transposes to a sideline examined in Line B and not a very good one for White at that.

c) 7 't!Vcl i2lb4! 8 iLlbS iD8a6 9 a3 i2lc6 10 �h6 �g4 11 �xg7 �xg7 12 't!Vd2 't!VdS was again good for Black in G.Kuzmin-L.Stein, Sochi 1970.

d) 7 e3 cS 8 dxcS i2lb4 gives Black obvious compensation; . . . i2ld3+ is the intention and it is unlikely that White can even aspire to retain the extra pawn on cS. 7 ... i2lb6

8 �g5

The only way to justify White's �d2 idea. Instead 8 e3 is pretty lame and after 8 . . . i2lc6 9 .ibS eS 10 �xc6 exd4 1 1 i2lxd4 bxc6 Black has a very active posi­tion, while 8 .if4 is best met by 8 . . . cS ! 9 dxcS �xc3+ 10 bxc3 (or 10 l:Ixc3 't!Vxd1+ 1 1 �xd1 tt:Ja4 12 l:Ia3, J.Partanen­T.Halmeenmaki, Finland 2001 , and now 12 . . . l:Id8+ 13 �cl tt:JxcS 14 e3 tt:Je4 with the initiative) 10 . . . 't!Vxd1+ 1 1 1Ixd1 tt:Ja4 12 i..h6 l:Ie8 13 c4 i2lc6 when Black

9 0

was doing very well in V.Smyslov­H.Ree, Wijk aan Zee 1972. 8 . . . h6 9 �f4

9 �h4 gS 10 �g3 tt:Jc6 1 1 e3 trans­poses. g . . . tt:Jc6 10 e3 gS 11 �g3 g4!

In this forceful way Black ensures that . . . eS will come with some effect. 12 i2ld2

12 tt:Jh4? ! is even worse in view of 12 . . . eS! 13 dxeS (13 dS? i2lb4 14 e4 �f6) 13 . . . tt:JxeS 14 �e2 �f6! with a clear ad-vantage. 12 . . . es! 13 dxes iLlxes 14 Wicl tt:Jds

Black has assumed the initiative and now 1S �xeS �xeS 16 �c4? tt:Jxc3 17 bxc3 't!Vd6! left White in big trouble in S.Martinovic-J .Timman, Amsterdam 198S, while even the superior 16 't!Ve4 �xc3 17 bxc3 l:Ie8 18 't!Vd4 �fS gives Black a powerful initiative.

c2) 6 �gs

In this very rare line White will fol­low up with l:Icl , aiming to force Black to trade on c3. 6 . . . h6

The inclusion of this move is useful

because it restricts the influence of White's dark-squared bishop. 7 �h4

Naturally 7 �d2 is not very consis­tent; for example, 7 . . .. :LJb6 8 e4 �g4 9 �e3 0-0 10 a4 aS 1 1 �e2 i2lc6 12 dS i2lb4 13 �d2 (J .Stocek-J .Borisek, Nova Gorica 2004) 13 . . . �xf3 14 gxf3 e6 and Black is better. 7 . . . cs 8 .l::i.c1

8 e3 is again inconsistent. Black ob­tains a good position against an IQP with 8 . . . cxd4 9 exd4 0-0 10 �c4 i2lb6 1 1 �b3 i2lc6 ( 1 l . . .�g4! ? 12 d S i2la6, intend­ing . . . iDeS is another good plan) 12 dS �aS 13 0-0 (T.Taylor-G.Antal, Buda­pest 2003) 13 . . . i2lxb3 14 �xb3 �g4, generating annoying threats. 8 . . . i2lxc3

Black must be careful here not to overlook that the obvious 8 . . . cxd4?? loses to 9 iLlxdS . 9 bxc3 �e6

An important move. The bishop is going to dS, from where it eyes both

4 cxds Cbxds with o u t 5 e4

flanks and controls some important squares. Moreover, the a2-pawn is made to feel a little sensitive. 10 �a4+

Similar is 10 �d2 i2lc6 1 1 e3 .idS when Black was fine in I.Sokolov­L.McShane, Selfoss 2003, while 10 e4? ! asks too much from the position and 10 . . . 0-0 (lO . . . �aS is also strong) 1 1 �e2 �xa2 12 0-0 �e6 hardly gives White any compensation for his pawn. 10 . . • i2lc6 11 e3 o-o 12 �e2 �ds

Black has equalized and can play for more, perhaps with a kingside ad­vance; for example, 13 0-0 .l::i.c8 14 .l::i.fd1 cxd4 1S cxd4 g5 16 �g3 �aS 17 �bS �xa4 18 �xa4 fS saw White come un­der pressure in A.Lahiri-T.Vakhidov, Chennai 2004.

Conclusion

The Nadanian variation enjoyed some popularity a few years ago, but it has clearly been defanged. However, S �d2 is an important system and the weapon of several grandmasters. Against it, Black must not waste time in purposefully attacking White's cen­tre and I feel that Black's chances are by no means inferior, especially since in many variations White's lag in de­velopment is a telling factor. Finally, the two rare lines examined after S i2lf3 �g7 do not present any danger; com­mon sense should suffice to obtain a good position against them.

9 1

C h a pte r S eve n I The Russ ia n System

1 d4 tt::lf6 2 c4 g6 3 tt::lc3 ds

Here we will consider:

A: 4 ctJf3 .¥i.g7 5 �b3

B: 4 �b3

Line A is the Russian System, while Line B is a rare transpositional attempt. The Russian System is a critical branch of the Gri.infeld since White intensifies the pressure against d5, practically forcing Black temporarily to abandon the centre .

A) 4 tt::lf3 �g7 5 �b3 dxc4 6 �xc4 0-0 7

e4

If left unhindered, White will now develop freely, stabilize his central su­periority and deprive Black of counter­play. The alternatives are much less critical:

a) 7 g3? ! is inconsistent with White's development scheme: 7 . . . �e6 8 �a4 (8 �d3 tt::lc6 9 �g2 tt::ld5 10 �d1 tt::lb6, fol-

9 2

lowed by . . . �d7, is fine for Black) 8 . . . tt::lc6 9 �g2 tt::ld5 10 0-0 tt::lb6 1 1 �d1 (so far S.Reshevsky-W.Lombardy, New York 1960) 1 1 . . .tt::lxd4 12 tt::lxd4 �xd4 13 �xd4 �xd4 14 �xb7 M.ab8 15 �a6 �c4 leaves Black slightly better.

b) 7 il..g5 �e6 8 �5 tt::lc6 gives Black the initiative .

c) 7 �f4 is tried on occasion. A sim­ple and sound way to meet it is 7 . . . c6! 8 e4 (harmless is 8 �a4 �f5 9 e3 tt::lbd7 10 �e2 h6 1 1 0-0 tt::lh5 with equality, while 8 �3 �a5 9 e3 tt::lbd7 1 0 �e2 tt::ld5 1 1 .i.g3 tt::lc5! intending . . . tt::le4 gives Black the initiative) 8 . . . b5 9 �3 (9 �d3 �a5 ! 1 0 �e2 b4 1 1 tt::ld1 c5 ! sees White's cen­tre collapse) 9 . . . �e6! (planning . . . �a5 and . . . b4) 1 0 �c2 (in E.Hadet­C.Marcelin, Issy Les Moulineaux 2000, White preferred 1 0 d5 cxd5 1 1 tt::lg5, but now 1 1 . . . tt::lc6! is promising for Black) 1 0 . . . �a5 and again Black has the initiative .

After 7 e4, Black usually opts for 7 . . . tt::la6 and then . . . c5, or the more ambi-

Th e R us s i a n Sys tem

tious 7 . . . a6, intending . . . b5 and . . . c5. complex play. Overall, I feel that Black is walking a tightrope in these systems (especially so after 7 . . . a6), while the deep theoretical knowledge necessary to employ them is somewhat off-putting. Instead I have opted for a less violent set-up:

7 ... tt:lc6!? Black opts for piece play against

White's centre . Such a strategy is often possible when White is unable to sup­port d4 with a pawn and I see no reason why it shouldn't work here.

Not that this system is a sideline; on the contrary, it bears the stamp of the great ex-world champion Vassily Smyslov and was Black's initial reac­tion to the Russian System in the Fifties and Sixties. 7 . . . tt:lc6 has never been refuted and is still employed at the top level, although it is not currently espe­cially fashionable . However, the direc­tion of fashion towards the two afore­mentioned alternatives is mostly due to the influences of Garry Kasparov (for 7 . . . tt:la6) and the Hungarian school (7 . . . a6), who have both always pre-ferred to strive for more direct and

First of all with 7 . . . tt:lc6, Black eyes d4, greatly restricting White's develop­ment options. He intends to gain time by harassing the white queen with the important manoeuvre . . . tt:ld7-b6, while the light-squared bishop is destined for g4. At the right moment, a central break will occur, with dark square domination often a resultant feature. Sometimes after . . . tt:ld7-b6, Black will even strike with . . . f5; a concept we've already seen in the 5 i.d2 variation.

Our preferred system can also arise after 7 . . . i.g4, but like Rowson, I feel that the suggested move order is pref­erable, as it strongly discourages 8 i.e3, as we shall see . White has now tried several moves:

Al: 8 .ie3 A2: 8 ds A3: 8 .igs A4: 8 es AS: 8 .if4 A6: 8 h3 A7 : 8 .ie2

9 3

P lay th e G r ii nfe /d

A1) 8 �e3? tt:Jg4!

A key idea behind not hurrying with . . . �g4. 9 0-0-0

Alternatively: a) 9 e5 �e6 10 't!Vc5 (D.Lima­

E.Tsuboi, Brasilia 2000) 10 . . . a5 threat­ens . . . tt:Jb4 when White is already on the defensive.

b) 9 �e2 tt:Jxe3 10 fxe3 e5 1 1 d5 (or 1 1 0-0 exd4 12 exd4 tt:Jxd4 13 .l:!.ad1 tt:Jxf3+ 14 .l:!.xf3 WigS with the initiative) 1 l . . .ctJe7 12 0-0-0 a6 13 �b1 Wid6 14 't!Vb3 b5 15 .l:!.cl �d7 and in J.Murray­F.Olafsson, Reykjavik 1975, Black had the upper hand.

c) 9 .l:!.d1 ctJxe3 10 fxe3 e5! sees Black immediately strike on the dark squares: 11 d5 (or 11 dxe5 't!Ve8 12 ctJd5 tt:Jxe5 13 't!Vxc7 �h8 with excellent compensation as . . . f5 is coming - just look at the poor white king! ) 1 l . . .ctJe7 12 �e2 (V.Raceanu-N.Gerard, Bucharest 2002) 12 . . . a6 and Black, who intends . . . 't!Vd6 and . . . f5, has the initiative. 9 ... tt:Jxe3 10 fxe3 es 11 �b1

Instead 11 dxe5 't!Ve8 12 ctJd5 tt:Jxe5 leaves Black with an unquestionable

9 4

positional advantage.

11 ... �g4 12 dS tt:Je7 13 h3 �d7 14 h4 tt:Jc8 15 hS tt:Jd6 16 't!Vb3 bs 17 hxg6 fxg6

Black enjoys the initiative, O.Jakob­sen-D.Howell, Copenhagen 2006.

A2) 8 ds? ! This commits White's central pawns

too soon. 8 ... tt:Ja s

9 't!Vd3 White has tried a number of other

options, but in each case Black strikes quickly at d5 with . . . c6 and gains a good game.

a) 9 �c5 c6 10 dxc6 (and not 10 b4? cxd5 ! 11 e5 ct:Je4 12 ct:Jxe4 dxe4 with a clear advantage) 10 . . . ct:Jxc6 1 1 �e2 (J .Richardson-S.Ernst, Lichfield 2000) 1 l . . .ctJd7 12 �a3 ctJde5 13 ctJxe5 ctJd4 with an edge.

b) 9 �4 c6 10 dxc6 ctJxc6 11 �a4 ct:Jg4 with the initiative.

c) 9 �a4 c6 10 dxc6 (instead 10 �d2 cxd5 1 1 ct:Jxd5 ct:Jxd5 12 .ixa5 b6 13 .l::i.d1 bxa5 14 .l::i.xd5 �c7 is obviously good for Black, while 10 b4? ct:Jxd5 ! 11 ct:Jxd5 cxd5 ! ? 12 e5 �c7 just wins) 10 . . . ct:Jxc6 1 1 �e2 C2Jd7 1 2 .ie3 C2Jb6 1 3 �5 .ig4 14 .l::i.d1 �c7 15 0-0 e6 16 a4 was seen in T.Nemec-D.Koval, Bratislava 1991, and now simply 16 . . . .l::i.fd8 ensures equality.

d) 9 �5 c6 10 dxc6 ct:Jxc6 11 .ie2 a6 12 �3 (M.Pukklia-P.Jantti, Helsinki 2000) 12 . . . .ie6 13 �c2 ctJb4 14 �1 �c7 and Black is better. 9 . . . c6

10 dxc6 Of course, 10 b4? ct:Jxe4! is a blunder

and White should also avoid 10 .ie3 cxd5 1 1 exd5 �f5 12 �d2 ctJe4 13 ct:Jxe4 �xe4 14 .l::i.d1 .l::i.c8 15 �d4 �xf3 16 �xg7 .ixd1 17 �xf8 (J.Thomassen-K.White,

Th e R u s s i a n Sys t e m

Gausdal 2005) when Black can win with 17 . . . �xf8 18 �xd1 �6. Finally, 10 �g5 cxd5 1 1 exd5 .if5 12 �d2 .l::i.c8 13 .ixf6 .ixf6 (T .Kuzmak-P .Staniszewski, Bielsko Biala 1990) is much better for Black, who has two strong bishops. 10 . . . ct:Jxc6 11 �xd8 .l::i.xd8

Black has a slight development lead. 12 �e2 b6 13 o-o .ib7

Here 14 e5 ctJd5 15 ct:Jxd5 .l::i.xd5 gives Black the initiative. Thus White pre­ferred 14 �f4 in W.Uhlmann-V.Korch­noi, Stockholm Interzonal 1962, but after 14 . . . ct:Jb4 15 e5 ctJe4 16 ctJxe4 �xe4 17 .l::i.fd1 ctJd3 18 .ixd3 .ixd3 Black had a slight edge due to his bishop pair.

A3) s .tgs h6 Renewing the problem of where to

station the bishop. 9 �h4

This leaves the bishop away from the centre, which means that White will have to push d5, but the alterna­tives also fail to impress :

a) 9 �xf6 �xf6 10 0-0-0 .ig7 gives Black good long-term chances with his extra dark-squared bishop; he should continue with . . . �d6 and . . . a5 .

9 5

P lay th e G r ii nje l d

b ) 9 i.e3 ct:Jg4 10 e5 was seen in A4) 8 eS !? ctJd7 V.Epishin-A.Ledger, Gibraltar 2003, and now the typical 10 . . . i.e6 1 1 'iVa4 i.d7 12 'iVb3 a5, intending . . . a4 or . . . ct:Jb4, gives Black good play. g ... ii.g4 1o d s

10 .. . i.xf3 The direct 10 .. .lba5 can also be

played and is similar: 1 1 'iVb4 i.xf3 12 gxf3 b6 (Black can also equalize with the forcing 12 . . . c6! ? 13 0-0-0 cxd5 14 i.xf6 i..xf6 15 ct:Jxd5 .Mc8+ 16 �b1 ct:Jc6 1 7 ct:Jxf6+ exf6 18 .Mxd8 ctJxb4 as White's kingside pawn structure is weakened) 13 i.b5 a6 14 ..ia4 'iVd6! and Black had the initiative in A.Guseinov-D.Zagor­skis, Pardubice 1995. 11 gxf3

Instead 1 1 dxc6 b5 12 ctJxb5 i.xe4 13 .Md1 'iVb8 14 i.xf6 exf6 is promising for Black, as is 1 1 i.xf6 i.xf6 12 dxc6 b5 13 ctJxb5 i.g5. u . . . ct:Jes 12 'iVe2 ct:Jhs 13 i.g3 c6

With a typical position for this line and one in which Black stands well; White's lack of development is a cause for concern.

9 6

9 i..e3 Black was comfortably placed after

9 i.e2 t2lb6 10 'iVc5 i.e6 1 1 i.e3 h6 12 .Md1 t2Jd5 in J.Colchado-M.Aguilar, Lima 2002. g . . . ct:Jb6 10 'iVcs

Best, whereas 10 'iVb3 .i.e6 (also promising is 10 . . . a5 11 a3 ii.e6 12 'iVc2 a4, K.Stead-V.Feldman, Gold Coast 2001 ) 1 1 'iVd1 ( 1 1 'iVc2 ct:Jb4 12 'iVd1 ct:Jc4 13 .tel c5 gave Black a clear edge in Z.Birovljevic-M.Podobnik, Pula 2000) 1 1 . . . 'iVd7 gives Black an edge, with . . . .MadS and .. .£6 being the plan. 1o ... as ! 11 .i.e2

White should avoid 1 1 a3? a4! when suddenly a nasty . . . l:ra5 is threatened: 12 i.b5 l:ra5 13 l:rd1 (M.Miranda­E.Tsuboi, Curitiba 1999) 13 . . . i.d7 pre­pares . . . ct:Ja7 and leaves White in real trouble. 11 ... ct:Jb4 12 o-o c6

12 . . . ct:Jd7!? is also interesting: 13 'ifu5 (13 'iVc4 C2Jb6 14 'iVc5 repeats the posi­tion) 13 . . . c6 14 'iVa4 ct:Jb6 15 'iVd1 ctJ4d5 was equal in B .Gulko-V.Gavrikov, Tal­linn 1983.

After the text, 13 ct:Je4 i.f5 14 ct:Jfd2? is a serious mistake: 14 . . . ctJ6d5 15 C2Jg3 i.e6 16 a4 f5 (16 . . . tDxe3 ! ? 17 fxe3 i.h6 was also very good for Black in V.Milov-Ma.Tseitlin, Tel Aviv 1994) 17 exf6 exf6 and Black i s much better; . . . f5 is next. Better is 14 lDg3, although after 14 . . . i.g4 Black has equalized.

AS) 8 i.f4 A critical move . The c6-knight is

now pinned and White threatens d4-d5, so Black must react purposefully . s . . . ct:Jhs ! 9 i.e3 i.g4

In this way Black mangles White's kingside structure and can then hope

Th e R u s s i a n Sys t e m

to gain a positional advantage, particu­larly on the dark squares. In return White will play for either a kingside attack or a direct central breakthrough, leading to some rather complex and dynamic positions. 10 0-0-0

The most aggressive choice, but also the most risky. Alternatively:

a) 10 d5 CtJe5 11 tDxe5 i.xe5 12 g3 i.£3 13 l:rg1 c6 (S.Pedersen-L.Schan­dorff, Aalborg 2000) and Black cannot possibly be worse.

b) 10 l:rd1 ? ! i.xf3 1 1 gx£3 e5

12 dxe5 (or 12 d5 CtJd4 13 .ixd4 exd4 14 CtJe2 i.e5 15 C2Jxd4 'iVf6 with an obvi­ous advantage; here 13 l:rxd4 exd4 14 i.xd4 i.xd4 15 'iVxd4 was an ambitious positional exchange sacrifice in Y.Razuvaev-Y.Kotkov, Moscow 1969, but I don't see the point as 15 . . . c6 opens lines for the black rooks and appears rather promising) 12 . . . 'iVh4 13 .ie2 (13 e6? ctJe5 highlights the weakness of f3) 13 . . . tDxe5 and Black enjoys the initiative.

c) 10 e5 is rather anti-positional, but White hopes to exploit the position of the h5-knight. However, after 10 . . . CtJa5

9 7

Play t h e G r il nfe ld

1 1 iVa4 c5 12 .l:Id1 tt:lc6, Black was able to dismantle White's centre and to solve all his problems in Z.Kozul­L.Gutman, Graz 1987. 10 .. . iLxf3 11 gxf3 es !

Aiming for the aforementioned dark square control. 12 ds

Instead 12 dxe5 'ifu4 13 e6 fxe6! (13 . . . tt:le5 14 exf7+ .l:Ixf7 15 'iVb3 c6 is given as promising by Rowson, but after 16 iLc4 tt:lxc4 17 iVxc4 White just seems to be better) 14 iVxe6+ �h8 gives Black good compensation. 12 ... ctJd4 13 f4 ctJf3 14 iLe2

White can also win the exchange, but 14 iLc5 tt:lxf4! (14 . . . exf4? 15 'iVb3! threatens the unpleasant iLe2) 15 iLxf8 .i.xf8 gives Black excellent play on the dark squares. He also gains good pros­pects after 14 f5 ctJd4 15 fxg6 hxg6 16 h4 c6 (16 . . . c5 can also be tried; 17 iVxc5 .l:Ic8 18 iVa3 .l:Ie8 19 �b1 a6 20 .l:Icl b5 21 tt:le2 �£6 gave Black promising com­pensation in R.Markus-C.Lupulescu, Subotica 2003) 17 �b1 iVd6, intending a pawn-storm on the queenside with . . . b5 and . . . aS .

9 8

14 ... exf4 1 5 .ixf3 fxe3 1 6 .i.xhs gxh5 17 fxe3

Or 17 £4 'ifu4 with some advantage. 17 ... iVd6

The great scope of the g7 -bishop gives Black excellent compensation for his split kingside.

A6) 8 h3

Preventing . . . .i.g4 and facilitating iLe3, but this gives Black a significant tempo. 8 ... ctJd7

8 . . . e5 ! ? is another way to exploit White' s loss of time: 9 dxe5 (9 d5? tt:ld4, intending 10 tt:lxd4? exd4 1 1 iVxd4 tt:lxe4! , is good for Black) 9 . . . tt:ld7 (but not 9 . . . .ie6? 10 ex£6 .ixc4 1 1 fxg7 �xg7 12 i..xc4 and White's pieces outclassed Black's extra queen in J.Van den Berse­laar-J .Hulin, Bethune 2003) 10 e6 (or 10 iLg5 tt:ldxe5! 1 1 'iVd5 'iVe8 with unpleas-ant threats) 10 . . . fxe6 11 iVxe6+ �h8 12 �d5 tt:lb4 (12 . . . .l:Ixf3?! also deserves at-tention, but is ultimately inadequate : 13 gxf3 ctJd4 14 i..e3! tt:lc2+ 15 �d2 c6 16 �d6! tt:lxa1 17 �cl �a5 1 8 �bl iLxc3 19 bxc3 tt:lc2 20 i..cl with a clear advan-

tage for White) 13 �3 lbd3+ 14 i..xd3 lbc5 and Black has good compensation for the pawn. 9 i..e3

9 d5 is inconsistent: 9 . . . lbce5 10 lbxe5 lbxe5 11 �3 e6 12 i..e3 (12 i..e2 exd5 13 exd5 c6 14 i..e3 cxd5 15 'tixd5 i..e6 16 'tixd8 .l:i.fxd8, threatening . . . lbc4, was prom1smg for Black in G.Levtchouk-S.Barbeau, Quebec 1987) 12 . . . exd5 13 exd5 (13 0-0-0 c6 14 exd5 cxd5 15 .l:i.xd4 transposes, while 15 'tixd5 'iff6 gives Black the initiative) 13 . . . c6 14 0-0-0 cxd5 15 .l:i.xd5 was T.Jugelt-S.Joachim, German League 1995, and now after 15 . . . 'tic7 16 .l:i.c5 �8 and . . . i.e6, Black has the upper hand. g . . . lt:Jb6 10 'tics

This is almost always the best re­sponse to . . . lbb6, as other moves fur­ther misplace the queen and lose con­trol of the position. Indeed, 10 'tid3? ! £5 ! threatens . . . lbb4 and is awkward for White:

a) 11 i..e2? f4 12 i..d2 lbxd4 was a disaster for White in K.Hjornevik­T.Bae, Oslo 2003.

b) l l .l:i.d1? lbb4 (or 1l . . . fxe4 12 Cbxe4 i.e6, threatening both . . . i..c4 and the a2-pawn) 12 'tid2 fxe4 13 lbxe4 i..f5 14 ·'ii'xb4 i.xe4 15 i..e2 'tid6 with the ad­\'antage, W.Uhlmann-E.Jimenez Zer­quera, Tel Aviv Olympiad 1964.

c) 11 e5 f4 12 i.cl lbb4 13 'tid1 1 R.Canaza-H.Retamozo, Lima 2004) I 3 . . . i..e6 gives Black an edge.

d) 11 'tid2 fxe4 12 lbxe4 i.f5 13 lbc5 i'd5 (Black is beautifully centralized and has a promising position; . . . e5 is on

Th e R u s s i a n Sys tem

the cards) 14 i.e2 (or 14 .l:i.cl e5 15 lLJa4 Wh8 which gave Black a clear advan­tage in I.Bender-M.Konopka, Plzen 2001) 14 . . . .l:i.ab8 15 0-0 i.xh3 16 gxh3 .l:i.xf3 17 i.xf3 'tixf3 with a powerful initiative and great positional compen­sation for the exchange. 1o ... fs !

Again, this is the most incisive way to play. 11 .l:i.d1

Instead 11 .ic4+ Wh8 threatens . . . f4, while 1 1 e5 is met by 1 l . . .f4! 12 i..d2 a5 (threatening . . . Cbb4) 13 i..d3 (13 i..b5 lbb4 14 0-0 c6 is better for Black) 13 . . . i..e6 and Black stands well, but not 13 . . . Cbd7? due to 14 "ifc4+ Wh8 15 Cbd5. u . . .fxe4 12 lt:Jes !

Both 1 2 Cbxe4 i..e6 and 12 lbg5 e 6 in­tending . . . .l:i.f5 are better for Black (Suetin) . 12 .. . 'tid6 13 Cbxc6 bxc6 14 lt:Jxe4 'tids !

Black has equalized, S.Lputian­Y.Balashov, Moscow 198 1 .

A7) 8 i.e2 White elects simply to develop and

refrains from immediate action.

9 9

Play t h e G r ii nfe ld

8 ... lLld7

9 ds !? An interesting attempt to achieve a

slight edge. Instead 9 i.e3 can now be safely played, but Black is in time to obtain counterplay: 9 . . . lLlb6 10 'iWc5 (or 10 'i'd3 f5 11 �d1 f4 12 i.cl ..tg4 with at least equality for Black, while here 1 1 e5? f4 12 i.cl i.g4 creates great problems with the d4-pawn) 10 . . . f5 1 1 �d1 (alternatively 1 1 e5 f4 1 2 ..td2 i.g4 13 lLlg5 was seen in S.Bromberger­M.Keller, Bad Wiessee 2003, and now 13 . . . 'iWd7 14 h3 i.xe2 15 lLlxe2 �adS supplies strong pressure; 11 d5 lLle5 12 lLlxe5 i.xe5 13 f4 i.xc3+ 14 'iWxc3 fxe4 15 0-0-0 'iWd6 is less clear-cut but still better for Black, although he must be careful about the dark squares around his king) 1 l . . . fxe4 12 lLle5 (12 d5? exf3 13 dxc6 fxg2 14 �g1 'iWe8 is clearly bad for White, but possible is 12 lLlxe4 i.e6 with equality) 12 . . . 'iWd6 13 lLlxc6 bxc6 14 lLlxe4 'iWd5 and Black is fine. 9 . . . lLJces 10 lLJxes

10 'iWb3 lLlc5 intends .. . f5, forcing White on to the defensive. 10 .. . lLJxes

1 0 0

The correct recapture, whereas Black's position was a mess after 10 . . . i.xe5 1 1 i.g5 lLlb6 12 'iWc5 f6 13 i.h6 �f7 14 f4 in J .Lautier-J .De la Villa Garcia, Pamplona 2000. 11 'iWb3 e6 12 o-o exds 13 exds �e8

Also worthy of attention is 13 . . . 1Vh4! ?, preparing . . . lLlg4. 14 i.e3 a6

Having prevented an annoying lLlb5, Black intends . . . 'i'h4 with threat­ening activity.

B) 4 'iWb3 dxc4 5 'iWxc4 i.g7

Play will now normally transpose to the Russian System, but White also has

a few harmless alternatives. 6 e4

Instead 6 'i!kd3 lt:Jc6 7 lt:Jf3 0-0 is equal, whereas 6 i.£4 c6 can easily turn out well for Black:

a) 7 e3 0-0 8 lt:Jf3 'i!t'a5 is fine for the second player.

b) 7 .l:!.d1? ! 'i!kaS 8 .i.d2 was played in the famous game M.Euwe-A.Alekhine, World Championship (Game 4), The Hague 1935, and now 8 . . . �6 9 .tel i.f5, as suggested by Suetin, is good for Black.

c) 7 lt:Jf3 0-0 8 e4 b5 9 'ifh3 (or 9 'ii'd3 'i!kaS! 10 i.e2 b4 1 1 lt:Jd1 c5 ! ) 9 . . . .i.e6! (Kasparov's suggestion) 10 'i!kc2 'i!t'a5 and Black has the initiative.

d) 7 e4 b5 8 'i!t'd3 b4 9 lt:Ja4 (or 9 lt:Jce2 .i.a6 10 'ilkc2 0-0 1 1 i.d2 'i!t'b6 12 lt:Jf3 li:Jbd7 13 lt:Jg3 i.xfl 14 lt:Jxfl c5 and Black is much better in view of White's exposed king, E .Werner-I.Zalys, corres­pondence 1953) 9 . . . .i.a6 10 'ilke3 0-0 favours Black. 6 . . . 0-0

7 li:Jf3

Th e R us s i a n Sys te m

Once again White has a number of lesser options:

a) 7 .i.e2 lt:Jc6 8 lt:Jf3 transposes to Line A7, whereas 8 .i.e3 e5 9 lt:Jf3 exd4 10 lt:Jxd4 (N.Vanderhallen-R.Swinkels, Vlissingen 2005) 10 . . . lt:Je5 threatens . . . lt:Jfg4 and leaves White in trouble.

b) 7 f3 lt:Jc6 8 .i.e3 e5 9 d5 lt:Jd4 is better for Black.

c) 7 e5 li:Jfd7 8 lt:Jf3 lt:Jb6 9 'i!kcS .i.e6 intends . . . lt:Jc6 and . . . aS.

d) 7 h3 lt:Jc6 8 .i.e3 lt:Jd7 is unpleas­ant for White, in view of the corning . . . lt:Jb6 and . . . f5 .

e) 7 .i.f4 lt:Jc6 8 d5 (8 .l:!.d1 lt:Jd7! 9 li:Jf3 lt:Jb6 10 'i!t'c5 i.g4 1 1 d5 i.xf3 12 gxf3 lt:Je5 13 i.e2 'i!t'd6! 14 'i!t'e3 f5 ! gave Black the initiative in Y.Yakovich-P.Svidler, St Petersburg 1993) 8 . . . e5! 9 i.g5 (9 .i.e3 lt:Jd4! 10 .i.xd4? exd4 1 1 'ilkxd4 lt:Jxe4! 12 'ii'xe4 .l:te8 i s winning for Black) 9 . . . lt:Jd4 and Black was better in T.Jobe­M.Bestvina, correspondence 2000. 7 . . . lt:Jc6

Play has returned to Line A.

Conclusion

The Russian System is one of White's most threatening options. Against it I have opted for a positionally-oriented but still dynamic system, focusing on piece play against White's centre and exposed queen. White cannot fully con­tain Black's counterplay, and even his more ambitious tries do not succeed in reaching more than a complex position with mutual chances.

1 0 1

C h a pte r E i g h t I �a4+ Syste m s

1 d4 lt:Jf6 2 c4 g6 3 lLlc3 d S 4 lt:Jf3 Should White wish to check, he is

recommended to delay it for a move. Quite simply the immediate 4 Vi'a4+ is less precise in view of the forceful re­sponse 4 . . . �d7 5 Vi'b3 lt:Jc6 ! .

Now . . . lt:Ja5 i s threatened and after the logical sequence 6 lt:Jf3 ( 6 Vi'xb7? �b8 7 Vi'a6 lt:Jb4 simply wins for Black, while 6 cxd5 lt:Jxd4 7 Vi'd1 lt:Jb5 8 �d2 lt:Jxc3 9 �xc3 c6 10 e4 �g7 11 dxc6 �xc6 12 �d3 0-0 13 Vi'e2 lt:Jh5! leaves Black with much the superior devel-

1 02

opment, I .Kan-P.Dubinin, USSR Ch., Leningrad 1947) 6 . . . lt:Ja5 7 Vi'b4 lt:Jxc4 8 lt:Jxd5 lt:Jxd5 9 Vi'xc4 lt:Jb6 1 0 Vi'c2 i.g7 Black has the initiative (Botvinnik). 4 ... �g7 5 Vi'a4+ �d7 6 Vi'b3 dxc4! 7 Vi'xc4

Of course, White can't play 7 Vi'xb7? due to 7 . . . lt:Jc6! 8 �f4 (8 Vi'a6? lt:Jb4 9 Vi'xc4 lt:Jc2+) 8 . . . �b8 9 Vi'xc7 Vi'xc7 10 �xc7 �xb2 when Black i s much better, T .Sorri-I .Koskimaa, 1981 . 7 ... 0-0

correspondence

We have now reached a position very similar to the Russian System, with Black having played the extra move . . . �d7. White hopes that this move is actually harmful for Black, by virtue of releasing some of the pressure against d4 and restricting the light­squared bishop's options. However, a tempo is a tempo and Black can make good use of his extra move with a quick . . . b5.

White's main options are now:

A: 8 .i.f4 8: 8 e4

Instead 8 e3 ctJa6 !? (8 . . . .i.e6 9 'ika4 c5 is a simple equalizer) 9 "ifb3 (9 ctJe5? is a mistake due to 9 . . . c5! 10 dxc5 .i.e6 intending . . . ctJd5, while 9 .i.e2 c5 10 0-0 cxd4 11 exd4 "ifb6 also gives Black the upper hand) 9 . . . c5 10 .i.xa6 bxa6 1 1 0-0 cxd4 12 ctJxd4 'Yi'a5 13 .i.d2 .l:!.ab8 14 'Yi'c2 .l:!.fc8 left Black for preference in E.Griinfeld-L.Pachman, Vienna 1949.

A) 8 .i.f4 A valid option, just as in the Rus­

sian System. 8 ... tt:la6

This seems the most logical. Black will play . . . c5 next and is ready to follow with . . . b5 if needed, while the c8-square is now available for his major pieces. 9 e4

Alternatively, 9 l:td1 ?! c5 10 dxc5 'ika5 11 e4 .l:!.ac8 ( 1 1 . . . .l:!.ad8 ! ?, intending . . . .i.e6, offers Black good play and is probably a better choice) 12 e5 .l:!.xc5 13

'Wi a4 + Sys t e m s

"ifb3 (13 exf6? .l:!.xc4 14 fxg7 .l:!.e4+ 15 .i.e3 .l:!.d8 was winning for Black in E .Ermenkov-Ghizdavu, Varna 1973) 13 . . . ctJe4 14 .l:!.xd7 .l:!.xc3 15 bxc3 tt:Jac5 16 "ifb4 'Yi'xb4 17 cxb4 tt:lxd7 reaches an equal position. g . . . cs ! 10 es

The 10 d5 of I .Solomunovic-I.Saric, Pula 2004, should be met by 10 . . . b5! 11 tt:Jxb5 tt:lxe4 12 'ikxe4 .i.xb5 13 .i.xb5 'Yi'a5+ 14 ctJd2 'ikxb5 15 ctJc4 "ifb4+ 16 .i.d2 "ifb7 with an edge in view of the coming . . . e6. Note that 10 dxc5 'Yi'c8 is also fine for Black. 10 ... tt:Jhs 11 .i.e3 cxd4 12 'Yi'xd4 .i.c6 13 .i.e2 .i.xf3 14 .i.xf3 'Yi'xd4 15 .i.xd4 ctJb4 16 0-0-0 tt:lf4

Black has equalized, A.Riazantsev­E.Sutovsky, Sochi 2005 .

8} 8 e4 bS !

9 'ikc5 c6 plans an awkward . . . tt:la6, so White usually chooses between:

81: 9 'Yi'b3 82: 9 tt:Jxbs

1 03

Play th e G r u nfe ld

81) 9 'i¥b3 cs

Logical, and now we reach an im­portant tabiya for this variation. 1o dxcs

Two critical alternatives are: a) 10 i.xb5?! .i.xb5 11 liJxb5 (1 1 'ii'xb5

cxd4 12 li::\e2 a6 13 'iVd3 e5 is just better for Black) 1 1 . . .liJxe4 12 0-0 cxd4 13 'ifc4 liJd6 14 'iVd5 liJd7 15 liJbxd4 liJb6 16 'ifh3 liJbc4 17 l:d1 was Y.Anikaev­V.Malishauskas, USSR 1983, and now 17 . . . l:tc8 leaves Black clearly better.

b) 10 e5 li::\g4 and now: b1 ) 11 h3 cxd4 12 hxg4 dxc3 13

'ii'xc3 i.xg4 gives Black the initiative. b2) 1 1 'iVd5 cxd4! ( 1 1 . . .li::\c6 ! ? 12

"Bxc5 l:.c8 ! is an interesting suggestion of Kasparov' s) 12 li::\xd4 (12 'ii'xa8 dxc3 is much better for Black) 12 . . . 'ii'b6 13 i.xb5 i.xe5 and, according to Kasparov, Black has a serious advantage.

b3) 1 1 i.xb5 cxd4 12 li::\xd4 i.xb5 (12 . . . i.xe5 !? is also interesting: 13 .i.xd7 li::\xd7 is clearly better for Black and after 13 'ii'd5 'ifb6! - Kasparov -White's position looks very loose) 13 liJdxb5 a6 14 li::\a3 (safer is 14 "Ba4 li::\xe5 with just an edge for Black) 14 . . . "iVd4!

1 04

15 'iVc2 (Black gains a strong attack af­ter 15 0-0 "Bxe5 1 6 g3 �5 - Kasparov) 15 . . . li::\c6 16 'ii'e2 'ii'xe5! 17 "ib'xe5 ( 17 li::\c4 "ib'xe2+ 18 li::\xe2 liJb4 19 0-0 l:!.ac8 again leaves Black much better) 17 . . . li::\gxe5 18 0-0 liJd3 19 l::tb1 l:tab8 20 l:td1 l:.fd8 21 Cit>£1 £5! and Black was clearly on top in R.Hiibner-G.Kasparov, Brussels 1986. 10 . .. li::\a6 11 es li::\g4

12 .txbs Alternatively, 12 e6 li::\xc5 13 exf7+

(13 �4 li::\xe6 14 'iVxg4 liJd4 15 'iVxd4 .txd4 16 li::\xd4 b4 17 li::\e4 eS wins for Black) 13 . . . Cit>h8 14 'ii'a3 (U.Osieka­M.Pein, Lugano 1986) 14 . . . �6 is clearly better for Black, and 12 i.f4 liJxcS 13 'ii'd5 "ib'b6 14 'ii'd2 i.c6 was again much better for Black in R.Goletiani­M.Neubauer, Port Erin 2005. 12 .. . l:tb8

As Adorjan observes, Black has good compensation here; he now threatens . . . li::\c7.

82) 9 li::\xbs li::\xe4! 10 'ii'xc7 The only way for White to justify

his opening play, but a risky venture nonetheless. Alternatively:

a) 10 'iYd5? c6 1 1 'iYxe4 Af5! leaves White in big trouble.

b) 10 tbxc7? tbc6 11 t:Dxa8 'iYa5+ 12 tbd2 (or 12 Ad2 tbxd2 13 tbxd2 tbxd4 14 tbc7 Ac6 with the advantage) 12 . . . t:Dxd4 13 'iYc7 'iYf5 14 t:Dxe4 'iYxe4+ 15 Ae3 tbc2+ and Black was already winning in S.Kiselev-V.Kozlov, Moscow 1986.

c) 10 'iYc2 was seen in V.Golod­J .Gustafsson, Dieren 1999, and is best met by 10 . . . Ac6 11 tbe5 AxeS 12 dxe5 'iYd5 13 f4 l:!d8 with a strong initiative.

d) 10 Ad3 is tame and after 10 . . . tbd6 Black clearly has no problems.

10 . . . tbc6 Davies suggests 10 . . . tba6? 11 'iYxd8

l:!fxd8, but I don't see what to do after the calm 12 a3, intending Ae2 and 0-0. 11 Ad3 tbd6!

This move makes great sense to me. Black exchanges White's active pieces and increases his lead in development. Instead 1l . . .tbb4 12 Axe4 Axb5 13 'iYxd8 l:!axd8 is often quoted as a simple equal­izer, but I beg to differ. After the accurate 14 a3! Ac6 (or 14 . . . tbd3+ 15 Axd3 Axd3 16 Ae3 Ae4 17 �e2 Axf3+ 18 �xf3 Axd4 19 Axd4 l:!xd4 20 l:!hd1 with an endgame

'ilV a4 + Sys t e m s

edge; note that the b4-square i s unavail­able to the black rook) 15 axb4 Axe4 16 l:!xa7 Axf3 17 gxf3 White can definitely hope for more than a draw. 12 'iYxdS l:!axd8 13 t:Dxd6 exd6 14 Ae3 tbb4!

Giorgadze feels that White is much better after 14 . . . Ag4? ! 15 0-0-0 Axf3 16 gxf3 Axd4 17 Ae4 Axe3+ 18 fxe3 tbb4 19 l:!d4! and I have to agree. 15 Ab1 Abs

With the white king stuck, Black has at least good compensation for the pawn. He will most likely obtain the two bishops with . . . tbd3+, while White has trouble activating his rooks.

Concl usion

The lines in this chapter are hardly dangerous for Black. The check must be met by . . . Ad7 and then Black can make use of his extra move, in com­parison with the Russian System, to play a quick . . . b5 . It turns out that the tempo gained by Black is more signifi­cant than the slight disharmony caused by the position of the bishop on d7.

1 05

C h a pte r N i n e I iL f4 Syste m s

1 d4 ti:Jf6 2 C4 g6 3 ti:Jc3 d S 4 i.f4 Rather than immediately occupy

the centre with his pawns, White opts initially for piece play. 4 . . . i.g7

There are now three main options:

A: 5 .l:!.c1 B: 5 e3 C: 5 etJf3

White also has a number of less im-

1 0 6

portant options; a situation typical, I'm afraid, of the i.f4 systems in general . There is a fair amount of theory, albeit mostly harmless for Black!

Those 5th move alternatives: a) Immediately capturing on c7 is

not advisable: 5 cxd5 tt:Jxd5 6 tt:Jxd5 'ii'xd5 7 i.xc7 tt:Ja6 8 i.e5 (8 i.f4 'ii'xd4 9 'i¥xd4 .ixd4 regains the pawn with ad­vantage, V.Kucera-P.Kozak, Prague 2005) 8 . . . i.xe5 9 dxe5 'ii'xe5 10 'ii'd2 i.g4 and Black has a strong initiative.

b) 5 i.e5 is best met by 5 . . . dxc4! 6 e4 (or 6 e3 tt:Jc6 7 'i¥a4 0-0 8 .ix£6 .ix£6 9 i.xc4, M.Aadrians-J.Dekic, Budapest 1 996, and now 9 . . . i.d7 when Black threatens both . . . ti:Jxd4 and . . . e5, with a strong initiative) 6 . . . tt:Jc6 7 i.xc4 (7 f4 0-0 8 .ixc4 tt:Jg4 is better for Black) 7 . . . tt:Jxe5 8 dxe5 'i¥xd1+ 9 .l:!.xd1 tt:Jg4 10 ti:Jb5 i.xe5 1 1 ti:J£3 i.d6 12 .l:!.xd6 (Black is also on top after 12 h3 tt:Je5 13 tt:Jxe5 i.xe5 14 .l:!.d5 .i£4) 12 . . . cxd6 13 tt:Jc7+ �d8 14 tt:Jxa8 i.e6! and Black is better (Uhlmann).

c) 5 'iVa4+ �d7 6 'Yib3 tt:Jc6! is a strong response to the check, just as in the previous chapter, and comes with a double threat:

7 e3 (instead 7 cxdS? tt:Jxd4 8 'iVd1 tt:JbS 9 tt:JxbS �xbS 10 e4 �xfl 11 �xfl 0-0 was horrible for White in A.Rothman-S.Reshevsky, New York 1946, and very risky is 7 'iVxb7 �b8 8 'iVxc7 'iVxc7 9 �xc7 �xb2 10 0-0-0 �b7 11 i.f4 tt:Je4 12 tt:Jxe4 dxe4 13 e3 eS with abundant compensation, since White's king is a target and his kingside still asleep) 7 . . . tt:Ja5 8 'Yib4 tt:Jxc4 9 �xc4 dxc4 10 'iVxb7 'Yib8! 1 1 'iVxb8+ �xb8 (the queen exchange has left White with the difficult task of protecting all his weaknesses, while Black also benefits from the bishop pair) 12 0-0-0 �b7 13 i2Jf3 �g4 14 �es 0-0 15 �d2 �fb8 16 �b1 �f5+ 17 �a1 tt:Je4 18 tt:Jxe4 �xe4 and Black was at least equal in N.Spiridonov-V.Jansa, Polanica Zdroj 1979.

A) 5 �c1 A tricky move, discouraging both

. . . cS and . . . 0-0. However, Black has an

�!4 Sys t e m s

enticing possibility a t his disposal. s . . . tt:Jhs!?

I t is surprising just how much dis­ruption is caused by this move! 6 �gs

The only good reply. Rowson ex­plains why simple moves won't suffice :

a) 6 e3?! tt:Jxf4 7 exf4 dxc4 and Black is better.

b) 6 �eS? ! �xeS 7 dxeS d4! gives Black the initiative .

c) 6 'iVd2? ! tt:Jxf4 7 'iVxf4 dxc4 again with the initiative.

d) 6 �g3 tt:Jxg3 7 hxg3 dxc4 8 e3 0-0 9 �xc4 cS! 10 dxcS i2Jd7 with good play on the dark squares.

e) 6 �d2 is relatively preferable, but also rather passive. A.Lauber­J.Gustafsson, German League 1999, con­tinued 6 . . . c5! 7 e3 (7 dxcS d4 8 tt:JdS is well met by 8 . . . tt:Ja6, while 8 tt:Ja4 can be met by Krasenkow's 8 . . . �d7!?, followed by . . . ..ltc6) 7 . . . cxd4 8 exd4 dxc4! (the resultant IQP position is promising for Black) 9 dS (9 ..ltxc4 0-0 10 dS trans­poses, whereas 10 i2Jf3? ! �g4 11 0-0 tt:Jc6 12 dS tt:Jd4 is good for Black) 9 . . . 0-0 10 �xc4 i2Jf6 1 1 tt:Jge2 ( 1 1 i2Jf3 allows

1 0 7

Play t h e G r ii nfe la

1 1 . . . .1Lg4 12 h3 .ltxf3 13 Vi'xf3 lt:Jbd7, fol­lowed by . . . lt:Je5 or . . . lt:Jb6 and then . . . �c8) 1 l . . .lt:Jbd7 12 .lte3 and now sim­ple and good was 12 . . . lt:Je5 13 ..tb3 .ltf5 14 .ltc2 .ltxc2 15 �xc2 Vi'd6, increasing the pressure against d5. 6 . . . h6!

The bishop must be kicked.

7 .lth4! Again the most consistent, whereas

7 .ltd2 dxc4! 8 e3 enables Black to retain the c4-pawn by playing 8 . . . ..te6! (seeing that lt:Jg5 is no longer possible for White) 9 lt:J£3 (Black was better after 9 lt:Je4 0-0 10 lt:Jf3 ..td5 1 1 Vi'c2 b5 12 b3 cxb3 13 axb3 c6 14 lt:Jc5 f5 15 ..te2 lt:Jd7 in W.Heinig-A.Colovic, Bad Woris­hofen 2005) 9 . . . c6 10 lt:Je4 ..td5 11 Vi'c2 b5 12 lt:Jc5 ..txf3 ! 13 gxf3 Vi'd5 14 b3 cxb3 15 axb3 Vi'xf3 16 �g1 0-0. With . . . e5 coming, I don' t see any compensa­tion for White. 7 . . . dxc4! 8 e3 ..te6

It is not clear how White is planning to recover the pawn on c4. 9 .lte2

Instead both 9 lt:Jf3 lt:Jd7 10 d5 .ltg4 1 1 .ltxc4 0-0 12 0-0 lt:Jb6 13 .lte2 g5 and

1 0 8

9 lt:Je4 0-0 10 lt:Jc5 .\td5 1 1 e4 b6 12 exd5 (J.Szabolcsi-J .Chabanon, Eger 1992) 12 . . . bxc5 13 dxc5 lt:Jf4 leave Black bet­ter. 9 . . . lt:Jf6! 10 lt:Jf3

Of course 1 0 Vi'a4+? c6 1 1 .ltxc4? b5 is impossible, while 1 0 ..tx£6 is well met by 10 . . . exf6! followed by . . . f5 (Rowson). 10 ... c6

White's problem is not only the pawn minus, but also the fact that the extra c4-pawn greatly restricts his pieces. u lt:Jes

Alternatively, 11 0-0 (J.Horvath­T.Fogarasi, Zalakaros 1994) 1 l . . .lt:Jbd7! is fine for Black, as 12 lt:Je5? lt:Jxe5 13 dxe5 Vi'xd1 14 �fxd1 lt:Jg4 drops the e5-pawn, while 1 1 a4 lt:Jbd7 12 lt:Jd2 lt:Jb6 13 0-0 0-0 14 .ltxf6 exf6 15 lt:Jce4 (I.Stavrianakis-A.Sismanis, Ermioni 2005) 15 . . . ..td5 allows White to recover the pawn, but at the cost of the bishop pair; Black stands well. 11 ... bs 12 f4

12 0-0 lt:Jd5! 13 lt:Je4 ..txe5! 14 dxe5 lt:Jd7 is good for Black. 12 ... ttJds

Now the 13 'ilid2 of A.Dreev-P.Leko, Wijk aan Zee 1996, can be met by 13 . . . gS ! ? 14 fxgS �xeS 1S dxeS hxgS 16 �xgS 4Jd7 17 0-0 lt:JxeS when Black is better. White might thus try and pre­pare lt:Je4 with 13 �f2 !?, but after 13 . . . 4Jb6! ? 14 lt:Je4 4J6d7! 1S b3 lt:JxeS 16 fxeS �dS 17 �f3 0-0 he is badly strug­gling for any compensation following Black's purposeful play.

B) 5 e3 0-0!

The alternative is S . . . cS, but that al­lows White some pretty dangerous possibilities which I prefer to avoid. Furthermore, it seems that Black has no reason to fear the capture on c7, after which he can obtain fully adequate play, as we shall see below.

White can now continue developing or grab on c7:

B1: 6 cxds 82: 6 4Jf3

The rare alternatives at this juncture scarcely merit much attention:

i.j4 Sys t e m s

a ) 6 �eS e6 7 4Jf3 4Jbd7 8 �g3 (M.Botvinnik-V.Smyslov, World Cham­pionship, Moscow 19S4) 8 . . . cS ! ? 9 cxdS lt:JxdS 10 �e2 cxd4 1 1 lt:JxdS (or 1 1 exd4 lt:J7f6 intending . . . 4JhSxg3) 1 l . . .'tliaS+ 12 'ilid2 'ilixdS gives Black an edge; . . . b6 and . . . �b7 is the plan.

b) 6 h3 cS 7 4Jf3 (7 dxcS 'tliaS 8 'ilia4 'ilixcS is also fine for Black) 7 . . . cxd4 8 exd4 4Jc6 9 �e2 dxc4 10 �xc4 'ilVb6 gives Black a promising position.

c) 6 'ilVb3 does nothing to discourage 6 . . . cS ! and then 7 dxcS (7 cxdS cxd4 8 exd4 4Jbd7 is a line of the Panov­Botvinnik Attack with White having played an unusual and early �f4; that gives Black good play after 9 �e2 4Jb6 10 �f3 �fS 1 1 .l:i.d1 aS or 1 1 lt:Jge2 �d3 ! ) 7 . . . 4Je4! 8 cxdS (8 lt:Jxe4 dxe4 9 lt:Je2 'iliaS+ 10 lt:Jc3 lt:Ja6 is better for Black, while there are several ways for White to go wrong here: 8 �xb8? dxc4 9 'ilic2 lt:Jxc3 10 bxc3 �fS, 8 lt:Jge2? lt:JxcS 9 'ilic2 eS 10 �g3 �fS and 8 lt:JxdS? 'iliaS+ 9 'ilVb4 'ilixb4+ 1 0 lt:Jxb4 �xb2) 8 . . . 'iliaS 9 lt:Jge2 lt:JxcS gives Black the initiative and excellent compensation after both 1 0 'tlic4 eS! and 10 'ilid1 4Jba6 ! .

d ) 6 .l:i.cl i s more notable, in that it discourages . . . cS and threatens the c7-pawn at a moment when . . . dxc4 is not an option. However, Black has a strong riposte in 6 . . . �e6!, after which 7 4Jf3 (7 'ilVb3 b6 8 4Jf3 cS gives Black good and active play; for example, 9 cxdS lt:JxdS 10 lt:JxdS �xdS 11 �c4, J .Michenka­P.Svanda, Czech Team Ch. 199S, and now 1 l . . .i.xf3 12 gxf3 cxd4!? 13 �dS lt:Ja6 14 �xa8 'ilixa8 with the initiative) 7 . . . dxc4 transposes to Line C2, below.

1 09

Play t h e G r ii nfe ld

81) 6 cxds White takes up the challenge and

accepts the offered pawn. 6 . . . tLlxds 7 tLlxds 'ii'xds 8 i.xc7

Unsurprisingly 8 lLlf3 c5 9 i.e5 tLlc6 10 .ixg7 'lt>xg7 just cedes the initiative to Black.

8 . . . tLla6! The best move, developing with

tempo. Instead . . . tLlc6 makes much less sense here as it doesn't gain time against d4. 9 i.xa6

Retreating the bishop is too compli­ant: 9 Sl..g3 e5! 10 Sl..xe5 Sl..xe5 1 1 dxe5 'ii'a5+ 12 'ii'd2 tLlb4 (now . . . �d8 is threatened) 13 lLlf3 �d8 14 tLld4 tLlc2+ 15 We2 'ii'xe5 16 'ii'xc2 �xd4 with excel­lent compensation; the c8-bishop will be developed with tempo and White's king is rather misplaced. 9 . . . 'ii'xg2 10 'ii'f3 'ii'xf3 11 lLlxf3 bxa6

A very interesting situation has arisen. Black is strong on the light squares, but White can easily evacuate the long h1-a8 diagonal after which he retains a firm hold on the key e5-square, thereby preventing Black from

1 1 0

freeing his g7-bishop with . . . e5. Before White can consolidate, with something like 'lt>e2 and lL\d3 followed by dou­bling on the c-file, Black must aim to gain counterplay. In particular, he can enforce the . . . e5 break with the help of .. .£6, in order to open the position for his bishop pair. 12 �g1!

A necessary precaution, preventing . . . i.g4 and thus preparing 'lt>e2. Alter­natively:

a) An immediate 12 'lt>e2? ! i.g4 leaves White awkwardly pinned; for example, 13 �hg1 (or 13 h3 ii.h5 when . . . f5-f4 is on the agenda) 13 . . . i.h5 14 �g3 �ac8 15 �cl e5 16 'lt>d2 .ixf3 17 �xf3 exd4 18 exd4 ii.xd4 and Black was much better in Chan Peng Kong­D.Semerene, Turin Olympiad 2006.

b) 12 'lt>d2?! Sl..b7 is another rather unpleasant pin.

c) 12 0-0? ! avoids any nasty pins, but does nothing about . . . e5 : 12 . . . £6 13 �acl ii.b7 14 tt:ld2 e5 15 tt:lb3 �£7 16 �c3 exd4 17 tt:lxd4 £5 left Black better in N .Karaklajic-S.Gligoric, Belgrade 1962.

d) 12 �cl f6 doesn't change the situation much: 13 �g1 (13 'lt>e2? ! i.g4! is again annoying and after the 14 �c4 of Y.Pelletier-M.Illescas, Pamplona 2003, Black has 14 . . . e5! ? 15 dxe5 �ac8 16 �xg4 - 16 ex£6? Sl..h5 ! 17 fxg7 Sl..x£3+ 18 Wd2 �fd8+! 19 Sl..xd8 �xd8+ 20 'lt>c3 ii.xh1 wins for Black - 16 . . . �xc7 with excellent compensation - Illescas) 13 . . . SI..b7 14 'lt>e2 �f7 15 �c5 (V.Erdos­T.Fogarasi, Budapest 2005) 15 . . . e5 will be followed by . . . �e8 with a strong at­tack for Black.

12 .. . f6 13 �e2 l:l.f7 14 l:l.ac1 .id7 !

Now the bishop can move to b5 with check, while it may also block the path of White's d-pawn (after a subse­quent . . . e5, d5-d6) . 15 tt:Jd2

Note that 15 l:l.c3 does not prevent 15 . . . e5 ! and after 16 d5 (16 dxe5 .ib5+ 17 �el fxe5 leaves Black actively placed in view of 18 tLlxe5? l:l.xc7! 19 l:l.xc7 .ixe5) 16 . . . .ib5+ 17 �d2 .if8 (the bishop switches to a better diagonal) 18 a3 l:l.c8 19 l:l.gcl l:l.d7 20 .ixe5 (20 e4?? .ih6+ drops a rook) 20 .. . l:l.xd5+ 21 �e1 l:l.xc3 22 .ixc3 Black has at least equal­ized. 1s ... es 16 ds .ibs+ 17 �f3

Despite White's creative play with his king, I feel that Black has a good game here. Now 17 .. . £5 18 b3 e4+ 19 \¥;>g2 .ib2 20 l:l.c2 .id3 21 l:l.c6 .ib5 22 l:l.c2 .id3 23 l:l.c6 was roughly level when T.Radjabov-V.Ivanchuk, Morelia 2006, was agreed drawn here. Black should also consider the untried 17 . . . l:l.d7! ? when 18 e4 (or 18 d6 f5 with the initiative) 18 . . . £5 threatens . . . .ih6 and gives him good activity.

Sif4 Sys t e m s

B2) 6 tt:Jf3 cs !

7 dxcs The only critical move should White

want to prove anything in the opening. Instead, allowing Black to exchange on d4 either results in a typical IQP posi­tion with good chances for Black, or in the concession of central space. 7 ... 'iVas

Black aims to regain the c5-pawn and also threatens . . . tt:Je4. One may have seen that many games with a 5 . . . c5 6 dxc5 .ig7 move order also reach this position, but I prefer our 5 . . . 0-0 move order: (i) the capture on c7 is then harmless for Black, as we saw in

1 1 1

P lay th e G r ii nfe /d

Line B1; and (ii) in the alternative 5 . . . c5 move order White retains the danger­ous option of developing his king' s knight to e2. 8 l:rc1

The only serious option for White. He has many alternatives, but they are again all rather harmless:

a) 8 cxd5?! tt::lxd5 9 i.e5 tt::lxc3 10 'i!Vd2 .ltxe5 1 1 tt::lxe5 'i!Vxc5 12 'i!Vxc3 'i!Vxc3+ 13 bxc3 b6 14 .te2 .tb7 15 0-0 f6 16 .\t£3 .ltxf3 17 tt::lx£3 tt::la6 18 a4 l:rfc8 gives Black an obvious advantage, as he will double rooks against the c3-pawn.

b) 8 .lte2 tt::le4 9 0-0 tt::lxc3 10 bxc3 dxc4 11 .ltxc4 'i!Vxc5 was also good for Black in A.Kofidis-M.Turov, Halkida 2000.

c) 8 'i!Va4? ! 'i!Vxc5 9 'i!Vbs (9 .\te2 tt::le4 10 l:rcl tt::lxc3 1 1 bxc3 dxc4 12 .txc4 .td7 13 'i!Vb4? l:rc8 14 tt::ld2 'i!Vxb4 forced White to resign in A.Bely-E.Gasanov, Kharkov 2004, as 15 cxb4 b5 wins a piece, while 9 cxd5 tt::lxd5 10 tt::lxd5 'i!Vxd5 11 i.c4 'i!Vh5 12 'i!Va3 tt::lc6 13 0-0 .tg4 gives Black the upper hand) 9 . . . 'i!Vxb5 10 tt::lxb5 tt::la6 11 l:rd1 .lte6 12 tt::lfd4 (or 12 tt::lg5 i.g4 13 f3, C.Cacco­E.Arlandi, Cremona 2005, and now 13 . . . .td7 14 cxd5 h6 15 tt::le4 tt::lxe4 16 fxe4 tt::lc5 gives Black very good com­pensation for the pawn) 12 . . . .td7 13 .teS l:rfd8 14 .te2 (14 cxd5 tt::lxd5 15 i.xg7 'lt>xg7 16 .\te2 e5 is also better for Black) 14 . . . dxc4 15 .txc4 tt::lc5 16 0-0 tt::la4! and Black was on top in H.Rossetto-S.Gligoric, Mar del Plata 1950.

d) 8 tt::ld2 blocks the vulnerable e1-

1 1 2

aS diagonal, but looks rather passive: 8 . . . dxc4 9 .txc4 (the other options are weaker; for example, 9 tt::lxc4 'i!Vxc5 10 .te2 .te6 1 1 tt::ld2 tt::ld5 12 tt::la4 'i!Va5 13 .txb8 l:raxb8 14 0-0 l:rfd8 won material due to the threat of . . . tt::lxe3 in A.Dunkelblum-R.Wade, Munich 1954) 9 . . . 'i!Vxc5 (it now becomes clear that White benefits more from the l:rcl of our main line than from tt::ld2)

10 .te2 (White rather lacks a good move here; for instance, 10 0-0 tt::lh5 1 1 .\txb8 l:rxb8 12 'i!Vb3 tt::l£6 1 3 l:racl, H.Sorensen-C.Ekeberg, Gausdal 2001, and now 13 . . . b5 gives Black the initia­tive, while 10 'i!Vb3 tt::lc6 11 'i!Vbs 'i!Vxb5 12 .ltxb5 tt::lb4 13 0-0 tt::lfd5 14 tt::lde4 tt::lxf4 15 exf4 .ltf5 was good for Black in A.Dunkelblum-R.Byrne, Leipzig Olym­piad 1960) 10 . . . tt::lc6 1 1 l:rcl 'i!V£5 12 h3 .\te6 13 'i!Va4 a6 14 g4 'i!Vc5 15 0-0 tt::ld5 and Black is slightly better, E .Ungureanu-H.Glauser, Lugano Olympiad 1968.

Returning to the more critical 8 l:rc1 : 8 . . . dxc4! 9 lt..xc4 'i!Vxcs

White now has several options, but none trouble Black unduly. The black

queen may appear somewhat exposed on c5, but there is no concrete way of exploiting her position.

10 ..11l.b3 The most natural move. Alternatively: a) 10 'i'e2 ..11l.g4! is awkward: 11 CDb5

(or 1 1 0-0 CDc6 12 ..11l.b3, as in S.Aarland­M.Tal, Riga 1954, and now 12 . . . 'i'h5 13 h3 ..11l.xf3 14 'i'xf3 'i'xf3 15 gxf3 CDa5 gives Black an edge in view of White' s weakened structure) 1 1 . . . 'i'b4+ 12 'i'd2 (12 Wfl CDbd7 is worse for White) 12 . . . 'i'xd2+ 13 tt:Jxd2 CDc6 14 h3 �d7 15 �h2 .l:tfd8 16 CDb3 .l:tac8 17 0-0 a6 18 CD5d4 CL'lxd4 19 CDxd4 b5 with at least equality for Black in B.Kelly-R.Dineley, Turin Olympiad 2006.

b) 10 'i'd4 'i'h5 11 ..11l.xb8 .l:txb8 12 'i'xa7 was tried in V.Golod-Ma.Tseitlin, Ashdod 2004, and now the typical tac­tic 12 . . . ..11l.h3 ! promises Black excellent compensation.

c) 10 b3 CDc6 11 0-0 'i'a5 12 CDb5 �f5 13 a3 .l:tac8 14 b4 'i'b6 15 �c7 .l:txc7 16 CDxc7 'i'xc7 17 b5 (M.Kustanovich­M.Lurie, Petakh Tikva 1997) 17 . . . e5 18 bxc6 bxc6 leaves Black a pawn up for

iL/4 Sys t e m s

no real compensation. d) 10 ..11l.e2 CDc6 1 1 0-0 'i'b4 12 CDa4

.l:td8 13 'i'c2 ..11l.f5 14 'i'c4 (H.Urday­S.Roa, San Sebastian 1991) 14 . . . .l:tac8 and Black has the initiative.

e) 10 'i'b3 ! ? tt:Jc6 threatens . . . tt:Ja5 and now: 11 CDb5 (this would be a seri­ous problem if White's f3-knight was on e2, protecting the .l:tcl, but here Black has nothing to fear; note too that 1 1 tt:Jg5 CDh5 12 ..11l.xf7+ Wh8 13 0-0 tt:Jxf4 14 exf4 CDd4 leaves a number of White's pieces rather misplaced and liable to drop off) 1 1 . . . ..11l.e6! (renewing the threat of . . . CDa5 and exploiting the loose rook on cl) 12 CDc7 �xc4 13 CDxa8 was seen in I.Walch-C.Horvath, As­chach 2002, and now 13 . . . CDa5 14 'i'c2 .l:txa8 15 CDd2 .l:tc8 16 b3 'i'c6 17 bxc4 'i'xg2 gives Black a strong initiative.

f) 10 CDb5 ! ? ..11l.e6! is once again a good defence against the discovered check and now:

fl ) 11 ..11l.xe6 'i'xb5 12 ..11l.b3 (or 12 i.c4 'i'xb2 13 0-0 CDc6 14 'i' a4 CDd7 15 .l:tb 1 CDb6 16 .l:txb2 CDxa4 17 .l:txb7 CDa5 18 .l:tc7 CDxc4 19 .l:txc4 CDc3 with an edge for Black in M.Boehnke-O.Brendel, Berkel

1 1 3

Play t h e G r u nfe ld

2003) 12 . . . 'Llc6 13 'iVe2 'iVb4+ 14 'iVd2 'iVxd2+ 15 'Llxd2 'Llh5 1 6 �d5 'Llxf4 17 exf4 'Llb4 18 �e4 �xb2 gave Black a clear extra pawn in M.Dlugy­G.Kasparov, Saint John (rapid) 1998.

£2) 11 Sl..e2 'iVb4+ 12 'iVd2 'iVxd2+ 13 'Llxd2 (M.Ruiz Vinals-F.Steenbekkers, Calvia 2004) 13 . . . 'Lld5 14 b3 'Llc6 is again advantageous for Black.

f3) 11 'Llc7 (critical, but Black is happy to 'lose' the exchange like this! ) 11 . . . �xc4 12 b3! (instead 12 'Lld2? b5 13 'Llxa8 'Lld5 14 'Llxc4 bxc4 15 'iVa4 'Llxf4 16 ex£4 l:k8 wins for Black, while here 13 b3 e5 14 'Llxa8 ex£4 15 bxc4 fxe3 16 fxe3 'iVxe3+ leaves him with a clear edge; note too that 12 'Llxa8? ! 'iVa5+ 13 'iVd2 'iVxa2 14 'Lle5 .ie6 also favours Black) 12 . . . 'Llc6 13 'Llxa8 'iVa5+ 14 'iVd2 'iVxd2+ 15 'Llxd2 was seen in M.Dlugy-M.Chib­urdanidze, Brussels 1987; now 15 . . . �d5 16 'Llc7 �xg2 17 .l:i:g1 .ih3 leaves Black with the initiative for Black and good compensation for the exchange.

Returning to 10 Sl..b3:

10 . . . 'iYas 11 o-o 'Llc6 Black intends to continue with

. . . �f5 and the centralization of his

1 1 4

rooks, after which his pieces will be the more active and his position the more comfortable. This is why White often decides to unbalance the position, as we will see in the lines which follow.

12 h3 A necessary precaution since 12 'iVe2

allows the strong 12 . . . 'Llh5 ! disturbing the f4-bishop: 13 �g5 (both 13 �g3 'Llxg3 14 hxg3 .ig4 and 13 'Lld5 'Llx£4 14 exf4 e6 15 'Llc3 'iVb4 are good for Black) 13 . . . .ig4 14 �h4 (or 14 'Lld5 e6 15 'Lle7+ 'lt>h8 16 'Llxc6 'iVxg5 17 'Llcd4 il..xd4 18 exd4 'iVf4 19 'iVe5+ 'lt>g8 with equality in T.Roussel Roozmon-V.Mikhalevski, Montreal 2005) 14 . . . 'iVb4! (accurate and strong) 15 'iVc4 'iVxc4 16 �xc4 .l:i:ac8 17 .l:i:fd1 �f6 18 .ixf6 'Llxf6 19 h3 �xf3 20 gxf3 .l:i:fd8 and Black was most certainly not worse in M.Petursson-J.Smejkal, Thessaloniki Olympiad 1988.

Note, too, that 12 'Llg5 h6 13 'Llge4 'Llh5! (V.Tukmakov-L.Stein, USSR 1970) creates similar problems, while 12 'Lld5?! 'Llxd5 13 'iVxd5 Sl..xb2 14 .l:i:b1 Sl..g7 15 .l:i:fcl �f5 16 e4 �g4 was a completely unjustified pawn sacrifice in H.Tikk­anen-E.Hermansson, Gothenburg 2005.

12 . . . -tfs 13 'i'e2 13 tt:Jd4 looks like a way to activate

White's forces, but things tum out dif­ferently after 13 . . . .1td7! . Then 14 'i'e2 tt:Jxd4 1S exd4 e6 16 .ltd2 'i'b6 17 .Ufd1 .ltc6 18 .ie3 'i'aS was equal and agreed drawn here in A.Karpov-G.Kasparov, World Championship (Game 9), Lon­don 1986. White has also tried 16 .ieS, but after 16 . . . .ic6 17 .Ufd1 .Ufd8 18 'i'e3 (or 18 dS?! exdS 19 'i'f3 d4 20 'i'f4 lLlhS! 21 'i'x£7+ Wh8 22 .ltxg7+ tLlxg7 23 tLle2 .Uf8 24 'i'c4 'i'gS and Black's attack is very strong - Ivanchuk) 18 . . . .Ud7 19 'i'gS (Black defends and gains the advantage after 19 dS tLlxdS 20 .ixdS .ixdS 21 .ltxg7 Wxg7 22 tt:Je4! ? .Ue8 23 'i'f4 'i'd8) 19 . . . 'i'd8 20 tt:Ja4 h6 21 'i'g3 lLlhS Black was clearly on top in M.Petursson­V.Ivanchuk, Reggio Emilia 1989/90. 13 . . . tt:Je4

Now Black's pieces are very active and White's queens ide is beginning to come under some pressure. Thus White now often opts to sharpen the struggle .

14 tt:Jds ! This i s clearly best. White tries to

create counter-threats and intends to

Jij4 Sys tems

sacrifice an exchange on c6. Instead 14 tt:Jxe4 .ixe4 i s a little too compliant by White and leaves him the side strug­gling to equalize:

a) 1S .Ufd1 'i'hs 16 .ltc2 (16 .Uc4 .idS 17 .Uc2 .ixb3 18 axb3 .Ufd8 19 e4 'i'aS 20 .Ucd2 .Uxd2 21 .Uxd2 .Ud8 22 .Uxd8+ 'i'xd8 23 .lte3 a6 gave Black a small structural advantage in U.Andersson­D.Navara, European Team Ch., Goth­enburg 200S, although the legendary Swede managed to hold the draw) 16 . . . .txc2 !? ( 16 . . . .1txf3 17 'i'xf3 'i'xf3 18 gxf3 .Uad8 19 .te4 .txb2 20 .Uxd8 .Uxd8 21 .Uc2 .ia3 22 .ixc6 bxc6 23 .Uxc6 f6 equalized comfortably for Black in L.Portisch-K.Arakhamia Grant, Roque­brune 1998) 17 'i'xc2 'i'bs with good play for Black.

b) 1S tLlgS .idS 16 .ic7 (or 16 .ltxdS 'i'xdS 17 b3 .Uad8 18 .Ufd1 'i'fS with at least equality) 16 . . . 'i'xc7 17 .txdS (D.Sahovic-Z.Lanka, Yurmala 1978) 1 7 . . . 'i'aS 18 .ixc6 bxc6 is roughly equal, although Black's queenside pressure and superior minor piece gives him some hope of a possible grind.

c) 1S tt:Jd2 .idS 16 .ixdS (or 1 6 tt:Jc4 'i'bS 1 7 .Ufd1, Nguyen Anh Dung­M.Hoffmann, Budapest 1999, and now 17 . . . .Ufd8 with the initiative) 16 . . . 'i'xdS and Black was again at least equal in V.Hort-W.Uhlmann, Moscow 1971 . 14 .. . es !

A good response. In this rather theoretical position,

White faces an important choice be­tween entering the complications by taking on c6, and meekly retreating his bishop:

1 1 5

Play t h e G r ii nfe ld

821 : 15 .l:txc6 822: 15 �h2

15 �g5?! is rather dubious in view of 15 . . .<�:Jxg5 16 tt::lxgS 'iVd8! ; e .g. 17 tt:::lx£7 (or 17 tt:::l£3 e4! 18 tt::ld2 tt::laS ! and Black is better since he can answer 19 g4 with 19 . . . tt::lxb3! ) 17 . . . .l:txf7 18 tt::lc3 �d3 19 .l:tfd1 (Black is also for preference after 19 �xd3 �xd3 20 .l:tfd1 e4 21 tt::la4 .l:td8) 19 . . . 'iVxe2 20 tt::lxe2 .l:te8! 21 �xf7+ �xf7 22 tt::lg3 e4 and Black is much better.

821) 15 .l:txc6!?

1 1 6

This operation turns out unsuccess­fully, but it took several years to reach such a conclusion. Black used to ignore the rook and instead capture on f4, but my opinion is clear: the gauntlet should be taken up! 15 ... bxc6! 16 tt::le7+ Wh8

17 tt:Jxc6 Instead 17 tt::lxe5? �xeS 18 tt:lxc6

fails to the pretty 18 . . . �d2! when White ends up clearly worse in all lines: 19 �xd2 ( 19 �xeS+ f6 is also very good for Black) 19 . . . tt::lxd2 20 �xeS+ f6 21 �c3 (Black wins after 21 .l:td1 tt:Jxb3 22 axb3 fxeS, while 21 �d6 tt::lxfl 22 �xf8 .l:txf8 23 �xfl .l:tc8 24 tt:le7 .l:tcl + 2S We2 �e4 leaves him with some advantage) 21 . .. tt::lxfl 22 Wxfl �d3+ and Black was much better in I .Sokolov-A.Shirov, Wijk aan Zee 1999. 17 . . . �b6 18 tt:Jcxe5 �e6!

A very important move, neutraliz­ing the powerful b3-bishop. It was this discovery which turned the assessment of this critical variation in Black's fa­vour. 19 �xe6

Alternatively: a) 19 1i'c2 .ixb3 20 1i'xe4 .ie6 was

preferable for Black in V.Gavrikov­A.Kochiev, Tallinn 1997.

b) 19 l:i.cl l:!.ac8 20 l:!.xc8 l:!.xc8 21 tt:Jxf7+ (E.Frolik-W.Sauermann, Wuert 1 988) 2l . . . .ixf7 22 .ixf7 g5 with the initiative for Black.

c) 19 l2Jc4 1i'a6! also fails to improve matters for White: 20 ctJd4 (20 tLlg5? tL'lxg5 21 .ixg5 l:!.fc8 22 l:i.cl l:!.c5 wins ma­terial - Ftamik) 20 . . . .id5 21 l:i.cl l:!.ac8 22 'ilffl (22 f3 ctJd2 23 1i'xd2 .ixc4 is also good for Black) 22 . . . l:!.c5 23 ctJa3 l:i.xcl 24 'ilfxcl l:!.c8 and Black was on top in I.Shliperman-M.Ashley, New York 1999.

d) 19 ctJd4! ? (probably White's best alternative to 19 .ixe6)

19 . . . .ixb3 20 l2Jxb3 (instead 20 axb3 �adS 21 l:i.cl, L .Kwatschewsky­M.Neubauer, Hartberg 2004, and now 21 . . .l:!.d5 gives Black the initiative) 20 . . . 1i'e6 21 ctJf3 l:!.ac8 22 l:i.d1 l:!.fd8 23 �xd8+ l:!.xd8 24 l2Jfd4 1i'd7 25 1i'f3 l:!.e8 26 1i'e2 and White, with two pawns for the exchange, was just about holding on to equality in H.Karlzen­:\1.Carlhammar, Gothenburg 2004.

Sif4 Sys t e m s

19 . . . 1i'xe6 20 1i'c2 Similar are both 20 b3 and 20 a3

l:!.ac8; White is struggling slightly to demonstrate full compensation. 2o . . . fs

21 1i'a4 Instead 21 b3?! g5 22 .ih2 l:!.ac8 23

l2Jc4 g4 24 hxg4 fxg4 25 ctJfd2 ctJc3 26 ctJb1 1i'f5 27 1i'xf5 ctJe2+ 28 Wh1 l:!.xf5 gave Black an edge in J .Levitt-A.Shirov, British League 2005, while 21 1i'c4 1i'xc4 22 lLlxc4 l:!.ac8 23 lLlfe5 Wg8 is not dis­similar to what we have already seen, and after 24 f3 g5 25 .ih2 .ixe5 26 l2Jxe5 Black is surely not worse. 21 . . . gs 22 .ih2 1i'e8

Again trying to trade queens and then exploit the extra exchange in an endgame. That would not be desirable for White, but he must be careful in any case; not only are his knights vulner­able, but his bishop can become en­tombed on h2. An example of that final point is 23 ctJc6 l:!.c8 24 ctJfd4 (V.Kosyrev­K.Sakaev, Internet blitz 2004) 24 . . . l:!.f6 25 l:i.cl f4 (Krasenkow), blocking in the h2-bishop and securing Black a clear ad­vantage. Instead U.Adianto-M.Roiz, Biel

1 1 7

Play t h e G r ii nfe ld

2006, continued 23 'i¥a5 l:!.b8 24 'i¥xa7 (24 tbc4 l:!.b5 25 'i¥xa7 g4 26 lDd4 l:i.d5 27 'i¥b7 l:!.d7 28 'i¥b5 f4 was better for Black in A.Iljushin-V.Belov, Sochi 2006) 24 . . . l:!.xb2 25 lDc4 and now Black should have played 25 . . . l:!.c2 26 .lte5 l:!.f7

with at least equality.

B22) 15 .lth2 ct:Jcs

16 e4 Trying to establish some sort of

foothold in the centre, but abandoning control of the d4-square. Instead 16 .ltc4 i s strongly met by 16 . . . e4! when

l:!.cd1 'i¥d8 20 .ltc7 'i¥d7 21 .lte5 .lte6 22 tbe3 .ltxc4 23 Cbxc4 'i¥d5 with advan­tage in R.Loetscher-O.Brendel, Swiss League 2003.

b) 17 ctJd2 ctJd3 with an edge for Black.

c) 17 Cbe1 (J .Dobos-T.Hillarp Pers­son, Recklinghausen 1999) 17 . . . ctJd3 18 Cbxd3 exd3 19 .ltxd3 'i¥xd5 20 e4 .ltxe4 21 .ltxe4 'i¥xa2 22 �xc6 bxc6 23 l:!.xc6 l:!.fe8 and the b2-pawn will fall .

d) 17 Cbg5 (E.Magerramov-V.Jakovl­jevic, Abu Dhabi 2006) 17 . . . 'i¥d8! 18 Cbxf7 l:!.x£7 19 Cbc7 ctJd3 20 �xf7+ �xf7 21 Cbxa8 Cbxcl 22 l:i.xcl 'i¥xa8 and again Black has the advantage.

e) 17 Cbh4 �e6 18 .ltc7 (18 �d6? loses material to 18 . . . Cbd3 ! ) 18 . . . b6 19 l:!.fd1 Cbb4 20 Cbxb4 'i¥xb4 21 �xe6 fxe6 22 �d6 g5 ! 23 �x£8 l:!.x£8 24 b3 gxh4 and Black had a clear advantage in the game Bu Xiangzhi-S.Ganguly, Internet (blitz) 2006. 16 . . . l:!.ad8!

Further undermining the position of the strongly-placed d5-knight.

Black is better in all lines: 17 t:Dxes a) 17 ctJd4 Cbxd4 18 exd4 ctJd3 19 I believe that this forcing sequence

1 1 8

is White's best option here, but yet again he has tried a number of other approaches:

a) 17 �c4? �xe4 18 b4 tt:Jxb4 19 tt:Jxb4 �xf3 20 'iVxf3 'iVxb4 is very good for Black.

b) 17 exfS?! tt:Jxb3 18 .l:i.xc6 'iVxdS 19 Ilc7 (19 f6? t2Jd4 20 tt:Jxd4 exd4 21 fxg7 Ilfe8 wins for Black) 19 . . . tt:Jd4 20 tLlxd4 exd4 21 fxg6 fxg6 with an obvious edge for Black in view of his powerful passed d-pawn.

c) 17 'iVe3? ! tt:Jxb3 18 axb3 (18 .l:i.xc6 IlxdS! 19 exdS bxc6 20 axb3 'iVxdS is no improvement for White) 18 . . . �xe4! (a promising queen sacrifice) 19 l:.cS 'iVxcS 20 tLlf6+ (20 'iVxcS .l:i.xdS was much bet­ter for Black in P.Lukacs-L.Ftacnik, Stara Zagora 1990) 20 . . . �xf6 21 'iVxcS .ltxf3 22 gxf3 .l:i.d4 and Black's advan­tage is quite serious, especially as the h2-bishop is shut out of play.

d) 17 .l:i.xcS 'iVxcS 18 exfS .l:i.xdS 19 .ltxdS 'iVxdS 20 .l:i.d1 t2Jd4 21 tLlxd4 exd4 again leaves Black with an obvious ad­vantage.

e) 17 .l:i.fd 1 ! ? �xe4 18 .l:i.xcS �xf3 19 1/Ve3 �xd1 20 .l:i.xaS tLlxaS 21 �xd1 (or 21 tt:Je7+ 'et>h8 22 �dS f6 23 'iVxa7 tLlc6 24 .ltxc6 bxc6 2S tLlxc6 Ila8 with some ad­vantage for Black) 2l . . . .l:i.xdS 22 �f3 tLlc4! (Rowson suggests 22 . . . IlbS 23 b3 tLlc6, but this doesn't seem so clear to me after 24 �xc6 bxc6 2S 'iVxa7) 23 'iVxa7 IlaS 24 'iVxb7 .l:i.xa2 and again Black has more than enough for the queen. 11 . . . tt:Jxes 18 tLle7+ 'et>h8

The position remains rather unclear, but Black is certainly not worse. After 19 �xeS .ltxeS 20 exfS tt:Jxb3 21 axb3

i..j4 Sys t e m s

�f6 2 2 fxg6 fxg6 he gains the advan­tage, and even White's best try of 19 exfS tLlxb3 20 axb3 .l:i.fe8 21 b4 'iVxb4 22 �xeS 'iVxe7 23 �xg7+ 'et>xg7 24 1/Vxe7 .l:i.xe7 2S fxg6 hxg6! (Rowson) leaves Black with the more active rooks and a slight edge in the endgame.

C) 5 tLlf3 This position frequently also arises,

of course, via a 4 tLlf3 �g7 S �f4 move order . s . . . o-o

As in Line B, the c7-pawn is not our primary concern, although in this par­ticular case White is ill-advised to cap­ture it.

1 1 9

Play th e C r ii nfe ld

6 l:tc1 Dissuading . . . c5 . Alternatively: a) 6 e3 c5! transposes back to B2. b) 6 'iVb3 dxc4 7 'ifxc4 transposes to

note 'c' to White's 7th move in the Rus­sian System.

c) 6 cxd5 ctJxd5 7 ctJxd5 (Akobian's 7 ..ie5 is hardly an earth-shattering idea; Black can reply with either Naka­mura's combative 7 . . . ..ih6!? or 7 . . . ..ixe5 8 ctJxe5 ctJxc3 9 bxc3 ctJd7 10 CiJxd7 .ixd7 11 e4 c5 with equality) 7 . . . 'ifxd5 8 Jlxc7 (both 8 e3 c5 and 8 'iVb3 'i¥'xb3 9 axb3 c5! already see Black wresting the initiative) 8 . . . ctJc6 9 e3 Ylf5 sees Black mobilizing very quickly and with threats against the c2-square. Now:

cl) 10 .i.d3? is a blunder: 10 . . . J.xd3 11 'ifxd3 l:tac8 12 e4 'ifd7 (J .Slacik­Z.Zvan, Latschach 2001) and Black wins material with . . . ctJb4.

c2) 10 a3 l:tac8 1 1 ..ig3 ..ic2 ! 12 'ife2 ctJa5 13 ctJd2 e5! saw White facing seri­ous problems in S.Shestakov-V.Zilber­stein, USSR 1974.

c3) 10 'ifa4 �d7 1 1 .i.a5 ( 1 1 ..ig3 ctJxd4! 12 �xd7 ctJxf3+ 13 gxf3 Ji.xd7 was similar in J .Ochkoos-I .Zugic, On-

1 2 0

tario 1997) 1 1 . . .ctJxd4! 1 2 �xd7 ctJxf3+ 13 gxf3 Jlxd7 with an obvious edge, M.Dietze-P.Keres, Prague 1943.

c4) 10 Jle2 l:tac8 11 ..ig3 ( 11 ..if4 e5 12 ctJxe5 Jlxe5 13 dxe5 'ifxg2 14 Ylf3 'i¥'h3 does not help White much) 1 1 . . .�a5+ 12 �d2 (or 12 Wfl e5 13 dxe5 l:tfd8!? 14 'iVb3 ctJxe5 15 ctJxe5 .ltxe5 16 ..ixe5 �xeS with advantage in J.Heltzel­R.Ris, Hengelo 2002) 12 . . . ctJb4 13 0-0 (similarly unpleasant for White was 13 Ji.d1 ctJd3+ 14 Wfl �xd2 15 ctJxd2 ctJxb2 in H.Alavi Hour-D.Harika, Tehran 2002) 13 . . . l:tc2 with a significant superi­ority for Black m Y.Anikaev­T.Georgadze, USSR 1973. 6 . . . dxc4!

Because 6 . . . c5 is not enticing, as is usually the case when White has played l:tcl, this is the correct way of beginning counterplay.

White now has a choice between an ambitious central advance and a more restrained approach.

C1: 1 e4 C2: 1 e3

C1) 7 e4 With this White ensures the imme­

diate recovery of the pawn, but allows pressure against d4. That results either in the mangling of White's kingside structure or the ceding of the d4-square to the black knight. In both cases Black obtains good play on the dark squares. 7 ... i.g4! 8 .i.xc4

White must capture, as both 8 ite3 c5 9 dxc5 1i'a5 and 8 d5 c6! 9 .i.xc4 cxd5 10 exd5 ct::lbd7 11 0-0 .ticS 12 Jtb3 ct::lc5 (Adorjan and Dory) leave Black on top. 8 . . . .1txf3

White now faces a critical choice:

(11: 9 11Vxf3 (12: 9 gxf3

(11) 9 1i'xf3 Maintaining structural integrity at

the cost of allowing Black plenty of activity. 9 .. . ct::lc6

9 . . . 1i'xd4? is unplayable, of course, due to 10 ct::lb5. 10 ds ct::ld4

$.j4 Sys t e m s

11 1i'd3 Alternatively, 1 1 iVdl c5 12 0-0 a6

13 .tiel b5 14 .ltfl e6 15 .ltg5 h6 16 i.e3 (F.Lipinsky-D.Gross, Melk 1999) 16 . . . l:ie8 is fine for Black, while 11 1i'e3 c5 12 dxc6 (12 0-0 a6! 13 a4 1ib6 gives Black the initiative on the queenside) is a logical approach, but Black need not abandon his central outpost: 12 . . . bxc6 13 0-0 ct::lh5 14 i.g5 (I.Farago­M.Szekely, Budapest 1969) 14 . . . 1i'd6 15 l:ifdl l:iad8 16 ct::le2 c5 gives Black good play; the d4-knight is a powerful piece. 11 . . . ctJd7 !

The knight aims to later gain a tempo on the white queen and cru­cially e4-e5 is now prevented. 12 0-0 cs

I feel that Black has a good position here. He has a strong foothold in the centre and can expand on both wings. 13 i.b3

13 dxc6 bxc6 14 1i'g3 1i'a5 15 l:ifdl ct::lb6 gives Black great activity, while 13 a3 a6 is no improvement for White. 13 . . . l:ic8 14 itg3

14 ct::le2 ct::lxb3 15 1i'xb3 11Vb6 is fine for Black.

1 2 1

Play th e G r u nfe ld

14 . . . a6 15 f4 b5 Black's initiative grows after 16 e5

lt::lf5, threatening . . . c4 and . . . lt:Jc5, but 16 Jtf2 allows Black to strike back power­fully at White's seemingly imposing centre: 16 . . . e5! 17 dxe6 c4! 18 exf7+ l::!.xf7 19 �d1 (as Adorjan and Dory observe, 19 �d2? lt:Jc5! 20 l::!.cd1 l::!.d7 wins mate­rial) 19 . . . lt::lb6 20 Jtc2 l::!.xf4 and Black is for preference due to his active pieces, A.Feuerstein-V.Simagin, correspon­dence 1966.

C12) 9 gxf3 A more ambitious choice . White re­

tains his central influence by keeping his queen on d1, but now Black can initiate play against the split white kingside with . . . lt:Jh5 and . . . e5, eyeing the f4-square.

9 .. . lt::lhs ! 10 i..e3 e5 11 dxes 1 1 d5?! lt:Jf4 followed by . . . c6 and

. . . lt:Jd7 is not what White wants. 11 . . . i..xe5

Now the struggle revolves around whether White will be able to advance his e- and f-pawns or whether Black will succeed in keep them blockaded.

1 2 2

12 �xd8 In view of his weakened kingside,

White is well advised to exchange queens. Otherwise, . . . �f6 or . . . �4 greatly improves Black's chances; for example:

a) 12 'i!Vb3 lt::lc6 13 �xb7 �f6 14 .lte2 lLJd4 15 lLJd5 �4 saw Black' s pieces invade with some effect in A.Dreev­B.Alterman, Elista Olympiad 1998. That continued 16 'i�Vb4 (both 16 b3 a5 17 lt::lxc7 l::!.fb8 and 16 Jtd1 lLJe6 are also good for Black) 16 . . . lt:Jxe2 17 '>¥txe2 and now 17 . . . �3 would have left White under serious pressure; . . . f5 may well follow.

b) 12 .i..h6 lt::lc6! ? is a promising ex­change sacrifice: 13 .ltxf8 �xf8 14 lt::le2 lLJf4 15 lt::lxf4 .ltxf4 16 l::!.c3 (I.Rogers­A.Timofeev, Amsterdam 2005) 16 . . . l::!.d8 and Black has excellent long-term compensation due to his superb dark square control. 12 ... l::!.xd8

13 ctJe2 Most consistent with the positional

struggle taking place. Indeed, I don't really see the point of 13 b4?!, especially

since the strong reply 13 . . . lt:if4! 14 lt:id5 lt:ig2+ 15 �fl (15 �e2? c6! 16 lt:ie7+ �f8 17 ..ltc5 lt:if4+ 18 '<t>e3 i.d6 gives Black some advantage) 15 . . . lt:ixe3+ 16 fxe3 c6 leaves Black better in all lines, S.Lpu­tian-V.Ivanchuk, Elista Olympiad 1998.

White has also tried 13 0-0, which is reasonable but not really incisive enough: 13 . . . lt:ic6 14 i.d5 (or 14 lt:id5 tt:la5! 15 ..ltg5 l::!.d7 16 ..ltb5 c6 17 b4, M.Bosboom-J.Van de Martel, Rotter­dam 1999, and now Ftacnik's sugges­tion of 17 . . . cxb5 18 bxa5 l::!.e8 gives Black an edge) 14 . . . lt:id4 15 ..ltxd4 ..ltxd4 16 lt:ib5 (16 l::!.fd1 ..ltxc3 17 l::!.xc3 c6 18 .tb3 l::!.xd1+ 19 ..ltxd1 l::!.d8 20 ..ltb3 lt:if4 left Black with some advantage due to his control of the d-file in S.Nagle­D.Musanti, Dos Hermanas 2003) 16 . . . c6 17 lt:ixd4 cxd5 18 l::!.fd1 lt:if4 with at least equality for Black. 13 .. . lt:ic6 14 0-0

14 f4 is more critical, but not really justified: 14 . . . ..\txb2 15 l::!.b1 lt:ia5 (15 . . . ..\ta3? is suggested by Rowson, but it fails to 16 l::!.xb7 l::!.ab8 17 l::!.xc7! l::!.b1+ 18 CLlcl ..ltxc1 19 0-0 and White wins) 16 �xb2 lt:ixc4 17 l::!.xb7 was seen in E.Relange-M.Palac, Bastia (rapid) 1998, and now 17 . . . lt:if6 18 lt:ig3 l::!.ab8 gives Black a powerful initiative. 14 .. . CLld4!

Black has occupied the central out­post and clearly stands well. 15 lt:ixd4 ..\txd4

Now any exchange of bishops on d4 would, of course, leave Black with complete control over the dark squares, and the presence of a knight on f4 would be a telling factor in any end-

�!4 Sys t e m s

game. Thus White must continue accu­rately, as he did with 16 i..d5! ..ltxe3! 17 fxe3 c6 18 ..ltb3 l::!.d2! 19 l::!.f2 l::!.ad8 20 l::!.c2 l::!.xc2 21 l::!.xc2 �f8 when the posi­tion was about level in A.Beliavsky­P.Leko, Dortmund 1998.

C2) 7 e3 In my opinion this is a more critical

line. Should White recover the pawn on c4 with ease, he will have prevented both . . . c5 and . . . e5 while maintaining a firm hold on the centre . Black must not allow himself to be forced into a pas­sive position and so should fight to hold on to the pawn. 7 .. . ..lte6 !

1 2 3

Play t h e G r ii nfe /d

s lbgs Taking up the challenge, and now a

fairly forced sequence follows. The al­ternatives are harmless; for example:

a) 8 Jte2 c5 9 0-0 (9 dxc5 lbbd7 is equal) 9 . . . lbd5 10 lbxd5 .i.xd5 11 .ltxc4 cxd4 12 exd4 lbc6 gives Black useful pressure against the IQP.

b) 8 lbe5 c5 ! 9 dxc5 (or 9 .ltxc4 .ltxc4 10 lbxc4 cxd4 1 1 exd4 lbc6 with an edge for Black - I. Gurevich) 9 .. .'ihd1 + 10 lbxd1 (Black is also better after 10 �xd1 lbh5! 1 1 lbxc4 lbxf4 12 exf4 lba6) 10 . . . �c8 11 .i.xc4 was seen in I.Rogers­I.Gurevich, Hastings 1993/94, and now Black could have obtained an obvious edge with l l . . .�xc5! 12 b3 ( 12 .i.xe6?! �xcl 13 lbxf7 �f8 14 e4 �a1 15 e5 lbe4 is even worse for White) 12 . . . .i.xc4 13 �xc4 lba6. S . . . .ltdS 9 e4

White is essentially committed to continuing his forcing strategy. Instead 9 lbxd5? lbxd5 hits the f4-bishop and is good for Black; for example, 10 .ltg3 c5 ! 1 1 itxc4 cxd4 12 �3 (S.Furman­V.Korchnoi, USSR Ch. 1954) 12 .. .'�'a5+! 13 We2 lbb6! leaves Black clearly better, as pointed out by Efim Geller. 9 ... h6! 10 exds hxgs 11 i..xgs lbxds 12 .ltxc4

Instead 12 �f3?! lbb6! 13 �xb7 itxd4 14 lbb5 itxb2 15 �d1 lb8d7 16 lbxc7 �b8 was a disaster for White in A.Schneider­P.Hardiscay, Debrecen 1975. 12 ... lbb6 13 .ltb3 lbc6!

Black has the more compact struc­ture and strong pressure on the dark squares, which fully compensates for White' s bishop pair and central space

1 2 4

advantage, and now White must solve the problem of the d4-pawn.

c21: 14 ds C22: 14 lbe2

c21) 14 ds This abandons d4 and e5, but in­

creases White's spatial advantage and aims to later drive back Black's central­ized knight. 14 ... etJd4 15 0-0 �d7

16 .i.e3 Black's plan is obvious: he will

bring his rooks to d8 and e8, after

which a capture on b3 will leave dS hanging. The text counters that by im­mediately harassing the d4-knight, but White can also aim for counter­pressure against e7:

a) 16 �e1 �fe8 17 h4! ? (Xu Jun's idea; both 1 7 .i.e3 lt:lxb3 18 'iVxb3 .ixc3 19 bxc3 'iVxdS, as in S.Halkias-J.Borisek, Terme Zrece 2003, and 1 7 a4 lt:lxb3 18 'iVxb3 aS 19 �cd1 .ixc3 20 'iVxc3 lt:lxdS see White lose his d-pawn without gaining any real compensation) 17 . . . aS 18 a4 lt:lxb3 19 'iVxb3 .ltxc3 20 �xc3 lt:lxdS 21 �h3 �a6 and White has some compensation for the pawn, but there is no immediate threat and Black's re­sources seem entirely adequate to me.

b) 16 h4 (G.Giorgadze-A.Shirov, Cala Galdana 2001) 16 . . . �fe8 1 7 �e1 aS is fine for Black, who intends to further advance the a-pawn.

c) 16 a3 lt:lxb3 17 'iVxb3 was seen in I .Rogers-P.Wolff, San Francisco 1991, and now Rogers points out that Black could have equalized with 17 . . . .\txc3 18 �xc3 (18 'iVxc3 lt:lxdS 19 'iVd4 f6 is good for Black) 18 . . . 'iVxdS ! . 16 .. . lt:lts

The downside to White's last move is that he must now allow the exchange of his dark-squared bishop, although doing so does at least enables him to keep dS fully protected. 17 �e1 lt:lxe3 18 �xe3 as 19 a3 a4

This position was seen in the com­puter game Isichess-Deep Sjeng, Reykja­vik 200S . After the forced 20 ita2, Black should have played 20 . . . Jth6 21 �h3 �g7!, followed by . . . �h8 to extinguish White's kingside play; Black's better

�/4 Sys t e m s

bishop and superior structure should have their say in the future.

C22) 14 lt:le2 Trying both to retain some influ­

ence on the dark squares and restrict the black knights.

14 . . . as ! Black wastes no time in dynami­

cally gaining his knights some squares. 15 a4

Practice has shown that White must block the further advance of the a-pawn in this way, but now he also has weak­nesses on the queenside. Instead 1S 0-0? a4 16 .ltc4 (R.Hartoch-J.Timman, Leeu-

1 2 5

Play th e G r u nfe /d

warden 1971 ) 16 . . .'�Jxd4 17 .!te3 'Llxc4 18 l:l.xc4 'Lle6 leaves Black with an extra pawn, while 15 a3 a4 16 �a2 l:l.a5 is a typical plan in this line and one which worked wonders in V.Huuskonen­J.Klovans, correspondence 1971 : 17 'iVd2 l:l.b5 18 l:l.c3 'iVd6 19 h4? (19 l:l.d3 'Lla5 was better, although still good for Black) 19 . . . 'Llxd4! 20 l:l.d3 l:l.d8 and Black had again netted a pawn. 15 .. . l:l.c8!

By defending c7 Black renews the threat against the d4-pawn. 16 .!te3

Consistent, but this fully concedes the initiative to Black. However, 16 0-0 is rather an admission of failure by White, who now hopes to equalize the position with multiple exchanges. In­deed, after 16 . . . 'Llxd4 17 'Llxd4 'iVxd4 18 �xe7 (G.Giorgadze-B .Avrukh, Bugojno 1999) 18 . . . l:l.fe8! 19 �c5 (and not 19 �a3 c5 20 'iVf3 c4 21 �c2 'iVdS when Black is better) 19 . . . 'iVxd1 20 l:l.fxd1 �xb2 21 .l:':l.b1 �f6 22 �xb6 cxb6 23 �d5 Black's extra pawn will vanish, leading to a rather level endgame. 16 . . . es !

1 2 6

1 1 dxes Again White cannot hope for any­

thing more than equality, so exchanges are the way to go. Instead 17 d5 'Lld4 gives Black a strong centralized knight, and after the 18 'Llc3 (18 'Llxd4? exd4 19 �xd4 �xd4 20 'iVxd4 l:l.e8+ 21 'it>fl c5 22 'iVg4 'iVf6 leaves White in a quandary; Black is attacking the b2-pawn and . . . c4 is also a threat) 18 . . . 'iVh4 19 �a2 c6 20 dxc6 l:l.xc6 21 0-0 of A.Dreev-Ni Hua, Shanghai 2001, the simple 21 . . .'Llc4 22 �xc4 l:l.xc4 ensures Black of an edge. 17 . . . 'iVxd1+ 18 .l:':l.xd1 'Llxes

White must now give up his bishop to prevent . . . 'Llec4, and after 19 �xb6 cxb6 20 0-0 l:l.fd8 the position is equal, G .Giorgadze-S.Atalik, European Team Ch., Batumi 1999.

Concl usion

We've covered several highly dynamic lines in this chapter. Black must be pre­pared to sacrifice the c7-pawn, although its capture always gives him good com­pensation. However, should White play less incisively, Black can counter quickly with . . . c5, activating his pieces and often gaining a useful initiative. Line C is perhaps the most critical, but by switch­ing his approach to capturing on c4, followed by playing actively on the kingside and in the centre, Black still gains a good game. Overall in these �f4 lines, a certain accuracy is required from Black in the opening phase, but if he knows his theory, he can count on a fairly promising position.

C h a pte r Te n I jLgs Syste m s

1 d4 lt:if6 2 C4 g6 3 lLic3 d5 White can now develop his bishop

to g5 immediately or first wait a move:

A: 4 lt:if3 .ig7 5 .ig5 B: 4 .ig5

A) 4 lt:if3 .ltg7 5 itg5

This system does not threaten the vi­ability of the Griinfeld, but it is a sound and solid way of meeting our ambitious opening. White aims for simple yet

somewhat annoying development, as he increases the pressure against d5 by threatening to exchange on f6. In re­sponse, Black is best advised to counter with . . . lt:ie4xc3. White then hopes that his central bastion on d4, further rein­forced by e3, will neutralize the Gri.in­feld g7-bishop. However, there are some downsides to White's method of development: his pieces find themselves not especially well coordinated and his dark-squared bishop can become loose. 5 . . . lt:ie4

1 2 7

Play t h e G r ii nfe l d

Clearly the most principled move, and Black will now gain counterplay, no matter what White does about his dark-squared bishop.

A1: 6 h4 A2 : 6 1i'c1 A3: 6 ..if4 A4: 6 cxds AS: 6 ..ih4

A1) 6 h4?! This is rather overambitious, since

White will sorely miss his dark­squared bishop once Black attacks in the centre . 6 .. .Ct:lxgs 7 hxgs dxc4!

8 e3 Alternatively, 8 1i'a4+ c6 9 1i'xc4 ..ie6

10 1i'd3 lt:la6 11 g3 lt:lb4 12 1i'd2 c5 13 lt:le4 ..ixd4! ? (seemingly a risky pawn­grab, but tactically justified) 14 a3 lt:lc6 15 e3 .idS! 16 l:th4 ..ig7 17 0-0-0 'ifb6 18 1i'e2 0-0-0 19 'ifh5 1i'xb5 20 ..ixb5 h6 21 gxh6 l:txh6 22 l:txh6 ..ixh6 23 lt:lfd2 lt:ld4 24 ..ic4 lt:le6 25 ..ixd5 l:txd5 and Black was winning in J.Granda Zuniga-

1 2 8

I .Sokolov, Wijk aan Zee 1989. 8 . . . cs 9 ds

Instead both 9 ..ixc4 cxd4 10 lt:lxd4 0-0 and 9 dxc5 1i'a5 10 1i'a4+ 1i'xa4 1 1 lt:lxa4 lt:la6 are better for Black. 9 . . . bs !? 10 lt:lxbs 1i'b6 11 ..ixc4 a6 12 lt:la3 1i'xb2

Black obviously has some advan­tage.

A2) 6 1i'c1 Another rather unnatural move.

6 . . . h6! Simple and good. Play will now re­

semble Line A3, but with the white queen misplaced on cl . On the other hand, Black has some difficulty castling in view of the hanging pawn on h6, although this turns out to be of less importance. 7 ..if4 lt:lxc3 8 bxc3 cs 9 cxds

Instead 9 ..ie5 ..ixe5 10 lt:lxe5 cxd4 1 1 cxd4 (but 1 1 1i'f4 f6 12 lt:lxg6 l:tg8 13 1i'xh6 'if'a5 14 l:tcl dxc3 15 cxd5 c2+ 16 1i'd2 1i'xd2+ 1 7 'lt>xd2 l:txg6 i s good for Black) 1 l . . .lt:lc6 leads to equality, ac­cording to Adorjan. 9 . . . 1i'xds 10 e3 lt:lc6 11 h3

White should avoid 1 1 ..ie2? ! g5 ! 12 ..ig3 g4 13 C£:lh4 ..if6 (Adorjan), espe­cially since he cannot then play 14 0-0? due to 14 . . . �5. u . . . il.fs 12 ..ie2 l:l.c8

�g5 Sys t e m s

..ixc4 tZ:lc6 (or 10 . . .'iYa5 ! ? 1 1 0-0 tZ:ld7, L.Gofshtein-Z.Kozul, Zagreb Zonal 1993) 1 1 0-0 tZ:laS 12 ..ie2 b6 ! ? with good play, especially in the case of 13 dxc5? ! ..id7.

b) 9 l:l.b1 hopes to hinder Black's queenside development, but is hardly problematic: 9 . . . cxd4 10 cxd4 tZ:lc6 1 1 �a4 !? (or 1 1 ..ie2 dxc4! 12 ..ixc4 tZ:laS 1 3 ..id3 ..ie6 14 �e2 a6! and Black will follow with the standard plan of . . . b5 and then occupy c4) 1 l . . . ..id7! 12 �a3 (if 12 l:l.xb7 e5 ! ) 12 . . . ..ig4 ! (E.Bareev­A.Beliavsky, Linares 1992) and White's centre is under pressure, while it is not clear whether he will manage to castle anytime soon.

13 �a3 9 . . . cxd4 10 cxd4 �xds 11 ..ie2 tZ:lc6 12 Inferior is 13 �d2? ! g5 14 i.g3 cxd4 0-0

15 cxd4 0-0 16 .l:':l.cl e5 (I .Zaitsev­V.Tukmakov, Yerevan 1982) when Black is better, while 13 0-0 g5 14 ..ig3 0-0 (Adorjan) gives Black the initiative. 13 .. . 0-0

Black has fully equalized .

A3) 6 ..if4 This retreat aims to prevent the

. . . e5-break. It is more solid than the retreat to h4 (Line AS), but also less annoying for Black since he can now quickly counter in the centre. 6 . . . C£:lxc3 7 bxc3 cs 8 e3 o-o 9 cxdS

Alternatively: a) 9 ..ie2 dxc4 (9 . . . tZ:lc6 10 0-0 cxd4 1 1

cxd4 dxc4 12 ..ixc4 occurred in L.Portisch-V.Kramnik, Biel Interzonal 1993; a model game for handling such positions, but it seems to me that White can prove a slight edge in this line) 1 0

Both sides have developed sensibly so far and now Black's play focuses around the . . . e5-break. Should White prevent that, Black can always switch to the alternative plan of occupying c4. 12 . . . ..ifs !

In my mind, clearly the best move; Black both prevents l:l.b1 and increases his central control .

1 2 9

Play th e G r u nfe ld

13 'iYa4 With this move, White intends to

kick the black queen away from the centre with .Ucl-c5 and then either pressurize Black's queenside or ad­vance in the centre. This is quite a dan­gerous plan, so Black has to respond accurately. The alternatives are less challenging:

a) 13 'iYb3 if..e6! (the bishop will be excellent on d5) 14 'i!Yxd5 (14 'i!Yxb7? ! tt::lxd4 15 'i!Yxd5 tt::lxe2+ 16 �h1 if..xd5 is good for Black) 14 .. . if..xd5 is equal, al­though Black can prepare .. . e5 and the long-term chances are with him.

b) 13 .Ucl ?! is met by 13 . . . 'i!Yxa2 14 d5 .Uad8! 15 if..c4 'i!Ya3 ! and Black is on top; . . . tt::la5 follows, leaving White without any real compensation for the sacri­ficed pawn.

c) 13 tt::lh4! ? is a favourite of Zvja­ginsev. White aims to strike back in the centre, but this plan has two significant drawbacks: d4 is weakened and the white minor pieces are rather loose.

Play might continue 13 . . . if..e6 (in-stead 13 . . . if..c8 ! ? threatens . . . tt::lxd4 and practically forces White to retract his

1 3 0

last move and repeat the position; for example, 14 'i!Ya4 tt::lxd4 ! ? 15 exd4 'i!Ye4 16 tt::lxg6 hxg6 17 if..e3 if..g4! gave Black the upper hand in V.Zvjaginsev­P.Leko, Tilburg 1998) 14 if..f3 (this was White's idea, but now his knight is rather offside; Black is also fine after Rowson's suggestion of 14 .Ub1 'i!Yd7! ? when 15 'i!Ya4? fails to 15 . . . tt::lxd4) 14 . . . 'i!Ya5 15 d5 (or 15 .Ub1 if..c4! ? 16 .Ue1 if..a6 - Rowson) 15 . . . .Uad8 16 e4 f5 ! when Black breaks up White's centre and has excellent play. 13 ... 'i!Yas !

A n endgame i s more than acceptable here to Black. Not only does he have a queenside pawn majority, he is also quite likely to achieve . . . e5, liberating the g7-bishop. Furthermore, that piece will often spring to life should White advance his own central majority. 14 'i!Yxas

Less accurate is 14 'iYb3? ! 'iYb4! (in­tending to plant the knight on b4 after an exchange of queens) 15 .Ufcl .Uac8! (and not 15 . . . .Ufc8? 16 .Uxc6! ) 16 h3 a6 ! 1 7 'i!Yd1 .Ufd8! which is good for Black since . . . e5 is on the cards.

In P.Van der Sterren-A.Huzman, Wijk aan Zee 1993, White radically tried to prevent that break, but . . . eS actually followed in any case: 18 g4? (instead Rowson offers 18 j_d3 j_xd3 19 �xd3 eS 20 j_gs .l:i.d7! 21 .l:!.ab1 �aS 22 �3 exd4 23 exd4 j_xd4 24 lLlxd4 �xgS with a clear advantage, and 18 �fl �aS!? also favours Black) 1 8 . . . j_e4 19 lt:Jd2 eS! 20 j_gs (White was proba­bly counting on this, but having won the central battle, Black can afford to sacrifice some material to pursue his initiative) 20 . . . exd4 ! ! 21 .l:!.c4 (White also loses after both 21 lt:Jxe4 dxe3 22 j_d3 j_xa1 23 .l:!.xa1 lt:JeS and 21 j_xd8 .l:!.xd8! 22 .l:!.c4 �2 23 �cl dxe3 24 �xb2 j_xb2 2S .l:i.d1 lt:Jd4) 21 . . .�2 22 .l:i.acl dxe3 23 j_xe3 j_ds 24 .l:!.4c2 �eS 2S �fl �f6 26 �c4 lt:Jb4 27 .l:!.c3 lt:Jxa2 0-1 . 1 4 .. . lt:Jxas 15 j_c7

Alternatively, 1S .l:i.fcl .l:!.ac8 16 .l:!.c7 (or 16 lLld2 lt:Jc6! 17 j_f3 eS! 18 dxeS ctJxeS 19 j_xb7 .l:i.cd8 ! when White was unable to parry the threat of . . . lt:Jd3 and was in huge trouble in G.Hertneck­V.Anand, Munich 1996) 16 . . . .l:!.xc7 17 iLxc7 lbc6 18 .l:i.cl lLlb4!? and Black is still creating small problems, even in this simplified position. Indeed, Black's superior piece activity means that, while this and similar positions may be objectively equal, he is the side with the better practical chances. 1S . . . lt:Jc6

With equal chances, but also plenty of play left.

A4) 6 cxds This line is rather different to the

.il...g5 Sys t e m s

others in this chapter a s White does not waste any time safeguarding his bishop. While this may seem a serious concession on his part, one must not forget about White's lead in develop­ment. 6 ... lt:Jxgs 7 lt:Jxgs e6

The pawn sacrifice 7 . . . c6 does not convince me at all . Instead the logical text move should be played, and now White can continue tactically or posi­tionally:

A41: 8 �a4+ A42 : 8 lLlh3 A43 : 8 ctJf3 A44: 8 �d2

A41) 8 �a4+ This tricky sideline, full of tactics,

actually results in a tedious, equal end­game almost by force! 8 . . . c6!

I would love to suggest the roman­tic 8 . . . j_d7?, but it seems badly insuffi­cient to me: 9 �3 �xgS 10 �xb7 0-0 1 1 �xa8 j_xd4 12 .l:!.d 1 ! lt:Jc6 1 3 �7 �eS

1 3 1

Play th e G r ii nfe ld

14 .U.xd4 'Llxd4 15 e3 'Llf5 16 i.d3 with a clear advantage for White . 9 dxc6 tt:Jxc6 10 'Llf3 i.d7 !

Threatening . . . 'Llxd4. 11 1i'd1!

A good test of Black's resources. In­stead 1 1 0-0-0?! is too risky due to 1 l . . .b5 ! 12 'Llxb5 0-0 13 1i'a3 1i'h8! 14 e3 .U.c8 15 'Llc3 aS with excellent counter­play for Black, M.Cebalo-B .Lalic, Za­greb 1993. 11 . . . 1i'b6 12 1i'd2 'Llxd4

The start of the aforementioned equalizing sequence. Should Black wish to keep the game more alive, the following idea of Harts ton's may be worth a punt: 12 . . . 0-0 ! ? 13 e3 e5 14 d5 'Lld4!? (inferior is 14 . . . 'Llb4 15 a3 'Lla6 16 e4) 15 i.e2 (or 15 exd4 exd4 16 'Lle2 .U.ae8 17 .U.d1 .U.e4 with dangerous play for the piece) 15 . . . 'Llxe2 16 1i'xe2 e4 1 7 'Lld4 f5 and Black has good practical compensation for the pawn. 13 0-0-0 .U.d8 ! 14 'Llxd4 i.c6 15 e3 es 16 1i'e1!

White's idea, but Black is scarcely troubled by the check. 16 ... exd4 17 exd4+ 'lt>f8 18 ds i.xds !

1 3 2

I'm not convinced by 1 8 . . . i.d7?! as an attempt to play on: 19 1i'e3 1i'xe3+ 20 fxe3 .U.c8 and it's not clear how Black can play for a win. 19 .U.xds .U.xds 20 1i'e7+!

The only move. 20 'Llxd5? 1i'xb2+ 21 'lt>d1 1i'h1 + 22 'lt>e2 'iVbS+ wins for Black. 20 ... 'lt>xe7 21 'LlxdS+ 'lt>d6 22 'Llxb6 axb6

This was all played (and agreed drawn at this point) in A.Shirov­B.Ostenstad, Gausdal 1991 . Black even has a very slight initiative, but it would take quite some effort to convert it.

A42) 8 'Llh3 This was once popular as a means

of pressurizing the pawn on d5 and thereby preventing any . . . c5 ideas . However, this knight manoeuvre doesn't actually prevent . . . c5 at all! 8 .. . exds 9 'Llf4 o-o 10 e3

10 g3?! is the alternative plan but leaves d4 at Black's mercy; for exam­ple, 10 . . . .U.e8 1 1 i.g2 'Llc6 12 0-0 'Llxd4 13 'Llfxd5 c6 14 'Llf4 i.g4! 15 f3 i.d7 16 e4 1i'b6 was very promising for Black in V.Kovacevic-V.Jansa, Amsterdam 1973.

1o . . . cs ! Accepting the challenge - and why

not?! 11 dxcs d4! 12 exd4 ii.xd4 13 ii.e2 tt::lc6 14 o-o 'iVgs

Black's activity definitely compen­sates for the pawn and may even count for more. 15 g3

If 15 tt::ld3 ..tf5 and c5 will drop even­tually, while 15 tt::lfd5 ..te6 is equal . 1s .. . ii.fs 16 .Uc1 .Uad8 17 'iVb3 ii.es

Now 18 tt::lfd5 .Ud7 19 f4 ii.d4+ 20 Wh1 'iVd8 21 ..tf3 ..txc5 is good for Black (Gutman), and after 18 h4 '*h6 19 tt::lg2 (White should avoid both 19 tt::lcd5?

J..g5 Sys t e m s

ii.e4 and 19 tt::lfd5 .Ud7) 19 . . . .Ud7! ? (im­proving over the 19 . . . .Ud2?! 20 'iVxb7 of J.Hebert-L.Gutman, Hastings 1984/85), Black prepares . . . tt::ld4 and it is clear that White has been forced on to the defensive .

A43) 8 tt::lf3

With this retreat, White opts for quiet play along the lines of the Queen's Gambit Declined. In fact, the resulting positions are very reminiscent of the Tartakower Variation of that opening. However, I believe that White's position is rather bad in practi­cal terms; his only real plan is a minor­ity attack on the queenside, but Black can stop that without undue effort. Black, on the other hand, can easily complete his development and will then look to attack on the kingside; an aggressive plan which has often scored well in practice . s . . . exds 9 e3

An important alternative is 9 b4 which is actually White's last chance to play this advance unhindered, but he doesn't prove ready to back it up after

1 3 3

Play th e G r ii nfe ld

9 . . . iYd6 ! :

a ) 10 a3 0-0 1 1 e3 c6 12 ..lte2 ..ltf5 13 0-0 lt:Jd7 (the benefit of Black's 9th move is smooth development, as well as White's inability to set his queenside pawns in motion) 14 lt:Ja4 (watch how Black now puts a stop to White's queenside play) 14 . . . a5 15 iYb3 b5 16 lt:Jc5 a4 17 iYc3 lLlb6! was Y.Seirawan­G.Kasparov, Dubai Olympiad 1986. Black will follow up with . . . lt:Jc4 and it is obvious to the naked eye that his position is the more pleasant.

b) 10 iYb3 (a more risky choice and one suggested in Palliser's Play 1 d4!)

10 . . . lt:Jc6! and now: b1) l l lLlb5 iYe7 12 .l::i.cl ..ltg4 13 iYe3

iYxe3 14 fxe3 0-0-0 and Black is much better.

b2) 1 1 .l::i.cl a6! 12 a3 (12 e3 lt:Jxb4 13 a3 lLlc6 14 lt:Jxd5 - and not 14 iYxd5?? iYxa3 - 14 . . . 0-0 15 lt:Jc3 is safer and ap­proximately equal) 12 . . . lt:Jxd4 13 lt:Jxd4 (Black wins after 13 iYxd5? lt:Jc2+! ! 14 �d1 lt:Jxa3) 13 . . . .ixd4 14 e3 ..ltg7 15 lt:Jxd5 c6 1 6 lLlb6 ..lte6 with an edge for Black.

c) 11 e3 iYxb4 12 iYxb4 lLlxb4 13 �d2

1 3 4

reaches a critical position in which I quite like the natural 13 . . . ..1tf5 ! ; for ex­ample, 14 a3 lt:Jc6 15 lt:Jxd5 ( 15 .ib5 a6 16 ..ltxc6+ bxc6 17 lt:Ja4 �e7 18 .l::i.hcl .l::i.hb8 ! is also good for Black) 15 . . . 0-0-0 and Black's lead in development gives him the advantage; . . . lt:Ja5, with tempo, and . . . c5 may follow. 9 . . . as

White will now find i t hard even to push b4. 10 .ie2 0-0 11 0-0 .l::i.e8 12 a3 ..ltf8!

Black insists on preventing b4 and transfers his dark-squared bishop to its best square, namely d6. The following play is rather unforcing so I've elected to highlight a few standard plans and ideas:

13 ctJe1 Instead 13 lt:Je5 c6 14 ..ltg4? ! (a typi­

cal mistake; White wants to reduce Black's attacking possibilities, but this both costs too much time and weakens White's light square control) 14 . . . ..1td6 15 ..ltxc8 iYxc8 16 lLld3 lLld7 17 iYf3 iYd8 (Black has fully equalized) 18 b4? ! (and here comes the other common error: White weakens c4) 18 . . . lt:Jb6 19 lt:Jc5

�c7 20 h3 4Jc4 gives Black a pleasant edge, R.Vaganian-P.Wolff, New York Open 1990. 13 . . . c6 14 4Jd3 i.d6 15 b4

This weakens the c4-square, and maybe now is the time for 15 4Ja4 i.f5 16 ..ltg4!? . 15 . . . �e7 16 �b3 bS ! 17 .Ufc1

17 .Ufe1 i.f5 18 4Jc5 4Jd7 19 i.fl 4Jb6 (or even 19 . . . 4Jxc5! ? 20 bxc5 i.c7 with a clear plus) 20 bxa5? ! 4Jc4 21 a4 b4! saw Black instructively take over the queenside initiative with some ef­fect in A.Kakageldiev-I.Gurevich, Biel Interzonal 1993. 17 .. . ..\tf5 18 bxa s .Uxas

Both sides have some coordination difficulties, but Black enjoys a clear target on a3, the bishop pair and some chances on the kingside.

A44) 8 �d2 Clearly the most critical try.

s . . . exds 9 �e3+ �f8 10 �f4 The only dangerous move; for ex­

ample, 10 g3 i.f6 1 1 h4 �g7 gives Black no problems. 10 .. . ..\tf6

jLg5 Sys t e m s

Unfortunately 10 . . . �f6?! runs into some tactical problems: 1 1 �xc7 4Ja6 12 �g3 4Jp4 13 .Ucl ! i.f5 14 e3 with a solid edg{ 11 h4 h6!

The knight should be kicked back. 12 4Jf3

The tactical shot 12 4Jxd5?! has claimed some victims, but I really doubt its soundness. Let's see how to meet it: 12 . . . ..\txg5 13 �e5 .Uh7 14 hxg5 4Jc6 15 �e4 ii.f5 16 �f3 4Jxd4 17 �a3+ �g7 18 4Je3

18 . . . �xg5! (the correct capture; infe­rior is 18 . . . hxg5 19 .Uxh7+ �xh7 20 0-0-0 �f6 21 �c3 c5 22 4Jxf5 �xf5 23 e3 4Jc6 when according to Rowson Black is clearly better, as in S.Skembris­J .Smejkal, Thessaloniki Olympiad 1988, but after 24 i.c4 �g8 25 .Ud6 White has a lot of pressure) 19 �c3 .Ud8! (a clear improvement over the oft-suggested 19 . . . �f6 20 0-0-0 c5 21 4Jxf5+ 4Jxf5 22 ·e3 when White has good compensation) 20 .Ud1 (Black stands very well after 20 4Jxf5+ �xf5 21 e3 - or 21 .Ud1 c5 22 e3 �c2 - 2l . . .�c2 22 .Ud1 �xc3+ 23 bxc3 4Je6 24 .Uxd8 4Jxd8) 20 . . . c5 21 �xc5

1 3 5

Play th e G r u nfe ld

if..e6 22 'iYc3 'iYf6 23 g3 .l:i.hh8 and Black has the initiative - one look at White's king will suffice to confirm this! 12 .. . '1t>g7

Black must aim to catch up in de­velopment as much as possible before White breaks with e4 or advances with g4-g5. If White continues slowly, Black regroups with .. .liJd7, . . . c6 and . . . �8. 13 0-0-0 if..e6!

Black is now all set for . . . c5, as well as to counter the e4-break.

14 e4 Alternatively, 14 g4? ! c5 ! 15 e3 lbc6

16 .id3 cxd4 17 exd4 �8! is better for Black, but White has also tried 14 e3 and now:

a) The intended 14 . . . c5 leads to a draw: 1 5 dxc5 'iYa5 16 l2Jd4 'iYxc5 17 .id3 lLlc6 18 lLlxe6+ fxe6 19 'iYg4 if..xc3 with perpetual check, as White cannot play 20 'iYxg6+? due to 20 . . . '1t>f8 21 bxc3 'iYxc3+ 22 'lt>b1 �4+ 23 'lt>c2 lbd4+! 24 exd4 .l:!.c8+ and Black wins.

b) 14 . . . lLld7 15 g4 .ie7! 16 'iYg3 (16 e4 is suggested by Chandler but just looks bad to me after 16 . . . dxe4 17 'iYxe4 lb£6 18 'iYxb7 .l:i.b8 or 17 lLld2 if..d6 18 'iYxe4

1 3 6

lb£6 19 'iYxb7 .l:i.b8 with some advantage in both cases) 16 . . . .id6 17 'iYg2 c6 and Black is better; his queenside play, with . . . a5-a4-a3 the immediate intention, is much more potent than anything White can drum up on the kingside. 14 . . . dxe4

15 'iYxe4!? Instead 15 lLlxe4? ! .ixa2! (Black can

and should grab this important pawn) 16 g4 (16 d5 l2Jd7 17 lLlxf6 lLlxf6 is great for Black) 16 . . . l2Jd7 17 J.d3 c5! ! (a pow­erful move which virtually refutes White's concept) 18 if..b5 (Black is much better after both 18 dxc5 .l:!.c8 and 18 g5 if..xd4) 18 . . . .id5! 19 dxc5 (19 .ixd7 .ixe4 20 'iYxe4 'iYxd7 21 g5 .ixd4 22 lLlxd4 cxd4 gives Black a clear plus) 19 . . . .ixe4 20 .l:i.xd7 'iYa5 is excellent for Black, G.Schroll-V.Epishin, Vienna 1991 . 1S . . . c6

Black mustn't be too ambitious here: 15 . . . c5? ! 16 .ic4! .ixc4 17 dxc5 is good for White. 16 lLles lLld7

I believe this is White's best option in the whole 8 cxd5 variation, but it' s still nothing great. There are some

prospects of a kingside attack, particu­larly on the light squares weakened by . . . h6, but that is fully counterbalanced by Black's strong central influence.

As) 6 if..h4 White aims to exploit the fact that

the e7-pawn must be protected in order to force Black to surrender the centre. The pressure against e7 does prove annoying at times, but on the other hand, the bishop's inability to return to the queenside leaves that sector of the board very exposed. 6 .. . 'Llxc3 7 bxc3 dxc4!

In my view, definitely the best way of gaining counterplay.

Since 8 .l::i.b1 ?! a6, intending . . . b5 and . . . c5, is good for Black, White' s choice is really between:

A51: 8 e3 A52: 8 e4 A53 : 8 �a4+

A51) 8 e3

il...g5 Sys t e m s

This pretty much turns White's opening play into a pawn sacrifice . 8 . . . bs !

Holding on to the pawn in the most straightforward manner. 9 a4

9 if..e2 will generally transpose as White cannot really make do without a4. 9 . . . c6

As Rowson points out, this defence is playable only here, and not in the 4 if..g5 system (in the analogous position axb5 . . . cxb5, �f3 would just win mate­rial) or in Line A3 (when the attack on the b8-knight would result in the col­lapse of Black's queenside). 10 if..e2

Alternatively: a) 10 'Lld2 a6 1 1 if..e2 is another

route into our main line. b) 10 axb5 cxb5 1 1 'Lle5 if..b7 (but not

1 l . . .if..xe5? 12 �f3 'Llc6 13 �xc6+ if..d7 14 �e4 and White has too much com­pensation) 12 .l::i.b1 (or 12 �1 a6 13 'Llxc4 �c7 14 'Lld2 �xc3 15 �cl �xcl+ 16 .l::i.xcl 'Lld7 which favours Black) 12 . . . �d5 !? (12 . . . if..xe5? is a mistake in

1 3 7

Play th e G r ii nfe ld

view of 13 .l:i.xb5!, but Black can also consider 12 . . . a6 13 lt:Jxc4 .ie4 14 lt:Jd2 i..xb1 15 'i:Yxb1 0-0 with advantage) 13 f3 f5 and Black is clearly on top. 10 . . . a6 11 lt:Jd2 o-o 12 i..f3 .l:i.a7 13 o-o

Both sides have continued logically since the pawn sacrifice, but already White's compensation doesn't look too convincing.

13 . . . h6!? A waiting move which slightly im­

proves Black's position while waiting to see what White is up to. Instead Kas­parov, when faced with this line, opted for 13 . . . i..f5, intending either to place his bishop on d3 or provoke e4, and that is a good alternative: 14 .l:i.e1 (14 e4 ..ltc8! 15 e5 i..e6! is good for Black) 14 ... h6! ? (simi­lar to my main suggestion, whereas 14 . . . ..\td3 15 CLJb3 i..f5 16 CLJd2 i..d3 was A.Sorin-G.Kasparov, Buenos Aires simul 1997; now White should have continued to repeat, while Black should have preferred 16 . . . h6 - he is better, af­ter all ! ) 15 e4 (15 ..ltg3 e5! 16 ..ltxe5 i..xe5 17 dxe5 .l:i.d7 18 .l:i.a2 'i:Yc7 is very good for Black) 15 . . . ..\te6 and Black threatens . . . b4; White really feels the absence of

1 3 8

his dark-squared bishop from the queenside. 14 GtJe4

Similar is 14 'i:Ycl f5 ! 15 .l:i.d1 (or 15 'i:Ya3 g5 16 ..\tg3 e5 17 'i:Yc5 'i:Ye7 and Black is better) 15 . . . 'i:Yc7, intending . . . e5 with advantage. 14 . . . fs 15 ct:Jcs 'i:Yd6

With . . . e5 coming, Black is clearly for preference.

A52) 8 e4 This, in comparison with 8 e3,

leaves d4 less well covered and thus more vulnerable to . . . c5 or . . . e5 breaks. s . . . bs

9 a4 c6 10 if..e2 Another option is 10 'iVb1 ! ? a6 1 1

if..e2 ti:Jd7 1 2 0-0 ti:Jb6 13 'iVb4 (13 aS ti:Jd7 14 ti.d1 0-0, with the idea of . . . 'i:Ye8 and . . . eS, is promising for Black) 13 . . . tt::lxa4 14 if..xc4 cS 1S 'iVb3 0-0 16 if..dS ti.a7 17 ti.fd 1 'i:YaS 18 dxcS e6 19 if..c4 'i:Yxc3 20 ti.xa4 which was rather unclear in N.Ristic-V.Tukmakov, Cet­inje 199 1 . 10 . . . ti::ld7

Black can also play 10 . . . 0-0 1 1 0-0 ti::ld7! 12 'i:Yc2 ti::lb6 13 aS ti::ld7, when he is again somewhat better. 11 o-o ti:Jb6 12 as

Now Black should avoid getting in­volved in 12 . . . ti::la4? ! 13 ti.xa4! bxa4 14 'i:Yxa4 when White's strong centre gives him good compensation for the ex­change. Instead the simple 12 . . . ti:Jd7 13 'i:Yd2 0-0 leaves White with insufficient compensation for the pawn.

A53) 8 'i:Ya4+

White opts for the immediate re­covery of the pawn. This line forms the core of White's anti-Griinfeld reper­toire in Grivas' recent and very de-

lig5 Sys t e m s

tailed book. Grivas himself has played this line several times with success and is its leading theoretician. Nonetheless, having spent quite some time analys­ing this line, I have come to the conclu­sion that it is perfectly satisfactory for Black, as the following analysis will prove. 8 . . . 'i:Yd7 !

The correct interpolation, otherwise White regains his pawn with the more harmonious position. 9 'i:Yxc4 b6!

Black will now both gain time with . . . if..a6 and eventually force White to lose his castling rights, due to the very likely possibility of an exchange on fl . Note, too, that the key element in Black's strategy is the . . . cS break.

10 e3 Alternatively: a) 10 'iVb3 if..a6 11 e3 transposes to

our main line. b) 10 ti::leS if..xeS ! 1 1 dxeS if..a6 12

'i:Ye4 'i:Yc6 13 'i:Yd4 (or 13 'i:Yxc6+ tt::lxc6 14 f4 0-0-0 and White's development diffi­culties grant Black lasting compensa­tion) 13 . . . 0-0 14 if..xe7 ti.e8 1S if..h4 (or 1S

1 3 9

Play t h e G r u nfe ld

ii.f6? tt::ld7 16 l:td1 tt::lxf6 17 exf6, T.Sielicki-P.Bobras, Internet blitz 2003, and now 17 . . . l:te4! with a large advan­tage) 1S . . . tt::ld7 16 f4 iDeS with good compensation; Black has strong pres­sure and White faces ongoing prob­lems both developing and coordinating his pieces.

c) 10 ii.g3 is too tame and leads to trouble: 10 . . . cS !

1 1 e3 (or 1 1 iDeS ii.xeS! 12 .i.xeS ii.a6 13 "iVb3 0-0 14 .i.xb8 - White hardly benefits from either 14 e3 .i.x£1 1S �xfl tt::lc6 or 14 l:td1 tt::lc6 1S ii.g3 l:tfd8 -14 . . . l:taxb8 1S e3 ii.x£1 16 �xfl, A.Alif­erenko-V.Nedeiko, Alushta 200S, and now 16 . . . eS! leaves Black a little better) 1 l . . ..i.a6 12 "iVb3 .i.xfl 13 �xfl (Black was better after 13 l::tx£1 tt::lc6 14 �e2 tt::laS 1S "iVb2 0-0 16 l:tfd1 l:tac8 17 l:tacl 'iVa4 18 �fl l:tfd8 19 'ife2 tt::lc4 20 �g1 bS in D.Sahovic-J.Dorfman, Lvov 1984) 13 . . . 0-0 14 �e2 tt::lc6 1S l:thd1 (1S a4 l:tac8 16 l:thd1 tt::laS 17 "iVbS 'ife6 18 �fl 'iVc4+ 19 'ifxc4 tt::lxc4 20 CLJd2 also gave White nothing in L.Bass-A.Mikhalchishin, Bu­dapest 1989) 1S . . . tt::laS is equal. 10 ... .i.a6 11 'i¥b3 ii.xf1 12 �xf1 0-0

1 4 0

Note that 1 2 . . . cS? ! i s inaccurate in view of 13 dS! 0-0 14 l:td1 ! as Grivas' games have shown; for example, 14 . . . eS 1S dxe6 'iVxe6 16 'iixe6 fxe6 17 c4 gave White an edge in E.Grivas-W.Schmidt, Novi Sad Olympiad 1990. 13 �e2

White must now resort to artificial castling, which is quite time-consuming, but Black must act energetically to ex­ploit it. 13 ... cs !

14 dxcs This is Grivas' main line, but Black

has some improvements, as we will see. Note that White cannot play 14 dS? due to 14 . . . c4 1S 'ifxc4 l:tc8 when he faces serious problems. Instead, objec­tively best and definitely safest is 14 l:thd1, even though after 14 . . . cxd4 1S cxd4 tt::lc6 16 a4 (or 16 l:tacl tt::laS 17 "iVb4 e6 18 a4 l:tfc8 and Black was fine in F .Gheorghiu-K.Sundararajan, Biel 1999; he can follow up with . . . "iVb7-e4) 16 . . . l:tac8! ? (or 16 . . . eS ! ? 17 dxeS 'ife8 with equal play; instead 17 dS? e4 1 8 dxc6 exf3+ 19 gxf3 'iVh3 i s bad for White, while 17 tt::lxeS? tt::lxeS 18 dxeS

'iig4+ just loses) 17 J::!.acl tt::laS 18 'li'bs 'li'b7 19 ..t>£1 e6 Black has equalized ef­fortlessly. 14 . . . tt::la6!

Black typically ignores the pawn in order to speed up his development, whereas 14 . . . bxcS? ! 1S .l:!.hd1 'ifc7 16 J::!.ab1 just gives White some advantage. 15 .l:!.hdl

White cannot solve his problems by simple means: both 1S cxb6 iDeS ! and 1S c6 'iic7 followed by . . . iDeS are good for Black.

Grivas has suggested instead 1S J::!.ad1 ?!, but it seems very risky to me. White's queenside is now abandoned and, moreover, any king retreat to f1 may well block the h1-rook; I don't think that White can afford this waste of time. After 1S . . . 'iic7 16 c6 (16 cxb6 axb6 17 .i.g3 'ifc6 is also unappetizing for White) 16 . . . tt::lcS 17 'iic4, Black must decide how best to prepare the recap­ture on c6:

a) 17 . . . e6?! 18 tt::ld4 eS?! ( 18 . . . .l:!.fc8 19 .ltg3 ! eS 20 tt::lf3 restricts White to an edge) 19 tt::lb3 'ifxc6 20 tt::lxcS bxcS 21 J::!.dS J::!.ab8 22 .l:!.hd1, as in E .Grivas-

i..g5 Sys t e m s

E.Kondou, Athens 199S, shows White's main idea.

b) 17 . . . .l:!.fe8! appears to give Black excellent play. After 18 tt::ld4 a6 (but not, of course, 18 . . . eS?! 19 tt::lb3 'ifxc6 20 tt::lxcS bxcS 21 e4 when White is better) 19 .ltg3 eS 20 tt::lb3 'ifxc6 21 tt::lxcS bxcS, White has managed to implement his plan, but both his king and h1-rook remain misplaced. Indeed, even ex­ploiting his control of dS doesn't really help White; for example, 22 J::!.dS (or 22 'iYdS 'li'bS+ 23 c4 'iiaS 24 'iid2 'iia4 2S .l:!.cl .l:!.ad8, intending . . . 'ifc6 and . . . fS, with an obvious edge for Black) 22 . . . 'li'bS! 23 J::!.xcS 'li'b2+ 24 ..t>f3 'iid2

when the white king is in some danger; Black will attack with moves like . . . 'ifd7, . . . e4 and . . . hS, and I would be surprised if White can survive the onslaught.

c) 17 . . . .l:!.fc8 ! ? is an untried idea I came up with, albeit probably inferior to the simpler 17 . . . .l:!.fe8. Now 18 ..t>£1 ?! 'iixc6 19 .ltxe7 'li'b7! reveals Black's idea; the capture on e7 doesn't come with tempo and . . . tt::le4xc3 will follow with some effect. Instead 18 tt::ld4

1 4 1

Play t h e G r u nfe ld

(White should also avoid 1 8 .l:!.d2? �xc6 19 �xe7 'li'b7! 20 .i.xcS .l:!.xcS 21 �a4 bS) 18 . . . eS 19 �g3 'iie7 20 tt::lf3 ! (White's only chance is to attack eS, otherwise he will have to suffer with his king stuck in the centre; for example 20 tt::lbS? a6 21 tt::ld6 .l:!.xc6 22 'iidS �c7 23 �h4 tt::la4 24 tt::le4 .l:!.c8 2S .i.d8 'li'b8 26 �d7 �f8 27 tt::lgS tt::lxc3+ 28 'it>fl .l:!.6c7 29 .i.xc7 .l:!.xc7 30 �d8 'li'b7 and White re­mains hampered by his inactive king's rook) 20 . . . 'iie8 21 .l:!.dS 'iixc6 22 tt::lxeS (and not 22 �xeS? bS 23 'iid4 f6 24 .i.d6 tt::la4 2S c4 �xc4+ 26 �xc4 .l:!.xc4 which is very good for Black) 22 . . . �e8 23 .l:!.cl bS 24 'iid4 tt::la4 intends . . . tt::lb6-c4 and gives Black excellent compensation for the pawn (analysis by Dembo/Grivas).

Returning to 1S .l:!.hd1 :

1S . . . �c7 ! Less accurate was 1S . . . 'li'b7? ! 16 c6 !

�xc6 17 .i.xe7 in E.Grivas-P.Leko, Dortmund 1992. 16 cxb6

Note that 16 c6 iDeS 17 �c4 e6! 1 8 �fl ( 18 tt::ld4 �xh2) 1 8 . . . 'iixc6 19 .l:!.acl .l:!.fc8 regains the pawn with advantage. 16 . . . axb6

1 4 2

17 �g3 Alternatively: a) 17 .l:!.ab1 iDeS 18 'iixb6 �xb6 19

.l:!.xb6 tt::la4 20 .l:!.b3 tt::lxc3+ 21 .l:!.xc3 .i.xc3 22 �xe7 with a slight edge for Black.

b) 17 'it>fl iDeS 18 'li'b4 .l:!.a4 19 .i.g3 �c6 20 'li'b2 tt::le4 (20 . . . .l:!.fa8! ? 21 tt::ld4 'li'b7 also makes a lot of sense) 21 il.e5 'iic4+ 22 �g1 .i.xeS 23 tt::lxeS �xc3 leads to full equality.

c) 17 a4 iDeS 18 'li'b4 .l:!.aS 19 �g3 (19 �fl .l:!.fa8 20 .i.g3 'iia7 is equal) 19 . . . eS! (shutting the white minor pieces out of play) 20 �f1 (20 tt::ld2 .l:!.fa8 21 tt::lc4? tt::lxa4! 22 'it>fl ? �f8! 23 tt::ld6 tt::lxc3 24 .l:!.acl .l:!.cS 2S .l:!.d3 tt::le4! 0-1 was the fa­mous game J.Lautier-V.Ivanchuk, Ter­rassa 1991, but here White can again maintain equality with 21 �f1 tt::lxa4 22 .l:!.acl ) 20 . . . .l:!.fa8 21 .l:!.ab1 tt::lxa4 22 c4 is also fairly level, albeit perhaps a touch more pleasant for Black. 17 . . . 'iixc3

Black must avoid 17 . . . eS?, since 18 �fl iDeS 19 'iic4! .l:!.a4 20 �dS is some­what better for White, but he can also consider 17 . . . 'iic6 ! ? 18 �fl iDeS 19 'iidS (19 'li'b4 .l:!.a4 20 'li'b2 tt::le4 21 .i.eS �c4+

22 Wg1 ii.xe5 23 tt::lxe5 'iVxc3 is also equal) 19 .. .'iYxd5 20 .l:!.xd5 ii.xc3 21 .l:!.cl ii.f6 22 .l:!.c2 .l:!.a6 with equality, as in T.Bosboom Lanchava-J .Gustafsson, Amsterdam 2001 . 18 'i¥xc3 .i.xc3 19 .l:!.ab1 tt::lb4

The position is completely equal, F .Gheorghiu-E.Liss, Biel 1 994.

B) 4 ii.gs

The immediate bishop move, intro­duced into high-level practice by Tai­manov in 1970, has become very popu­lar recently, thanks to the efforts of several young Georgian and Armenian grandmasters, most notably Aronian,

i..g5 Sys t e m s

Jobava and Sargissian. White seeks to exploit the fact that while tt::lf3 may not be entirely essential at this early stage of the game, Black can hardly refrain from . . . .i.g7. Still, it seems to me that Black's prospects are in no way infe­rior. He should respond in the same manner as in Line A, and indeed the world's elite has settled on this ap­proach. 4 . . . tt::le4

The most consistent reply. 5 .i.h4

White's main choice by far. Alternatively: a) 5 tt::lxe4?! dxe4 leads to trouble as

the g5-bishop is exposed: for example, 6 e3 h6 7 .i.h4 c5 gives Black the initia­tive, while 6 'iVa4+ 'iVd7 7 'iVxd7+ tt::lxd7 8 0-0-0 ii.g7 9 f3 h6! 10 .i.e3 e5! 11 d5 f5 (C.Crouch-J .Rowson, London 1997) is promising for Black.

b) 5 cxd5? ! tt::lxg5 6 h4 tt::le4 7 tt::lxe4 'i¥xd5 8 tt::lc3 'i¥a5 already gives Black already the better prospects thanks to his bishop pair.

c) 5 .i.f4 tt::lxc3 6 bxc3 .i.g7 usually leads back to Line A3; for example, 7 e3

1 4 3

Play t h e G r u nfe ld

cS ! (Black must be ready to answer cxdS with . . . cxd4) 8 li'lf3 or 7 cxdS �xd5 8 li'lf3 c5 9 e3 cxd4 10 exd4 0-0 and in both cases play has transposed. s . . . lt:Jxc3 6 bxc3 dxc4!

7 e3 Instead 7 e4 cS 8 i..xc4 i..g7 is basi­

cally an Exchange Variation with White's dark-squared bishop mis­placed on h4; the pressure on e7 is much less relevant than White's inabil­ity to support his centre with this bishop. Play might continue: 9 li'lf3 li'lc6 10 0-0 0-0 1 1 d5 li::laS 12 i..e2 'ib'd6! ? (I like this move which prepares . . . e5, followed by . . . fS; note that the aS­knight prevents White from harassing the blockading queen with li'lc4) 13 'ib'd3 (13 i..g3 e5 14 'ii'a4 b6 15 li'ld2 i..d7 16 'ib'c2 i..h6! 17 .l:!.fd1 f5 18 exf5 gx£5 gives Black a healthy initiative) 13 . . . eS! 14 li'ld2 i..h6! 15 .l:tfb1 fS 16 f3 (if 16 .l:tb5 b6 17 l:.xaS i..xd2! ) 16 . . . b6 17 i..£2 .l:tf7 18 li'lc4 li'lxc4 19 'ii'xc4 i..d7 20 a4 fxe4 21 fxe4 l:.af8 and Black has good counter­play. This wasn't all forced, but it is a good indication of how Black should play.

1 4 4

Note that 7 lWa4+ �d7 8 �xc4 b6 leaves White with nothing better than to transpose with 9 li'lf3 to Line A53 . In Wang Yue-A.Timofeev, Taiyuan 2006, he tried the independent 9 i..g3 cS 10 i.e5, but Black was fine after 10 . . . i.a6 11 'tlfb3 f6 12 i..xb8 l:.xb8 13 li'lf3 cxd4 14 cxd4 e6 15 e4 i..xfl 16 Wxfl aS . 7 . . . i..e6!

White will now have a hard time regaining his pawn. However, he tends to view his opening play as a long-term positional pawn sacrifice: his solid cen­tral structure blunts the g7-bishop and Black can't easily strike back in the cen­tre. For his part, Black should make the recovery of the pawn as difficult as possible, while also interfering with White's development and trying to create counter-threats.

White must now avoid both 8 i..xc4?? i..xc4 9 'ib'a4+ bS and 8 'ii'a4+? c6, and so usually chooses between:

81: 8 li'lf3 82: 8 .l:tb1 83: 8 'i!Vb1

81) 8 ctJf3 ii.g7

The alert reader will have noticed that we have transposed to Line A51, but with White having enticed Black into meeting 8 e3 there with 8 . . . .i.e6, rather than our preferred 8 . . . b5. How­ever, even with the bishop on e6 this line is pretty harmless for Black. 9 'i¥b1

Instead 9 l:!.b1 b6 transposes to Line B2, while 9 .i.e2 0-0 10 tt:Jg5 .i.d5 1 1 e4 h6 12 exd5 hxg5 13 ii.xg5 'i¥xd5 14 h4 c5 gave Black the initiative in I. Ivanisevic-E.Sutovsky, Internet (blitz) 2006. 9 . . . b6

Clearly mistaken are both 9 . . . c5? ! 1 0 'i¥xb7 cxd4 1 1 'i¥xa8 and 9 . . . 0-0? ! 1 0 'iixb7 tt:Jd7 1 1 tt:Jg5 when White i s much better. 10 tt:Jgs

Attacking the e6-bishop is the most critical move. White can aim to regain his pawn immediately with 10 tt:Jd2, but after the 10 . . . 0-0 (another good op­tion is 10 . . . c5) 1 1 ii.e2 (White's only serious option is to continue develop­ing without paying attention to the c4-

i.g5 Sys te m s

pawn; instead both 1 1 ii.xc4? ii.xc4 1 2 tt:Jxc4 'iid5 and 1 1 tt:Jxc4 c5 are good for Black according to Y ermolinsky and A.Livshits) 1 l . . .c5 12 .i.f3 cxd4! 13 cxd4 of A.Huzman-A.Yermolinsky, Tbilisi 1985, Black can obtain very good play with 13 . . . c3 14 tt:Jb3 tt:Ja6. 1o ... .i.ds

Black should not allow the ex­change of this bishop; for example, 10 . . . 'iid7? ! 11 tt:Jxe6 'iixe6 12 'iYh5+ c6 13 'iixc4 'iie4 14 .i.g3 tt:Jd7 15 l:!.g1 a6 16 .i.d3 left him much worse in H.Grooten-F.Nijboer, Rotterdam 2000. 11 e4 h6!

12 ctJf3 Black is not troubled by either 12

tt:Jh3 .i.b7 13 .i.xc4 0-0 14 0-0 (or 14 e5?, D.Arutunian-T.Sanikidze, Tbilisi 2006, and now 14 . . . c5 is better for Black) 14 . . . c5 or 12 exd5 hxg5 13 'iYh5+ c6 14 dxc6 a6; in both cases he has a good position. 12 ... ii.e6 13 .i.e2 0-0 14 0-0 a6 15 a4 c6 16 'iib4 cs 17 'iia3

17 dxc5 bxc5 1 8 'i:Vxc5 g5 19 .i.g3 .i.xc3 is fully equal for Black. 17 ... gs

1 4 5

Play th e G r u nfe ld

Now 1 8 .ltg3 g4 19 tt:leS cxd4 20 cxd4 'iixd4 was very good for Black in J .Obsivac-T.Oral, Czech League 2002, but even 18 dS .i.xdS 19 I:i.fd1 gxh4 20 I:i.xdS �c8 most certainly doesn' t leave Black any worse.

82) 8 I:i.b1 This is a fairly popular move, but

not one which creates any real prob­lems for Black. He is often able to re­tain the extra pawn on c4 or will use the time White spends recapturing it to strike back in the centre. 8 ... b6

9 tt:lf3 Alternatively: a) 9 .lte2 .lth6 (but not 9 . . . c5?? which

is a blunder: 10 .ltf3 .i.dS 1 1 �a4+ tt:lc6 12 �xc6+) 10 tt:lf3 0-0 11 0-0 c6 12 'iic2 bS with the initiative for Black.

b) 9 tt:le2 .lth6 10 .ltg3 0-0 is also good for Black

c) 9 tt:lh3 .lth6! (less accurate is 9 . . . .i.g7 10 tt:lf4 .ltdS 11 f3 !? with the initiative, while 9 . . . .i.xh3 10 gxh3 �dS is the critical approach but not the most effective, since 11 'iig4! ? fS 12 .ltg2 �aS

1 4 6

13 �e2 �xc3+ 14 �d2 �xd2+ 15 �xd2 c6 16 I:i.hcl gives White good compen­sation) 10 e4 (or 10 .ltgS .ltg7 1 1 tt:lf4 .ltdS 12 'iic2 'iid6 13 .i.e2 c6 14 e4 h6 which was fully equal in N.Pedersen­V.Varadi, Budapest 2006, while Black stands very well in the case of 10 tt:lgS? .i.dS 1 1 e4 .ltxgS) 10 . . . f6! ? 1 1 dS .ltxh3 12 gxh3 �d7 with complicated play, but good prospects for Black. 9 • • • .1tg7 10 tt:Jgs

White does not even equalize after the alternatives; for example, 10 tt:ld2 c6 1 1 'iif3 fS 12 .ltxc4 .ltxc4 13 tt:lxc4 �dS and Black has the initiative, or 10 .lte2 c6 1 1 0-0 0-0 12 e4 bS 13 tt:lgS (V.Malakhatko-I .Khmelniker, Vlissin­gen 2005) 13 . . . .\td7 14 £4 (14 tt:lf3 fS is similar) 14 . . . �a5 15 �c2 cS with the upper hand for Black. 10 ... .\tdS 11 e4

Instead 11 'iia4+ c6 12 .ltxc4 (Black is somewhat better after both 12 e4 h6 13 exdS hxgS 14 .i.g3 'iixdS and 14 .ltxgS 'iixdS 15 .lte3 bS) 12 . . . 0-0 13 .i.xdS cxdS 14 0-0 'iid7 15 �a3 tt:lc6 left Black better in L.Garcia Galeote-I.Valles Moreno, Collado Villalba 2004. 11 ... h6!

The standard reaction in this type of situation. Now 12 exdS hxgS 13 .ltxgS (or 13 .ltg3 'iixdS 14 .lte2 c6 15 0-0 fS with advantage) 13 . . . 'iixd5 14 .lte3 c6 gives Black a clear advantage, E .Kahn­T.Fogarasi, Budapest 2000.

83) 8 �b1 White's most popular and challeng­

ing choice. The queen takes aim at Black's weakened queenside light

i..g5 Sys t e m s

squares and hopes to help White regain was prematurely agreed drawn in his pawn on c4. A.Dreev-B.Avrukh, Gibraltar 200S)

s .. . 'iVds A seemingly very risky move, as it

exposes the queen to harassment either by e4 or ctJe2-f4. However, the queen can move away to aS with tempo if necessary. We must not forget, too, that Black is bringing another piece into play and he can now play . . . il.fS, taking control of the b-file. Overall, it tran­spires Black has enough resources to prevent White from consolidating his position. 9 ctJf3

White can again turn his attention to the bishop on e6 with 9 tt:Je2, but this is now well met by the paradoxical 9 . . . .\tc8 ! . This rather unnatural retreat in fact completely takes the sting out of White's set-up, due to the threat of . . . gS should tt:Jf4 be played:

a) 10 .:t:J£4? 'iVaS 11 �4 (11 �2? gS just wins a piece) l l . . .'iVxb4 12 cxb4 bS! is very strong in view of the possible continuation 13 tt:JdS tt:Ja6 14 a4 c6 IS axbS cxdS 16 bxa6 .l:!.b8.

b) 10 a4 il.h6 11 il.g3 (this position

l l . . .cS 12 �S+ .ltd7 13 'iVxcS 'iVxcS 14 dxcS ctJa6 IS ctJd4 tt:JxcS 16 .ltxc4 il.g7 leaves Black better.

c) 10 .ltg3 il.g7 11 .ltxc7 0-0 gives Black a lead in development and the initiative; . . . eS will follow.

Finally we should note that 9 a4? ! tt:Jd7 10 tt:Jf3 'iVaS 1 1 �2 tt:Jb6 12 �S+ 'iVxbS 13 axbS .ltd7 transposes to the note to White' s 11th move, below. 9 . . • 'iVa s

I like this forcing move more than the usual 9 . . . tt:Jd7. It is true that Black has also been doing fairly well after that, but White can generate some play; for example, 10 .lte2 il.fS 1 1 �2 .ltg7 (l l . . .tt:Jb6 12 a4 aS 13 0-0 .ltd3 14 .ltxd3 cxd3 IS e4 'iVxe4 16 .l:!.fel gave White very good compensation in T.Radjabov­E.Sutovsky, Rishon Le Zion (blitz) 2006) 12 0-0 eS 13 �4 cS 14 'iVxc4 'iVxc4 IS .ltxc4 .l:!.c8 saw Black go on to win in L.Aronian-P.Svidler, Dormund 2006, but at this stage I feel that White was a little better.

10 'iVb2

1 4 7

Play t h e G r u nfe ld

Retaining some influence on the b­file. Less incisive is 10 'iVc2 CL:ld7 1 1 CL:\d2 CL:lb6 12 �e2 0-0-0 and after . . . hS White's dark-squared bishop will be in trouble, in view of the threat of . . . gS and . . . h4. 10 . . . ctJd7 11 ii.e2

Instead 11 a4? is a mistaken idea. White wants to exchange queens, but his queenside pawns remain targets and his development backward: 1 1 . . .ltJb6 12 'iVbS+ (or 12 lLld2 0-0-0 13 ii.e2 gS 14 ii.g3 hS with the initiative for Black) 12 .. ."�xb5 13 axbS �d7 14 �g3 �xbS 15 ii.eS f6 16 �xc7 ltJdS 17 �aS e6 and Black's positional advan­tage is unquestionable, not to mention his extra pawn.

1 1 ltJgS is also insufficient: 1 1 . . . ii.dS 12 e4 h6 13 exdS hxgS 14 ii.xgS 'ifxdS with a clear plus; Black will follow up with . . . 0-0-0 and a central break. 11 . . • CtJb6 12 0-0 �g7 13 CtJd2 cs 14 ctJb3 'iVbs 15 'ifa3 'ifa4

16 'ifxa4+? was seen in G.Sargissian­M.Vachier Lagrave, Turin Olympiad 2006, but this seems entirely wrong. The

1 4 8

game continued 1 6 . . . CtJxa4 1 7 ltJxcS ltJxcS 18 dxcS and now Black should have simply played 18 . . . 0-0-0 when c3 is hanging and . . . l:td2 imminent. How­ever, even after 16 'ifxc5 f6 Black has the initiative. He will play .. :�£7 next and then centralize his rooks, while White must attend to his b3-knight and his pieces are uncoordinated.

Conclusion

We've covered a number of lines in this chapter. Black must respond accurately to the �gS systems, but should he do so, he can obtain plenty of counterplay. After 4 CtJf3 �g7 5 �gS ltJe4, 6 cxdS is tricky since White obtains a lead in de­velopment, but with careful play Black should be the one to benefit should a sharp skirmish ensue. More demand­ing is 6 ii.h4, after which Black really has to capture the c4-pawn. White can then immediately recapture the pawn, albeit at some cost in terms of devel­opment, or consider his opening play a pawn sacrifice. However, in the safer former case, Black fully equalizes, while the latter gives him good play.

More troublesome, at least for an unprepared black player, is the imme­diate 4 .i.gS. Black should again re­spond with . . . ltJe4 and an exchange on c3 after which the resulting positions are pretty complex. Following up with the popular 8 �1 is probably White's best try, but is still quite risky for him and currently 8 . . . 'iVd5 looks like an ex­cellent response.

C h a pte r E l eve n I The So l i d 4 e 3

1 d4 tt:Jf6 2 c4 g6 3 tt:Jc3 d 5 4 e3 With this solid choice White essen­

tially refrains from any kind of theo­retical discussion. He intends simply to develop and relinquishes any hope of gaining a theoretical advantage. Most of the lines examined in this chapter either involve natural and unambitious development or a specific (but usually artificial) scheme to discourage the standard . . . c5-break. Almost any legal move is available to White at every step and it is naturally impossible, as well as rather pointless, to cover every possibility. Instead we will focus only on specific set-ups where a less obvi­ous treatment is required. Black play­ers should just remember that their main idea is to castle and play . . . c5; in most cases that should be enough to give you a good position!

One very important point about these lines is that, after Black plays . . . c5, there often occurs a total clearance of pawns from the centre. In that case

Black is often left with the initiative, thanks to his actively-placed pieces and particularly the g7-bishop. Therefore, do not fear this simplification of the position! 4 . . . .i.g7

White usually now continues to de­velop solidly, although he can also pressurize d5:

A: 5 'iVb3 B: 5 ctJf3

1 4 9

P lay t h e G r u nfe l d

The alternatives are rather rare: a) 5 .id2 aims to bring a major piece

to the c-file and thereby discourage . . . c5, but what is the bishop doing on d2? A.Aleksandrov-V.lvanchuk, Halkidiki 2002, continued 5 . . . 0-0 6 J:cl (or 6 'iVc2 tt:'lc6, intending . . . ctJb4 and . . . .i£5, when 7 tt:'lxd5 tt:'lxd5 8 cxd5 'iVxd5 9 tt:'lf3 .if5 10 i.d3 i.xd3 1 1 'iVxd3 J:ad8 leaves Black on top; 6 c5 mechanically prevents . . . c5, but is met by 6 . . . b6 7 cxb6 axb6 and an even stronger . . . c5 follows) 6 . . . c5! 7 dxc5 tt:'la6 8 tt:'lxd5 (instead 8 tt:'la4 is well met by 8 . . . d4 9 exd4 'iVxd4 10 ctJf3 'iVe4+ 1 1 .ie2 J:d8 with advantage, while 8 cxd5 tt:'lxc5 9 .ic4 .if5 10 tt:'lge2 .id3 1 1 .ib3 was seen in R.Cusi-A.Yermolinsky, Oak Brook 2003, and now 1 l . . .'iVb6 threatens both . . . ..ia6 and . . . tt:'lxb3; Black is better) 8 . . . tt:'le4 9 f3 tt:'lxd2 10 'iVxd2 e6 1 1 tt:'lc3 and now Ivanchuk could have obtained good compensation on the dark squares after 1 l . . .'iVh4+ 12 g3 'iVf6 or 12 'iVf2 'iVh6.

b) 5 b4 is an idea seen more often af­ter 5 ctJf3 0-0 . Here it can be effectively countered by 5 . . . 0-0 6 'iVb3 (or 6 ..ib2 b6! 7 'iVb3 c5! 8 tt:'lxd5 tt:'lxd5 9 cxd5, as in J .Johansson-N.Crickmore, Olomouc 2002, and now 9 . . . cxd4 favours Black) 6 . . . a5 ! 7 b5 (7 bxa5? dxc4 8 ..ixc4 tt:'lc6 is much better for Black) 7 . . . tt:'lbd7 intend­ing . . . dxc4 and/or . . . e5, and the imme­diate 7 . . . e5 ! ? is also possible.

A) 5 'iVb3 This is a respectable move, with

ideas similar to those in the Russian System. Here White is able to recapture on c4 with the bishop, so . . . dxc4 is not a good idea. On the other hand, 4 e3 has

1 5 0

greatly restricted White' s possibilitie� allowing Black simply to strengthen d� and complete development.

s ... e6 6 iVa3 With this very unnatural move

White intends to disrupt Black'� smooth development by preventin§: castling, and this system is covered ir. great detail by Timothy Taylor in hi� recent Beating the King's Indian an,;

Griinfeld. In my opinion, White canno: possibly treat a chess opening like this especially one as dynamic as the Gri.in­feld ! White's idea violates all the cardi­nal rules of the opening and its nature is very one-sided; the queen is mis­placed on a3 for any other purposes the minor pieces still at home . . .

Botvinnik liked t o meet this idea with . . . 'iVe7, which of course should suffice for equality, but I feel that it's a crime to exchange the white queen: Black should gradually prepare the . . . c5 break and not allow himself to be provoked by White's play. Indeed White should really prefer 6 ctJ£3 transposing to Line B 1 . 6 .. . tt:'lc6!

The very simple idea behind this move is . . . t}je7, facilitating castling. Black, it is true, cannot break with . . . c5 anytime soon, but he has all the time in the world to prepare it. 7 t}jf3

7 cxd5 exd5 brings about a Carlsbad structure, but this particular version cannot possibly be good for White. One example went 8 .ib5 ..ltd7 9 t}jf3 t}je7 10 £Lxd7+ t}jxd7 ( 10 . . . 'ifxd7 i s also fine) 1 1 £Ld2 0-0 12 0-0 t}jb6 13 b3 t}jfS 14 'iVb2 .l:te8 with at least equality in A.Kotov­K.Honfi, Sarajevo 1966. 1 . . . t}je7

Black will now castle, perhaps ex­change on c4 to free d5 for the e7-knight, and prepare . . . c5 with . . . b6 and . . . .ib7. In all the lines that follow Black easily equalizes the play, and in practi­cal terms these 'equal' positions offer Black excellent long-term chances and much the easier play. 8 £Ld3

The most natural development and Taylor's main preference.

Alternatively: a) 8 .ie2 0-0 9 0-0 (or 9 ..ltd2 b6 1 0

Th e S o l i d 4 e 3

0-0 ..ib7 1 1 cxd5 exd5 12 .ia6 ..ltxa6 13 'ifxa6, O.Jovanic-B .Predojevic, Kusa­dasi 2006, and now 13 . . . c5 followed by . . . t}je4 is equal) 9 . . . b6 10 cxd5 t}jexd5 1 1 t}jxd5 ( 1 1 e4? t}jxc3 12 'ifxc3 t}jxe4 1 3 'ifc6 'ifd5 1 4 'ifxc7 .ib7 i s excellent for Black) 1 l . . .exd5 12 b4 t}je4 13 .ib2 ..ltg4 14 l:He1 (I .Khenkin-B.Alterman, Tel Aviv 1994), and now Black should play either 14 . . . c6 or 14 . . . a6 with equality.

b) 8 ..ltd2 0-0 9 J:cl t}je4 10 ..ltd3 (or 10 t}jxe4 dxe4 1 1 t}jgS f5 12 .i.c3 e5 1 3 dxe5 t}jc6 with the initiative) 10 . . . t}jxd2 1 1 t}jxd2 b6 12 0-0 .ib7 looks promising for Black, who is about to blast open the position.

c) 8 'ifa4+ c6 9 'ifc2 0-0 10 ..lte2 (I.Farago-V.Jansa, Baile Herculane 1982) 10 . . . dxc4 11 Jl.xc4 c5 is equal . 8 . . . 0-0 9 0-0 b6

10 cxds Again not the only try, but 10 e4

dxe4 1 1 t}jxe4 ..ltb7 12 .ig5 t}jfs 13 J:ad1 h6 14 .ixf6 ..ltxf6 15 t}jxf6+ 'iVxf6 16 Jl.xf5 'ifxf5 is a t least equal for Black, and 10 c5 bxc5 1 1 'ifxc5 t}jd7 12 'iVa3 c5 13 ..ltd2 (13 dxc5 'ifc7 14 b4 aS is prob­lematic for White, as Taylor notes)

1 5 1

Play t h e G r ii nfe /d

13 . . . c4 14 i.e2 e5 leaves Black the one idea behind them and should be pulling the strings. treated with some care. 10 ... exds 11 .1i.d2

This occurred in the game T.Taylor­V.Mikhalevski, Las Vegas 2006. Instead 11 e4 dxe4 12 tt:'lxe4 tt:'lfd5 intends . . . h6 and . . . .1i.e6, with an initiative for Black, while 1 1 b4 hopes to prevent . . . c5 for good, but it doesn't promise anything: 1 l . . . a6 12 �3 (12 b5?! .1i.f5! 13 i.xf5 axb5 14 �3 tt:'lxf5 15 'ii'xb5 is just better for Black) 12 . . . .1i.f5 13 .1i.xf5 tt:'lxf5 14 a4 c6 15 i.a3 (15 b5 axb5 16 .1i.a3 bxa4 also fails to solve White's problems) 15 . . . b5 16 tt:'le5 �6 and with . . . tt:'ld6-c4 com-ing, Black is better. u . . . .1i.fs 12 .1i.xfs

Also promising is 12 i.e2 tt:'le4. 12 . . . tt:'lxfs 13 .Mac1 tt:'ld6!?

Black stands well and will follow up with .. J�.e8 and . . . c6. He can later play a knight to e4 and seek to expand on the kingside, whereas White does not have any obvious way to proceed.

B) 5 ctJf3 0-0 White now has a choice of several

set-ups. Most of these have a specific

1 5 2

81: 6 'ii'b3 82: 6 b4 B3: 6 cxds 84: 6 i.d2 Bs: 6 .1i.d3 86: 6 .1i.e2

Note that 6 c5? ! is, of course, met by 6 . . . b6 7 b4 aS when White's pawn chain collapses.

81) 6 'ii'b3 e6

White again discourages . . . cS, as his pressure on the dS-pawn keeps Black's pieces tied up. Thus Black should opt to complete development with . . . b6 and . . . i.b7, slowly preparing . . . cS. 7 i.d2

Quickly completing queenside de­velopment. The alternatives have also failed to impress; for example:

a) 7 i.e2 b6 8 cxdS exdS 9 0-0 i.b7 10 a4 aS 1 1 .l:!.d1 tt:Jbd7 12 i.d2 'fie7 and Black stands well.

b) 7 i.d3 cS! 8 dxcS tt:Ja6 9 'fia3 b6 10 cxb6 axb6 11 0-0 i.b7 leaves White's pieces rather loose and Black with good compensation.

c) 7 cxdS exdS 8 i.d3 b6 9 0-0 cS sees Black effortlessly achieve his break. 7 . . . b6

8 .l:!.c1 Again White has several alterna­

tives: a) 8 cxdS exdS 9 .l:!.cl (or 9 tt:JeS i.e6

10 i.e2 tt:Jfd7 1 1 tt:Jxd7 'fixd7 12 'fia4 cS 13 'fixd7 tt:Jxd7 14 0-0-0 cxd4 lS exd4 .ixd4 16 i.h6 i.g7 17 i.xg7 �xg7 18 �xdS .l:!.ac8+ 19 �bl ltJcS and Black was better in D.Jacimovic-I.Nepomniach-

Th e S o l i d 4 e3

tchi, Feugen 2006) 9 . . . i.b7 10 i.e2 tt:Jbd7 1 1 0-0 cS 12 dxcS tLlxcS 13 'fia3 tt:Jfe4 14 .l:!.fdl (S.Nikolic-V.Jansa, Ger­man League 2001 ) 14 . . . 'fie7, intending to centralize the rooks, is equal.

b) 8 i.e2 i.b7 9 0-0 tt:Jbd7 10 cxdS (10 .l:!.acl cS 1 1 cxdS tLlxdS is equal) lO . . . exdS 11 .l:!.fdl (Black managed to wrest the initiative after 1 1 tLla4 tt:Je4 12 .l:!.acl .l:!.b8 13 i.b4 .l:!.e8 14 i.bS aS lS i.el .l:!.e6 in M.Bertok-P.Benko, Stock­holm Interzonal 1962) l l . . . .l:!.e8 12 i.el c6 with typical and easy play for Black.

c) 8 .l:!.dl i.b7 9 i.e2 tt:Jbd7 10 0-0 cS ! 1 1 cxdS tt:JxdS (G.Barbero-M.Ghinda, World Team Ch., Lucerne 198S) and Black has achieved his aims.

8 ... .ib7 9 .ie2 tt:Jbd7 10 cs c6 11 0-0 Instead 11 cxb6? axb6 12 a3 cS 13

h4? ! tt:Je4 was horrible for White in I .Kincs-M.Konopka, Hungarian Team Ch. 1994. The text is better and was seen in G.Stahlberg-J .Iliesco, Mar del Plata 1943, but after l l . . .tLle4 12 tLlxe4 dxe4 13 tLlel eS Black again has the up­per hand.

82) 6 b4

1 5 3

Play t h e G r ii nfe ld

Makagonov' s variation, which has received some recent attention. 6 ... i..e6!?

I like this move, forcing a clarifica­tion of White's plans. Instead 6 . . . b6 is the traditional move, but 7 .ia3 c5 8 bxc5 bxc5 9 �cl ! has been causing some problems of late. 7 cs

Committal, but best. Instead: a) 7 tt:lg5 £Lf5 ! 8 .ie2 (8 cxd5 h6 9 e4

hxg5 10 exf5 tt:lxd5 11 tt:le2 e5 gives Black a clear advantage) 8 . . . a5 9 b5 c5 10 bxc6 tt:lxc6 gives Black the initiative.

b) 7 tt:le5 dxc4 8 .ixc4 (8 tt:lxc4 tt:ld5 9 i..b2 tt:ld7 is fine too) 8 . . . .1i.xc4 9 tt:lxc4 tt:ld5 10 'i!Yb3 tt:lc6 1 1 tt:lxd5 ( 1 1 .1i.a3? tt:lxd4 12 exd4 tt:lxc3 13 'l?Vxc3 .1i.xd4 14 'l?Vcl i..xa1 15 'liixa1 'l?Vd3 was disastrous for White in V.Rabrenovic-S.Atalik, Vrnjacka Banja 1999) 1 1 . . .'liixd5 12 0-0 a6 is comfortable for Black.

c) 7 'i!Yb3 a5 8 b5 (or 8 c5 tt:le4! 9 .ib2 as in I .Drasko-I.Sokolov, Sarajevo 1987, and now 9 . . . axb4 10 'liixb4 tt:la6 1 1 .1i.xa6 bxa6 gives Black a small edge) 8 . . . c5 (so Black achieves his break after all ! ) 9 bxc6 (9 i..a3? dxc4 10 .1i.xc4 a4! nets a

1 5 4

piece and 9 dxc5 tt:lbd7 10 i..a3 tt:le4 i;: also very promising for Black) 9 .. . bxct 10 c5 (10 'i!Yb7 tt:lbd7 11 'liixc6 �c8 re­covers the pawn with a good position 10 . . . tt:lbd7!? 1 1 'liia3 .1i.g4 12 i..e2 'liic:­intends . . . e5 and Black has no prob­lems, to say the least. 7 ... i..g4

8 h3 8 i..e2 allows Black to prepare . . . e5

at ease: 8 . . . tt:lbd7 9 0-0 c6 10 i..b2 'liic7 1 1 a4 b6 12 cxb6 axb6 13 'i!Yb3 tt:le4 14 �fdl e6 15 h3 i..xf3 16 i..xf3 tt:ld6 and the idea of . . . f5-f4, as well as control of the c4-square, gave Black the advantage in B .Crane-M.Micklethwaite, correspon­dence 1998 . 8 ... i..xf3 9 gxf3

White preferred 9 'liixf3 in M.Bens­dorp-N.Narings, Amsterdam 2005, and should have been punished by 9 . . . e5! 10 dxe5 tt:le4 11 tt:lxe4 i..xe5 12 �bl dxe4 13 'liixe4 .1i.c3+ 14 �e2 tt:lc6 with a powerful initiative. 9 . . . c6 10 f4

10 e4 tt:lh5 (or 10 . . . dxe4 1 1 fxe4 e5) 1 1 .1i.e3 e5 intends . . . f5 and leaves Black on top.

10 ... tt::Jbd7 11 Si.d2 Instead 11 'ifc2 b6 is good for Black,

as is 1 1 b5 cxb5 12 i.xb5 a6 13 i.xd7 tZ:lxd7 intending . . . b6.

The text was preferred in V.Ako­bian-J .Rowson, Philadelphia 2002, and now after the logical 1 l . . .b6 once more, Black intends . . . 'ifc7 and . . . aS with good play.

B3) 6 cxds This line, Taylor's other suggestion,

was played by Keres and remains quite dangerous for the unwary. 6 ... tt::Jxds

7 Si.c4

Th e So l i d 4 e 3

Obviously the only critical move. Instead 7 tZ:lxdS 'ifxd5 8 i.d3 c5 equal­izes very easily and 7 Si.d2 also fails to impress: 7 . . . c5 8 'ifb3 (or 8 dxc5 tt::Jxc3 9 Si.xc3 i.xc3+ 10 bxc3 'ifaS 1 1 'ifb3 'ifxc5 12 Si.c4 tZ:ld7 and Black was fine in R.Dujin-L.Van Wely, Dutch League 2001) 8 . . . tt::Jxc3 9 i.xc3 (9 bxc3 tZ:lc6 10 i.e2 was preferred in I .Gavrilov­S.Petrov, Struga 2002, but after 10 . . . cxd4 1 1 cxd4 e5 12 dxe5 tZ:lxeS 13 �d1 Si.e6 14 'ifxb7 �b8 15 'ifxa7 tZ:ld3+ Black has excellent compensation) 9 . . . cxd4 10 tZ:lxd4 (or 10 i.xd4 i.xd4 1 1 tZ:lxd4 e5 1 2 tZ:lf3 tZ:lc6 with advantage) 10 . . . e5 (improving over the 10 . . . tt::Jd7 of J.Bosch-E.L' Ami, Dutch Ch., Hilversum 2006) 1 1 tZ:lf3 tZ:lc6 12 �d1 'ife7 and Black intends . . . .ie6 or . . . i.g4, with a good game. 7 ... tt::Jxc3 8 bxc3 cs 9 0-0

Note that 9 h3 'ifc7 prevents 10 'ife2? in view of 10 . . . cxd4 1 1 cxd4 i.e6 12 Si.xe6 'ifc3+. g . . . 'ifc7 !

The most accurate, highlighting the loose position of the c4-bishop. It is surprising how much this simple move

1 5 5

Play th e G r iJ nfe /d

disorganizes White. 10 tt:ld2

Alternatively: a) 10 'ii'e2 allows the annoying pin

10 . . . i.g4 and then, for example, 11 i.a3 tt:ld7 12 i.b5 (or 12 �ab1 tt:lb6 13 i.d3 c4 14 i.c2 tt:ld5 with the initiative -Gipslis) 12 . . . a6 13 i.xd7 i.xd7 14 i.xc5 b6! 15 i.a3 i.b5 16 c4 i.xc4 17 �fcl �fc8 and Black has the advantage (Bagirov) .

b) 10 i.b3 leaves the bishop mis­placed: 10 . . . tt:lc6 1 1 i.a3 b6 12 'ii'e2 tt:la5 13 i.a4 i.b7 14 e4 a6 15 �ad1 �ac8 with easy equality.

c) 10 i.e2 b6 11 i.a3 i.b7 12 �cl tt:ld7 is also fine for Black. 10 ... tt:ld7 11 'ii'e2 b6 12 i.b2 i.b7 13 �ac1

This was seen in V.Ilinsky­M.Konopka, Elista Olympiad 1996. Now Black can play 13 . . . a6 and with . . . b5 coming, he has the upper hand.

84) 6 i.d2 Karel Opocensky' s line and one

which Black should meet head -on. 6 .. . cs !

1 5 6

7 dxcs Instead 7 cxd5 cxd4 8 tt:lxd4 tt:lxd5 i'

simply equal . 7 . . . tt:la6 8 cxds

White cannot play to hold on to the c5-pawn: for example, 8 tt:la4 i.d7 9 b4� tt:le4 10 �cl tt:lxd2 1 1 tt:lxd2 tt:lxb4 L 'i¥b3 'ii'a5 13 a3 tt:la6 14 tt:lc3 d4! wa' superb for Black in A.Fodor-Z.Timar correspondence 1984. 8 ... tt:Jxcs 9 i.c4

So White holds on to the d5-pawr with all his might. Black cannot imme­diately regain it, but he can exploit tht: placement of White's pieces to generatE a very dangerous initiative. 9 .. . a6! 10 a4

Preventing . . . b5. Indeed, the forcefu: 10 b4? ! meets with retribution in thE form of 10 . . . b5! 1 1 bxc5 ( 1 1 i.e2 tt:lce-± 12 0-0 i.b7 - Adorjan - is very good for Black, who regains d5) 1 1 . . .bxc4 12 e-± (or 12 'ii'a4? tt:lxd5 13 'ii'xc4 tt:lxc3 H i.xc3 i.e6 15 'i¥b4 �b8 16 'ii'a3 'ii'd3 and Black wins) 12 . . . e6! when White has many weaknesses as Black opens the position with crushing effect; for ex­ample, 13 d6 i.b7 14 �cl �c8 15 tt:la-± tt:lxe4 16 i.b4 (or 16 �xc4 tt:lxd6! 16 . . . tt:lxc5! 17 tt:lxc5 �xeS leaves Blacl,. with some initiative and advantage. 10 ... i.g4 11 0-0

Alternatively, 1 1 h3 i.xf3 12 gx£3 ( 12 'ii'xf3 b5 13 axb5 axb5 14 �xa8 'ii'xa� 15 tt:lxb5 'ii'a4! 16 b3 tt:lxb3 17 ..ixb3 'il'xb3 is much better for Black) 12 . . . e6! 1 13 dxe6 fxe6 gives Black good compen­sation in view of White' s weakened kingside and the threat of . . . tt:ld3+. 11 ... �c8 12 h3

We have been following E .Ubilava­O.Romanishin, Tbilisi 1986, and now 12 . . . t:Z:lce4! 13 hxg4 .l::l.xc4 leaves Black with sufficient activity to at least regain his pawn.

BS) 6 �d3

This and 6 �e2 both allow . . . c5 and . . . cxd4, bringing about a position very similar to a reversed Tarrasch Defence. White's extra tempo is not really rele­vant in the standard IQP position that arises, and both here and in Line B6, I will concentrate chiefly on some the­matic examples. 6 . . . cs 1 o-o

Th e S o l i d 4 e3

Instead 7 cxd5 cxd4 8 t:Z:lxd4 t:Z:lxd5 9 t:Z:lxd5 (and not 9 'iVb3?? t:Z:lxc3 10 'l?Vxc3 e5 1 1 t:Z:lb5 e4, winning a piece) 9 . . . 'l?Vxd5 1 0 0-0 .l::l.d8 already gives Black some pressure, while 7 dxc5 dxc4 8 �xc4 'l?Vxd1 + 9 t:Z:lxd1 t:Z:le4 10 0-0 t:Z:ld7 11 c6 bxc6 12 �e2 aS leaves him with the initiative. 7 . . . cxd4 8 exd4

8 t:Z:lxd4 allows the dynamic 8 . . . e5 9 t:Z:lb3 e4 10 �e2 dxc4 1 1 �xc4 'liVc7 12 �e2 t:Z:lc6 with an edge. 8 • • • t:Z:lc6

We have reached a reversed Tar­rasch. Now 9 c5 �g4 gives Black good play, as does 9 h3 dxc4 10 �xc4 �f5 1 1 �g5 .l::l.c8, while the aggressive 9 cxd5 t:Z:lxd5 10 �e4 �e6 1 1 t:Z:lg5 (R.Perisic­Z.Licina, Bosnjaci 2003) fails to im­press: 1 l . . .t:Z:lxd4 12 lZ:lxe6 lZ:lxc3 13 t:Z:lxd8 (13 bxc3 t:Z:lxe6) 13 . . . t:Z:lxd1 14 lZ:lxf7 lZ:lxf2 15 .l::l.xf2 .l::l.xf7 16 .l::l.xf7 'it?xf7 with an edge for Black.

86) 6 �e2 In view of . . . .ig4 possibilities, this is

the most accurate way of entering the reversed Tarrasch.

1 5 7

P lay th e G r ii nfe ld

6 . . . cs ! 7 o-o cxd4 8 exd4 tt:Jc6

9 i.gs White has tried a whole host of

other moves, but none really troubles Black:

a) 9 h3 �£5 10 �e3 dxc4 1 1 �xc4 :cs 12 jLe2 'ifd7 and Black is pressuriz­ing the d4-pawn.

b) 9 :e1 dxc4 10 �xc4 b6 1 1 jLgS i.b7 12 a3 h6 13 jLf4 e6 was fine for Black in G.Giorgadze-V.Ivanchuk, Minsk 1986.

c) 9 CDe5 dxc4 10 CDxc6 bxc6 11 jLxc4 'ifd6 with an edge for Black, as d4 is

1 5 8

weaker than c6. d) 9 c5 i.g4 intends . . . tt:Je4.

9 . . . dxc4 10 i.xc4 10 d5 tt:Ja5 1 1 b4 cxb3 12 axb3 �g4 is

promising for Black, in view of the standard tactic 1 3 b4 :c8! , or simply 13 . . . i.xf3. 10 .. . i.g4

With good play for Black, B .Vercammen-M. Turcan, dence 1998.

Conclusion

correspon-

The e3 systems are pretty harmless for Black. White's plan of simple de­velopment is not aggressive enough and allows Black to employ the stan­dard . . . c5-advance without any prob­lems. More interesting are White's at­tempt to discourage . . . c5 or to prevent it altogether, with moves such as 6 b4 and 6 'ifh3. Black can then shore up d5, and slowly prepare . . . c5, or try to clar­ify the central tension; he has nothing to fear in either case.

C h a pt e r Twe l ve I Offbeat 4th Move Alte rnat ives

1 d4 tt'lf6 2 c4 g6 3 tt'lc3 ds Some very rare moves have been

tried here, mostly in an effort to avoid theory. However, these moves are hardly threatening; in fact Black is usu­ally able to obtain the better chances.

A: 4 g4 B: 4 h4 C: 4 f3

A) 4 g4?!

Such a move can be hard to take se­riously, of course, but an accurate re­sponse is still required to highlight its deficiencies. White hopes to destabilize Black's central ir"!.fluence, but. . . 4 ... dxc4!

. . . Black is happy to resolve the cen­tral tension, and now the g4-pawn is en prise. 5 h3

Instead Black is better after 5 g5 tt'ld5 6 'Yj'a4+ (6 e4 tt'lb6 7 lte3 h6 8 f4 hxg5 9 fxg5 ltg7 was also excellent for

1 5 9

Play th e G r u nfe ld

Black in R.Moravcik-J .Pribyl, Sala 1995) 6 . . . c6 7 'it'xc4 .i.g7, while 5 e4 .i.xg4 6 f3 .i.c8 7 .txc4 .i.g7 8 eS tt:lfd7 (J.Pietrasanta-R.Bagirov, Le Touquet 1998) leaves him with an extra pawn. s .. . tt:lds ! 6 e4

Note that 6 .i.g2 c6 is just bad for White, and 6 'it'a4+ c6 7 'it'xc4 tt:lb6 8 �d3 .i.g7 9 tt:lf3 tt:la6 gives Black a seri­ous initiative. 6 • • • tt:lb6!

Black simply holds on to the extra pawn in view of 7 .i.xc4? tt:lxc4 8 'it'a4+ tt:lc6 9 �xc4 �xd4. 7 .i.e3 .i.g7 8 f4 tt:lc6 9 tt:lf3 e6 10 'ii'd2 �e7 11 .i.e2 .i.d7 12 o-o .l:r.d8 13 .l:r.ad1 .i.c8 14 'ii'e1 fs

This logical sequence left Black clearly better in J .Obsivac-J.Pribyl, Czech League 1998.

B) 4 h4!?

More of a test than our previous variation. 4 .. . cs !

I feel this is best. White's 4th move doesn't contribute to the central strug­gle in any way and so Black is fully

1 6 0

justified to strike there. s cxds

White has often preferred 5 dxcS, but after 5 . . . d4 6 tt:lbS tt:lc6 7 e3 eS

8 exd4 (or 8 tt:lf3 .i.xcS 9 exd4 exd4 10 .i.d3 a6 1 1 tt:la3 0-0 and Black is clearly better, J.Bertin-A.Rodier, French League 2002) 8 . . . tt:lxd4 9 tt:lf3 (instead 9 b4 aS 10 .i.gS .i.e7 1 1 tt:lxd4 exd4 gave Black excellent compensation m M.Cebalo-M.Sebenik, Pula 2004, while after 9 tt:lxd4 exd4 10 .i.d3 .i.g7 1 1 .i.f4 0-0 12 tt:lf3, Y.Gozzoli-K.Van der Weide, La Fere 2004, Black can play 12 . . . �a5+ with the initiative) 9 . . . .i.xc5 10 tt:lbxd4 exd4 1 1 .i.d3 .i.g4 Black is on top, S.Porat-A.Jerez Perez, Andorra 2001 . s • • . tt:lxds 6 dxcs

Black also gains the initiative after 6 hS .i.g7 7 hxg6 hxg6 8 .lir.xh8+ .i.xh8 9 e3 (G.Kundrak-G.Kosa, Szombathely 2003) 9 . . . cxd4 10 exd4 tt:lc6 1 1 tt:lf3 .i.fS . 6 • • . tt:lxc3 7 �xd8+ �xd8 8 bxc3 .i.g7 9 �d2

Or 9 .i.d2 tt:la6 10 c6 bxc6 11 e4 tt:lcS 12 f3 .i.e6 again with the initiative for Black, B .Szabo-J.Sondermann, Buda­pest 2006.

9 . . .Cba6!? 9 . . . .if5 10 £3 CLJd7 11 e4 ke6 12 c6

bxc6 was played in A.Zaitsev­V.Smyslov, Sochi 1963, and also looks promising; as does 9 . . . CLJd7!?, intending 10 c6 CLJ£6 1 1 £3 CLJhS with compensa­tion. 10 c6 ct:Jc5!

Davies' idea. White's position looks decidedly unharmonious and Black has good compensation.

C) 4 f3 The most serious option in this

chapter. White aims to build a big pawn centre, although the way he goes about it looks somewhat ugly. 4 .. . c5 !

Again I believe that this direct cen­tral counter is best. Black aims to ex­ploit the weakened dark squares in White's camp. Now 5 kgS kg7 6 e3 (or 6 dxcS d4 7 CLJbS CLJc6 8 e3 h6 9 .i£4 eS 10 CLJd6+, as in J.Jambrich-D.Schwarz, Bratislava 1998, when 10 . . . �£8 1 1 kg3 'iVaS+ 12 'iVd2 CLJb4 leaves White in a bad way) 6 . . . cxd4 7 exd4 CLJc6 8 CLJge2 dxc4 leaves Black with an extra pawn,

Offb e a t 4 th M o ve A l te rn a t ives

M.Lopes-A.Ferreira de Souza, Juiz de Fora 1966. Therefore, White must choose a capture in the centre .

C1: 5 cxd5 C2: 5 dxc5

C1) 5 cxd5 ct:Jxd5 6 e4 Alternatively, 6 CLJa4 .ig7 7 CLJxcS

CLJc6 8 CLJb3 CLJb6 9 e3 (I.Botvinnik­A.Greenfeld, Israeli League 2002) 9 . . . e5 and Black has the initiative. 6 . • • ct:Jxc3 7 bxc3 .ig7

8 .ib5+ Instead 8 ke3 CLJc6 9 kbS (9 CLJe2

1 6 1

Play th e G r u nfe ld

cxd4 10 cxd4 'ii'a5+ 1 1 'ii'd2 'ii'xd2+ 12 Wxd2 0-0 is simply an inferior version of the Exchange Variation; White's kingside development is seriously hampered by the need to keep d4 pro­tected) 9 . . . 0-0 10 i..xc6 bxc6 11 'ii'd2 (D.Dela Rosa-V.Di Fonzo, Geneva 1994) 1 l . . .i..a6 keeps the white king stuck in the centre and should thus be very good for Black. 8 . . . tL'lc6 9 i..xc6+

Also good for Black is 9 tLle2 cxd4 10 i..xc6+ bxc6 1 1 cxd4 0-0 12 i..e3 i.a6, I .Kincs-L.Vadasz, Balatonbereny 1995. 9 • • • bxc6 10 tLle2 i.a6 11 i..e3 o-o 12 dxcs 'iWas 13 o-o l::tfd8 14 'i:Vc2 i..d 3

With a large advantage for Black, T.Avontuur-G.Cools, Brasschaat 2006.

C2) 5 dxcs The most critical move.

s . .. d4 6 tL'lbs tL'lc6!

7 e3 Alternatively, both 7 i..g5? a6 (Row­

son) and 7 i..f4 e5 8 i..g5 i..e7 9 i.xf6 i.xf6 10 lL'ld6+ 'it>f8 (Davies) are very good for Black, as is 7 e4? a6 8 tLla3 e5 9 i..d3 tL'ld7! (Lechtynsky) .

1 6 2

7 .. . es 8 exd4 tLlxd4 Black stands well in the centre. In­

deed, he will emerge with some advan­tage should he be able to recapture on c5 with impunity . 9 i.gs

Another suggestion of Rowson's is 9 b4 b6! 10 i..g5 i.e7 leading to com­plex play, while Black has the initiative after 9 tL'lxd4 exd4 10 i..d3 i.xc5 1 1 tL'le2 0-0 12 0-0 tL'ld7 13 tLlg3 (V.Lainburg­M.Konopka, Augsburg 1997) 13 . . . 'ii'h4, intending . . . f5 . 9 . . . i.xcs!? 10 i..xf6

White preferred 10 tLlxd4 in D.Nestorovic-N.Misailovic, Budva 2003, but now 10 . . . i.xd4 1 1 tLle2 h6 12 tL'lxd4 hxg5 13 lt:lb5 �xd1 + 14 .l:hd1 <Ji;e7 fa­vours Black. 10 .. . 'ii'xf6 11 lt:lc7+ �f8 12 tLlxa8

White's knight raid has won a rook, but all his other pieces are on their starting squares and his king is very exposed. Black enjoys very dangerous compensation. 12 ... i..b4+

12 . . . �g7! ? is also interesting, as after 13 i..d3 'iWe7! with the idea of . . . i..f5,

Black has a strong attack. 13 �f2

13 . . . e4! Suggested by Lechtynsky and a

very strong move. Now 14 'i'cl e3+ 15 'i'xe3 tt:lf5 should just be winning for Black, as I don't see a defence to the threats of . . . .ic5 and . . . 'i'xb2, and no better is 14 g3 tt:lxf3! 15 tt:lxf3 .ic5+ 16 �e1 'i'xb2 . Relatively best is probably 14 ke2! tt:lf5 15 fxe4 tt:lg3+ 16 tt:lf3 tt:lxe4+ 17 �g1 .ig4 when Black re­mains a rook down, but White's sur­vival task is immensely difficult, if not impossible. White's king position, lack of dark square control and poor coor-

Offb e a t 4 th M o ve A lt e rn a t ives

dination continue to cause him serious difficulties. Indeed, Black is surely for preference here !

Conclusion

Only 4 f3 can trouble Black in any way, since both 4 g4 and 4 h4 are weakening and obscure at best. After 4 f3 Black must react vigorously with 4 . . . c5! and fight for the centre. White can then suf­fer from a lack of development and dark square weaknesses; key factors brought to light by our critical main line in which Black obtains a very strong attack for the rook.

1 6 3

C h a pte r T h i rtee n I The F ia n c h etto Syste m

1 d4 l'2Jf6 2 C4 g6 3 g3

The fianchetto system sees White not displaying any early ambition in the centre, preferring to develop qui­etly. However, this solid choice should not be underestimated as Black can often be left without any counterplay. Black has a little time at his disposal to disturb the balance, before White cas­tles, and he must aim to exploit it.

We should note that White can also play g3 after 3 tZ:lc3 d5 4 cxd5 (then 4 . . . tZ:lxd5 5 g3 �g7 6 �g2 transposes to

1 6 4

a position we will consider after 3 g3 j_g7 4 j_g2 d5 5 cxd5 tZ:lxd5 6 tZ:lc3), but in that sequence 4 g3? ! is inaccurate and can be met by the immediate 4 . . . dxc4 ! .

White must now act immediately to recover the pawn, but Black gains excel­lent counterplay with 5 \\1Va4+ (5 �g2 c6 allows Black to retain the extra pawn) 5 . . . tZ:lfd7! (this is always the correct re­sponse to the check from a4) 6 j_g2 (the 6 j_f4 of K.Schulz-L.Ftacnik, Altensteig 1987, is well met by 6 . . . tZ:lc6 7 \\1Vxc4 tZ:lb6

8 \\1Vd3 i.g7 9 e3 0-0 intending . . . eS and/or . . . tt:ldS) 6 . . . kg7

and now: a) 7 e3 0-0 8 \\1Vxc4 cS ! (Black exploits

the exposed position of the white queen to wrest the initiative) 9 tt:lge2 (or 9 tt:lf3 cxd4 10 tt:lxd4 tt:leS 1 1 \\1Ve2 tt:lbc6 12 tt:lxc6, as in A.Kotov-P.Keres, Zurich Candidates 1953, and now the correct recapture is 12 . . . bxc6! 13 0-0 \\1Vb6 with a serious initiative, in view of the coming . . . ka6; while 9 dxcS tt:leS! sees Black jump into d3) 9 . . . tt:lc6 10 0-0 cxd4 1 1 tt:lxd4 (1 1 exd4 tt:lb6 1 2 \\1Vb3 kg4 13 i.e3 \\1Vd7 is again good for Black) 1 1 . . . tt:lxd4 12 exd4 tt:lb6 13 \\1Ve2 ke6 14 .l:!.d1 \\1Vd7 15 kgS h6 16 i.e3 tt:ldS 17 .l:!.acl .l:!.ac8 18 \\1Vd2 'it>h7 19 b3 .l:!.fd8 and Black was ob­viously better in H.Steiner-P.Trifunovic, Mar del Plata 1953.

b) 7 tt:lf3 tt:lc6 8 \\1Vxc4 (8 ke3 tt:lb6 9 \\1Vc2 kfS 10 \\1Vd2 0-0 1 1 0-0 \\1Vd7 12 .l:!.fd1 .l:!.fd8 13 .l:!.acl .l:!.ab8 was great for Black in E .Ragozin-D.Bronstein, Mos­cow 1947) 8 . . . tt:lb6 9 \\1Vb3 0-0 10 e3 (or 10 dS tt:laS 11 \\1Vb4 tt:lac4 12 0-0 aS ! 13 \\lVbS a4 and, with . . . a3 threatened, White is in trouble) 10 . . . i.e6 1 1 \\1Vd1

Th e F ia n c h e tto Sys tem

( 1 1 \\1Vc2 tt:lb4 threatens . . . i.c4) 1 1 . . .kc4 12 tt:ld2 kd3 13 a3 i.xd4! and Black had snatched a pawn in P.Machacek­V.Jansa, Havirov 1970. 3 . . . i.g7 4 kg2 d5

Now White has a major choice be­tween exchanging on dS and allowing Black to take on c4.

A: 5 tt:lf3 B: 5 cxd5

In both cases we will focus on a modern approach for Black: delaying castling in favour of more pressing matters. This approach, first brought to my attention by Vlastimil Jansa's excel­lent Dynamics of Chess Strategy, is en­tirely justified by the relative slowness of White's set-up: he doesn't seize the centre at once, leaves the c4-pawn un­protected and is not trying to mount a kingside attack. White's play may be solid and sound, but it's also somewhat loose; a factor we will exploit with some concrete play!

Finally, White also has the rare 5

1 6 5

Play t h e G r ii nfe /d

lbc3. This should be met in similar vein to 5 l2Jf3: 5 . . . dxc4 6 'iVa4+ l2Jfd7 7 'iVxc4 (or 7 l2Jf3 l2Jc6 8 0-0 l2Jb6 9 'iVc2 0-0 10 .if4 .ig4 with a clear advantage, E .Guseinov-Ni Hua, Artek 1999, while 7 e3 0-0 8 'iVxc4 e5 9 l2Jge2 exd4 10 exd4 l2Jb6 gives Black an edge) 7 . . . l2Jb6 8 �3 (8 'iVd3 'iVxd4 9 'iVxd4 .ixd4 10 l2Jf3 .ig7 1 1 l2Jb5 l2Ja6 holds on to the extra pawn) 8 . . . 0-0 9 l2Jf3 il..e6 10 'iVc2 l2Jc6 1 1 0-0 l2Jxd4 12 lZlxd4 'iVxd4 13 .i1Lxb7 l::!.ab8 14 .ig2 l2Jc4 15 l::!.d1 l2Ja3! 16 'iVd2 'iVxd2 17 l::!.xd2 l2Jc4 18 l::!.c2 l2Jxb2 19 .ixb2 l::!.xb2 20 l::!.xb2 il..xc3 and Black again enjoyed an extra pawn in B.Dantas­E.Tsuboi, Brazil 1998.

A) 5 l2Jf3 dxc4!

True to our strategy, we capture this pawn. White can now either try to recover it immediately or ignore it for the sake of development:

A1: 6 l2Ja3 A2: 6 'iVa4+ A3 : 6 0-0

1 6 6

Instead 6 l2Jbd2 b5 7 a4 c6 8 l2Je5 l2Jd5 9 0-0 0-0 10 l2Je4 a6 1 1 l2Jc5 (Mik­halchishin and Beliavsky's suggestion, whereas 1 1 b3? .if5 ! 12 bxc4 il..xe4 13 .ixe4 l2Jc3 14 'iVd3 'iVxd4 was a total disaster for White in M.Todorcevic­Z.Kozul, Cetinje 1991) l l . . .'iVd6 is again fine for Black, who has retained his extra pawn.

A1) 6 l2Ja3 The most common move in the po­

sition with both sides castled, but Black can directly exploit the difference: 6 . . . cs !

Immediately challenging White's central presence. Black intends to fol­low up with . . . l2Jc6, when the impor­tance of not wasting time on castling becomes obvious. 7 0-0

Relatively best. Instead: a) 7 l2Jxc4?! l2Jc6 ! 8 'iVa4 (8 dxc5? !

'iVxd1 + 9 'it>xd1 .ie6 10 l2Jfd2 - or 10 l2Jfe5 0-0-0+ 1 1 il..d2 l2Jg4 12 l2Jxg4 .ixc4 with fantastic compensation due to the many open lines - 10 . . . l2Jd7 1 1 .ixc6 bxc6 12 l2Ja5 l2Jxc5 13 f3 l::!.c8 14 'it>c2 0-0

15 .l::!.e1 was seen in A.Karpov­V.Tkachiev, Bastia (rapid) 2002, and now 15 . . . l2Ja6 leaves White in trouble as Black intends . . . l2Jb4+ and . . . .l::!.fd8) 8 . . . cxd4 9 l2Jfe5 (or 9 l2Jxd4 \\i'xd4 10 il..xc6+ bxc6! 11 \\i'xc6+ \\i'd7 12 \\i'xa8 0-0 13 l2Ja5 \\i'b5 14 \\i'xa7 il..a6 which was good for Black in A.Maric-A.Brustman, Yerevan Olympiad 1996) 9 . . . it..d7 10 l2Jxc6 bxc6 1 1 ltJe5 ltJd5 12 \\i'xd4 \\i'c7 13 il..f4 g5 14 it..xd5 \\i'a5+ 15 it..d2 \\i'xd5 16 \\i'xd5 cxd5 17 l2Jxd7 'it>xd7 18 il..xg5 il..xb2 was L.Debnar-D.Navara, Par­dubice 2001; Black is better in view of his central superiority and more active pieces.

b) 7 \\i'a4+ l2Jc6 8 l2Je5 (8 dxc5 c3 9 l2Jb5 l2Jd5! is good for Black, as pointed out by Jansa) 8 . . . it..d7 9 l2Jxc6 it..xc6 10 il..xc6+ bxc6 1 1 dxc5 c3 12 0-0 0-0 13 bxc3 (I .Balinov-A.Brustman, Aschach 1997) 13 . . . \\i'd5 intends . . . \\i'h5 when Black has good attacking chances, as the absence of the fianchettoed bishop weakens White's king position.

c) 7 dxc5? \\i'xd1 + 8 'it>xd1 l2Je4 (Jansa) is excellent for Black.

7 .. . l2Jc6 8 dxcs

Th e F ia n c h e tto Sys t e m

Critical; White prepares t o recap­ture on c4, whereas 8 \\i'a4 0-0 9 dxc5 c3 gives Black the initiative. 8 .. . \\i'as !

Thus Black sets out t o recover the c5-pawn. 9 ctJd4

This hits c6, but Black can ignore it. Instead 9 l2Jxc4 \\i'xc5 10 b3 \\i'h5 intend­ing . . . it..h3 is equal (Savon) . 9 ... 0-0!

10 il..e3 ! ? The most dangerous move. White

indirectly protects c5 and is set to emerge a pawn up ahead, but Black crucially retains some dynamic re­sources. Otherwise 10 l2Jxc6 bxc6 1 1 l2Jxc4 ( 1 1 it..xc6 .l::!.b8 ! 1 2 it..f4 \\i'xc5! 1 3 il..xb8 \\i'xc6 1 4 il..xa7 il..h3 15 f3 \\i'a6 ! 1 6 it..f2 l2Jg4! leaves Black much better) 1 1 . . .\\i'xc5 12 b3 (now ka3 appears somewhat awkward, but Black should remember that the c6-pawn is often sac-rificed in this line) 12 . . . \\i'h5 (intending a primitive attack with . . . it..h3 and . . . l2Jg4) 13 it..b2 (13 kxc6?! il..g4! is promising for Black - Jansa) 13 . . . it..h3 (White must now seek a way to extinguish Black's

1 6 7

Play t h e G r iJ nfe /d

initiative) 14 i.xh3 'iVxh3 15 i.x£6 i.xf6 speeding up his development and thus 1 6 l:!.cl l:!.ad8 1 7 'iVc2 l:!.dS 1 8 'iVe4 l:!.fd8 seizing the initiative. 19 'iVg2 'iVe6 led to equality in V.Dydyshko-V.Jansa, Czech League 2002. 10 ... lt:Jg4 11 lt:Jxc6 bxc6 12 i.d4

12 lt:Jxc4 'iVa6 13 l:!.cl i.e6 14 'iVc2 oc­curred in M.Gagunashvili-D.Navara, Athens 2001 , when Black should have played 14 . . . 'iVxa2 with good prospects. 12 . .. l:!.d8 13 e3 ke6

14 l:!.c1 Instead 14 kxc6 l:!.ac8 1S 'iVa4 (or 1 S

i.g2 l:!.xcS with an edge, a s Jansa points out) 1S . . . 'iVxa4 16 �xa4 i.xd4 17 exd4 l:!.xd4 (Marin) is good for Black. Thus he preferred the text move in M.Marin­V.Jansa, Andorra 2000, and now Black should play 14 . . . lt:JeS 1S 'iVe2 l:!.ab8 16 �xeS ( 16 l:!.c2 i.fS is also good for Black) 16 . . . �xeS 17 lt:Jxc4 �xc4 18 'iVxc4 l:!.xb2 1 9 �xc6 'iVxa2 (Jansa) with the advantage.

A2) 6 'iVa4+ lt:Jfd7! A typical reaction to the check.

Black will now gain time against the white queen after . . . lt:Jc6 with . . . lt:Jb6,

1 6 8

7 0-0 Other possibilities: a) 7 'iVxc4 lt:Jb6 8 'iVc2? ! (instead 8

'iVcS?! lt:Ja6 9 'iVa3 i.xd4 10 lt:Jxd4 'iVxd4 1 1 0-0 c6 12 b3 f6 13 e3 'iVcS 14 'iVb2 'iVhS was very good for Black in A.Minasian-P.Nikolic, European Ch., Kusadasi 2006, and even after 8 'iVd3

. lt:Jc6 9 lt:Jc3 0-0 10 �f4 i.fS 1 1 'iVbs �g4 Black's ongoing pressure is very awk­ward) 8 . . . lt:Jc6 9 e3 (or 9 0-0 �g4 10 e3 0-0 1 1 h3 i.e6 12 l:!.d1 lt:Jb4 13 'iVcS aS 14 b3, J.Jirka-S.Ganguly, Nakhchivan 2003, and now 14 . . . 'iVd6 intends . . . a4; the a1-rook is a serious liability for White) 9 . . . 0-0 10 0-0 �fS 1 1 e4 �g4 12 dS kxf3 13 dxc6 �xg2 14 'i!>xg2 bxc6 1S 'iVxc6 'iVd6 16 'iVc2 l:!.fd8 17 lt:Jc3 'iVd3 1 8 'iVxd3 l:!.xd3 with a strong initiative, E.Kahn­J .Horvath, Budapest 199S.

b) 7 lt:Jbd2 cS 8 dxcS (8 lt:Jxc4 cxd4 9 �gS 0-0 10 'iVa3 lt:Jc6 1 1 l:rd1 h6 was good for Black in V.Poddubnyi­A.Yermolinsky, St. Petersburg 1987) 8 . . . lt:Jc6 9 'iVxc4 'iVaS 10 0-0 lt:JxcS 1 1 lt:Je4 lt:Jxe4 12 'iVxe4 0-0 was agreed drawn

here in O.Romanishin-M.Palac, Ljubl­jana 1997; Black clearly has no prob­lems.

c) 7 e3 0-0 8 �xc4 QJc6 9 0-0 eS 10 dxeS QJdxeS 1 1 QJxeS QJxeS 12 �c2 ( 12 �3 c6 13 QJc3 �6 14 �c2 .ii.e6 1S b3, T.Kottmair-Z.Szabo, Balatonlelle 2004, and after 1S . . . �cS the c3-knight is very unstable and Black much better) 12 . . . c6 13 QJc3 �d3 14 �xd3 QJxd3 1S .l:!.d1 QJxcl 16 .l:!.axcl Si.g4 17 f3 .ie6 with an obvious long-lasting edge, L.Reutova­V.Grebionkin, Ekaterinburg 1996. 7 . . . QJc6

Threatening to keep the pawn with . . . QJb6. 8 �XC4 QJb6 9 �C2 0-0

White must now attend to the mat­ter of the d4-pawn. 10 .i:!.d1

Instead 10 h3 QJxd4 11 QJxd4 �xd4 12 �xc7 QJdS 13 �aS .ie6 left Black with an obvious developmental lead in S.Mariotti-M.Stean, Las Palmas 1978, while 10 QJc3 QJxd4 1 1 QJxd4 �xd4 12 QJbS �c4 (as in M.Bezold-M.Konopka, Munster 1992) is a typical equalizing sequence in this line.

Th e Fia n ch et to Sys t e m

1o .. . .ifs This does not waste time as the e4-

advance is loosening. 11 e4

Black seized the initiative after 1 1 �3 aS 12 QJa3 a4 13 �c3 eS 1 4 Si.gS exd4 1S �d2 �d7 16 .ii.h6 .l:!.fd8 1 7 .ixg7 'it>xg7 18 QJbS d 3 i n V.Smyslov­M.Stean, Teesside 197S. 11 ... .ig4 12 d S .ixf3 13 .ixf3 QJd4 14 �d3

Black enjoys a promising position with a strong centralized knight, C.Lamoureux-N.Sulava, Lido Estensi 2002.

A3) 6 0-0 A more ambitious approach; White

disregards the c4-pawn for the time being. 6 . . . c6!

With this move (instead of the stan­dard reply 6 . . . 0-0), Black prepares to retain the pawn on c4 or at least to greatly inconvenience White with its recovery. Once again we will see the advantage of concrete opening play in this line.

1 6 9

Play t h e G r ii nfe ld

7 tLlc3 A gambit continuation which, how­

ever, is hardly promising. Instead: a) 7 tLla3 bS simply leaves the a3-

knight offside: 8 tLleS tt:JdS 9 tLlc2 (Black was also somewhat better after 9 e4 tt:Jc7 10 i.e3 0-0 11 'i¥c2 cS 12 l:tad1 cxd4 13 i.xd4 tLle6 14 i.xa7 'i:Yc7 1S i.xb8 l:txb8 16 tLlf3 i.a6 17 l:tfe1 b4 in P.Li.ichtmeier-

12 l:tcl .idS leaves Black with the more active pieces, A.Petrosian-l.lbragimov, Kazan 1997.

d) 7 tLleS is powerfully met by 7 . . . tt:Jg4!

which destabilizes White's central control. White is now worse in all lines: for example, 8 tt:J£3 0-0 9 tt:Jc3 eS 10 dxeS 'i:Yxd1 1 1 l:txd1 tLlxeS 12 tLlxeS ii.xeS as

M.lvanov, Baunatal 2002) 9 . . . .tb7 10 e4 in A.Raetsky-I.Glek, Zurich 2001, or 8 tLlb6 1 1 'i¥f3 0-0 12 tLlg4 'i¥e8 13 tt:Jh6+ , tLlxc4 'i¥xd4 9 'i¥c2 0-0 10 tt:Jba3 'It>h8 14 eS tLl8d7 1S l:te1 tLldS and Black (V.Bologan-A.Volokitin, Sarajevo 200S), was a healthy pawn up in A.Kharlov- and now Ftacnik notes that 10 . . . 'i:YcS 1 1 J .Horvath, Vienna 1996. ii.f4 (11 h3 tt:Jf6 12 e4 'i:YhS is similar)

b) 7 'i¥c2 bS 8 tLlbd2 (8 tLleS tLldS) 1 1 . . .tLla6 is good for Black. 8 . . . 0-0 9 b3 cxb3 10 'i:Yxb3 ii.b7 11 a4 a6 1 . . . 0-0 12 i.a3 tLlbd7 13 e4 tLlb6 14 aS tt:Ja4 1S l:tacl l:tc8 16 l:tfe1 cS 17 dxcS 'i:YxaS 18 c6 l:txc6 19 l:txc6 ii.xc6 with a big advan­tage for Black, Su.Polgar-V.Anand, Monaco (rapid) 1994.

c) 7 a4 0-0 8 tLla3 (or 8 tLlc3 tLldS 9 e4 tt:Jb4 10 ii.e3 ii.g4 11 h3 ii.xf3 12 ii.xf3 cS 13 dS tt:Jd3 14 'i:Yd2, A.Matnadze­E.Zaiac, Antalya 2002, and now 14 . . . tLla6 with some advantage) 8 . . . tt:Ja6! (Black develops freely and prepares . . . cS) 9 tt:Jxc4 ii.e6 10 b3 cS 11 ii.b2 l:tc8

1 7 0

Black can finally afford to castle! 8 e4

Black was obviously better after 8 h3?! b5! 9 'Lle5 a6! 10 e4 (10 'Llxc6 'Llxc6 11 .txc6 .txh3 is a clear extra pawn, and 10 a4 b4 1 1 'Lla2 aS 12 'Llxc4 .te6 13 b3 .td5 - Jansa - gives Black the initiative) 10 . . . .tb7 11 l1e1 (or 11 .te3 'Llbd7 12 'Llxd7 'Llxd7 13 e5 e6 14 'Lle4 c5 15 'Llf6+ .txf6 16 .txb7 l1b8 17 exf6 l1xb7 18 dxc5 'Llxf6 with a clear advantage) 1 1 . . .'Llbd7 12 'Llxd7 'Llxd7! 13 e5 e6 14 'Lle4 c5! ! (Anand's powerful play deserves admi­ration and is fully in the spirit of the Gri.infeld) 15 'Llf6+ (15 'Llxc5? 'Llxc5 16 dxc5 'iYxd1 17 l1xd1 .txg2 18 'it>xg2 .txe5 wins for Black) 15 . . . .txf6 16 .txb7 l1a7 in O.Romanishin-V.Anand, 3rd match­game, New York 1994.

Likewise, 8 'Lle5 .te6 9 e4 'Llfd7 10 'Llxd7 ( 10 f4 'Llxe5 1 1 dxe5 'iYxd1 12 l1xd1 'Lla6 13 .te3 f6 14 exf6 exf6 15 h3 l1fe8 16 'it>h2 .tf8 17 .td4 f5 was more of the same stuff in D.Kopec-K.Sasikiran, Hampstead 1998) 10 . . . 'iYxd7 1 1 .te3 'Lla6 12 'iYd2 l1fd8 13 l1ad1 'Llb4 14 'Lle2 'Lld3 gives Black a huge advantage, J .Richardson-A.Whiteley, London 1996. 8 .. . .tg4 9 h3

9 .te3 'Llbd7 (threatening . . . e5) 10 e5 'Lld5 (Jansa) is much better for Black. g . . . .txf3 10 .txf3 'Lla6 11 .te3 'Llc7 ! 12 a4 eS !

A well-timed central break. Now 13 dxe5 'Lld7 regains the e5-pawn with a big plus, but even after 13 .te2 'Llg4! 14 hxg4 exd4 15 .txd4 .txd4 16 .txc4 'iYf6 17 l1cl l1ae8 Black was clearly for pref­erence in F .Vallejo Pons-Z.Almasi, Pamplona 2000.

Th e Fia n c h et to Sys t e m

B) s cxds The most challenging move, elimi­

nating any possibility of . . . dxc4. s ... 'Llxds

White now has three different ap­proaches to the position:

Firstly, he can play as in a normal Exchange variation with 6 'Llc3, practi­cally forcing the exchange on c3. How­ever, this is the least troublesome line for Black, as he can then develop quickly and comfortably. White counts on a strong central set-up (c3, d4, e3) and pressure on the queenside, but this plan is not dangerous and Black can neutralize it without much effort.

1 7 1

Play t h e G r ii nfe ld

Alternatively, White can play the direct and aggressive 6 e4, exploiting the fact that Black cannot exchange knights on c3 . He will then follow up with tDe2, avoiding any pin with . . . .ig4. This line is quite threatening and requires precise handling by Black, but the resulting positions are very double-edged.

Finally, White can opt for the most popular move, 6 tiJf3, with which he aims first to complete his development and only then to advance in the centre. In my view, this approach is the most difficult for Black to meet, as it com­bines ambition with solidity. White often ends up with a space advantage in the centre and piece pressure on Black's position, so I propose to meet this system head-on with a dynamic central break advocated also by Jansa. Here the importance of delaying cas­tling, prevalent throughout the 5 tDf3 dxc4 variation, is again a key feature .

On to the theory!

81: 6 t2Jc3 82: 6 e4 83: 6 t2Jf3

81) 6 t2Jc3 ttJxc3 7 bxc3 c5 Black's main task in this position is

to counter the pressure applied by the g2-bishop. A second important aspect of the position is the weakness of the c4-square in White's camp; a weakness caused by the development of the light-squared bishop on g2. While playing to occupy c4 is standard in

1 72

several lines of the Exchange Gri.infeld, here it is even more effective than usual : Black can set about occupying this square with moves such as .. .'Ylfa5-a6, . . . i.e6, and . . . ttJc6-a5.

8 e3 Preparing tDe2. A different set-up

occurs with 8 tiJf3 and now 8 . . . ttJc6 9 .ie3 (no other move fits in well with White's chosen set-up: for example, 9 d5 t2Ja5 is already better for Black, as pointed out by Rowson, and 9 0-0? ! is a dubious pawn sacrifice since 9 . . . cxd4 10 cxd4 tDxd4 1 1 tDxd4 'i¥xd4 12 'i¥xd4 .ixd4 13 .:.b1 i.b6 is good for Black) 9 . . . 0-0 10 0-0 cxd4 1 1 t2Jxd4

(a typical recapture in this line, re­leasing the g2-bishop at the cost of some structural damage; 1 1 cxd4 ii.e6 12 'Yi'd2 �d5 13 l:i.ab1, as in A.Beliavsky­C.Balogh, Austrian League 2005, is just equal after 13 . . . 'Yi'd7) 1 1 . . .'Lla5 12 'Yi'cl ! (trying to evacuate the bishop from e3 in a natural and productive way; both 12 l:i.b1 'Llc4 13 �cl e5 14 'Llb5 'Yi'a5 and 12 'Yi'd3 a6 13 l:i.acl 'Yi'c7 14 c4 ii.d7 15 'Llb3 ii.c6 are equal) 12 . . . ii.d7! (simply aiming to pile up on the c-file) 13 �h6 l:i.c8 14 �xg7 'it>xg7 15 l:i.d1 'Yi'c7 16 'Yi'a3 l:i.fe8 17 l:i.acl occurred in J.Hajtun­I.Polgar, Hungarian Ch. 1966, when 17 . . . b5 would have left Black with an obvious initiative. 8 ... '2:lc6 9 'Lle2 �d7 10 o-o l:!.c8

I believe this to be Black's most reli­able set-up. Note that Black has imme­diately opted for this regrouping of his queenside pieces rather than castle: time is important in chess and, as we have already seen, that is especially so in the Fianchetto Gri.infeld ! The point of Black's move order is to be ready to respond to White's plans, as will be­come apparent.

Th e Fia n c h e tto Sys tem

11 dxcs Critical, and White often accepts a

weak c-pawn in return for bringing his knight to d4, but we will now witness an instructive method of play for Black. Thus White might try something else -not that the alternatives are especially impressive:

a) 1 1 ii.d2 0-0 12 l:i.cl 'Lla5 13 'Llf4 ii.c6 14 �h3 l:i.b8 15 c4 e5 16 'Yi'e1 'Llxc4 ! 17 'Lle6 'Llxd2! gave Black a decisive material advantage in Gilb.Garcia­V.Smyslov, Havana 1962.

b) 11 a4 'Lla5 12 e4 0-0 13 d5 e6 14 l:i.a2 exd5 15 exd5 l:i.e8 saw Black have the better of it in S.Gligoric­V.Korchnoi, Yugoslavia-USSR match 1967. Indeed, Gligoric was probably impressed by Black's play in this game, as the later 12 d5 0-0 13 'Yi'c2 'iVb6 14 �a3 'Yi'a6 15 l:i.fb1 l:i.fd8 favoured Black, who had obtained control of c4 in M.Cuellar-S.Gligoric, Sousse Interzonal 1967.

c) 11 'Llf4 0-0 12 l:i.b1 cxd4 13 cxd4 (13 l:i.xb7 e5 14 'Lld5 ii.e6! is given as very good for Black by Adorjan and Dory) 13 . . . b6 14 �d2 e5! 15 dxe5 'Llxe5 16 �b4 l:i.e8 17 'Lld5 a5 ! was another successful outing for Black in Iskov­V.Jansa, Svendborg 1981 .

d) 1 1 ii.a3 'Yi'a5 ! 12 'iVb3 'Yi'a6 13 'Llf4 b6 14 l:i.fe1 'Lla5 15 'iYh1 'Llc4 was typi­cally great for Black in E.Geller­D.Bronstein, Amsterdam Candidates 1956.

e) 1 1 l:i.b1 b6 12 dxc5 bxc5 13 c4 'Lla5 14 'Yi'c2 0-0 15 �b2 �f5 with easy equal­ity, V.Kirillov-A.Suetin, USSR 1961 . 11 ... '2Ja s 12 .ta3 .tbs!

1 73

Play t h e G r ii nfe ld

Ensuring that the knight will never reach d4. 13 l::te1

Instead 13 'iVxd8+ l:txd8 14 l:tfe1 Axe2 15 l::txe2 ctJc4 left Black with a large advantage in Hausner-I.Stohl, Zlin 1995, while Stohl notes both 13 Af3? l2Jc4, intending . . . l2Je5 with some advan­tage, and 13 l2Jd4 Axfl 14 'iVxfl 0-0 15 �5 'iVc7 when Black i s again better. 13 ... 'iVxd1 14 J::taxd1 l2Jc4

Now 15 Acl is met by 15 . . . Ac6!, in­tending 16 e4 Jk.a4 with a clear plus, and 15 i.b4 aS 16 Axb7 l:tc7 17 Ag2 axb4 18 cxb4 0-0 leads to a complicated position in which Black is certainly not worse, having an extra piece!

82) 6 e4 '2Jb6 7 tt:Je2 es ! I much prefer this central counter to

the other popular option, 7 . . . c5, as it makes more sense. White will push d4-d5 in either case, after which Black will challenge the d5-pawn and an ex­change will take place on d5. Following that, the outcome will be decided pri­marily by the viability of the passed d5-pawn.

1 74

Having played . . . e5, Black will at some point be able to blunt the g2-bishop with . . . f5 and . . . e4, whereas in the case of the 7 . . . c5 line the scope of this bishop is greatly enhanced. I feel that this factor alone should be enough to tilt the balance in favour of 7 . . . e5. 8 ds c6 9 o-o o-o

White' s other problem in this varia­tion is what to do with his e2-knight: the e5-pawn greatly restricts this piece and moving it to c3 leaves the other knight without a good option. Mean­while, after an exchange on d5, Black's main positional aim is to blockade the passed d5-pawn with a knight, most

usually with . . . 'Llb6-c4-d6. He then has nothing to fear and can proceed ac­tively with a pawn advance on the kingside and in the centre. An unhur­ried approach to the position by White would, of course, enable Black to ac­complish this aim. Thus White must play aggressively and apply all his might to advancing his passed pawn; a strategy which dominates the resulting variations. 10 'Llbc3

Alternatively, 10 'Llec3 cxdS 11 exdS ii.fS 12 a4 'iVd7 13 l:te1 'Lla6 sees Black clearly untroubled: 14 aS (or 14 'Lla3 l:tfc8 1S �e3, B .Jobava-L.Valdes, Ubeda 2001, and now 1S . . . 'LlcS with the initia­tive) 14 . . . 'Llc4 1S 'Lle4 �xe4 ! (establish­ing a dark square blockade is the top priority ! ) 16 l:txe4 l:tac8 17 l:te1 'LlcS 18 'Llc3 fS 19 ii.fl 'Lld6 20 ii.d2 e4 and in J .Fedorowicz-P.Wolff, New York 1998, Black had fully accomplished his plan and stood better.

Likewise, 10 'iVb3 'Lla6 11 ii.e3 cxdS 12 exdS ii.g4 13 'Llbc3 l:tc8 14 h3 ii.xe2 1S 'Llxe2 'LlcS 16 'iVc2 'Llcd7 was at least equal for Black in N.Edgzveradze­J.Rowson, London 1999; the long-term chances are on his side. 10 ... cxds 11 exds

1 1 'LlxdS 'Llc6 gives Black simple equality. 11 ... '2Ja6 !

The correct way to develop, eyeing the b4- and cS-squares as well as leav­ing the path of the c8-bishop un­blocked. 12 b3 !

Clearly the most dangerous move.

Th e Fia n c h e tto Sys t e m

White intends to support the advance of the dS-pawn with �a3. Note also that with this move White prevents the typical . . . 'Llc4-d6 manoeuvre. Instead, the immediate march of the d-pawn fails to convince: 12 d6 'LlcS 13 ii.e3 'Lle6 14 b3 fS ! 1S f4 (1S 'LldS f4 16 ii.xb6 axb6 17 'Lle7+ 'it>h8 18 'Llxc8 l:txc8 was much better for Black in L.Johannessen­J .Gustafsson, Hengelo 1999; Black will play . . . 'Lld4 at some point and the d6-pawn will drop) 1S . . . e4 and here a draw was agreed in J .Lipka-J.Banas, Slovakian Team Ch. 199S, but of course Black is for preference. 12 ... ts !

Black must meet White's aggression with some of his own! This move also clears a square for the soon-to-be­attacked rook.

13 .ta3 Instead 13 a4! ? aims to show that

the b6-knight is misplaced. Black should again play actively with 13 . . . f4! 14 aS (14 gxf4? exf4 1S �xf4? ii.xc3 16 'Llxc3 l:txf4 - Rowson - just wins for Black, while 14 f3 'Llb4 1S i.a3 'Ll6xdS! ? 16 'LlxdS 'LlxdS 17 �xf8 �xf8 i s a very

1 75

Play t h e G r u nfe /d

interesting exchange sacrifice; Black's prospects are surely greater, in view of both the coming . . . 'Lle3 and the passive white pieces) 14 . . . f3 ! 15 axb6 fxg2 which gave him the better chances in D.Rogozenko-L.Ftacnik, Hamburg 1998. 13 .. J:!f7 14 d6

The only efficient way of keeping the pawn protected since 14 'Llb5 e4 15 'Llec3 (or 15 .l:!.cl .i.h6 16 .l:!.c2 'i¥xd5 and Black has captured the important pawn) 15 . . . 'Llc7 16 .l:!.cl 'Llxb5 17 'Llxb5 a6 again sees White lose it for insuffi­cient compensation. 14 ... .te6 15 ct:Jds 'Llxds 16 .i.xds 'i:Ye8 !?

This idea was suggested by Rowson and seems quite promising to me: Black intends to increase the pressure against d6 with . . . .l:!.d8 and . . . .i.f8. At the moment White's pieces appear quite active, but no concrete blow is appar­ent and the e2-knight is still a problem. Overall, I'd rather play Black.

83) 6 'Llf3 'Llb6! As will become clear, this immediate and unprovoked retreat is correct for several reasons.

1 7 6

Black's next move will be . . . 'Llc6, at­tacking d4, and the omission of castling gains him an important tempo: after 7 'Llc3 'Llc6 8 d5, 8 . . . .txc3 comes with check and so d5 drops, while after 7 0-0 'Llc6 8 e3 e5 White cannot push d4-d5 . Thus Black manages to achieve . . . e5 without any difficulty, and can then choose whether to saddle White with an IQP (with . . . exd4) or to invite com­plications with the ambitious . . . e4.

White's choice here is actually a rather important one, although the two lines can easily transpose.

831: 7 CLJC3 832: 7 0-0

There's also 7 a4! ?, a rare but inter­esting idea, aiming to secure the b5-square for a knight. Black does best to 'succumb' with 7 . . . a5 ! when 8 'Llc3 'Llc6 9 0-0 (or 9 'Llb5 'Llb4 10 0-0 c6 1 1 'Llc3 0-0 and now 12 .i.f4 ii.g4 is equal, whereas 12 'Lla2?! 'Llxa2 13 .l:!.xa2 ii.g4 14 e3 e5 15 dxe5, as in E.Eliskases­B.Larsen, Beverwijk 1959, and now

1S . . . tt::ld7 is simply better for Black) 9 . . . 0-0 10 ii.f4 tt::lxd4 1 1 tt::lxd4 eS 12 tt::ldbS (R.Ruck-Y.Pelletier, Buk 1996) leads to an interesting and complicated position. This idea deserves more tests.

831) 7 tt::lc3 tt::lc6 8 e3 Actually 8 dS? ! doesn't lose a pawn,

but is still not to be recommended:

8 . . . ii.xc3+ 9 bxc3 'ti'xdS 10 i..f4 (oth­erwise White struggles for compensa­tion: 10 0-0 'ti'xd1 11 .l:i.xd1 f6 12 tt::ld4 tt::lxd4 13 cxd4 gS 14 f4 h6 1S .l:i.d3 i..fS 16 .l:i.c3 0-0-0 didn't give him much in S.Lindblom-L.Ogaard, Norwegian Ch. 197S, and 10 'ti'xdS tt::lxdS 11 tt::ld4 tt::lxd4 12 cxd4 c6 13 ii.h6 f6 14 h4 'it>f7 1S .l:i.cl .l:i.d8 16 ii.d2 ii.e6 17 hS gS left him struggling in C.Guimard-J .Rubinetti, Buenos Aires 1977) 10 . . . 'ti'xd1+ 1 1 .l:i.xd1 ii.e6 (White can now regain his pawn, but at some expense in terms of time) 12 ii.xc7 (12 tt::leS tt::laS 13 c4 f6 14 cS tt::la4 1S ii.d2 0-0-0 16 i..xaS .l:i.xd1+ 17 �xd1 fxeS was simply good for Black in J.Yrjola-J.Ehlvest, Kuopio 1992) 12 . . . tt::ldS 13 i..eS tt::lxeS 14 tt::lxeS 0-0-0 1S 0-0 f6 16 tt::ld3 tt::lc7 17 tt::lcS .idS and Black was

Th e F ia n c h e t to Sys t e m

clearly o n top i n V.Milanovic-D.Antic, Kecskemet 1990.

Note, too, that 8 0-0 will be exam­ined via 7 0-0 tt::lc6 8 tt::lc3, while 8 i..f4 sees White try to prevent . . . eS, but the bishop is not well placed on f4 and falls prey to tactics: 8 . . . 0-0 9 0-0 (instead 9 e3 ii.e6 10 0-0 h6 11 h4 'DdS 12 tt::lxdS ii.xdS 13 'ti'c2 fS 14 'ti'cS e6 1S .l:i.fd1 .l:i.f7 16 a3 aS 17 .l:i.acl was solid and logical play from White in O.Moisieev­A.Cherepkov, Moscow 1949, but Black has his fair share of the play after 17 . . . a4) 9 . . . tt::lxd4! 10 tt::lxd4 eS 1 1 tt::lc6 (White preferred 11 tt::ldbS exf4 12 'ti'xd8 .l:i.xd8 13 .l:i.fd1 in M.Chiburdanidze­A.Brustman, Yerevan Olympiad 1996, and now 13 . . . i..g4 would have been strong) 1 l . . .bxc6 12 'ti'xd8 :xd8 13 i..gS f6 (Y.Schwartz-R.Ruck, Zurich 200S) and Black already stands well; his pieces are well placed and the possibil­ity of . . . tt::lc4 can be awkward. s . . . es

White must now make a fundamen­tal decision: allowing an exchange of pawns on d4 is not very promising as Black is well placed to deal with the

1 7 7

Play th e G r ii nfe ld

IQP, but on the other hand, the d5-advance may well prove premature. 9 d5

The alternative is 9 0-0 exd4 (9 . . . .i.g4!? will be covered in line 'B322' ) 10 exd4 (10 'Llxd4 is just too compla­cent; for example, 10 . . . 'Llxd4 11 exd4 0-0 12 d5 .i.f5 13 .i.f4 'i¥d7 14 �e1 �fe8 15 'iYh3 .i.d3 16 �xe8+ �xe8 and Black had assumed the initiative in P.Masak­M.Smistik, Moravian Team Ch. 2003) 10 . . . 0-0, and now 11 .i.g5

1 1 . . .f6 (this has been condemned in some sources, but I believe it is best, whereas 1 1 . . . .i.f6?! is surely bad; Black will sorely miss his dark-squared bishop around his king and, for example, 12 d5 .i.xg5 13 'Llxg5 'i¥xg5 14 dxc6 bxc6 15 .i.xc6 1 6 .i.xa8 .i.xfl 17 'iYxfl �xa8 18 �d1 ! should be a t least somewhat bet­ter for White) 12 .i.f4 (or 12 'iYh3+ 'It>h8 13 .i.e3 'Lla5 14 'i¥d1 'Llac4 15 .i.cl .i.e6 and White hadn't got anywhere in Z.Jasnikowski-R.Palus, Zakopane 2000) 12 . . . .i.e6 13 �e1 .i.f7 14 �cl occurred in O.Cvitan-B .Jaracz, Nova Gorica 2004, when Black enjoys a very sound posi­tion after the natural 14 . . . 'i¥d7.

1 78

9 . . . 'Lle7 Other knight retreats are tried occa­

sionally, but throughout this chapter I will focus on this natural move. Black attacks d5, thereby forcing e3-e4 (which temporarily blocks the g2-bishop and concedes some dark squares); then plays . . . c6 and exchanges on d5, even­tually aiming for the blockade on d6. White can prevent the full realization of this plan, but only at the cost of his prized d-pawn. 10 e4 .i.g4

Exchanging the light-squared bishop on f3 is an important part of Black's plan. 11 h3

Alternatively, 1 1 0-0 (as in J.Murey­Z.Ribli, Reykjavik 1975) 1 1 . . .0-0 12 'iYh3 (or 12 a4 'Llc4) 12 . . . c6 13 .i.g5 h6 leads to equality, while 11 a4 c6 12 a5 'Llbc8 13 a6! ? (13 0-0 cxd5 14 exd5 was also interesting in C.McNab-M.Bakalarz, European Team Ch., Gothenburg 2005; Black should now play 14 . . . 0-0 with complex play) is critical, but ultimately not so effective: 13 . . . b6 14 'i¥a4 b5 15 'iYh3 0-0 16 h3 .i.x£3 17 .i.x£3 ex d5 18

CLJxd5 CLJxd5 19 'ti'xd5 'ti'xd5 20 exd5 l2Jd6 and Black had both realized his plan and retained good chances in V. Borovikov-S. Voi tsekhovsky, Barlinek 2001 . 11 . . . kxf3 12 'ti'xf3

Black easily equalized after 12 Jtxf3 c6 13 kg5 h6 14 kxe7 'ti'xe7 15 a4 aS 16 l:ta3 0-0 17 0-0 l:tfd8 in J .Horvath­P.Popovic, Gleisdorf 2000. 12 .. . c6 13 o-o cxds

14 exd s Recapturing with the knight is less

incisive, as an exchange of pieces makes the resulting d-pawn weaker: 14 CLJxd5 CLJbxd5 15 l:td1 (or 15 exd5 CLJf5 ! -the blockade on d6 is the top priority -16 'ti'a3 'ti'd6 17 b3 'ti'xa3 18 .i,xa3, as in N.Pert-A.Zhigalko, Turin Olympiad 2006, and now the simple and logical 18 . . . �d7 suffices for equality) 15 . . . 0-0 16 exd5 (16 .i,e3 was preferred in A.Adly-K.Van der Weide, Wijk aan Zee 2006, when 16 . . . f5 ! ? 17 exd5 'ti'd6 18 l:tacl l:tac8 19 l:txc8 l2Jxc8 i s unclear, but Black has maintained the blockade on

Th e Fia n ch e tto Sys t e m

and now simply 18 . . . l2Jd4 !? would have given Black a promising position. 14 ... l2Jfs

A critical moment. Black threatens . . . l2Jd6, after which his positional supe­riority will never be in doubt, as was shown in several games in the Sixties and Seventies (in similar positions) . Therefore, White has only one option: 15 d6! l:tb8 16 l:td1 0-0

We have reached a very sharp posi­tion. Black is planning to round up the d6-pawn after blocking the d-file with . . . CLJd4, and so again White must act. 17 lte3

Black is already better after 17 l2Je4 l2Jd4 18 l:txd4 exd4 19 kg5 f6 intending . . . l:tc8, while the 17 'ti'd3 of L.Van Wely­P.Leko, Wijk aan Zee 2001, should be met by 17 . . . 'ti'd7!? . 17 ... l2Jd4 18 kxd4 exd4 19 lLJbs l2Jc4 20 'ti'ds

The key test; 20 l:tab1 CLJxd6 21 CLJxd4 (Krasenkow) would be both equal and an admission of failure from White. 20 ... l2Jxb2

d6) 16 . . . 'ti'd6 17 ..id2 CLJf5 18 'ti'b3 was Now 21 l:td2 l2Ja4 22 CLJxd4 CLJc3 !? 23 seen in O.Cvitan-D.Navara, Pula 2003, iVf3 l2Jb5 24 CLJxb5 kxa1 shows that

1 79

Play t h e G r ii nfe ld

White cannot afford to play anything; 8 .. .l2Jxd4! 9 tt:Jxd4 'iVxd4 Black is better. Thus the ever-creative Levon Aronian preferred 21 .!:l.xd4 ! ? against Bologan (in Merida 2005), but ultimately this, too, is insufficient for advantage. Bologan accepted the of­fered material and had to suffer for it, but Krasenkow has pointed out the way for Black to fully extinguish White's pressure: 2l . . . a6 22 tt:Jc7 b5 23 .!:l.e1 tt:Jc4! (the start of the round-up! ) 24 .!:l.dd1 .!:l.b6 ! 25 'iVc5 .!:l.xd6 26 .!:l.xd6 'iVxd6 27 'iVxd6 CtJxd6 28 tt:Jxa6 with equality.

832) 7 o-o tt:Jc6

In this theoretically-important posi­tion, we have a final divide:

8321: 8 CtJc3 8322: 8 e3

8321) 8 tt:Jc3 This pawn sacrifice has been tried

several times, but unsuccessfully: Black is not greatly troubled by White' s slight initiative. Naturally, the proof is in the eating!

1 8 0

10 tt:Jbs The trickiest move, although still

nothing special. White now manages to regain his pawn, but only by exchang­ing several pieces.

Alternatively: a) 10 'iVxd4 .ixd4 11 CtJb5 .ie5 12 .if4

J.xf4 13 gxf4 �d8 14 .!:l.fd1 + has been a popular choice, but now White's pres­sure comes to a standstill : 14 . . . tt:Jd7 15 .!:l.d2 (15 a4 a6 does not impress; neither did 15 .!:l.acl c6 16 CtJd4 e6 17 b4 a6 18 a4 �e7 19 b5 axbS 20 axb5 cS 21 CtJb3 .!:l.d8 22 .!:l.a1 in L.Pantsulaia-A.Areshchenko, Yerevan 2004, and now with the simple 22 . . . .!:1.xa1 23 .!:l.xa1 c4 24 tLlaS tt:Jb6 25 tt:Jxb7 .!:l.d2 Black achieves a large advan­tage) 15 . . . c6 16 CtJd4 (or 16 .ih3 e6 17 CtJd6 �e7 18 .!:l.ad1, as in U.Rohde­R.Swinkels, German League 2005, when 18 . . . CtJf6 ends White's initiative) 16 . . . a5 17 .!:l.cl e6 18 .!:l.c3 �e7 19 a3 CtJf6 leaves White with insufficient compensation for the pawn, Z.Sturua-M.Makarov, Helsinki 1992.

b) 10 a4 'iVxd1 11 .!:l.xd1 c6 12 aS CtJc4 13 a6 0-0 14 axb7 .ixb7 15 .!:l.d7 .ic8 16

Th e Fia n c h e tto Sys tem

.!:l.xe7 .!:l.b8 was fine for Black in 14 ... .ixa4 15 .ixcs .!:l.b8 16 .if3 .ib3! N.Davies-E.Liss, Rishon Le Zion 1994. 1o ... iVcs ! 11 a4

And not 11 'iVb3? ! .id7! (Rowson) which is very good for Black.

11 ... Lt:lxa4! 12 'iVxa4 Instead 12 Lt:lxc7+ Wifxc7 13 'iVxa4+

.id7 14 .if4 .ixa4 15 .ixc7 .ic6 is much better for Black, as pointed out by Rowson, and 12 'iVd5 (V.Filippov­R.Kempinski, Swidnica 1999) 12 . . . 0-0! 13 .!:l.xa4 'iVxd5 14 .ixd5 .id7 15 .!:l.xa7 .ixb5 16 .!:l.xb7 c6 is fine for Black. 12 ... .id7 13 .ixb7

White can abandon any hope of an advantage in order just to regain his pawn with 13 Wila5 Wifxb5 14 Wifxc7 0-0 15 'iVxb7 'iVxb7 16 .ixb7 .!:l.ab8. 13 .. . .ixbs 14 .ie3

Black also equalized after 14 Wila5 .!:l.b8 15 .ie3 'iVf5 16 Wifxc7 (16 .if3 a6! 17 Wifxc7 0-0 18 'iVxe7 .ixb2 19 .!:l.ad1 'iVf6 ! ? saw Black successfully come out of the complications with an advantage in M.Grabarczyk-R.Kempinski, Polish Ch., Sopot 1997) 16 . . . 0-0 17 .!:l.a5 .ie5 ! in R.Skomorokhin-A.Mikhalchishin, Lviv 1996.

With a few accurate moves Black has equalized . 17 .ic6+ �f8 18 .!:l.xa7 .ie5 19 f4 .id6 20 .id4 f6 was agreed drawn in R.Ruck-P.Szekely, Hungarian Team Ch. 1997, while 17 .ixa7 .!:l.d8 18 .ic6+ �f8 19 .ic5 .if6 i s also equal.

8322) 8 e3 es

9 Lt:lc3 White can also exchange on e5, hop­

ing to profit from his central pawn ma­jority. Ulf Andersson has often played like this, but Black has nothing to fear, providing he doesn't just play any ran­dom move that comes to mind !

1 8 1

Play t h e G r ii nfe ld

R.Hi.ibner-G.Kasparov, Cologne 1992, instructively continued 9 dxe5 (9 Lt:lxe5 Lt:lxe5 10 dxe5 �xd1 1 1 .l:'txd1 i.xe5 is the same, of course) 9 . . . �xd1 10 .l:'txd1 Lt:lxe5 11 Lt:lxe5 ( 1 1 Lt:ld4 c6 was equal in L.Polugaevsky-Letelier, Mar del Plata 1962) 1 l . . .i.xe5

12 Lt:ld2 (plenty of other options have been tried; for example, 12 f4 i.g7 13 e4 i.g4 14 .l:'td3 i.e2 15 .l:'td2 i.a6 16 e5? ! was somewhat overambitious in Elarbi-M.Bakalarz, Turin Olympiad 2006, and was punished by 16 . . . f6 17 exf6 i.xf6 18 Lt:lc3 c5 19 Lt:ld5 Lt:lxd5 20 i.xd5 0-0-0 when Black was much bet­ter; 12 Lt:lc3 can be met by 12 . . . 0-0 or 12 . . . c6! ?, intending to keep the king in the centre - we are, after all, close to an endgame - and 13 e4 i.g4 14 f3 ii..e6 15 i.h6 .l:'td8 16 f4 .l:'txd1+ 17 .l:'txd1 i.xc3 18 bxc3 f5 was fine for Black in A.Garcia­J.Fernandez Aguado, Spanish Ch., Lleida 1991) 12 . . . c6 13 i2lf3 i.g7 14 Lt:ld4 i.g4! 15 f3 i.d7 16 i.d2 c5 17 Lt:lb3 Lt:la4 and White was already under serious pressure.

After 9 Lt:lc3, play has returned a well-known position (well-known to

1 8 2

us, that is!) , where the fully adequate 9 . . . exd4 was examined above (see the note to White's 9th move in line 'B31' ) . However, that is not Black's only op­tion and he can also try the following: 9 . . . i.g4!?

The pressure on d4 is now almost unbearable and White must act. 10 h3

The alternative 10 d5 i s well met by the counterblow 10 . . . e4! 11 dxc6 ( 1 1 Lt:lxe4 �xd5 12 �xd5 Lt:lxd5 13 h3 i.xf3 14 i.xf3 0-0-0 15 Lt:lc5 b6 16 Lt:lb3 aS 17 e4 Lt:ldb4 18 i.g5 a4 19 i.xd8 .l:'txd8 was a simul game of Kasparov' s as Black and a good illustration of Black's pos­sibilities) 1 l . . .�xd1 12 .l:'txd1 i.xf3! (the correct recapture, aiming to later ex­ploit the weakened light squares in White's position) 13 i.xf3 (13 cxb7 .l:'tb8 14 i.xf3 exf3 15 e4 Lt:lc4! sees Black wrest the initiative) 13 . . . exf3 14 Lt:lb5 0-0 15 cxb7 .l:'tab8 16 Lt:lxc7 (16 Lt:lxa7 c5 is similar) 16 . . . .l:'txb7 17 Lt:lb5 Lt:la4 when Black has good compensation for the pawn; White's queenside is under strong pressure and the £3-pawn causes his king problems. I feel that Black's

chances are preferable. 10 .. . .i.xf3 11 .i.xf3

Or 1 1 �xf3 exd4 12 exd4 0-0 13 d5 Cbe5 14 �dl �e8 15 a4 Cbbc4 16 b3 Cbd6 17 �a2 �f6 18 Cbe4 Cbxe4 19 .i.xe4 �adS and Black had centralized well without any problems in K.Langeweg-V.Hort, Wijk aan Zee 1975. 11 ... exd4 12 exd4 o-o

White is now forced to advance his d-pawn, thereby blocking his light­squared bishop and conceding the cen­tral dark squares to Black's knights . 13 ds Cbd4 14 .i.g2 �e8 !

15 h4 After 15 jLf4?! Cbc4 16 b3 g5 17 jLd2

Cbd6 18 �5 h6 19 h4 gxh4 20 jLxh6 .i.xh6 21 �xh6 Cb4f5 22 �£4 hxg3 23 fxg3 �£6 24 �acl �g7 25 �f3 �e7 White was struggling in K.Urban­V.Jansa, German League 2000; his pieces are badly coordinated, unlike Black's which are dangerously placed. 1S • • . Cbc4 16 b3 Cbd6

Establishing the standard blockade. L.J anjgava-A.Mikhalchishin, Pavlodar 1987, continued 17 i.g5 �d7 18 �cl h6

Th e Fi a n ch e t t o Sys t e m

19 .i.f4 �e7 2 0 �d2 ®h7 21 �eel �ae8 and with simple and logical moves Black had obtained the advantage.

Concl usion

The concept of delaying castling has dominated this chapter and is, in my opinion, an essential tool at Black's disposal to disrupt White's smooth development. Black should take on c4 if allowed and then either comfortably hold on to the extra pawn or, should White spend time recovering it, imme­diately strike in the centre with . . . c5.

After an exchange on d5, inviting a structure similar to the Exchange Varia­tion is unappetizing for White; Black is able to fully extinguish White's pressure with accurate play. The line 6 e4 Cbb6 7 Cbe2 is more dangerous: Black should meet it with . . . e5, rather than . . . c5, and aim to blunt White's fianchettoed bishop. Finally, in the main line with 6 Cb£3 Black again does best to delay cas­tling, preferring a quick . . . e5 . It seems to me that the crucial tempo gained by this enables Black to obtain good play.

1 8 3

C h a pte r Fo u rt e e n I An g l i n g fo r a

Sa m isc h : 3 f3

1 d 4 ctJf6 2 C4 g6 3 f3 There is nothing deep about this

move : White wants to play e4 and take play into a Samisch King's Indian. Black can of course reply with 3 . . . d5 and after the exchange there and 5 e4, retreat the knight to b6. Then a position similar to the 5 i!.d2 variation arises. However, I feel that f3 is a much more useful move than i!.d2 and that this difference gives White some aggressive and dangerous options.

Retreating the knight to b6 is by no means the end of the world for Black, but I would like to suggest in this book a very different, visually stunning, and apparently quite sound way of meeting this dangerous system: 3 . . . es!?

The originator of this outrageous idea is, I believe, the great Gri.infeld guru Andras Adorjan. It came to prominence when Peter Leko em­ployed it in 1997 against Vladimir Kramnik, even succeeding in winning

1 84

the game. Unsurprisingly, therefore, much of the analysis in this chapter stems from Adorjan and Leko.

Black's idea is to strike at the dark squares, which have been weakened by 3 f3, and to exploit the fact that the f3-pawn significantly hinders White's natural development. In a sense, it feels like replying to 2 c3 in the Sicilian with 2 . . . d5, although here, of course, things are much more complicated. Still, I have great faith in this move! 4 dxes

Though the most natural, there are other moves. Practice has also seen:

a) 4 d5 is not a very good idea, I feel, as it allows Black to develop his dark­squared bishop actively on c5: 4 . . . e4 ! 5 Cbc3 i.c5 ! ? with promising play. Black even has other options, including 5 . . . d6! ? 6 i.g5 h6 7 i.xf6 (7 i.h4!? is per­haps better) 7 .. .'�xf6 8 fxe4 i.g7 9 Cbf3 0-0 with good dark square compensa­tion, and 5 . . . exf3 6 exf3 d6 7 'iVe2+ 'iVe7 8 'iVxe7+ 'lt>xe7 is a simple equalizer.

b) 4 e4 allows Black to develop quickly: 4 . . . exd4 5 'iVxd4 Cbc6 6 'iVd2 i.g7 7 Cbc3 0-0 8 b3 d6 9 i.b2 Cbe8 ! ? ( 9 . . . Cbd7, intending . . . Cbc5 and . . . f5, is also fine) 10 0-0-0 f5, with counterplay, is the standard reaction and a sound one at that, but even better might be 6 . . . i.c5 ! 7 Cbc3 0-0 and I feel that Black already has the upper hand.

c) 4 i.g5 h6 5 i.h4 Cbc6!

(hitting where it hurts most, namely on the dark squares, but Black can con­sider, too, both 5 . . . exd4 6 'iVxd4 i.g7 7 'iVe3+ 'lt>f8! 8 Cbc3 d6 9 'iVd2 i.e6 10 e4 Cbc6 when . . . Cbxe4 is threatened and Black has good play, and 5 . . . g5 6 i.f2

A ng l ing fo r a S d m is c h : 3 j3

Cbc6 7 d5 Cbe7 8 e4 d6 9 h4 g4, which led to complicated play in M.Luch­T.Studnicka, Czech League 2004) 6 d5 (instead 6 e3? is just bad in view of 6 . . . exd4 7 exd4 i.e7 8 Cbc3 d5!, and 6 dxe5 Cbxe5 7 'iVd4 i.b4+ 8 Cbc3 d6 does not really trouble Black) 6 . . . Cbd4 and :

cl) 7 e4 c5 8 Cbe2 d6 leaves Black with a well-placed knight on d4.

c2) 7 Cbc3 c6 ! ? with pressure against White's centre (but not 7 . . . i.c5? 8 g4! or 7 . . . c5? ! 8 d6! ) .

c3) The ambitious 7 g4 i s very risky since 7 . . . h5 ! 8 e3 (8 g5? Cbh7 wins the g5-pawn) 8 . . . hxg4 9 i.x£6 'iVxf6 10 exd4 i.b4+! (the best move, although 10 . . . exd4 11 'iVe2+ 'lt>d8 12 Cbd2 'iVh4+ 13 'lt>d1 d6 can also be considered, with compensation for the piece) 1 1 Cbd2 (other moves are clearly bad: 11 Cbc3? exd4 12 a3 'iVh4+ 13 'lt>e2 dxc3 14 axb4 cxb2 15 .!:l.b1 g3 with a winning position or 11 'lt>f2? e4! with a clear plus) 1 l . . .gxf3 ( 1 l . . .'iVh4+ 12 'lt>e2 exd4 is also promising) 12 'iVxf3 'iVh4+ 13 'lt>e2 'iVxd4 14 'iVe3 'iVxe3+ 15 'lt>xe3 f5 and Black has quite good compensation for the piece . 4 . • . C£Jhs

1 8 5

Play t h e G r ii nfe ld

White must now meet the threat of . . . 'ifu4+, and not with 5 f4 'ifu4+ 6 g3 'bxg3 7 'bf3 'ifu6 8 .l:!.g 1 'bxfl 9 .l:!.xfl 'bc6, which leaves him too exposed.

A: 5 g3 8: 5 'bh3

A) 5 g3 Simple and logical, but now White's

position looks even more loose. 5 ... tt:Jc6!

Another move order is 5 . . . d6 6 exd6 (or 6 lt:Jc3 lt:Jc6 7 exd6 �xd6 8 J.g2 0-0 9 'bh3 J.e6 10 lt:Jd5 'bb4! and Black is already better) 6 . . . J.xd6 7 f4 lt:Jc6, transposing to the note to White's 7th move below. 6 f4

After the 6 e4 lt:Jxe5 7 �e3 �g7 8 lt:Jc3 d6 9 f4 lt:Jg4 10 �d4 0-0 1 1 tt:Jge2 of D.Mergvelashvili-M.Vachier Lagrave, Chalkidiki 2003, Black should seize the initiative with 1 l . . .f5 . 6 .. . d6!

By means of this pawn sacrifice Black opens up the position to exploit

1 8 6

White's multiple weaknesses. Indeed, it looks like a very favourable version of From's Gambit (1 f4 e5 2 fxe5 d6) ! 7 �g2

Or 7 exd6 �xd6 8 J.g2 (8 lt:Jc3 J.e6 9 e4 Vfiie7, followed by . . . 0-0-0 is promising for Black, and 8 'bf3 �g4 9 J.g2 Vfiie7 10 0-0 0-0-0 1 1 J.d2 J.c5+ 12 �h1 'bf6 is also good) 8 . . . 0-0 (8 . . . J.e6 !? is another option) 9 lt:Jc3 .l:!.e8 10 'bf3 �e6 1 1 lt:JdS �c5 12 e3 'bf6! 13 lt:Jxf6+ (or 13 lt:JgS J.xd5 14 cxd5 'bb4 15 0-0 h6 16 'bf3 lt:Jbxd5 with a clear advantage for Black; Predojevic's 15 . . . lt:Jbxd5 16 �h1 J.xe3 is good too) 13 . . . Vfifxf6 14 Vfifb3 'baS 15 Vfifbs �xe3! leaves Black much better. 7 ... �e6!

Ignoring the pawn for the sake of de­velopment, and preferable to 7 . . . dxe5? ! 8 �xc6+! bxc6 9 Vfiixd8+ �xd8 10 fxe5 �g7 1 1 'bf3 J.g4 12 .l:!.fl .l:!.e8 13 'bc3 (Predojevic) with an edge for White. 8 exd6

Again White has some not particu­larly inspiring alternatives:

a) 8 Vfiib3? 'bd4! 9 Vfiixb7 .l:!.b8 10 Vfiixa7 (10 Vfiie4 dxe5 wins) 10 . . . lt:Jc2+ 1 1 �f2 J.d7 with some advantage for Black.

b) 8 �a4 dxeS ! (also possible are 8 . . . �d7 and 8 . . . .id7 9 exd6 .ixd6 10 'Llf3 0-0 1 1 0-0 'Llf6 12 'Llc3 �e7, with com­pensation in both cases) 9 ii..xc6+ bxc6 10 �xc6+ .id7 1 1 �e4 'Llf6! ( 1 l . . ..ig7 is possible, but the text is stronger) 12 �xeS+ fi.e7 13 'Llf3 0-0 14 0-0 .!:i.e8 and Black's initiative is growing rapidly.

c) 8 fi.xc6+ bxc6 9 �a4 dxeS 10 �xc6+ .id7 1 1 �e4 'Llf6! 12 �xeS+ .ie7 was promising for Black in Magalash­vili-B .Predojevic, Kallithea 2003. 8 • • • fi.xd6

Black has good compensation.

B) 5 CLlh3

A ng l i ng fo r a S ii m is c h : 3 j3

The most ambitious move, defend­ing against the check (which is now met by g3) and preparing .igS. s . . . 'Llc6 6 ii..gs

Instead, 6 e4 'LlxeS 7 fi.e3 (or 7 fi.e2 fi.cS with the initiative) 7 . . . .ig7 8 'Llf2 (8 'Llc3 dS 9 cxdS .ixh3 10 gxh3 0-0 gives Black good compensation) 8 .. .fS leaves Black very active; 6 e3 d6 7 'Llf2 .ie6 is similar to what we've already seen; and 6 'Llc3 'LlxeS 7 g4 (Black has the initia­tive after both 7 .igS f6 8 fi.h4 fi.cS and 7 e4 fi.cS 8 f4 'Llc6 9 CLldS 0-0) 7 . . . 'Llg7 8 .if4 'Llc6 gives Black good play after the regrouping . . . 'Lle6 and . . . fi.g7. 6 • . . fi.e7 7 .ixe7

Black gains at least equality after 7 .ih6 'LlxeS, intending . . . d6 and . . . .ie6. 7 ... �xe7

A critical position. If Black recovers the eS-pawn he will obviously stand well, but preventing this is not easy at all for White . 8 'Llc3

Once again 8 f4? is too weakening: 8 . . . d6! 9 exd6 �f6 10 dxc7 .ixh3 11 gxh3 �4+ 12 'it>d2 �xf4+ 13 e3 �xc7 with a clear advantage.

1 8 7

Play t h e G r ii nfe ld

In H.Stefansson-M.Krasenkow, Anta­lya 2004, White preferred 8 g4? ! which is best met by 8 . . . Cbxe5 ! (8 . . . Cbg7 9 Cbc3 Cbxe5 10 Cbd5 'ifu4+ 1 1 Cbf2 Cbe6 12 e3 was the game, when Black should have played 12 . . . f5 13 gxf5 gxf5 threatening .. .f4 with good play - Krasenkow; while 9 f4 d6 10 exd6 'ifu4+ 1 1 Cbf2 i.xg4 12 dxc7 i.e6 13 e3 �c8 14 Cbc3 �xc7 gave Black good compensation in A.Kuzmin­M.Krasenkow, Amsterdam 2004) 9 Cbc3 (9 gxh5 d6 gives Black a strong attack and another enticing possibility is 9 . . . Cbxf3+! ?) 9 . . . Cbf6 with an edge. 8 ... 0-0!

It is better to delay the capture on e5, since 8 . . . Cbxe5 9 Cbd5 is unpleasant, as is 8 . . . 'iVxe5? ! 9 g4.

9 e4 Some important alternatives: a) 9 Cbd5 'iVxe5 10 'iVd2 Cbe7! 11 f4

'iVd6 does not allow White to establish control: 12 g3 (or 12 e4 Cbxd5 13 cxd5 �e8 14 e5 - 14 i.d3 'iVxd5 - 14 . . . 'iVb6! 15 i!.e2 d6 16 i.xh5 gxh5 with an obvious advantage) 12 . . . c6! 13 Cbxe7+ (13 Cbc3 'iVxd2+ 14 Wxd2 d5! with a clear plus) 13 . . . 'iVxe7 14 i!.g2 d5! 15 cxd5 i.xh3 16

1 8 8

i.xh3 �ad8! 17 i!.g2 Cbf6 18 0-0 �fe8 is given by Leko and Adorjan as favour­ing Black, which seems correct, and they also offer 18 . . . 'iVc5+ 19 Wh1 Cbxd5 as an alternative.

b) 9 f4 is now met with 9 . . . 'iVb4! 10 'iVd2 (instead 10 'iVb3 d6 11 'iVxb4 Cbxb4 12 0-0-0 dxe5 13 fxe5 Cbc6 14 Cbf2 CbxeS 15 Cbb5 Cbxc4 16 Cbxc7 �b8 was equal in E.Postny-A.Szeberenyi, Budapest 2000, but Black might consider 15 . . . i.d7! ?) 10 . . . d6! 11 exd6 i.xh3 12 gxh3 �adS when Black has the initiative and in­tends 13 Cbd5 'iVxd6.

c) 9 g4 Cbg7 10 f4 d6 is a by-now­typical motif, as is the similar 10 . . . 'iVb4 1 1 'iVd2 d6.

d) 9 'iVd2 Cbxe5 10 e4 d6 11 Cbf2 i.e6 leads to a normal position for Black, and 1 l . . . f5 12 f4 Cbc6 13 g3 Cbf6 14 i.g2 i.e6 15 b3 fxe4 16 0-0 i!.xc4 17 �fe1 i!.f7 18 Cbfxe4 was also fine for the second player in V.Mihajlovic-M.Rabrenovic, Sutomore 2004.

9 .. . 'iVxes Inferior is 9 . . . Cbxe5 10 Cbd5 'iVd8 11

f4 Cbc6 12 i.e2 Cbf6 13 Cbf2 d6 14 0-0 according to Moiseenko.

10 'i�Vd2 d6 Black is ready to strike with . . .£5,

while White is lagging in development.

11 f4 Critical, but Black's resources are

fully sufficient. However, it's not clear what White should prefer:

a) 11 t'bf2 f5 12 0-0-0 fxe4 13 t'bfxe4 ..tf5 14 t'bf2 t'bf6 15 g4 ..td7 16 g5 t'bh5 17 t'bdS .!:!.xf3 18 .!:tel llxf2 19 'i!Vxf2 'i!Vxg5+ (Predojevic) gives Black excel­lent compensation for the exchange.

b) 1 1 g4 t'bf6! was pointed out by Leko and Adorjan (and is superior to l l . . .t'bg7 12 t'bdS!? ..txg4! 13 fxg4 'i!Vxe4+ 14 �e3 'i!Vxhl 15 t'bf6+ Wh8 with a very complicated and unclear position), when play might continue 12 0-0-0 �c5, with a good game for Black. 11 • • • 'i!Vd4!

Not l l . . .�c5 12 t'bf2 f5 13 t'bd5 with an edge. 12 �xd4 t'bxd4 13 0-0-0 cs !

A further accuracy; 13 . . . ..txh3?! 14 llxd4 ..te6 15 f5 ! leaves Black in trouble.

A n g l i ng fo r a S ii m isch : 3 f3

14 g3 t'bf6 15 t'bf2 t'bg4! Ensuring that Black quickly com­

pletes his development. Predojevic now analyses 16 t'bxg4 ..txg4 17 .!:!.d2 a6 18 ..tg2 llae8 and concludes that Black has good counterplay.

This is correct in view of the excel­lent knight on d4 and the possibility of . . . fS .

Conclusion

3 f3 i s an ambitious move, but Black can hit the nail on the head with the amazing 3 . . . e5! ? . White's weakened dark squares and stilted development justify Black' s approach. Indeed, White must be careful not to end up in a hopeless situation, and even with best play he cannot reach anything more than a normal position.

This whole line is under-researched and there is plenty of scope for further independent analysis. Best of all, 3 . . . e5 works!

1 8 9

I n d ex of Va r i at i o n s I

I ntroducing 7 t2Jf3 cs

1 d4 tt:lf6 2 c4 g6 3 tt:lc3 d5 4 cxd5 tt:lxd5 5 e4 tt:lxc3 6 bxc3 i.g7 7 tt:lf3 c5 8 i.e2

8 h3 - 14 8 i.b5+ - 1 6

8 .. . tt:lc6 9 i.e3 - 14 9 d5 - 13

The Critica l 8 .i:.b1

1 d4 tt:lf6 2 c4 g6 3 tt:lc3 d5 4 cxd5 tt:lxd5 5 e4 tt:lxc3 6 bxc3 i.g7 7 tt:lf3 c5 8 �b1 8 ... 0-0 9 i.e2 b6 10 o-o i.b7

10 . . .'ii'c7 - 2 1 11 'Yi'd3

1 1 d5 i.xc3 12 'Yi'b3 - 23 12 i.c4 - 24

11 ... i.a6 11 . . . 'Yi'c7 - 27

12 'Yi'e3 12 'Yi'c2 - 28

12 ... 'Yi'd7 12 . . . e6 - 30

13 d5 - 33 13 i.xa6 - 32

1 9 0

i.e3 Systems

1 d4 tt:lf6 2 c4 g6 3 tt:lc3 d5 4 cxd5 tt:lxd5 5 e4 tt:lxc3 6 bxc3 i.g7 7 tt:lf3

7 i.e3 c5 8 'iV d2 'iV a5 9 �cl - 42 9 �b1 b6

10 �b5 - 45 10 i.b5+ i.d7

11 i.d3 - 46 11 i.e2 - 48

7 ... c5 8 i.e3 'Yi'a 5 9 'Yi'd2 o-o 10 �c1 10 �b1 - 36

10 ... cxd4 11 cxd4 'Yi'xd2+ 12 tt:lxd2 - 40 12 �xd2 - 37

The Fashiona ble 7 i.c4

1 d4 tt:lf6 2 c4 g6 3 tt:lc3 d5 4 cxd5 tt:lxd5 5 e4 tt:lxc3 6 bxc3 i.g7 7 i.c4 7 ... c5 8 tt:le2 tt:lc6 9 i.e3 o-o 10 0-0

10 �cl cxd4 11 cxd4 'Yi'a5+ 12 �fl 'Yi'a3

13 'Yi'd2 - 54 13 h4 - 55 13 �c3 - 56

13 'i¥b3 - 56

1o . ..l2'la5 10 . . . itd7

11 �b1 - 64

11 �cl - 69

11 itd3 b6 12 �C1 e5 13 dxc5 ite6 14 C4 - 6 1

14 cxb6 - 60

White's 7th Move Alternatives

1 d4 lLJf6 2 c4 g6 3 lLJc3 d5 4 cxd5 lLJxd5 5 e4 ctJxc3 6 bxc3 itg7 1 itb5+

7 ita3 CLJd7 8 CLJf3 c5 9 itc4 - 72

9 'i¥b3 - 73

7 li'a4+ li'd7 8 li'xd7+ - 74

8 'i¥b3 - 76

7 itg5 - 79

7 .. . c6 8 ita4 0-0 9 CLJe2 c5 10 0-0 lLJc6 11 ite3 lLJa5 12 �b1 - 78

12 dxc5 - 78

4 cxd s tbxd s without 5 e4

1 d4 lLJf6 2 c4 g6 3 CLJc3 d5 4 cxd5 lLJxd5 5 itd2

5 lLJa4 e5 6 dxe5 CLJc6 7 a3 - 82

7 CLJf3 - 83

5 CLJf3 itg7 6 itd2 - 89

6 itg5 - 90

5 ... itg7 6 e4 CLJb6 7 ite3 0-0 8 ite2 8 h3 - 85

8 f4 - 86

8 itb5+ - 86

I n dex of Va ria t i o n s

8 ... lLJc6 9 lDf3 - 88

9 d5 - 87

The Russian System

1 d4 CLJf6 2 C4 g6 3 CLJC3 d5 4 CLJf3 4 'i¥b3 - 1 00

4 ... itg7 5 li'b3 dxc4 6 li'xc4 o-o 7 e4 lLJc6 8 itf4 - 97

8 ite3 - 94

8 d5 - 94

8 itg5 - 95

8 e5 - 96

8 h3 - 98

8 ite2 - 99

'iVa4+ Systems

1 d4 CLJf6 2 C4 g6 3 CLJC3 d5 4 ctJf3 itg7 5 li'a4+ itd7 6 li'b3 dxc4 7 li'xc4 o-o 8 e4

8 it£4 - 1 03

8 ... b5 g li'b3 - 1 04

9 CLJxb5 - 1 04

�f4 Systems

1 d4 ctJf6 2 C4 g6 3 CLJC3 d5 4 itf4 itg7 5 CLJf3

5 �cl - 107 5 e3 0-0

6 cxd5 - 1 1 0

6 CLJf3 c5 7 dxc5 li'a5 8 �cl dxc4 9 itxc4 li'xc5 10 itb3 li' a5 1 1 0-0 CLJc6 12 h3 it£5 13 li' e2 lLJe4 14 CLJd5 e5

15 �xc6 - 1 1 6

1 5 ith2 - 1 1 8

5 ... 0-0 6 �c1 dxc4 7 e3 7 e4 .i.g4 8 itxc4 itx£3

9 li'xf3 - 1 2 1

1 9 1

Play th e G r ii nfe ld

9 gxf3 - 122 7 ... .i.e6 8 l2lg5 .i.d5 9 e4 h6 10 exd5 hxg5 11 .i.xg5 lt:Jxd5 12 .i.xc4 l2lb6 13 .i.b3 lt:Jc6 14 d5 - 124

14 l2le2 - 125

i..gs Systems

1 d4 lt:Jf6 2 c4 g6 3 l2lc3 d5 4 l2lf3 4 .i.g5 l2le4 5 .i.h4 l2lxc3 6 bxc3 dxc4 7 e3 .i.e6

8 l2lf3 - 145 8 �bl - 146 8 �1 - 146

4 .. . ..tg7 5 ..tg5 lt:Je4 6 ..th4 6 h4 - 128 6 �cl - 128 6 .i.£4 - 129 6 cxd5 l2lxg5 7 l2lxg5 e6

8 �a4+ - 13 1 8 l2lh3 - 132 8 l2lf3 - 133 8 �d2 - 135

6 .. . tt:Jxc3 7 bxc3 dxc4 8 �a4+ - 139 8 e3 - 137 8 e4 - 138

The Sol id 4 e3

1 d4 lt:Jf6 2 c4 g6 3 l2lc3 d5 4 e3 4 ... .i.g7 5 l2lf3

5 �3 - 150 5 ... 0-o 6 .t.e2 - 157

6 �3 - 152 6 b4 - 153

1 9 2

6 cxd5 - 155 6 .i.d2 - 156 6 illd3 - 157

Offbeat 4th Move Alternatives

1 d4 l2lf6 2 c4 g6 3 l2lc3 d5 4 f3

4 g4 - 159 4 h4 - 160

4 . . . c5 s dxc5 - 162 5 cxd5 - 1 6 1

The Fianchetto System

1 d4 l2lf6 2 c4 g6 3 g3 .i.g7 4 .i.g2 d5 5 cxd5

5 l2lf3 dxc4 6 lt:Ja3 - 166 6 �a4+ - 1 68 6 0-0 - 169

5 ... lt:Jxd5 6 l2lf3 6 l2lc3 - 1 72 6 e4 - 1 74

6 ... l2lb6 7 0-0 7 l2lc3 - 1 77

7 . .. l2lc6 8 e3 - 1 8 1 8 l2lc3 - 180

Angl ing for a Sa misch: 3 f3

1 d4 lt:Jf6 2 C4 g6 3 f3 3 ... e5 4 dxe5 l2lh5 5 l2lh3 - 187

5 g3 - 186