Steve Giddins EVERYMAN CHESS First published in 2012 by Gloucester Publishers plc (formerly Everyman Publishers plc), Northburgh House, 10 Northburgh Street, London EC1V OAT Copyright© 2012 Steve Giddins The right of Steve Giddins to be identified as the author of this work has been as serted in accordance with the Copyrights, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, electrostatic, magnetic tape, photocopying, recording or otherwise, without prior permission of the publisher. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. ISBN: 978 1 85744 694 4 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 EC1V OAT tel: 020 725 3 7887 fax: 020 7490 3 708 email:
[email protected]; 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. Printed and bound in in Great Britain by Clays, Bungay, Suffolk. Bibliography 7 Preface 9 1 Pawn Endings- Introduction 13 1: Cohn-Rubinstein, St Petersburg 1909 14 2: Brinckmann-Rubinstein, Budapest 1929 18 3: Randviir-Keres, Pamu 1947 21 2 Knight Endings- Introduction 24 4: Znosko Borovsky-Seitz, Nice 1931 25 5: Pillsbury-Gunsberg, Hastings 1895 29 6: Lasker-Nimzowitsch, Zurich 1934 31 7: Kramnik-Shirov, Tal Memorial, Moscow 2007 35 3 Bishop Endings- Introduction 38 8: Yanofsky-Pinkus, Ventnor City 1942 40 9: Averbakh-Furman, USSR Championship, Odessa 1960 43 10: Larsen-Polugaevsky, Le Havre 1966 46 11: Kotov-Botvinnik, USSR Championship, Moscow 195 5 52 12: Kurajica-Karpov, Skopje 1976 56 Moscow 1969 82 19: Flohr-Capablanca. World Ch'ship. St Petersburg 1909 126 28: Capablanca-Tartakower. Vancouver 1971 68 16: Torre-Jakobsen. World Championship. Chigorin Memorial. Bled/Zagreb/Belgrade 1959 180 . Candidates T'ment. Moscow 1957 97 22: Flohr-Botvinnik. Gothenburg 1920 62 14: Plater-Botvinnik. Moscow 195 1 107 5 Rook Endings. British Championship. Germany/Austria 1910 173 38: Tal-Keres. Amsterdam 1973 74 17: Henneberger-Nimzowitsch. Buenos Aires 1927 113 25: Alatortsev-Chekhover.Introduction 1 56 33: Botvinnik-Minev. New York 1924 131 29: Flohr-Vidmar. Moscow/Leningrad 1933 101 23: Botvinnik-Bronstein. World Championship.4 Bishop versus Knight Endings.Introduction 111 24: Alekhine-Capablanca. Winterthur 1931 79 18: Smyslov-Tal. USSR Championship. Candidates Match. Swiss Ch'ship. World Ch'ship. Morecambe 1975 169 3 7: Schlechter-Lasker. World Championship. World Championship.Introduction 60 13: Reti-Rubinstein. San Sebastian 1912 124 27: Spielmann-Rubinstein. Vienna 1898 166 36: Webb-Miles. Tbilisi 1937 118 26: Schlechter-Rubinstein. USSR Championship. Baguio City 1978 151 6 Queen Endings. Moscow 1984 90 21: Botvinnik-Smyslov. Varna Olympiad 1962 145 32: Korchnoi-Karpov. Tilburg 1981 141 31: Botvinnik-Fischer. Amsterdam Olympiad 1954 158 34: Marshall-Maroczy. Nottingham 1936 137 30: Andersson-Miles. Ostend 1905 1 63 3 5 : Pillsbury-Burn. Moscow 1947 65 15: Fischer-Taimanov. Moscow 193 5 86 20: Karpov-Kasparov. Candidates Match. U SA 1907 193 42: Andersson-Stean. USSR Championship. World Championship. St Petersburg 1914 190 41: Marshall-Lasker. Brissago 2004 230 Index of Games 23 3 . FIDE World Cup. Khanty Mansiysk 2005 222 49: Karpov-Kasparov.Introduction 184 39: Timman-Velimirovic. Riga 1958 219 48: Carlsen-Azmaiparashvili. Moscow 1963 211 46: Fischer-Petrosian. Rio de Janeiro 1979 185 40: Alekhine-Lasker. World Championship. Moscow 1984 226 SO: Kramnik-Leko. World Championship. Buenos Aires 1971 216 47: Petrosian-Bannik. Interzonal.7 Rook and Minor Piece Endings. London 1883 202 44: Barcza-Soos. Sao Paulo 1979 197 8 Multi-piece Endings.Introduction 200 43: Englisch-Steinitz. World Championship. Varna Olympiad 1962 207 45: Petrosian-Botvinnik. Lars Bo Hansen (Gambit 2006) Grandmaster Chess Strategy. Andre Cheron (Berlin Frohnau 1952) 7 . Alexander Panchenko (Yoshka-Ola 1997) lskustvo Endspily a.und Handbuch der Endspiele. Vasily Smyslov and Grigory Levenfish (Batsford 1971) Studies and Games. (Progess in Chess 2002) Lehr.com November 2000) Basic Chess Endings. Yuri Averbakh and llya Maizelis (Batsford 1974) Knight Endings. Jon Speelman (Batsford 1988) Practical Rook Endings. 2nd Edition. Reuben Fine (McKay 1941) Practical Chess Endings. Yuri Averbakh and Vitaly Chekhover (Batsford 1977) Barcza Gedeon: Elete es Jatszmai. Mikhail Marin (Quality Chess 2004) Sec rets of Minor P iece Endings. Vasily Smyslov (Moscow 2006) Rook Endings. Mark Dvoretsky (Chesscafe.Bishop Endings. Mark Dvoretsky (Russell Enterprises 2006) The Battle of the Heavy Pieces. Paul Keres (Batsford 1974) Sec rets of Chess Endgame Strategy. Jurgen Kaufeld and Guido Kern (New in Chess 2001) Learn from the Legends. Yuri Averbakh (Batsford 1975) Pawn Endings. Jan Timman (Cadogan 1996) Chess the Hard Way!. Yuri Averbakh (Batsford 1976) Queen and Pawn Endings. Yuri Averbakh (Batsford 1977) Bishop v Knight Endings. John Nunn (Batsford 1995) Teoria I Praktika Shakhmatnykh Okonchanyii. Daniel Yanofsky (Pitman & Sons 1953) Analysing the Endgame. Viktor Korchnoi. Istvan Bilek (Sport Budapest 1990) Dvoretsky's Endgame Manual. with just a cue ball. which shows that chess is not the only game where such a study of the basics pays off. were absolutely right to recommend that beginners start their study of chess with the endgame. and also to in struct the reader. his father only allowed him one ball. was tu tored in his youth by his martinet father. because it is by studying simplified positions. by presenting some of the finest practical examples of endgame play in master chess. such as Capablanca. 9 . the Australian star. short of making it sing and dance. The present book is somewhat different. we will look at the different types of endgame. and what they do and do not "like". because this phase of the game has always been my favourite aspect of chess. who in the 1930s and 40s developed the game of bil liards to such a level that he practically killed it as a competitive event. and this expertise was the foundation of his later immortality at the game. Walter could do just about anything imaginable with it. It seems that for the first three months. practicing. In this context. The Greatest Ever Chess Endgames aims to entertain. The older masters. In chess. Obviously. himself an excellent player. Walter Lin drum. As with other books in this series. that one can develop an understanding of the capabilities of the pieces. with only a few pieces on the board. Although we all enjoy seeing spectacular sacrificial attacks. there are many textbooks on the endgame which deal with the basic theoretical positions. Young Wal ter was forced to spend up to twelve hours each day. Vasily Smyslov is an excellent example of a player who learnt his chess expertise by starting with the endgame. I recently came across an interesting story. The complexities and fine points that can be hidden in positions with only a small number of pieces on the board never cease to astonish. and see how they have been played by the world's greatest masters. the cue ball! As one can imagine. by the time he had spent that long. it is in the endgame that the true depth and subtlety of chess is revealed. Through the medium of so annotated examples.I was delighted when Everyman asked me to write a book on the greatest ever endgames. Whilst I would accept that adjournments. One only has to look at the epic analytical achievement of Jan Timman and his second Ulf Andersson in Game 39. One other thing I should comment on. as far as pos sible. To take just two examples. this process has been taken even further. applicable to all endings. and the use of third parties to assist with analysis of adjourned games. Clearly. Rubinstein's famous rook endings against Lasker (St Petersburg 1909) and Alekhine (Karlsbad 1911) have both been omitted. is the age of the games presented here. as far as possible. because I felt them to be so well-known that it would be impossible to say anything original about them. In recent years. where I have included a summary of some fascinating analysis published in the Russian journal 64 a few years ago. Where I have chosen other famous examples. The paucity of games from the past 20 years to be found in this book reflects this sad fact. I think there is also another. The reader will note that. I have tried. and many long be fore that. as being a battle between two individuals over the board. One example is Capablanca-Tartakower. there are three general endgame principles. and there are bound to be some games missing from the book which readers might expect to see here. this reflects my own interest in chess history. and rever ence for the great masters of the past. to add something new.Th e Greatest Ever Ch ess Endg a m es I should say a word about the selection. with a book such as this. most of the games were played no later than the 1980s. For now. with the adoption of Fischer-style increment time limits. which are worth mentioning here: 10 . which is the change to tournament regimes and time-limits over recent years. more objective reason. To some extent. These have enabled me bring a degree of certainty to some positions that was previously unachievable. and will be games which have been published in other endgame books in the past. to see what we have lost by getting rid of adjournments. which almost all authorities agree have further damaged the standard of endgame play. However. New York 1924 (Game 28). However. as far as the selection is concerned. does contradict the basic philosophy of chess. A great many of the games in this book were played in the days of slow time-limits (40 moves in two and a half hours being the norm). Where appropriate. or at least to pull together recent analytical discoveries published in more obscure sources. with a few excep tions. and with ad journments. At the beginning of each chapter. a lot of the examples presented will be familiar to many readers. which summarize the main principles and characteristics of the different types of endgame. I have also made use of the seven-man endgame tablebases. to pick at least a few slightly less well-known examples. I think it is clear that abolishing adjournments has had a hugely negative impact on endgame play. which have only recently become available. I have added some introductory remarks. I have tried. or to avoid any suggestion of a wasted tempo. and exact calculation of specific variations is re quired. such as in Henneberger-Nimzowitsch (Game 17). you will see that a long bout of patient manoeuvring usually culminates in a specific. which needs to be accurately calculated. so as to allow penetra tion (or. This may be as sim ple as counting a passed pawn race in a king and pawn ending. Whatever the precise means. the pace of the endgame is slower than the middlegame. but this is a major fallacy. quite often. a second front is nearly always required. in this book. Endgames always become concrete! This rule. The majority of the time (rook endings are a frequent ex ception). and have little place in the ending. and there is not the same pressure to take immediate action. and the need for patient manoeuvring. merely the threat thereof) by a piece. in favour of taking minor prophylactic measures. If you look through the endings in this book. 3. marginally improving the position of a piece. and I hope the reader will gain a similar enjoyment from reading the book. Many players tend to assume that tactics and calculation belong in the middlegame. but if you cannot or will not calculate variations. Such play is absolutely typical of the endgame. P refac e L The principle of two weaknesses In almost all of the endgames you will see in this book. 2. or it may involve much more complex calculations. so as to stretch the de fender's resources beyond their limits. which to some extent may seem to contradict point 2 above. Instead. opening a file. playing small strengthening moves. especially the minor piece endings. A single weakness is rarely enough to lose by itself. Do not hurry! Time and time again. etc. This has been an immensely enjoyable book to write. Despite the slower pace of endgames. was taught to me years ago by GM Jonathan Levitt. con crete tactical finish. patient manoeuvring and gradual strengthening of the position tend to be the order of the day. you will see players delaying decisive action. This may involve creating another passed pawn. If it serves to stimulate an interest in the endgame amongst 11 . nearly all endgames eventually reach a phase where they become concrete. the key to winning the po sition is for the stronger side to create a second front. Spending my days going through some of the finest practical endings ever played has been a great way to pass the last few months. or simply weakening a single square. you will not be a successful endgame player. who between them dug out and supplied details of Larsen's Danish commentary on Game 10. Rochester. Bernard Cafferty. for some interesting analytical suggestions in Game 25. the eventide imbibing of whose products provided such splendid refreshment from my daily labours on this book! Steve Giddins. who provided results from his seven-man endgame tablebases (and also John Nunn and Harold van der Heijden. who put me in touch with Jakov in the first place). Laithwaites Wine. Per Skjoldager and J�rn Erik Nielsen. but by no means least. and finally. Calle Erlandsen. for their advice. U K. for supplying photocopies of Dutch-language material pertaining to Game 16. Jakov Konoval. January 2012 12 . technical support and general encour agement. Last.Th e G rea test Ever Chess E n dg a m es players who hitherto thought it dull. Gerard Welling. Matthew Sadler. who clarified the probable repetition rules applying in Game 35. acknowledgments are due to the following: John and Byron of Everyman. I will be even more delighted. widely ac knowledged as one of the very greatest 1. to lose a tempo Finally. of a tactical breakthrough to establish a 3. King activity. to establish a passed pawn. endings. overall endgame technique. Game 1 critical in all sorts of endings. to be relatively insignificant. as are protected passed is able to realize a miserable-looking pawns. which appears at first sight from combinational breakthroughs. Al are probably the most concrete of all. The main concepts of king and In this section.King and pawn endings are the basic 4. it is generally essential to be endings is absolutely basic to good able to calculate everything in detail. This is frequently endgame players of all time. The ability in other endings. but with 2. King opposition. Game 3 sees an elegant example and transfer the move to the opponent. combined with threats is a standard technique. Outside passed pawns are often passed pawn. of the sort which are common endings after simplification. but never sees a typical demonstration of how to more so than when the kings are the exploit the vastly more active king. Triangulation. we present three clas pawn endings are well known. Game 2 shows a similar theme. There is usually such an ending. Two are by be summarized only briefly here: the great Akiba Rubinstein. In king and pawn to understand and play king and pawn endings. 13 . both the possibility of a combinational break direct and distant. and will sic king and pawn endings. Above all. only pieces remaining. and varia where a long phase of play consists of tions are everything. Opposition. are key ideas in such through. king and pawn endings building blocks of the endgame. numerous games fin little scope for general strengthening ish up being decided in king and pawn moves. by means of which Black decisive. in though there are not that many games that tempi are usually vital. Such passed pawns often result extra pawn. of triangulation. because their rather cooperative.i.l:. The de We1 . Advance his king to h 3.Th e Greatest Ever Chess E n dg a m es 1 d4 ds 2 lLlf3 cs 3 c4 dxc4 4 dxcs weakness i s frequently not enough to 'iii'xd1+ 5 .ig4 7 .hc1 l:tac8 15 . the secret of winning frequently lZ:lxc4 17 .xc1 of one weakness. Replace the white pawns draw and has systematically hoovered on a2 and b2.ib3 l:thd8 16 lLlc4 fare.ixf3 20 gxf3 lLld6 21 . However. because of the well-known principle. a single exchange all the kingside pawns. he hold the king and pawn ending despite proves unable to defend..ti>e7 12 .xcs 9 b4 .ti>d2 lLlc4+ 23 .l:lc8 22 fender's forces are tied down to defence ..ti>e2 . forcing the ending proves the more correct. just as in war 14 . the white king to defend the h2-pawn The reason White loses this ending is passively. at the right moment. lZ:lbd2 . so as 14 . as the defender can concentrate all e6 8 a3 .ixes lZ:lxes and simply hold.xc4 lLle4 19 lies in opening a second front. But Rubinstein's assessment of 1. In this position..i. what is less apparent to the inexperienced eye is that his queen side White's play hitherto has been pawns are also weak. Such a change of front frequently proves enough to break the defence. confident that he can not be losing this ending. as Black cannot effectively attack them). utilizing his superior piece mo bility. Instead.. In chess.l:. and he would probably off the pieces. it is obvious that White has a weakness on the kingside.id6 10 . but as it is.ies 13 . 2.ib2 lLlf6 11 of his defensive resources on that point. to say the least.ie2 . and then the attacking 25 Wxc1 side switches his attention to the other target. of as possible. Advance his king side pawns as far especially important in endgames. namely the isolated pawn on h2 (the doubled pawns on the f-file are not really an exploitable weakness here.ti>xd1 lLlc6 6 e3 . advanced position makes them vulner Cohn was obviously looking only for a able to attack.ixc4 lose.xc4 l:txc4 18 .ixc4 l:txc4 24 :c1 J:. "two weaknesses". and. Black's plan the defect in his kingside pawn struc consists of three steps: ture. after 3 1 �c7 b 5 ! Black is far too quick. instead... Although some "30-ply" (to 25 �6 26 <ii>d 2 ••• use computer speak. i. b5 at once. 29 .es! This is an important move. always assum ternative is to seek counterplay. e.. the variation runs as fol lows: 26 <ii>c 2 <itl>g5 27 <ito>c3 <ifi>h4 28 <ifi>c4 'it>h3 29 <iti>c5 �xh2 30 <ii>d6 <ii>g 2 31 b5! (the best try..e. <ii>gl calculation shows that this idea fails. Thus. h 5 32 <ifi>c7 h4 33 <iti>xb7 h3 34 <ii>xa7 h2 3 5 b6 h 1'ii' 36 b7.) 31 . 15 . ending. because Black is able to exchange queens at the end and win on the king side. be fixed.. winning the black queenside pawns 26 .g. Now they can not advance. weak nesses should. See 36 . <ifi>g4. one might won 39 �c7 'ii'xb8+ 40 <iti>xb8 �xf3 and Black der why he did not play 29 . and the text does that with the white doubled f-pawns.. by ing one has sufficient time on the bringing his king to c7 with the idea of clock.. His only al such a line at the board.. winning a pawn."it'a1! 37 b8'ii' 'ii'x a3+ 38 <iti>b7 'iYh4+ ing Black's next move. P aw n En dings to leave the black king nearer to the wins easily. As we will see many times in this book.. .. and this completes step one of Black's plan. exf4. Exploit this last factor to race the simple variation that the player often king over to the queen side and win the needs to calculate in a king and pawn white queenside pawns. it is not so hard to work out with his opponent's plans.�g5 27 <ii>e 2 <ifi>h4 28 �1 �h3 29 and queening his own pawn. 32 <iti>b7 h 5 33 'it>xa7 h4 34 <ii>b 6 h 3 3 5 <ii>xb5 h2 36 a4 h1'ii' etc.. Now he starts the advance of his kingside pawns. This is a good example of queenside than his opposite number. 15 moves for White has no choice but to fall in each side). the kind of very long. wherever possible. but would always be met by .. because a subsequent f4 It looks as though White is fine. but relatively 3. The white king has just made it across in time.. fol now comes the sting in the tail: lowed by . However. fxg4. 3o. when Black wins in elementary help. followed by winning the white pawns on the queenside. pointed out by sive tactic of oscillating his king be Mark Dvoretsky.fs 32 'it>h1 gs 33 'it>g1 hs 34 'it>h1 g4 30 'iii>h 1 White could try to anticipate Black's next by playing 30 b5 himself. plan is hindered. .g 5 32 'lii>g 1 40 fxg3 'it>xg3 and wins.. .Th e Greatest E ver Chess En dg a m es but this would have permitted 30 f4.. but it makes no particular difference. play would have been much more in 31 'iii>g1 teresting after the natural 35 fxg4. bs 35 e4 Rubinstein decides to fix the enemy This move loses rather easily. fashion. is the slightly counter tween gl and hl.hxg4. After the further moves 36 'it>g1 f4 37 exf4 exf4 38 'iii>h 1 we reach the crucial position. as they are both highly in structive and also not devoid of histori cal interest. h 5 33 'iii>h l h4! ( a key preparatory Interestingly. which also wins. White is reduced to the totally pas Now the simplest win. though.. although only with some accurate play by Black. almost all move) 3 4 'it>g1 g4 3 5 fxg4 'it>xg4 36 'it>g2 books pre-Dvoretsky give only the move 35 ... by following his plan: 36 'lii>g 1 Averbakh then gives the following e 4 37 'iii>h l h 4 38 'it>g1 g3 39 hxg3 hxg3 winning plan for Black: 31 . because it weakens the f4-square...h 3+! (the point) 37 'it>g1 'it>f3 and after which the second stage of Black's wins easily. it is worth looking at the variations after 3 5 .. 16 . hxg4. Although superfluous in view of the simple win after 3 5 . The only other move intuitive 3 5 ... 31 ... and the fact that the pawns have been advanced even further merely weakens them still more. Black will still continue with the plan of ad vancing and eliminating the kingside pawns. 36 .fxg4 (away from the cen he could try is 3 1 e4 but this does not tre!).. but queenside weaknesses anyway.. this does look like the logical culmination of Black's plan.bxa4 46 <iiic 1 is also drawn. continues 39 fxg3 fxg3 40 hxg3 �xg3 and adds the conclusion "and the success of Black's plan appears in full relief'... as is <Ji>d2 <Ji>d4 45 Wc2 <Ji>c4 and Black has 4S. who also pointed out the queen on a1 first.) 46 the opposition. Back at the last diagram..h4 37 <Ji>g1 g3 38 hxg3 hxg3 0-1 bining threats of switching his king to White clearly loses the e4-pawn. Averbakh.Wg5 41 h 3 gxh3 42 Wh2 Wg4 43 <iiig 1 <oitf4 44 Wh2 We4 45 <iitxh3 Wd3 46 Wg4 <iite 2 47 Wg3 a6! and wins) 40. 36 fxg4 hxg4 37 Wg1 e3! 38 fxe3 e4 39 and instead play 38 . P aw n Endings the queenside or still going after the h2-pawn: 40 <iiif1 (the other try is 40 <iith 1 when Black's reserve tempo on the queenside proves decisive: 40. but as Jon Speelman pointed out in Analys This is zugzwang... give way: 46 <Ji>d2 <Ji>b3 47 <Ji>e3 <Ji>xa3 48 Speelman attributed this find to Jona <iiif4 <Ji>xb4 49 'it>xg4 as and Black will than Mestel. White in fact draws Wh4! wins the h-pawn) 42 ..<iitxb4 46 axbs <iitx bs 47 <iiic 1 etc. for one.fxe4! 36 fxe4 tually needs to abandon his original Speelman also points out the line plan of liquidating the king side pawns.. com 36. Indeed. g3? as winning for Black..f3! 39 Wg1 Wh4 <Ji>h1 g3 winning... thereby taking con correct winning plan for Black at move trol of the white pawn's promotion 38.. after 41 <iiig 1 <iiif3 42 <iiif1 Black gets to the queenside with just �e3 43 <iiie 1 <iitd 3 44 <iitd1 <iiic3 45 a4! a6 one tempo to spare) 43 We1 <Ji>e4 44 (4S ...'it>hs ! 41 We1 WgS ! Here. the white king must axbs axbs 47 <iitc 1 <iiixb4 48 Wb2 etc. square on h8.. 4 2 <iiif1 (42 Wd2 ing the Endgame... 17 .. and aim for a subtle zugzwang. several generations of com mentators have mistakenly given 38 . Black ac 3 5.. Wf4! (now this position. We hxg6 17 l:tae1 . .:tds 24 f3 .i.xe2 f6 20 <M1 <M7 21 h3 gS 22 .i.e8 55 .i. 39 'ii>d 2 l:.c7 42 provided by this type of pawn struc- 'it>d1 il.l:ld7 25 'ii>d 2 as 26 pawn structure includes fixed doubled a3 a4 27 'ii>d 3 . At present.e3 c6 vanced queenside pawns create the 11 .c7 l:tc8 32 i. Once again.fe8 15 .cs 28 :e1 .i.l:td2 .xf4 'it>xf4 potential dangers facing White in such positions. .d3 ds 12 'ii'd 2 .i.bs consists of the more active king. but it 61 .e6 57 gS i.f4 l:.e1 c4 54 il..l:.c8 48 b4 (with the threat of 68 .xe6 'iii>xe6 60 i.l:.g3 Black's a-pawn.d6 40 .g3 . promot- l:td1 'iii>c6 49 .i.g3 66ctf2'1ii>e4 llh8 33 J:te2 :tas 34 .f2 i.:tes so l:te2 l'te6 51 h4 ing a pawn) and if 68 axb4 then 68 .e s 43 'iii>c1 'it>g6 44 'ii>b 1 eMs 45 ture..Ws 62 <ii>c1 i.i. would be the line 66 .c7 pawns.l:te2 'ii'g6 16 'ii'xg6 wing.i.Th e Greatest Ever C h ess E n dg a m es 1 e4 es 2 lbf3 tbc6 3 lbc3 lbf6 4 ..e3 g4 56 fxg4 :!.cS+ 'iii>d 7 59 .c6 41 . however.i.i.U.xd3 possibility of a breakthrough on that 14 'ikxd3 ..c3! 'ii>d6 52 hxgs fxgs 53 l:.i.i.:tc8 30 i.i. . establishing a passed pawn.l:r.c7 %b7 35 i..e3 .f4 63 <iir>d 2 'iii>xgs is worth seeing as an illustration of the 64 We2 'ii>g4 65 .U.i. Rubinstein's advantage 18 .i.d4 g6 this line fails to the move 68 c3!.d8 29 .l:.a 7 knocks out the last barrier in the way of l:. c3.i.g3 l:..e7 9 o-o o-o 10 i.d4 67 cxd4 l:td2 <iir>e6 46 l:te2 <iir>d 6 47 .fs 13 l:tfe1 .d8 31 il.f2 cs opportunity for a pawn breakthrough.i..i.xe2 19 will see this motif in the game itself. 58 .d2 ...g3 and it is especially common when the bs 23 <ii>e1 . The lbd4 5 lbxd4 exd4 6 es dxc3 7 exf6 other factor here is how his more ad li'xf6 8 dxc3 i.l:.a7 37 l:te2 i.f8 18 .g3 An example of the sort of tactical lld7 36 lle1 .b6 38 . as is the case here. •• As in the note to White's 68th. Now the pawn on g2 prevents the white king maintain ing the opposition. 73 .. P aw n En dings 67 'iti>e2 gS 68 <M2 71 �d2 If 68 g3. c3) 72 ... 19 .. then Black wins in the same If 71 axb4 there follows 71. d4 69 cxd4 bxc3+ �c4!! (this is the refinement - �d4 70 'iti>d2 �e4 71 �e2. because after 74 g3 the kingside is closed and Black has no space for his outflanking manoeuvre..a2.b4! 72 c3 (the only defence against 72 . 72 c3+ �e4 73 'it>e2 and now again the breakthrough idea applies: 71.. by 68 .. 74 <M2 �g4! This is the point.. the black king stops White's passed b pawn) 73 �d2 a3 74 �cl �xc3! (threatening 7S....c3 72 way as in the game. and thus forcing White's reply} 75 �bl �xb4 76 �a2 �c3 77 'lto>xa3 �xc2 and wins. etc..b3! and Black wins by virtue of having the op position. The position would be drawn.. The key effect of Black's last two moves in this variation has been to force White to weaken the d3-square. tion. 68 d4! . but he can be outflanked on the 69 cxd4 �xd4 70 �e2 b4! king side.. Now White's king has the opposi The same idea..g4?? would be a dreadful mis take. 71 b3 ...<M4! 73 . so now Black's king can enter after 73 'iti>f2 �d3. the key effect of Black's play is to force a fatal weakening of d3.. .tf5 'iitf3 80 �e5 �e3 81 �d5 c. 79 �g3 79 'IW1 g3 is hopeless.. 79. after 81 �4 �f2 82 'it>e4 �e2 83 �d4 'iitd2 84 �xc4 �c2 etc.�g2 78 �xg5 �xg3 79 . 20 .t>xg2 0-1 75 Wf1 After the alternative 75 g3.t>d3 82 �c5 c.�g1 80 'iitxg4 c. 77 Wf1 �h2 78 Wf2 The b2-pawn falls and Black will queen his own b-pawn. etc.g. 1s ..'it>h3 76 �f3 'iith 2 77 'iit g4 (77 'IW2 g 4 wins) 77. g3 +?? 79 'iitf3. 79 .. �h1 .. Black would only have a draw. Black wins as follows: 75 .'iitg 3 76 �g1 g41 The vital reserve tempo allows a further outflanking of the white king......Th e Grea test Ever Chess E n dg a m es 78.. Care is always required in the end game! After 78 .. e.. �g1 80 �xg3 �f1 81 'iiff3 'iite 1 82 g4 'iitd2 83 g 5 �c2 84 g6 �xb2 85 g7 �xa3 86 g8'i¥ b2 87 'i¥b8 b1'ii' ! (simplest) 88 'ili'xb1 stalemate.t>c2.. bs a6 8 Black's passed cS-pawn. should the black 33 . the h .l:te2 lle7 2 S �d2 f6 26 dance for the opposition.d1 �as 34 .l:.xe4 40 . lbefs . .. h7-h6.i.l:.d7 17 all. h s.l:. it does not look all that valuable.xd3 4S �xd3 the game for Black.a4 lbes 9 'iie 2 lD7g6 10 lbd2 lbf4 11 spite these first impressions.i.i.l:.i.xf4 'ifxf4 12 g3 'iig4 13 'i!i'xg4 lbxg4 pawn in fact turns out to be vital after 14 h3 lbes 1S . but only thanks to some extremely nifty footwork by his king which would not be out of place on Strictly Come Dancing! 4s . but since it is the backward pawn on h7.i.xc6 24 .xe4 41 cxds ...i..l:.d7 breakthrough .i. Black has an extra pawn.xg4 the reply gxh s .i.d3 These factors prove sufficient to win .he1 gs 28 fs .:!e3 l:the8 27 .xe4 .l:tes 29 c4 there is also the potential tactical bs 30 b3 �b7 31 �c3 �b6 32 �d4 . the white king is after 4S .lies s .c6 20 provides a reserve tempo with .i.d7 22 f4 lbc6 23 which can play a vital role in a king lbxc6 ..i.i.�b6 46 �c4 aS 47 a4 Black quite active in the centre..bS+ 43 'iii>e 3 .c2 d6 16 o-o-o . There are two reasons. de . and he has a would be forced to expend his precious 21 .i.i. One is that it lbf1 0-0-0 18 lbe3 :!deS 19 nd2 . 42 hxg4 .i.'iitb s! This is a particularly interesting po sition.i. P aw n E n dings 1 e4 es 2 lLlf3 lbc6 3 d4 exd4 4 lbxd4 passed pawn on dS to set against .e3 'iif6 6 c3 lbge7 7 . as it seems unlikely to produce a passed This is the only move to win.xd4 21 lbxd4 . since pawn.c2 �b4 3S a3+ �as king be located within the square of 36 .i. Secondly. Meanwhile.l:.. However.xf1 44 .i..f1 cs+ 37 �d3 bxc4+ 38 bxc4 ds 39 the resulting white passed pawn after g4 . .. 48 'it>c6 49 d7 'iii>xd7 so 'it>xcs 'it>e7! .... pawns..h7-h 5 means that the white king is effectively held in an invisible prison.• 52 . since after 48 'it>c3 'it>c7 49 'it>c4 52 'it>e4 'it>d6 so 'it>bs Wxds 51 'it>xas.. because after 51. whilst 46 a4+ 'it>b6 47 'it>c4 as the white h-pawn is stopped by Black's Now we have the same position as king (see the introductory remarks). but with White to This is the key point of the whole end move .. bounded by the square of the gS-pawn. The threat of . and shows that Black's extra h pawn is not the irrelevance it may have first seemed.a crucial difference.Wc6! imprisoning the white king in front of his a-pawn....g..Th e Greatest Ever Chess E n dg a m es reserve tempo with 47.. Black wins with the neat S1 . e.Wf7 Now 52 .h s ! 52 would only be a draw. in the previous note.. The point of the 22 . As will be A vital moment.Wh6 and ... 52 'it>d4 h S ! 53 gxh s Wg7! (with the idea of 48 d6 simply . gxhs g4 the g-pawn promotes.h s is again a threat.. this time with a quieter follow-up.h s ? 53 gxh s Wg7 is not possible because of 54 Wf3 'iii>h 6 55 'iii>g4 and the white king is in time to s1 'iii>d s defend the h S-pawn. and cannot go left of the c-file or further forward than cs... 51.'it>xhs) 54 'iii>d s g4! Keres shows that White has no al 55 'it>e4 Wh6 56 Wf4 'it>xhs winning eas ternative to this exchange of passed ily..h6. ing.. The move 51 'it>bs is come clear later.'iW8! Now 52 . the resulting position not possible. . showing several important themes.h s 54 gxh s gxhs 'it>xfs 59 'it>xas g4 60 'it>bs g3 61 as 'ifilg7. 53 '1t>e3 56. but. g2 62 a6 g 1'ii' and wins... once again because of the strike by the h pawn: 56 'it>c4 'itte s 57 'it>bs 59 .. His 57... h51 Fittingly enough.. 23 . to be able side the square of the gS-pawn! 58 to meet the threat of 5 3 ....hS! Now the white king is out king has to stay on the e-file. 60 gxh5 'it>xf5 61 'ifi>f3 'it>e6 As Keres points out. but now we see that 57. but allows un this too has its drawbacks. 'it>xa4 59 'it>ds '1t>b3 60 'it>e6 etc. 56 �e4 Going the other way also loses. Black still has his reserve tempo in his pocket. crucially.'it>b4 also wins. necessary counterplay after 58 'it>d4 53 'ifi>e71 ••.g4+ 62 '1ii>g 3 '1ii>g s 63 h6 although Black is still winning after 63 . Keres' The black king changes direction move maintains Black's dominance and heads back to the centre..'�xh6 64 'it>xg4 'it>g6 65 'it>f4 fs. 62 'it>g4 �7 63 �f5 '1ii>g7 o-1 A highly instructive endgame. must come to meet him and take the 58 '1t>d3 'it>e5 59 'it>e3 opposition. . this is again simpler than 61 .. Black takes the oppo sition and the white king must give way. P aw n E n dings text is that White is in zugzwang.. White and wins without any alarums.'1t>c5 57 'lt>e3 'ittd 51 The only move. 54 'it>e4 'it>d6 55 'it>d4 h61 And here it is. it is the h-pawn which administers the coup de grace. only see a relatively rare example of a double the kings can use triangulation to lose knight ending. certainly not shows a filigree endgame between two to the same extent. The knight is not able to lose a greater piece activity enables Black to tempo by itself. but the presence of knights still opposite number and set a passed alters the position. achieved in a position of equal material. White has a space advantage. Knights hate rooks' pawns! notorious ending of two knights versus Knights are quite clumsy when fighting pawn. we see the 2. and is they are especially poor against a very rarely handled correctly in practice. a knight sacrifice. Finally. As in many endings. because of the knight's discernible mistake from his opponent. is often a istics of such endings are: common feature of such endings. This is because the edge In Game 4. Game 6 always do to other pieces. This is often an impor squeeze a win out of a position with tant factor. Finally. In Game 4. a win is a tempo and transfer the move. 4. The knight's short range renders In this section. and means that if some equal material. rook's pawn. Dr Seitz gives a textbook of the board seriously restricts the demonstration. This is one of the most fascinat against passed pawns in general. as is structure. 24 . which proves enough. king activity this time also with a symmetrical pawn is a big factor in knight endings. of the game's immortals. Game 5 sees a classic knight's agility in a way that it does not tactical battle of passed pawns. in which 3. to deflect its ings. knight endings.Knight endings are usually considered clumsiness in fighting against a passed to be the closest to king and pawn end pawn. we present four outside passed pawns especially strong. 1. space generally. Here. but ing theoretical endgames of all. once again. we form of zugzwang position arises. without a clearly s. in Game 7. The main character pawn decisively in motion. l:Id4 38 poser. two knights cannot force mate lL'ld1 bS 13 f4 lL'lg8 14 es 'W/c7 15 exd6 against a bare king. 27 . in the early years ttJe2 l:. on each file..if1 <t>h6 35 dary Russian analyst and study com lLlc3 h4 36 ..id3 lL'ld6 51 . lL'lb6 53 . generations of computer endgame ta- 25 .:txd3 32 endgame was carried out by the legen lL'lc3 lL'lf6 33 ttJxbs hs 34 .id7 9 occurs rarely in practice. and he even worked out precisely �d1 lL'lh4 62 . However..l:be2 ment is remarkable.l:txg3 .if3 �e7 54 hs gxhs ss .ig7 5 nically fascinating in chess. when John Nunn published Secrets of Minor P iece Endings...ixhs He identified the basic plan.b3 lL'ld3 40 Wg3 lL'le4+ 41 of the 20th century.ig2 .d2 39 l:.e2 . because at the cru 1Wxd6 16 ttJec3 fS 17 a4 a6 18 axbs cial moment the defender's king will be axbs 19 .l:.. of course.l:.l:lf2 lL'lb4 28 l:!.l:!.g2 hxg3+ 37 .b8 10 .. as he analysed the 44 Wxe2 lL'ld4+ 45 �d3 lL'lxb3 46 �c2 endgame single-handedly. Alexey Troitzky. the tech lL'ldS+ 56 �f3 lL'ld4+ 57 �g3 lL'lxb4 58 niques for corralling the defender's .ie2 lLlcB 52 h4 build. so it is possible to force mate. for new discoveries thrown up by the early This ending is one of the most tech.. with no prior lL'ld4+ 47 �d3 lL'le6 48 �e3 Wg7 49 b4 published theory or practice on which to �6 so .:te2 . in order for the position to be a win. Knig h t En dings 1 e4 cs 2 lL'lc3 lL'lc6 3 g3 g6 4 . and. This was part of a series of books in which Nunn updated and cor rected established endgame theory. no computer help.. the stalemate does not 'ifxd4 lL'lxd4 25 l:td6 ttJxc2 26 ..l:Ixe6 lL'lg8 31 . Troitzky's achieve �3 lL'le1+ 42 �e3 lL'lc2+ 43 <t>f3 ..a6 e6 20 'ii'f2 . if the defender cxd4 22 lL'le2 ttJge7 23 lL'lxd4 'ii'xd4 24 has a pawn.xd7 l:lfe8 arise.ie3 h6 11 'iid 2 �h7 12 known.id4 21 .ixd4 stalemated. although it ttJge2 d6 6 d3 lL'lf6 7 o-o o-o 8 h3 . The biggest compliment to the quality of his work came in 1995.ig6 'it>e6 59 �2 lL'lds 60 'it>e1 lL'lf3+ 61 king. but almost all of the credit for solving this ending goes to Troitzky. As is well �h2 l:.l::t bd8 29 l:txd8 Almost all of the research on this hd8 30 .. A few later analysts tidied up some loose ends..txfs+ lL'lxfs where the defender's pawn needs to be. which will then advance. and allowing for the re 2. Troitzky lov-Lilienthal. only 3-4 times playing fore have a black pawn): a move that the tablebase shows to be White wins. and con sists of three stages: 1. which is as fol onship 1941)! In the present game. even for top Obviously. he brings over the blockading ing is extremely complicated and far knight to effect the mate. and it varies depending again. he reached it not uniform. and never one bad pawn is no further forward than shown enough to add more than a handful of in the following diagram. and still failed to win it (Smys on which file the pawn stands. He obvi the case. the line ously did not learn much from this. Effectively. he has to release the blockade ending arises. the pawn is behind the king into a corner. 26 . 3. rather more often than it is won. USSR Absolute Champi identified the full line. the end corner. For various reasons. When it came to two knights versus pawn. and so the position is a and the other knight. and this in turn means Lilienthal reached the endgame. the leading grandmaster Andor to give mate. The stronger side blockades the We can see from this that. present case. when he commences grandmasters. using his own king "Troitzky Line". when this stage 3. the tablebase added nothing of real significance to the earlier analysis! The basic technique for winning the endgame is always the same. providing the black sub-optimal. Having confined the king in the Despite Troitzky's analysis.Th e Greatest Ever Chess E n dg a m es blebases. for ex promote before the knights are ready ample. but that the pawn needs to be blockaded was unable to win it (see Norman far enough back to ensure that this is Lilienthal. A few It is important that the pawn cannot years after the present game. Hastings 1934-5). from easy to handle well. Dr lows (the diagram assumes that White Seitz plays the position in almost ex is the stronger side and that we there emplary fashion. win. "after discovering that the pre-database work of Troitzky and other analysts was astonishingly accurate". it tends to be drawn on the pawn. extra moves to the mating process. He then drives the defender's versed colours. in the enemy pawn with one of his knights. Nunn decided to omit this ending from his book. In practice. behind which the pawn must stand is because six years later. else White will just push his pawn (Black can never afford to capture it. but with the white king on the back rank . and �as. note. �cs 67 �a6 ll:\d7! 68 �as which complicates the job of driving �b6! 69 �a6 ll:\c4! 70. pre lbe4+ venting the white king from fleeing. The two knights create an im 67 �b4 ll:\ed6 68 �b3 �c5 69 �c3 penetrable barrier on the kingside. 75 �e2ll:\ge3 This is an interesting formation.. 'ito>d3 would mate five moves more quickly than the text. Black would con the centre and exerts less influence. It oc curs again. much later. the defender's king into the corner.. �b7 �d6. as it is a very typical one for this ending.. tinue 66 .. of course). The pattern shown over these last few 10 �d3 �d5 11 �e2 �c4 72 �3 ll:\f6 moves is worth remembering.. etc. One fac 66 �C4 tor which makes this position slightly The main alternative defence was to easier than some is that the blockading flee towards the a8 corner with 66 knight is quite centrally placed. so the job of corralling the white king has to be done by the black king and the This is a good illustration of the last other knight. fence. With a rook's pawn.see moves 83-8S below. As a general rule of thumb. The knight on fs can never move until stage three is reached. but this is of little practical significance here. that this is actually the weakest de such as d4 and e3.. 66 ll:\e4 . as Black has plenty of moves to spare within the so-move allowance. although the tablebase shows thus helps to control some squares. 73 �e2 �c3 74 �d1ll:\g4 This is one of Seitz's few inaccura cies in this whole endgame. The table base shows that 74. in the present game. two knights are We can already see how the white less effectively placed when they defend 27 . Knig h t En dings 63 �d2 ll:\f6 64 �c3 �d5 65 �b4 �c6 king is gradually pushed back. and shortens the winning proc the blockading knight is further from ess by several moves. If instead 87 Wh2 Wf3 88 'it>g1 ..!be3+ 90 'iti>g1 'it>g3 91 fS .!Dd1 96 f6 ••• ..!Dge3+ 89 'iti>h2 'ittf2 90 'iti>h3 'ittf3 91 <ii>h 2 'iti>g4 92 'iti>h1 'iti>g3 93 'iti>g1 Now the knight heads to c3..!bd3 76 'iti>e1 'it>d3 77 'ittf2 . who scarcely put a foot wrong. 87 .'it>e3 81 '1tth 3 'ittf3.Th e Grea test Ever Chess E n dg a m es each other. 8o.. g4 and h4.!bd1 and mate next move.!bc3 79 'ittf2 'iti>e4 80 'ittf1 The encircling is almost complete..!bd3 85 'iti>g1 �e2 86 'iti>g2 . 84 'ittf1 such as seen at move 72.. If now 80 'iti>g2 then 8o. 78 'iti>e1 ..!Df3+ 93 'it>h1 . here .the two knights 84. and after the forced reply ..!bd1 83 'iti>e1.. Game 16) should have come from players who did not even hold the grandmaster title.. 93 .. and are more effective on to d1..'ittf3 81 '1tte 1 'iti>e3 82 'ittf1 . and it is time for the final stage: administering mate.!Dd5 89 Wf1 and now it is already time for stage 3: 89 ... the knight moves to take away e1: we see an exception .. 28 .!bg4 88 'iti>g2 .!Des 92 f6 .!Dh4 94 f5 . to cut the white king off from d1. g3. adjacent squares. creating a barrier.!Df3+ 95 'iti>h1 ..!bf2 mate (0-1) A splendid effort by Seitz.!Df2 Now the king is finally trapped in the comer. thus still con 'it>g1 fining the white king in the comer.. It is interest ing that two of the best technical dis plays in this book (the present game and Torre-Jakobsen.!Df2 87 control g2. However. This This is another typical formation. shows that excellence in the endgame The knight on f2 stops the king going can be achieved by assiduous study. f8 18 l:. in en tance to be able to play this. However.d2 :fc8 16 We2 e6 17 Now the knight jumps to the other l:[hc1 Ji. axb4? then .xc8 :xeS 19 J:c1 lbc1 20 side. this would Ji.xc4 28 tLlf4. tering this position. 27 f5! 31 tLlxc6 'iitxc6 29 .d2 'iitf8 22 �b4 <J. neutralizing White's protected This is the real point of White's play.g7 5 The key breakthrough.a5 This is a classic example of an ex tremely dynamic endgame... so the reply is forced. this is clearly forced.c5 a6 24 b4 f6 2 5 g4 Ji.xds 12 tLlxds 'ikxd5 13 'i'xd5 cxds 14 21 . There is no time here for quiet manoeuvring or the gradual strengthening of the position.'it>d6 30 fxe6! tLlxc6 lated a truly wonderful finish. The threat is tLlf3 tLlf6 6 Ji. Pillsbury had calcu 29.. If 30 . but the changed central pawn structure makes all the difference.xc1 �d6 21 Ji. when Black will be unable to tbd5 9 f4 Ji. 28 . Once again. 31 e7 'iitx e7 32 c7 wins.xds defend dS.e6 10 'iib 3 b5 11 Ji. Black's resignation. passed pawn. Black is threatening to bring his knight 29 c6!! to c6. Knig h t En dings 1 d4 d5 2 c4 c6 3 e3 g6 4 tLlc3 Ji... threatening the advance of the c Ji. One move earlier.e7 23 pawn. which he Black is being led down a narrow had to see pretty much all the way to path of only moves.xcs 26 bxc5 have been ineffective because of the tLlb8 reply in the game. when he would have but he had to see a considerable dis nothing at all to fear. g5 28 tLlb4! tbd3 tLld7 15 Ji. Ji...d3 o-o 7 tLle5 dxc4 8 Ji. he had to realize that this pawn 44 d6 +! and White will queen first. It is a rare case of a king and pawn end ing in which central passed pawns prove stronger than outside ones. he should be up a passed pawn on the kingside. 32 e4! dxe4 33 dS+ 'itd6 34 �e3 Now... 30 . and his king can cope with the enemy pawns. who had to have cal king over to mop up the queen side. Black threatens 37 gxhs a3 to help himself to the e6-pawn. Black has to try to make something 38 'itc4! fs 39 h6 f4 40 h7 1-o of his queen side pawns. play that anyone should think that endings continues 37 'itc4 b3 38 axb3 a3 39 are boring. His two cen tral passed pawns are secure. we begin to see the whole of Pillsbury's idea. setting one who does think so. 'ite7. Perish the thought After the more natural 36 .Th e Grea test Ever Chess E n dg a m es White's pawns are faster) 40 gxfs hS 41 b4 g4 42 bS h 4 43 b6 g3 This is the first stage of White's grand conception. at first sight highly unfavour with check. else White will A justly famous "white knuckle simply take on e4 and then bring his ride" by Pillsbury..b3. 34.. is in fact winning. culated virtually the whole of the above 35 �xe4 a4 36 'itd4 hs sequence before entering the knight Desperation.. when it is he who would be winning. able. but in order for it to work.. but if there really is some 'itc3 fs (the only remaining try. ending. and pawn ending. but White has another breakthrough. but converted by this example. at last.b4 Threatening 38 . but he is lost anyway. xd2+ 7 'ifxd2 o-o 8 king into the centre.tt::\c 2+ 40 cJi>f3 simplified and material level. and the text gives the Black king a penetration square on b3.l:td3 .t>es 38 :Xd2 f6 15 gxf6 gxf6 16 i.l:le4 :Xe4 3 5 fxe4 39 a3?! Every pawn move creates a weak ness.xh3 20 "Passed pawns must be pushed"! truch3 tt::\f3 21 �dd1 exd4 22 tt::\xds . dealing with rooks' pawns. 23 tt::\df4 .tt::\b4 41 a3 repulses centralize his king before White will be the knight. However. White's choice was motivated by the fact that Although the position is heavily after 39 tt::\h 3. pecially important in knight endgames. It is still far from clear that able to do the same. 31 . Black starts by bringing his .cJi>f7 36 cJi>c1 c...xh4 33 . points out that White can instead play the more active knight.f4 31 l:.d2 tt::\e 7 6 tt::\b s .l:.fs Black sets his main asset in motion.xe3 d:xe3 32 .l:.l:. Knig h t En dings 1 e4 e6 2 d4 ds 3 tt::\c3 i. 39 as! •.i. The h-pawn is es 39 tt::\h 3 would have saved the game.h2 :t!4 29 f3 l:te3 30 l:!. in this case to the c3 tt::\fs 9 g4 tt::\h4 10 gS cxd4 11 cxd4 excellent outpost es.i:tbf8 24 tt::\d 3 tt::\ces 2 5 tt::\xes :Xes 26 tt::\g1 tt::\g s 27 h4 tt::\e6 28 ....e2 l:!. there is 39 .h3 fxes 17 �e3 h5 �7 %1b8 18 tt::\xe6 l:tf6 19 tt::\c 7 i. �6 12 o-o-o 'ii'a s 13 cJi>b1 'ii'x d2 14 3S.b4 4 es cs 5 As usual. Black has tt::\b4 41 a3 tt::\d3 42 b4 tt::\e 1+ 43 'ite2 a number of important positional ad tt::\c 2 winning material..tf6 37 �d2 c. but it probably offers slightly improved because knights have great difficulty chances thereof.. Keres vantages: the outside passed h-pawn.i. and time to 40 cJi>d2! when 40.l:txe3 tt::\d 4 34 . but has not changed the essentials of the position. Now White always has to worry Once again..tt:Je6! which forces 41 lDf2 (else Dvoretsky makes the very sensible 41. Black fixes the weakness of the enemy queenside 41 'it>d3 ?! pawns. so.'it>d6 White's knight raid has somewhat disrupted Black's piece formation..... and his h-pawn is ready to advance further..lDe1+ 42 'it>e2 tt:Jg2 43 'it>f3 lbh4+ 44 pawns (see move 52).. Keres points out that 41 about the black king entering via b3. but in view of the resource pointed out in the following note. when Black the black knight landing on c4.tt:Jf4! maintains great pressure. Keres instead recommends so 'it>d4 40. 51 lDh3 a4! See the note to move 39. .. to prevent 41 .. 4s .aS-a4... fixing the white queenside 41. or 'it>d2! is a better defence. Black re-establishes his pieces in a secure central arrangement.Th e Grea test Ever Chess E n dg a m es Setting up the later possibility of knight to d4 and try again. the queenside pawns being fixed by the move . it should have led to nothing.'it>f6 46 tDh7+ 'it>g7 47 tt:Jgs 'it>f6 48 lbh7+ 'it>e7 49 tt:Jgs tt:Jes 40.....lDc2+ The text-move works in the game..tt:Jgs wins the e-pawn) and now suggestion of so b3 here.. has nothing better than to return the 52 lDf4 h4 53 lDh3 b61 32 . as-a4.. 'it>e3 tt:Jg6 45 tt:Jgs 40lDh3 45 lDf2 lDf4 would actually reach the position given in the note to move 40. In order to avoid this. but it is still with the white knight on h3..lt:les+ 58 Wc3 would not get any �b6 57 lt:lf4 tt:Jg6! (such deflecting sac..ti'd4 61 Wb4 Wxe4 62 �xbS the into 56 lt:\g6+..lt:\c6+ and .. 'iii>d6 63 tt:Jgs is no help.. Nimzo rifices are a typical idea in knight end witsch plans to play the moves . now 57. rather than one... but it is certainly clear that White vents the black king coming to es.... White should whereas after the text. Wc5 57 �d3 b4 tt:Jh1 Wxe4 61 �d2 is drawn) 56 lt:\e6+ 57 . stead play Dvoretsky's recommenda 56 �d3 'it>es 57 �e3 lt:las is an illus tion 5 5 Wc3 ! when it is not clear if Black tration of the comment in the note to can win. ings) 58 lt:lh3 (58 tt:Jxg6? h 3 and the . because it pre line. therefore retain this resource and in 55 lt:lc6+ 56 We3 .. b5?! 54 lLlf4 tt:Jc6+ 5 5 Wc3 and now 59 b3 axb3 60 'it>xb3 tt:Jes (after the otherwise desirable ss .lt:les+ is a real threat.. by ad tt:Jgs lt:\c4 62 tt:Je6+ vancing his b-pawn in two moves.. "Do not hurry!" 54 lLlf4 b5 62 .. lLlh4 h2 58 tt:Jfs+ �es 59 lt:lg3 �f4 6o 56 .bs..�cs (Dvoretsky). Wb6 (62 .. note to move 53 shows. The black knight arrives on c4 with ss .... rather not 100% clear if this is enough... a sample variation is 53 .. 33 . He can retain some pressure move 51. However....�cs....tt:Jg6 56 tt:Jxg6 h3 57 with decisive effect. view of the fork tt:Jg6+.. white a-pawn is too dangerous) 61 Nimzowitsch sacrifices a tempo. where.. As the �a3 68 �bl and the ending is a draw. he realizes knight cannot get back to stop the that the knight check is best delivered pawn) S8 . wcs (ss .�es runs into 64 tLlf3+) 63 lt:lgs �as 64 a4 lt:ld2+ 65 �c2 lt:lxe4 55 lLlh3? (the last chance) 66 tt:Jxe4 �xa4 67 tt:Jgs This is the decisive mistake. he is lost. Kn ig h t En dings A typical endgame subtlety. in has excellent drawing chances.. the white There may be improvements in this knight is useful on f4.'. Ex than f4. This is shown by the variation tending this line.. since 63 .wes walks 60. Now Black trades his a-pawn for the A beautifully played ending by enemy e-pawn. 62 c. Desperately passive. Nimzowitsch. and one of the most fa wins the white knight. 34 . but after 60 The proximity of the edge of the board 'litd3 l2Je6 he would have to play 61 Wc2 means that the knight has too little anyway. 64. and now the attack The centralized black knight is ide switches to the white e-pawn (two ally placed. both denying the squares f4 weaknesses!). e.'litc4 'iii>xh 3 67 b7 lDcS+ is the main point. more importantly..'it>f3 65 b5 c. and also ttJgs and the h-pawn marches forward.Th e Greatest Ever Chess E n dg a m es 58 axb4+ 'i¥txb4 59 'litc2 l2Jd4+ 60 'iii>b 1 knight i n trying t o defend against a RP.. we see the helplessness of a chess history. trapping the knight. 62 'litd2 Wd4 63 l2Jf2 and gS to its opposite number. 66 b6 61.Wc4..l2Je6 The black king is heading for g2.g.t>a3 'lifd4 63 'litxa4 'litxe4 60.. ing in to b3. In this last mous knight and pawn endings in phase. stopping the enemy b-pawn.. and. to stop the enemy king com room for manoeuvre. The alternative 61 'it>c2 is met by 61 ..t>g2 0-1 61 'it>a2 White decides to go after the a pawn. 64 b4 64 l2Jg1 'it>e3 is no help. in an attempt to create some counterplay... After 30 lt:lb4 (30 lt:la7? b4 20 l:tac1 lt:lb6 21 lt:le1 �ac8 22 lt:ld3 leaves White's knight trapped on a7) l:i. but that side of the board. . without it being all that i.. The knight anticipates White's threat of 33 'it>b4. .xg2 26 'it>xg2 the game. and White will have more space. l� M�fn. . i. . knight endings. Even so.•.'IJU�c�A.xd8 knight. once he advances e2-e4. Kramnik wins it against his powerful there are no weaknesses to attack on 35 . begun at move 28. .. In this one.d s 18 .��. the position aiming to seize additional space on should objectively be a draw. 33 tLlb4 lt:lc7 34 lt:lf1 'iti>d7 3S lLle3 tLle7 Four-knight endgames are not very common. .as 27 lt:lc6 a4 28 e4 lt:le8 lt:lbd7 12 lt:lb3 i.:>in8 29 �f3 lt:le7 looks a logical lUeS 15 �xd6 cxd6 16 'tWas . the black queenside pawns are somewhat weak. 1 lt:lf3 dS 2 d4 lt:lf6 3 c4 e6 4 g3 i.. • ·• . :"'. His pieces are more active.. .xc2 23 .g2 0-0 6 'it'c2 dxc4 7 'ii'xc4 a6 8 �f4 clear what Black does wrong..<)��t MO:Sc..'.e 7 5 opponent. . space is often a significant factor in White starts a kingside pawn advance.l:i.d6 9 o-o bs 10 11Vc2 �b7 11 lt:lbd2 26.fc1 28 . .cb8 17 approach. .J<�aii. At the moment...... White has a small. despite the mate rial equality and symmetrical pawn formation. . · .l:tc8 24 l:lxc8+ lt:lxc8 25 White retains the same small plus as in lt:lb4 Si. .'J. Gtm�e't' ·•• .·. Kn ig h t En dings 'J.e4 13 'i!Vd2 'ii'e7 14 l:.l::txc2 .. . A typical plan in such positions. 29 'it>f3 �fs 30 'it>e3 lt:lc7 31 �d3 'iti>es 32 'lt>c3 lt:la6 Now we see the point of Black's knight manoeuvre. As 36 g4! Botvinnik observed many years ago. challenging the white lt:lbd2 i.l::[c 2 'iid 8 19 'ifxd8+ l:i.ow iOO'l' . but defi nite advantage. 'it>b6 Black's 38th ..Th e Greatest Ever Chess E n dg a m es that flank.. White does retain an edge af tack ds with .. defending ds and attacking prove Black's drawing chances. 36.. al serious trouble. note to Black's 3 5th... in establishing a passed pawn on the 42 exds! kingside.gs!? The text cedes White an outside major It is always difficult to know how to ity on the kingside. Later in the game. kingside he has a potential passed but it is not forced.by removing the white 43 .. his f6-pawn would prove a (Game 15). but Shirov was pre react in the defender's place . Black after 39 gxfs exfs 40 exfs tbcdS+..tbf4 or ... The white knight can though the simplifications should im come to f6.. Fritz 11 suggests pawn on the g-file. tbe7. White succeeds exds White still enjoys a clear plus. h7.tLlexdS+ 42 tion.fS ! ? which initially seems fine for very difficult.�c7 43 gS This is the main tactical point of 43 .. Shirov nuisance to defend.see also sumably concerned that after 40 . to hold back e4-pawn. but after ter 41 tbxds tbxdS+ 42 'it>d2 gxf4 43 the strong reply 44 tbg4! Black still has �e2 tbe3 44 tbxf4 tbxfs 45 'it>d3.. but this does weaken his posi little difference. . but the space advantage 39 tbxf4 es 40 dxes fxes itself can represent a worthwhile plus.. Black's defence is 38 . decides to fight for space and not allow 41 lLlfdS tbcxdS+ his opponent a free squeeze on the Taking with the other knight makes kingside. White is better on both of the g-pawn. after 41. thanks partly to this advance Suddenly.dxes the comments in Fischer-Taimanov 41 tbhs.. whilst on the The exchange is not bad in itself. tbg6 looks natural. His king still threatens the en 37 tbd3 f6 38 f4! gxf4 emy queenside pawns. Another difficult choice for Black... wings. he gets the use of ds for his the white h-pawn and threaten to at knights. With his king tied down on the 36 . 42 . Jbs and the d-pawn pro and White's kingside pawns will be motes. 56 �e6 lt. 47.Jf4+ (this coun terattack is insufficient.Je1! (holding 50 h7 lt. whilst after 47...Jh6+ 60 The king march to e4-fS-f6-g7 wins 'it>g6 lLlg8 61 h s and wins. 37 .d4 59 'it>f7 lt.Jf3 lt.Je7. pressed to hold the position. 'itfs.Jc3+ 51 'it>fs e4 52 g6 The only way to stop the h-pawn.Jf8 loses at once to 48 tt.g. the A highly impressive squeeze by black queenside pawns are still not Kramnik.li.Jgs e2 54 li...lt.. Meanwhile.. 44. stronger than Black's unsupported cen 48 g6 hxg6 49 h6 g5 tral ones.. can bring his king to e4 and start push ing on the kingside.Jf4 47. Black would be hard far away.Jd1+ etc. 51 . White 'it>xh8 b3 57 'lt>g 7 b2 58 li.Jg8 51 tt. e..Jfs lt.Jg6 47 h5 ple..Jg6 51 �d3 1-0 up the d-pawn) S8 .Jxh7.Je2 so lt..Jf6 'lt>cs 46 b4+ axb3 47 axb3 b4+ 48 'it>d3 li. easily. and a knight sacrifice on h7) 49 'lt>e4 lt. e.g. li.Jxds 55 g7 lLle7+ albeit only temporarily. Kn ig h t En dings other flank..Jh8 5 4 'it>f6 '1t>c3 s s �g 7 b4 56 If Black does not exchange.... e3 53 tt. :�b6 45 li.li.Jc2-a3 will win the bS-pawn. as the moment his king strays too pinpoint a clear error by Black. For exam 45 axb3 'it>a5 46 h4 lt.Jg8 57 h4 dS 58 lt. in which it is very hard to safe.Jxd6 b4 52 tt.Jfs 'it>b6 49 lLle7 followed by so g6. but there is no hope after the passive 48. the simplest win is 48 �d3 'iii>b4 49 'iii>e4 �xb3 so tt..lt.Jf8 49 'it>e4 followed by h2-h4.li.�b4 52 'it>e4 'iii>xb3 5 3 44 b4! axb3 �fs li. so . combined with game assets are important here too . etc. its characteristic of bishops is their colour activity is restricted and it risks becom blindness . viz. against a world-class op ing the pawns on the squares of the ponent.hop Et\di. allows Yanof passed pawns. too. The ending illustrates espe- 38 . 1. Chapter Three Bis. depending on degree of control over those squares whether the bishops operate on the too. the defender suffers from the bad Same colour bishop endings are the bishop. Finally. This game also shows a typical purposes: device in such ending. In same colour bishop endings. the sacrifice of the bishop for several pawns. same colour squares.a bishop can only cover ing a "bad bishop". king sky to win a position with equal mate activity. bishop endings is the placement of the Game 9 sees White able to win by pawns. Black is place one's pawns on squares of the handicapped by his queenside pawns. If the pawns become fixed on the posite colours. A particular issue in rial and symmetrical pawns. although here. In both cases. the player can establish a distinct categories.ngs' Bishop and pawn endings come in two other colour. The usual end Black's dark squares. By plac advantages. or squares of op 2. This serves two bishop. bishop means that it can only control who exploits the smallest of positional half the squares on the board. The consequent weakness of more straightforward. The colour-blindness of the Game 10 is a virtuoso display by Larsen. which means that its influence is always limited to Game 8 shows a typical case. where only half the squares on the board. pawn weaknesses. squares of one colour. opposite colour from those on which fixed on the same colour squares as his one's bishop travels. White's more active king. virtue principally of the much more one should almost always strive to active king. the key same colour squares as the bishop. which quently allows the defender to draw illustrate this. whilst the enemy bishop cannot notorious for their drawing tendency. Bishop Endings cially well the battle between mutual his bishop. in order to cre colour bishop endings is the pawn ate his second passed pawn. This secures them from at passed pawns. colour. several files away from the first. or even more pawns example of a combinational break down. tack. By defini port whichever pawn the enemy tion. Game 12 placement. the stronger side usu endings is that of the fortress. tions with bishops of opposite colour. one of the finest-ever in same-colour bishop endings. attack them. Since ally needs to create a second passed bishops only control squares of one pawn. The stronger side fortress based on the colour squares then needs to get his king over to sup which his bishop controls. weave a magical win against a grand the defender should usually put his master. through. although this should not be exagger In order to win an opposite-colour ated. we have two classic fight on those squares. he will have an extra piece over bishop is stopping. in posi handlers of opposite-bishop endings. since his own bishop can defend Opposite-colour bishop endings are them. 39 . Contrary to the usual rule shows Karpov. his opponent when it comes to the In this chapter. where Black is prepared to An important detail about opposite sacrifice several pawns. from a completely drawn posi pawns on the same colour squares as tion. and this fre opposite-colour bishop endings. it follows that opposite This will usually force the defender to coloured bishop endings allow the de block one pawn with his king and the fender the possibility of setting up a other with his bishop. Game 11 is a famous positions two. The fundamental concept in such bishop ending. c8 38 tt:Jd4 his bishop over.. f4 (in view of the next note.e6 37 lLlb3 i.d 1 lLlf8 10 .i. wins the d-pawn) 49 more active king and the better bishop.f7 46 �as <Ji.. However.g s (48.Th e Greatest Ever Chess Endg a m es GtJme B ..h6 (the 40 .td7 30 'iW1 'it>f8 31 �e1 �e7 32 'it>d2 forced anyway..xc7 lLlxc7 29 lLld4 argues that the move is practically .....te2 c5 22 bxc5 bxc5 23 dxc5 tt:Jxc5 24 43 . but then he allows the white tt:Jxd4 42 exd4 . by and Black is close to zugzwang. The threat is g2-g3.. fol 'it>d6 33 �c3 f6 34 �b4 . i. . wise.e6 39 . Yanofsky proves other 48.td3 tt:Je6 20 'ii'b 2 l:tac8 21 kingside pawns on light squares ..e8 9 Black also has two of his kingside . but Yanofsky i.l::tc7 18 stage of his plan is to fix all three black b4 g6 19 . very fine play.tc6+ <Ji.i.l::txe7 26 l:.i.ie8 35 .f7 36 a4 i..e7 5 on dS.i.. to defend the kingside .'it>a8 49 . ibg5 o-o 6 e3 tt:Jbd7 7 Vi'c2 c6 8 a3 .txc4 lLld5 12 pawns on white squares.i.b5 lowed by h4-hS.d3 �ec8 16 lLlxd5 exd5 17 i.D!tallqfSkJ.d7 44 g3 .l::!.ic8 king into a6. Yanofsky gives the varia tion 43 . The first ..i:..c7 47 �a6 'it>b8 48 as Yanofsky notes that the game was adjourned in this position.i.i.xc8+ Coming rather quietly.td3 dxc4 11 .d7 40 h4 tt:Je6 41 .e8 4S .bs .d3 f5 1Wb4 tt:Je6 25 1Wxe7 .. 49 i.b8 so �bs.f5 .d3 .xc8 27 l:. and Yanofsky i.i. and the other competitors in the tournament all thought it was drawn.d3 h6 principally because of the black pawn so �bs and �cs is simpler) 49 .i.i.i.b5 pawns. Black can try to bring .xe7 'ii'xe7 13 0-0 b6 14 .Piltll� · Vef1tnar city 1942 1 d4 lLlf6 2 lLlf3 d5 3 c4 e6 4 lLlc3 .l::!.c1 l:tc7 28 . when the a-pawn or d pawn will fall.. which blocks in his own prelate.c1 ibb7 15 hurries to capitalize on this. White has the advantage of the followed by �cs..l::!.i. White's 48 hs main plan is to play h4-h5.�b6 so aS+! �c6 41 .e 8 and there is other wing.d7 46 . loss of material.i. as Nigel Short can now prepare his h4-h5 break. keeping no defence to the threat of .....e8. Yanofsky therefore drives the enemy bishop off its ideal square.. White will have a pleasant choice between ex changing pawns.. but if he prepares that with 46 ..b S+ �d6 51 i..Wc6 �bs-cs.g6! which probably draws. but now the bishop loses the h-pawn. note to move 43.i. i. The king and pawn ending is of At present. and after 48 . 48.e8..a4! able to reach the e8-h S diagonal..bs! .i. If instead 46 .i.i. 46.c8 The latter creates various tactical Black chooses to maintain the guard threats involving the promotion of the of a6..i. White is the full "Volkswagen".. 49. defending the .e6 4S g3 i.i..i. from where it is able Wxd7 so WcS �e6 51 Wc6..i. then Black has 46 . for example 48 ..i.e 6.. or pushing on to h6...d7 49 . 47 .x hs moves in on the queenside instead: 47 Now Black has a further weakness �as Wc7 48 Wa6 �b8 49 i..c8? so i..�d7 ally placed on d7. h s would obviously create a fa 44 f4 tal weakness on g6.xg6.i.. gxh4! 50 gxh4 47 i.gxhs If Black does not capture.. in view of the threat 44 .i. calls it: so ....i.. the option of coming to f7. the black bishop is ide course lost after 48 .. as White's With the enemy bishop no longer king gets stuck on a6) and now so . then White 49 i.i.e2..bs Immediately fixing the pawns on there would already be no way to avoid light squares. forcing it to choose an inferior post.e6 5 1 ..e 2! ... creating a fatal weakness on g6. Bis h op En dings computer points out 49 ..e6 49 h6 chance to come to e8.e8! followed by 47 . .e 8 as in the on fs.i...i. to swing over to either wing. He tries to make things difficult.Th e Greatest Ever Chess E n dg a m es Yanofsky points out that 50.b 5 63 a6 would also have won (63 . Yanofsky gives the further line 64 .c8 the Canadian Champion..c8 56. and if the 64 <j.'ot>a6 This condemns another pawn to leads to a comic state of helplessness oblivion..e8+ 'ot>d6 52 'ot>b5 62 . move without allowing the white king 62 ..h5 'ot>d6 54 i.. collection (alas..e 2 i.i. but Yanof 56 i.c8 67 ..d7+ 58 'iii>a 6 <j.i.a4+. other move. the position is ob white king into a6.a6 'itc6 for Black after 51 i...e 6 61 <j. then White frees his king by means of 51 i. and can be highly 60 i... played endgames.e 6 66 . but Yanofsky's own book gives bishop moves to d7 or e6.i.a6.c2! h5 recommended. After 64 a6.c8 62 i.xh5 <j.'it>d6 69 'ot>b6..i. .dl 'itc6 the enemy camp. but once these days) contains a number of finely again zugzwang defeats him.i.d 3! 65 .'ot>e7 53 i.b 7 52 i...e6. h6 57 i.....g6 .xh5 i.c 2! sky prefers not even to allow Black that Zugzwang again. Either way.d3 i...i.. The of truth by one move.f7 'itc6 But that only staves off the moment 68 i.i. The king cannot forlorn hope of counterplay...i.... Now Black must pawn promotes after 68 . it allows the the text.xa6? 64 ..e6 64 a6 1-0 into c5. 52 .e8 i.i. make a concession. 42 .... who was a Black has been forced to surrender a noted endgame expert.i. a bibliographical rarity by boxing in the white king.. i. His best game pawn.. but if 60 ..f7 i.b 7 55 i..i.i.g 8! with another zugzwang. . There is only one viously hopeless for Black...a4+ 62 i.c7 59 'itxa7 .. the bishop cannot go to b7 be Megabase gives the final move as cause the f5-pawn drops. A splendid piece of precision play by 57 .'ot>c7 63 . White is gradually able to inch into 61 .e2+). c8 18 to do so.xa3 'ii¥x a3 27 minor piece endings .fdl ..b2 lt:le4 13 llc1 �e8 14 lt:lc3 lt:lxc3 15 seems unlikely that he should be able i.e2 i.f7 31 b4 i.i. Mean.g6 .b2 .d6 22 i.c8 42 a4 49 g41 i. 54.. . � � .a 3 25 . ..�d6 so Wf2 We7 51 �g3 �f8 52 tration of that rule applying in a bishop Wh4 i.e8 41 �cs i. ·. .d6 12 the kingside (two weaknesses!}.l:f. .F�•a·> ..l:[ec8 24 nally.i.c8 35 f3 �e7 36 Averbakh's handling of the end �2 �d6 37 .c8 Fixing weaknesses . "' ' t [ � � .i.l:f. Black has two weak pawns Tying the black bishop down to the on a6 and c6.S. to the defence of which defence of a6. it is a sig nificant advantage to have the more active king.. Fi i.t.d31 endgame. exd4 Wf8 34 i.e6 17 . ' . White will penetrate with his d4 dS 6 cxds exds 7 i...AverJatti..xcs 9 o-o o-o 10 b3 a6 11 lt:la4 i. . 1 c4 cs 2 lt:lf3 lt:lc6 3 e3 e6 4 lt:lc3 lt:lf6 5 while.d 3 h6 44 aS+ �b7 45 . ··� ·�.xc3 .i. It . .i. Bis h o p En dings .i.l:[xc6 bxc6 26 .l:te1 1.U.c7 16 'ii'd 3 i.' ' � ' ' ' ' .e7 47 .hampionsllifiSerrii-fil(..c8 54 . �d2 Wb6 40 a3 hr. . we see another typical device in g3 i. In almost any endgame.c1 .e 2 i.e 9 .b7 38 We3 Wc7 39 game is technically flawless. lJSSRr. · Y...xd4 i. ...f3 .l:f..now the enemy 46 i. and White can prepare to infiltrate them with his king.e6 32 'it'd4 'iVxd4 33 pawns. but with the aid of zugzwang..�7 43 . qd�ssa 1960..i... 'ii'bl "ifie7 19 'it'a1 f6 20 lt:ld4 lt:lxd4 21 his king makes it all the way to h8.i.the sacrifice of a �d4 'ii'a s 28 �a4 'ii¥b6 29 'ii'f1 l:ta8 30 piece.d6 8 dxcs king on the weakened light squares on i. and this is a perfect illus 49 .e6 53 Whs . .as .l:f..lxe1 48 �xe1 �c7 kingside pawns are fixed in their place.. to net a winning collection of 'itd3 i. his bishop will be tied down.l:!.d 7 43 i..· ' y . ..c6 23 . h51 Preparing a further zugzwang..g6+ �8 62 .<. White sees that his bishop will no longer need the square fS. since s s . in order to destroy the of the earth..'iii>e 7 68 h4! (not 44 .. on an empty square (always the most aesthetic form White's king has made it to the end of sacrifice). Desperation... Now the next stage is to black kingside pawn structure. Setting up the final breakthrough.c8 64 f5 i.fs.ti>xe8 67 'iii>xg7 h5 g7.b7 results in his bishop being completely immobilized after 56 i. If 67....i.. 59 �f8 60 'iit h 7 �7 61 i....d7 65 i.i. force Black to undefend the pawn on 66. so he can advance his pawn there.d31 65 ... 'iith 8 66 i.•.. to squeeze Black further. Black is in zugzwang and must give ground. 55 .�8 56 �g6 �gs 57 i.i..d7 63 f4 Gaining further ground.. 57.The Greatest Ever Chess E n dg a m es 62 . 55 h31 Zugzwang.�f8 58 i.c8 Once again. The bishop sacrifices itself.e2 �g8 59 i..f1 63 . Now Black's king must give way.e81 A beautifully elegant final blow. i. A beautifully elegant ending by the great Soviet endgame expert.'1W8 70 gS �g8 71 g6 'iW8 72 �gS .c8 74 'iiifi> 4 is no better. ..d3 78 'iit>xc6 'iit>f7 79 'iiid i> 6 . . White's own king to penetrate to the queenside ..f7 76 �es .i..i. and a highly n ..i.i.d7 73 f6 8o bs! 1-o Now the a-pawn will decide. then The two connected passed pawns after 79 .. typical endgame ideas. If the bishop stays on the a6-f1 di 68 �xf6 hxg4 69 hxg4 agonal.d7 74 'iitfs �g8 75 g7 .... Bish op En dings 68 �xh6?? �f7 with a draw) 68 ..i.e8 instructive demonstration of some 73 .. 45 ...i.fs will promote.i.c4 White wins by 80 g81ii'+ ! will divert the black king and allow �xg8 81 �e7.two weaknesses again ! 69 .i.i. to prevent White's next.g6 77 and now simply 69 g S ! and the f-pawn �d6 . White has a space advantage. the game is instruc to create a target on f4.i. Secondly. if one has the determination 25 h3 �8 26 �2 gxf4 27 gxf4 f6 28 46 . the tive in demonstrating that even appar downside is the weakening of the black ently drawish positions can be won. the first stage of the answer is dxe4 16 tt:Jxe4 tt:Jxe4 17 dxe4 .Th e Greatest Ever Chess Endg a m es 1 lt:Jf3 lt:Jf6 2 g3 ds 3 i.b2 h6 7 d3 . the black bishop on h7 is currently shut out of play by the white e4-pawn and will take some time to emerge into play. and also hopes the latter phase.l:.U. the latter is b3 e6 6 i.l:txd1+ 22 �xd1 'iVd8 23 'Wixd8+ couple of small straws.fs s to try. White does have a 21 i.e7 8 e3 i. 24 'it>g2 gS At a casual glance. If one asks "How does one win 'iVe2 as 10 a4 tt:Ja6 11 lt:Jc3 tt:Jcs 12 tt:Jes such a position against a strong oppo 0-0 13 f4 lt:Jfd7 14 lt:Jxd7 �xd7 15 e4 nent?".g2 c6 4 o-o i. and it is indeed hard to believe that anyone could win such a position against a world-class grand master such as Polugaevsky. the position looks dead drawn.ad1 'Wic7 19 'it>h1 l:txd1 20 l:txd1 l:td8 In playing on.f3 . opponent.fd8 18 "By not agreeing a draw!" . h-pawn. where it was attacked by the handling of the ending. from which to i. . which later becomes very im with only the smallest error from the portant. Quite apart from Larsen's superb from g7. this is one of the reasons why I have selected it for inclusion in the present With this. However. Black removes the pawn book.h7 9 the key. In the first instance. However. thanks to his pawns on e4 and f4.xd8 weave his basket. especially in white queen's bishop. Firstly. a3 .te2 with the idea of transferring the pressure after 39 eS+ fxeS+ 40 fxeS+ bishop to c4 and threatening f4-fS. . 36 �e3 . White commences Black continues to use the weakness a plan of going after the black h-pawn. 38 .tfS . to tie White down and hamper 29 .i.... of f4.tel Wd6 35 .txd6 Wxd6 42 <ifi>d4 eS+ it is White who has weaknesses on the dark squares.tg6 29 ...e2 b6 40 .td4 33 ...�d7 30. Bis h op Endings �e3 .i. but White could 37 . �e7 was also possible.lib6+ 37 .d4 30. :iiif7 would keep the enemy 37.�e7 30 .lte4..c7 bishop out of f8..i. since after 41 .. <ifi>c7 41 ...i... With the light-squared bishops hav- 47 .lib2 Now the exchange would be less dangerous for Black.c5+ �d7 39 .i.txe4 .txf5 nerable to a later white king march to exf5 h4...c5+ his attempts to make progress. Now the h-pawn is potentially vul 43 ..d3 fxe4+ 44 .tc5+ 42 'iii>f3 f5 Passive defence with 42 ..td6 41 . when White retains some .... Polugaevsky prefers to Black's position would be rather pas retain the dark-squared bishops.ta3 .lixd4+ 38 <it>xd4 <ifi>d6 was the strengthen his position by means of 31 alternative..tb2 �e7 With this move.... be sive and still somewhat uncomfortable.i. lieving that it will be harder for White 31 .tc7 34 ..tb6+ 32 Wd3 h5 to penetrate the black position with both bishops on the board... 41.i..tf5 45 . but Polugaevsky prefers to clarify the position. . The position ter his opponent had been able to ana should still be drawn.d2 Polugaevsky later wrote that he had reached this position in his adjourn ment analysis. to his endgame technique. Wf7 venting the white king from coming 49 i. is a remarkable tribute problems to solve.d 8 We4 51 i.d6 so i. 47 . convinced that the strength of his passed f-pawn should be sufficient to draw..e1 had been adjourned around move 41.... in par ticular..h6 53 'iii>xhs f4? This natural move proves to be the decisive mistake. Passive defence by 48 . It is worth noting that the game 54 �g4! i. .... 46 �g3 i.xf4+ 52 Wh4 i. but Black has lyse it at home. draw. followed by �h4.. even af vulnerable after Wg3-h4. pre terattack...te3? 55 i.x a5 f3 56 Wg3 wins so Polugaevsky was now following his easily.g 7! White's plan is i. 49 i.gs i. and had stopped here. .�e6 48 i.h 6-g5. This simple variation illustrates 48 ...g 5 i. although it takes some superbly subtle play by Larsen to 48. Polugaevsky later claimed Polugaevsky puts his faith in coun that 53 .c5 should also be adequate to back to help fight against the f-pawn..xb6 i. his ability to continue to find ways to pose the maximum problems for Black. which can prove managed to win the position. 54 .�d5 prove this.<it>f3 ! would have drawn.. and..e7 47 i. The fact that Larsen still ries over his h 5-pawn..Th e Greatest Ever Chess E n dg a m es ing come off.. Black still has some wor home analysis. .f1'ii'+ 58 �xf1 . 58 . However.. Bishop En dings a key point about the ending . square e2.. and cannot stop both passed pawns. the only route for the Instead. the black king returns to perform defensive duties in stopping the white h-pawn... that Larsen's play was a flawless dem 6S ..Black's is true.ixf2 63 h7 This is the point. After able to patrol effectively with his 58 . remains on c2.. as the following variation shows: 59 ... but is note to Black's 62nd move. �g s. 'iitg 3 ! wins at once) 59 h S ! (59 as? b3!) 55 h4! f3 56 �h3! Black simply cannot cope with both passed pawns. .. by setting up a king is the square e3.ie7? Preparing to advance the h-pawn.. al h8'ii' b3 69 'i!Vh7+ 'iitc 1 70 'it'xa7 b2 71 though the first part of the statement 'it'cs+ and White wins in standard fash- 49 ...g..ixb4 axb4 60 bishop. 56..ixh4 59 .�d2 66 a6 �xc2 67 a7 .xf2 ...ic5 'iti5 Since he is unable to support the further advance of his e-pawn.'it>e3 60 ii..'iii>g s (59 .g. . 58 b4! wins. the second is not.ixa7 68 onstration of that fact.. All previous annotators of this ending.... . which White is second passed rook's pawn.f2 is not . 59 . In view of the f-pawn looks very dangerous.. e. Now S6 .. have assumed that break 64 b4! axb4 65 as and the bishop Black was lost after his 53rd move. dangerous because of 57 �g2 and after S7 .. ... Larsen's last only so if the black king can reach the has to be considered a serious mistake.ib6 cs 61 as c4 62 a6 b3 63 a7 b2 64 a8'i!V b1'iV 65 'ii'd s+ forces mate. 58.�f4 57 .cs+ �e2 (or 60.�e4 59 h5 The black king has finally gained ac cess to the square e3. axb4 (58 .. without allowing .id4 and the h-pawn pro motes) 61 h6 f2 62 Ji..id4.id2 60 �g3 �e4 61 �f2 also wins) 60 ..ixb4? 59 ...ixas White wins easily. e.. but it is now too late. including Now there follows the decisive both of the players. All the while the white pawn which should cost the win. .'�d2 61 h6 �xc2 62 . .f6? 'iit e 3 65 along which to control e3. i.-tes?? 62 .i. by forcing the black king i n front of the b-pawn and then bringing his own king closer. the white h-pawn has on account of 63 i.. including two for the last move. Both Larsen (in the Danish magazine Skak bladet 1966 pp 77-79) and Polugaevsky 6o i. i... threat of 64 . and the h-pawn threatens to ad 6o.e3! is the way to draw. and in close off the e3-square once again. 59. awards Larsen's last four during the preparation of this book. it has not marks. Larsen adds the moves 63 . apparently overlooked by both After the text move. .g7. Polugaevsky. king.. to Larsen's best games collection.d4 is also not winning for White.b6!) renews the queen ending.h4 i..e3 61 i.. who was extremely indeed. . and 62 64 h6 winning for White. here.i..xg7 64 merciless computer points out that hxg7 f2 when Black should draw the 63 ... he may well be losing) 64 . Black has a simple draw else.d4 62 . Black is defini players.. as far as I am aware.e3 generous in his commentary on this etc. of moves after 64 i. In Grandmaster Preparation) dismiss this.cs (in his Preface to the Russian edition of The bishop returns just in time.d2 62 ... when the e3- advanced one rank closer to promo square is again denied to the black tion. and leaves White 62 i.. analysis.d4 vance.i. i. This possibility only came to light ending.h6 (64 i. the meantime. i. -tcs ! (or 63 . However..i.Th e Greatest Ever Chess Endg a m e s ion. moves a total of five exclamation and.f8 i.'iit e 3 is again a threat... 63 i...g7 fails to 63 . the h6? 'iite 3 63 ii. previously been pointed out anywhere but in fact. at the board and in subsequent tively lost.h6 with nothing better than a repetition The bishop finds another diagonal.. 'iit e 3.d2! 50 .. . but it is too late. whilst the black will be too much for Black's own king has no waiting moves that do not bishop. despite Again. 63 . and also preparing to annex the as 68..tg3 .c3 73 as 1-0 66 .d 2 68 i. If Black's bishop bishop for his e-pawn....h4 i.txas 'it>e3 69 C4! Black will finally win the white And here it is. i..We3 his king into contact with the f3-pawn... be driven into zugzwang. 65 i...f2 70 i. . Instead...f2 69 ... A lovely final touch. it allows one of the white because the two passed rooks' pawns pawns to advance. . Bish op En dings Anticipating the threat of 63 .tf4 with a draw. . Black will the lapse at move 58.. all of lured to cs..th4? i. 51 .g s 67 after 71 .. 71 h6 'ii?e 3 72 h7 'it>d3 73 as and wins. after which White will have two Or 68 . . the most precise.ixf2 cJi>xf2 70 c3 i...tf4 64 .x c3 passed rooks' pawns... lose the f-pawn.xf2 cJi>xf2 71 h6! 66 cJi>g4! All beautifully done.g3! Superlative play by Larsen.cs pawn. White brings 71.ixh6 72 as . h 6 67 i. Now that the black pawn has been with gain of tempo. closing the a7-g1 diagonal. 72 h7 i....tf4. moves.e1 'it>e2 69. White's subtle play hitherto would Black's bishop cannot stop the a-pawn come to naught after 66 . e4 The defender essentially has two de fensive schemes.xe3 i..The Greatest Ever Chess E n dg a m es 1 d4 ds 2 e4 e6 3 lLle3 lLlf6 4 lt::lf3 e6 5 e3 of opposite-bishop endings is the for lt::lb d7 6 i. a single h3 "iVg6 28 'ti'f2 hs 29 'ith2 as 30 i.. and rely on his bishop to defend the other flank.xd4 19 'ii'xd4 i. on 'ii'e8 40 :d2 'iixa4 41 1i'h4 l::te 2 42 I.e s b4 32 .lxe2 the other flank. duction to this section. but with 21 . the main theme so that it can defend them.. One is to use his bishop to stop the passed pawn and his king to defend the other wing.d3 i.d 4 i.e4 35 l:tddl li'fs 36 "iVe2 nesses" principle is paramount: Black 'ii'g6 37 'fif2 a4 38 :XeS+ l::txe8 39 bxa4 has to create a second passed pawn. although it has been re 47 <it>g3 <tA7 48 h4 produced in countless endgame text As usual in opposite-bishop end books.d2 i.a 3 passed pawn is frequently not enough. b5 31 i.. 52 . is to deploy his king to stop the passed pawn.e4 26 'iid 2 'ii'g4 27 other pieces on the board.:eel l:. the "two weak i. In this case.xg2 46 i.de8 33 i. the winning plan is usually to break through to the passed pawn with the stronger side's king.. and in this game we see a classic 9 b3 'fie7 10 "iVe2 es 11 exds exds 12 example of breaking down the defen dxes lt::lxe s 13 lt::ld4 lt::lx d3 14 'it'xd3 'ii'e s sive fortress. Vxg2 i.es 18 Black has an extra pawn.. which Kotov uses here.l:te3 lt::le4 22 l:tee1 b6 23 l'Ud1 f6 24 opposite-coloured bishops and no lt::le3 lt::lxe3 2 5 l:. so as to overstretch the 'ii'xe2 43 1i'g3 "ii'x a2 44 i. practice.fs 20 i.e2 34 I:!. In that case. the key to This is one of the most famous op success for the stronger side is to cre posite-coloured bishop endings in ate a second passed pawn.b4 'ii'd 7 outside passed pawn at that.b4 7 0-0 0-0 8 i... 15 f4 'fie7 16 l:tae1 �d8 17 :e2 i. the defender places his pawns on so instructive..xf6 'ii'xg2+ 45 defence. and force the win of the bishop.. it bears repeating here since it is ings. As we said in the intro the same colour squares as his bishop. and. The second defen sive plan... and an lt::la 4 i.d4 In such situations.d 2 i.d6 tress. by This natural move is the fatal error.'it>xe3 59 'it>xb3 'it>xf4. On the 'it>xf4 62 'it>c3 is the same dead-drawn face of it. with ac matters. but it takes only one slip by Kotov to 59 �cs? allow a combinative breakthrough. ing play.b2 60 'it>c2 Wxe3 61 'it>xb2 The bishop is heading for e6.. this is previous note. but in reality. �f6 56 �f6 'iW3 57 . Black wishes to have his bishop After the text. if Black's king ing position would be a dead draw. Meanwhile... to con Now the white king heads off to the trol the square g8. passed pawn. The point of this last queenside. as 59. this appears a more passive ending we have already seen in the square than e4...g6 49 'iW2 advances to b3. Instead.. for two reasons: one of the most famous combinations firstly. to defend the b-pawn when it in endgame literature. endings. and secondly. as the presence of passed curate play by White that is the case. Bishop Endings 48. His bishop holds could win a second pawn with the h4-pawn.. not so. 53 . it is It looks as though Black can make no not so much the number of pawns that further progress. the square g5. but the result to defend the e3-pawn. After 59 �e7 . in such endings. and can also come to d4 58. gress.te7 b3 58 Wc3 . Black scheme in operation. factor will become clear in the follow 49 'it>e6 so 'it>e2 'it>fs 51 �d2 'it>g4 52 •. and. pawns.tfs! Wd2 Black has no way to make pro The start of a deep regrouping idea. the stage is set for on the a2-g8 diagonal.. The bishop had to maintain guard over 52 'it>g3 53 �e7 'it>h3 54 . the white king typical of opposite-coloured bishop stands guard over the enemy b-pawn. attacks it. which Black secures a decisive second allowing a combinative breakthrough. as we will see.tf6 'it>g4 55 •. indeed..te6! Here we see White's defensive As Botvinnik pointed out. or 64. and leave the bishop to control 60 hxg5 is obviously hopeless. just by remaining . The bishop can control them approach. it is worth pointing 63 'it>d2 'it>h3 ! ! (the obvious 63 .. so that the white bishop cannot defend the h4-pawn along that diagonal.i.. However.65 .i.f6 and the pawn is whilst they remain unsupported by stopped) 64 'it>e2 (if the king does not their king.xd4 is interesting. then just 64 ..Th e Greatest Ever Chess E n dg a m es on the short diagonal e6-g8. so as to preserve its colleague on b3..i. 60 fxg5 The white pieces now need to swap roles. 'it>g3 62 g6 �xh4. ues 61 ... A second pawn sacrifices itself. However...'it>g 3? out. but the idea soon becomes clear.i.. so the white king must try to get across to stop that pawn.'it>g2 65 . .. that the apparently fails to 64 .f6 h4! and 54 . If left unaided.'it>g4 65 dangerous white connected passed 'it>el and the white king gets across in pawns on f4 and g5 are no threat at all time . it is es pecially nice how the d-pawn advances to a square which is covered by no fewer than three white pieces! 61 exd4 Taking with the king is obviously not possible. but the alternative 61 s9 gs!! . the bishop will be lost for the h-pawn... presents Black with a lethal passed h Black is able to foil this by subtle play: pawn... 60 d4+! .. in that case. The game contin At first sight.e5+! 'it>g2 ..'it>g2 wins) 64. As Bot vinnik notes with some pride. and the unsupported b-pawn. this seems pointless. The purpose is to close the d8-h4 diagonal. h4 followed by without any trouble. . Bishop En dings the bishop is fatally overstretched.Wg3! g-pawn posing any threat. Botvin The white king can get across to nik points out that accuracy is still re stop the h-pawn.. pawn. but after 66 �f2 �f4 quired..idS+ o-1 The climax of the combination. the bishop also stops the white 61. h-pawn and drawing. 62 .. Black wins the h4-pawn and sets up a winning second passed pawn... 55 . do. of opposite-coloured bishop endings. leaving Black with the "right" rook's there is nothing more that White can pawn for his own bishop. controlling the square g8 (see the note teaches one many of the key principles to move 58): as well as defending its b. In this position.ia3 Wxh4 63 Wd3 Wxgs 64 We4 h4 65 Wf3 ... if 61.if2 keeping his wins the white bishop for the b-pawn. the importance of the black bishop which in the space of a few moves. we can again see A minor masterpiece by Botvinnik.Wg4? White has the de Black's king goes to the other flank and fence 62 dS ! .ixds 63 . After the text. as Karpov also points out. the bishops of opposite colour.fe1 'ili'f7 19 'i!Vb2 b6 20 cs bS on an aggressive square.xc6 dxc6 6 lbc3 i. and with However. This succeeds because the pawn on d3 prevents the black king having any way to break into the white position . 34 g3?1 Karpov points out that here.. or on the next move. it seems it is psychologically extremely difficult that Black cannot hope to win.a3 l:te8 30 l:!. Now there are 21 . once again.Th e Greatest Ever Ch ess E n dg a m es 1 e4 es 2 lbf3 lbc6 3 i..d6 7 d4 i. so lbfs 15 lL'lxfs i. He therefore challenges the black pawn at once.even coming round the queenside to bS achieves nothing. it takes only one or two in opposite-coloured bishop endings.xfs 16 'ii'd 2 i.. Karpov shows that by extremely side pawns on light squares.b s a6 4 i. Black seems to passed pawn....b4 8 33 .. pawn is securely blockaded. White could have estab lished a fortress by playing 34 g4 fol lowed by 35 h3... In general.xe8+ hardly afford to play 34 h3.b4 point. How to decide on a plan of fixing the king ever. Mednis even went so far as to claim The pawn structure means that that White also draws by the immedi Black effectively has an extra outside ate 34 h3. imperceptible errors from White to the defender seeks to place his pawns 56 . h41 lbxes lbxe4 9 o-o i. since White's three have no way to force a breakthrough. pawns in the centre are securely held thanks to the vital pawn on d3 which by Black's two.d s 23 f3 :xe1+ 24 threats of a possible .. leaving all 'it'xe8 31 'it>f2 'ikg6 32 'iid 3 �xd3 33 his kingside pawns fixed on light cxd3 squares.b4 as 22 i.h4-h3 at some lbe1 b4 25 a4 bxa3 26 'i!Vxa3 a4 27 i. 5 i.a3 lbd6 12 c4 f6 13 lbg4 lieS 14 lbe3 breakthrough is on the kingside.a4 lbf6 cost him the half point. whilst it seems that White can 'ili'g6 28 'it'c3 hS 29 i.e6 17 'ili'c3 Karpov immediately places the h-pawn 'tWd7 18 l:!.xc3 10 bxc3 o-o 11 It is clear that Black's only possible i. the passed keeps the black king out.d2 i. Even so. subtle play.i. The text is a very nice echo of Karpov's 47 tLlg2 against Kasparov (see Game 20) . as discussed in the previous note..sometimes..�xh4.i.. after which i. his commentary on the game that he rejected this.. in this par 35 gxh4? �g6 is an example of ticular position..once again..b 2 'it>xh4 Black has the plan of infiltrating with there is the saving blow 42 d5!. he can penetrate via h 3.b2 ble.a2 3s ..a 3 .f5..i. 37 �e3 3S 'it>e3 Karpov pointed out that..d5 40 i.. and bring his king to h5 first.. After 36 i.a 2 40 37 .�hs 38 �xf5 i.a3 (37 �g3 i. as 34 ... his king via the centre as well) 37 .Wf7! This is the only way to create seri ous winning chances.xf3 39 pick up the vital d3-pawn. However. Karpov explained in jected it. because he was afraid it would push White towards the plan of playing g4... after 35 .. it is cnecked away by the bishop. This is a nice psy chological point .. and only then it is not so surprising that Kurajica re play . and after 37 . he is prepared to give up a pawn. out in the note to White's 37th. i. Bishop En dings on the same-colour squares as his In view of the possibility pointed bishop.i. so that they can be defended.'it>hs 38 i. Indeed.b 2 i.b 3 threatens to f5..c 1 i. because the white king is attacking �h5 37 .. it can even pay to choose a slightly sub optimal move oneself! 36 �f4 �g6 34.b1 41 �e2 ..b4 gs 39 �2 i. g5? 35 gxh4 gxh4 leaves Black no way of breaking in with his king . in order to clear a path for his king to walk into the en emy king side..fs Karpov probes with his bishop. 3 7 gxh4! seems possi what Karpov is seeking. �g6 36 g4 White still has time to build the afore mentioned fortress.f7 41 i. in order to provoke an error from the opponent..... it The necessary defensive plan here is seems that Black would do better to therefore highly counter-intuitive.the moment it reaches f4. pas- 57 .. not considered by Karpov.... the alternative.. without committing him his extra f-pawn. if SS .the black bishop ii.Th e Greatest Ever Chess E n dg a m es sibly also seeking to make the ad lows White to stir up counterplay with journment..i.xf4 a3 so i.. draws. ss .d2 a2 51 i..b2 We6 51 i. After 44 We2... .'iti>e6 56 i. 58 .b2 g3+ 53 hxg3 h2 54 'iii>g 2 breakthrough to that in the game) fxg3 55 Wh1 which Karpov claims would still hold the draw after 44.. g3 48 hxg3 hxg3 49 f6 �g6 50 f4 Wxf6 51 Wf3 .xc7 'iii>x cs 60 i..g.a6 45 �e3 f4+ 46 gxf4 g4 47 fs!. is the immediate 44.i. but we can show the 42 .a 3 WfS 57 i.. 47 This panic reaction seals White's �f2 'iit>fs (or 47 .ic1 i... ex ploit the fact that this advance is not check.as 'iii>d4 59 i. .e6 43 Wf2 i.. a3 (the only attempt to make progress. to block things up with 45 g4+. . and must. which al.c1 Wg4 58 'iii>g 2 Black achieves noth This is the difference from the varia ing) 56 i. Now the only winning chance is the sacrifice 4S ..�h3 48 �g1 g4 49 fate..c 3 draws) 48 tance was 44 We2! (not 44 We3? when 'iit>f1 g4 49 �f2 h3 so i.txg4 46 fxg4+ 'iii>xg4 but 44 d5? this does not appear sufficient.b4 Wxd3 58 tion after 44 We3? .i.a3 44. The only move to prolong resis . variation here.c8 draw by continuing: 47 .f4 when White can..xg3 has left the c8-h3 diagonal.. Karpov stops the self on the king side.f4+l 45 gxf4 g4 forces a similar �ds 52 i.i...xd3 52 'iii>xg3 We6 53 �g4 when the f pawn ties down Black's king and pre vents him coming to assist his a-pawn.xa3 Wxd4 57 . e. g.i.a3 �d7 52 .i.. Bishop En dings 44...i.a3 cal. The key issue zugzwang. as Black 51 .xg4 looks promising.b4 �c6 53 .c1+ . which means that 51.i.i.i.f5 48 'it>f2 gxf3 49 �xf3 Si. .. a pawn �h1 �g4 59 �g2 �fs 6o 'it>f2 'it>e4 61 sacrifices clears a path for Black's king.c1 �h3 55 'it>g1 . Black just �a6 48 ii.. followed by an attack on bishop attacking the f3-pawn) 49 .i. 47 .h5 'ite8 51 ..a3 57 'it>g1 i.i.>g4 53 �b2 c6! the black king can infiltrate with his As Mednis points out. 59 . "do not hurry" move. indicates that there is no need. cxd5 45 d4 f4 46 gxf4 .e4+ After 46 g4+ the piece sacrifice 50 . this is a typi king.i. but Mednis . because of the threat of the terplay by c6.d3 (this ties the white eliminates any slight danger of coun king down. and after 53 'it>g1 'it>e2 the a now is the absence of a white pawn pawn costs White a bishop.i.. 46.b2 has a sound and logical positional win. Zugzwang. via the queenside.b2 .t. 'it>e1 i.'it>xf4 52 . 51 �g1 �f3 52 'it>f1 h3 renews the starting with 46 .a3 the c7-pawn.. rJilf7 (now the king marches in) so .>f2 'it>g4 46 .b4 54 i.. from the square d3. e.g.i.g6 56 �h1 ii.i. Once more.g4! White will be forced to allow the black king to the queenside.. 58 The key idea.i.d1 0-1 'itbs with a simple win. e.b2 'it>d3 etc. 47 'it>g2 .t.g4 62 . 'it>g6. The knight has no 3. and the posi In this chapter. . the conventional sense. A bishop can only at game practice. pawns. Game 16 is another modem classic. Conversely. · Bishop<versus Knight Endings The battle between bishop and knight for example where there are weak is one of the most interesting in end pawns to attack. It shows how effective a ten. al 3. + +Ch�pter< Four . in blocked positions. Games 16 and 17 illustrate cases where This is also the case in positions with the bishop is bad and the knight is supe pawns in only one sector of the board. above. In such posi examples where the bishop shows its tions. the short-stepping knight duce a superb piece of long-range calcu does not suffer from its lack of long lation to convert his advantage. squares on the board. 2. Game 17 sees Nimzowitsch pro since then. the knight is usually superior. the other 32 be Game 18 is a great example of point ing inaccessible to it. we start with three tion is relatively open. and a knight can often and the bishop not especially "bad" in be superior to the bishop in such cases. Which piece is superior tack weak pawns if they are fixed on depends on a number of factors: squares of the right colour. a bishop is frequently stronger superiority over the knight. in than a knight. Although the position is open such problem. The longer range of the bishop ever squares they try to hide. rior. a bishop knight can be in a position with pawns can only ever control half of the all in one small area. the "colour-blindness" though one undeservedly neglected in of a bishop should also not be forgot the literature. particular. when the bishop is obstructed by to obtain an overwhelming pawn mass. comes to the fore when there is play on both sides of the board. and includes the typical piece sacrifice. whilst range. the presence of 60 . whereas a knight can hunt them down on which 1. is a modem classic by Fischer. Even in an open position. Game 15. However. Bis h op vers us Kn ig h t Endings pawn weaknesses in Black's camp al Finally. cover them all. This frequently arises. which makes the bish Games 20 and 21 are two of the ops together a formidable weapon. it is the vantage than a single bishop versus a bishop which emerges on top. knight. this reason. and blindness of a bishop. Whereas a single Capablanca's defensive strategy in bishop can only cover half the squares Game 19 is a model of how to handle on the board. The reason two bishops are so fender having an isolated queen's strong is related to the colour pawn (IQP). with the de knights. For greatest world championship endings. whilst in the other. with two examples where the bishop Game 19 looks at the standard ending pair proves its superiority over the of knight vs bad bishop. the knight out-manoeuvres the and knight is frequently a greater ad bishop. two bishops versus bishop In one. 61 . two bishops together can such positions. we round off this chapter lows the knight to show its superiority. xd6 'ifxd6 16 "i¥e3 b6 17 tt::ld 2 tt::le 7! tt::l e 3.fe8 24 f4 f6 25 tt::lf3 rJ.i.i. In addi superior to the knight when there are tion.fe1 .i. and an pawns on both sides of the board and open position.ls�s " s s s " " ".g2 tt::lxc3 7 bxc3 d6 8 exd6 32 Wf3 gets nowhere.:SA.xe8 �xeS 33 . the enemy king coming in to fs.. flank.u · 1 e4 c5 2 tt::lf3 tt::lf6 3 e5 tt::ld 5 4 tt::lc 3 e6 5 gested that after 3 1 d4! r:l.f4 ..l:.. . ft. creating the threat of 34 ..xe2 l:!. so as to 62 .td6 tt::lg 2!) 32 r:l.. because after . which leaves his the reason why the bishop is usually pieces tied to the defence of c2.�fs g3 tt::lc6 6 . but 35 We3 h5! 36 tt::l h 2 this weakening of the kingside pawn White struggles to maintain his structure proves fatal.lati. hs Reti quite naturally seeks to prevent 35 g s . the biggest factor is the weakened This ability to influence both flanks i s white pawn structure.d7 32 . the bishop This is a classic case where the switches its attention to the other bishop is superior to the knight..e7 30 We3 �e6 31 g4? Stronger than the immediate 34.Uae1 l:.l::i.td7! With c2 solidly defended.xh3 36 gxf6. with pawns on both wings.i.d2 �c6 33 tt::l g 2 �bs 34 15 .i.hs. pawn phalanx on the kingside..h s there is the strong retort 33 12 'ifd2 'W/c7 13 l:. "?s" .d6 (31 .i.a4! 21 l:!. 34 tt::lf3 We7! 29 tt::le1 r:l.f7 26 �f2 l:!.e2 tt::ld 5 22 .. . . Dr Euwe sug.� = �<)t11e1tt>ura·t�?. Here. the superior mobility of the position i s open..Th e Greatest Ever C h ess E n dg a m es ·aorrre .te7 11 ..xd5 exd5 23 31 .ad8 14 . White would still have good 18 tt::lc4 fl/c7 19 �e5 fl/xe5 20 tt::lxe5 chances of holding . the bishop proves decisive...xe2+ 27 .e3 Ji. 'it>d6 32 h3 g6 33 'iir>d 2 ...txd6 9 0-0 0-0 10 d3 .e8 28 l:!.i. this time with more justi fication. trying to keep lines closed on the queenside..'it>d6 37 �e2 40 . Once again. �d6 42 lt:Jf3 i. as follows: 38 .c6 43 �e2 a6 (pre- ter defence.. g S ! 41 fs (41 fxgs fxg s is no better) 37 d4 was again suggested as a bet 41 .... appears to be holding.. 37. Fine analyses a win for Black.<it>c6 38 'it>d2 'it>bs 39 �d3 �c8 (playing for zugzwang. Black is perhaps better advised to try the direct approach at move 39.... In view of this... Fine (also quoted by Aver paring a decisive transition into a king bakh) dismisses this as follows: and pawn ending) 44 �f2 37 .. Wxb4 57 We4 Wa3 substantially here by 44 gxhs gxhs 45 58 'itxd4 'itb2 winning..... but White can improve axb4+ (56 a4 b3) S6 .i.d4! 38 cxd4?! 38 c4 is a more tenacious alterna tive. Bis h op vers us Knig h t En dings deny the black king entry squares. A very nice 63 . .�g4 46 lt:Jg3) 46 lt:Jg3+ 36 gxhs gxhs 37 h4 there would an �xf4 47 lt:Jxhs+ �es 48 <ii>e 3 when he obvious entry route via fs.. 44....Wc6 47 'it'f3 �d7! 48 �e4 �d6 49 Wf3 wes so cxbs (or so �g3 bxc4 51 dxc4 d3! 52 cxd3 Wd4) so.xf3 ! 45 �xf3 bS! 46 �e4 (White's king is now forced back by means of triangulation) 46 ... <ii>a4 40 dxcs bxcs 41 c4 d4 42 c3 dxc3 43 'it>xc3 hxg4 44 hxg4 <ii>a 3 also looks promising) 40 �d2 �c4 41 dxcs bxcs 42 lt:Jf1 d4 43 cxd4 'it>xd4. but the direct 39 ..hxg4 39 hxg4 �e7 ! 40 �d2 (the white king cannot stray too far towards the kingside.. After h4 �e4 (4S... axbs 51 a3 �ds 52 �g3 Fine now ends his analysis with 44 C4 53 �f3 C3 54 'itg3 WcS 55 �f3 b4 56 lt:Je3 �e4 "etc". be cause the c2-pawn drops after .. �a4) 36. e6! 46 'it>f3 i. Now there is always the tactical threat of . 38.xe4 46 dxe4 b4! 47 �d2 bxa3 48 40.. for queens (Fine)..1i.g2! stalemating the knight..a2! 47 40 lt:Jxg4 i.ds 42 a3 bs 43 lDf1 as 44 li:Jd2 entry....a4! An instructive moment. ter via c5-b4. 4S li:Je4+ 39 �d2 hxg4 40 hxg4 Or 45 li:Jb1 . destroying �xc3 i.. gs. b4. 64 .cxd4 44. 41 ...i. 4S .xg4 41 hxg4 is another li:Jd2 b4! 48 axb4 a3 and the a-pawn lost king and pawn ending after.i.. etc.. example. the white phalanx and securing a king 41.. which shows Black's winning plan very clearly.Th e Greatest Ever Chess E n dg a m es piece of analysis..i..c6 41 �e2 'it>c1 gS! 0-1 Fine also points out 41 c3 dxc3+ 42 The final thematic break. se Now the black king threatens to en curing a passed a-pawn. .b3 'ti'd4+ 13 'ifilh1 c4 14 c3 and also taking the cs-square from 'it'xd3 15 'ii'xd3 cxd3 16 tLlf2 l:td8 17 White's knight.c7 27 is what stays on the board that mat 'iifi>g 3 b6 ters. tLla5 12 . Botvinnik the open file.i.i. and Black also controls rooks to start counterplay.:l. Once again. they cannot really use it to where the bishop is superior to the penetrate effectively. we have a typical case open file. Driving pawn majority on the kingside.l:.f7-f5 would leave White's queenside majority contains a a weakness on e6.. it 24 h3 !tac8 25 'iifi>g1 �8 26 'iifi>h 2 . Black nails the b3-pawn to king and bishop will be too much for the spot. Botvinnik wins without therefore exchanges the rooks off.xe3 20 28 'it>h2 . and allow the white doubled pawn. Black has a healthy knight on e4. In addition.xe3 .i. fident that in a pure minor piece end 23 a5! ••. the tLlc6 8 d3 e6 9 o-o . con undue difficulty.tc4 queenside pawns on dark squares. whilst the knight away by .e3 . not what comes off. The combined force of weaknesses..tc5 18 �xd3 . .ae1 tLlxb3 23 axb3 29 . our favourite device of fixing more active. controlling d2.i.cd7 29 �g1 l:. because of the knight.d7 19 . 65 .:d1! ••• Although Black's rooks have the Once again.l:..e7 10 f4 o-o 11 l2Je4 opposite colour from Black's bishop. B is h op vers us Kn ig h t En dings 1 e4 c5 2 tLle2 tLlf6 3 tLlbc3 d5 4 exd5 Completing the arrangement of his tLlxd5 5 ttJxd5 'ti'xd5 6 t2Jc3 �d8 7 . ing his king will be able to become Again.tb5 21 l2Je4 h6 22 . White's king and knight. l:td1 .i. e7 33 .l:td3 38 l:te3 cause Black will simply win the queen side pawns with ...Jie4 Driving the knight away from its guard of b4.i.... be 36 .xe3 36 cj. or having his knight tied permanently to the defence of the b3-pawn (the pawn fixed by Black's 23rd move!). this 41. so Botvinnik heads for the pure minor 66 .xh 3. thus preparing the way for the black king. by putting his bishop on dS. Then White would have the unpleasant choice of playing c4. establishing a 3S ..e 2 f6 34 �2 l:.. 42 g3 e5 43 fxe5+ fxeS 44 lt:Jc2 . Black will at some point be able to in duce this move anyway. Naturally..x e3 il.l:. 45 tt:Je1 <iitcs 46 <iii>e 3 i.l:. �c5-b4. 30..c6 31 li:Jc3 .l:txe1+ 32 ..d3 35 h4 Creating another weakness. Black takes the fs challenge the black rook.. 3s ...Th e Greatest Ever Chess E n dg a m es 30 C4 piece ending. hs there can be no question of White en Fixing weaknesses! tering the king and pawn ending.fs 47 li:Jf3 38 . by al 40 lt:Jxe2 <iitd 6 41 li:Jd4 lowing the black king a later penetra tion route via cs and b4. .. Jhe3 39 cj.x e3 i. but he wants to be able to A typical idea. This weakens the queenside.g61 time on g4.. passed pawn in the centre. However.e3? would lose a pawn after thereby preparing es.xg2 40 lt:Ja4 is less effective for Black.xg2 3 7 li:Ja4 i.xe2+ There is nothing for the rook to do..l::i..xe1 cj.l:!.l:!.e3 l:td2+ 37 �e2 . 39 l:Ie2 l:!.�d2+ Now 38 .. ate 3 5 .. The immedi square away from the enemy knight. ."knights hate rooks' stop. Threatening 61 . gives up the passed e-pawn.. ss..c2 Now the white king must retreat. the superiority of a bishop over a knight in such positions. preventing the white king marching ss �c3 b4+ 59 �d3 i. 49 �f3 i.i. Now it defends its king side pawns against the white king. and impossible for White to trouble stopping ... that the bishop still influences Freeing the way for the b4-pawn to events on the rest of the board. Note.il. a4-a3.. Now the b-pawn is doomed.tfs Now there is no defence against the advance ofthe queen side pawns. forcing the white king back.fS+. A ate a passed a-pawn.. knowing that this passed pawn on the a-file will be much is the pawn that the knight has most stronger. pawns!" In order to do so. S6 �d4 . but Botvinnik prefers to cre win the white queenside pawns.. 53 ltJc4 bS 54 tbd2 a4 ss �ds 62 . Forced... after which there is no defence to .. A model technical display by Bot vinnik.xb3 so �e4 i.d3 The bishop continues to perform sterling service. 57 �d3 �C5 too. here advance.fS+ 62 �e3 il. i.a3 56 bxa3+ �xa3 is objectively .ds 60 lbb1 i..i.1l. s2 �d4 . in order to winning. and the pawn promotes.c21 0-1 Depriving the knight of the square b3.e61 into e4 to start counterplay.xc4 51 'it>xes 61 tbd2 .i. with crystal clar ity. Black ss .g21 48 . Bis h op vers us Kn ig h t En dings 47 <ii>b41 ••.h31 Transformation of advantage. 67 . he sets up a 48 tbd2 zugzwang. which shows.. 'iii' d6 but then rial distribution in Chapter Seven. White and the majority of the ending is pure threatens artificially to isolate the cs bishop vs knight. sidered other examples with this mate The alternative is 24. "Game :ts " �-t� . Here.xc5 pawns on both sides of the board em dxc5 15 f3 i.i.bs 26 White's advantage is based on the as Black has a weakness on a6. Although in this book.g 2 ltJf6 8 o-o than rook and knight. and and bishop vs rook and knight.xd1+ 22 'iixd1+ 'iid 7 23 24 l:Ie5 b6 'iixd7+ 'i£. the presence of h6 12 l:i. It may fact that the rook and bishop are gen.. we have con sion.i.g4 14 .. this Marin points out that after 2S a4 (of particular example is placed here. be course.f4 d6 11 "fid2 open positions.xd5 phasizes that advantage.d8 17 ltJd5 ..e6 16 f4 .." " � " < . 'i£. t�ridfdat�"Matl...tt fGame 4)#" \/a�c�9lera'!J>i' 1 e4 c5 2 lLlf3 ltJc6 3 d4 cxd4 4 ttJxd4 erally a more powerful combination "fic7 5 ltJc3 e6 6 g3 a6 7 i.i.ad1 e5 13 .>d8 21 "fie2 J:. Fischer activates his pieces. especially in ttJxd4 9 "fixd4 .Th e Greatest Ever Chess Endg a m es ". sequently penetrates. not 2S i. be this was still a better defence than 68 ...x b7? l::tb 8 which gives cause the rooks are soon exchanged Black active counterplay). was in view of the fact that it on precisely something of a Fischer speciality.e3 i. As the first 18 exd5 e4 19 :fe1 :xd5 20 :xe4+ stage. with rook the light squares on the queenside.i.. ". and these squares that the white king sub he won a number of famous examples. pawn by 26 as.>xd7 This leads to a severe weakening of This type of endgame. it may be that This is just about the purest and best of Black should have resisted this conces all..c5 10 i. whilst after 2S ..l:f. ..'Lle7 34 h3 White's medium-term plan is to gain space on the kingside by advanc ing his h-pawn. The immediate 34 h4 would bring his knight to the square d6. venting it later coming to b4 or d4..e3 'Llb8 30 l::[d 3+ Wc7 Now White has secured a potential 31 c3! entry route for his king.f8 27 �g2 knight is on its way to c6.l:t.l:!. g6.dS+ fol side light squares in the game.. 32 .. via d3-c4-b5.... 25 �f1 as 26 �c4 28 Wf3 lLld7 29 l:. from allow Black to blockade things with where it would defend bS and f7.. h s followed by .l:. 33 . as 34 .. when his 69 ..'Llc6 32 lie3 Still trying to make it hard for Black to get his knight to d6. so White Starting the process of activating "dominates" it with the c3-pawn. but he has to be care 27.... lowed by . As usual. 'Lla7. the immediate 27 ...'Lle7 and 32 ... The text pre vents both 32 . 'Lle8 al hoped that his king and knight would lows White to get his rook to the sev be able to cover the weakened queen enth after 28 . see several times in this book. pre the king. judgement call. Bis h op vers u s Kn ig h t En dings in the game...Wd6 33 a4 Fixing the queenside weaknesses once and for all. Black would like to endgame.d7+.tbS+ Wd8 29 .'iti>d6 ful. Taimanov no doubt However. This is a nice move. but it is a very hard well as attacking the bishop on c4. rushing is not good in the In principle. The black 26 . of a type we will and also placed his bishop very actively. 31. .h6-h5 himself. attacks the black pawns there. when the 3s .. we criticize him for creating weak However. so as to of his kingside pawns fixed on light have a choice of d.l:.lt:Je7). it is not obvious why Fischer the white bishop. h4-h5... but this sort of di chose the second rank. to d6. will criticize him for pas tive.Th e Greatest Ever Chess E n dg a m es knight is well placed to jump to f5 and the defender faces in such endings. g6 36 g4 pushing Black back again. because there is no pawn attacked sive play. White methodically increases his space advantage on the kingside. which would give White the 38 l:!.d3+ �c7 37 :ds line 35 h4 �d7 36 . h5 threat of f4-f5 will force Black to play Black decides to prevent the move 38 . and now returns.lt:Jc6 rest assured that later annotators. 35 h4 37 .. in view nesses in his own position! Who'd be a of the previous play. it seems more logical.f5 himself.. Instead..or e-files.l:te5 playing . and meanwhile.lt:Jf5 would be less effec self included. 70 . cramp his kingside with h4-h 5.d2 subsequent break g4-g 5. .. my Now 34 . in view of I1e7+.. he can 34. White would continue 35 . where they are vulnerable to tedly.. to play 34 . as in the game. and on d6. but this does not really matter. The downside The rook has done its job on the of his move is that he ends up with all fifth rank... Fischer takes he sits passively and allows White to things more slowly. Admit squares.tb5+! r3ic7 37 h5. If from there.. Marin then gives the 36 .. lt:Jc8 grandmaster? heading for d6. g6 White will play 38 . since after 37 . rather than the lemma is typical of the decisions that third.fs This move cannot be delayed for long.. Fischer immediately responds to the Black is still not ready to put the knight changed situation on the kingside. whilst if he fights for space by on h4..tb5! (stopping . Now a later g4-g5 is on the cards.. l:te2 �d7 40 . and a threat to The knight on d6 would stop the white penetrate with the king on the queen king penetrating. As it is.t>d3 ti'Je7 Note that the other factor Black must always take into account is the possible transposition into a king and pawn ending.tbs �d6 42 'it>e2 .t>a6 Note that..l:.. ing ending. (the first zugzwang) 49 .l:td3! 46 .... Note that Black would still kingside pawns on light squares.�C7 so . despite all his efforts. In the result could do so without losing his kingside. 43 ...txc6! �xc6 47 �c4 �c7 48 WbS �b7 49 c4! 71 .�C7? 46 . on the kingside pawns.. and the game would side light squares (two weaknesses!) be drawn.t>c7 Black has still not managed to get his knight to d6! 41 . which en able his king to penetrate by means of zugzwang. Now that he has fixed the enemy fending g6.'iii>d 5 43.. and if he rooks can be exchanged.t>ds 52 b3! (the second) 52 . zugzwang. he is never will eventually place Black in able to achieve this.. Marin points out that this was already a threat. a combination of attacks he would be able to set up a fortress. Indeed.Jte8 This is the next stage of White's Now the knight is tied down to de plan..... because White has two reserve tempo moves with his queenside pawns. 46 .. though.. 'iii>c6 51 �a7 .e3 g6 .t>c7 44 �xd6 Wxd6 45 . 'iii>c6 53 �b8 and the white king wins all of the en emy queen side pawns.. Bis h op vers us Kn ig h t En dings 38 :f6 39 . the like to get his knight to d6. Thus: 4S .. or the g6- pawn. it is White to move .c7 Black.it would be mate in one! How ever..f7 'it>b7 White again uses zugzwang to force an entry with his king.. 'it>d6 48 'it>c4 'it>c6 49 i.dl '1t>b7 56 ... The point is that later in the ending. he 47 ..ds tt:Je7 Now there is no value in putting the knight on d6. but the bishop can.tt:Jd6+ 53 'it>a6 achieves nothing for S6. since the white king is already amongst Black's vitals. and this time with real effect . 57 . '1t>a7 when 53 i. Marin analyses 57 c4! as slightly more accurate than waiting moves with the bishop.f7+ Now we have another position of The black king is driven back zugzwang.a 7 Renewing the mate threat by 54 ..b3 <j.ltf3+ At last. whilst the white bishop proves especially well placed on f3.c6+ 'it>c7 52 i.. 52 . tt:Jg8 (Black's only hope is to seek counter play with his knight) 58 Wc6 tt:Jf6 59 'iii>d6 tt:Je4+ 60 'iii>e 6 tt:Jxg3 61 Wf6 and Black's temporary pawn plus will soon 72 . this time via c6....c7 56 'it>a6 tt:Jd6 57 c4 and Black will lose either the b6... The alternative is 56 .. 51 i...Th e Greatest Ever C h ess E n dg a m es 47 i.b6-b5. Black has no counterplay with . Black is threatening to get his knight to d6.tt:Jc8 also leads to zugzwang after 55 i. Since it is White's move.'1t>b7 so 'it>bs tt:Jcs triangulate to do this. 55 ..<j.e8+ needs to lose a tempo and transfer the And again.d5+ rJi. 54 i.. move to Black His king is not able to 49 ...i... .b 5+ <Jilc7 61 i.<Jitd8 6S.f7 tbe7 61 il.axb4 66 cxb4 tbc8 67 as tbd6 68 bs tbe4+ 69 �b6 �c8 70 �c6 <Jitb8 71 b6 1-0 73 . Bis hop vers us Kn ig h t Endings be wiped out by the loss of all of his 62 i. and if the king 60 i. 'it>c6 60 i... The culmination of White's play.e8 comes across to help...xg6! kingside pawns.. pawns on its own.tbxg6 63 <Jilxb6 Wd7 64 Wxcs tbe7 Black's 54th move.. 57 �a6 tbc8 ss i. White's king Yet again. 65 b4 59 i.... Black has run out of penetrates to the other wing (two moves. white bishop.. 61.. The 59 .tbd6 59 c4 is the same harvest of black pawns. His last chance is to attack the weaknesses!).e 8 is black knight cannot hold the passed again zugzwang... zugzwang position as in the note to 62 ..c4 tbc6 The rest is straightforward.d s tbe7 The bishop is sacrificed for a decisive 58 . Th e Greatest Ever C h ess E n dg a m es 1 e4 g6 2 d4 i.g7 3 t2Jf3 d6 4 i.e2 tt:Jf6 5 This is a classic case of a knight vs t2Jc3 o-o 6 o-o c6 7 a4 'Wic7 8 as es 9 bishop endgame where the limited dxes dxes 10 i..e 3 l2Jg4 11 i..cs .l:!.d8 12 number of pawns on the board means tt:Jd2 i..h 6 13 tt:Jcb1 l2Jd7 14 i..e7 l:te8 15 that the bishop is unable to exploit its .i.h4 tt:Jgf6 16 t2Jc3 i..g7 17 t2Jc4 .l:te6 18 longer range. Although at the moment, 'it'd2 i..f8 19 .l:.fd1 i.b4 20 f3 .l':.b8 21 there are pawns on both flanks, that .:ta4 i..e7 22 i..f2 bs 23 axb6 axb6 24 will not be the case for long, as the b l:.da1 bS 25 l:ta7 l:!.b7 26 l:Ixb7 i..xb7 27 and c-pawns will be exchanged off. tt:Jas tt:Jb6 28 tt:Jxb7 'i¥xb7 29 l2Jd1 tt:Jfd7 With just three pawns each remaining, 30 'W/as i..cs 31 i..xcs tt:Jxcs 32 'iia 7 .l:.e7 all on one flank, the knight's shorter 33 'W/xb7 Itxb7 34 'iitf2 f6 35 �e3 <j;{f7 range is no problem. 36 b3 fs 37 i..d 3 f4+ 38 �e2 t2Je6 39 The other big factor, naturally, is the t2Jc3 tbd7 40 b4 gS 41 �d2 hS 42 l2Jd1 fact that all the white pawns are on t2Jb6 43 c3 �e7 44 t2Jb2 �d6 45 i..e 2 light squares, the same colour as his l::rd 7 46 �c2 <i;;c7 47 .l:ta7+ <i;;c8 48 l:!.xd7 bishop. Furthermore, the very ad �xd7 49 'it>d2 �e7 so i..d 1 tbc7 51 i..b 3 vanced black pawns also create addi tt:Je8 52 h3 t2Jd6 53 <i;;d 3 l2Jd7 54 �d2 g4 tional problems for White. As we will 55 hxg4 hxg4 56 tDd3 g3 57 �e2 tt:Je8 see, the possibility of a knight sacrifice 58 �d2 tt:Jef6 59 'iii>e 2 �d6 60 c4 bxc4 on f3 means that White's bishop 61 i.xc4 t2Jb6 62 i..b 3 tt:Jfd7 63 'it>d2 t2Jf6 spends much of its time tied down to 64 �e2 t2Je8 65 t2Jb2 tbc7 66 l2Jc4+ the squares e2 and dl. tt:Jxc4 6 7 i.xc4 Black's systematic and precise play makes this endgame a classic. It is even more impressive when one takes into account that, when the game was played, the black player was not even an IM - in fact, he secured his IM title at this same tournament! 67 �d7 ..• The first stage of Black's plan is to bring his king to b6. 68 <i;;d 3 �c8 69 'itc3 �b7 70 'it>b3 'it>b6 71 <j;;c 3 74 Bis h op vers us Knig h t En dings Stage two is to improve the knight's Note that White cannot maintain position. the king opposition with 80 �b4? be 71../i)e8 72 i.e2 t2Jf6 73 i.d3 tL'lh7 74 cause of 8o ... lbc2+ and the knight Ji.f1 ttJgs 75 �c4 t2Je6 76 �c3 lL'ld4 comes to e3, winning at once. For now, d4 is the ideal square. 8o...�as 81 �c3 �b6 82 �d2 �c6 83 From there, the knight ties down the i.e2 white bishop, which cannot move Now 83 �c3 �cs would bring about without allowing ... t2Jxf3. Stage three is the position after White's 79th move, to exchange the queenside pawns, but with White to move. Since the which is necessary, in order to give the black king cannot be prevented from black king a pathway into the enemy reaching b4, Torre seizes the chance to position. transfer his bishop to a better square. 77 �d2 cs 78 bxcs+ �xes 79 Wc3 83 ...�cs 84 i.d1 �c4 85 i.e2+ �b3 86 i.d1+ 'lii>b 2 The position has many resem blances to a king and pawn ending, Now it is zugzwang again, because since often the bishop is stalemated. 87 i.a4? allows 87 ... tLlxf3+. This means that typical king and pawn 87 �e1 �c3 88 i.a4 �d3 endgame ideas, such as opposition and Black gradually inches his way in, triangulation, play a part. The present but White's position still contains some position is a case of zugzwang, since resources. White, to move, would have to give 89 i.d1 �e3 90 �f1 �d2 91 ..ta4 way and allow the enemy king forward. But now what? This is again a posi Thus, Black wishes to transfer the move tion of zugzwang, but with the wrong to his opponent, but since the knight player to move. White to move would cannot lose a tempo, this has to be ac have to retreat his king still further into complished by a king triangulation. the corner, but first, Black has to find a 79 ...�b6 80 �d2 way to lose a tempo. 75 Th e Greatest Ever C h ess E n dg a m es ing e3. 9S ...�c3 96 .ia4 lbc4 91 .. .lt:Je2 The direct plan of bringing the knight to e3 fails, because 91 ...lLlc2?? 92 But now he cannot do so. The black .ixc2 �xc2 is a drawn king and pawn king has had to give ground, but it does ending. Even though Black can outflank not matter, as the white king cannot his opponent and gain the opposition, come forward to e2. the best he can achieve is stalemate, as 97 iLbs the white king can just run to hl. Not 97 �e2? when 97 ...lLle3 wins 92 .id1 the g-pawn. Torre uses all of his resources, in 97 ...lLle3+ 98 �g1 �d2 99 1La6 �e1 cluding stalemate tricks! 92 ...lLlc3 93 1Le2 The black king continues to edge ever further into the white position, 93 ...lbb1 but now a further winning plan is 93 ...lbxe2 is again stalemate. needed. The key idea is to sacrifice the 94 1Lc4 lLla3 95 iLb3 knight on f3 at an appropriate mo He must prevent the knight reach- ment. This is achieved as follows: 76 Bis h o p vers us Kn ig h t En dings 1. Black transfers his knight to d2 10S ..ib3 stops Black's plan of bring and his king to e3. ing the knight via as, but after 2. When White brings his bishop to 1os ... lt:Jd6 (1os ... tt:Jas? 106 ..ids) 1o6 g4 or hs, to protect f3, Black can put his .ii.d s tt:Jbs 107 i.c4 lt:Jd4 the knight gets king on e1. to d4, and White is again in zugzwang; 3. Then Black transfers the knight to either the knights gets to check on e2, e2 with check, forcing the white king to or it comes to b3 and d2. h1. 1os tt:Jas 106 ..ia4 lt:Jc4 101 ..ibs lt:Jd2 ... 4. Then the black king comes to f2. Now the black king needs to come s. Finally, Black brings his knight to to e3. e1, when White cannot both defend 108 ..ia4? the g2-pawn and stop the sacrifice of This does nothing to hinder the king the knight on f3. transfer, but Lars Bo Hansen demon Although this plan seems long strates that it cannot ultimately be winded, White has no counterplay, and prevented anyway: 108 ..ia6 Wd1 109 in such blocked positions, multi-stage ..ibs Wc2 (the king intends to come plans like this are a common feature. round via c3-d4-e3) 110 .ii.a4+ 'it>d3 111 100 ..id3 lt:Jd1 101 ..ia6 lt:Jc3 102 ..id3 ..id7. White fights to stop the knight reaching b1 and, hence d2. 102 tt:Ja4 103 ..ibs tt:Jcs 104 ..ic4 lt:Jb71 ... Following the defensive plan out lined in the note to move 99. The bishop comes round to defend the f3- The knight takes the scenic route. It pawn, but now Black wins as explained is heading for as, when White will not above: 111 .. .<it>e2 112 .ii.g 4 lt:Jb3 113 be able to prevent both ... lt:Jb3 and ..ihs lt:Jd4 114 i.g4 'it>e1 (Step 4, above) ...lt:Jc4, either of which will enable the 11s ..ihs lt:Je2+ 116 'iith 1 �f2 111 ..ig4 knight to reach d2. (117 ..ie8 lt:Jd4 118 ..ia4 stops the 1os ..tbs knight reaching c2, and then e3/e1, but 77 past 40 years. 78 ..a4 118 i....bS+ <iii>e 3 110 i.lt:Jd4 109 i.c6 g2 Completing step s. 119 .g4 lt:Je1.lt:Jxf3 winning) 117 .hs lt:Jc2 119 i. 108. White must ei 0-1 ther lose g2 or allow 120 i.. logical planning and precise execution.<ifi>e2 and one of the classic endings of the Now the sacrifice on f3 decides.h3 lt:Jxf3 121 An absolutely magnificent piece of gxf3 <iii>xf3 with a simple win.Th e Greatest Ever Chess E n dg a m es allows 118 ....lt:Jxf3+! 111 gxf3 <iii>xf3 112 i. d2 is met to the defence of the two weak pawns by an exchange on d2. which is far from simple. to bad bishop ending.f1 fs 25 . followed by 79 . The fact of Black's stop its opposite number penetrating.l:!.i. This will force Thus we reach a "pure" knight vs the enemy king to remain on e3.d6 28 .rl.g3 g5 19 f4 h6 20 l:!. Put his king on dS. deeper thought reveals that the only possible penetration with the king is via the central light squares dS and e4.l:!.he8 23 .f1 . to defend the c3-pawn.l:!. since . great positional superiority is not in 2.i.i. Nimzowitsch conceives the plan as fol lows: 1. In fact.dc2 �d7 32 . Put the knight on b1.. Now Black has to find a way to penetrate with his king.f3 g4 24 .e7 14 .xc3 44 bxc3 Thus.l:.'>t>c6 45 �e2 lDf6 46 �e3 lDe4 ltJxe2 .l:. the first stage is completed easily enough.l:!.ad1 �7 21 l1d2 bs 22 . In prin logical plan is to bring the knight to e4.i.i.i.l:!.f4 . the on el.g3 c4 16 ltJes lLlxes 17 .xh1 37 l:txh1 .I:!c1 h4 29 l:i.e1 l:.i.:tfel lDds 15 .l:!.d7 7 .xb4 41 �f1 �c6 42 .xd7+ liJbxd7 breaking through with the black king.i.bs+ . from where it will tie the white bishop if it is White to move.t2 . ciple.xh1 38 �xh1 b4 39 �g2 a4 40 cxb4 i.a1 hxg3 34 hxg3 as 35 l:lcc1 l:lch8 36 l:th1 .i.. but the question is whether he c3. Clearly.h7 33 .i.c8 12 o-o e6 13 .dd1 .h8 30 l:!.d2 �e7 31 .l:.i.e1 'iit b s 43 liJc3+ . in which case the knight needs to move.i. This will force the bishop to stand can break through and win. 8 lDf3 'ii'a s+ 9 c3 'ii'a 6 10 'ii'e 2 "ii'xe2+ 11 44. and then to seek a way of 5 liJg3 cs 6 .:tgs 26 g3 hs 27 'it>g2 . attacking doubt. such a position will be zugzwang.xes f6 18 . Bis h op vers us Kn igh t En dings 1 e4 c6 2 d4 ds 3 liJc3 dxe4 4 ltJxe4 liJf6 on c3 and g3.l:!.i. �e4.. ss ...e l li:Jb1 he is in consequences of White's next.. too.d2 61 i... he also entombs his own knight.. if the black pawn were already to bl only when the bishop is on e1.Th e Greatest Ever Chess Endg a m es .b 2! lt:Jxd2 56 �xd2 �e4 57 �e2 a3 !. so the finish requires extremely accu rate and deep calculation by Black. In ad dition.. ss.'it>d6! 57 �e2 �c6 58 �dl Black has duly lost his tempo.. zugzwang and loses after 55 i. 80 . 'it>ds 59 �c2 �e4 6o 'it>xbl <ii>f3 51 'it>e2 li:Jd6! 52 'it>e3 li:Jbs 53 i. his knight to bl. 58 �e3 �d5 would can now embark on stage 2.b 2 tempo by triangulation. Black is going to promote.a 1 Now Black must again triangulate. which means that Black must lose a 54 . 54 i. but in order to do so.c 1! but in entering this whole plan starting White finds the only way to put up with 55 . the knight must come tempo. a3. After 54 i..d 2 li:Ja3 Clearly. However. li:Jb1 55 i.a3 56 i.. and The only chance. Note White. he had to calculate the resistance. bringing lose more simply.. Black must spend the reserve tempo with his a-pawn...... he would only draw. White has adopted the only other 47 �e2 �dS 48 �e3 �d6! 49 �e2 'it>c6! means of defending his c3-pawn. so as to transfer the move to White and force him to allow the black king in to e4. will obtain a passed the vital importance of this reserve pawn. on a3. S6.. Now so 'it>e3 �ds his bishop can still be paralysed.. but as so often 72 dS exdS+ 73 <it>xds f4 74 c4 f3 75 cs in such cases.i. f2 76 c6 f1'iV 0-1 62 <it>xg3 63 as Wh2 64 a6 g3 65 a7 g2 ••• A fantastic tour de force of deep 66 a8'ii g1'iV+ 67 Wxb2 and accurate calculation. 62 a4 69 'it>a3 'it>f3 70 .txa3 Wf3 63 not even earlier! .c5 g3 64 a4 g2 65 d5 ! and White 68 'iVxg2+ Wxg2 promotes his a-pawn. there is a sting in the tail.'ili'g2+1 •• The superb point of Nimzowitsch's And here it is. if loses after 61 .. . The pawn ending is an easy win.. Bis h op vers us Kn ig h t En dings 61 axb211 ••• 67 . Nimzowitsch had to play.t>b4 Wxf4 71 Wxc4 �e3 Now it is a footrace. 81 .'it>xg3? 62 . Declining the bishop actually have seen this at least at move 55. .b2 l2:\xc3 10 i.ad1 and the superi ority of White's knight over the bishop is clear. which can attack the pawns l2:\f3 i. After the text.i. yet he exchanges it for the where a knight is fighting against a enemy knight.f5 19 f3 i..C4 White's bishop on g2 looks a very This is a typical scenario in endings fine piece. then there can follow 1S .. which leaves him with a long-term advantage..xc6! 17 .g 4 is a real threat.xc3 i.i. Black is forced to play his position.. and indeed.l:te1 i. Although rook and bishop them onto squares of the same colour often prove stronger than rook and as his bishop.g 2 ltJc6 4 ltJC3 g6 5 knight.f5 14.i.fxd1 i. thereby gradually re knight.i.g7 6 o-o o-o 1 b3 d5 8 cxd5 lLixd5 regardless of which colour square they 9 i.i. Smyslov reports that Tal offered a draw with his last move..c8 20 l:.c 2 82 .... pawns are often favourable for the 18 l2:\c5 .a6 18 lLid7 l::tfe8 19 ltJxcs .xe2 17 . 16 ...e6 14 i. and there are weak pawns on that the broken black queenside pawns the latter's side... bxc6 15 lLif3 f6 If Black tries to exchange bishop for knight.. li'xd1 12 .Th e Greatest Ever Chess Endg a m es 1 C4 lt:Jf6 2 g3 C5 3 i.xc3 11 dxc3 stand on. Smyslov understood bishop. but Smyslov anticipates it. But White has the opportunity to carry out a somewhat unexpected exchanging manoeuvre.l:.i. In order to defend his would represent a permanent defect in weak pawns. 13 l2:\h4 . 16 l2:\d2 l:tfd8 17 l2:\e4! Early simplifications have resulted in the endgame being reached very quickly.. positions with such weak stricting the latter's mobility.g4 16 lLies . . gradually increases the pressure.e6 83 ... but after 24 a4 Black is entirely queenside. 22 b4 .... gxfs 31 exfs . cxb3 20 axb3. to stop a knight check on d6.i. Thanks to Black's last move.l:.C 7 29 lbd6+ �e7 30 ltJxfS+ (30 lt:lbs . Bishop vers us Kn ig h t En dings Black could exchange off the dou This move weakens the sixth rank.d7 28 lt:ld6+ c.t>el Now 23 . 27 . the rook must stay on the d-file.l:txd8+ . 24 . but it transpires that Black is unable to 24 lt:lb7! make any great use of it. Instead. however..t>e7 29 g4! Another typical "little" endgame move....l::.i. White controls the key entry squares.�c8 as "more care leave him with two weaknesses on the ful". while White ponent to cede control of the open file. and can only wait.l:td2 was a real threat. 27 lt:lb7! 23 .xd8 21 c.es?! This is better than 27 e4 when Black can avoid loss of material by 27 .b7) 30.. After the text. but which Black soon has cause to regret.. The black bishop is forced to leave the bl-fS diagonal.d7 2s lt:las! Suddenly..:l. Tal forces his op passive.fs 23 c.f7 21 .l:. J:td2 is pointless after 22 'tt>e 1..l:.xcs 28 lt:lb7 . 29.l:. 2s .t>f2 cJ.... .:tds 20 .l:.. that would expose his a-pawn and Smyslov gave 23 .. so his next move is forced. bled pawn by 19 .. two pawns are attacked.. Smyslov immediately strikes...cs 26 bxcs .b7 and his active rook ensures that he can draw the rook ending without any trouble. af ter which the rook could establish itself on the seventh.. thereby allow ing the white rook to occupy the b-file. xc7+ <3. 33 h6 •. and so he preferred to 36 tbe8 fS post the knight in the centre.xcs was still not a threat. be alternative was 33 tt:Jas..l:. which soon drops off. a black pawn is plays on both flanks.. which this on d6. However. 'it>d7 32 . His further plan lanx... Black would create counterplay the fS-pawn falls. before going 32 lbxb7 'it>d7 after the cs-pawn. which is causing h7-pawn.c6 can be met 31 . at some 38 ..xc7 33 tbe8+ is by 34 tbe8 fs 35 gS. Tal havoc in the enemy position.<3.. on the kingside.. The 35 .fs is now square arrangement of knight on as answered by 36 gxfS gxfS 37 f4!.fs also yet another example of the agility of runs into 34 gS. where it As at move 17. the extra pawn on cs is not so time weakens the dark squares.. therefore prevents g4-g5. Smyslov continues the variation is then to bring his king to b4.. pawn. 37 . in both cases fixing the the white knight. ingly hard for Black to eliminate this 30 1Ib1 1Ic7 31 l::tb 7 pawn. when the dark cause of 36 tbe4+.Th e Greatest Ever Chess Endg a m es 29 . <3.<3.... 31. but once again the 37 tbg7 . whilst 3S . securely defended. forc and pawn on cs enables White to keep ing a breach in the es-fs pawn pha his extra pawn safely.tc2 30 tbxc4 l:tc7? 31 tbe3 wins agility of the knight makes it surpris material. whilst 33 .td7 38 gxfs gxfs 84 . exf4+ 39 'it>xf4 'it>xcs 40 tbg7 and point..xb7 The immediate 33 ...ds 38 tbe8! (threat tbc7+) Smyslov was worried that. 33 tbd6 Smyslov describes this as a particu 34 <3..l:. White is "collateral damage" elsewhere on the now just in time to defend his extra board....f2 'it>c6 35 'it>e3 as larly difficult moment in the game. but now White gains a tempo to activate his king. .. With the knight chased onto a white square. without suffering significant Thanks to his 29th move. 'ii>d s Tal resigned without resuming. White's king's knight executed the tour g1-f3-h4-f3-d2-e4-cS-b7-aS 40 c61 b7-d6-b7-d6-e8-g7. the ru moves in many of the variations .or fs-pawn drops in return. with additional Like that of Mark Twain. the game was adjourned and 39.. In the course of this game.c8 44 fxes+ lito>xes 45 tt:Jf4 . The great lesson of this endgame is the re markable agility of a knight. either the es. If the pawn is taken.td7 46 tbg6+ <iitf6 47 h s and the white king advances (47 .. even on an open board.. Here.. A beautiful ending by Smyslov.i.<iitg s 48 tt:Je7 queens the c-pawn). 40... Black cannot maintain his pawn barrier against the white king...-tcs 41 c7 'iitd 6 42 h4 1-o 39 f41 The same strategy as in the previ ous variation.'it>xcs 40 fxes 'ii>ds 41 �4.td7 there could follow 43 tt:Jh s .. 42 .. when there are weak pawns to attack... Bis h op vers u s Kn ig h t En dings greatly exaggerated. After Or 39 . .a mours of the c-pawn's death have been real "workhorse" if ever there was one! 85 . i. from an other Capablanca game.U. Not easily arise.xe7 'iVxe7 10 t2Jxd5 exd5 11 blocked by its own d-pawn. White's extra strates that such a position should be possibility of infiltrating via cs ren tenable. result in Black obtaining an IQP. such as cs. as well as other openings. .e6 17 �ac1 �fc8 18 ltc2 t2Jxd3 19 24 �d2 'iitd 6 25 Wc3 b6! 'iitxd3 ltxc2 20 �xc2 �8 21 �d2 l:!. White has a knight against the 5 . did commit a later inaccuracy. Many lines of .e7 6 e3 o-o 7 'ii'c 2 c5 8 cxd5 inferior light-squared bishop. by themselves. It is essential that he arranges his pawns so as to deny the white king potential entry squares.i. and its accurate execution.i. and if things go wrong. In this respect. and Capablanca du1y won (in fact. for the win.d3 ficient. granting 86 . the QGD. but in Aobr� lies in Black's ad the present game Capablanca demon vanced queenside pawns.Th e Greatest Ever Chess E n dg a m es 1 d4 d5 2 c4 e6 3 t2Jc3 tLlf6 4 tLlf3 tLlbd7 pawn. This game .i. he As well as the weakness of the ds.b5 'ii'x d2+ 15 'iitx d2 a6 16 .i. it is useful to compare this position with a similar one.i.d3 cxd4 12 t2Jxd4 'ii'b4+ 13 'it'd2 tLlc5 shows that such advantages are insuf 14 .c8 This is a key element in Black's de 22 l:!.g5 . A large majo rity of such ting� 1936. with the right defensive plan dered the position untenable for Black. which is t2Jxd5 9 . an endgame such as this can This is capab1anca-Reshevsky.c1 . played a year after the present encounter: I include this game because it is a textbook example of how to defend a type of endgame which frequently arises in practice.xc1 23 'iitxc1 We7 fence.i. The key difference from endings end in a win for White. . because the natural tendency is to want to keep the pawns on dark squares.. but in this instance. if Black adopts In the present game.ixe6 43 fxe6 '>t>d6 44 e7! rJ.f7 39 g4 '>t>c6 40 tt:Jg2 '>t>d6 41 tt:Jf4 Wc6. Capablanca.. Once again. The pawn on f6 keeps White's pieces out of e5 and g5. 29 tt:Jd2 i. fights against this. This shows that Black must not allow White to achieve the f5-g4 vs f6-g7-h6 pawn structure on the kingside.id7 27 tLlf3 f6! kingside pawns on f5 and g4. If we assume for the present that Black just marks time with his king. way for White to play... Capablanca purely waiting tactics on the kingside.c8 30 tt:Jb1 . and then putting his knight on f4. ter missed. which the lat n hS! .... in the present game. avoids this and keeps his pawns back to The plan (actually suggested by llya cover the dark squares. Black pursues the strat egy of using his pawns to protect the entry squares. but these vicissitudes do Averbakh discusses an interesting not alter the general assessment).ie6 31 tt:Jc3 �c6 Now White can strike with 42 tt:Je6! 32 a3 h6 33 g3 and he is winning the king and pawn ending after 42 . Rabinovich) involves setting up White's 26 f4 . Bis h op vers us Kn ig h t En dings Black a drawing chance.. This is a counter-intuitive decision.d7 46 b4 etc.x e7 45 �xd5 rJi.. we could get the following: 33 .�d6 34 tt:Je2 '>t>c6 35 tt:Jg1 �d6 36 tt:Jf3 �c6 37 tt:Jh4 �d6 38 f5 i. by putting his pawns on h5 and g6.. 28 �d4 a5! Still continuing the strategy of plac ing his pawns on dark squares. Capab- 87 . .... but Capablanca's move is good l2Ja4 c... 41 lLlf4 .xfs 40 lbxds Black has two bishop manages to prevent this.e2 47 lLlf2 .e8 42 l2Jxd5 . rate. but his 39 .f1! material by 40.Th e Greatest Ever Chess E n dg a m es lanca is proved entirely correct.. and would save a tempo over the 34 b4 axb4 35 axb4 c.i. 44 lLlc4+ c.ig8 . 40 l2Je2 48 lLld3 The only try.i.i. pawns attacked.t>c7 38 lLlc3 c.tf7 is m arginally more accu- 88 .tc6 forces Black to defend very accurately..i.. . 48 .. tion.....i.. but Black has calcu lated the pawn ending very accurately. because after Black would still be in trouble...t>e6 45 l2Jb2 39 f5! gxf5 If the white knight could reach f4. and he cannot save his 45 .txd3 49 'it>xd3 40.td7 Averbakh points out that 40 . This is forced.d7 because of 41 Keeping d3 and h 3 under observa l2Jxf6 .i.t>d6 enough.txbs 42 lLlds <it>c6 43 lLle7+. White has no way to progress by 41 l2Jf4 . ..xb5 43 l2Jxb6 quiet play.t>d6 36 b5 g6 37 game.i.b5! 46 ttJd1 . but his next breakthrough . ..or S1 . and a key lesson in how to han dle such positions. so.�d5 54 �f4 �e6 leaves him without another. Having said that.h4! (now timed just right. but if so 'it>e2 they do stumble into it. �ds 52 51 �f2 h4! draws. if they can avoid it. 89 . 54 exf4 wins) 51 �e1! (exploiting the Sl �ds 52 �f3 �es Yz-Yz ..... but with White to move.. so 53 . the here. 'it>ds? loses after so 'it>d2! �es 51 h3 (so . position....�ds 52 'it>f2 ! After 53 h4 White has used up his �e4 53 �e2! and White wins.52 gxh4 f4 53 h s ! <Ms time for 54 h6. because 52 gxh4 'it>f3 �es 53 h3! �ds 54 �f4 �e6 s s h4! f4 53 hS fxe3+ is check. king to f4.. White aims to get his 52 h S fxe3+! drawing.. Model defensive play by Capab lanca.'it>esl Now Black has the key zugzwang Careful footwork is required. they should If so �d2.. last tempo move. Bis h op vers us Kn ig h t En dings This final stage of the game belongs white king is too far back) 51 gxh4 f4 to Chapter One.. it is clear that great accuracy is required from Black. Players of the IQP are definitely advised "not to start from this position".. Black's only defence is remember Capablanca's master-class so. �e4 51 �e2 h4 . 49 . whilst 55 e4 fxe4 56 �xe4 fS+ 57 'it>f4 �f6 is a draw..�e4 49. distant opposition) 51 . and the ending contains many pitfalls.. so White has no and zugzwang .. ..e7 7 o-o o-o 8 tt:Jc3 tt:Jc6 after 28 . Black can hope to attack the 27 h3 h5 28 a3 backward white pawn on b2.. we include this game in of the ending.e6 18 b5 32 'it>g2 l:I7c5 J:.d4 .c5 16 i.xc5 17 tt:Je3 . 1i'b3 tt:Ja5 13 'i!Vc2 i.U.l::td 4 'ilic7 26 1i'xc7 l:!.l:Ixc4 ing a target. However.e3 .l:!. a nightmare endgame.d6 aS 35 ..d7 25 .dxc7 33 l:.g5 cxd4 10 tt:Jxd4 h6 11 .l:!.. it pawn grab 28 tt:JxdS? after which Black cannot get to the b-file very easily.bxc4!? bar his king.c8 15 28.. White takes his first big decision Once again. are trained on the weak was worth considering. exchanging a pair of the present section.. I suspect Naturally.dxc4? loses a pawn at once after squared bishop.i.i. because it soon rooks.k!xds 1:tc2.ad1 'ii'c8 19 'i!Va4 l:td8 20 l:. 33 ..l:i.x ds . Note that this is only possible simplifies to a pure minor piece battle. and dS...fd1 tt:Jc4 22 tt:Jxc4 l:txc4 23 'iVa5 . and his inferior light 33 . 34 .xds 30 l:!.g6 29 e3 'it>g7 30 �h2 .tt:Jxds 29 i.e8 12 . because Black cannot recapture with Black's Tarrasch Defence has yielded the d-pawn (see next note)..xc4 After some preliminary manoeu vring.xc5 1:i..l:tc5 24 1i'b6 l:.i. All of White's pieces.i.xds 31 ..b6. there is no point in Kasparov was deterred from this by the White's jumping in with the premature awkward position of his rook on cS. He faces a thankless defensive task.. 90 .xds 9 i..f3 . with the IQP be 33 .Th e Greatest Ever Chess E n dg a m es 1 d4 d5 2 c4 e6 3 lt:Jf3 c5 4 cxd5 exd5 5 obtains counterplay in the rook ending g3 tt:Jf6 6 i.. In return for pawn and Black is condemned to passiv the blockaded hanging pawns on c4 ity.l::.g 2 i..g4 14 ti:Jf5 l:!.i.d3 a6 21 l:.c4 31 i. from the latter's famous as the basis for some counterplay. On the face of it. Kasparov miliar with several famous games with therefore keeps the pawn structure the same pawn structure.. are analysed in detail by Marin. 91 . which the h3-pawn. Black has the "bad" light-squared shortly be weakened by the advance bishop. Bis h op vers us Kn ig h t En dings since it is tied to the defence of dS.. d6. rooks..xd4 36 exd4 36. such "mutual white f-pawn "dominates" the enemy IQP" structures can still confer a sig knight. especially cs. which is necessary in order to further point in his favour is the pres free the white king from the defence of ence of the black pawn on bS. White's decision looks a little strange.e 2 l:. That game. The Botvinnik games showed. from where it will attack both targets on a6 and dS. umph in 1963-4. including intact and hopes to use the open c-file Saidy-Fischer.. 34.�e7 37 lt:Ja2 The knight aims for b4.�f8 35 i.c8 38 tt:::lb4 �d6 39 f3 ••• blockading square d4. would run into e3-e4. as to e4 and g4. as well as threatening sometimes to invade via c6. The change in the pawn structure is worth commenting upon. since now he also has an IQP.. A h3-h4. depriving it of the active jumps nificant advantage on White. However. es "clean sweep" US Championship tri pecially if his d-pawn falls. as this will reduce Black's poten whose treatment of the subject can be tial counterplay. and other 34 l:td4 classic examples of this type of end White is happy to exchange both game. The latter square will here. as several classic number of times in this book. when. has weakened the queenside dark whilst bringing the black king over to squares.. in front of the We have seen such pawn moves a enemy IQP. and has lost the use of the classic 37 i. obstructed by the dS-pawn. highly recommended. to free the rook from its defensive Karpov will no doubt have been fa duties. which is that when analysing. 44 i.gs hand.. but rather than reducing its value.lbg8 40 h4 lt:Jh6 •.d3xfs. by vourable bishop-for-knight trade. 47 lt:Jg2!! 41 �f2 lt:Jfs 42 lt:Jc2 f6 One of the most striking moves in world championship history. Rather than recapturing automatically on h4. so that his king can get into the black position. and iso lated. which leaves White with a classic knight vs bad bishop position.d3! most strong players tend to effect a Seizing his chance to make the fa pawn exchange.b1 46 b4 gxh4 Black in his turn responds to White's 39th move. now simply scooping the g S. such as that on h4. the remaining pawn is left on h S. fixed on a light square. It is heading for fS. However. off the board in a single motion of the 43 . whereas Black has no such option with White's knight. Karpov sacrifices a pawn. but in return. by seeking greener pastures for his knight. however. Black gets two of his kingside pawns off white squares.xfs i.Th e Greatest Ever Chess E n dg a m es 39 . Black is resigned to the minor piece exchange.and g3-pawns that the g6-pawn has been weakened. 43 i. The prob lem is. This is for a curious practical reason. that White can always threaten to eliminate the knight by i.xfs 45 lt:Je3 i. in order to open additional lines of entry on the kingside. Karpov's move was found during adjournment analysis. rather than physically playing 92 .. in a strange way it makes it even more im pressive. from where it will attack the weak white pawns on d4 and g3. covering es and gS.. he has "widened the h4-square. which tend to improvement. so .td1 54 �g6 It is notable that White is not averse 48 .. h3 does not really help. 52 .. By elimi king will still have potential access to nating the h-pawns. 'ite7 51 ltJxhs one side of the board. end- 93 .. endings. because Black will. of course. and Black cannot chances. However. White's 47th move can now be seen. and his king 48 �xg 3 now has a potential entry route round the back of the black position. as after The effect of White's remarkable the capture of the h 3-pawn.i... which the exchanges leave all the pawns on is lost for him.. with a white pawn on h4 and black pawn on hs.�e6 to an exchange of f-pawns. White will take on hS increase the defender's drawing next move anyway..g.. the front" on the kingside.. 47. This is especially true when enter the king and pawn ending. this move would have been impossible. It is not clear how White can make any further pro gress. hxg3+ 47 .. Usually in Black gives back the pawn at once.. As i. Of course. via the h file.. It can get as far as f4 but no further. i. s3 �hs . the stronger side prefers to but 48 .. board when one is analysing. keep the f-pawn on f6... �e6 S1 lDf4+ �d6 52 �g4 .xh s? 52 �xhs Wf7 53 'ith6 and the a result. so.f7 50 'ith4 is no avoid pawn exchanges. which 49 lDf4+ 'itfs so lL!xhs makes it even harder to spot the idea of 47 lDg2 ! ! ..g6 49 ltJf4 i. Bis h op vers us Kn ig h t En dings the moves 46 . e.�e7 move does not physically arise on the 54 �g6 �e6 s s f4 fs 56 'itgs etc. the position after Black's 46th black king is outflanked after 53 .c2 Stopping the white king coming to fs. The point of Karpov's idea is that af ter 47 gxh4 his king would not have a way into the black position.gsxh4 and 47 g3xh4. and gradually bad bishop even with very limited ma inch his way into the black position. as the knight is our beloved principle "Do not hurry!". as we will see. where it hampers and appears to draw fairly simply. However.. and he will win simply. �d7 58 tt:\g7 (taking the e6-square away from Black's king) 58 ..i.i.. it allows the black king a path into the 55 . it deprives when the time is ripe. For the 55 . is that terplay. White should leave the d5. the ing his way in on the dark squares) text is a serious error. it can be hard to resist the temptation to win a pawn 61 . especially adept in such positions with and it is rare indeed to see the 12th all the material on one wing. .tf3 66 tt:\e7+ �c7 67 tt:\xd5+. via d5 and c4. as terial. As we World Champion err in this respect. pawn on the board..e4 61 tt:\f5+ 55 tt:\xd5+? Even for a World Champion at the height of his form. in this instance.Th e Gre a test Ever Chess E n dg a m es ings with knight vs bishop are some error is a perfect example of breaching thing of an exception.. Black's counterplay. and only take it Compared with the game.. where material is usu on f5) 62 �e5 (White is gradually inch ally key. even more so in a minor pawn ending is lost after the exchange piece endgame.tf3 63 tt:\e3 �c6 64 �e6 . In fact. have already seen in the game Torre The correct winning method was to Jakobsen. the knight can overcome a annex only the f-pawn.�e7 tt:\e8 (threatening tt:\c7+) 57 . the king and with check.. �d7 (of course.. which could and 62 .ciite 6?! white position.�e7 59 �f5 �d6 60 �f6 . Karpov's White of the extra option of going after 94 ... After tt:\f5 . he can take would have been a tragedy.e4 65 should have cost Karpov the win... The prob the d5-pawn without allowing coun lem with the text.txf3 was sug time being. that that White has full control. gested by Geller and Mikhalchishin.�d6 56 tt:\xf6 . follows: 55 tt:\h 5 ! . all on one side of the board...txf3 56 tt:\xf6 �e6 57 54. Now his superlative play hitherto. g. i.t>c3 59 'it>f5 <.. counterplay by .b 7 59 'iii>e 3 i.. 56...Wd5 consolidate his two-pawn advantage. White makes progress after 67 .g4 68 tLlc5 ii.f5 67 lLld3 ii.c 6 60 �d3 ii. 58 tLlc7+ Wxd4 59 tLlxa6 'iii>c 3 and White loses his a-pawn. the analogous line 56 lLlxf6 i....t>e3 <.t>xf6 'iii>d 6 63 . . White ever. tLle4+ 'it>d5 62 tLlc5 i. were also possible.b7 61 king into b3. It seems that White but it turns out that it is still very hard can then win if he can arrange the to convert. move a3-a4.d7 68 but it is hard to make progress with the lLlc5 i..g2 The immediate 63 .. 56 tLlc7+ By comparison with the previous note. White 63 lLld3 does not appear to have any way to make progress..c6 i..e4+ 66 �c3 6o. and the black king is out of play.'iii>d 5. bxa4 70 l2Jxa4 and the tablebase con square.i. 'it>b3? is bad. How block the bishop's control of d5..'it>d7 57 tLlxa6 i. him then to sacrifice the bishop for the but he cannot stop Black obtaining remaining white pawn. but waiting tactics by 64.�b3? is again bad because of 65 This tempo gain allows White to Wd3. i. it is still not clear how has managed to hold onto his material... Black must time his active counterplay for a moment when his king can eliminate both white queenside pawns..t>d 5 6o <. be cause after 64 tLle5 i..h1 61 <.c8 (Marin).xf3 57 tLle8 would be met by 57 .d5 65 �d3 �xa3 66 Wc3 White has consolidated.. 64 tLle5+ <.c8 69 a4 Black plays his bishop over the e4.. e.g 2 58 d-pawn without allowing the black 'it>f4 i.t>f4 64.g..xf3 58 <.t>c4 62 tLlc5 ii. in this line. White will simply bring his knight round to e3 and set his passed pawn in motion. e. 65 'iii> d3 i. and after 57 'it>f5 i. 95 . Bis h op vers us Kn ig h t En dings the a6-pawn. allowing White now has two extra pawns.. so that a later tLle4 does not firms that this is a win for White..e8..<it>d5 and Black has enough counterplay to draw. g2 69 'it>f2 .xe2! 72 'it>xe2 'it>xd4 when the pawn ending results in a draw.i. 'it>d5 poses only slightly greater problems. and after several pages of analysis of this position. 68 tDd6+ 'it>b3 69 tDxbs 'it>a4 10 tDd6 1-0 who in turn built upon a previous A magnificent endgame.g2 67 .. de spite his two extra pawns. Marin found a key im.i. Black is loses. 61 tDts i.g4! 71 'it>e3 ...b7? <iSi>d3 i..h1 ! 67 helpless. as his king is driven back from ttJfs <t>ds.. as Marin demonstrates: 68 66 . Whereas the latter errors.. of an exhaustive analysis by Marin. ... Karpov's 47 tbg 2 ! ! is one of the had claimed that White is winning this most famous moves in world champi position.i.i.i.. onship match history.Th e Greatest Ever Chess Endg a m es 65 tDg6 <iSi>c4 66 tDe7 provement for Black..i.a8 69 tDe3+ 'iite6 70 lDd1 and It seems that this move definitely once the knight reaches c3. One key point is that Timman's 68 tDg3 .h 3 70 tDe2 can be met by 70 . 96 . This position was the subject dS and White is preparing a3-a4. he concluded that White cannot win. The critical line was 66 . despite the analysis by Timman. . Bis h op vers us Knig h t En dings �Qme �i "' .l:r. on c3. This . .· . Black manages to formulate a winning plan. the black king will not be able to get d4 o-o 6 l2Jc3 ds 7 cxd5 cxds 8 l2Jes b6 9 past the queenside barrier represented . the knight on c2 is l2Jc4 31 l2Jd1 �f7 32 b3 l2Jd6 33 �g2 hs invulnerable against the single enemy 34 'iii>h 3 l2Je4 35 g4 hxg4+ 36 �xg4 fS+ bishop. 40 l2Jxc3 bxc3 41 l2Je1 �h6 though. 37 �h3 .ltg2 .tf6 38 l2Je1 �g7 39 l2Jd3 l2Jc3 Notwithstanding these problems. and during the interval Botvinnik offered a draw. 42 lDc2 .l:. .ltg7 13 . it is securely block White also has the alternative plan of aded.tb7 10 . but here.-. .· .i. . .with two bishops. .tgs . with Black sealing his last move.h3 l2Jd6 19 blindness" of the bishop poses a prob �h2 as 20 ltfe1 bs 21 l2Jd1 b4 22 l2Jf2 lem . Black could .W�Hik�v. Clearly. .l::txc8+ l:txc8 2 5 hope to dislodge the knight with his . His first step is to annex the white pawn on h4. . besides passive defence.· .��1 . Smyslov declined and went on to fashion a superb win. However. · · ·· ·· ·· .txf6 .·.tg7 5 c2.c1 f6 14 l2Jf3 l2Jc6 15 e3 'iid 7 is an example where the "colour 16 'i!Ve2 lDas 17 h4 l2Jc4 18 .e7 Here the game was adjourned. with 28 'it'xc2 l:txc2+ 29 . .wor:td Chamf)i<>f:tship (Ga� .txf1 2 6 .s� . However.i..xf1 'ii'c6 27 l2Jd3 'i¥c2+ light-squared bishop. the first thing which strikes one about the 43 'iii>g 3 position is the advanced passed pawn Of course. Then he will transfer his king to the queenside and set up a zugzwang.ta6 23 'ii'd 1 lUeS 24 . .1n# Mo��ciw :.txf6 11 o-o e6 12 f4 by the knight and the pawn on b3.l::tf2 l:txf2+ 30 l2Jxf2 only one bishop. . and once the white knight gets to establishing a passed pawn of his own 97 .tf1 . 1 l2Jf3 l2Jf6 2 g3 g6 3 c4 c6 4 . 'iii>h4 49 tbc2 �h3 50 tba1 'ith2 Now the a3-pawn will be an addi This is the key to Black's winning tional target. because of the advance 'iii>xfs i. but this now gives Black the extra option of retain ing the a-pawns.d6 48 tbe1 Once again. Thus.f5-f4 undermining his d4-pawn. 43 .c7 51 ttJe1 'itg s 52 tbc2 .d 8 58 tbc2 i. White could abandon his waiting tactics and set up the queenside passed pawn. 48 a3 'iii>h 4 49 b4 a4 SO bS i...�f6 51 tbc2 es 52 dxeS+ 'itxeS 53 The point is that the white knight can tbe1 d4 54 exd4+ �xd4 55 'iii>f4 �c4 56 not leave c2. in the hope of creat opposite number and penetrate to the ing counterplay.f2 57 tbc2 'itb3 and wins. Black uses zugzwang to lov. His king aims to get behind its �d3 allows the black king in: S9.b6 5 4 tbc2 'iii>f6.. 59 a3 48 .'itf2 98 .f6 chances of survival. have offered some chances of holding....�h5 44 'itf3 �xh4 45 tbe1 g5 46 fxg5 'itxg5 47 ttJc2 i. Smyslov gives the line 43 a3 'iii>h s 44 b4 axb4 4 5 axb4 i.g3+ 52 'itf3 i. 59 plan. Compared with analogous variations... according to Smys Once again. although he offers no analysis) make further progress with his king. On the other hand...b6 51 'iii>f2 i..xh4 4 6 b S i.h4 53 tbc2 �g1 Black is able gradually to inch his way in. .tas 5 3 tbe1 i. This would indeed white queenside from the rear. so. 54 'iii>e 2 'itg2 55 tba1 i.Th e Grea test Ever Chess E n dg a m es on the queenside. the extra pair of pawns on a3 and a4 enhances Black's winning chances.d8 47 'iii>g 3 gs 48 fxg s Wxg s 49 'iii>f3 i...e7 56 tbc2 �g3 so tbe1 (so tbb4 offers some 57 tbe1 i. it is better to retain the Pushing the pawn with 60 a4 also a-pawns. since a dxeS .d6 65 lbc2 White has eliminated the passed c3- f4 66 �xc3 f3 67 lbe3 f2 68 �d3 Wf3 69 pawn..g s wins the e3-pawn. much as in wins. the note to White's 63rd in this varia tion) 60 . . 'it>e1 61 lbc2+ �d1 62 lba1 J.f5-f4) 68 lbb4 f4! 69 exf4 .xe3 66 lbxe3 �xe3 67 b4 axb4+ 68 �xb4 f4 69 as f3 70 a6 f2 71 a7 f1'iV 72 a8'iV 'i!Vc4+ 73 �a3 'ii'xd4 is 1ost for White.i.b6 71 �XC3 �Xf4..gs 62 lbc2 i... b) 63 lbc2 i. loses.i.xd4 70 Wd3 .i. 63 'it>d3 63 bS also loses: 63 .c7 67 lbc2 .e 7 and now: a) 63 'it>xc3 �e2 64 lbc2 i.. In typical fashion.d8 64 lbe1 i..i. 6o a4 .i.. and the queen ending after 65 a4 .b4+ 66 Wd3 Wb2 67 lbc2 'it>xa2 and Black wins the b3-pawn and the game.xes and now: knight move allows 63 . once again because White falls 61 lbe1 i. Bish op vers us Kn ig h t E n dings 60 lba1 (60 �xc3 'it>e2 followed by lbf1 i. a) 63 lbc2 d4! 64 exd4 i.e7 60 b4 ..d6 61 lbe1 es 62 White is again in zugzwang.. b) 63 Wd3 d4! 64 exd4 i..d6 64 lba1 �c1 65 Wxc3 .f6 into zugzwang: 60.. but now the two connected 99 .i.as 65 lbc2 �g4 66 lbe1 i. .f4.i...d6 and the second passed pawn decides...tb6 (now Black is ready for . 59 i.f4! 70 dS �g2 71 �e2 �g1 and �gS wins the e3-pawn. as in the variation 65 lbe3 f4! 66 liJfS+ 'it>g4 67 lbxd6 f3+ and White can not stop both pawns. Once again. Black can even sacrifice a piece. We3 73 'it>b4 d4 Mao would have been proud! 74 tt:Jes d3 75 �xa4 d2 76 tt:'lc4+ 'iii' e 2 77 tt:Jxd2 Wxd2 78 Wb4 es 79 a4 e4 80 as �d8 81 b6 e3 and wins.The Greatest Ever Chess Endg a m es passed pawns in the centre decide: 72 deserves a diagram: Black's king has tt:Jc6 (or 72 'iii'd3 'iit e s 73 tt:'lc6+ �d6 and completed a "Long March" of which the pawns will roll) 72 .• bs �d8 67 tt:'lb4 �b6 68 tt:Jc2 . as does 70 tt:'lc2+ Wd1 71 tt:'la1 'iii'c1 72 tt:Jc2 Wb2.W2 64 tt:Ja1 �d8 65 tt:Jc2 i.i.gs 66 .<.. The final position of this variation 100 ..as 69 tt:'lb4 �e1! 0-1 White resigns as 70 'iitxc3 We2 wins. 63 . e. ltJa4 followed by . two bishops versus two 'iWc2 o-o 13 ltJc3 . the position should objectively cxd5 l::txc1+ 19 1'6xc1 1li'd8 20 o-o l:lc8 21 be drawn.b4 4 1Wc2 c5 5 stronger side's task. which also complicates the Black passes up another chance to 101 . Flohr himself suggested 29 .xc3+ 7 'iWxc3 ttJxc5 8 f3 here White has the "whole" bishop d6 9 e4 e5 10 . dxc5 ttJa6 6 a3 i.U. for a couple of reasons.. there is always the possibility of an exchange of bishops.. examined in Chapter Eight.ie3 'iitd 8 27 We1 <i. . it is ex 'i'd2 "W/c7 22 l::tc1 "it'xc1+ 23 �xc1 l:txc1+ tremely hard to defend as Black. On the other hand. · 1 d4 ltJf6 2 c4 e6 3 ltJc3 i.. i. king centralization is one of the first steps in the ending. leading to an ending where rook and bishop are stronger than rook and knight. · .. · ..txd5 18 Overall. but in practice. White does not have that possibility.l:..te6 12 pair. · . he loses the bishop pair and it is not clear how his king can ever break through effectively on the queenside..c7 28 1it>d2 As usual.22 S$r·¥·�� · � match·�e.te3 'iic7 11 ttJe2 . 24 i.fc8 14 . This example is rather more compli 30 g3 ttJb6 31 'iitc 2 cated than the Englisch.ltJc5 29 b4 ttJcd7?! This is a little passive..Steinitz one. Bis h op vers us Knig h t En dings Gome. . with rooks on the board. As we will see in Englisch-Steinitz. rather than bishop and knight. Here. In the first place.te2 a6 15 .c1 knights..b7-bS. ttJcd7 16 'ii'd 2 'iib 8 17 ltJd5 .ttJbd7 Steinitz.xc1 'IW8 25 �f2 �e7 26 . The second factor here is that the posi tion is more closed than in Englisch 31 .··. If White exchanges the knight on a4 for his light-squared bishop.>MoSt:��m�gr:a� 1933. there are no rooks on the board. 28. 39 . should his Now we see another reason why pieces stray too far away from the Flohr placed his bishop on b2. To that end.Th e Greatest Ever Chess E n dg a m es establish his knight on a more active 38 i. indeed. He can only await developments..lt:la4.. it "dominates" the enemy knight n . 3s . The kingside pawn Now Black must always reckon with advance f4 can come at any moment. knight passive .... c3 or d4. either by b4-b5.c1 'it>ds 35 . there. tional space on the queenside. Flohr first places queenside. This is classic ing the a-pawn. and which he 3S. the rest of the game.. it does not .lt:lc7 40 ...e7 39 h4 Once again.. Before bringing the bishop to h3. or the tacti each of his pieces on the most effective cal trick i. 41 f4 102 . and also seizes addi penetrating to c8 at some moment.in positions where the op as. This the queen side for the moment. Now the other bishop readies itself 32 a4 to come to an active diagonal on h3..lt:lb6 33 as plays the ending. getting ready for the endings. Flohr seizes every op portunity to strengthen his position.tc4 the bishop takes up position on the against Englisch.lt:lbs must always bear in mind. never threat that hangs over Black's head for hurry. he advances the h-pawn.. so he withdraws the knight to a defensive role on the other flank.lt:lbd7 34 i. technique . It may never happen . the possibility of a breakthrough on the so rather than rush it. Note how slowly and patiently Flohr 32 . From queen side.. one should never lose sight of advance f3-f4.xa6 followed by bs.fl square by 31 .tt:Je8 36 'it>d2 lt:lc7 37 �e3 <J. possible counterplay by the opponent.see Steinitz' ... preventing it penetrating White cannot make any progress on White's position at a3. As always in such long diagonal.but it is a ponent has no counterplay.. and it is relative position of bishop and knight is clear that his immediate plan is to ad a standard device for keeping the vance on the other flank.. gaining a little more space. promot squares he can find.tb2 at bs.th3 lt:les Black has nothing active to do on the queenside. Now White puts paid to the idea That will create a potential threat of once and for all. Now he has to reckon on the possibility of an ex change on es. his knight is tied to the defence of the h7-pawn. Now until eventually one straw breaks the there is an immediate threat of 42 camel's back. more small worries onto the defender. Once again. gs. That not giving White a passed pawn on fs. If he plays 42 . similar this scenario would be to the Now the idea is to meet a subse one we outline in Englisch-Steinitz. One is f4-f5.. the text weakens the light but it hung over White's head and squares.. winning the es-pawn. threat also never actually materialized.. and after hxgs. establish happen..h3 42 i.. and if he recaptures with the f6-pawn. 44 i. Note too how 43 h6 . Bis h op vers us Kn igh t En dings Finally. White tries to provoke a further weakness in the enemy pawn structure on the kingside. 42 g6 . this advance comes. i. trating via dS-c4-b3 with his king.. the light square g6 is seriously weakened. quent fs with ..the probing is forcing his opponent into stronger side gradually piles more and small concessions. 43 i. especially g6. This is later on.xd7. 41 f6 . for Black to worry about. but it would be another threat ing an outside passed pawn.cl! 103 . and Now White has other possible ways there would be the long-term possibil to attack the kingside. ity of the white king penetrating via which would set up the possibility of g4-h S(fS)-g6. the white king may have an entry route via g 5 to h 6 or f6. which may tell added to his defensive burden. Botvinnik chooses to keep the light squares covered...fs! This is a very important and typical move. However. Note how White's gradual how such positions are won . to where Black created the threat of pene be able to recapture with the h6-pawn.. but he weakens the dark squares instead. it may never fxg6 followed by g4 and h s.h6. If Black does nothing. dxes from where it has prevented the enemy Doubtless played with a heavy knight creating any counterplay from heart. the i. with a discovered attack on h6 knight! .te8 (47 .f6.c8 (see pawn move . It is ment.ttJxes loses to 46 i. However. to in that. 4S fxes! and for the fact that he has gradually Another great example of switching improved his position and placed his to concrete play at just the right mo pieces on the best possible squares.gs+ and tion on b2..... tics happening. /.Th e Greatest Ever Ch ess E n dg a m es The bishop has done its job on b2. Now when the knight has no square and the king can white king moves. there will be a threat not defend it without losing the other of fxes....fxes runs fended and the enemy knight back on into a very nasty tactical point: 46 Wf3 e8. 4S .. tDC7. 44.h s 47 i.. since his g7- weakened h6-pawn. and has also pressurized the es pawn on dS. Taking Black is lost: after 47 ..... with almost no tac the pawn structure.. Even so. Flohr pounces on a small tactical also a tribute to his tactical sharpness. 4S.tDg7?! Flohr points out that the knight is not well placed here. and it is true that no such immediate catastrophe would have followed if he had played 44 . so as to force the weakening forced. despite 40-plus moves of pa force a favourable transformation of tient manoeuvring. whilst 4S ..ttJf6 48 advantage of Black's last move. . he does not miss the 104 .h6! and he is ac bishop immediately eyes the freshly tually losing a piece.. and that 44.. the bishop no longer has any func (attacking h6) 46.c8 again wins) 48 i. detail inherent in Black's last move. This might seem like a bit of bad luck for Black.something else for Black to worry about. the move is pawn. With es so solidly de note to move 38!).tDc7 would have been a little better.. since this gives White a passed bS. this variation is a logical triumph of White's superb play. so Flohr redeploys it. because of . 48. "two weaknesses" stuff.i..h6 Forcing the knight back to a passive square. so the white king comes back to deal with it.. hs 47 .i.li:Je8 46 �f3 Tactics work both ways.i.. 49 g4l 46.classic comes along..e6 105 ..e3 And now White forces the opening The small tactical interlude at move of further lines for his bishops..e3 tt:Jbs 52 �e2! Black is threatening to create coun terplay on the queenside.. S2 . 48 . Flohr points out that White must not play 46 �d3? because of 46 ..i. still far from over. but now son (as well as because of .tLic7 Flohr points out that the trick 52 .xd7.i. 49.xg4 tt:Jc7 51 ..... in the form of a protected when his h-pawn would be hopelessly passed pawn on d5. but the game is weak. so his next move is forced.. Black 45 has brought White a major addi cannot allow the exchange on h5.. Bis h op vers us Knig h t Endings tactical opportunity the moment it kingside.c5+. and the that the king has strayed from the c7-knight cannot move.hxg4+ so i.c5+. tLih 5! when Black sud denly has counterplay. White strikes there . his d7- 47 .i. 53 �d3 Black is now in zugzwang. tLic3+ 53 �d3 tt:Jxe4 fails to 54 i.'1td6 knight cannot move for the same rea Stopping a potential i.c8). His king cannot move because of . tional trump. . After 6o.d2 1-o A magnificent example of patient technique by Flohr. �d6 64 i.g5 lbh5 62 i. thanks to 64 .. lbxhs 64 i. 61 i.lb7f6 57 �c4 Wc6 58 i.lbf6 65 i..xh s lbxb4+. attention is required to the very end. After the text. but it is too late to do him any good. e4+! 65 i. b5+ 59 �d3 lbe7 6o i. 64...d2 Black has avoided material loss. the b4-pawn is defended and hS cannot be stopped anyway. Botvinnik clearly White methodically prepares the learned a lot from this painful lesson. Knights are notoriously bad at stopping rooks' 63 i.. advance of the passed pawn. 58 ..g6 63 ..xf5 lbxd5 56 i.C3 and wins. unnecessary complications after 56.d21 pawns.. The hasty 64 hS? would throw away the hard-earned win.e1 e4+ 67 �d4 lbgh5 68 i..Th e Greatest Ever Chess E n dg a m es followed by i. but now White has a monster outside passed pawn on the h-file. 63 ..lbxe4 61 Wxe4 Wd6 62 hS We6 63 i.c8 Wc6 66 i...f3 lbg3 53 ..g41 Once again.xe4 lbxh s.f7 winning the g6-pawn. 106 ..f5 54 exf5 gxf5 55 i.e4+ lbed5 Desperation is all that is left.. Black finally has the chance to eliminate one of his tormentors..f5 'litd6 69 i. and 18 years later it came in handy.. and this one will cost Black a Do not hurry! 63 hS would allow piece. :: .DJit�-le&J '..ixc1 38 23rd game saw Botvinnik trailing by �xc1 l2Jxb3+ 39 <iii>b 2 lZ:Ias. . ' 1 d4 tiJf6 2 C4 g6 3 g3 c6 4 . cxd5 cxd5 6 tbc3 .' ' . technical posi l:tfe1 tL'la5 24 �f1 ..ia3 35 'it>d1 i.xc8 33 l:tc1 l:txc1 34 l2Jxc1 Had it not been for these. Black has a temporary initiative.!:.'�f. Unfortunately. < '' ' >'�i-'' i ' . He was greatly Bronstein's tL'ld6 21 i. whilst the black queenside pawns are doubled and weak.c8 12 o-o tL'ld7 13 tbe2 �b6 14 i.f1 f6 31 l:ted1 l2Ja5 32 .. '' ' ' ' 7 .. . ' . >..ac1 tbac4 23 superior in such dry. 'x.. . and.�f7 when White has nothing after his title in the event of a 12-12 tie.. but this is one of for White's knight.xc1? played under the auspices of FIDE. championship history... Bis h o p vers us Kn ig h t En dings � ' ""' 5:.3 '.ig2 d5 5 must-win situation.x.if8 28 a4 tL'lc7 29 i.d2 in rather a sterile manoeuvring game.xh3 tL'lc6 9 i.·� . it was a 36 'it>xc1 l2Jxb3+ 37 �c2 l2Ja5 107 .. and could have maintained the balance fairly easily. or 36 �c2 one point. a positional trap awaits Black: he can win a pawn in this position.ig7 2 5 g4 tbc6 26 b3 tions..xh3 8 A symmetrical opening had resulted i. by giving up his bishop It may not look it.c7 .g2 e6 10 e3 o-o 11 i. ..f8 20 f3 Botvinnik. and Botvinnik suggested that Black using the "new" system of 24-game could maintain equality with simply matches. thereby presenting the most dramatic positions in world White with the "whole" bishop pair.l:!. In this position.. It was the first Bronstein could not resist.. ' .l:. with the champion retaining 3S . How ever.xc8 thanks to Bronstein's endgame lapses.si ��> ['.ie1 tL'la5 19 tL'ld3 i.ixa6 bxa6 37 i. ' . game with the white pieces. with a new world champion crowned. indeed.w(. the struggle would already have been done and dusted several games earlier. ': :' ' • "' '.f2 i.c3 but despite the match situation. world championship match to be 34.. ' 's t:ilme 2. ' '. White has the bishop pair.flc:l'c'tiam'�Jbnsbm·foafue�3{¥t>tt<>* ta. ' ' ' � � ' '. ·. I:!. had already gained l2Jb5 27 �e2 . this :fd8 15 tL'lf4 tL'lf6 16 �b3 tbe4 17 �xb6 was probably not so unwelcome for axb6 18 . and with this being his final l2Jb4+.tg7 7 tL'lh3 i..g3 several extra half-points in the match l2Ja6 30 i.h6 22 . This either 36 . d3 �g6 self.and i. but the weaker text-move. The pawn itself is after 46 i. Alas. Salo Flohr was left to spend all night Botvinnik criticized this as playing analysing the position after 42 i. that Botvinnik whis pered to Flohr (using the affectionate diminutive form of the latter's name}... Now the h7-pawn whilst Botvinnik went to his room and will be left weak. liter ally moments before taking his seat for the resumption. analysed the true sealed move by him 4o gxfs gxfs 41 i.tZ:Ic6 (or 42 .b3 and Black will lose the d-pawn the good news ends. vinnik's lifelong suspicion.. White has a clear advantage. the Bronstein camp was unaware of this 108 . Salomonchik. but there .a2 tZ:Iab4 45 Black has won his pawn.i. Botvinnik had not sealed 42 and is of no real significance. at this moment. whilst i.. sealed move.i. "Actually. that he had sealed another move! Poor 39 fs?! . he was so concerned about a "leak". this fact is associated with one of and Black's knights lack secure central those stories. the game was ad journed overnight.d6. Amazingly.Th e Greatest Ever C h ess E n dg a m es 42 . 38 c.a2 will follow) 43 exds exds 44 i. Despite the material balance.f4! - dominating the knight .i.a2) 42 .bl! (threatening 43 exds exds 44 i.bl.d6? At this point.tZ:Ic4 43 i. with Botvinnik seal ing his 42nd move. which demonstrates Bot posts. His second in this match was the noted endgame expert Salo Flohr. the miserable doubleton on the b-file. Flohr was almost in tears. It was not until the next day..bl. who quickly established that White was winning after the move 42 . ever. How White now has the whole bishop pair. so humiliated was he at the obvious lack of trust! Back at the game. into White's hands. I sealed a dif ferent move!" Botvinnik later wrote that. that he did not tell Flohr White seeks to open the position.tc3 <M7 39 e4 that would alert his opponent to the As always with the bishop pair. however.. Finally. natural though the text 48 i. bishop on d6 unnecessarily. The idea is to meet 44 the necessity of defending fs.fxe4 attention to Botvinnik's actual sealed All that remains is this exchange.d6. it is not the strongest..t>gs and .bl. he missed the defence which might have made him the next world champion. Had nothing...h4+. which Botvinnik admitted that he only found at 8. possibly because they had too tbac7 49 Wc2 Wf7 so 'i..tes moves. they might have found the de which at least frees the black king from fence 43 .hs 4S h3 changes move is. say with 44.tba7!. Black is now in zugzwang. It would appear that they did not Black is in no danger) 46 .tbe7 loses a pawn after 4S i... 1 09 . e6) 4S . His king must maintain the defence of fS. then it loses the chance to reach e6. which cannot have helped his chances.lteS+! (driv black king to get across to e6. to defend ing the king away from contact with the dS-pawn.<&t>g s.d6 tba6 47 exds exds However.. and if it moves.xb4 lbbs+) 47 . 42 tbc6 43 i..a2. allows the and 44 ..... As a result. Black appears to be holding (Botvinnik).ttJbs 48 . could he? Bronstein had not looked at the game continua tion at all..a2 lbcb4 47 i...t>b3 tba6 and much faith in Botvinnik's accuracy .tg31! simple 42 i. 43 ..�6 and White continues 4S exds exds 46 This is the difference from 42 i. 44...txa6 'i...lta2 with the surprising night convincing themselves that their 4S . Bis h op vers us Kn ig h t En dings drama.bl.lbab4 is met by 4S .ooam on the morning of the re sumption.bS! 46 as (46 .txdS bxa4 47 . Bronstein and his team devoted some 44. However.tbl.b3 followed by 48 i. move.b1 ••.wgs 46 i. i..b4+! 47 'Ot>d3 take the precaution of analysing other (or 47 i. This is the key winning idea. The extra tempo wasted on placing the 44.. surely The Patriarch could not miss the 44 .. and had spent a discouraging exds exds 4S ...ltxb7 man's position was hopeless after 42 lbbs+ 48 'Oti?b2 tbxd6 49 . Bronstein Fixing the hS-pawn as a target.. much to also securing gS for White's bishop. when White should win without White can win a pawn at once with 51 much trouble. bs with 61 i. of meeting 6o ..f3 �fS Smyslov recommended 60 decent central posts. on the other 52 . was psychologically scarred for .lt::l 8 e7 57 i.xb6 and White is winning...if3 lt::l6 e7 life by the experience.b6-bS.g5 1-o h4 Black is in zugzwang. but this does not seem SO . and thought for 40 minutes and..xc6 bxc6 Botvinnik also analyses so.lt::le 7...d3 scores.id3+ �f6 55 hand.c1! (60 .. After 57 .txhs lt::le4+ 61 'it>c4 lt::lx gs 62 50.c8 is a threat.i.... i... when 62 as..b6-b5.. David Bronstein. endgame enabled Botvinnik to level the 51 i. Indeed.x ds greater freedom. lt::lfxd4 59 i. some commenta the white bishops are.. title..xe7 (keeping up the pressure with 51 This dramatic and textbook two-bishop ii..'iii>x e7 52 i.. One can see at a glance how powerful he resigned..Th e Greatest Ever Chess E n dg a m es now the white bishops gain much Or S6 ..te2 'it>g6 56 . lt::lc 6 58 ii...lt::la b8 49 i.. The black correct..c7 lt::lfs 58 i. .f4 h5 47 exd5 exd5 48 57 i. 45 fxe4 h6 46 i..h3 is also good) Sl.g 5+ Wf7 50 i.f4 lt::lbc6 52 ii.te2 'it>g6 54 .. and on c6 and d6.lt::lc8 53 .lt::la 7 hxgs 'it>xgs is less clear) with the idea Black returns to the idea of .xds lt::ld6 59 knights never succeed in obtaining any i. lt::ld7 is prevented.. tors even claimed later that he could 48. the surprise of most of the spectators..g 6.f5 still have drawn by placing his knights Now 51 i.. 1 10 . and a draw as Black in the final Preventing the counterplay with game meant that he retained his world . but with the Rook endings tend to be rather stronger side's rook at the side of the more "concrete" than minor piece end pawn. and lengthy. Games 26 and 27 are two immortal Because of their relatively concrete. in order to activate the rook. they conceal an enormous num mous and instructive of rook endings. further examples of the exploitation of 111 . The most notable fea strategic plans rather less common. and also the most difficult. it dealt only with posi that "rooks belong behind passed tions containing rook and pawn pawns".Rook endings are the most common of and we will see this theme repeatedly all endings. ture of the game is the study-like fin Rook endings are much more about ish. multi-stage ample. and loses. accurately. Games 29 and 30 are even two. This contained 3 5 2 pages. which shows the subtleties con specific variations. with long spells of patient ma but Black soon goes wrong in this ex noeuvring. Game 25 shows a case with against rook! the same material. most famous example of the dictum but despite this. it is hard to formulate knowledged as the all-time great mas many rules of thumb in rook endings. Both show the im but the best piece of general advice one portance of the more active rook. This is the single Endgames. This should usually be a draw. it is worth sacrificing a pawn or than material. Game can give is that activity is almost al 28 is another classic. ings. where the active ways crucial. widely ac tactical nature. ber of subtleties. One illustration of the final game of the Alekhine this is John Nunn's book Sec rets of Rook Capablanca match. In the great majority of rook and king prove more important cases. which is why they cealed in an apparently simple posi are so much harder to analyse and play tion. Despite an outwardly simple appear We start with one of the most fa ance. in the examples in this chapter. endings by Rubinstein. ter of such endings. in one of their vinnik. Game 31 is a ters of rook endings) squeezes out a classic battle between Fischer and Bot win against Karpov. 112 . we close the chapter with Korchnoi (another of the greatest mas two titanic struggles. in which the latter rescues half most dramatic world championship a point from a seemingly lost ending. whilst in Game 32. the great Viktor Finally.Th e Greatest Ever Chess E n dg a m es the more active rook. battles. a2 cxd4 15 in the next example.d4 l:r.a8 world title.f6 53 l:r.:txd4 'iftg7 51 as I:ta6 52 . This is the textbook demonstration Now the black rook is tied to a6.l::!. pawn.xc4 li:Jds 10 i. so this was the endgame 37 !ta2 .xe6 'ii'xe6 23 'ii'a s li:Jc4 24 'ii'xa7 li:Jxb2 this was game 34 of the Alekhine 25 l:.e7 7 li:Jf3 o-o 8 i. White li:Jxd4 g6 16 litc1 J. whilst passed pawn in a rook ending.f6 49 'ir'd4 54 .a41 'ii'xd4 50 .gs li:Jbd7 5 the black rook on a4.l:.b8 36 . as mentioned 12 li:Jg3 cs 13 o-o li:Jb6 14 J.xe4 .:te2 l:f. 'ii'h8 47 'i!t'b6 'ii'a 1 48 'iftg2 .l:!. outside the white pawn will advance.l:i.xe7 'ii'xe7 11 li:Je4 li:Jsf6 defend as once thought.d7 'ii'b1+ 42 �h2 'i!Vb8+ Cuban and gave the chess world its 43 g3 l:tfs 44 'ili'd4 'i!Ve8 45 .e6 22 of great historical significance.d3 dxc4 9 be a draw.a6 The vital first step. with the the white rook can make tempo moves stronger side's rook behind the passed back down the a-file.:as 38 'ii'c7 'ifa6 39 'ii'c3+ 'ifth7 which dethroned the great "invincible" 40 l:td2 'ii'b6 41 .l:.xc8 :Xc8 26 'ii'xb7 li:Jc4 27 'ii'b4 Ita8 Capablanca world championship 28 Ita1 'i!Vc6 29 a4 lt:Jxe4 30 lt:Jxes 'ii'd6 match. Alekhine led 5-3 in decisive 31 'ii'xc4 'ili'xes 32 Ite1 li:Jd6 33 'ii'c1 'ii'f6 games. as required. but as it is.ac8 18 e4 is winning by force.d7 17 'ii'e 2 . because i. the game would e3 c6 6 a3 i. The ending is also es 19 li:Jf3 'iftg7 20 h3 h6 21 'ili'd2 i. If the white rook were on a6 and White's plan is as follows: 113 . else of how to exploit an extra. needing six wins to take the 34 lt:Je4 lt:Jxe4 35 l:. although it is not as easy to J.:tds l:. The rook takes up position behind its passed pawn. R o o k En dings 1 d4 dS 2 c4 e6 3 li:Jc3 li:Jf6 4 i.l:tds l:tf3 46 h4 fourth world champion. as follows: 57 h 5 54.. Yet it is open to the objection that it weakens the square g5 and so allows the white king an avenue via this square into and White wins in similar fashion to black's kingside pawns.. Black will not be able to hold both fronts: 59 f4 h5 60 'it>d4 'it>d6 61 l:1a3 ! 56. to prevent the white king to an examination of this very move. rather as in the final phase of the game continuation) 58 hxg6 fxg6 and White will now be able to create another passed pawn. a waiting move by the white rook here. Once the kings stand in opposi analysis is too extensive to reproduce tion. Just as in the game. this move has hardly ever been commented upon.. although this ending has been examined in almost every major end game textbook. umn anyway. Readers book was being written. 'it>d5.. supported at the crucial mo By a curious coincidence. h5!? I have always thought it curious that.Th e Greatest Ever Chess Endg a m es 1. pawn. or into the kingside (two analysis immediately exploits Black's weaknesses!}.'it>f6 55 'it>f3 'it>e5 56 'it>e3 'it>e6 (or 57 . German end. but the conclusion is that either it allows its opposite number White still wins... as indeed oc the game: his king will either penetrate curs later in the game. Erich Korber. and can be found in Muller's col will force the black king to give way. wishing to see the complete analysis 114 .. pawns. this time on the king side. just as this ment by a switch of his rook. The black king will have to stand one of his Chess Cafe endgame columns in opposition..h 5. failure to play 56 . It therefore on the queenside and support the a seems logical to suggest 56 . or will attack the kingside keeping the pawn on h6. Bring his king over towards the game expert Dr Karsten Muller devoted pawn. The full 2.. German amateur.. The main line of the into b5.g5 58 f4 gxf4+ 5 9 'it>xf4 and White's king invades the kingside. reaching b5 and forcing the advance of prompted by analysis supplied by a the pawn. com.. However. 62 .. and the black rook will endeavour to hold off the white king.Wc5 61 �c3 Wb5 For the time being. raid on the black kingside pawns.chesscafe. 59 .l:.l:. His king holds up the passed pawn. Black's king must give way.a21 Step 2. White repeats moves. zugzwang will prove to be the straw that breaks the camel's back 67 f4?1 "Do not hurry!" Alekhine is here guilty of breaking that cardinal rule..e5+ 66 <ii>h 6 �f5 Step one has been accomplished... as so often.. so his only defensive idea is to "change the guard" on the queenside. it does make it harder.'it>ds? 60 �b4 obviously loses much more quickly. R o o k En dings should visit www.. blocking the passed pawn. 59 . because the resulting king and pawn ending would be hopeless. whilst his rook defends the kingside pawns. and Continuing the next stage of the look up the Karsten Muller article plan. Black cannot take the a-pawn. Now the white king will mount a "News on a classical rook endgame".d6+ 63 We5 l:te6+ 64 <io>f4 Wa6 65 �g5 . 6o �b3 Before advancing his king.. and dated 19 October 2011.. seek to establish a decisive material 57 <ii>d 3 �d5 58 �c3 'it>c5 advantage on that side of the board. I am sure a 115 .l:.�b5 59 . As usual. the king will occupy a6. 6o. and although it does not imperil the 62 Wd41 win. Black has achieved his best possible defensive formation. ds 73 �g7 l:tfs 74 1134 �bs zugzwang. The a-pawn will fall.l::!.. J::tc 5 73 'ittf6 . whilst the black ish 67 'iii>g 7! l:lf3 68 �f8! l:tf5 69 �g8 ! rook has plenty of squares on the sec Triangulation..Th e Greatest Ever Chess E n dg a m es Rubinstein or an Ulf Andersson would much of its mobility. after which the h-pawn falls and his remaining kingside pawns are weak.... 6 9. but then 72 �g5 ! (72 �xh5? f4+) and Black must lose a pawn.l:!..) 75 gxf4 l:tf5+ 76 �g6 rook onto the second rank. White still has another way of breaking the enemy resistance .:ta3 68 .undermining his kingside pawn structure with a pawn break.f5 11 'it>g7) 10 'it>g7! .l:tf5 11 f4 zugzwang. 68. 70 fS! This is the key idea.. where he . .. l:tc3 was a threat. because of the have played the precise zugzwang fin need to defend g3.Jlcs! sive play also loses. His rook has lost the opposite flank. but the black king Now we can see that White has is too far away to avoid devastation on complicated things.l:tc7 69 'iii>g7 l:td7 7S lle4! The decisive switch to the kingside.lk5 75 'ittf6 Seizing his chance to re-position the .gxf5 71 �h6 f4 71 .. 68 . 116 .l:tf6 7 0 f4 ond rank and cannot be driven into l:. 74. e.g.l:tf3 (or 6 9. Black must cap ture... ktd5 was another defensive try. Fortunately..xf4 77 'iitxh5 and wins.d 5 76 l:tf3 ! etc.. 70.... Clearly. has enough squares to avoid 12 gxf4 l:!... 72 .l:td5 74 'iitxf7 f4 (pas 67. the only hope. e3 'iii>x as 81 l:th3 etc.l:ta1 78 'iii>x hs :g1 79 .l:r... 1 17 . decisive.e7 1-o Black could still pose practical prob Of course. but not with the black king cut base confirms that White is winning off so far away. although the table drawn... Rook En dings 75 .'�a6 .• with the transfer of his rook to the h 7S . 77 .'iii>x as? 76 l:teS+ is immediately file by 80 l:r.gs l:th1 76 'iii>h 6 llxas 80 lUs 'iii>b6 81 l:txf7 'iii>c6 82 .'iii>b 7 77 lle7+ 'iii>a6 78 :tes material distribution are theoretically l:txf4 79 'iii>xh s f6.l:.l:tes . many endings with this lems by 76 . .xd6 32 'i'ixd6 37 'iiig 2 l:.Ub3 'i¥a6 21 l:ta3 'ii'b6 22 l:.. subtleties that Black needs to be aware The position should be a draw.l:!.d2 'Wie7 does just that. ponent..g2 l2Jc6 6 l2Jf3 e5 7 dxe5 l2Jxe5 8 from the path of righteousness. l:!.l:!.xd2+ 9 'iVxd2 'it'xe5 10 l2Jc3 o-o eventually pays the inevitable price for 11 o-o d6 12 . Despite to the side of the passed pawn. it This is an ending played by two makes Black's life more difficult. and l2Jxe5 i. The theory of this ending has re course. l:Ie2! (Sadler) is probably more . but soon he is tempted 5 i. Initially.Xa7 39 .l:!.. cxd5 'fib6 15 .l:Ie1?! Although this move does not lose.txc6 bxc6 19 .txc6 18 . there are a few make it worth its place in this book. following some behind the passed pawn.b3 'it'a6 23 accurate. White's play was superb.a1 38 h4 g6 39 'iitf3 'ilixd6 33 .l:. and al standard way to draw the position though he had some help from his op would be 39 .. when the attack on f2 slows l:!.l::!.e6 30 e4 l:!.. by players whose names will not be found permitting White to transfer his rook amongst the all-time greats..l:txd6 c4 34 bxc4 .d1 over to the queen side.xd6 l:!.:t. he detailed researches by a Swiss player. even here.c3 c5 17 36 .l:.b4+ 4 i.d3 .fe8 27 e3 h6 28 'ii'd 3 Wic7 29 .a3 'i!Vb6 24 .a6 .ad8 25 b3 'i!Vc6 26 down White's plan of bringing his king ..xc4 35 . the final stages of the ending However. as Black can transfer his rook cently been updated. of of. The this. hs followed by . 'iiig 7-f6.l:tec1 c5 20 36 .c8 31 l:!. 1 18 .d7 16 ..Th e Greatest Ever Chess E n dg a m es 1 d4 l2Jf6 2 c4 e6 3 g3 i..Jle1+ dxc6 .:.xe4 36 .ac1 i.l:.:tfe1 'i'ic5 13 lLld5 lLlxd5 14 his sins..d5 .l:.l:. fS! . 60 f4+ �g4 61 fS l:lb2+ 62 �c4 l:tc2+ 63 �d4 l:lc7 64 f6 h3 65 . ... It seems 119 . 43 .g.. J::te s! as with 47 .. 47 . Af he thinks this is the decisive error.d1 43 �e3 Now White's plan is clear... Fine then continues the the clearest way to draw.d7! .. .hs h4 White immediately seizes the op portunity to transfer his rook to a bet ter location. e. He will gradually bring his king over to the queenside. al sky's Endgame Manual for a detailed though it is close.g7 . This appears correct.U.l:f. Dautov.f4 allows 49 . Fine analyses 49 ...c1 45 Wd3 �g4 46 49. .e2! And now he also cuts off the black king. 54 �c4 hS 55 summary of their findings.l:Ic8 etc..a1 l:tc2 . 40.b6 'it>xf4 S6 l:txg6 J:Ie2 57 �bS l:!.l:f.:te8 section in the second edition of Dvoret with a draw. which White will necessarily have to abandon in order to carry out his plan.g8+ �f3 66 l:.c4) 49 l':tb2 l:tel so a4 sky and others. by sheltering the monarch behind the rook. I can recommend the f4 51 gxf4 �xh4 52 as �g4 53 a6 . Black must meet this by preparing counterplay on the king side.l:f. Dvoret 48 ...d2 I:.h6 �gs s9 l:!.l:!.�f5 44 ..a2 58 40 l:.:i.l:!..l:.'ifi>g7 41 1:td2 'it>f6 42 l:. alongside the pawn. R o o k En dings Johannes Steckner. but I ter 47 'ifi>c3 Black will start counterplay believe it is. augmented by variation 48 �b3 l:r.a1 48 �e3 l:te1+ 49 �d2 42.d1+ 47 'it>e2 Fine does not make clear whether Biding his time for the moment.bl+ (the immediate Grandmasters Muller.l:!. so there is no time to lose.l::td8+! (not 54. base confirms that the position is a here Black's rook is on a1 rather than d1.l:.l:la3 Compared with the line after White's (instead of 53 'it>b4).U. The rook switches roles yet again.gs? This is a further mistake.hxg 5 62 'iitb 7 .... Fine merely writes that this "was better".l::ta2 l:tas 52 'it>c3 f4 53 'iitb4 . . but af By comparison with the position exam ter 61.l::!.l:. Black has just White looks to be winning) 55 �es enough play to draw.a4+ �hs and draws. Even so. . it answer Black's . l:ld2 56 gxf4 . returning to the counterattack ss 'it>cs . Another try here is Sadler's 53 . runs as follows: so a4 fs 50 �C3 51 .. seems to me that White is now winning..b8+! Black's counterplay has hit a temporary brick wall.a3 . here the white rook is excellently induce Black to exchange on g3.fs. but S3 .xa7+ the table ined in the note to White's 47th move. because he can always stabilizes White's kingside.'iitxhs 62 'iitb 7 . or 61 hxgs (Fritz initially thinks 61 h S saying whether it is sufficient to draw.a8 60 'iftc6 gS so. have been able to .. and ined in the note to White's 47th) and it the position is a draw.b2 56 as llxf2 57 gxf4 l:txf4 58 a6 l::tf8 59 a7 .f8 Fine's analysis.U. . The idea is to try to 47th. draw.fS-f4 break with lk4. nor 59 a7 l:la8 60 �f6 g s ! 61 hxgs hxgs 62 my silicon assistants. without analysing further.U...l:.Th e Grea tes t Ever Chess En dga m es rather illogical to switch the rook back seems that Black can hold..:txf2 57 as l:txf4 58 a6 ..a8 54 �d4 .a8 54 l:i.. in which neither I.l:tas 55 with his king so far advanced amongst 'it>e4 h s 56 gxf4 'it>xh4 57 lla1 and White's kingside pawns. 1 20 . offers better winning chances. after which Black is definitely lost.. find any flaws. which placed on c2. after in front of the enemy passed pawn. Black should have played so .. Clearly.l:. although it takes some study-like play by White to prove it. whichever way White recaptures This allows White to gain a tempo with on a7) 61.:txa7+ 63 51 �b2 (51 �b3 llb1+ is the line exam 'iitxa7 'it>f4 64 'it>b6 g4 65 'it>cs g3. xa7 'iii>xf2 is a draw......J::!.g. 51 hxgs hxgs 52 'iii>b 2 l:.:es 59 a6 :as 60 'iti>cs with a simple win . If his king strays too far towards 53 .. Black's only hope is to give up his rook for the a-pawn and win the white f-pawn with his king. R o o k En dings This appears to hold up Black's counter See the last note . after 61 .a6 57 'iii>b4 and passed pawn.b 6+ 54 'iii>a3 f4 is still met by 55 pawn. e.Black has no counterplay.. 61. about to drop. 61 'it>cs! 61 'it>bs? :xa7 62 l:..l:!. ..:.g. and there is also one less pawn each on the kingside. which Now we begin to see White's prob enables White to squeeze out a win. whereas after the text.xa7 Wxf2 63 'iii>d4 'iti>e2 64 l:!.g s 56 as .a2+ 'iii>e l 65 'iii>e 3.e1 53 a4 Now Black's counterplay is a couple of tempi slower.e6 53 �b3 �3 pawn. Black has no real threats on the king 56 as f3 57 a6 Wh3 side) 54 as f4 55 gxf4 'iii>xf4 Obviously.a2 .fs 54 Wb3 f4 55 gxf4 gxf4 the a-pawn. about to become a dangerous l:k4. lem..c8+! This is obviously the only chance.l:!. at a moment when White must recapture with the rook. earlier variations. Black will be able to sacri fice on a7 and draw the resulting king and rook vs king and pawn position.l:. the white king is close enough to get back and stop the f-pawn. e.el 52 a4 l:.xa7 62 . s s.compared with play enough for White to win. there is no white h Sl. In order to do this. and no black h (53 .. else White simply marches his king to 121 . Black needs to sacrifice on a7.ki. 58 a7 l::te8 59 l:la2 l:ta8 60 'iii>c4 Wg2 56 'iii>b4 g s 57 hxgs hxg s 58 ki. etc.. whilst his rook is also ready to 63 �f1 .l:.e2 idea only works when White has brought his rook to c2.. whilst his However. preparing to use the rook to shelter his king against the lateral checks. 122 ..a crucial detail.e81 .. After the fur ther moves 66.. Likewise. the l:!.l:.. a wait ing move such as 63 ..e6+! 64 �b7 l:le2 !.e7+ 65 'it>b8 :e8+ 66 l:tc8 and the a-pawn queens. because the Now we see the difference. it may not be obvious why this makes any difference to White's dilemma. Wf1? would fail to 64 �b7 l:..e2? fails to 64 a8"i¥. Black fxe2 65 a8�+ is check . but we will see that it does. The obvious way for White to proceed is 63 l:l. . 63 .. must return his king to g2.. in certain circumstances. At first sight.:. Thus.. After 63 �c6 ! ! Black has no choice Now Black is prepared for lateral but to wait with his king.. after 63 I:tc2 by 63 .I!e2? when 64 l:txe2 the back rank.. which places Black in zugzwang.. since the white It looks as though Black must re king blocks the long diagonal.e8 63 Wc611 A magnificent move. . Similarly. so 65 sign. but now he reveals that he still l:!. 62 �b6 Now this is playable...xc8+ 67 Wxc8 �xf2 68 a8� White wins the resulting queen vs 66 l:tc21 bishop's pawn position. 65 a8'i¥ l:!.. where it is undefended. The rook checks 63 . since the rook on a2 is defended.xe2 fxe2 66 a8'ii' is not check.. Black can instead defend own rook defends the f-pawn.xc2 is also a draw.. not the which is why the king stands badly on immediate 63 . 64 �b71 I!e7+ 65 'it>b6 l:. 62 l:!. The pawn is only on the sixth rank rather threat is 67 Wb7 l::te 2 and now 68 . whilst main checks after 63 �b7 l:te7+ 64 Wb8 J:r.l:Ic1+! than the seventh. Consequently.c8+? 64 �b7 both lose at once. come to e2.l:.e8+ taining the attack on the f-pawn.Th e Greatest Ever Chess E n dg a m es b7 and wins the enemy rook.c2. whilst has another shot in his locker.. J:te6+? 64 �d7 and 63 . e8+ 69 l::tc8 :xeS+ 70 Wxc8 well the subtleties that can be con 'it>xf2 71 a8'ii' cealed in an outwardly simple position.�e7+ As we know. 67 ..e2 but after the text.. moving the king to the back rank loses after 67 . and illustrates very 68 Wb8 .l::!.. 67 �c711 It is still too early for 67 'it>b7? .�e2 White forces the king 67 . is a model of pre 'it>g2 72 l:. Moving 71.. with rook and two pawns l:te2 69 Ik6 .l:. deprives Black of the resource in front of the pawn by 73 "ii'e4+ 'it>f2 .h7+ 69 'it>b8 and then brings up his king.l:te2? 68 . as we know. The only other king 73 Wd7 'it>g3 74 'it>e6 f2 75 'ii'h 1! 1-0 move is 67 .. J:te2. but now the king A really magnificent endgame by does not attack the f-pawn.'it>g2 If we compare this position with that arising after Black's 62nd move.c2 etc.h8.his rook is already on c2.l:!.l:!... cise calculation.. so now 68 'it>b7 l:. and Black is Alatortsev.h8+ 70 l:tc8 wins.l:!...'it>f1 68 Wb7 And. Black once again finds himself in zugzwang. fxe2 69 a8'ii'+ is again check.. 'it>h2.. His conduct of the final a crucial tempo too slow after 68 �b7 stage.. R o o k En dings 66. 123 . for example with After 72 . we see that White has gained a tempo .. . whilst 67 .x e2 cal win.e8 70 a8'ii' l::txa8 71 'it>xa8 against rook and one. this is a theoreti l:te2 69 �cl+!.We3 12 'ii'd s �f2 the rook off the e-file.. l:!.. . i...l:lh3 hs 25 ltf3 es 26 dxes :Xes tion.. which can yield a Trying to keep the king out with 32 passed pawn.l:tf6 Now the threat is 30 . 30 .txc6+ peerlessly accurate.e7 7 lL'lxf6+ lL'lxf6 8 Rubinstein's handling of the ending is i. whereupon Black will be able to use his greater piece ac 32 C4 tivity to exploit his kingside pawn ma Schlechter tries to get his pawns jority. to create counterplay. side is a healthy one. In also hopeless after 32 ...l:.d6+ fol pieces means that they will be unable lowed by 33 .Th e Greatest Ever C h ess E n dg a m es 1 e4 e6 2 d4 dS 3 lL'lc3 lL'lf6 4 i. l:tf2.bs+ c6 11 . this makes all the difference. At the same time.'iti>e4 33 cJo>el l::tf3 124 ....xe7 Immediately fastening onto the <Jitxe7 18 <Jitb1 lthd8 19 l:thg1 g6 20 lL'les white pawn weakness and forcing the lL'lb4 21 c3 l2Jc6 22 l2Jxc6+ .b7 10 i.l:.U.c8 16 o-o-o lL'lds 17 i.xc6 12 lL'lxc6 'i!i'd5 13 lL'les 'ii'xg2 14 1i'f3 28. but the passivity of his 'iti>d3 is simply met by 32.. White's majority on the queen moving at once.l:te61 'i!i'xf3 15 lL'lxf3 . whilst Fine shows to support the passed pawn as effec that passive defence with 32 <Jitd2 is tively as Black is able to do with his.'iti>e4. Black frees the path for his king to advance down the e-file. Black has no material advan tage..gs dxe4 5 the end.e2 'ite6 31 'iti>c2 <Jites Here. lL'lxe4 lL'lbd7 6 lL'lf3 i.U.d3 b6 9 lL'les i. 27 ..l:!.xc6 23 lld3 enemy rook to take up a passive posi ltds 24 . White's two isolated pawns will result in his rook being tied down to passive defence. which forces another passive white response. but he has the superior pawn formation.e3 l:txe3 28 fxe3 29 l:le1 . a6+1 0-1 A nice final touch. with no release of the pressure. R o o k En dings 34 'iti>d2 g 5 3 5 c4 g4 36 b4 h4 37 c5 bxc5 be-passed pawn..l:l. His rook will ments: even if White creates a passed have to be sacrificed for the black g pawn out of his queenside majority. 46 .. 36 . which holds up White's counterplay for a couple of vi tal moves. whereas Black's kingside ma jority is about to produce a passed pawn.l:. with a simple win. meanwhile. bxcs 'ii. in demonstrat White's only hope is to use his king ing what we said in the opening com to support the c-pawn. This 37 'iltc4 variation is instructive. ..f5! and the c-pawn is lost.g3 38 hxg3 hxg3 39 WbS bxcs 40 adequately. Now the white king must spend two tempi transferring to d5. enemy's.l:tc2.g2 :g6 41 . passive pieces are unable to support it 37 . and holding up the 38 bxc5 l:.g1 32 'ii. before his c-pawn becomes a real threat.. but it is going nowhere because it lacks support. his pawn... powerfully supported by its king. 125 . both protecting his own soon-to.>e4 33 b4 gS 34 Wc3 g4 35 cs h4 ••• We see the same phenomenon again. Black annexes the white rook: 42 'ifi>c4 g2 43 'iitd 5 'iltf2 44 The active black rook does double l:txg2+ llxg2 45 c6 'iti>xe3 followed by duty.>f3 41 . White has his passed pawn. ibs a6 4 .l:.hs 10 d3 o-o 11 tt:'lbd2 ds 12 ment.a2! is the 126 ..xas bxc3 29 \Wxc3 tt:'le4 30 cause he has an outside passed pawn.ixgs 20 . is the great lesson of this g5 . 36 . exds tt:Jxds 13 tt:'lf1 . On the principle instructional viewpoint.xc2 40 :xc2 against two..ic2 .l:r.e1 .fe8 27 view.a4 tt:'lf6 in common i s that they involve White 5 o-o .ie7 16 tt:Jxes tt:Jxes 17 lixe5 tt:'lb6 18 ending.ig6 15 above all. these do not 41 l:r. Indeed. and.l:.ib3 d6 8 c3 .f6 14 g4 .ab8 26 l:tes . they are quite that "rooks belong behind passed helpful.i.l:r.U.i.i. because the thing they all have pawns".ixgs 'ifxgs+ Fine makes the point about this po 21 �g4 'ili'd8 22 tt:'lg3 tt:'lf6 23 'iff3 1'id7 sition that. from an ment of the ending.i.l:tds 'ife7 32 l:.a2 38 .l:tc7 . be l:. This.e2 33 whereas Black's 3-2 kingside majority 'ii'g 3 'ifd6 34 'ii'xd6 cxd6 3 5 . However. which means that he will be the first to take an active position with his rook. Although modern analysis has revealed several drawing opportunities for White.c3? really detract significantly from the This is already the first critical mo value of the ending.xg6 hxg6 White also has four pawn islands 39 l:!. d4 tt:'ld7 19 .i.l:.l:!.ie7 6 l:te1 b5 7 . it is Black's move. tt:Jxe4 l:rxe4 31 .1c2 l:!. crucially.xb2 contains a doubled pawn. even at the cost of material.. from a structural point of 24 �h2 as 25 a3 l:.g4 activating his rook at the critical mo 9 h3 .g1 b4 28 l:!.l:tgc1 'it>f8 37 .Th e Greatest Ever Chess Endg a m es 1 e4 es 2 tt:'lf3 tt:Jc6 3 . 40 .a8! This is one of Rubinstein's most fa mous achievements in the field of rook and pawn endings.cs . it is obvious that 41 . White ought to stand better. a8 54 41 J:ta4 42 . It is understand .xg7 . but it appears that he was sim White's all-round advantage in piece ply mistaken.l:. �e7 43 �3 �e6 44 'iii>e4 gS 45 ...l:tf4 58 l:i.. but the further variation S6.l:.l:txf2 60 ..al f6 46 l:ta2 fS+ 47 �d3 �ds 48 'iii>c 3 �c4+ 49 �b3 ! l::txd4 so a4 l:td3+ 51 <il>b4 and 53 l:I.l:td3 •.'ii. as his pieces are simply too passive.l:. Rook E n dings move White would like to play.as and Black cannot make progress.l:tb4 44 as . As Mark Dvoretsky has observed of endgames in general.l:txdS+ liti>a4 61 �d3 with a drawn rook and pawn vs rook position (Fine).l:. the down. and felt that he would have black king does guard duty on the en more drawing chances with equal ma emy passed pawn.b7 45 a6 lia7 46 �f4 �e7 47 'iii>e4 �e6 48 l:. 1 27 . The analysis of the result activity yields sufficient counterplay to ing position can be very lengthy.>c 6 52 as liti>b7 This time.e7+ liti>a6 55 .g4 57 liti>b7 l:txa7+ problem is that it involves a pawn sac 58 �xa7 f4 59 It.. with the as passive set-up. weak pawns have a tendency to get ex changed off for one another. This is confirmed by the strengthen his position.l:txd6 g2 62 llg6 and White gives his able that Spielmann should have been rook for the two black pawns.g6 ds 59 lld6 . .h4+ 56 �c3 c.. but here.>xas 57 .e2 ! An important moment. Analysts such as Fine and Smys lov/Levenfish also consider a more sub tle approach by Black..l:txd4 43 a4 . and Black is free to sessment "=".l:txgs . S3 .h8 53 a6 l:.. too. terial. but a draw.. Smyslov Now White has adopted an entirely and Levenfish stop here. reluctant to play the ending a pawn b) Sl . a) Sl .i.xh3 54 .d2 g3 60 fxg3 fxg3 61 rifice after 41 . but that too seems only enough for a draw: 42 . l:ta4.l:. especially rook endings.. a7 �e4 55 liti>bs liti>f3 56 liti>b6. The now: white rook gives up supporting the passed pawn and switches over to hunting Black's weak pawns.U..l:. typical line runs as follows: 42 <il>g3 . �xh3 52 as .. ..xh4 l:ta2+ and standpoint. objectively speaking. whereas..e2! 69 .t>e2 l':tb6! 57 �d2 . in which the strength of this position is indeed winning for the outside passed pawn ensures Black." With the white rook immobilized.. but advance in computing power to prove Smyslov and Levenfish continue the that the simplifying variation wins for line with 47 . So. it 4S ..l:[h6 d3 64 'it>g2 l:.l:tf6 Wd2 70 l:. ..t>e2 gS! able seven-man endgame tablebases.Th e Grea test Ever Chess En dg a m es 42 .b7 is dismissed by it has taken many decades and a major Fine as "a fairly certain draw". since 46 .t>e7 43 .g3 d4 55 ltf3 l:Ia6! 56 . 52 'ii?c 2 leads to a lost pawn ending after 52 .. Jte2+ 53 l:.b2+ 61 Wel 'ite4 62 �1 f4 63 .t>ds ss ltf4 'iite s ! 59 I:tf3 fs Go l:.l:tb3 l:i.l:txd4? 46 .l:txa3 49 h4 Black.t>c3 51 . it is 46.g6 'ite3 71 :d6 llf2+ 72 Wgl :xf3 and wins. The tablebase's main line runs as White at least a draw. would have to fight hard to hold this 47 We3 'itc4 48 l:td3 dS 49 Wd2 l':ta8 position... SO 'itc2 The first point of Black's last is so We3 llb8! and the rook will penetrate to b2. follows: 52 �el dS 53 . Even so.t>e6 44 'iW3 .. Rubinstein could have won by taking on d4.l:le6 l:. over-the-board gxh4 SO l:ig4+ .h3 l:..in our opinion.. From a practical... his slower choice allows White a further drawing chance later in the ending... his choice of 46 .b4 65 f3+ Wd4 66 l:!. so.t>c4 54 46 .f6 is very claim that " . Rubinstein starts probing with his own rook....f6 48 :xg7 .Jta7 51 'iitd 2 kte7! Now it is zugzwang. �xd4 47 l:.l:Ixd2+ 54 'it>xd2 Wb3 ! so White must surrender the d4-pawn... 1 28 .t>g3 .d6+ 'iti>c3 67 'it>fl .l:[h3+ ...l:[b2 68 .l:!.. White understandable.f6 hard to criticize him too severely.. as we will see.t>e3 ! forces a king and can now be stated with certainty that pawn ending. hoping for a zugzwang.d2 ..t>ds 45 With the aid of the recently avail . xf6 �c3 and to stop the enemy passed pawn as far the d-pawn looks too strong. After his passive choice in the game. 55 Ita1 �c4 On balance. 52 l:tc3+ �xd4 53 a41 . then.. �b4 57 l:tbl+ �xa4 58 �d3 l:tc5 59 �d4 l:tc2 60 l::i. 61 . but it is obvious that he has a much infe rior version of what he could have had Continuing this analysis further..i.l:te6+ terplay to a minimum. it is always useful blebase win) 63 .• �c3 64 Itxd5 l:td3 65 l:tf5 Itxh3 is a ta In such positions. too. This is the second chance in the game that he could have done so. The king will take over the task of blockad ing the white passed pawn. he would probably have Now it is too late to give up material done better to "go active" with 56 in order to get the a-pawn marching: 129 .l:!. . so as to reduce coun here.l:txf2 61 l:txg7 as offering White drawing chances. Rook En dings l:!cl+.:td2+ 62 �c5 �b3 63 l:!. d4 64 l:. whilst the black rook will win another pawn on 56 �e3 the kingside.. but even back as possible. 56. he will have no further chances.d4+ 57 'iitd 2 Itf5! The start of the decisive stage.l:!... Spielmann defends passively again.d7 54 . Smyslov and Levenfish then ana lyse 56 .W3 68 l:lf6+ �g3 69 :g6 draws.la3 At least White has managed to get his rook behind the passed a-pawn.a5! ..a6 �d3 66 'iitd 5 'iite 3 67 . it appears that Spielmann could probably still have drawn by activating his rook at this moment.b 7 . 58 �e1 but here...f7 (63 l:!..a7 54 . .a2. by playing 41 l:!.. 65 l:. :as 62 �d2 <iitc s and Black simply I:tc1 l'1a4 69 .. example of the more active rook. flank (doubled g-pawns notwithstand 6g.Wc6 70 l:th7 .l:tb2 60 a6 Ilb8 61 6S l:te2 . tivate his rook at the cost of a pawn.l:lxh3 64 as l:th8 different points White could have 65 a6 l::ta8 66 a7 �bS wins as in the drawn. 1 30 . and still a marvellous 65 a6 l:th8 66 a7 .Uxh3+ 63 Wxd4 .Th e Greatest Ever Ch ess E n dg a m es 58 as .l:ta6-c6. ending..l:th1 trades his d-pawn for the a-pawn.•.f4 66 �e3 Wb6 67 I:tc2 Wb7 68 a7 .a7 71 We4 <iitd 6 72 ing).f4 72 g6+! .. 62 . 61 lila2 The last tactic to wrap things up. had he seized the chance to ac note to White's 58th move.Uxf2+ 59 <iite 1 .l:ta8 67 <iitf1 ..... then the combined force of the 0-1 black king and d-pawn decide: 61 <iitf1 One of Rubinstein's most famous . �5 ss.l:!. Black just plays . How 61. the analysis shows that at two 62 Ita3 <iitb4 63 l:ta1 .l::th4 62 <iio>d 3 ever.l:txa7! etc.l::th4+ 64 'i.:�b4 59 We2 <iita s 60 l:ta3 l:i.l:!.td3 l'!xa4 This is the single biggest lesson of the Now the ending is a theoretical win..Uh4 62 <iitg 2 <iitb4 63 .l:1a1 d3 64 as d2 rook endings... with If he maintains the cut-off on the c a winning 3V1 majority on the other file.. If the white king defends the h 73 <iitxg6 :xh7 74 �xh7 <iite s 75 <iitg6 g4 pawn. but in fact. passed g-pawn. 15 "ifh3 . This is an extremely famous and also highly instructive position.ld7. What game continuation. if joined by the white king.i.>h4 l:.b7 8 o-o �e8 9 force. l::tx c3+ 37 'ii. it probably looked to the spectators as though Now 38 g6 is less effective after things had gone wrong for the great 38 . 41 l::tg 7+ 'it>f8 42 l'lxc7 f4! 43 �gs 131 . and Black's rook so active.l:txe8+ 'it>xe8 27 h5 journal 64.l:te8!.. whereupon Black can now take on c3 in better circumstances than the game: 36 . Alexander Gol 34 i. Cuban legend.xe7 .d3 i. 'it>g8!?.f3.!2Jxc3 11 bxc3 'it'xe7 this happens. Capablanca This is the big difference from the knew exactly what he was doing. not to be confused with the US l:tc6 31 g4 lbc4 32 g5 lbe3+ 33 �f3 lbf5 resident Grandmaster. .xf3 13 �xf3 lbc6 14 l:tfb1 l:.. The black rook has he had understood was that White's managed to return to the back rank.. a few extra black pawns 12 a4 i. combined with the with the fs-pawn still a distraction for rook's control of the seventh rank.>f6 .h7 master. Once 1i'e2 lbe4 10 i.f7 23 h4 d5 24 cxd5 exd5 In issue 12/2003 of the Russian 25 1Wxe8+ �xeS 26 .t:l. Rook En dings 1 d4 e6 2 lbf3 f5 3 c4 lbf6 4 i..i.. With the white c-pawn indefensible.xf5 gxf5 din) suggested the paradoxical move 3S . Now White has nothing bet ter than 36 l.g5 i.ae8 on the queen side will be an irrelevance.h1 'it>fS 30 .e1 1i'd7 19 e4 fxe4 20 �xe4 g6 21 g3 35 <fitg3! �XC3+ 'iWS 22 �g2 l:. White.l:[f6 28 hxg6 hxg6 29 . Vladimir Goldin (a Russian .e7 5 would be turned into an irresistible lbc3 o-o 6 e3 b6 7 i.t:tf6 16 f4 lba5 11 "iff3 d6 18 Play continued: l:!.lhf4+ 39 c.>gs l::te4 40 c.l:. and thus allowing the white king into Thus far was all given by Goldin in f6. a well-known theoretical draw - l::tf 7+! 'iitg 8 42 l:!.g4 can win against the toughest defence (48 .'iitg 8. For example. several months after Goldin's analysis appeared.b 1+ 54 'it>c6 l:.. over whether White which appears to win.c1+ 55 Wb6 l:Ib1+ 56 <Ji. as the silicon beast eliminated the black f-pawn earlier) persists in claiming White is winning ! 43 . However. In the position of the last diagram.xd4 51 36 . is winning..l:lg6+. he should continue 38 a) 52 'iitxds? l:. 1 32 . 5/2004). Some 15 his original article. White has known to Fritz 12. when he claimed a years ago.f7+ 'iit g 8 45 ltxf4 l:.g4! is the Vancura po 'it>hs ! .l:th4 b s and Black establishes h i s own passed pawn. e. . as confirmed by the table base. But Zaitsev flared up.l:.b7! :g4 SO l:txb6 l:. Karpov's former trainer Igor Zaitsev.d7 . again in the pages of the found the much stronger 48 WC7! Russian journal 64. Iig4 46 .l:lxa7 .c7 .xd4 so as l:tb4+ 51 deserves the credit for suggesting this 'lites l:tc4+ and now: defence).g.U. one of the Soviet chess school's leading analysts for many years.l:!..c1+ 57 'iitd 8! I1a1 58 l:txds and Black will not succeed in reaching one of the drawn rook and rook's pawn vs rook positions. published a reply (see 64.:!. Vladimir Goldin as wins) 49 'iitxb6 .f8! 46 . with devastating effect.l::!..xc7 l:r.l:Ie4 44 Wf6! . in which he pointed out that White still seems to be winning after 3S . .a6! (once again.. 4S .. 53 l:.. who is Wf8 46 l:txa7 Iixd4 is less good) effectively playing with an extra king..xf4 39 'iitg 6 'iin8 40 Wf6 lle4 41 sition. by removing the black rook from a posi tion to give check along the sixth rank.l:lc1 and White.. and is in no danger (Goldin).Th e Greatest Ever Chess E n dg a m es a6 44 l:..'iftf8 49 l:.l:tf4+ 45 'it>es (45 'it>g6 b) 52 'it>bs ! .e8 43 'iftxfs (com something which appears to be un pared with the previous line. a large analytical discussion draw after 48 'it>xds 'iftf8. 36 'it>h4 l:tf3? This makes things easy for White. 48 ...l:txgS+ 47 Wd6 . :tg4! . with his king in front of the 'it>c6 l:r.Barsky ..even if he loses the c l::tb 8 ..l:th3+! 43 <t>g4 b3!! 44 :t.. thus preventing the white king marching into the black po sition in the way it does in the game.bs l:th2 46 .. e. analysed by Kasparov 48 .i:.f7+! (this move.xds b2 45 .bs 38 axbs axbs 39 <t>gs b4 40 . issue 7/2003. and secondly. is the best try.l:!.g..l:tg3+! 42 <t>hs (42 'iitf6? . 133 .<t>g8. the black rook remains ready to check on c6. Thus. it starts the process of creating a passed pawn on the queenside..c2+ 47 <t>d7 <t>xfs 48 g 7 l:tg2 etc. Black can reach the Philidor po to hold.places White in a curi ous zugzwang. 40.d7. where he must lose a pawn) and now the leading Russian study composer Oleg Pervakov found a study-like draw with 42 .l:txfs . 44 fS .:t.i:..l:Ib7 <t>g7 47 <t>gs I!g2+ 48 �s This move has two purposes: firstly. b) White's second try is 41 l:.. R o o k E n dings Now White has two tries: a) 41 . after the obvious 37 g6 there fol lows 37 . after .e2+ 45 <t>xds �6 46 sition.) white pawn.l:txb2 I!d6+ 52 <t>es <t>xg6 with a theo 40 <t>xfs b3 41 �es b2 42 l:th8+ �g7 43 retical draw . <t>h6! 49 ds l:td2 so 'iitf6 l:txds 51 in 64.c2 Black has enough counterplay pawn.. Jk6+) 40 ....l:.b8 1Ic2 ! 43 'it>g s position to a draw: 41 .l:.. Kasparov analysed the b 3 41 ... l:tg2 4S l:!.b6 is the threat) 49 tion: 44.xc7 holds) 47 d6 l:...l:..'iii'g 8! (threatening 40.Uc6! (49 ..l:.l:. b2 44 'iii'xfs l:.... After 45 dS (threat 46 d6) 4S ..) l:txb2 'iitx g6 so d7 .. However...lh6 ! and the rook gets behind the a-pawn in time.bxa4 is met by 39 'iii'xfs a3 40 I.c6 41 l:th8+ 'iii'e 7 42 l:!.�a3! drawing..b3 42 l::rx ds b2! (43 .l:th3+ 40 'iitg s l:lxh7 41 gxh7 'iii>g 7 when it is Black who wins!).xb3? 'iitf6) 43 ... 'iii>g 8 (now 38 . his position. White fares no better with 38 as �c6 39 g6 b4 40 'iith 6 134 . e.. Now after 38 axbs axbs 39 'iit g 6 (39 g6? fails to 39 ... a6.Th e Gre a test E ver Chess E n dg a m es but here.. l:Ic2 ! (even 46 ...b7 'iii'g 7! with a posi and White has no way to strengthen tion similar to the previous variation. White should start with 37 'iith s which the black king already stood on bS.g7+ 'iii'f8 40 llf7+ 'it>g8 41 llf6!.cS+! 48 'iii'e4 (48 'iit e 6 �c6 49 .g7+ 'iitf8 41 l:lf7+ 'iii'g 8 42 l:tf6 b4 43 �a6 43 . and then play 38 'iitg 6!!.a7+ 'iitd6 46 'iitf6 etc. 'iith 6 47 d6 l:txg6 48 dxc7 �c6 49 l:txb2 .xc7+ l:txc7 so dxc7 bl'ii' 51 c8'ii' 'ii'e4+ is a drawn queen ending) The preliminary pawn exchange on bS is weaker.Uh8+ �e7 42 l:!. This all resulted from meeting 36 . in 36 . a6 with 37 g6. too. 39 l:!. 40..g2+ 46 It was again Igor Zaitsev who found 'iii>xfs White has effectively gained a the winning path for White after tempo over the previous variation.d6 and a draw. 38 ... h6.. this is still not enough to win: 46 ...l:.l:tbs llc2! 44 l:.l:. there is 39 . ...a8 a2 43 l:.g.. Another try analysed in the pages of 64 is 37 'iith s bs.xa6 :c2 44 g6 with a winning posi 48 .. analogously with the game.g2 43 . in view of Zaitsev's winning line. as it would expose the lanca's task immeasurably more diffi white king to checks from behind. rook and pawn threatening mate. After 39 �f6 which were one of the key features of We8 40 l:Ixc7 l:!.b4 42 l:txa6. at the same time combined force of king.g.bs 38 axbs axbs 39 g6?. secures his king from lateral checks.... . Leaving the 37 g6! black pawn on the board.. this commits the rook too early. the difference being that Black no longer has a rook check on h3. White will White's idea is seen with full clarity. in this variation.g4 42 llxa7.. His capture the a6-pawn. .. and White can now win by sim ply 38 g6. a6. as was the case in the line with 37 . which Black had in the analogous position in the 38 axbs variation... The big difference with the line beginning 38 axbs is that here. .xd4 is no better. Note how. The conclusion of this long analyti cal diversion is that White appears to 39 'itrf6! be winning even after 36 .a3.:xd4 40 'it>f6 .xa4 41 g7 . but this Note that 39 'it>xfs? would be a seri move would have rendered Capab ous mistake. force of king.. 39 . e. White will win the black rook with 43 The only remaining question is g8"ii' next move. the excellently placed white rook on the sixth rank 3 7 l:txf4+ 38 'it>gs l:!.. 38 ..l:!. and the essence of from the exchange on bs. White has an extra pawn on a4. whether. Rook En dings Now we see the point of refraining straightforward... so as to stop the white king entering on g6. cult than in the game.l:. rook and passed g-pawn will again be too strong.l:i. Black can meet 37 'it>h s with the immediate 37 ...'it>f8 43 'it>f6 and the g-pawn will promote. and his combined is too strong.. Black's defensive plan. For example. 41 . to serve as an Now things are much more "umbrella" behind which to shelter the 135 . 42 .l::tf4+ etc..llc6 (instead of 37 .bs).. However.e4 . This cuts out the defence 42 . just as in the game continuation.. supported by its king and rook.l:..xc7 l:te8 42 cJi>xds Wf7 48 cJi>d6 a6 49 cJi>c6 etc. have added a great deal of in 45 g7+! �g8 terest to this classic endgame.l:te4 43 'iitf6 l:tf4+ 44 �e5 l1g4 46 l:. <ii>xf5 .. The analytical discoveries made over the last 15 years. 47 �xd5 l:lc1 48 �d6 l1c2 49 d5 l:tc1 50 l:lc7 l:!.. J:tg5+ 47 �d6 is no help.xa7 l:lg1 46 . as cJi>c6 will pick up the b-pawn.cJi>g8 40 :g7+ cJi>hS 41 l:i. from which generations of players have learned the strength of a passed pawn on the sixth rank.Th e Greatest E ver Chess Endg a m e s white king.xg7 46 l::tx g7 cJi>xg7 47 39.. is a common theme i n rook The king and pawn ending i s hope endings.a1 51 �c6 l:lxa4 52 d6 1-0 A very famous ending.. less after 4S .. 136 . by various Russian players. the defender made things clamped by g4-g5...i. the the text is positionally double-edged.tc2 'i¥g5 17 f3 .tg5 .xc1 tbc6 30 tbxc6 . Flohr demonstrates that White can win by force".tg4 14 holding back the black c-pawn.ad1 11if6 20 .l:!. easier by failing to seize the opportu 36 b4 .. but. since Black cannot af ings. It is true that Just as in Games 26 and 27. and claimed that after it.l:.td3 g6 12 dxc5 tbxc5 13 o-o . stronger side's advantage is based on because it weakens es and allows the the superior pawn formation.l:tc1 tbc5 23 . 1 37 .fS anyway.b3 .l:tb8? nity to activate his rook when he had This passive move really does leave the chance.a5 l:la8 e3 o-o 6 tiJf3 lbbd7 7 'ii'c2 c5 8 cxd5 35 �d4 tbxd5 9 . Flohr gives one of the classic demon 3S . R o o k En dings 1 c4 e6 2 tbc3 d5 3 d4 tbf6 4 .l:!. 'i£tc7 36 g4 actually forces the reply strations of how to handle such end 36 .td7 18 l:tfe1 �Ud8 19 l:. just as in Spielmann ford to allow his kingside to be Rubinstein. " . .l:tc8 31 . Black in desperate straits.l:!xc1 29 .ac8 15 'ii'd 2 a6 16 ..te7 5 32 'i£te2 �e7 33 �d3 �d6 34 .. As we will see. Instead. it prevents tively than their opposite numbers.. g2-g4. which white king a potential entry on the allows him to post his pieces more ac square..txe7 'it'xe7 10 tbxd5 exd5 11 White has achieved his first aim. and Black's real error came on the next move.. this is not true.l:ted1 "i!Vb6 24 tbe2 tbd7 25 'i¥d4 'ii'xd4 26 tbxd4 tbe5 27 b3 'iW8 28 'iW1 . of ....l:tc5 bxc6 35 .. As Fine himself acknowledges.f51? Fine criticized this. On the other hand. tbd4 ..txa4 tbxa4 22 .ta4 21 . without pre senting White with a passed d-pawn. by bringing over his king to effect because of the strength ofhis passed d a "changing of the guard". but this proves in a) The immediate 39 ..gS!.e2. Vidmar tries to terplay: defend passively. cause after 40 exf4 �e2 41 g4 White 37 a3 l:.l:. whilst the alternative 40 g3 can be met by the ingenious 40 .a3. we see here that White's the black king abandons the centre.a8 will establish a winning passed pawn by f4-fS.Uc4!. but passed pawn is actively supported by in rook endings. They conclude that Black has sufficient counterplay after 41 l:.e4..c3 l:.l:. as White lacks a strong passed pawn. tinue his policy of seeking active coun Instead of all this.. free to roam. and has a target on e3. gests as stronger still the move 39 . rook.. in order to activate his rook still further) 40 exd4 .xc6 l:i. Dvoretsky sug l:. b) Smyslov and levenfish analyse 39 ..f4. It is true that example.'it>b6. .l:. g4! 41 f4 .'ifr>c7.. which appears to yield Now Black has to decide how to con enough counterplay to hold. with the idea of 42 l:lc3 . The idea is to activate the black rook. planning ...f4? fails. After 37 'iii>c s 'iii>b 7 the black rook is their king..g7 and argues 1 38 . As in the Schlechter-Rubinstein 36 . whereas Black's passed just so crucial that this must take prior kingside pawns lack the support of ity..l:i.Th e Greatest Ever Chess Endg a m es Black should take steps to activate his that White maintains the advantage. preparing the break .l:f. be adequate. The point of all these lines is to activate the black rook at all costs. activating the rook is his pieces.xh2 43 a4 gS but Dvoretsky continues this line with 44 ... c) In view of his improvement at the Play might continue 38 Wd6 l:.xc6 l:. Black sacrifices a pawn.e8 39 end of this variation.c7+ '>t>b6 45 . Now 40 lic3 f4 41 exf4 gxf4 42 . . d4!? (another very typical move for rook endgames. with pawn.e2 gives Black enough counterplay.xg2 42 l:i. e.l:..gxhS+ 45 <ii>x hs is also hopeless. "widening the front"..... but while White low from 45 'it>f3?! :f8+ 46 'it>e4 �f4+. move. move 36.. to take up opposition vis-a-vis 52 h4 . thus eliminating possible 'itf3 . typical "do not hurry" Now Black has a passed pawn. self admits that it remains unclear 38. the 4S . Rook En dings 38 e4! gerous that White cannot win. 4S. h6 42 h4 'ite6 43 'itg4 :taB 44 hS! chances after 40. Smyslov and Forcing a breach. implement the defensive plan indi 41 �4 cated above.l:.f2) 41. and Flohr's execution is highly rook.l::tg 7 c4 centre.e7 as! (offloading one of the weak pawns and preparing to acti vate the rook along the a-file) 43 l:.. Now the h6-pawn is fixed as a weakness.h8 so l:txg6+ 'itbs 51 .'it>b6 42 . but still offers some defensive 41 ... White chance is Dvoretsky's more restrained has to clear some space to give his king approach 41 h4 'itb6 42 g4 but he him a route into the enemy position. Levenfish then give 41 �es (41 'ifilf4 gets 44.xb4+...l:th4 47 h3 lth6 (the rook goes black counterplay. its enemy.. This is another typical idea in such Probably White's best winning positions.fxe4 39 fxe4 dxe4 40 <ii>xe4 Zb7? whether White's advantage is enough Here. White takes control of the which provides vital counterplay: 46 square f4.J::tf8+! 42 'it>gs ..g..g8 46 g4!. so as to free his by force. and transfer his king to Now the position really is winning defend the a-pawn. brings up his king to exploit this. but 44. such as would fol temporarily passive.l:.xh7 axb4 44 axb4 l:.a4 45 l:lg7 .gs nowhere after 41..'itc7!..a7 46 'itf3 black b-pawn starts rolling) 48 'itf4 cS Now the white king returns to the 49 'itgs l:. Black misses his last chance to to win.l:.l:tc8 and Black's pawn is so dan.. It is less effective now than on instructive. 45 g3! A very nice... 139 .l:!. :as 47 �e4 l:!a7 48 �d4 �d6 49 hopeless king and pawn ending. required.c8+ Now Ith8 is menaced.. Now the c-pawn is lost.Th e Greatest Ever Chess E n dg a m es 46.... no matter how :a7.e5+! line 52 . 51.l:.e8 53 l:!. the stage is and now his king penetrates on the always reached when play becomes queenside: 53 �d4 �e6 54 �cs and concrete...�c6 The point of Flohr's last move is the 50 l:i. an endgame may involve.l:te5 �c6 56 51 .>d6 ample of the value of accurate calcula If so .i. The 55 . 51 l:!.>g6 'it>xb4 59 'it>xh6 as 60 side or the other.xe7 �xe7 53 �es is a ..i....cs ! Itc7 S2 l:tas! tion in the endgame..h8 cxb4 54 llh7+ �b8 The final. �c7 53 l:i. in time to stop the a-pawn. J:te7+ 52 . decisive manoeuvre.clock time and psychological effect.c5 53 .c.l:lf7+ 58 l:tf6 1-0 140 .l:. moves.l:txc5 . �e4 �e6 52 l:td8+! Once again. a little repetition of Precise to the very end. and accurate calculation is wins. White's last two moves have much slow and patient manoeuvring transferred his rook to as with tempo...l:txa7 �xa7 56 axb4 �b6 57 �fs black king must give ground to one 'itr>bs 58 c..�b6 54 . whereupon the white �xg s a4 61 h6 and White queens just king will enter.i:th7 55 .i:te6+ Wb5 57 'iirf5 .Wf6 then 51 .. for the usual reasons . 52 . A neat ex 50 . b6 28 f4 �6 29 �3 es 30 fxeS+ lbxes+ 31 lt:Jxes �xes 33 b4! Fixing bS.g7 4 e3 e6 S lt:Jc3 mean endgame player. without a discernible mistake from his opponent (who was also no 141 . combined with the march of the white h-pawn. White's plan is to probe the black king side.g2 i. This difference in activity proves enough for Andersson. It seems barely credible that White 33 J:lb7 34 h4 . R o o k Endings 1 c4 cs 2 g3 g6 3 i.xg2 17 �xg2 lt:Jc6 18 lt:Je4 .xd3 27 lbxd3 .c3 .d7 1S 0-0 i.l:!.d6 2S l:tb3 bs 26 l:!.d2 o-o 14 l:lc1 i. as well as securing the white rook in its active position.• should be able to win this position by force.ac8 20 i. with the extra island on e3. lt:Jc6 6 lt:Jge2 lt:Jge7 7 d4 cxd4 8 lt:Jxd4 ds 32 lieS+ �d6 9 cxds lt:Jxds 10 lt:Jxds lt:Jxd4 11 lt:Jc3 lt:Jc6 12 'i¥xd8+ lt:Jxd8 13 i. In common with the standard "two weak nesses" approach. by the way).d3 l:!. by threatening a king penetration via f4-g5.fc1 l:Ib8 22 a3 as 23 l:Ic3 fs 24 tt:Jcs l:!. the great endgame grinder of the modem era. to win the position.l:.c6 16 11c2 i.fd8 19 . and White even has the nominally weaker pawn struc ture.xg7 'ittxg7 21 . as a result of which his rook is able to take up a more active position.U. But it is his move.l:r.i. Material is equal. a4 tions and keep them hanging over The best chance. Although the de Black's head... 35 axb4 would simply reduce his hopes 36 . or even. in some extreme passive play by Black can be met by the cases. 38 .. in this case.. for example..gxhs not think he will be able to hold the would leave the black pawns too weak. I do from coming with check.l:. I think Andersson would g6.. has two options: either to exchange on Instead. 3S �f4 �e6 Naturally. White now 41 .l:td7 •• Kaufeld and Kern.a6 fxe4 43 �xe4 .. The beauty plan of e3-e4.�f6. after which the black kingside would cave in.. 3 S . their suggested fol low-up for White is unconvincing: 39 g4 gS+ 40 �f3 fxg4+ 41 'itxg4 and now it seems to me that White has lost 36 hS much of his advantage after the move With his pawn on h S... 'ite7 41 e4 �d7 42 That means White can retain both op.h6.g. criticize this bid for activity. creating a weakness on g6. The text move fixes the white weakness on a3.e7 44 �ds. which pro �f7 40 h6! (40 hxg6+ hxg6 41 �g s . J::!. g6-g S permanent thorn in Black's flesh.. a rook sacrifice on g6. since he can stop .. followed by a king advance to have to choose between these two op. and instead recommend 38 .l:. whilst . 40 . forcing the exchange of of White's position is that he does not pawns. position. in their recent book on Ulf Andersson. h6? would just walk into the undermining blow 36 hs..e7 vides various options...l:!. or to have answered 38 . e.'itf6 with 39 . axb4 cult.. 34 .. such as attacking is much less convincing). thereby making the lat fender is generally recommended to ter's defensive task much more diffi exchange pawns. With the pawn on h6 a tions just yet.. However. so that Black has the idea of activating his rook at some point and going after that pawn..c6+ advance the pawn to h6... �6 37 l:tc6+ �f7 38 l:lcs of counterplay. Continued the h7-pawn. 142 . cs.Th e Greatest E ver Chess Endg a m es 34. l:i.d7.l::i.b 6 .'it>g8 41 h6 is much like varia 49 . After 41 . . with the differences that the white pawn on h6 is again an enormous strength.l:..d7 'it>g8 46 'it>f4!. whereas it is very counter l:tg7+ �h8 45 l:. However. White retains some win ning chances.b7) 48 'it>f6 l:f. After the further moves 46. and meet book where a timely bid for activity ing 41 .. are hopelessly passive.d 3 40 :b7+ 'it>f6 'it>g s l:txe3 (47 . the various threats (including �a8-h8) All in all.xbs .e8 40. now we see the great power rook ending.l:. it is possible that 38. Rook En dings That only leaves the active bid 40.. White is slower in winning the bS-pawn. l:te6 48 .xbs l:txa3 43 l:tb7+ 'it>f6 44 �b6+ 'it>f7 45 �a6 (44 �xh7 gS+! 45 'it>f3 g4+! forces the white king back and draws)... I am certainly not surprised In the light of these variations. All of Black's pieces was slightly better than Miles' choice. could have saved the weaker side in a However..:e7 already seen plenty of examples in this stopping the e3-e4 advance.l:.l:.g7+ 'it>h8 so �b7 Black is again tions examined above.e4+ 42 'it>f3 �es. We have next thought is the defence 40.a2 143 .l:ta7 .l:. and the white king marches into f6. 41 l:txh7 gS+ 42 'it>f3 .l::i. Miles' choice is therefore of the pawn on h6: 43 l:tb7+ 'it>g8 44 natural..b6 with 41. and clear if he is actually winning by force...l::i..l:txa3 43 ... although it is not quite White will push his pawn to gS. but on the other hand.l::i.'iii'f6 will be too much. hopelessly passive.Ue4+ 47 39 . intuitive to allow the white pawn to h6 before seeking activity. one's that Miles played what he did.l:!.d 3 42 . This is similar to the game... a3+! 'ite4 from a position which many club play 52 'it>g4 llg1+ 5 3 llg3 (Panchenko) is ers would probably assess as a com also winning for White. plete draw..b1 51 l:.c6 Defending the a-pawn is no help: ...g1+ 51 'it>f2 !lh1 52 . 1 44 .b1 grandmaster with consummate ease.l:tb7 59 �g3 1-0 47 .l:!.l:..l:..lla2 48 h6 .l:txa4 <lte4 49 'itg3 . and he Black actually threatens mate in resists further mainly out of inertia.h4 54 l:.'it>xe3 so h6 ..g2+ 47 'it>h3 .g6 'it>e4 55 b5 'it>f5 56 li!.l:.l:.l:.l:.e7 58 b6 . one! 50. 49 .Th e Greatest E ve r Chess En dg a m es 44 :ta6+ 'it>e5 Black's position is hopeless.a1 49 bS is the sim An almost magical display by plest way to win.l:tb2 'it>g3 . Andersson.l:te4 57 'it>f2 .a6 'it>d5 53 45 g4 fxg4+ 46 'itxg4 l:r. . who beats a world-class 48 . It is a perfect example of 50 h6 the value of activity in rook endings. and was of great 37 .g7 s some years earlier. 5J..d1 Despite being Black. which Black would find it hard to cross. this plan cannot be realized. 145 .xe7+ (37 i.xd7 tt::lxd7 Fischer certainly thought he was 32 Ji.ac8 20 rook and bishop combination on an d6 exd6 21 exd6 Ji. In principle. R o o k En dings 1 c4 g6 2 d4 tt::lf6 3 tt::lc3 ds 4 tt::lf3 Ji..te2 tt::lc6 10 :d1 tt::lb6 11 'iWcs matches against Smyslov..tc8 'iitd8 he actually loses personal and political significance for the bishop! them both.d7 31 li:!.c7) ing World Champion. since the black knight lid3 tt::lc 5 27 . 36.xe7 38 . Then he would establish a barrier.l:!. Fischer had White still wants to get his bishop seized the advantage from the open to the important square c2.c8 li:!..g4 8 .txes .e4 38 i.l:i. The fact that White has the 'it'xf4 18 .i. and this may well have influenced his play adversely .e7 . in princi ple. White would be happy to exchange rooks. bishop to c2 and pawn to f4.'>t>g7 37 li:!. 33 .:. mous of both players' careers..l:i.txf4 tt::lxc5 19 tt::lxc7 l:.e1 Wf7 34 'iitg 2 tt::lc 5 35 ..l:. from where ing. bishop on g4.txf3 13 gxf3 :tfd8 14 ds Black is just a good pawn up and tt::le 5 15 tt::lb 5 'ikf6 16 f4 tt::le d7 17 e5 should win.l:.te3 one of his world championship tt::lfd7 9 ..d5 b6 24 Ji.e3 e5 28 .e7 36 l:!.f3+ Botvinnik points out that..xb2 22 o-o tt::l bd7 23 open board offers him little compensa l:.xe5 29 has such an excellent post on cs. l:f. 'it'd6 12 h3 .g4 .he thought it was going to be a little eas ier than was really the case. when preparing for 'ii'b 3 dxc4 6 'iWxc4 o-o 7 e4 Ji.c3 . refuting over the board some old it dominates the black knight and eyes opening analysis Botvinnik had made Black's kingside pawns..xe5 l:txd6 30 l:i.f3 tt::le6 25 tt::lxe6 fxe6 26 tion in this case. Unfortu This game was one of the most fa nately. After the sequence 36 played when Botvinnik was the reign '>t>f3 hs! 37 . if he could get his king to d4. It was because of the poor position of the their only meeting over the board.Ue3 .l::tc7 winning easily. .b4 He considered this would have won 42 a3 l:td4 43 f3 as when White is vir without undue trouble. and accord ing to Botvinnik. just play a couple of waiting 40..Jle7?! 146 . Botvinnik was expecting overnight. Botvinnik recommended 41.:tel.. This is a very impulsive decision.txe4 l:txe4 43 . and then king on d6..c7 was more accurate here. 38. 41 'it>g3 42 .e1 when the bishop makes no sense at all to take a commit cannot come to c2 because of 39 i. tually in zugzwang..l:!..l:!. sis. imply ing that it was tantamount to an insult for White to play on.d4?! The players had now passed the time And this allows White to achieve control and were about to adjourn exactly that.. 41 4Je4+? . adjourn the game. for 43 . .c 2 'it>f6 40 �f3 �gs proach for those days was to do noth Here Botvinnik recommended ing. this decision to exchange mi nor pieces was also very impractical. Fischer clearly thought the position was an easy win..l:i. In such circumstances.•..a3 43 I:.Th e Greatest Ever Chess E n dg a m es which Botvinnik severely criticized his opponent.. he made this very clear in the manner of his play.. Quite apart from the fact that he greatly underestimated White's resources in the resulting rook ending.. defending the knight. The pragmatic ap 39 i. .l:i.c2? tal decision like this immediately be . it the stronger 38 .�es with the idea of placing the moves. fore adjournment. and find the winning plan in home analy then advancing the queenside pawns. it also forms the start of coun played simply 43 . as we pointed out.xh 7 ! ! is a great move analyses 4S . As Botvinnik nected passed pawns.:cs own passed pawns.xb4 axb4 46 f4+ <itfs 47 will be able to harass the black king �3 'it>e6 48 'it>e4 'it>d6 49 �d4 bs and with checks from b6 and b7. Once the black the king and pawn ending is winning queenside pawns start to move.U. allow Black to improve his position sys and apparently keen to demonstrate tematically.. but it is also one 48 h4+ �h6 (48 .Jba4 48 h4+ 46 llf7 l:tas 48 . the king will have to ap 44 . White will create counterplay with his 4S .:lc4 46 as! bxas 47 . R o o k En dings Fischer was in no hurry to adjourn. in l:txh7 mate) 49 :ld7 as a draw. Jhh4?? 49 f4+ 'it>hs so more demonstration of the fact that.U.xh71! Botvinnik cites the following varia- This idea is the key to White's de.. after which Black cannot win. White rook endings. is counter-intuitive. as 44 l::tb 3 l:.b4 when terplay for White. However.f7 a6 and a wonderful idea. and there is no reason for this to both his opponent and the spec White to hope to draw. 47 .l:..f3 l:tc7 45 a4 proach the queenside. and his hurried play cause it presents Black with two con soon spoils things. Passive play with 47 . be player as Fischer.. 47 . After removing it.l:. tion as typical of the ideas outlined in 1 47 . here Black could have will see. In order to Black has the opposition.. White for Black: 45 .. activity is almost always will play 50 .l:!.l:!. Botvinnik Of course.. avoid these. and Fischer sealed his move. preferable to passivity. fence. The text. thereby aban Here the game was finally ad doning his g6-pawn. journed. rook endings are noto mously found by Geller during over riously difficult.�fs 47 ..l:td6 and take the a6-pawn. However. The strongest move. even for such a great night analysis.f4 would considering his position easily won.. fa tators.. l:tg7 Immediately Black undefends the g pawn. First. bs? out that White was drawing.b7+ and White draws. and for some 15 One of them..g7+ We6 60 . so. �6 49 l:1b7 claimed the main line of the analysis (holding up the queenside pawns) was found on the plane back to Mos 49. as the game continuation exercise to pupils in his chess school. the variation is a fascinating analytical story. a certain 13-year old years afterwards. the arguments raged called Garry Kasparov.xg6 b5 53 h 5 148 . by the threat of a check on g1..a1 Defending g6 tactically.Th e Greatest Ever Chess E n dg a m es the previous note: 48 .. shows.l:r.<it>d4.. Fischer claimed to refute this and show that Black was winning after all...l:l. In My 60 Memorable Games.g5+ Wc6 58 . way for White to draw! Fischer first claimed that Black was Let us follow the twists and turns of winning after 51. but Botvinnik then shot back and pointed s1.l:!... found another over whether Black could win here. But then Euwe suggested the much stronger try 51. Botvinnik pub lished analysis. White goes after it. After the game. Botvinnik's after the Olympiad (Botvinnik actually initial analysis ran: 52 .:txg6 bs 53 We3 as 54 �7 53 . The text move. 49 .. even in The critical moment....l:tf7+ <it>es so . missing the reply.g6+ Wd7 59 .'it>ds.b6+ draws after 52 . whilst keeping White's king at bay for the time being. since �d3 l:ta3+ 5 5 <it>c2 l:ta2+ 56 Wc3 . but immediately this famous analytical duel. 51 Wf3 57 l:. showing a draw for White. definitely only Botvinnik gave the ending as a training draws.l:ta5 (the only way to set the pawns cow!) it was established that White in motion) so Wg4 b5 51 f4 a6 52 . Then. All in all.l:!. in 1977.l:!.l:.l:!. finally. Fischer's line.a7 etc. The text brings the black king over.l:. which leads to very long and complicated variations..xf2 running the black king towards the queenside allows too rapid counterplay. 'ili'd1+ 66 We3 . .'ti'b3+! 65 'itre2 (65 'iii>g 2 'i!Vds+ 66 f3 'ii'd2+) 6S .g6+ (in order to understand 51 .b2 60 h7 b1'i!i' 61 h8'iV but now Black starts a winning attack with 61 ..'ii'e4+!) 59 . because after 61 l:tg7 as 62 . 5 5 l:tg4+! (55 h7? llh1 wins) s s .. /�'dS. but Fischer continued with the se quence 64.gS+ <itc6! (the best try. as he has retained both queenside pawns) 61 .l:f. As a result.l:.i:i.g.. 'itrd4. it is coming across to stop the queenside instructive to look at what happens if pawns..b7? .g7+ 'iii>a6.'iti'b4 57 ..c7+ 'itrd2 62 .l:f. <ita3 62 !lh4 b2 63 h7 b1'ii' 64 h8'i¥ and draws..b1! 149 . Black cannot launch a winning attack on the enemy king. as analysed by Fischer: S6 . Black's idea is a "long march". coming forward allows the white rook to de fend from behind the queenside pawns..b4 54 h6 b3 emy rook coming to h1 . <itas 60 . . .. and there is no queen check on e4.. the note to move 59).l:tb7 and draws) 57 .. compared White's rook getting to b7.g7! b2 58 h7 . commencing with a rook check.there follows 59 .l:. as opposed to 59 . Rook En dings (the point of S1 .l:!gS+ <ita4 61 l:tg4+! (now it is too late to switch be hind the b-pawn.l:....citcs 56 . Botvinnik's analysis stopped here. is that the black rook is boxed in by its own king. so his only hope is counterplay White plays 59 'it>g2? stopping the en on the other flank) S3 .h1 59 l:rxa7 Wb3 60 l:tb7+ <itc2 61 l:.l:. via b7 and back up the a-file. so as to stop The big difference here..l:tg6+ 'iii>b 7 58 .l:!. to similar variations above (e....h1 Black wins.. is to keep the white king from White's defensive idea in the game. bs h3+ ss .tih2 as 59 fs �c7 6o tion. He plays a num 'it>d3 "ii'f1+ 70 �d2 (but not 70 'it>e3? ber of further moves out of inertia.l::!.t7 63 The overall conclusion is therefore Iia6 �g6 64 :c6 a4 65 .a3+ 53 �g2 gxhs 54 l:tgS+ ••• 67 'ii'f8+ (67 'ii'c 3+? l:tb3) 67 . tically deep analysis to prove it. Black loses his b-pawn. 52 .l:i.xg6? tinued for some 15 years after it was �d4. and Black cannot strengthen his posi 57 . but the text move is much simpler.l:. 68 'iVcs ! White successfully holds. but �3+!! winning) 70. However. e.. l:Ibs Wd6 61 f6 �e6 62 .d3 67 .l:!.l:!a6 �f7 66 l:.. White draws by a hair's breadth endgame.g.. thanks to the ongoing battle of tions just examined above.l:!. in the posi 54 �d6 55 .i..c4! l:tb3+ 68 :c3 .c6 that the position is a draw after 47 There is no way for Black to . counting only on 52 ..l:. as Fischer realized that his .. Botvinnik showed that after The point.l:.Th e Greatest Ever Chess E n dg a m es in those lines. after which the ending with two rook's The final word on this amazing pawns against one pawn is a known game was spoken by the young Garry theoretical draw.b6+ <J.l:. As we have published analyses of this fascinating seen. instead of Botvinnik records that his oppo Fischer's 67 "it'f8+..U. White can also draw with 67 this point. which transposes into the varia played. Kasparov. although it takes some fantas strengthen his position.'iVe1+ 69 win had slipped away.a6 a3 68 'iii>g1 Yz-Yz ••• s2 hs! One of the most titanic battles in Fischer had missed this simple tac chess history.xh7!!.'ii'xf2+ 71 'it>d3 the draw is now inevitable.xbs h4 56 f4 �c6 ••• tion of the last diagram. adequate though nent's face was as white as a sheet at that is. . 68 . 66 .�a2 with the conclusion that "the white king cannot hide from a deadly check". 1 50 . and a drama that con tical blow. who found that. f8 36 ltJd2 ltJd6 37 hausted all of the possibilities.d6+ 'it>e4!.l:.l:. and it is l:r. At this hxg4 llh8 53 l:txg6 'it>ds ! 54 ..l:..ae1 ... Korchnoi trailed 5.l:.h6 than 25 pages of analysis (a quarter of 28 ttJdb3 . but Korchnoi says he did not wish to allow the black king to become active after 47.e 7 18 The complexity of this ending is ltfb1 ltJf6 19 as a6 20 ltJa4 i.l:. he analyses 48 a6 bxa6 49 . In his book.ae1 i. However. he admits that he had not ex ltf1 i. White has a 43 ltJc3 :xe2 44 l:.d8 46 ltJxbs axbs a tactical breakthrough based on the advances d4-d5 and as-a6. but objectively better!" 151 .l:!.fe1 h6 38 .. play resumed the next day.l:. Korchnoi devoted no fewer .xe2 i. 'ii'xfs "ifd7 16 �xd7 ttJxd7 17 a4 i.if8 29 ltJd2 i.l:. but after 51. Even 31 g4 i.xa6 l:. 47 f5 The alternative was 47 l:ta2.ce8 25 .f8 21 ttJcs such that. White has maintained pressure cludes that "it is not clear who stands throughout the game.g4 13 tLld2 .h6 35 f4 i. and the possibility of .h6 30 h3 'M7 the whole book!) to this endgame.d 3 ttJbd7 8 lLlf3 overnight adjournment analysis by :t.d8 33 ttJdb3 ttJbs 34 then. For example. so great is 4.xcs 45 bxcs space advantage. in his book on rook end 1le7 22 Wf1 ltJe8 23 'it>e2 ltJd6 24 Wd3 games. the raid on This ending occurred in a key game Black's kingside by SO l:ta7 hS 51 �g7 of the 1978 world championship match picks up a pawn. which somewhat eases Black's defensive task.e1 g6 26 :e2 f6 27 l:.. with the first to six wins taking the Black's activity here that Korchnoi con match.f1 l:lb8 39 lla1 . 5 5 We2 b4 56 .hxg4 52 between Karpov and Korchnoi..e7 6 e3 o-o 1 i..'it>e6. After s �gs i.xf6 . this is just about his only plan.:tc8 14 �fs �xfs 15 onds.l:!.l:.be8 40 an ending to which he returns periodi .b8 41 e4 dxe4+ 42 ttJdxe4 ttJbs cally even to this day.xf6 both players and their teams of sec 12 b4 i. setting up the break aS-a6.l:i.f8 32 f3 .c8! and concludes that White is not winning.l:!.e8 9 'i!Vc2 c6 10 0-0 ltJf8 11 �xf6 i. R o o k En dings 1 c4 e6 2 ltJc3 dS 3 d4 ltJf6 4 cxds exds this ending should be a draw.h3+ stage in the match. . but always seems to have just enough resources to stay alive. 48..g8 Korchnoi notes that.. he deliber possible outpost for his rook on e6. and has a least as well. He spends some 14 pages of his book.Th e Greatest Ever Ch ess E n dg a m es 49 �C3 This is a one of the most important moments of the game. in order to get Kar pawns to target on fS and h 3.e8! Now the king and pawn ending is a draw. White had two stronger pos sibilities: 49 l:!. whilst after a rook move.l:i. was actu Black's compact unit of king. Consequently..a2 and 49 I:te6. himself.. blance to Capablanca-Tartakower.. Black has weak second-rate move. at the board. and can pov away from his adjournment analy immediately activate his rook on the sis and force him to solve problems by freshly opened g-file.l:r..:. rook and ally a deliberate psychological trick. and he felt sure that Karpov's powerful 41 . One typical variation runs as follows: 49 lla2 l:tg3+ so �e4 llxh3 51 a6 ..es ! and "it is White who has t o think about how to make a draw!" This position bears some resem Korchnoi admits that the text move.h4+ 52 �d3 l:th3+ 53 'it>d2 . Korchnoi relates that he first thought he was winning with 49 dS (with the idea 49 . he and his force. even However.. which is potentially stronger team had failed to find a win for White. than White's extra pawns. Black is constantly balancing on a precipice. passed pawn representing a fearsome During overnight analysis.l:i.h2+ 54 �c3 and now 152 .gxfs 48 gxfs team of seconds (who included Tal) Now White has confined the black would have analysed the position at king to a considerable extent.d8! so d6 .l:!.cxds so l::te 6) until he saw the defence 49 .. .:l. such as Sl llg2 there is S l. with which looks a little strange. showing some of the amazing varia tions that could follow after these two moves.. objectively Which he does! speaking.. in return. ately played a somewhat vague. but the strongest of the al 54.l:!.d7 65 �f1 55 .e6 be l:k7+ <t>e8 60 <Ji>b6 I:!.f1 62 cause his king is cut off) 53 . if anything.xfs 61 c6 .l:id3 57 l:txc6 l:txd4+ draw..l:. is ac tually the weakest..l:i. the <t>b8 <Ji>e6 66 c81i'+ ..e s l:te1+ 63 �d6 b1'i¥ 64 Korchnoi analyses three of them. 49 .f3+ 54 'iitd2 b4! 5 5 l:. and thanks to his h-pawn.. 153 .xc6 as and Black's own truth of Dvoretsky's observation that passed pawns also become very strong.t>d3 (Korchnoi gives 53 l:f.. l:t..a6 as for one another.. weak pawns tend to get exchanged off e. although perhaps the most obvious to the club player. and <Ji>xf4 b2 62 <:J.a4 54 'iiie4 l:..l:tb1 61 Black had other options here. R o o k En dings 49.g1 is also analysed to a loss by Korchnoi.Uds.l:.e8 ••• 58 l:tas b4 59 .l::txb1 65 c6 etc... l:!.h4+ 54 <:J.l:i.l:f.gS ! and suffices for a <t>b4 bxa6 S6 l:t.xh3 51 a6! bxa6 52 ds cxds 5 3 c6 when his c-pawn is enormously strong.l:!.b7 .l:i. 59 53 . White is able to ef fect his breakthrough ideas after so 'iiib4 .J!th3+ (54 ..c s .xa2?? 5 5 axb7) 5 5 ternatives is 49 ..xas <:i.. but notes that pawn and even be a pawn up. One line then is so . White himself has to seek a draw after 49 .xa4 b3 60 l:tb4 .b1+ 63 <:i.l:lc4+ ss �xds! ltc1 56 'iitd 6 etc. l:f.c1 64 c7 <:i.. Black 56 <Jo>es f4 57 J::ta 7+ �g6 and now draws this ending. l:txb1 . This variation shows the a6! bxa6 52 l:f.g.. Jig3+.. Black will win the f "objectively stronger". but the White has no winning chances after white c-pawn is immensely strong.e 6 l:txfs 51 58 <t>xbs .l:. Korchnoi analyses this to a win after 5 3 .a6 fS+ and.xc8+ 67 <Ji>xc8 fS black pawns are stronger..l::!. .a7 .U. 53 .xa6 . b4 55 d5 cxd5 56 . 50. losing a vital tempo. in order to 54 'iii>b6 hold) 63 ..d6 (not 59 l:.f.g.. Karpov makes a significant White's plan is to penetrate with his error..l:. 56 . e. 50 lld2 After so .. Wc8 57 lld3 .a5? 51 'itt b4 Finally.l:. this proves fatal.e8.:te3+ 51 'iii>b4 l::txh3 5 2 a 6 bxa6 53 .. 58 c6 .xa6 l::td 3 and Black suc ceeds in drawing..b3 60 l::tg 7 'it>d8 61 c7+ :xc7! 62 :xc7 b2) 59 . but also aids his own b-pawn) 59 . position. In such a king on the queenside..b3 60 l:. Korchnoi points out that after 54 58 l:...Uxd5+ Pushing the c-pawn at once fails af ter 56 c6+ 'it>d6 57 !lc2 b3 58 c7 l:. 154 .. ..e4! Anticipating the threat of the ad vance 51 ds. Once again. as Korchnoi shows.l:.e6+ 'it>d8 (Black can even play for a win with 62 .l:tc4! after which White cannot win. He should 51.l:I...Black's rook does not just defend.d6+ with a draw.d7? when White even loses after 59 .Th e Grea test Ever Chess En dg a m es mate. 57 .l:ta2 Black can safely take the h-pawn: so...d6+ We7 62 .tc3 ! (this is the point . when Korchnoi demonstrates that White has to play a whole series of only moves...xf6 Wd8 61 ..Wf7.l:i.l:!.. 54.. and he is now lost..'itte8 52 a6 bxa6 53 'it>a5 'it>d7 have played 57 .g3 b3 59 'it>c6! Wxa6 Wc7 White can no longer achieve This is the move Karpov had over his central breakthrough: 5 5 d5?? l:ta4 looked. the black passed pawns throw a large spanner in the works. l:i.xa3 .xfs 68 �g3 .xf6 fatal error.a8+ <.c4 64 c7 �c1 nobody can hope to analyse them with (setting up checks when the white rook complete certainty.l:!.. Rook En dings 59.l:i. and analyses 62 .a4 .xa3 and White wins..l:tf4 65 ..l:i.l:r. 155 .d1+ 68 <it>e6 lle1+ 69 �xf6 .a6+ <.xb3 J::!... but is also l::tg7+ �c8 71 l::!.l:i. h 7 1-0 losing after 60 . maximum practical difficulties for the 60 J::!.. especially in a leaves the d-file) 6 5 h4 h 5 66 �d4 a3 67 game situation.l:!. Handing over the b-pawn with 64.:.l:i. a4 63 c6 .x b3+ �a7 61 �b7+ �a6 62 .l:. which eventually induced a �a7 63 �bs a4 64 l::!.b4 61 .xh6 a3 66 .f6 69 . . Korchnoi's success l:r.c1 70 came from the ability to pose the l:r.b6+ opponent.U. Korchnoi ity of such positions is enormous.t>d7 71 ..�b8 The rest is clear.l:i.<it>d8 is more tenacious.d3+ �e8 A marvellous example of a purely 62 �d6 and the c-pawn is much practical rook endgame. .t>b8 67 check is equivalent to resignation.g8+ �c7 10 59 .U. The complex stronger than Black's a-pawn.. we see a case more moves checking. Tony Miles gives 156 . the passed pawn ad much the more active queen. but rather than the usual out The two big factors in queen end side pawn majority. pawn. which poses unusual dif Unlike a rook. piece" endings are highly tactical in if he is to defend successfully against nature. as the stronger side endings. thanks to the enormous attack an enemy passed pawn. must rely on checks. The vast scope for times in the examples that follow. in which many of the key ideas vance a passed pawn one square.Queen endings are the ones that enormous influence. a queen can force home ficulties. without vances another square. then and themes of queen endings are the defender spends the next ten or shown. In Game 34. Finally. board. This means that the de three examples of endings with queen fender. The first two show across to help. Such "heavy with the ultimate threat of perpetual. as is possession of the on a central square. 35 shows Black exploiting an extra etc. the queen exerts initiative. checks means that queen endings are Here. his extra pawn is ings are passed pawns and king safety. and we will see this theme many in my experience. In Game 33. queen and rook endings. In Game 36. against the We then round off the chapter with enemy queen. on all sides of the frighten the average player the most. Game other long sequence of checks starts. a passed pawn on its own. A typical scenario sees him ad queen. the where Black wins by virtue of having checks run out. in the centre. we examine three pure queen often very long. King safety is paramount in the activity of the queen. if he cannot get his own king plus other pieces. and then an any material advantage. Eventually. When placed such endings. we see a theoreti makes very slow and painstaking pro cal example of queen and pawn vs gress. ing power of the queen and rook com The other issue in queen endings is bination. we see a rare Game 37 is justifiably the most famous example of the ending queen plus heavy piece ending in the whole of knight against queen plus knight. Here. in Game 38. Q u e e n E n dings an impressive exploitation of the supe championship history. and the most dramatic games in world Keres fashions a precisely-played win. chess literature. rior activity and safer king. 157 . as well being one of too. the initiative is a big factor. whilst Finally. l:!..hg8 34 . 55 g8'ii' al'if 56 'ikg4+ was 57 'ifxe6 2.l:!.h8 33 rules of thumb: . fxe6 fxe6 28 �f7+ 'iiic6 29 i.gf8 37 .hg8 36 1.e .l:txb2 44 gS a4 45 g6 queen.l:!. .Th e Greatest E ver Chess Endg a m e s 1 lt:lf3 lt:lf6 2 c4 c6 3 d 4 dS 4 lt:lc3 e 6 5 e 3 demonstrated the winning method for lt:lbd7 6 Ji.xg3 53 hxg3 a3 54 g7 a2 principles of such endings.l:!. The stronger side should usually 158 .l:!.xf7 38 J::!.f3 'iii b4 so l:.b7 10 lt:lxf6+ gxf6 11 e4 .l:!.l:.e3 �ds 30 Broadly.tc7 .td6 plexities of queen and pawn vs queen 12 o-o .d3 dxc4 7 Ji.l:. it is not always a simple matter .. i.b1 46 �h4 .g7 J:.xh7 lt:lc6 31 . and even though the 15 lt:\c4 'ii'c7 16 f4 cs 17 a3 fS 18 axb4 whole endgame has now been "table lt:lf6 19 lt:las c4 20 lt:lxb7 cxd3 21 lt:lxd6+ based".d3 b4 such an ending in this game.xf7 .g7 .l:. The tablebase confirms that the present position should be a draw with best play.afl . 3.. inaccuracies in the play.c2+ 43 Wg3 . where its influence is at its this type of endgame at the time this greatest.txd4 �xd4 41 'iiif2 as 42 answering a check by interposing his g4 . fewer chances with a knight's pawn. This 'it>bs 49 . The com 9 lt:\e4 i.l:. and the chance of cross game was played.g1 47 �hs 'iiic4 48 l:tf4+ so force an exchange of queens.l:tf7 . The stronger side's queen usually takes up a post in the centre of the Relatively little was known about board.l:. and fewest of all with a rook's pawn.f4+ 'it>bs s1 l:lf3 idea lies behind several of the other 'iiib4 52 :g3 l:.tf4 . so as himself to give check.dd8 32 .l:.l:tg8 13 'it'e2 'i!i'as 14 lt:\d2 o-o-o are enormous..tes lt:lxd4 40 . One of White's main weapons in . we can point out the following I:!..c8 39 such endings is the "cross-check".xc4 b S 8 Ji.l:. and I:!. Botvinnik first 4. Whilst there are no hard and fast rules as to which positions are won and which are not. and despite a few checks is maximized.l:th7 :h8 35 . it is true to say that the stronger side generally has most win ning chances with a centre pawn or bishop's pawn.xd6 22 'i!i'xd3 fxe4 23 'it'e2 'ii'b 7 24 fs to divine the underlying point behind lt:lds 25 'ifxe4 lt:lxb4 26 'ifxb7+ 'iitxb7 27 many of the computer's moves. the a1-corner in this position).�a4. This does not mean that An inaccuracy in turn. With a knight's pawn as here. especially principles (3) and (4). after which the position is a win for White. too. pawn at least as far as the sixth rank in 61 'ii'f5+?! such positions. words. albeit in 65 moves. This maximizes the chance of for White starts once his pawn gets to cross-checks. 57 1i'h8+ 58 �g6 'i¥c3 59 g4 ••.'ii'd 3+?? with the cross-checks. 60 g5 we will see how Botvinnik handled the ending.. In my opinion. the tablebase confirms that Black has defended well enough to keep the position within drawing bounds. Armed with these basic principles. Thanks to the position of his king and This. Black has five possible drawing moves here. heading It should be noted that White can the black king towards the safety zone not be prevented from advancing his in the south-west corner. but his last move is an inaccu racy. the sixth rank. of course. The ruthless 159 . is 60. is motivated by the de queen.. It should be noted that he had the benefit of adjournment analysis. cross-check 60 'ii'f5+. 60 'ii'd 4?! ••• Thus far. it can be assumed that his team mates will have contributed. and since this was an Olym piad. Q u e e n En dings try to place his own king on the same the position is lost or that the defender or adjacent rank or file as the enemy has erred. but the most logical of these. none of this diminishes the value of Botvinnik's achievement in this game... The real technical problems king. the 59 1\Vd2 ••• defender's king should generally head This is a typical example of the ap for the corner diagonally opposite to plication of the principles enunciated the pawn's queening square (in other above. following principle (5) above. White is in a position to meet fender's desire to reduce the chances of checks such as 59 . 5. because 64...Wa5 65 'ili'e5+ king should seek a position on the There is nothing wrong with the same or the adjacent file as its opposite immediate 65 g6 (indeed. two involve following In itself..:ili'cl+ 72 �5 1Wc8+ 73 �g5 ii'd8+ move. All any human more centralized square. but it is the start of a short king towards the safe corner: 63 . and 67 . the tactical features of these positions are so great that they are always going to be beyond a hu man player's capacity.. after which �f5! when there are no safe checks. the basic principles and apply them as 65.. 61 ..'ii'd 2+? 68 'ii'f4+.Th e Greatest Ever Chess E n dg a m es tablebase shows that White has only Black's last. However. which goes runs into the cross-check 65 'ili'f3+.. Now his king is forced away one move to win here... way..'it>a4 62 Wh5 'ili'h8+ 63 Wg4 'ili'h1?! 67 . White finds the most accurate n.�a4 66 g6 11i'd1+ 67 Wg5 far as possible.Jii'd 8+ Once again. the proximity of the opposing kings means that cross checks are in the air. underlining the drawback of 74 1i'f6?! 1 60 .'1t>b3. it even wins number. but Botvinnik's player can be reproached for such mi move is more logical from the human nor imprecisions.he first puts his queen on a plexity of such endings... both after 67 .�a3? rather illogical 61 �h7.. and that is the from the safe corner. against the principle that the white 64. given the huge com viewpoint . He should prefer 71 �g4 or 71 64 'ti'f4+! 'iii>g 5. the assessment swings over to "win 68 Wf5 'i'ic8+ 69 �4 'ii'cl+ 70 'ife3 ning" once again. I do not think either two moves quicker). ..'ii'g 1+ 68 Another imprecision.Wa3 phase during which Botvinnik loses his or 63 . Of Black's three 'i!Vc7+ 71 'ii'e5 drawing moves. this move does not spoil principle (5) above and heading his anything.. in accordance player can hope for is to understand with principle (3)... the black king should emu late the birds and gratefully migrate to Now there are no more profitable the kinder climate of the south. first time round. eight years ear However. In so doing. 'i¥h8+ 77 'it>g4.'it>a3 ! . way. advance his g-pawn again inside the so moves allowance). also be noted that the plan of placing which again endangers the win.<iita s? •• Returning the favour. were now on a2 or al. It should A fundamental error of principle. since after 76 . completely out of ammunition.pinning the pawn. The tablebase confirms that the is attested to by the fact that the win ending is then a draw. Black is not yet 79 :�as 8o 'it>gsl . as he had made exactly that mis White can win with several continua take in the same ending. Black soon runs out 1 61 . the difficulty of such endings lier. had he taken his chances earlier duces the enemy queen's scope for in the endgame. 74 �5.. This the white king in front of his pawn move simply drives the black king with 74 'it>h6? 'ii'h4+ 75 <itg7 is much closer towards the safe comer. Botvinnik found the 77 'ii'f4+? most accurate move. 77 . this does not mean it is a draw . It is clear that Minev did not understand the point about where his king was best placed. Q u e e n En dings We have actually reached the same defensive technique in such queen position as after Black's 67th move. and number. and so re been. This was a lesson that Botvin tablebase confirms that the position nik had been forced to learn the hard should be drawn after this. even though checks. as it could have he maximizes its influence. of course. One can see how much easier takes the chance to centralize his his defensive task would be if his king queen with gain of tempo. Obviously. he is out of checks. Minev resorts to another typical piece positioning. The endings . In Now we see the effect of Botvinnik's stead.. However. Soviet master Ravin sky. but one which makes Botvinnik's task easier.. and the weaker. against the tions. the main one being 77 'it>g4. a forgivable error for his time.. Black once again suffers from his king 78 'ii'd 2+ 'it>a4 7 9 'ii'd4+ being on the same rank as its opposite Botvinnik follows principle (3).White will. Instead. with checks.. n . still takes 60 moves from this position! 74 Ji'ds+ 75 'ili'fs 'ili'd8+ 76 'ii?h s 'ii'e8 •• (However. ld 6 'ili'h2+ 91 �cs 1-0 checks. apart from his check... 1 62 . own game against Ravinsky years be 8s .. and all the possible checks are defeated by a cross 8S �d4 check. the possibility of a of the checks. the two kings stand close together. and this time he cannot pre vent the pawn advancing further. but the human instinct is to improve the queen position first. square of the pawn. to seek the elusive cross theoretical practice. cross-check means that Black is out of 89 'iid s 'ii'e 2+ 90 'ii.'ii'h 6 + 89 �d s is again the end Once again.. and im square from queening. 83 . Mark Dvoretsky gives the ending in his adjournment analysis.. note "87 �dS?! 'ii'c 8! . Without the aid of approaches closer towards its opposite either tablebases or any substantial number. endings.. Another characteristic idea for such Despite a number of inaccuracies. the tablebase shows that the immediate 83 g7 wins slightly more quickly. also controlling the queening square of the pawn. 87. g7.'ii'd1+ 86 �cs 'ii'c1+ 87 �d6 fore. but the table " identifying all the key elements of the base shows that this wins for White in winning method in such positions.... as it perfectly illustrates the elements of the winning method.Th e Greatest Ever Chess E n dg a m es of checks. 1i'd2+ 88 �e6 "i¥a2+ 8o.Ji'e7+ 81 �5 'ii'f8+ 82 �e4 88 . With the pawn only one this is a classic endgame.. whilst the superbly placed exactly the same number of moves as white queen also controls the next the text. �a4 84 g7 'ii'h 1+ The final position deserves a dia gram. 82 . the white king mensely instructive. 'it'h6 83 'ii'e s+ Once again.. The white que�n occupies a dominating central square. Botvinnik effectively "solved" this Curiously. by loosen ing up the enemy king position. 1 63 . and continue 3 5 h4 in an attempt to create some counterplay.. too passively placed to exploit this.xd8 2 5 'it>xf1 h6 26 'ifc2 involves a pawn sacrifice after . but it proves enough to win. 28 'fid8+ 'it>h7 29 'ifd3+ g6 30 'ili'c3 'ili'd5 31 a3 'ifd1+ 32 'ili'e1 'ili'd3+ 33 Wg1 and although the black king has 33 'fie2? loses a pawn at once after been exposed a little. .l:txd1+ 27 1i'xd1 "jWc5 34 . to be a very small posi tional advantage.l:. White remains 33 .e7 5 33 .g5 i.1Lb3 i..e4 . and here he gives a perfect demonstra tion of how to exploit what appears. Q u e e n En dings 1 d4 d5 2 c4 e6 3 ltJc3 liJf6 4 i.l:!. . this seems inadequate after 3 5 .l:. The alternative 34 b4 ltJxf1 24 . 'ifxa3 (Panchenko only analyses 3 5 .d3 dxc4 7 i. .h 5 36 "jWe3) 36 h 5 "ii'd 3! 37 hxg6+ 'it>xg6 (Dvoretsky). on the face of it.c2 ...J!i'c2 e3 liJbd7 6 i. because the weakness of White's back rank prevents the defence 35 'ife3. His only real plus here is the weakness on e5.'ih'b2..xe7 ltJd2 23 'ili'e2 passive move..g5 liJf8 17 i.ad1 'fic7 13 i. Panchenko suggests that White should sacrifice the pawn any way.xc4 a6 8 liJf3 Taking up a dominant position with o-o 9 o-o c5 10 'ii'e 2 b5 11 . and the greater queen mobility that results therefrom. Maroczy was famous as an out standing specialist in queen endings.f4 1i'b6 15 34 'i!Va1 ltJe5 l::tac8 16 i.'i'b1+.1Lxe4 18 It must have been hard for White to ltJxe4 cxd4 19 exd4 ltJg6 20 1i'f3 ltJxe5 swallow the need for such a grievously 21 dxe5 ltJxe4 22 i. How ever.fd8 14 i.b7 12 the queen.1Lxd8 .. bxa3 43 'ii'f6+ 'iit g 8 44 'ii'd8+ Wh7 45 'ii'e 7 'ilfhl+ 46 Wh2 'ilf5 when White can still resist after 47 'i!i'xa3) 43 'ii'f6+ Wg8 44 'i!Vd8+ 'lt>h7 45 'ile7 'ilb1+ 46 'lt>h2 'ii'f5 47 'ii'b7 'it. as his queen cannot move without dropping the b2- pawn..'ii'f2 zugzwang. queen is totally immobilized... 35 g3 a4 36 f4 Both defending the e5-pawn. b3! (more accu rate than the analogous line 42 .. seeking to open up the black king for possible counterplay. ing the white queenside pawns by 40 'ti'g1 'ii'x b2 41 "ii'c 5 means of ... undermining the white king Such tactics occur frequently in side structure.h5-h4.xf7 45 'ii'c 7+ 'iite 6 ••.g. The black passed pawn 37 . the key c5-d4-e3 ! idea for Black is to establish a 39. e. Now it is a "real zugzwang". exf5 43 e6 bxa3! 44 exf7 White has run out of pawn moves.. and also to have the chance of a subse quent f4-f5.. queen endings..b4! . 36. 44 e7 a2 and Black will promote 39 'lt>h1 with mate..... Or 39 'ii'a 2 'ili'dl+ and the white 44 'it. f8-e7-d7-c6- As so often in endgames. h5 38 h4 'it>g7 will be decisive.g7 wins comfortably. but ing his king all the way into the heart Black escapes the checks. 37 h3 37 Whl 'ii'f2 sets up the threat of 41. There White has managed to flush the en would be nothing to stop Black march emy king out of his hiding place. This will be achieved by fix Now White must lose material. The other attempt to do so is 42 'ii'e 7 when Dvoretsky's 42 ..'it>g8! White is in zugzwang. as 42 .Th e Gre a test E ver Chess E n dg a m es 34 .... a5-a4... 42 f5 A desperate attempt to create coun terplay and chances of perpetual check. by following 1 64 ..a5! of the white position. Q u e e n Endings the same technique we saw in Botvin �2 also sees the checks run out. nik-Minev - he runs his king into the 47 ... a2 48 liVeS+ 'iitd 5 49 'iVd7+ 'iiie4 50 region of its opposite number, where 1\Vc6+ 'iiie 3 51 'ii'c 5+ 'ii'd4 52 'ii'a 3+ 'ii'd 3 upon the threat of a cross-check even 0-1 tually causes White to run out of checks. White loses the king and pawn end 46 li'c6+ 'iite5 47 1i'xa4 ing after 53 iVxa2 'iff1+ 54 'iith 2 iVf2+ 47 'iii'c 5+ 'iiie4 48 'iii'c4+ 'iiif3 49 'ii'd 3+ 5 5 'it'xf2+ 'iitxf2 56 'iiih 3 'iiig 1. 1 65 Th e Greatest Ever Ch ess E n dg a m es 1 e4 e6 2 d4 dS 3 tLlc3 tt:Jf6 4 ii.gs dxe4 3 s..:�ds+ 36 f3 es 5 tt:Jxe4 tZ:lbd7 6 tLlf3 i..e 7 7 tt:Jxf6+ tt:Jxf6 There is no particular value in grab 8 ii.d3 o-o 9 iie2 cs 10 dxcs �aS+ 11 c3 bing the a-pawn. After 36 ... 'it'xa2 37 'ii'xcs 12 o-o lidS 13 l:i.ad1 ii.d7 14 tt:Jes 'ii'c8+ Wh7 38 'ifxb7, Black would sim .ie8 15 l:i.fe1 .l:.ds 16 c4 lidd8 17 'ti'f3 ply have exchanged a pair of pawns, .l:.ac8 18 'ii'h 3 g6 19 :e3 tLld7 20 ii.xe7 without making any progress in solv 'ii'xe7 21 tt:Jg4 hs 22 tt:Jes tt:Jxes 23 .l::!.xes ing his main technical problems. Burn l:!.xc4 24 l::tde1 .l:.cd4 25 ii.c2 iVf6 26 li'g3 instead leaves his queen on its power l:!.g4 27 iVc3 ii.c6 28 g3 l:.gd4 29 .U.se3 ful central post, and starts advancing 'ii'g s 30 ii.e4 ii.xe4 31 l:txe4 l:.xe4 3 2 his pawn majority. .l:Ixe4 l:i.d1+ 33 l:.e1 'it'd2 3 4 .l:!.xd1 37 h4 b6 38 a3 as 39 a4 "ii'c s 40 'ii'd 3 'it'xd1+ 35 Wg2 Wg7 41 b3 �d4 42 'iic 2 'it'e3 In this example, we see a case Black places his queen as actively as where the stronger side has an extra possible, attacking the pawn on f3, pawn, but not an outside majority. This thereby tying White down to its de poses special problems of conversion, fence. since advancing the majority to create 43 li'b2 'it>h6 44 'it'c2 Wh7 45 'ii'b 2 'litgB a passed pawn inevitably means expos 46 'ti'c2 'it>g7 47 'ii'b 2 Wh7 48 'i!Vc2 Wh6 ing one's own king, and thus increas 49 'ifb2 'it>h7 50 iVc2 Wg7 51 'ifb2 ing the risk of perpetual check. Burn By my count, this is the third time carries out the winning plan to perfec the position has been reached (see af tion. ter White's 43rd and 47th moves)! I 166 Q u e e n En dings have not been able to find reference to 63 'ii'b 5 e4 64 fxe4 fxe4 65 1i'g5+ �f7 this in any published source, including 66 'ii'f4+ �e6 the tournament book, but I believe the As usual in queen endings, centrali explanation may lie in the wording of zation of Black's forces is the key to the rules at that time. The tournament protecting his king from checks, when book states that the tournament was it steps into the open. played under the rules stipulated in 67 'it'g5 'i!Vb2+1 Bilguer's Handbuch. The latter did con tain a version of the threefold repeti tion rule, but unlike the modern ver sion, there was no reference to a posi tion occurring three times. Instead, the rule referred only to threefold repeti tion of the same moves or sequence of moves, so I assume that it was not open to Pillsbury to claim a draw here. 51 .'i!i'd4 52 'ii'c 2 'ii'c 5 53 'ii'd 3 'ifc1 54 •• 'ii'b 5 'ifc2+ 55 'iW1 'ii'c5 56 "ii'd 3 'ii'd4 57 'i!Vc2 "ii'd 5 58 �g2 'itf6 A useful intermediate check, driving After a long bout of uncertain ma the white king to an inferior square. noeuvring, during which Burn was try 68 �h3 ing to decide on a plan, he makes the 68 'iii>g 1 leaves the g3-pawn unde first step towards further progress. fended. Pillsbury prefers to hide his 59 'ii'c 3 'ii'c5 60 "ii'd 3 �e7 61 'it'd2 'ii'd4 king on h3, but this is soon exploited. 62 'it'e2 f5 68 'i!Vf6 69 'ife3 �d5 10 'i!Vd2+ 'i!Vd4 71 .•• 'ii'g 5+ With his king now centralized, Black proceeds to establish his passed pawn. 71 'ii'e sl 72 'ti'e3 ••• 167 Th e G reatest E ver Chess E n dg a m es Now we see the value of the white queen ending, Black will win the a king having been lured to h 3. After 72 pawn, with a winning position. 1\Vxg6? "ili'e6+! forces a winning king 77 axb4 78 'ili'gS+ 'iii>d 4 79 'ili'f6+ 'it>c4 •.• and pawn ending. 8o 'ili'xb6 72 ...'it>c6 73 <iitg 2 �b2+ 74 'it>h3 �c2 White has his passed pawn, but Now 7S .. .'ii'd3 is the threat. White Black has two, both of which outpace cannot defend passively, and must seek the opponent's. counterplay. 80...'i&'f3 81 'ili'e6+ 'it>c3 82 'ii'e S+ 'it>b3 83 75 'ii'gs "ili'dS+ 'it>a31 75 ...'ife21 Black neatly uses the white passed The point of Black's play. As Fine pawn as an "umbrella", to shelter his points out, the g6-pawn is not very im king from the checks. portant, because even after taking it, 84 'iii>h 2 b3 85 1\Vcs+ �a2 86 as b2 87 White does not have a passed pawn. It 'i&'ds+ 'i!lb3 88 'i&'d2 is much more valuable for Black to get As in Botvinnik-Minev, pinning the his e-pawn going. passed pawn is a typical defensive de 76 'i¥xg6+ 'it>cs 77 b4+ vice in queen endings, but it does not An ingenious, if desperate, attempt enable White to resist for long here. to create a passed pawn. Fine gives in 88 ...'i&'d3 89 �f2 e3 90 'i¥f7+ 'ifb3 0-1 stead 77 'ii'fS+ (77 g4 'ii'f3+) 'it>b4 78 g4 Splendid technique by Burn. He 'i¥xg4+ 79 'i!lxg4 hxg4+ 80 'it>xg4 'it>c3 ! took some time making up his mind 8 1 h S e3 82 h6 e2 83 h7 el'ii' 84 h8'ii'+ exactly how to approach the task, but 'iii>x b3 8 5 'i¥g8+ 'it>a3 ! (8S ... 'it>xa4? 86 once he formed his plan, he executed it 'i¥a2+) and in this "second generation" with admirable clarity and accuracy. 1 68 xf3 21 vantage is that the white king is l:txf3 tt:\e4 22 tt:\xe4 'i¥xe4 23 ... when ings are amongst the most concrete of Miles (annotating the game in Chess.xd4! 27 'ii'xd4 . o �d6 13 tt:ld4 tt:lxd4 14 exd4 h6 15 As a result of its weakness. it is a clear �xc4 b5 10 �e2 �b7 11 tt:lb3 tt:\c6 12 o weakness here. Fritz 12 assesses ings. Black has a clear the position as equal. Amazingly. king safety is es pecially important with such powerful forces on the board.. In this example.. 24 �xd6 l:!. although the 5 tt:lc3 d5 6 e3 e5 7 tt:lf3 e4 8 tt:\d2 dxc4 9 white d-pawn is passed. Black's �e3 o-o 16 llc1 'ii'd 7 17 a3 .ed8 26 l:. to a ing that " . means that . Heavy piece end 25 l:!.i.l:!..�ds. this may well have 169 .f4? l:. but I know whose advantage.xd6 and. all endings..e.U. it is very uncom King safety is paramount in such end fortable". 2 5 l:. As we have seen in some stance! Nonetheless. Q u e e n En dings 1 tt:lf3 tt:lf6 2 c4 c5 3 d4 cxd4 4 tt:lxd4 a6 other cases in this book. conclud nate over general planning.l:tad8 18 pieces are more active.f4 . may not yet be lost. one with both queens and rooks on the board. although White's position greater extent than in other endings. A further ad 'ii'd 2 'i¥e6 19 f4 exf3 20 �xf3 .l:.l:tfe8 draughtier than its opposite number.i. i..d3 (26 and rook. thanks to his superior pawn judgement I would trust in this in structure. when combined.dl is the alternative.xd4 28 llxd4 concrete variations tend to predomi 'ii'e 1+ 29 �fl 'i¥e3+) 26 . The huge power of queen March 1976) gives 2S .f4 This is the first example we have seen in this book of a "heavy piece" ending. as noted above. rather than a strength. b3 should still offer good winning chances. Alternatively.. 28 .Th e Greatest Ever Chess E n dg a m es been White's best course. although 28. unlikely defensive try"...c8+! 'ii'xc8 29 �xe2.l:. as Black estab 31 'ii'c3 lishes two connected passed pawns on Miles mentioned 31 "ii'f4 as "an the queenside... Now 29 . because his points out the strong move 29 ... 'ii'e 2..l:... 21 Jle4l 28 ds . l:te2 by 28 that queen and two extra pawns only l:!.. relying on the d-pawn 26 . unable to improve on this analysis by and recommended 27 .. J:tg6 32 1Wf1.. the text seems more \i'g4+ 35 l:..e2 26 l:tf2 26 'it'xe2 l:txe2 is a nasty double rook ending for White. 1Wxd5. the silicon beast is Miles gave this as a serious mistake.l:txb2 29 :fxf7 l:tg6 30 g3 l:!. After 30 l:re4 'it'c1+ 31 �2 �g7 advantage. 36 years later. al 25 Ji'e2 •• though he may well be holding with best play. Fine writes sis that White can meet 27 ....l::tcc7 l:td1+ 30 l:tf1 l:ldd2 when Black is winning. "with a probable 27 l:td1?1 win".:tes 29 l:!. on the ba Britain's first Grandmaster. settling for the favourable double rook ending.'ii'c 1+ 30 beat two rooks "when the rooks are not l:tf1 is nothing..xf1 1We4+ fol logical.'ii'xd2 is another favourable for counter-chances.l::texds ! 30 lWxds 'ii'x d1+! forcing a winning king and pawn ending.l:tc7 is a better try. After 27 l:tc8+ �h7 Miles continued his analysis with 28 ..:txf7 l:txd4 29 ..• With the threat of 29. and which forces the rook onto a bad Black must find a plan to increase his square. g5! position is still very solid here.l:Ixf1 36 J:%.l::tf3 l:. Black can still answer 27 l:tf4 with 27 ..f6 33 ..f3 �hsl .l:tf4!."ii'g4 31 . He further ana double rook ending.d3 ..xg2+! 34 Wxg2 king exposed. The intention is 26 .. 30 l:. but the computer united or when there are connected 1 70 .g3 . White's position is uncomfortable.. but with White's lysed 32 . lowed by 37 . .Ua7 is almost the same ending as the the extent that there would always be main variation) 34 l:txd5! l:.ILg6! 35 'ii'x e2 .. .g3 l::txg2+ is met as yet.a sort of J:. ning.'�xg7 34 wrong.i:Ixd5!? which he re 1 was beginning to worry about what " ••• jected because of the spectacular coun the d-pawn might do if things went ter-blow 33 'it'xg7+!! 'it'xg7 (33 .xe2 36 d6 l:. This is a �h7 36 l:.xb2 careful play ... That is instructive from a practical viewpoint: 31.<iitf7. it is worth looking at another His reason for preferring the text is variation mentioned by Miles. but the second can be. and the pected Black is probably still winning calculation skills needed to play them after 33 . after 33 . Miles rejected the line example of the tactical resources avail because of 33 llel! although he sus able in heavy piece endings. but as a further nice win for Black.h5.xa6 b4 39 l:.f6..i:Ixd5 . yet again. . the computer thinks Black is winning ! 1 71 . Here.. and he finds the best chance to make life difficult. players' safety plays.xg2+ had the reassuring J:. because great accuracy...xg7+ �xg7 37 l:ta8 �a2 38 comment worth noting . .b6 bxa3 40 l::ta6 with a chessboard equivalent of the bridge theoretically drawn ending. .e8 should also The best move. Miles analysed with has considerable freedom.i:!g3 ......l:txg3 34 hxg3 l:i. 32 . by The other capture was 32 .l:Ixb2 36 .J:.�e2+! Once again... neither condition 3 2 l:!..xg6+ fxg6 3 5 .l:td7+ �8 quality of exposing the white king to 37 . whilst 33 'i¥d3 l:txg3 34 hxg3 'ii'e 2 31. the only real difference between Miles' analysis and that of Fritz 12 is the former's greater ability to assess the final position.. well. etc. Black which. Q u e e n En dings passed pawns".l:.:!.a2!. 3S .. ... Now 3 3 'ii'd4? loses at White is never threatening to trade once. 33 'iti>xg2 'ifxd1 34 �c8+ �h7 35 �fsl Simon Webb was well known as a resourceful player. �xg3 34 hxg3 'i¥e2 win both rooks for queen and pawn.'ii'g 6 32 .xb2 3 5 l:td8+ a perpetual if necessary".. . with their much side pawns . Two very strong GMs. Black makes a new queen after 46 less instant reaction. whereas the other sug The most notable feature of all is the gested Black's winning chances were depth and accuracy of his calculation. but it is especially true in king to d8 and start pushing his king heavy piece endings..'ii'f6. unfair to name them) had differing An instructive and excellently views. 172 .l:r. about the same as White's chances of We have emphasized the importance drawing. 38 �g2 fxg6 39 'iVe6 'iVb2+ 40 'it>g1 'iVxa3 41 d6 'ili'c1+ 42 'iiif2 'ii'd 2+ 0-1 I have to admit that I remain unsure about the extent of Black's winning chances here. al 37 .. llt'xb2 37 .xg6 ending after 37 :Lxg6 fxg6 38 llt'd3 'ii'd6 After 3 7 'ii'xf7 Miles intended simply offers "good chances of success.. so it would be d8� 'ilfxd8 47 'iVxd8 bl'ii' and wins. dead drawn. White resigned without resuming. the d-pawn. He opined that the queen 36 . prevent ing the white queen coming to d3. ."ilfe7! holding back calculated play. Miles' game up with some more accurately intention was 36 . Again the most accurate. and the threat of a white queen penetration to c6 or e6 ties down Here.'ii'd4+! left in the position". the game was adjourned and Black's king to a considerable extent. Miles wraps the toughest defence.Th e Greatest Ever Chess E n dg a m es 36 'it>g1? it is easy to make real progress. The d-pawn is very strong.. After 36 �h3. One thought the position almost played heavy piece ending by Miles. though there is obviously a lot of play 37. But here he slips and misses the After the text.but I am still not sure that greater tactical content.he can bring his book already. Black has some scope to tor of this aspect of endgame play in this ture his opponent . to support the d-pawn. whose opinions I since after 43 �g3 b4 44 d7 b3 45 1i'e8 asked (both only had time for a more or b2... tg5 o-o 8 dxe5 Material is level and both kings are lbxe5 9 . this is the start tions. d4 . side. it is . Lasker had given a marvellous demon 25 J1d7 26 'ifd1 'ife5 27 'i¥g4 �e8 •• stration of how to handle such posi As will become clear.e6 17 l:td3 Wle7 18 :f. weaknesses.:. 'ii'xf3 'it'xe7 12 lbd5 'i¥d8 13 .txd7 lbfxd7 10 it.. However.l:.xe7 lL'lxf3+ 11 fairly safe. and tries to find a way to work up some winning chances.g3 l:i.g6 19 l:i. it is White's more extended such endings. the Even Lasker's blunder is instructive. the position is just equal. and I therefore believe that the of a plan to create play on the queen game deserves its place in this volume..fe1 lbb6 15 'ii'c3 lbxd5 16 .xd5 and cannot really be considered to give . and if any showing the dangers that can lurk in thing. ders the game away! However. aiming to take advantage of 1 73 . as Lasker Schlechter's hitherto impenetrable ar outplays his opponent. if not overwhelmingly.. and he was doubtless This game is perhaps a slightly sur impatient to make a dent in prising choice for this book. and all the first four had been drawn. Lasker plays on.b5 lbf6 4 o-o d6 5 In essence. Even though he has no advantage.ad1 l:te8 but it is in no way vulnerable to attack. in pawn is easily defended. and should be held back.e8 20 h3 'iW8 21 Ibg6 hxg6 White's e4-pawn which is the nearest thing to a weak pawn in this position.te7 7 .ee3 White any advantage. and then blun mour.l:.i. and most instructive pawn on d6 was a significant weakness heavy piece endings in chess history.td7 6 lbc3 .l:. de structure that contains more potential spite a few inaccuracies on both sides.l:t. Indeed. I have 22 'ii'b4 c6 23 Wia3 a6 24 'ifb3 l:i. This was Game 5 of his ten-game match against Schlechter.. having won all pre vious matches comfortably. Q u e e n E n dings 1 e4 e5 2 lL'lf3 lbc6 3 . Lasker was unused to such a close battle in his world cham pionship matches. Before his blunder. Black has a doubled pawn.d8 25 c4 chosen it principally because it remains Schlechter clearly believed the black one of the best. 14 . . Lasker the weakened white pawns on each. Fritz is very optimistic Securing his queen against the about White's position in this entire thrust f2-f4. bxa4 36 'ti'a2 (at queenside.. evidently believing 'i¥c4 with a sharp position.c3 �a6 (39. as the king can become ex posed. .•• The rook now has possibilities of 36 a4 was the consistent move... one in the next note. black a-pawn is a definite trump. brings his king over to the left flank. However.l:1.. Dvoret Seeing the enemy king on the sky or Kasparov) 3 S .fs. and after 38 . it may have been more pru take an active part in the battle.f6 in order to avoid the possibility course.Th e G re a t e s t Ever Ch ess E n dg a m es White's 2 5th move. I like the idea of this flexible move. Black) 40 �f3! threatens the very dan 32 cxbs axbs 33 g3 gs gerous . This is dent to have prepared the move with standard endgame technique. he runs into an unexpected the shakiness of his king position retort."i!ibs? Black seizes the initiative on the 39 'iVd4+) 39 l:. causes him plenty of anxiety. 35 'it'd1 Here. . to However. variation. White could have tried to ex ploit Black's last with 35 a4! (a possibil 28 'ii'e 2 'it>d8 29 ii'd2 �c7 30 a3 �e7 31 ity not analysed at this point by such b41? commentators as Romanovsky...'iVbs 40 very sector of the board where White 'i!Vf7+ �b6 41 ii'd7 is again bad for thought he was pressing. but However. in which the that the king can prove vulnerable.. rec swinging to either rook's file.l:Ia8 3 7 1\Vxf7+ '>t>b6 38 storm on that wing.. of 34. probably not without good 34 'lt>g2 .e8?1 reason. has to be especially careful with such action. To that end. Schlechter starts a pawn tacking f7) 36 .. 39 �c3 is 31 bsl? ••• a threat.�b7 (38 .t:l. but in heavy piece endings. 3S f61 36 'i¥b3 . the 1 74 . to attack ommended by Tarrasch. .. he has a weakness jected this whole ending to a detailed to attack on a3 and the possibility of examination in his classic book Mittel the break ..bxa4 37 'i¥xa4 achievement of his queenside play - 'iitb 7 38 lla3 ? (my punctuation} almost any rook ending is likely to be 38 ... 38 . 38 . Fritz is stubbornly uncon 39 a4?! vinced.. both sides.. I am with the lat verdict of "dynamic equality" is nearer ter.. after 1 75 .. �xe4+ 39 �h2 �d4! 40 'i¥a6+ �c7 bad for White. so he needs to retain the 41 'it'a7+ �xa7 42 l:txa7+ 'it>b6 43 llxg7 queens and seek chances against l:te4 and Black should win the end Black's king.lta6? 40 bS!} 40 :d3. Kasparov's 38 �c2 is more sensible.. game.. c6-cS. It is a difficult position for unnecessarily. J:ta1 41 \t>h2 <3ic7 42 a bit here.'iie6}. it remains very be fatal.. but with best play it does not difficult for Black to strengthen his po look to me as though either side can sition. This is the one big spiel} then analyses 36 . . in search of counterplay.'i¥xb4?! 39 . to prevent White's next.. who analyses 40 axbs This weakening is forced.l:th8 39 . but a dubious sign by (39 . since the 'ii'x bs 41 'ifb3 (Fritz gets very optimistic rook ending after 38 'ii'g 4 'ii'x g4 39 about 41 :tc3 but in fact. This time..'ili'e6 37 'i¥d1 .a8 although he claims that Black retains here.�C4 . but Kas the advantage after 38 .lta8 (or parov sees no reason for this.a8 was given as "no worse" by 38 g4 Romanovsky.l:!. .. Possibly a Kasparov. Romanovsky advocates 38 . and starting thrashing about 'ii'd2 'ii'e s.:i. much preferring White after Awarded an exclamation mark by the further moves 39 I1c3 'it'hs Romanovsky. really claim to be better... where any error is likely to After the solid 39 'i¥d2.. 36.. It appears that Schlechter panicked the truth after 40 .. Q u e e n E n dings Russian master Romanovsky (who sub hxg4 favours Black... lb3 1 76 . 46. l:!.l:txbs 41 'ii'a 1 �b7 45 11Ve1 'ii'd 3?! 42 1lt'd1 'i!Vxb4 Black has won a pawn.l:. 46 l:lb4! Setting up a threat of1lt'a1..b1 48 'ii'a s+ �d7 49 same idea... agrees. Kasparov points out that this is a although White retains chances missed opportunity.b8! 44 'ii'b4 'iWbs is the best move..l:..:tbs 48 . After 40 axbs .bs.J�e5! �c7 44 'ii'xa8 �d7 the position is just a The rook defends very well along draw) 41. 44.. as the queen end ing after 46 .a 1!. 42 .l:!.cs?! 41.r.. However.b4 (Bologan) may was 46 .b4 as Kasparov agree that after 47 .Itbs runs into 45 'iWc4!. 41 l:Ib3 41 'i!Va1 :tb8 42 11Va2 ds 43 exds 'ii'x dS+ 44 'ii'xds cxds is extremely good for Black. l:!. and both Fritz 12 and tition 42 .l:t.. puter is probably correct that 39 .Th e G re a te s t Ever C h es s E n dg a m es 41.la3.. Once again. . Dvoretsky then analyses the repe too cooperative.. A safer option ... However.. . a conclusion with which Fritz Black is much better..b4 'tWas 47 1lt'b1 ..as. Romanovsky instead recommended 42 . Romanovsky dismisses be a stronger way to implement the this with 47 l:ta4 . the com the fourth rank.l:t..r. but suggested that maybe 43 'ii¥b4 cs defends.l::tb 1 46 . After 45 . but Black's 47th is far . White intends 'ii'b 3 and 'iWfS+ drawing.xbs 'i!Vxbs should be winning for Black.. passed over by pre against the exposed black king.b3 'ii'a6 44 .. taking aim at the b4. winning the queen ending after 47 . whilst 43 'i!Vb2? Lasker was unsure whether he was :txe4 is simply bad for White.'fi'a6 42 li'd4 This exposes the black king further With the threat of 'i!Vb2 (or 'i!Vb4) and and is extremely risky. 11t'a3 43 .l:.'�Vxb4! 42 l::txc6+ �xc6 43 'ilVdS+ 43 ..... vious commentators.l:i.e8! 43 ...l:.l::!.1lt'a4! 40 axbs 1lt'xb5 Black is clearly better.'i¥i'd7! best.. White needs to keep the queens on. pawn.l:ta4. e.b7 and now �h2 'ii'e 5+ with perpetual check. but his ingenious play shows the resources available in heavy piece endings.. Ji'c5?! As Kasparov writes..a7+ Wc8 51 actually loses the game. 'it'xe4+ 49 'it>h2 l:tb5 d5 51 exd5 cxd5 52 'iWa5 Now 40. .l:i.'ti'd7 53 'i!Vb4+ '1t>C7 54 'iWd4. who have almost unani 48 'iWa1?! mously labelled it the decisive mistake...�e2+! 53 'it>g3 �eS+ 54 �g2 d4 and I see no reason to argue with him..'iWe5+ 51 �gl 'iWel+ 52 �h2 d5 53 l:!.. Fritz prefers 48 'i!Vb4. 50 't1t'a2! Attacking both the c4. which seems to but in fact this does not appear to be guarantee sufficient counterplay. 54. The point of the check on e1 was to cover the as-square..and g7- pawns... He is winning the e4-pawn. 50.a8! Schlechter's position is very bad.'fie5+ is a threat. 47 l:ta4 C4 This is the point of Lasker's play. the case.Ji'b4 54 �g2 .'iWxe4+ 49 �h2 . .cs 50 l:...l:. Q u e e n En dings 'ii'x b3 48 'ii'x a5+ 'it>b7 49 'it'd8 'ii'e 6 50 f3 48 .l:i.. over the dec ades this move has attracted dozens of question marks from the various com mentators. Kas the text threatens 54 �a7+.. especially after 53 . l:txg7 and Black probably has nothing There are three main alternatives: better than 51 .. 56 'ii'e 3 'i!Vd6 was given as winning by 1 77 .g. parov has fewer doubts.. 52 . so that after 53 .l:!..a7+ l:!..b8 55 'ii'a 7+! . But 52 .'1i'e5+ 52 'it>g1 �a1+ 53 a) 54.b7 White does not have 54 'iWaS+. it is Black's next move which 48 . e 8! l:i. 55 'ii'a 6 Kasparov claims to have done ex actly this. over the board. such a hard fight. But the computer's preference quires extremely complex analysis to 56 '6'a6! appears to give White enough establish this. 1 78 ..l::!..C3 when he analyses prevent 'W/a5+.. 'ife4+ 60 �h2 'Wie7 and sacrificing queen for rook. although it re.. ... although for a defini that Lasker should have lost control of tive answer.. and believes White is draw 55 :bs? .l:txcs l:txc5 58 'ii'h 7 + b) Schlechter recommended 'it>d6 59 'W/xg7 c2 60 'fWxf6+ 'it>c7.g8! (this rook switch is a key resource still draw in two ways.. ing. Kasparov claims that Black can l:i. The overall conclusion seems to be that Black was no longer winning the position at this point. after powerful engine running for hours.and d-pawns. and relying now 61 �g2! (avoiding 61 'iVa5+ �d7 on his passed c.b8 (or 5 5 . one needs to leave a very the complications.'iVb6 57 'iVa3 'ii'b4 58 "fi'a6 "fi'e1 59 fight.Kasparov) 56 'ii'a 7+ l:tb7 57 'i:Va8. His first 62 'W/xd5+? "ifd6+!} and Black cannot line is Lasker's 5 5 . and White can make no progress. c) Dvoretsky's pupil Vadim Zvjagin sev came up with what the former de scribes as "the most solid solution": S4..c3 but after the strong reply 55 .l:ta5 56 'ii'e 2 "and White is OK" . It is scarcely any wonder play for a draw.c8+ �d7 57 .. l:i. White is holding..b7 but then 5 5 "ii'a 6! transposes to the lines just examined. in both cases by for White) 59 .l:!.The Greatest Ever Chess Endg a m es Lasker.. 54 . 56 . one of his variations going This move really does lose without a 56 .. lk5 (58 .a7+ 'it>d8 57 l:. it is stronger and appears to win: not clear that he has anything more a) 56 .l:!.. This does indeed seem to Lasker resigned without waiting for hold...l:lg7+ 'fie7 61 l:!.'iit b8 57 l:txg7 c2 58 llg8+ Wc7 than perpetual check. because the black rook is on b5 rather than c5) 58 . 1 79 ..l:tc8+.. �6!. Wd7 60 'fif5+ or 59 .'i!Vxa6 58 'fixa6+ Wc5. e. . entire Lasker-Schlechter match..'ii'xa7 57 'ii'xa7+ 'it>d6 58 'i!Vxg7 (White has gained a vital tempo over the line with 56 . Despite the mistakes. Q u e e n En dings However. .a7+! looks optimism about White's position.a6 (57 59 �f8+ 'fid8 6o 'fic5+..xg7 mate) 60 . c2? 59 �xf6+ wins after 59 . very similar A monumental battle. this remains one of the most instructive heavy piece endings one will ever see. Fritz 12's 56 l:!. which was one of the finest world champion ship matches ever played... inevitable in such a tense and complicated game. 59 �e6 'it>b7 (59 .. 'ii'f7+ Wb8 63 'ii'g 8+ Wa7 64 'iWh7+ 'it>a6 65 'fixc2. cl'fi 60 . winning.l:lg7+ forces 56 l:!. 56 'fic8+ 'it>d6 57 . he slipped and and despite the silicon monster's over-balanced at the crucial moment. and eventually establish a position that he should have won.xc7+ Wxc7 62 Now Black is being mated. Kasparov's second drawing sugges tion is Capablanca's recommendation 57 Ji'b6 58 'i¥a3 �c8 1-0 .d8+ 'ii'xd8 60 �xd8+ 'it>c6 is another..g. b) 56 .. Lasker's leg endary ability to balance on a precipice enabled him to outplay Schlechter from an equal position. just like the ending) 57 .'it>c7 60 'fic3+) 59 'fixf6+ Wd7 60 �xg5 c2 61 �cl d4 62 �f3 and again White wins. but for once... 'i!ff8+ We5 58 'ii'e 8+ Wd6 59 l:!.. 5 5 .. . l:Ixe8+ prefer to avoid this exchange.. Bl�d/Zagreb/Belgrade 1959 1 tbf3 d5 2 d4 c5 3 c4 e6 4 cxd5 exd5 5 tacking force.The G re a te s t Ever Ch ess E n dg a m es Game38.b3 33 .d3 26 'it>g1 36 ..:tc3 l:txc3 34 'i¥xc3 iie4 36 iixb5 36 'i!Vd3 leads to a lost knight ending after 36 .b6 35 iib3 b5! 12 lbxb6 axb6 13 tbd4 h6 14 i. and Keres uses this factor g3 tbc6 6 .tbc4.e7 8 tbc3 o-o gradually to increase his advantage..tbes is a highly unpleasant threat.Tai-P. White would 21 e3 'ii'd 5 22 exd4 l:txd4 23 .f4 �d7 Black is only too happy to trade the 15 a3 i.ig2 tbf6 7 o-o il.xc5 11 tba4 . tbc4 27 tbe1 . but there is cant positional advantage. from a crucial game by . in the shape nothing wrong with the text..JWxe3+ 37 �f1 'iff3+ 38 �g1 'ii'e 3+ Black has an extra pawn. 9 .l:Ib3 28 l:tc1 tbxe3 29 fxe3 'ii'e 5 30 tbg2 .h3 16 iid3 l:tfe8 17 �fe1 i. but it is dou 39 'it>f1 g5 bled.ie6 10 dxc5 i. he also has a signifi have won more quickly.g5 .'ii'xd3 37 tbxd3 tbas followed In this position. of White's weakness on e3 and his 40 tbe2 tbe5! weakened king position.i.tbd4 would value. Queen it takes a further 30-odd moves to plus knight are a highly effective at.l:i...l:tae8 20 i.. leaving 18 'it>xg2 .. M. and so in itself of diminished Keres suggests that 39 .d2 d4 him a "real" pawn up.Keres Candidates Tournament. the latter be Commencing a decisive attack. but tbxe8 24 'ii'e 2 tbd6 25 i. in the 1959 Candidates tournament.e 3 .xb2 31 iid3 'ii'e6 32 tbf4 .i..l:te4 19 lDf3 .l:!. achieve victory! 1 80 .... 36. al ing an especially important factor in though White's resources are such that endgames featuring queens.xg2 b-pawn for White's e-pawn.l:i. However. . which soon forces mate. 1 81 . He gives a full ac count of the analysis in his classic chapter in The A rt of the Middlegame.tt::l d1+.. and the white queen falls to 44 tt::ld 4 a knight fork. or 46 tt::lf3 �e2+. e.. and Keres was able to analyse the next phase of the game in detail. 49 ri.. in the face ofthe threat of 44 .tt::lf2+! tacking power of queen and knight Here.. if not the advantage.a striking illustration of the tactical tricks which can be concealed in endgames with queens and other pieces in tandem.. to harass the exposed white which Keres had fortunately spotted king.hs+ 49 'it>xgs 'iff6+ so 'iitxh s tt:Jes when there appears to be no de fence. But now White has the shocker 51 'i!Vd6! ! and it is he who is suddenly winning .g7 43 'ii'fs "iid 2 looks as though Black can mate quickly with 48 . together.g4 This was the sealed move. 46 tt::lc 6 'ii'd2+ 47 Wgl (or 47 �h3 tt::lf2+ 48 'iit>g 2 tt::l g4+ 49 �h3 h S) 47.fs 49 'iii>f 3? loses at once after 49 .. so Black must use the at 48.. 46..'ii'c 1+ 48 "iifl �cS+! winning..... the white a-pawn will give him very strong counterplay.'i!Vf2+ 47 �h3 �fl+ 48 ri. It 42 �c8+ ri. The key to Black's win is the attack on the white king. 44. If either of the pieces comes off..'i!Ve1+ 45 'iitg 2 �e3! 46 "fi'ds! There is no good square to which the knight can be removed. during his adjournment analysis. �e1+. there is a very nasty trap. Q u e e n E n d ings 41 �xb7 tt::ld 3 The only way to stay on the board.g. . so Black still needs to retain both pieces and use his initiative to increase his material gains... ceteris paribus. but Keres wrote that he preferred not to place his knight so far offside. We see Keres opponent.Th e G re a t e s t Ever C h es s E n dg a m es 49 .'ife5. and should win. 54 �bs �e4 The computer confirms that both sides have played the best moves since adjournment. since the slight weakening of to confirm that the discovered check is Black's pawn structure. 52 'liic7 �e7+ 53 'liteS lt:Je3 53 ... With only three 55 �b2 'litg6 56 '1i'b6+ pawns vs two. he still needs to play employing the same technique here. Even so. the endgames in this book that centraliza elimination of the a-pawn is a big gain tion of one's forces is a key method of for Black. all on one side..'fWe8+ 54 Wb7 lt:Jxh2 is also pos sible. because it removes a major reducing counterplay and protecting potential source of counterplay for the the stronger side's king. both the knight and the queen endings will usually be drawn. Nonetheless. Black has not been able to wrap the game up with a quick mate We have seen in the pure queen or material win..'liih 5 was although the computer allows us today stronger.f6 chances with 5 5 lt:Jf5+ 'it>g6 56 '1i'd3 and Keres argued that 56 .'ii'd 3+ so 'lites lt:Jg4+ 51 'liid6 �xa3+ not a serious threat after 56 . accurately to win. The respite would allow White to set up some counter S6 . . which the text- 1 82 .. Despite his seemingly decisive attack. it is understandable that Keres preferred not to enter such murky waters over the board.. 'Llg3.ifxg3 is also good. Q u e e n En dings move involves. Black exchanges queens his king. 1 83 . formidable attacking force.. However...'Llg4 74 �ds 'Llf2! 59 CLlc7 iVc2 60 iVd6 Covering h1 and getting ready to Keres points out that 60 �e6 is re advance the pawns.'i�Vd7+!... n. His extreme accuracy shows the 62 �b7 ifxg3 63 �xf6+ �hs resources that exist with queen and Black's position is a technical win. Keres. White has to seek 'Llfs 'Lle4 79 'ite6 'ii'g4 0-1 counterplay... 57 CLle6 'Llc4 58 'ii'a 6 'Lles 68 Wd8 68 'Llxh6? iVf8+ is one important point of Black's play. Black will force off the queens after 6o.. is also a threat. They form a but still requires some care.'�fs. knight working together. hs 69 'ifg6 'Lles 10 'ife6 'Llg4 11 Once again.. futed by 60. . this does not seem to be a significant factor..iVc6+.�xh2 61 'Llds 'iff2 8o... 68 . as is A superbly played endgame by 61.�3+... exchanging queens by 73 . but require 64 �e6 'Llg4 65 'Lle7 'ii'f3+ 66 �c8 �h4 excellent tactical control to utilize ef 67 'Llfs+ Wh3 fectively.. With another pawn 75 �c8 h4 76 iVes 'ire4 77 'ii'f6 'ii'f4 78 inevitably dropping. Black's centralized 'ii'g6 'Lles 72 'ii'e6 'ifd3+ 73 'Lld4+ forces do an excellent job of protecting Otherwise.. 61.. 59 . hence the text. gives White additional counterplay. and also one of considerable gether. Game 42 is an ending with two Game 40 also sees an ending with rooks and bishop on each side. Anders rate play. Together with his sec is when the knight manages to find a ond at the event. but here we have bishops being of opposite colours. SO moves to win. In general. In saw one such example in the early part Game 39. usual drawing tendency of opposite Once again. 1 84 . The an extra pair of rooks on the board. and this game is an example manages it. we have one of the most fa of the ending in Game 1S. and Game 41 is how he manoeuvres early shorten the winning process. Chapter Seven Rook and Minor Piece 'Endings In this chapter. We ings with rooks and minor pieces. Tim secure central outpost. Ulf Andersson. It is a difficult rook and knight. and one of the man managed to improve on estab main features of Lasker's deep play in lished endgame theory at the time. Despite equal material. in an end on to ensure that this does not become ing that was thought to take more than the case. although there are ending to understand. with the an extra exchange. the paucity of pawns coloured bishops is significantly dimin makes the material advantage ex ished when there are other pieces on tremely difficult to realize. mous practical endings of the last so rook and bishop cooperate well to years. thanks to the Game 41 is a classic example of rook superior pawn structure. with some extremely accu of that. but Lasker the board. and are generally better than theoretical significance. we look at various end plus bishop versus rook plus knight. but is well also plenty of exceptions. The main one worth the effort. son wins a masterpiece. He was .l::tg s+ �s 41 �g2 lids 42 son. with a rook's pawn.txes 26 discover its secrets before resumption b3 'it>g7 27 . It was played in the days i.g4 of adjournments.· I nter�<ln�t :Rio de Janeir(..:.f6 ner! With any pawn other than a rook's pawn. However.l:i.g8 22 ife2 . with even the �e2 'lites 55 .l 1�79 1 d4 dS 2 c4 e6 3 lLlf3 cs 4 cxds exds 5 Timman's chances of qualifying for the g3 lt:lc6 6 i. this is of course much more difficult to achieve.idS 6o ing whole pages to the ending.t>fs 48 bution to endgame theory.l:..f3 .tgs o-o 13 i.. .xf6 gxf6 14 l:tac1 past so years..txb4 30 l:txds 'ittg6 31 l:td4 as 32 g4 cs also fortunate enough to have as his 33 gxhS+ 'it>xhs 34 Itd5+ �h6 35 �c4 a4 second at this tournament the great 36 .. and between them the two GMs �e4 . which was crucial for This problem is exacerbated here.:!. .xf4 25 'iWxes .l:lhs 43 �f3 'it>e7 44 �g4 . with .t1 i..}g .xh3 20 lt:lxh3 schedule.h3 l:lhs+ 49 �e4 'it>e7 so �f4 'ittf6 51 caught the imagination of the chess 'it>g4 .g2 cxd4 1 lt:lxd4 .ac8 19 i.e7 61 �d4 i.. and thanks to the 18 lt:lf4 l:!..l:. by 1 85 .. White can sacrifice his rook for bishop and pawn.xes i..xg3+ 23 fxg3 of a week to analyse the position and . because at the right moment. Timman had the better part 'iVes 21 lt:lf4 l:.d6 29 b4 after the second adjournment.l:l.txe3+ 24 'it>g2 i.f4 fS+ 53 'it>f3 �6 54 loving Dutch public.· J.l:lh7 45 managed to make a significant contri ..h3 ii.f4 <iii>e s 57 mainstream Dutch newspapers devot . became one bxc6 9 \i'c2 "fle7 10 o-o i..e7 38 l:txf6+ �g7 39 Swedish endgame maestro Ulf Anders l:tfs l:!.f.t2 l:Ig7+ 46 'it>fs l:th7 47 l:!.t>f3 a3 37 lld6 l:!.xfs detailed analysis by GMs such as Don 63 �xfs+ 'i£te6 64 .d7 11 lt:lc3 of the most famous endgames of the lt:lf6 12 .d7 40 .t>g6 52 l:!. this material balance will usu ally be a fairly simple win.Ifdl l:tc7 28 . The game .. the king and rook's pawn vs king endgame will usually be drawn..O.t>ds ss <iii>d 3 ..Velitrtirovk·. since unless the enemy king is far away.l:tc4 <iitd s 56 . This game.b6 15 e3 hS 16 h4 �h8 17 lt:le2 il. R o o k a n d M in o r Piece En dings Qqme.:.tas 59 l:thl .xh4 62 �xfS+ l:!.l::tc2 i.tcs 8 lt:lxc6 Candidates' tournament. and win the resulting king and pawn vs king ending.Tin8nao.c4 . but in many lines it and Black is in zugzwang: 82 ...l::rc 3+. That is the theory..T h e G r e a t e s t Ever Ch ess E n dg a m es the fact that the white a-pawn is so far driven quickly into zugzwang. because he is ments found by Timman and Anders- 1 86 . Be Andersson's lasting contribution was cause of this.. according to the published the ory of the time. own king back to the a-pawn. to win the position.i. and we will not be able to do more here than present the main ideas and varia tions.<it>es?? 66 l:ta6 would be one journed for the second time.�d8 75 �e6 Wc8 76 Wd6 �b8 77 emy king to the h-file.b2 73 l:Ia7 �e8 74 requires a highly complicated manoeu �f5 Wf8 vre. 75 <it>e6 <it>g8 76 l:.<it>g7 69 We4 <it>f7 10 <it>f5 Wf8 11 releasing its opposite number. White can sneak home just file.. but in practice the ending is extremely subtle and difficult. one might expect the black king 8o Wd6! to head for the a8-corner. the position was a theoretical win. but in reality. the game was ad 6S . Logi positions. l:th7 �c8 78 �c6 �b8 79 �d7 Wc8 80 Then the white king returns to the l:tds <it>b8 81 'it>b6 <it>c8 82 l:td3 queenside and arranges the sacrifice of rook for bishop and pawn. A typical back.f7 i.<it>b8 83 �d8 mate. without 68.f3 i. so that even in the longest ing to avoid being forced onto the h variation. he has to stay out of the to find ways of shortening the winning a8 corner. Effectively. inside the so-move limit.. This <it>g6 . make the textbooks! 78 <it>e7 Wh7 79 l:tg3 <it>h6 66 <ilt>e4 .b2 65 l:tc6+ We7 At this point. This was the first of the improve this corner is bad for him.. based on a variety of zugzwang This move deserves comment.tb2 67 <it>d5 <it>f7 68 l:te6 White has achieved his first aim. and cut him off. Timman and 83 . . variation might be something like White's plan must be to drive the en 74.. by White gradually confines the black confining the black king on the h-file. king and drives him towards the edge but he now has to find a way to get his of the board.. One notable aspect of the ending is that.. Timman has published pages and pages of analysis on the ending.tc1 took more than 50 moves. and the way to ensure that the ending did not analysis started. or 82 . cally.c3 77 l:.c3 72 l:ta6 i. whilst at the same time try process. via c3) 91 . for example..id6 90 'it>d5 i.ies analysis.i..g2 . which allows various tactical ideas.l:. 'it>h5 85 l:t.c5.e 7+ (83 ..l:!. and Andersson found on Cheron's 82 . as presented in �h 5 85 I:tg8 "and White would have Cheron's classic book. �hs 81 'it>cs 84 'it>c4 Once again. 84.. . 'it>h4 92 l:tb3 i. or i. Now White threatens simply 82 Black has to transfer his bishop to 'it>b4.l:lg3 (stopping the bishop returning to 84 'it>b4 wins at once.. tage of having far more squares than 81 . After. R o o k a n d M i n o r Piece En dings son. when White cannot be prevented from sacrificing on a3. This stops the threat.tb2 .. since after a but on the other hand..l:.c 1 85 b2.f8 93 l:tf3 �b3 i...g.�h4 the two available on the a3-c1 diagonal.xa3. so Black's the a3-f8 diagonal.ic1 87 I:tc2 wins.id6 86 l:tg6 ..i. Established theory. e. White must full measure of the improvements he keep the black king confined.ib2 ing so �e6 �h 5 81 �f5 but which eventually takes an extra four moves to reach the position obtained after White's 84th move in the game....e ? (now the only square.. 8o... 82 Timman analyses 84 .f6 . The main line runs as follows: Timman points out that this is the 85 �b3 . much more indirect manoeuvre.xa3) 84 �d5 93 .e7 89 . Thus. in view of 84 .e 5 with the l:td3 'it>g4 83 �b4 'it>f4 84 .g6 i. had given a gained several tempi" (Timman).f4+) 94 �e6 ..b4 91 82 ..g3 and 86 .. start 83 'it>ds .g..l::.. 85 'it>b4/'it>b3 is threat ened.l:tg8 .... followed by 83 l:.xa3 . the bishop will rook move along the third rank. black king would come off the h-file... 1 87 .i.if8 87 'it>c4 �h5 88 most stubborn. i. That has the advan next is forced. after which the sacrifice fails. and 85 �xa3 'it>e5 86 �b4 'it>d6 and the gives fascinating variations showing the black king is in time.g 2 . e. z:tgs i.b 2 86 l::!.ixa3+ same basic idea as his main line. 84 . the now be undefended by the b-pawn.. i.l:.if6 there is 83 ..b4? 94 ... ..c1 91 1:tc3 . b) The other option is 95 .tc1? 90 .Ud3! ! .tf8 96 �f6! (the bishop gradually runs out of squares) 96 .tg7?? is impossible because of 98 .tc5 102 l:lg4! �h5 103 .l::i.... Black now has only two moves that do not lose material: a) 95 ...b4 98 lld5+ 'iZith6 99 lld3 'it>h 5 100 �f5 �h6 101 l:tg3 ....tf8 100 l:t. as Cheron had found. 106 ..l:i. the line J:!c1! wins the bishop after 101 . although once again Timman and Andersson had speeded up his winning process. Cheron had analysed positions with the black king on h3/h2/h1.i.�h 5 97 l:ld8 ii.d4.g 8) 98 lk4+ �g 5 (98 .i.d1. whilst other bishop moves lose the pawn after 107 ....e3 1 88 ..td6 104 'lt>e6 . but this last 102 l:..97 .l::i. and shown them to be winning... Thus far...i.b2 92 . this is a bad square for the black king.Ud3.. gave a very long and complicated 85 l:tg6 i.gs 86 'it>ds . �h 5 99 'iZitf5 .e 5 lose the bishop to 107 l:td5.Uc8 i.l:k4 .....i.�h4 Black's king is now forced down the board towards the h1 corner. which is the reason why the king is so bad on g4 .Th e G r e a t e s t E ve r Ch ess E n dg a m es squares... Finally..l::i.i.'1t>g4 but.l:i.tc1 87 'it>e4 .i.h 6 (forced.tf8 105 �7 . in view of 89 . which 88 'iZitfs �hs 89 .tf8 97 .b2 (though beautiful) variation. 89 ..c 5 or 106 ..c8 .d6 requires another 30 moves to win. Cheron is analysis by Timman)..... This is shown by 96 . but the text is quicker...id6 106 .h1+ 'lt>g3 103 l:tg1+) 99 �f7! and move was the big improvement found there is no defence to 100 i:tc3 (all this by Timman and Andersson.c 3 .l::i.g7 101 and now 95 . �h4 follows Cheron's analysis. the bishop runs out of 90 lid3 . similar zugzwang. but motely similar in a practical game here.tcs+ 99 Wc4 i. 1 89 .tf6 99 l:. it does not get hit with tempo by 92 .l:ta4 Creating a similar grandmaster achieving anything re zugzwang to that in Cheron's line...h3 when Black is in zugzwang and must move his bishop.i.xa3 and wins.b2 when 94 eral extra tempi for White.. and with increment time-limits. but in the present game this to d6 anyway.d2 Black's king.e7 Timman now gives 98 �2! (Cheron only considered 98 . 97 �e4 We2 98 'lt>d4 .) 99 . it was only possible 105 �c2 �2 106 . Andersson. White is then able to 100 l:[h3 i.i.f8 102 l:i.b2 theory.h4+ �g1 99 l:i. but now the king is cut off behind the pawn.c1 93 . would have taken Timman perilously 103 l:ta8 1-0 close to the SO-move rule.. 99 .. 94 l:th1+ �g3 95 l:i.d1 i. had analysed.h8 bring his king to e2-d1-c2 and prepare . as Cheron 102 .b2 A magnificent achievement by 101 . R o o k a n d M in o r Piece En dings Transferring the rook to e1. However.l:tb3 with another.tb2 102 'it>c2 �f3 'it>f3 �h2 reaches the position Cheron 103 'it>b3 �e3 104 .l:. Timman and his second. The continuation and a lasting contribution to endgame 102 .it is already attacking a3..l:te1 i.l::!....g2+ 'it>h1 100 �1 i. with no adjournments and placed ..U. This all translates into sev The main line is 93 . Now there is no defence to the sac the text shortens the process consid rifice on a3..tc1 100 �g1 �h3 95 �f4 'it>h2 96 l:lg4 �h3 97 'it>e2 �g3 101 �d1 ..d6 101 �b3 i..) 98 . .td6 a winning sacrifice on a3. 98 . since the black king cannot erably.l:td3+ 'it>f2 Forced. the white rook is much better nowadays....i. and there is some six moves faster than Cheron's no realistic prospect of a modern line.b4 96 ..l:tg3 'it>h2 102 ......xb2 leaves White because of adjournments.cs 103 l:lc8! forces the bishop shows. Of course.tc1 103 �e2 'lt>g2 104 'it>d1 i.'1t>h3 (Timman also analyses step onto the d-file. tbc3+.Th e G re a t e s t Ever C h e s s E n dg a m es Game40 A. The threat was 42 .e2 lbb6 35 cxb3 Lasker's tactics have paid off. in the l:.g1 . he had therefore played l:.e3 ii. e.id6 9 Ji.l:. In the tDb3 tbg6 12 ii.. this is forced.xb3 40 . We join the a4 30 bxa4 ttJxa4 31 es fxes 32 l:.g. that is only the start of the interest. Moving the rook away.d4 tbh4 17 in his typically provocative style. l:.f2 tbg6 20 fs ing straightforward equality in favour i.Aiekhine--Em. After a slow start. this game has been a rather tense affair.he1 b6 15 Ji.xh2 23 of stirring up complications..Jixb6 i. Lasker had found himself trailing the Cuban significantly. and in order to 190 .xb3 21 axb3 ttJf8 22 . The play was greatly affected by the tourna ment situation.xds 43 tbe6 Thus far.l:Ig2 ttJds 41 .eb8 33 tbe4 b3 34 l:.Jid6 19 ii. Lasker was involved in a big battle with Capablanca for overall tournament victory.. spurn l:. 42 .xes game just at the moment where l:..l:Id3!! 42 l:. 42 .Lasfcer St Petersburg 1914 1 e4 es 2 tDf3 tbc6 3 i..bs a6 4 Ji.d1+ 43 �c2 tbe3+ picking up the other rook... he dxc6 5 tbc3 f6 6 d4 exd4 7 'iVxd4 �xd4 needed to win practically every game 8 ttJxd4 .e3 tbe7 10 0-0-0 0-0 11 in the second half of the event.l:tb7. whilst 42 �cl walks into a snap mate after 42 . 41.cs Ji. in which Black has objectively had the worst of it..ie6 18 f4 .. lla1+ 43 Wc2 ltJb4 mate.e8 14 present game. allows 42 .es 16 i.xc6 have any chance of catching him. as ttJds 36 g4 h6 37 gs hxgs 38 tbxgs tDf6 White's last was a serious mistake 39 lle7 l:. However.xds Surprisingly.l:td7 which costs the exchange..h1 cxb6 24 llxh2 bs 25 lle1 tDd7 26 hope thereby of obtaining chances to tDd1 as 27 �h3 b4 28 tDf2 ttJcs 29 llhe3 outplay his opponent.f4+ 13 �b1 l:. R o o k a n d M in o r Piece E n di n g s Despite his extra exchange, Black's 47 .l:tf7+ wes 48 'iit>c2 l:.h6 49 li:Jd3+ Wd6 game is extremely difficult to win, as so .l:.fs l':f.b8 51 'iit>c3 'iit>c7 he is losing the g-pawn, with the result Black's king cannot really do any that there will be only one pawn thing active, because it is too vulner apiece. Fine points out that with an able to checks, so instead it remains in extra exchange, the most difficulties in the background, out of the way, and winning often arise with two rooks vs also takes on the job of guarding the c rook and knight, because the latter pair pawn. of pieces can cover all the entry squares 52 .l::tf7+ 'iit> b6 53 l:Id7 and make it very difficult for the Preventing Black from ganging up stronger side's king to penetrate. With on the knight with ... l':!.d8 and ....l:th3. so few pawns left as well, the difficul 53..J:th3 54 l:td4 l:tbh8 55 .l:.b4+ Wc7 56 ties in winning are at their greatest. 'iit>c2 .l:.8h4 57 .l:.b3 43 ...'iit>f7 44 l:txg7+ 'iit>f6 45 .l:.c7 .l:td6 Now the rook has been forced into a Obviously, Black needs to hang onto rather awkward position. his remaining pawn, at all costs. 46 li:Jcs 'iit>xfs 57...l':!.h2+ Fine points out that 57 ... c5 was As Fine explains, so long as White rather quicker: 58 .l:tc3 (58 li:Jxc5? .l::i.c4+ retains his rook, there is little that Black 59 'iit>b l l:lhl+) 58 ... Wd6 59 .l:.a3 c4 60 can do, whereas once the rook comes .l:.a6+ Wd5 61 li:Jc1 .l:.h2+ 62 Wbl .l::th 1 off, the position is an easy win for etc. Black. The latter's task is therefore to 58 'iit>c3 l:l4h3 59 .:.b4 1lh5 60 l:!.g4 l:.2h3 force the exchange of White's rook, 61 'iit>c 2 lids 62 li:Jf4 l:i.cs+ which he does by a combination of 62 ...l:td4?? 63 li:Je6+ would be unfor mating threats and pins on the knight. tunate! With Alekhine defending stoutly, this 63 'iit> b 1 l:[h1+ 64 �a2 l:f.as+ 65 �b3 proves a long process. .l::t b S+ 66 �C3 Wb6 191 T h e Grea test E v e r Ch ess E n dg a m es White is gradually running out of moves. 82 l:!.g1 l':.h3 83 l:td1 l:.dh8 Again threatening the rook ex change by ... l:th1 and ... ll8h2. 84 .l:i.g1 l:.8h5 85 �c2 .l:Id5 86 l:td1 .Ug5 For the moment, Lasker continues to manoeuvre and probe, without mak ing any real progress. This is a natural part of the practical playing of such an ending. The pressure of the clock, among other things, means that such an approach is frequently inevitable. 87 .l:i.d2? 67 lDd3 .l:.h3 68 'it>c2 l:Id5 69 l:i.b4+ This loses at once. Fine analyses the 69 lbf4 l:td4 would force the ex toughest defence as 87 .t:i.f1, after change of rooks at once, whilst 69 lDf2 which he shows a win, as follows: J:i.h2 70 l:.f4 .l:!.g2 leaves White tied up. 87 ....l::i.g 2+ 88 'it>c3 .l:!.hh2 89 .l:!.c1 cs 90 69 ... Wc7 70 .l:tb3 .l:.h2+ lbf4 .l:ta2 91 lbd3 l:La3 ! (now c4 is a 70...cs is again possible here. threat) 92 .l:!.b1 c4 93 lbc1 'it>cs 94 .l:.b2 71 'it>c3 'it>d6 72 l:i.a3 .l:i.g2 73 .l:i.a1 .l:tg3 74 (the exchange of rooks is now tactically .l:i.d1 permissible, because Black's last pawn Now the knight is pinned both would fall, but Black himself can now ways, and Black can start to activate his avoid it) 94....l:!.h3+ 95 'it>c2 c3 96 .l:!.b1 king. .l:!.a8! and there is no defence. 74../lJc7 75 J::td 2 �b6 76 l:Id1 'it>b5 77 87 ....l:.hg3 Wc2 And finally there is no defence Conceding further ground, but after against the exchange of rooks. 77 .l:td2, Fine points out 77 ....l:.cs+ 78 88 lbc1 88...l:Ig2 89 lbe2 Wb6 0-1 �b3 .l:tcg s and there is no defence A fascinating endgame, which illus against the exchange of rooks: 79 'it>c3 trates the difficulties that can lie in the .l:.g2 80 .l:.d1 .l:tg1 81 .:.d2 .l::i. S g2 etc. way of converting an extra exchange, 77 ...'it>c4 78 b3+ 'it>b5 79 .l:.d2 .U.h3 8o especially with few pawns remaining J:!.d1 .l:i.h2+ 81 'it>c3 l:td8 on the board. 1 92 R o o k a n d M i n o r Piece En dings Gome 41 �F.MarthaJI-Em.ulker World Championship (Game �), l)SA 1907 1 e4 e5 2 lt::lf3 lt::lc6 3 j_b5 lt'lf6 4 d4 exd4 elsewhere in this book, when discuss 5 0-0 j_e7 6 e5 lt'le4 7 lt::lxd4 0-0 8 lt'lf5 ing knight vs bishop positions, the infe d5 9 j_xc6 bxc6 10 lt::lxe7+ 'ii'xe7 11 .l:.e1 riority of the knight is only an issue, 'ii'h 4 12 .ite3 f6 13 f3 fxe5 14 fxe4 d4 15 even in open positions, if the knight g3 'ii'f6 16 j_xd4 exd4 17 l:rf1 'ii'xf1+ 18 lacks a good outpost. If it has a decent 'ii'xf1 .l:.xf1+ 19 'iii>xf1 central square, on which it can en trench itself, it can be the equal of the bishop. In this position, given a couple of moves, White can bring his knight to the excellent square b3. From there, it would attack the d4-pawn, control cs and potentially be able to jump to as in some situations. If White is given time to get his knight to b3, it would not be inferior to Black's bishop. For this rea son, Lasker plays a much deeper series of moves than the obvious bishop We have already seen one rook and check. He starts with: bishop vs rook and knight ending in 19 ....l:.b8! this book, Fischer-Taimanov (Game 15). As emphasized there, rook plus bishop are usually stronger on an open board, and with pawns on both flanks. Here, we have such a situation. As far as the pawn structures are concerned, here both sides have some defects. Black has doubled c-pawns, but White has three islands, and his e-pawn, although passed, is just a serious weakness. I think 99% of players would have played one of the bishop checks here, The first point of this move is to followed by bringing the black rook to force the pawn advance... the f-file. However, as we have stated 20 b3 1 93 The G r e a t e s t E ve r Ch ess E n dg a m es ... after which the white knight is i..d3 nets a pawn, but a worthless one . deprived of its excellent outpost! Ad After 24 t2Jf3 l:i.xc2+ 2 5 l:txc2 i.xc2 26 mittedly, the knight has a square on c4, t2Jxd4 i..xe4 27 t2Je6 i..b 1 28 a3 i..a2 29 but that is no good, because a subse ttJcs, White is in no danger. quent ... i..xc4 would shatter the white pawn structure. 2o...:bs! 21....l:.hS 22 'it>g1 cs Now the d4-pawn is defended, and ... i.b7 becomes a possibility, attacking This is the second point of Lasker's the weakness on e4. last move. Although his bishop is unde 23 t2Jd2 �7 veloped on c8, it is actually very well placed. It has two diagonals on which it can potentially jump out, so Lasker concentrates instead on activating his rook. What he has realized is that rooks do not depend solely on files for their activity - sometimes, they can operate very effectively on the ranks as well, as is the case here. 21 c4?! This transforms black's d-pawn into a powerful, protected passed pawn, 24 l:!.f1+? but otherwise, there was an unpleas This seems to be the losing move, ant threat of ... l:tcs, tying White down wasting a tempo to drive the black king to the defence of c2. However, Kas further towards the centre of the parov shows that 21 tLld2 should en board. Immediate counterplay by 24 a3 able White to hold, e.g. 21 .. Jks 22 l:!.c1 was essential, when Shereshevsky and the tactical trick 22 ...i..a6+ 23 'it>f2 showed that White should just be able 1 94 d 4 33 . cates Black's win ..l::i. <J.l:. 21 . Either way. R o o k a n d M i n o r Piece En dings to hold.l:tbl <J. so as to harass possible.a6 27 .. only ten moves after the start of Once again.. 24.a1?! to zugzwang.<J.fl White's position is already quite lost....l:!.c 3 35 'it>e3 and and his rook immediately seizes the then a waiting pawn move was also opportunity to use it. thanks 26 h4 . he may as well have done so at possibility in the next note.e s Now the king also occupies a superb outpost. Lasker uses the ranks the endgame proper.l:!. White's last de the white queen side pawns. effectively. e. 30 .f6 31 .. 24.We3 would have precluded the anyway... fensive lines will be breached.. In view of the fact that he ends up 35 as a6 36 li:Jb1 'it>xe4 37 'it>e1 ii.e1 d3 ••• With the white king cut off. once and saved a tempo.a5 2 5 . there is nothing to stop its opposite number 38 li:Jd2+ marching into the heart of the enemy 38 Wd2 followed by li:Jc3-a4 compli- position..l:!..xf6 gxf6 34 Wf2 c6 been closed. 30 'it>g2 White has been driven into com plete passivity.l:!.g.. although it is not obvious that he can do anything useful with that tempo.l:!.e2?! playing a3-a4 in a couple of moves 37 . The fourth rank has now 32 .i.xb4 etc..compared with the 1 95 .We7 25 a3 .e 7 26 b4 axb4 27 axb4 cxb4 28 ... but the third rank is open.g4 28 'it>f2 'it>e6 29 a4 <J.h6! 32 .. The immediate 34 .l:!. so Lasker now removes the white kingside pawns.�3 42 lLlc3 'i£i>xg3 43 l2'la4 f4 'i£i>f4 44 lt:Je6+ is far from clear) 43 lDcs ... there is n o chance of 38 .i. but providing Black Hope springs eternal.'iite 3 39 l2'lb1 fs 40 lt:ld2 hs 41 lt:lb1 White still doesn't take the hint. After successive moves. lDxcs f4 42 l2Jxa6 i.. and could have resigned consid erably earlier... does not permit his opponent several 47 . and try to pro b4 mote his a-pawn.. Marshall's last hope is to ar 44 l2'lxcs f3 45 lLle4+ �f4 46 ltJd6 cs 47 range l2'lc3-a4xcsxa6. . in which Marshall goes down without a fight.Th e G re a t e s t Ever C h e s s E n dg a m es game.. .xd3 46 a7 f2 47 a8'iV fl'ii' 48 'iWxc6 'i£i>xh4 Black should be winning. but this is far less clear than the game continuation.cxb4 48 cs b3 49 lLlc4 'lt>g3 0-1 196 .f3? 43 lDcs 41. 38 ..g 4! (42 .fs 44 a6 f3 45 lt:Jxd3! i.<�f3 39 lt:Jc3 fs 40 lt:Ja4 'i£i>xg3 41 that ever happening. White gains several tempi.... . In the absence of the rooks. fixing weaknesses is cru Black is seeking to establish coun cial.il..l:Ib7 24 llc5 l:tg8 the split black queenside pawns.hc1 l:!. 21 .xas 22 l:!.te3 h5 13 h4 . the less great will be the drawing tendency of the opposite coloured bishops.f6 23 a4 .cc11 This is an important moment. as? 21 bxas l:!.td2 �e6 Once again.e3 winning a pawn. 19 b4 25 f31 As always.l:i. g7-g5. R o o k a n d M in o r Piece En dings Game 42 U...txc6+ move is that Black cannot get rid of the bxc6 11 lLlc3 e5 12 .Anclef$son.. As stated above.Stean sao Pauro 1979 1 c4 l2Jf6 2 lLlf3 c5 3 g3 d5 4 cxd5 l2Jxd5 19..Jidb8 5 i. as? because of 20 . He will therefore 17 a3 .e6 bxas l:txas? 21 . and consists of 22 . . Andersson gives a veritable tour de force of technique. .xc3 .c5 .b5 22 .txc3+ 18 l:!. the drawing ten dency of the opposite bishops is greatly reduced..b6. the more pieces remain on the board.. 14 �d2 rJite7 15 l:!.g2 l2Jc6 6 d4 cxd4 7 l2Jxd4 l2Jdb4 8 A key tactical point of White's last l2Jxc6 �xd1+ 9 Wxd1 l2Jxc6 10 . Andersson therefore White's formal positional advan keeps both rooks on. terplay on the kingside by . with opposite-coloured bishops on the board. it is hard to believe that this advantage should prove sufficient for a win.l:!. How ever.d8+ 16 We1 i.M. tactics prevent Black ridding himself of the a-pawn: 20 .td5 have to remain with both queenside pawn weaknesses. there would of course be no winning chances at all. 20 .l:. tage is clear enough. so 197 .tb4 weak a-pawn by 19 . but with two pairs of rooks on the board. .l:k7+ 'lt>e6 40 .l:.ltcs l:!.cs bishop on the g1-a7 diagonal. so as to have the possibility of infiltrating with his king. Again we see the "repetition trick". against this. On the prin ciple of "two weaknesses". bringing time to execute such a manoeuvre. . White would normally try to keep the king side open.Th e G r e a t e s t Ever Ch ess E n dg a m es White takes prophylactic measures As always.l:txa7 etc. Andersson i s i n n o hurry.d4 'lt>f7 33 White gains time on the clock and un I:!.b3 ? just loses a pawn af ter 3 S I:!.. this would mean his pieces being tied to the defence of the pawn on f3.. 33 'lt>d2? ... and can only mark time.l:!.dd7 32 i. and then attack Black devoid of counterplay.e3 .g.. he needs the enemy bishop after 26 e4.xb4 39 . His king to a3.l:!. 37 . and placing the 39. With defend the b4-pawn.xd4+ 37 'it>e3.g. discoveries with his bishop from dS 2s .e7 38 . The text threatens to trap Before the king can leave e3. 26 'lt>f2 l:!. which involves playing the trate on the queenside weaknesses.l:. However.as 'lt>f7 37 l:!.lta2).l:.xc6+ 'lt>f7 36 I:!.c3 e4 28 .l:r.. square. 36 l:i.d1 l:tdd7 29 33 .U..fs (e. Black's potential king moves his last piece from a white side counterplay has been eliminated. 35 'it>c3 �e6 Black has no counterplay at all.l:txb4 . i.l:.c2-cS-aS. 30. In stead. and then playing .bc7 40 i..a6 . White has the a-pawn with all his might.b1 .b7 41 i.xb3 J:.ed7 39 . J::tdxb4 38 .l:.d2 30 f4!? An interesting choice... e.d6 31 .l:!. White now wishes to get his rook on but thereby frees his hands to concen d2 to as.b2 derlines Black's helplessness. so Andersson re At a stroke. thus in to ensure that Black has no dangerous ducing the reply.d8 27 i. 198 . his rooks to a6 and as.�e6 34 'lt>d2 �f7 'lt>e3 g6 Now 34 . Andersson closes the kingside. moving the bishop to e3 or plan now is to transfer his king over to d4. any hurry to play this move. 46 . but to cS. system king.a2 ••• Now he loses c6. Andersson just improves his A superb piece of patient. square and securing the post d4 for the and instead forces the exchange of a bishop. R o o k a n d M in o r Piece Endings 41 .l:f.l:.U.. sive error on his part! 1 99 .dc7 46 i. now the king returns to a more even harder to pinpoint a clear or deci central position. a typical Andersson ture on a7 with the bishop.b7 49 �C3 . but it is too late to make 4S l:. There is nothing wrong with the his grip.. and gradually advance the a immediate 48 .d7 47 l:i. pair of rooks. but there was no 48 �b2! defence anyway. Now it is clear that the a7-pawn is so l:tcas Itxa6 51 .bc7 42 l:tc2 l:tb7 43 �b2 l:tdc7 44 ••• 48.cs .d4 ••• any difference.. It has done its job on a3. White will just cap Once again. by cult to believe that the starting posi clearing the c-file for the rook to come tion of this ending is lost for Black. maintain touch. Conse 52 l:txc6 1-0 quently.b6 �a3 ltd7 45 e3 Black finally wearies of shuffling his Getting another pawn off a light rook to and fro along the seventh rank. by bringing it to the more central atic technique by Andersson.bc7 ••• S1 i. but nor is there pawn.l:txa6 doomed. It is diffi square c3.l:Icas.l:!.l:!. in combination. All are one of the most famous and instructive united by general endgame principles exchanging operations in all of chess . we look at a selec has a good.use of the king. Magnus than rook plus bishop. secure. do not hurry. and unrecognized experienced players. central outpost. pawns. This game also features some combination or other. even at Petrosian-Botvinnik world champion a young age . Steinitz gives a perfect illustra volve exchanging off the opponent's tion of how the stronger side should bad bishop. and one which shows how Karpov are the three examples that rook plus knight can be more effective spring immediately to mind. but many of the endgame specialists of the past-war era.Capablanca. Chapter Eight Multi-Pie£e ·Endings In this final chapter. Paradoxically. Fischer shows The material involved varies. Game 44 is a masterly dem Endgame technique is something onstration of the agility of knights. shows the opposite than the usual 2-3 pieces on each side. and the key. an unexpected exchange holds the key Game 43 is one of the very first tour to White's play. more one of the great. side of the coin. two weaknesses. using his pawns weaknesses. however. tion of endgames where we have more Game 46. great natural talents in chess have Game 45 is a masterpiece from the been outstanding technicians. and Game 48 is a 200 . in order to expose his handle such positions. creation of passed literature. but most the strength of the rook plus bishop involve rooks and minor pieces. by generally associated with older. the centre. Fischer and ship match. Game 47 is another game in which etc.to and knight. With rooks also on the success can therefore sometimes in board. There. Game 47 is one of the to deny the knight a secure outpost in great examples of this. when the knight Carlsen is another. bad nament games to demonstrate the su bishops are often very effective de periority of the bishop pair over bishop fenders of weak pawns. M u lti. as the then 14. whilst Game so sees Kramnik saving Finally. maturity which belies his years. with a fine positional from world championship matches. In squeeze in a must-win game. Karpov gives another mag year old Norwegian wonder displays a nificent display of his prophylactic play. we have two masterpieces his world title.Piece Endings superb example of this. Game 49. 201 . fd1 Steinitz carries out the plan to perfec The Russian endgame author tion. Ji. but An absolutely typical move in such as it is. static sort. b 5 g6 4 d4 exd4 15 ..Englistb-W.l:tfe8 17 'Llb3 b6! This ending is one of the very first.. The key point is that the white knight does not have a secure outpost in the centre.. it is all about slowly and methodically strengthening his position.. yet the possession of the two bishops gives Black a clear positional advantage.d7 20 2 02 . The key strategic idea for ing his pawns to restrict the activity of Black is to use his pawns to drive the the enemy knight. examples of the exploi tation of the bishop pair..e2 'Llg4 13 Ji.xg4 14 b2.g7 6 i. If it did. Panchenko suggested 19 'Lld4 i.e6 19 . Black begins the task of us defensive.The G r e a t e s t Ever Ch ess E n dg a m es Gome 43 B.. There 'Lle7 9 �d2 d5 10 exd5 'Llexd5 11 'Llxd5 would be no sense at all in taking on 't\Vxd5 12 Ji. he would have chances to battle for equality.l:tad8 5 'Llxd4 i. White 'Llb3 �xd2 15 'Llxd2 would regain the pawn and create counterplay with his active rook amongst the black queenside pawns. which is now de enemy knight out of the centre and prived of the square cs.Steinitz LOndon1883 1 e4 e5 2 'Llf3 'Llc6 3 i. as after 1S ..l:!. prevent it finding a decent post. he will be permanently on the positions. Instead.xb2? 16 l:tab1..xg4 i.e3 'Llf6 7 'Llc3 0-0 8 0-0 Already a notable choice. 16 c3 . and so it makes no sense for him to rush things or al low unnecessary counterplay. Black's positional advantage here is of a long-term. 18 h3 i. and still classic. Material is equal and the pawn structure symmet rical. not rush things .xd8 I:txd8 24 . which is As in almost all endgames.. Steinitz does not un dertake anything precipitate. the alternative was to allow Black to ad vance his own pawn to f4. Once again. he just con centrates on strengthening his posi tion. M u lti.. to avoid this. since it fixes a pawn on the same colour squares as his bishop. using his pawns to gain 20 �g5 space and drive back the enemy minor White hopes to induce Black to sur pieces.. bit by bit.. 23 l:!. the black certainly a better chance than White's king has a part to play and Steinitz passive and stereotyped choice. such as 203 .f7 This is not a move White will have played with any enthusiasm. Note how he does 19 cs! . Now the white knight cannot reach the centre via d4 and stands very pas sively on the b3-square. 26 ii. but Englisch prefers to fight for space.. starts to centralize it.e1 fS 26 f4 20. which would severely cramp the white position.f6 ••• Once again.l:!.Piece En dings a4! fighting for counterplay. 22 f3 g5 Continuing the job begun at move 17.f6 21 il.lte3 h6 but Steinitz is not afraid to shut his This prepares to seize further cen dark-squared bishop in for the time tral space by means of the advance being. 25 . It is a Hobson's Choice.f4 <J. The bishop will .instead. render the d-file by exchanging on dl..f6-fs.. emerge into activity again later. However. . the bishop pair forces weak pawn and maintains that arrange nesses. White is being remorselessly driven back wards. so as to defend a2. Steinitz just defends the gS endings .. This is to control the maximum amount of an absolutely typical scenario for such space.llc4 . thereby making its exchange a 27 g3 serious threat. but if he does not play this. so as knight comes onto the agenda. In all endings. since when the knight is attacked by . it will be forced to retreat to cl. further hampering his tion. which only the knight can de ment. side by side. which is knight will be forever tied to the de why the ability to calculate variations fence of a2.. Steinitz abandons Another white pawn is fixed on a his patient strengthening of the posi dark square... with piece that can stop this is the white the bishop pair and a space advantage. knight. 28 lt:Jc1 a4 29 a3 With this move. Now the knight is effectively para lysed. White must constantly reckon on its being exchanged off. 30 gxf4 .. Another absolutely typical space gaining pawn advance.The G r e a t e s t Ever C h e s s E n dg a m es exchanging on f4 (which would just now has a highly inviting route into the weaken the black kingside pawns) or white queenside. With the white a2-pawn now undefended. whereupon the weak white squares on the queenside could be sub ject to invasion. via dS-c4-b3.. In such positions. The only advancing to g4. aS-a4.. this gains in strength. 29 . 30 'it>f2 21 as! . and goes over to a concrete course bishop and weakening his light of action. his comes when this is necessary. fend. so now the possible exchange pawns often stand most effectively in a of one of the black bishops for that phalanx formation.. if it should move to e2. the moment squares. Note that the black king well is a crucial factor in successful 204 . l:.e8+ ••• vantage for dear life. lier. and instead transposes into an ending with rook and bishop vs rook and knight. Whereas most of the play hitherto 205 . whereas a strong Steinitz calculates that he can ex master will often be seen to exchange change rooks and win the minor piece one advantage for another.. thanks to his much more active 32 �xgs hxgs king and the fact that his bishop domi Now Black has a potential passed nates the knight.f4 himself.. Another plan...bS-b4. seek ing to activate the dark-squared bishop down the long diagonal h8-a1. while there is 33 �e3 �f6 nothing at all objectively wrong with Steinitz' decision.ds. 34...gxh4 35 gxh4 This is Steinitz' idea.Piece En dings endgame play. and we will see other examples in this book. 31 �xf4 �g5 34 h4 White himself breaks the Gordian Knot. and a possible tar the concrete play that I referred to ear get on h 3. However. ending. it was not essential in this specific case. He abandons the advantage of the two bishops. Such positional transformations are a typical device in exploiting an advan tage.. If he sits tight. was to carry on the process of gradual squeezing. Many club players have a tendency to cling onto a particular ad 3s .. among other things..l:. This is an example of pawn on the f-file. but his position is indefensible in any case.. Black can al ways arrange .. sug gested by Panchenko. by 30. This defends the cs-pawn and so prepares the advance . 'lites and . M u lti. Th e G re a t e s t Ever Ch ess E n dg a m es has been on general principles, with Desperation, but now Black liqui very little calculation, now it is all dates completely. about calculating concrete variations. 38 ....i.xe2 39 'it>xe2 'it>f4 40 c4 'it>g4 41 36 'it>f2 .l::i.xe1 37 'lt>xe1 'lt>es We3 White's h4-pawn will be lost to the 41 ...f4+! rampant black king, and to all intents But not 41 ... 'it>xh4?? 42 Wf4. Care is and purposes we have a king and pawn always required! ending, since the white knight is para 42 We4 f3 43 'it>e3 Wg3 0-1 lysed. Black has no need of the enemy h 38 lLie2 pawn, but just queens his own instead. 206 M u lti-Piece En dings Game 44 G.Barcza B.Soos.. Varna Olympiad 1962 1 lZ'lf3 tt:'lf6 2 g3 g6 3 b3 �g7 4 i.b2 cs 5 of the best examples. c4 tt:'lc6 6 i.g2 o-o 7 o-o d6 8 tt:'lc3 �b8 9 The early queen exchange looks like d4 a6 10 dxcs dxcs 11 l\Vxd8 l:txd8 just a tedious exchanging operation, but is actually the start of a serious initiative. 12 tt:'la4! This is why. Barcza has seen that Black has some difficulties defending his cs-pawn. 12 ...tt:'ld7 12 ...b6 leaves the rook tied to de fending b6, so that Black has trouble contesting the open file after 13 l:tadl. 13 �xg7 �xg7 The Hungarian grandmaster Gedeon Barcza is now largely forgot ten, but in his day, he was the "grinder par excellence", a sort of 1950s/6os ver sion of Ulf Andersson. Famous for his tt:'lf3-g3-i.g2 systems, he liked nothing better than to exchange queens early and grind away in a long endgame. Harry Golombek is alleged to have said of him that "Barcza has the most varied opening repertoire of any grandmas ter. He sometimes plays g3 on the first 14 �ac1! move, sometimes on the second, some A neat piece of prophylaxis against times on the third, and occasionally not Black's ... b7-b5 advance. Now Black until the fourth move", a joke Barcza cannot get rid of the annoying knight himself was very proud of! Barcza was on a4, which attacks his queenside and also especially fond of knights, and hampers the mobility of his pieces. won many games by remarkably adept Knights on the rim are not always dim, handling of his cavalry, this being one Dr Tarrasch notwithstanding ! 207 The G r e a t e s t Ever Ch ess E n dg a m es 14... b6 15 l:!.fd1 4Jf8 16 l:txd8 4Jxd8 17 "Come into my parlour", said the ttJes spider to the fly... Probing the c6-square. 17 ...f6 24 tbc7 as 25 tba6! l:!.e8 2S ...l:tc8 is met by 26 i.b7 driving 18 .l::!.d 1! .i.fs 19 ttJd3 i.d7?! the rook away, and securing c7 as an Soos takes radical measures to rid escape route for the knight. himself of the irritating knight, but his 26 tbe4! .i.c8 27 tbc7 l:!.e7 pieces end up awkwardly bunched on the d-file. Barcza suggested Black would do better to take on d3, al though he did not give any variations. I assume that the idea was 19 ... i.xd3 20 l::txd3 4Jde6, followed by ...bs. 20 4Jc3 4Jc6?! A further inaccuracy. Barcza's rec ommendation was 20... .i.c6 21 tbf4 .i.xg2 22 'it>xg2 4Jc6, although then 23 ttJa4 threatens ttJds. However, Fritz suggests the balance can be main 28 tba8!! tained by 23 ...4Jb4 24 l:td2 g S, which An amazing concept. The knight seems satisfactory for Black. jumps into the very corner of the 21 tbf4! board. Barcza's play in this game is very Now 22 .l:i.xd7 is the main threat. concrete and unprejudiced, and in that 21 ...tbd4 22 e3 ttJfs sense, bears some resemblance to com 22 ... 4Jde6? again walks into the tac puter play. The silicon machine would tic 23 .l:!.xd7!. not hesitate to play a move such as 28 23 ttJfds! e6 4Ja8, if it is effective, but for a human 208 M u lti- Piece En dings human player it requires a special ef dangerous after 32....tb7 33 lLixb6 fort to consider such a move. The point lLif3+. is that Black has serious trouble de 31 ... bs 32 cxbs lLixbs fending his b6-pawn. 28 ...l:f.d7 28 ...lLid7 loses the pawn after 29 lLid6 lLixd6 30 .l:.xd6. 29 .Uxd7+ lLixd7 But what now? 33 .ic61 But now, finally, White will win a pawn, in more promising circum stances. 33 lLid6 34 .txd7 .txd7 35 lLixcs •.. 30 lLic31 Preventing Black's intended ....ib7 and preparing to attack the b6-pawn again with ltJa4. Barcza analyses 30 g4 .tb7! 31 gxfS i.xa8 32 fxe6 lLies 33 lLid6 i.xg2 34 �xg2 �f8 35 lLic8 lLid3 36 lLixb6 'iiie 7! and suggests that, despite White's ma terial advantage, the knight ending is hard to win. Black's knight is very ac tive and threatens ...ltJc1, and even af 3S ....tc6 ter 37 lLidS+ �xe6 38 �3 lLic1 39 lLic3 Losing another pawn, but the <itr>es the two Black queenside pawns theme of Black's bishop lacking squares hold White's three. Black's activity is continued after 3S ... i.c8 36 lLib6, would give him drawing chances here. when the knights once again display 30...lLid6 31 lLia4 their awesome power. After 36 ... lLib7 Barcza points out that winning the (36 ...�f7 37 lLixc8 lLixc8 38 lLib7 wins b-pawn by 31 .tc6 lLies 32 .ta4 is too another pawn) 37 lLie4 (the bishop stm 2 09 even at lovely games.xa4 41 lZ'lb6 i. the black endings. even on an open board..�f7? is impossible because of 37 lZ'ld8+. and White is winning. It is remarkable how.f s 38 tl:ld2 tl:ld6 3 9 bishop finds itself dominated by tl:lxc8 tl:lxc8 40 tl:lc4. endgame.. White's equine artistry has netted A super game by Barcza. you h ave not one word of Hungar A neat final touch. The a-pawn costs the bishop his life.The G r e a t e s t Ever Chess E n dg a m e s has no squares!) 37 .. the as-pawn drops White's knights.We8 45 lZ'lds f5 46 lZ'lc3 Si. I can strongly recommend Barcza's the bishop's path back into the game best games collection.d1 42 a4 tl:lcs 43 fective in closed positions.e8 39 lZ'lc4 a4 40 to think that knights are only ever ef bxa4 i. It is a mistake 37 . White shuts off ian... including numerous fine this late stage of the game. which him two extra pawns and a winning shows the extreme agility of knights. 37 lZ'lb6 44.. Even if. which h as some via g4. 210 . as well...a4 47 a61 1-0 The final humiliation for Black's un fortunate prelate. like as �f7 44 f31 me. 36 tl:lxe6+! �g8 36 .lZ'le4 38 lZ'ld4 Si. l:lxd8 l:txd8 Such an early queen exchange often presages a quick "grandmaster draw". The knight The assessment of the position is attacks e6. Rather. However. was quick to scotch this rumour. stories begun to circu late that. for his knights the traditional "square in front of the isolated pawn". imme diately bringing the king into the cen tre. In fact.. stops the black king coming that White has a slight plus.ixc4 c5 8 d5 e6 9 great vulnerability of the isolated pawn dxe6 "ii¥xd1+ 10 �xd1 i.�f7. he explained. Instead. Petrosian rec ommended the move 13 .e2 dxc4 7 .l:. on the key square e4. 15 . after the game. that superior. thanks to to f7 and also prepares to settle later his superior pawn formation.e8 ••• 211 . how ever. Black's slight imprecision.d 1 . as Petrosian pointed out. the main advantage it fxe6 confers on White is the excellent out posts on es and/or e4. 12 �e2 tt:Jc6 13 . he would win it. but that was not the case here. Petrosian himself. M u lti-Piece En dings 1 C4 g6 2 d4 t2Jf6 3 t2Jc3 d5 4 t2Jf3 i.l::i. where it has a good square at e7.ie2 system in the opening.xe6 on e6. he had merely told his second that the prospect of this endgame position arising should not 15 tLlg5! deter him from choosing the quiet 5 e3 Immediately taking advantage of and 6 . Petrosian had predicted to his second that if this position arose. 14 .l::ta d8?! A small inaccuracy.. We will see in the fur ther course of the game that Petrosian makes excellent use of these squares.g7 5 ity does not consist so much in any e3 o-o 6 i.xe6 11 i. during his pre-match prepa ration. the bishop will just be control welcomed by the bishop on g7.e7 16. 18 l:tb1 The bishop on g7 has a fine diago nal..d2 19 a4 would probably have trans posed to the game. es would not be pieces. I suspect that he did not notice the computer's clever defence 23 .i. and will not be 16 lLlge4 as effective as it looks.:tc7.. and causes White some short-term inconvenience..lLlds Which temptation Black wisely re sists. After 19 .. How would have the option of retreating to ever. once White plays 23 b4! c4 212 . a very slight inaccuracy.i.. unpinning the c-pawn. ling an empty diagonal.i.xb4 cxb4 22 l:l. ltJxa2? 20 l:ta1 ltJb4 21 . argued that after 23 lLlgs Black's posi 17 lLlxe4 b6 tion would not be easy. 19.xf6 18 lLle4 the bishop 22 .i..h6 (intending 24 lLlxe6?! l:. 20 a4 l:tc8 21 b3 .The G r e a t e s t E v e r Ch ess E n dg a m es This is more or less obligatory. because of the pressure against b2. Pet mentators almost unanimously con rosian recommended 16 . ...i. . and e7.f8 22 . the com Again. b6..l:Ic1 .xb2 23 l:tb7 White is clearly better.ltJb4 19 ... 18 .. However.ltJxe4 In view of what follows. be b2-b3 and untangles his queenside cause the move 1S ..xa7 . so that demned this move and recommended after 17 lLlxf6+ ....i. Petrosian was not convinced.e7 or 24 e4 lLlb4!) but even there White can retain an edge after 24 h4.i. but Petrosian pre fers to tempt his opponent with the a pawn. .. after which the c4-pawn will prove in defensible. before was in assessing the position after the he played his 23rd move. Rather than pursue any of the 28 :c2 sharper alternatives.. 25 .i.e5 3 1 'iti>xc3 e6 for his king. M u l ti-Piece En dings Practically forced. e5.xa3 tL!xc3 30 l:txa7! ..xb4 . b) 24 . First.. Botvinnik prefers a quiet life. an enormous part in endgame play.c31 The start of a strong regrouping plan. After 29 lL!d2 c3 30 �4 213 . This is an text.a3 26 . Here.i. leaving advantage... too. the 24 b5 accurate calculation of variations is also a sine qua non of being a good endgame player. the bishop block ades the c4-pawn. .d6 �xa4 30 f3 and White has a small ad vantage.c2 lL!xc3+ 27 .l:tc2 c3 26 . Jk4 26 Wd3 �xa4 27 . The All of these variations had to be cal real difficulty of Petrosian's 23rd move culated and assessed by White.a3 (yet another idea is 25 ..l:txe4 29 i.b4 . The passed c-pawn looks quite other example of the importance of dangerous. and Black's short-term ini calculation in endgames. White retains some really allow the exchange on c5..i::txc3 . which Petrosian had also seen before move 23. but it soon becomes clear that he has underestimated the dan gers of his position.ta3 25 ..e5 28 . Whilst sche tiative needs to be accurately calcu matic thinking and manoeuvring play lated.. driving the knight away from d5. Black with a weak isolated c-pawn..l:.Wf7?1 2S i.xc3 i.i.l:txe4! 29 stubbornly with 28 . since he cannot �xb5.<.xc3 i. c3 25 . 24.b4 27 'iti>d2 :c4 28 i.t>e7? again with a small edge for White) 26 Black could have resisted more . Then White intends to follow up with lL!d2. in order to free l:.i.i.l:f.•.l:.c2 l:k4 27 Wd3 :xa4 28 l:ta2 . Petrosian suggests several ways for Black to play more ac tively: a) 24. g2-g3 and e3- e4. . The last hope of serious resistance was the rook ending after 33 . i. 39 ..d2 33 Wb3 l:txc2 34 'li>xc2 �d5 33 . Petrosian's line suggests that White should not waste a tempo on the move 31 f3..�d6?! White simply creates a passed pawn on 31 �c3 �c5 32 .l:tc4) 32 'li>d3 l:td8+ 33 'it>c4 l:.l:td1 offers better chances....d8+ 32 'it>c4 . I therefore looked at the immediate 31 �d3..l:.xh2 35 �d4! Suddenly..dl avoiding exchanges: he proposed 33 l2Jxc3 31 .g.... and the king and pawn could still have obtained good drawing ending reached after 37 ..d2 34 �b3 . e5. in Petrosian's line.. Petrosian suggested that move 33. 34 lDe4! Much more effective than 34 h3.. Black could still continue to resist by 30 l2Je4 i. but I still doubt that 36 g3 214 . 29 l2Jd2 C3 e.as 31 �d3 l:!dS+ 32 �c4 J:...h6 32 Wd3 �d8+ 33 �c4 king is so active. 34. so Black seems to have nothing better than 31 .. White has more chances than c7-square. This does indeed seem satisfactory for Black.d3.. 'li>c4 38 lLld2+ chances with 28 . i.The G r e a t e s t Ever C h e s s E n dg a m e s �e6 31 f3. as seen at the end of Petrosian's line. since Black's ning after 31 .i.xd2 30 'it>xd2 l1d8+ (30.l:td2 is worse) 31 �c3 the kingside with f3-f4.. Now c3 is hanging. Tal claimed White was win h e i s actually winning.e 7. all of White remaining pieces are beautifully coordinated and centralized.i..h1? 35 l2Jxc3+ 'it>c4 36 l2Je4 i.l:.x c3 34 �XC3. as also at However...J:i.. whilst Black's scattered units make an abject impression...'iii>b4 40 �d3 �xa4 41 �c4 and 29 .a5 (threatening 32 ... which also weakens the e3-pawn..xd2 39 �xd2 is winning for White..l:td2 34 'li>b3 l1xc2 3 5 �xc2 Wd5 36 �d3 The conclusion is that Botvinnik c2 37 �xc2..l:!.. 3s Wd7 ••• He has to stop the rook reaching the Here. .l:. ....f7 l:i. 39. 43 lld6+ �c8 44 'iiii>e 8 i. from beautifully elegant game by Petrosian. etc.e 7+ 39 •. 215 . where it proceeds to wreak havoc..i:!. gS. Realizing that the bishop is ide Black's position is collapsing... and the most ac curate.U...l:.es 40 l:.d7+ 48 <i.b4 37 �es .. f6.d 8 'it>g7 Or 42 .g8 1-o The white king has marched right into the heart of the enemy camp.xa4 44 li:Jf6+ 'it>d6 45 lidS+ rJilc7 46 . A ally placed for the defence on e7. 46 lt:lgs l:td8+ 47 <J.c7 45 l:tc6 :d1 4S ..i.Uhs+ 38 'iiifi> 6 i. 41.c6 l:!.a1 42 �e6 i.. M u lti-Piece Endings Petrosian chooses the most reliable where it defends key squares such as path to victory.h1 41 'it>f71 The sealed move. pares to drive it away with l:te6.xes .l:. Petrosian simply pre some counterplay.b4 43 . 36 i...xa4 46 lt:lgs wins a piece. 36 g4 hs would allow d6.l:.d4 and li:Jds is curtains.. part of White's plan.ie6 15 i. pieces on the board..fe8 17 i. and the d4-square 2. and a key the isolated dS-pawn is a passed pawn. White will have the reply b4-b5.xe7 18 b4! ••• White has the advantage of the bet ter pawn structure. play on the latter's part.xe7 . It is true that Another excellent move. In this . outside passed pawn. 16 l:. It secures control of cs for the is controlled by White. and assures White a superb outpost on cS for his knight. It fixes Black's weak pawn on a6. which potentially can yield an as.e 7 bishops is clearly in White's favour. The other enemy weaknesses. the exchange of dark-squared 9 cxd5 cxd5 10 exd5 exd5 11 l2Jc3 i. The point here is important aspect of the pawn struc that if Black ever tries to eliminate his ture is that White has a queenside ma a6-weakness by pushing the pawn to jority.e1 'ii'xa4 14 l2Jxa4 it leaves Black with a relatively bad . The move has which is usually an asset in the end three points: game. It sets White's majority in motion.. With inaccurate white knight.T h e G r e a t e s t E v e r Ch ess E n dg a m es 1 e4 C5 2 l2Jf3 e6 3 d4 cxd4 4 l2Jxd4 a6 5 Much stronger than 16 lDcs.id3 l2Jc6 6 l2Jxc6 bxc6 7 o-o d5 8 c4 l2Jf6 position. as 12 'fia4+ 'ii'd 7 13 l:. 16 i. could become a real asset for Black. establishing a strong passed pawn. the dS-pawn 3.c8 20 f3 ••• 216 . but We have spoken many times in this objectively it must at this stage be re book about the importance of fixing garded as more of a liability.c5! 18 'iW8 19 l2Jc5 .e 3 o-o bishop.l:. but here there are still many 1.i.. Black having three pawn islands against two. for Black's nominally "bad" bishop. and one which stunned all the watching GMs... It is the former that matters. to which the exchange. not at what comes off. great players like Fischer are al 21.. ever.•• rook will take control of the open c-file. White voluntarily gives up his great-looking knight on cs. its merits. and also bishop will be much more powerful on lets the white king approach the cen the open board than Black's rooks plus tre. The king clearly has his eye on the knight. and are not hidebound by preconceptions. hence Black's next.. The real difficulty of the 21 �es move lies in the fact that it would not Putting further pressure on the normally enter most players' heads to black position.l:i..d7 ways flexible in their thinking. as well as making possible the an apparently inferior bishop. whilst his queen's 20 l'..l:tc6 is a threat. J:td6 22 lZ'lxd7+11 One of the finest positional ex change decisions ever seen in top-level chess. after this the squares e4 and g4.xd7 23 l:tc1 Now 24 . it is not so hard to see white knight on cs. by attacking the pawn exchange such a good knight for such on dS. M u lti. How doubling of rooks on the e-file.i. seeking to eliminate the powerful change 22 lZ'lxd7. the key point is that what matters when exchanging pieces is what stays 24 l:tc7 217 . 2 3 . not what comes off nates" the black knight by controlling Fischer understands that.look at what stays on the board.:tea7? .. which "domi on the board. the weakness on a6. just as in Petrosian-Bannik below. 22 . His bishop continues to attack key central blockading square d4.Piece En dings A useful little move. his combined rooks plus knight might otherwise jump. 20. For lesser players. I can recommend the above principle ..l2ld7 looks like a much better Once one sees the results of the ex try. How ever. l::t..tt'ld5.. is helpless against the threat of 35 ..e 8 loses the a6-pawn after 2 5 bound.T h e G re a t e s t Ever Ch ess E n dg a m es Black has defended the sixth rank.e2 g6 It is hard to find a sensible move for 25 �f5 was a threat. Fischer simply With the white rooks doubled on strengthens his position further. 26 'iW2 As calm as you like. bring the seventh. 31 . 2 8. but he hence Black's next move. 25 .. 24 tt'ld7 . Black.. 30 '1t>d2 tt'lb6 25 i:. 218 ..:i. but no immediate concrete 31 �ee7 tactical win available.fs 2 9 �e3 d4+ but now the enemy rook invades the This allows the white bishop to seventh instead.. l::tx e8+ 'it>xe8 26 l:la7. ford to allow the white king to d4.. transfer to c4.h7. ing his king into the game.tLlds 32 l::tf7+ �e8 33 l::t b 7 tt'lxb4 34 26. Black has even won a pawn. when he would be practically move 24. but Black could not af so Black must take action against this. but White has a move first. With Black tied in knots. Now at least he has a threat of starting counterplay by 31 . hs 21 f4 h4 28 �f3 �C4 1-0 Now 29 �g4 is on the agenda. .•.:ee7 is threatened. the end is near.. �c7 15 'Lle4 'Llb6 16 i. obstructed by its considered that White has few.xd8 19 �dl l:txdl+ 20 <io>xdl h6 ble". knight to fs and king to e4. if any. but in are very good defenders of weak fact. and so aim to drops his bishop back to f8 and marks reach an ending with knight versus the time with his king.xc8 l:raxc8 17 b3 l:tcd8 White has achieved his "dream" po sition. Black sticks his king on The most obvious way for White to e6 and waits. of course.U. Petrosian actually ex chances of winning from here. With Black having what is in for this? The reason is because he clearly a bad bishop.Piece En din g s e**� . 219 . M u lti. Black just and bishop for knight. very difficult. the changes it off! It might seem as though main problem being that bad bishops he has taken leave of his senses. Even if White gets his play this position is to exchange rooks..g5 i.xb6+ <io>xb6 22 g4.i. whilst White has no way to deep decision.i. 1\Vxd8+ �xd8 9 . In Petrosian's own "bad bishop".h3 21 i.e 3 c6 13 h5 g5 14 ... if not impossi l:txd8 .. own pawns.e7 10 0-0-0+ 'Lld7 but we will assume he "comes quietly") 11 h4 f6 12 i. tJSSt Cnampitirtsfil 1 c4 e5 2 'Llc3 'Llc6 3 'Llf3 'Llf6 4 g3 d6 5 (Black could of course move his knight d4 g6 6 dxe5 'Llxe5 1 'Llxe5 dxe5 8 to c8. it is an exceptionally pawns. Black has the bad bishop. and 18 i. .. Let us assume that Black words.· :� T�Pefr �.c 5! White's pieces have great squares on This move is the single most striking e4 and fs. break through. .. "winning such a position would cooperates fully in allowing this: 18 be very. So why did Petrosian not go of the game. to avoid the further exchange. Petrosian chooses another 26 c5?1 ••. Even so. 34 l:tc7 35 'it>e4 l:. and find the most accurate plan. as a way of 220 .c6 29 l:td5 tbf8 30 21 tbe4 l:te6 22 g4 tbg3 tbe6 31 tbf5+ �e8 32 e3 tbc7 33 . 34 �d3 Now the king takes up position on e4. Recognizing that Black's This is a big concession. Black wants to fight for space on tive defender of his weaknesses on f6 the queenside. 23 tbd7 24 'iii>c 2 b6 25 l:!. so as to in square for his rook. it was probably chiefly motivated by gaining time on the clock and getting closer to the adjournment at move 40.. 18 !Ixd1+ 19 l:txd1 i. price he pays is very high. weakening formally bad bishop is actually an effec dS. 22 a5 23 l:td3 . and can only wait.f3 'iii>d 8 26 a3 •.c6 36 tbd6+ �e7 37 ••• tbf5+ �e8 38 tbd6+ �e7 39 tbf5+ �e8 This repetition is a standard part of endgame technique.. after . such repetitions are also very handy psychologically.xc5 20 tbxc5 �e8 . permanently threatening to pene trate via the weakened dS-square. plan entirely. Already.Th e G re a t e s t Ever Ch ess E n dg a m es Instead.. but the crease the vulnerability of these targets. �e7. and also to free a and h6.•• Preparing to increase the pressure against f6 by means of l:tf3. the black rook has had to take up a very awk ward post on e6.•• 27 'it>c3 'it>e7 28 !Id3 l:i. we see the difference. he exchanges it.:d1 tbe6 Black is totally passive. In this instance. when White would be able to analyse the position in the peace and quiet of home. In order to defend the f6-pawn. M u lti-Piece En dings showing the defender who is boss! into fS. of economiz 52 'it>e7 53 lt'le4 c.d5 into trouble with the premature raid Now Black is again in zugzwang and 43 �fs �7 44 lt'lh6+ �g7 and the cannot avoid material loss.. Now the e-pawn decides.f7 52 e4 h6 53 f3! .'1W7 44 lt'lh6+ �es 45 lt'lf5 lt'le6 50.lt'lb4 60 e5 lt'ld3 61 e6 1-0 Although the white rook looks more A great technical display by Petro active than its opposite number..l:!. 43 . pawns on f6 and h6. 46 �d61 59 .lt'lf8 49 'it>fs �e7 so . and this in tum 46.f5 51 <3.. 221 . try to make progress. Most of black position. whereas the black rook Black's subsequent problems stemmed is an important defender of the poten from the vulnerability of his weak tial penetration squares.i.>f5 g3 55 fxg3 .i.. much the same purpose...• shows his hand.>fs. If he allows the king dark-squared bishop. Now Petrosian 57 lt'lc7+ 58 'it>d7 lt'la6 59 e4 .. white knight is trapped... The exchanging decision on move not have any way of penetrating the 18 is especially noteworthy.d21 lt'lc3 lt'ld7 Sl lt'ldS+ <J.>f7 54 c... it does sian...i.l:txd6 47 lt'lxd6+ 'it>d7 48 lt'lb5 lt'lg7 was a direct result of losing his "bad" Stopping c.• ing time on the clock and keeping the g4 56 lt'lg5+ 'it>g8 57 'it>e6 defender guessing as to how White will The pawn on g4 is irrelevant.. then he finds himself in 40 a4 lt'ld8 41 lt'lh6 lt'le6 42 lt'lg8 lt:Jf8 43 zugzwang after 48. Do not hurry! He only runs himself 4 9 h6 lt'les 5 0 <3.xe5 fxg4 52 lt::lc3 Keeping the white king in its domi The last few moves have served nating central position. Black cannot put his king on e7 at once. and the possibility of tying White down some what to the defence of the a3-pawn. 21 . White is somewhat bet An excellent choice. 222 . Overall.. Most players ter. it could almost be called a queenless middlegame.xc3+ 14 l:txc3 21 . Carlsen.xf6 any of the great endgame masters 'ii'xf6 9 'ii'xf6 gxf6 10 lDf3 lDd7 11 lDh4 could have done.. is poten tially bad. The principal feature which grabs one's eye is the weakened black kingside pawns. i. indeed.tDbs 22 l:. because the black kingside pawns would have automatically plonked the are the most serious weakness on the knight on the glorious-looking outpost board. it does aged just 14 when this game was not actually attack Black's weaknesses. This is quite a complex endgame. handles the position as well as s i. because the black queenside pawns are all on light squares.c4 the enemy bishop. Black has well-placed pieces. given the number of pieces on the board.fs 7 'i!Vf3 i.. despite being cally impressive on that square..c1 lDb6 13 a3 i.l::i.. but it requires great technique on fs. 2 5th and 26th moves. but with the white knight pointing at fs.Th e G re a te s t Ever Ch ess E n dg a m es 1 d4 ds 2 c4 e6 3 l2Jc3 lDf6 4 cxds exds played.ih3 lbc8 15 f3 lDd6 16 g3 as 17 b3 a4 18 b4 White wants to avoid exchanging i.a 2 20 :c1 i. which. although apparently actively placed. The pawn on e3 could also come under 23 lDg2! attack. but although the steed is opti to show this.gs c6 6 e3 i.b4 12 . even in their prime. castling turns out to be a loss of tempi.g6 8 i.a1 o-o In view of Black's 24th.b1 19 �2 i. that by forcing the vulnerable after the opening of the b black king to occupy e7. c6-cS.c 21 In freeing the fS-square for the knight. 30 i.cs Black makes his break. shielded his e3-pawn from pressure 31 bxcs bxcs 32 dxcs l:!.. so Carlsen prepares to redeploy his knight to fs. too.c8 down the e-file. his knight is on its way to hS... 29. Note. 27 b6 ••• Black prepares the break . which opens some lines and gains his pieces a little more freedom.. M u lti-Piece En dings I remember John Nunn once telling me that blockading isolated pawns was rather overrated. in an attempt to loosen White's grip. which is not easily defended. White's pieces can also exploit the newly opened lines.. However.fs h6 26 l2Jhs �e7 21 g41 and Black now has further weaknesses Now the black kingside weaknesses at dS and a4 (the latter becoming more are fixed. to defend the f6-pawn. 25 ii.. 33 lLlfs 223 .�8 Anticipating a later lZ:lfs+. 23 !Ifd8 24 lLlf4 �f8 ••• The king heads back to e7. the bishop also takes aim at the a4-pawn. from where it will attack the f6-pawn.a7 29 lLlg31 The knight has done its job on hs. 30 . and that winning them was better! Magnus clearly agrees. it is a bigger target than the f6- pawn. 28 :he1 .. tying Black's rook to its de fence. Now that the black pawn has advanced to h6.l:!. White has file). but n Jlxcs . to exploit an other freshly opened file. The h6-pawn is already quently happens in such cases.c8 3 5 .•• l:. and so forcing the following exchange. It is rather nice how each time Black opens a file in this game. 34 1:teb1 See note to Black's 30th.xa3 Black has regained one of his pawns.l::[b 7! ••• Usually. but now Carlsen switches his rook to the other flank. Carlsen's tactical control is admirable.c2 technical problems in realizing his rna- 224 . in order to activate his bishop. 34 l:. The white rook makes use of the freshly opened b file. we see that "endgames always become concrete". and will attack the a4-pawn from b4.ca8 44 .txfs l:.i.Ub4 d4 .T h e G re a t e s t Ever C h e s s E n dg a m es The knight lands on fS with much White has netted two extra pawns. obtain some temporary activity. the There is no value in passive defence. As fre the case earlier.txa4 li'lc3 38 . that could potentially cause White 36 exd4 .•• 42 l::t h l! Threatening no more and no less than mate in one. but here so Azmaiparashvili jettisons his ds that would be inappropriate. best advice is to trade pieces. 38 hs 39 h3 hxg4 40 hxg4 li'la2 41 . greater effect than would have been and is objectively winning. One thing pawn.. Once more.te6 37 . White al ding material enables the defender to ready has bigger fish to fry. Carlsen immediately exploits it for his own rook! 42 -txfs 43 .Ub2 . but as it turns out. shed doomed. when material ahead. Instead. i. by Carlsen. Such a factor favours the knight (see Torre-Jakobsen).xa8.hxf7 A wonderfully mature performance The rest is easy. l:tg7+ �h8 S1 l:th7+ �g8 52 l:thg7+ �h8 44 lt:Jc3 45 l:th8+ �g7 46 l:f.l:.e7 45 l:.e. He threatens to double rooks on the sev 47 . 225 . M u lti. with a line such as 44 l:th8+ <i. and win more material.h7+ �g8 47 ••• 53 l:th7+ �gs 54 ds lle3+ ss �d4 1-o l:r.lita2+ 48 �e3 l:te8+ 49 �d3 lt:Jd1 so ••• enth.Piece En dings terial advantage is the "narrowness" of the position. and so White should be a little careful about relying on routine simplification and the promotion of the d-pawn. the fact that the pawns are all concentrated within three files.xa8 . and the f7-pawn is doomed. Carlsen uses the greater activity of his pieces to attack. . tt:lds.. The ing his pieces down to defence.a6.fc1! 19 . and ty 'itf1. attacking moves to free fl. but Karpov understands 12 'iixc3 'i¥xc3 13 bxc3 tt:ld7 14 c6 bxc6 that Black may start counterplay by 15 l:.. but in fact. The rook rook is now very active on as. ..i. when the knight can attack the cs-pawn from d3. u*�spiro{ ��� � � � 'Wbrld�tna""ionship(Game �71.gs h6 6 i. M�c:ow 1984 �� � 1 tt:lf3 dS 2 d4 tt:lf6 3 c4 e6 4 tt:lc3 i. Even so. White has a small plus: he can attack the cs-pawn in short order and assume some initiative as a result.. i.The G r e a t e s t Ever C h e s s E n dg a m es " � �'?' ' Gami¥1... so he overprotects the pawn. This should result in his pieces being a little more active than Black's.. but then 19 tt:les is un pleasant. but at the mo White has achieved his first aim..e7 5 question then i s where should the rook i.xa2? 20 c4! the enemy bishop is An extremely subtle move.bs! tt:ld7 class opponent.xc3 the open file. 226 .9' .i..b1 1B 'it>f1 ..xc4 'i¥as 11 o-o i.l:. 11 .xf6 .xf6 7 e3 o-o s 'ti'c2 cs move to? Most players would choose 9 dxcs dxc4 10 i. this runs into . in anticipation. i. tt:la4 or . which lost! would not enter many players' heads.e2 cs attacking the c3-pawn with . to defend the bishop on e2. against a world 19 . � �" '�"" �? � � ��latpov � � .i.ab1 lbb6 16 .i. and without the aid of Now we see another reason why a crude mistake from the defender. zo ..after 17 J::. His ment.l:tas Karpov intends l:!bs... so that White can play both of Black's isolated pawns.i. it is desperately hard to imagine White winning the position.c6 would have prevented White's next.ds 18 .. Karpov put his rook on cl . This position looks extremely equal. g5-g4. Another very typical Karpov move.. If 22. The c5-pawn is not going anywhere..c6 .i.xc5 Finally.l:Ifb8 21 c4 .l:.�8 27 lt:lxcs lt:lxcs 28 lixcs the tactic 25 lt:lxc5?? lt:lxc5 26 ... 26 lt:lxc5 lt:lxc5 27 . then 26 .xg2+. c5-pawn for the moment. M u lti-Piece Endings 20. dinates his pieces is beyond all praise. The knight heads for d3.. whilst at the same time defeating the 26 . White has won the c5-pawn. on b2 or (as we will see) b4. This is the true meaning of prophylaxis . 2S lt:ld3 gS Black seeks counterplay.d8 If he tries to defend the c-pawn with 24 ..l:tb4 23 . 227 .l:!. despite the piece deficit. before helping himself to the 24 lt:ld3 lk7 indirectly defends the first fruits of his superb play.a4 wins it anyway.l:i...b7 24 f3 the b-file.it is about more than simply taking defen sive measures against the enemy's threats.l:ta4.i. murky waters. from where it attacks � weakness on c5 and at the same time �ents the black rook tak ing up an active post in the white posi tion. because of 26.i. exchanging off White's active :dd2 with surprisingly strong coun rook Now nothing can stop the tempo terplay.xc5 Black has the Karpov anticipates the threat of remarkable resource 27 .. whilst at the same time lim iting the scope of the enemy bishop and freeing his own king from the de fence of g2..d1! he takes the pawn at once. 24. .. Karpov calmly seals up 23 .. and instead involves finding ways to improve one's own position. to attack White's kingside The manner in which Karpov coor structure and open the g-file.If. then after Another prophylactic manoeuvre.i...xc6 23 .. Rather rarily active black rook from being than allow his opponent to fish in such driven back by lt:ld3.l:!.b3! enemy's plans.c7 2 5 lt:ld3 .. Karpov calmly eliminates this possibility.b2! ? 28 .i..i.l:....b7.l:!. by prepar 22 lt:le1! ing . e 3 that an easier win was to be had by ere- 228 ... advantage of the weakening of the en 43 l:Ig7 emy kingside pawns.e7 41 fl. Exchanging pieces is generally a 37 . 34 e4!? Instead. exploit the temporary misplacing of On the other hand. 30.xh3 l:tf3+ gives Black a passed pawn on the e-file.. It also Indeed. This move in One might have expected Karpov to volves exchanging two pairs of pawns. The game way adjourned around this point.e4 43 �dl. the black h-pawn will be vul have lost coordination.Jid6 29 �e2 <l.l:...:!.l:lf8 44 . so as to eliminate which in principle is not necessarily to the counterplay this pawn might offer.. after 43 . especially l:. 42 g3 l:. go after the h-pawn.. an �the remaining moves were influenced by overnight analysis.fxe4 3 5 fxe4 �xe4 36 htxgs �fs 37 analysis had presumably convinced him <i.... at the end of the White's rook. to invade via f2 or fl.Th e G re a t e s t Ever C h es s E n dg a m es without allowing any active counter Panchenko suggests that White play.e8 The passed pawn can be blockaded securely enough after 42 .h4 .. However. and his king is nerable.e 7 30 l::td 1 g4.g4 h3 when there is still plenty of material on the board. White's pieces process. and Karpov's adjournment 34..b 2 it is not obvious how Black can which may be a source of counterplay.J::txd1 31 �xd1 �d6 32 .. 45 <i. h4 38 <i. The next stage is to consolidate should fix the h s-pawn with 37 h4 but and start on the process of converting Karpov was presumably reluctant to his advantage. cut off.. the stronger side's advantage.as fs 33 �e2 hS Black continues to seek active coun terplay on the kingside.l:.b 1 40 a3 good idea when material up.d4 eS+ 39 �c3 i. Kasparov transfers his rook to Karpov immediately tries to take the f-file. give the enemy bishop a secure post on 28 . �4 51 c7 l:!.xe4 The path of the c-pawn has been i.e6 continues the resis 49 .l:td7+ �e4 49 l:.xe4 60 .f5 when Black draws... and playing actively.. 56 �b7 . all of White's pieces are excel lently coordinated.l:. M u lti. .bs tance.i. 54 g41 Accuracy to the very end.h2 loses to 51 C7 hl'i¥ 52 ..l:....xe5.i.xc8 56 �xc8 i. White's rook.a4+ �ds 48 .h8 55 .c6+. which will give the 43 . c6 e... and the end is now 54. It is hard to pinpoint any obvious mis 53 i.1:i..xh2 46 white rook the h7-square again. 56. black unit on e4. After Kar pov's phenomenal play..l:. Threatening 5 2 . and the only really lost for him..Wf2 then simply 57 .xes �xg4 59 :e2 1-0 square h7 is no longer available to A really fantastic display by Karpov.l:td2+ �xg3 61 !:txh2 �xh2..Piece Endings ating another passed pawn of his own way to close its diagonal is to put a on the queenside.c 6! take by Black...g...h 1 e4! 58 a4 Wf2 59 i. 57 i.b2+ so �as l:!. it would have been an absolute tragedy to have fallen into the trap 54 �b7 l:th8 55 c8'ii'? ? 46 . 50..�c6 47 i.a 2 56 l:te1+! not far away..c8 52 �b6 We3 If 56 ...l:txa7 .l:!.JU8 44 . cleared of obstacles. C5+ 53 . supporting the passed c-pawn.d1 i. 229 . since the 57 l:!. yet it is impossible to be Now the bishop takes over the task lieve that the starting position was of covering the h-pawn.l:...h2 Now we can see the difference from the variation given in the last note.f2 45 �b4 . Here.e4+ �g3 58 l:i..h2. ....The G r e a t e s t E ve r Ch ess E n dg a m es 1 e4 c6 2 d4 d5 3 es i.g3 'Dxf2 is met by 21 title. This was the 14th and last kingside with 19 gs.id3 'Dbc6 13 i.:i. hs. so he too needed to win this 19.d4 'Dxd4 6 cxd4 �xd2+ 17 �xd2 White with the traditional knight ver sus bad bishop.e3 'iVa5 14 �d2 'bg6 15 fear the exchange on d3 .f1 i s perhaps the reflex move. However.xc5 9 but Kramnik did not like the reply 'bxcs 'iWa5+ 10 c3 'ii'xcs 11 'Df3 'De7 12 18 . he saw no reason to . kingside gives more dynamism to the 17 .d7 6 'bd2 c5 7 dxc5 e6 8 'Db3 i.it leaves . h5 This game is something of a parallel to Botvinnik-Bronstein (Game 23).fs 4 h4 h6 5 g4 1 8 i... i. gxhs when Black can hardly defend the White has a small. the opening of lines on the him down. the Black posi Structurally.c6 game to save his world championship 19 .. too..i. White has acquired a tion is very solid and has no real weak weakness on the h-file..i. 1s. Kramnik's move game (barely over half the number activates the rook on the potentially played in 1951!)..'Dh3 20 . vantage. so it will not be easy to break other hand. Besides. but definite ad g-pawn.'Df4 18 l:tac1 position... in the decisive game of the There is no value in blocking up the match. 230 . by a point. but on the nesses. thanks to more space and the 20 gxh5 'bxh5 better bishop. and Kramnik trailed open g-file. the world championship was 19 l:thg1! at stake. Here. . and 26 l:tb7 is down to the defence of f7 once his nothing after 26 . Annotating the game in New in Black finds it harder to cover f7 satis Chess 2004/8. and the key to success frequently involves exchang ing off the bad bishop. Kramnik gives. Black intends to challenge how in this line Black is so badly tied the rook with .l:tc7 llxa4 28 l:tb1 :xd4+ 29 obviously bad. but it appears that this may be the case. bad bishops are often effective defenders of weaknesses. so as to expose the weaknesses. 23 . 22 . as a sample this natural move may in fact be the variation. That is the case here. As we discussed in the context of Petrosian Bannik.d7 25 .i. the line 25 . 23 lZ:Igs This is the difference. 22 . with a mating attack.xa4 23 . The point of White's play is to an swer 24.:e7 b6 when he could find no seventh...i.l:thc8.hb8.. . Kramnik suggested that factorily.:c7 is Wxe8 27 .xb5!.e8 24 bS ••• 21 b4 a6 22 a4! White starts selZlng space on the other wing.i. Note direct win..'�d8?! Once the bishop on e8 is removed. but Kramnik recom 'it>e3 l:tc4 30 l:tcxb7 l:tc3+ 31 'it>d2 l:tc8 32 mended 22 .. It is strange bishop is gone... classic two-front strategy.'lt>e7! 23 b5 axb5 24 axb5 l:!..l:tf8 26 .l:.i.. Now there is no way to prevent a 231 ..xe8 decisive mistake.. axb5 with 25 .a7 and White will double rooks on the ... that allowing the rook into c7 should 24 lLif4 25 b6! ••• prove to be better than keeping it out..i. The f7-square is now vulnerable. the white king can 2S ..U..:tg1 34.. nik..l:.l:Ih4 41 li'lf7+ 1-0 White can prepare a breakthrough A fine technical squeeze by Kram with f4-f5 at his leisure.Th e G re a t e s t Ever Ch ess E n dg a m es rook invasion at c7.g1+ 39 �h s.. starts checking. but the other will replace it.a3 33 llxg7 tension. 35 Wf3 .t'd7 allows the tactic 31 li:Jxb7! after which the white b-pawn will prove unstoppable.i.c6 29 li'lxf7 �xh4 30 li:Jd6+ Wd8 30. 31 .... 232 ..c1+ .l:.l:ta3+ 36 Wg4 .C..U.li:Jxd3 26 'it>xd3 .llh3+ 38 ..g.l:.h7 will in 37 fs .g4 evitably lose. via f4-gS-f6 etc.a2+ 34.l:Ixb7 l:. played in conditions of enormous 32 �e2 l:.g6 and 40 �c7 . e.xc8+ �xc8 28 l:r...l:Ic8 march into the heart of Black's posi Black eliminates one enemy rook. 33 .xb7 31..U..lla1 leads to a lost rook and pawn endgame after 37 li:Jxb7+ .. tion.xd4 35 fs exfs 36 e6 threatens mate in one.l:Ixa4 34 f4! 27 ..i. 32 as 'it>e7 33 .d3 36 . Passive defence by 31 .. and Black's posi Kramnik pointed out that i f Black tion is objectively lost.l:.l:..xd4+ 38 �gs exfs 39 �6 . New York 1924 131 Carlsen-Azmaiparashvili. Tilburg 1981 141 Andersson-Stean. Varna Olym piad 1962 207 Botvinnik-Bronstein. Vancouver 1971 68 Flohr-Botvinnik. London 1883 202 Fischer-Petrosian. St Petersburg 1914 190 Andersson-Miles. Varna Olympiad 1962 145 Botvinnik-Minev. Buenos Aires 1971 216 Fischer-Taimanov. Odessa 1960 43 Barcza-Soos. World Cham pionship. Moscow 193 5 86 Flohr-Vidmar. Amsterdam Olympiad 1954 158 Botvinnik-Smyslov.Alatortsev-Chekhover. Swiss Ch'ship. FIDE World Cup. World Ch'ship. Moscow 1951 107 Botvinnik-Fischer. Candidates Match. Nottingham 1936 137 Henneberger-Nimzowitsch. Buenos Aires 1927 113 Alekhine-Lasker. Budapest 1929 18 Capablanca-Tartakower. Candidates Match. St Petersburg 1909 14 Englisch-Steinitz. Moscow 1957 97 Brinckma nn-Rubinstein. Khanty Mansiysk 2005 222 Cohn-Rubinstein. World Championship. USSR Championship. Sao Paulo 1979 197 Averba kh-Furman. Moscow/Leningrad 1933 101 Flohr-Capablanca. USSR Championship. Winterthur 1931 79 233 . Tbilisi 1937 118 Alekhine-Ca pablanca. World Cha mpionship. Moscow 1984 226 Karpov-Kasparov. Moscow 1984 90 Korchnoi-Ka rpov. Ventnor City 1942 40 Znosko Borovsky-Seitz. Moscow 195 5 52 Kra m n i k-Leko. Chigorin Memorial. Pa rnu 1947 21 Reti-Rubinstein. USSR Championship. Ostend 1905 1 63 Petrosian-Ba nnik. World Championship. Hastings 1895 29 Plater-Botvinnik. World Championship. Skopj e 1976 56 La rsen-Polugaevsky. Zurich 1934 31 Marshall-Lasker. USA 1907 193 Marshaii-Ma roczy. World Championship. USSR Championship. Moscow 2007 35 Kuraj ica-Karpov. St Petersburg 190� 126 Tal-Keres. Germany/Austria 1910 1 73 Schlechter-Rubinstein. Baguio City 1978 151 Kotov-Botvinnik. Tal Memorial. I nterzonal. Candidates T'ment. Gothen burg 1920 62 Schlechter-Lasker. Moscow 1969 82 Spielmann-Rubinstein. Moscow 1963 211 Pillsbury-Burn. World Ch'ship. USSR Championship. Rio de Janeiro 1979 1 85 Torre-Jakobsen. Brissago 2004 230 Kramnik-Shirov. N ice 1931 25 234 . Moscow 1947 65 Randviir-Keres. British Championship. Vienna 1898 166 Pillsbury-G unsberg. Riga 1958 219 Petrosian-Botvinnik. World Championship. World Championship. Bled/Za �b/Belgrade 1959 180 Timma n-Velimirovic.Th e G re a te s t Ever Ch ess E n dg a m es Karpov-Kasparov. Amsterdam 1973 74 Webb-Miles. Le Havre 1966 46 Lasker-Ni mzowitsch. Morecambe 1975 1 69 Ya nofsky-Pin kus. San Sebastian 1912 124 Smyslov-Ta l.