ATTACK WITH MIKHAIL TAL MIKHAIL TAL & IAKOV DAMSIKY EVERYMAN CHESSContents Mikhail Tal - A Personal Tribute Introduction The Main Indicator — King in the Centre Breakthrough in the Centre The Assault Ratio Invasion Trajectories Lines of Communication Outposts Eliminating Defenders At the Royal Court Destroying the Fortress Walls Co oOI AA WWD Answers to: What Would You Have Played? Postscript Index of Players List of Illustrative Games vi vii 32 42 51 14 89 105 116 128 167 179 180 183Mikhail Tal: A Personal Tribute YOUNG PLAYERS of today may find it difficult to appreciate what the name of Mikhail Tal means to chess devet- ees of my generation. In the late 1950s I became passionately interested in the game and began subscribing to Chess Magazine, where I first read about Tal. In particular I remember his exploits in the 1958 USSR Championship, where he was striving to retain the title he had sensationally won the year before. Going into the last round level with Tigran Petrosian, Tal had Black against Boris Spassky, who himself needed to win at all costs, in order to progress to the next stage of the World Championship cycle. Petrosian pra- dently agreed an early draw, but after five hours' play Spassky and Tal were sti] locked in battle, and their game was adjourned in a position that looked grim for Tal. I can stil! vecall the words of Salo Flohr, as they appeared i : "Tal and all Riga slept badly . As it happens, the reader can follow this game in Chapter 8, and see how Tal defended heroically, before finally breaking out with a de- cisive counterattack to win his second USSR Championship gold medal. The next year, 1959, saw the thril- ling Candidates Toumament in Yugo- slavia, in which Tal and Keres fought out a ding-dong battle, Tal in par- ticular taking huge risks in striving to win in virtually every gare, Even the legendary Botvinnik was unable to resist his fiery play, but, not long after wresting the World Championship in 1960, Tal suffered the first bout of ill- health that was te dog him for the rest of his life. And after losing the Return Match the following year, he was never again to contest the title, al- though he reached the Candidates stage on several oceasions. With the passing years his playing style inevitably mellowed and became more solid, although in almost every event he could be relied on to provide at east one combinational flash re- calling the Tal of old. I vividly recall one occasion in 1970 at the Moscow University Chess Club, when Tal came to give a talk about the Tecent "Match of the Century" in Belgrade, He spoke modestly and wittily to the packed audience, and a wonderful evening was rounded off by a 12-player lightning toumament, into which the grandmaster threw himself wholeheartedly. As one of the leading foreign players at the University Club, I was invited to take part, but in my game with Tal was too over-awed, and after losing dismally in under 20 moves ] felt too ashamed to engage the great man in conversation... The pleasure I have gained from transtating this book, and which { hope. will be shared by its readers, is tinged by the regret, felt throughout the chess world, that the incomparable Mikhail ‘Tal is no longer with us. Ken Neat July 1994Introduction ON THE ONE HAND, this is not a textbook. Because in chess, in the opinion of the authors, this name can be given only to an explanation of the rules — the bishop moves diagonally, the rook vertically or horizontally in any direction — and also, perhaps, to the principles of checkmating the lone king with pieces of different values. These are indisputable — they are axioms. Everything else in chess consists of theorems, and they need to be demon- strated. Sometimes they drag on for many years, and sometimes stretch far into the distant future. Indeed: does the truth lie in a strong pawn centre, or the attack on it? Who will give a definitive answer? And when? On the other hand, this is neverthe- less a textbook. Similar to those that set out the basics of versification, the metres and peculiarities, say, of Alexandrine verse. It is true that, after studying all this, you will not neces- sarily become an Alexander Pushkin, a William Shakespeare, or a Johann Goethe — geniuses themselves make paths in the endless world in the name of creativity, but... Neither Pushkin, nor Shakespeare, nor Goethe would have become who they were, had their knowledge of poetry (feelings and thoughts we introduce into this conver- sation in brackets; they are primary and lay at the very foundation) re- mained at the level of prehistoric, primitive people, who made single-tine vii drawings on their cave walls of the desired objects of their hunting, but nothing more... And so, dear readers, what you have before you is a guide. Since it is fully admitted that this is a field in which there are many more exceptions than rules, this book has the aim merely of suggesting: where and on which paths in the dense forest of variations are concealed those very indicators — imperceptible or altogether invisible — that will allow you to replace quiet forward movement with a much more rapid tempo, or to switch from a posi- tional struggle to an attack. It is not always explosive, sacrificial — on the queenside things very often do not come to this. But if the target of the pursuit becomes the king, the sacrifice becomes both a basic weapon, and the main “performer” of that army which is mounting the offensive. This is why reference points and prompts are so important, in deciding whether or not it is the right time for a sacrifice. The theory of sacrifices has already been studied — true, not especially often, and not very recently — by chess researchers, including grand- masters Rudolf Spielman in the mid- thirties, and Leonid Shamkovich in the late sixties. The classification of sac- rifices compiled by them was correct, although also not altogether complete: life in chess does not stand still. But the aim of this book is different: it is to demonstrate the practical applicationVili Attack with Mikhail Tal of these sacrifices, to give typical attacking procedures, and —- most important — to point out those very "symptoms" of a position, which indicate the existence, at its heart, of a combinational, sacrificial solution to the problem. However — and the reader can see this for himself — the correctness of an attack on the kingside may sometimes depend on whether the white a~pawn is at a2, a3 or a4... So then, let's be off. The authors have endeavoured to differentiate the sections as far as possible, although, of course, in practice attacks only on h6 or exclusively on the light squares do not occur, But the main idea may be suggested by one particular feature of the position — and it is this that gives the name to this or that chapter. Bach chapter is accompanied by illustrative games: at the demand of one author GD.) they are taken from the "personal collection" of the other (M.T.). The basis for this selection was the record number of special prizes, the vast majority for “brilliance in attack", which overcame the feeble resistance of one author against the other, and Attack was victorious. May the readers of this book also be victorious.1 The Main Indicator - King in the Centre “IT IS A PROFOUND MISTAKE to imagine that the art of combination depends only on natural talent, and that it cannot be learned. Every experi- enced player knows that all (or almost ali) combinations arise from a recol- lection of familiar elements." LD. This thought of Richard Réti (with certain doubts regarding the numerical grounds of "all or almost all combinations") can be considered in- disputable. And the king in the centre is a stable "familiar element" of pos- sible combinational attacks. M.T. At any event I can admit: as long as my opponent has not yet cas- tled, on each move I seek a pretext for an offensive. Even when I realise that the king is not in danger. LD. Is such an expenditure of time and effort justified? MT. Well... It is more likely that doubts are unjustified, as shown by the following instructive episode. It oc- curred in the 4th game of my Can- didates Semi-Final Match with Bent Larsen in 1965. After 1 e4 46 2 5 405 3 d4 d6 4 4)f3 dxe5 (to some extent a Larsen patent) 5 4)xe5 there suddenly followed 5...45d7. Had this been a simultaneous dis- play, 1 would have decided that my opponent had simply overlooked 6 &)xf7 &xf7 7 wyh5+, when, against his will, the black king is forced to "go for a walk". But Larsen could not have over- looked this, and I began examining possible variations. My inwition insis- tently kept telling me that the sacrifice had to be correct, but I decided to calculate everything "as far as mate", spent some 50 minutes, but then in one of the innumerable variations I found something resembling a defence, and... rejected the sacrifice. This was a be- trayal of myself, I saved the game only by a miracle after the adjournment, and in general I barely, in a “war of nerves", won the match. Of course, I should have begun the attack without thinking, and only after the forced 7...8e6 8 c4 5/6 9 d5+ Gd6 10 WE7 He5 11 Af4 studied the position. This is no place to give var- jations: in certain of them the black king is mated not even on the central files, but at b2 or al! And so that this should not sound totally unreal, here, cleared of a whole network of bran- ches, is a continuation which, though not of course obligatory, is very pretty — 11...c5 12 4c3 a6 13 Hdl!! g6 14 SAxeS+ Hxe5 15 d6 g5 (defending against 16 f4) 16 Hd2 Qf5 17 He2+ Sd4 18 Hedd!! Axed 19 well, with the threat of 20 Qe2+ @d3 21 wyh3+ Se2 22 vyb3+ Sb! 23 Hc3+ Gal 24 Ad3 Qxd3 25 Gd2+ Abl 26 Axbl mate. And so, the attack on the king. It includes many components. If there is2 Attack with Mikhail Tal a lead in development, if the king's pawn shelter is weakened or files and diagonals have been opened, if a se- cure piece outpost or a striking force has been created — all these are suffi- cient grounds for an offensive, and a discussion of them will follow. But for the moment — the assault on the king that is still in the centre. These are the most natural kingside pawn formations, arising almost in- variably in many semi-open and closed openings. They can be broken up only by a direct blow — by a sacrifice at e6 or f7. The offensive then proceeds either along the opened central files, or else along either the h5-e8 diagonal (if it has been weakened beforehand, this often serves as a “prompt" for the attack) or the a2-g8 diagonal. However, usually the attack is mounted down both central files. Nezhmetdinov-Kamyshov Russian Federation Championship Yarostavl 1957 This can be considered a classic ex- ample: a preceding pawn sacrifice for the initiative, an energetic exploitation of a lead in development, and — an attack with limited forces! 17 &\xf7!! @xi7 17..Hg8 18 WhS Bg6 19 4yh8! is also bad. 18 wyhS+ Se7 19 exdS e5 20 £4 wxd5 21 fxe5 £5 22 e6 df6 23 h4! Preparing mate at f7 or gS. 23...0.05+ 24 Qh wxeb 25 Wh6+ Black resigns. Rudolf Spielmann called such a sac- rifice "preventive" — after it the king does indeed lose the opportunity of ending up in its customary and therefore safest place, and is "invited" to face the full force of the attack. Most often this draught threatens not just a cold, but terminal pneumonia...King in the Centre 3 Kupreichik-Grigorian Leningrad 1974 By a temporary pawn sacrifice Black appears to have gained good counterplay on the g-file, and is intend- ing, at the least, after 17 4)f3 2\xh4 18 Wxh4 b5 (there is no point in hurrying to restore material equality; the threat of ...b4 forces White to waste a tempo) 19 a3 4b7 to complete his develop- ment, and, more important, to create the preconditions for a counterattack, in particular on e4. All this would be right, were it not for the fact that the black king is still in the centre, and that White has the possibility of "inviting" it to face the storm. 17 |xf7! Oxf? The interposition of 17...4\xh4 18 Wxh4 @xf7 changes little: after 19 Wh7+ Bg7 20 QhS+ Bf 21 wh8+ Bg8 22 Axf6+ Black loses immedi- ately, while 19...f8 is insufficient, if only because of 20 4\d5!, renewing the threat of 4h5. 18 xf6 Axf6 19 Bd5! Preventing the black queen from switching to the centre or the kingside, where it could have taken on the role of central defender. 19...b5 Not through choice: if 19..W4d8 White has both 20 Hg5 with the irre- sistible threat of {1h5, and 20 Axd6 We? 21 Hd5 exd5 22 Bdxf6+ wxf6 23 Hxf6+ @xf6 24 ¥d6+, which is also good enough to win. And 19... exd5 20 4)xd5 Wd8 21 Bxf6+ is sim- ilar to this second variation. 20e5 20 Bg5, as given above, would have been more quickly decisive, but on the other hand White now attacks “with checks". 20...dxe5 21 Hxf6+ gxf6 22 Hed+ ET 23 WE2+ He7 24 wyl6+ Ph7 25 Se5+ Hh6 26 Bd3 wel+ 27 Adi Whd 28 8h3 wxh3 29 {7+ Ph7 30 gxh3 Qd7 31 h4 Be7 32 nS Hf4 33 QE Dad 34 whos He8 35 wh8+ Sxf7 36 wyxa8, and White soon real- ised his material advantage. Tal-Sveshnikov 4lst USSR Ch., Moscow 19734 Attack with Mikhail Tal Objectively speaking, White does not need to hurry, since he has ach- jeved all that he could dream of: he has brought his central attacker, the knight at e5, into a striking position, he has his sights set on e6, and his queen's rook and queen are both active. In addition, the advanced h-pawn hinders Black's castling, and therefore 12 0-0 was perfectly possible, in order to begin the assault a move later. But what chess player will defer his enjoy- ment without extreme necessity? 12 &xf7! @xf7 13 Axe6+ G8 The bishop is taboo ~ 13...8xe6 14 Wed+ (a “left hook"), winning imme- diately. 140-0 14 eS Qxhl 15 exf6 is also suffi- cient, when both 15...gxf6 16 Hxd7 and 15...)xf6 16 Bxd8+ Hxd8 17 f3 give White a great advantage. 14..%c8 15 Bxd7 4xd7 16 Hdl Ac6 17 HdS tyb7 18 e5 Ye, and now White had a straightforward win by 19 Qxd7+ ¥yxd7 20 e6 Wxd5 (otherwise 21 4)c7+) 21 HxdS Qxd5 22 WeS5 Sc6 23 wWe7 (but not 23 Wyxe7? BfB 24 Ac7 4 £6) 23....b5 24 c4 Qxc4 25 9.d6, or the more prosaic 25 ¥¥c6+, with mate in three moves. Today the episode, known in its time as the "Argentine tragedy", has been largely forgotten. It cecurred in the 14th round of the Géteborg Interzonal Tournament, 1955, when by the will of the draw an unusual USSR v. Argen- tina match took place: Geller had White against Panno, Keres against Najdorf, and Spassky against Pilnik. Specially for that day, the Argenti grandmasters prepared an innovation in a well known variation of the - Sicilian Defence. Here Black struck the flank blow 9...g5, and after 10 fxg5 he continued 10...4)fd7, planning to regain the pawn and obtain the "eternal" eS square for © his knight. After this his f7 pawn ” would be securely defended, and time © would be gained for the development © of his queenside forces. Black's plan could have been fully justified, had not for an instant the hS- 8 diagonal been weakened and his e6 | insufficiently defended. The first to notice this was Yefim Geller. Abs! Aimed indirectly against the black ~ outpost at e5; to support it with the other knight from c6 will not now be - possible. 13...43e5 This continuation, prepared in ad- vance, loses, as does 13.97, played by the other Argentine grandmasters, The strongest defence was found much © later, after a debate in virtually all the _ 11 2yxe6(!) fxe6 12 whS+ BIB 13 -King in the Centre 5 world's chess magazines. It consists in 13...4hn7!, covering f7 in a different way, when after 14 0-0+ @g8 15 96 Hg7 16 Af? Qxh4 17 wxh6 Axf7 18 exf7+ @xf7? 19 wh7+ Ge8 White either gives perpetual check, or contin- ues the attack with 20 ¥yh5+ @f8 21 Hfl+ Qf6 22 e5, although here too Black nevertheless has a draw. 14 fg31 Essential accuracy. After the move order 14 0-0+ @g8! 15 4.23 the offen- sive could have been parried by 15... hxg5! But now it is all over! 14...2.xg5 Alas, here 14...@g8 is met by 15 GixeS dxeS 16 tyg6+, while 14.7 loses to 15 AxeS+ dxeS 16 0-0 wWe8 17 Hes! 15 0-04 He7 16 AxeS wyb6+ Nothing is changed by 16...9e3+ 17 Bhi dxeS 18 WxeS Ad4 19 dd5+ Wyxd5 20 wc7+. 17 @hi dxeS 18 Wi7+ Bd6 19 Had1+ wyd4 If 19.5 White wins quickly by 20 Wf2+, mating. 20 Bxd4+ exd4 21 e5+ GcS A rare, pure mate in the centre of the board results after 21...@xe5 22 We. 22 tye7+ 06 23 Axc6 Black resigns: 23...bxc6 24 Wa5+ Sc4 25 b3 mate. The other two games concluded in similar fashion. Of course, in the next diagram White has the possibility of further activating his forces - his king's rook, but in this time Black will endeavour to simplify the position, in particular by ...Ad8. Bringing up the knight by 17 most 4e4 is parried by the natural and strong centralisation of the queen — 17..¥¥e5!, and so White resorts to a typical solution: Nezhmetdinov-Suetin Russian Federation Ch., 1947 17 Qxe6! fxe6 18 tyxe6+ fe7 After the alternative defence 18... We? White has a choice between 19 Wb3 Qe7 20 Bgl (but not 20 Hel eS 21 Bxe5 Wyxe5 22 wWxb7 0-0, when the advantage passes to Black) 20...8f8 21 He4, and the more con- vincing 19 wf5S! c8 20 wWf3! Black still cannot castle, 20...Ha7 is met by 21 Se4, and after 20...Wyc7 all the white forces unite with 21 Hel+ 9e7 22 Hgt HfS 23 wWhS+ @d8 24 He? He8 25 4d5! wyaS 26 xe7! (or 26 eyxe8+) 26.,.eyxh5 27 4)xc6 mate! 19 Hed AeS 20 Af6+ HB 21 Hd7! Another typical decision: Black's central defender — his light-square bishop, is removed from the game, We Shall be studying this procedure in detail in a later chapter.6 Attack with Mikhail Tal 21..4xd7 22 Sxd7+ @e8 23 56+ Bas Or 23.8 24 Higl!, and by the threat of mate at g8 White wins both black rooks. But what is he to do now, since Black has prepared 24...47b6 ? 24 Be2t! This decides the game. 24.0986 25 Hdl Wxdl+ 26 Hxd1 Axf6 27 eyxf6+ He7 28 we7+ Hb6 29 c4, and a few moves later Black resigned. Sometimes it is not so easy to ap- proach e6, but any "customs duties” will be recovered on the way. There are many examples, of which one of the most striking is a game from more than half a century ago. Ravinsky-Panov Moscow 1943 The player with a lead in develop- ment is obliged to attack. But there is simply no other target, other than e6, and it appears to be quite securely co- vered by the black knights. Securely? 20 Hxd7!! é)xd7 21 )xe6 fxe6 22 Wre6+ The result is the same, the only dif- 5 ference being that White has had to : pay a higher price. 3 22.4.7 i The king cannot escape from the centre: 22..0)d8 23 Qe5+ Bc7 24. We6+ Bb8 25 Af4+ Bc7 26 AxcI+ ¥yxc7 27 ¥yaB mate. / 23 Hel wWe5 In this way Black coordinates his forces and takes control of the impor- tant g5 square, since it is clear that” White cannot manage without includ- - ing his reserves in the attack. The other defence against the i mediate mate, 23...2)b6, is weaker on account of the overloading of the black queen. White quickly wins by 24 §5 (24 e3 allows Black to defend by giving up his knight — 24... Yyd7; this is where the price, paid by White for the sacrifice on ¢6, tells) 24...c7 25 Ac6+ Hf8 26 He3, and Black has n adequate defence against 27 Bf3+. 24 b4! Destroying Black's coordination. With 24...Wxb4 he loses control over g5, and after 25 @ 25 he does not hav the defence 25...\f6. Nevertheless, ac- cording to analysis by grandmaster Reuben Fine, it was in this line that _ Black should have sought salvation: 25..¥yxel+! (once again reminding — White that in his attack he has not © sacrificed, as usual, a minor piece, but — a whole rook!) 26 Wxel 4\f6 27 Web - Exc2 28 Qcb+ iB 29 Wc8+ Hf7 30 Wyxh8 Axc6 31 Axf6 Qxf6 32 ¥yxh7, and there is still a lot of play to come. After the move in the game Black's king remains for ever under fire. 24.498 25 Wed tyc3 26 Hxe7+! |King in the Centre 7 Leaving the king totally exposed. 26...Qxe7 27 Ae5+ Od6 The alternative was 27.,.@e8 28 Wye2+! @f7 (28...Qd7 29 te7 mate) 29 Qd5+ Bg6 30 wet+ Bxgs 31 Wide PhS 32 {7+ g6 33 wh4 mate. 28 Wdl+! Gc7 29 Alf4+ Hb6 30 Wd6+ Ha7 31 wWe7+ Be? 32 Axc7 Yyal+ 33 Qf Ag6 34 ¥yc5+ Qb7 35 a5, and Black did not in fact succeed in avoiding mate, LD. Misha, the whole world knows of that unusual ease with which you "squander" material in an attack. J rea- lise that this ease is purely superficial, yet is there a limit to your generosity, is there an inner censor, which some- times places a veto on tempting attack- ing ideas? M.T. I hope that we are talking not about those attacks, that are calculated to the end, but about those, in the cor- ': rectness of which one is persuaded by experience, knowledge and intuition? I once discussed this with grandmaster Yefim Geller, and we generally agreed that in intuitive attacks there is a limit to the amount of material that can be sacrificed. LD. That is what the Tal of the early 1990s thinks, But what did, say, the Tal of 1960, the World Champion, think? M.T. Mmmm... If you were to give me that Tal now, I would show him what's what, No, Geller is right: if for an attack on the king you have to give up a piece for one or two pawns, and there is the prospect, even in the ab- sence of a mate, of winning two pieces for a rook, then there is nothing to think about. If you sacrifice a rook for a pawn, it would be good at least to be sure of perpetual check. Of course, this is the most abstract viewpoint, and in concrete terms you could have, for ex- ample, two minor pieces and positional compensation for a queen, LD. In short, if the beauty of a com- bination does not depend, as one of the old writers said, on the “thickness” of the sacrificed piece... M.T. ... then in an attack one still has to make one's demands commensu- rate with one's possibilities. And don't Took at me like that: I have been an ex- World Champion for 30 years! The e6 square may also provide the altar for a "free" offering: a piece in- trudes in here with the same idea - of enticing the black pawn away from f7, thereby opening the h5-e8 diagonal. Tal-Bilek Amsterdam Interzonal 1964 The rejoinder 15 )e6 is logical and quite correct: after 15...fxe6 16 Wyh5+ Qd7 17 ¥yg4 the black king's path to a7 is blocked, and White has numerous threats, including 18 ¥¥xg7 and 19 d6,8 Attack with Mikhail Tal or 18 d6 immediately. 15.496 16 )xg7+ YB 17 Hest Bed Here in the event of 17...fxe6 the black king is completely defenceless. 18 Bhfl 95+ 19 @bi bS 20 whs Af4 21 b3 a5 22 2yc7+! The quickest, although 22 a4 or even 22 g3 is quite sufficient. 22...¥9xe7 23 d6, and Black re- signed, since on 23...W¥d7 there would have followed the obvious 24 Hxf4 exfa 25 wes+. An alternative to the direct blow against a king in the centre is the at- tack on it from the flank. In practice this most often reduces to the attacking pieces breaking through to g7 (usually the queen} or on the a4-e8 diagonal (again, usually the queen). M.T. Of course, we are formulating things somewhat abstractly, rather in the style of Mikhail Moiseyevich (Bot- vinnik - LD.) but, strangely enough, a great variety of attacks reduce to this... LD. In your games too! Ae a Tal-Vooremaa Tallinn 1971 14 We The only possibility of maintaining the initiative, even if at the cost of a piece: after 14 fxeS Qxe5 15 c3 d6 Black has nothing to fear. 14...exd4 15 Yyxg7 HFS 16 e5 Ge7 17 £5 £6 : Otherwise White himself would ad- vance his pawn to this square. 18 4f4 Bf7 19 exf6! Hes The queen is clearly immune — after 19...Axg7 20 fxg7 @f7 21 £6 the threats of 22 4+ and 22 Qh7 are irresistible, and also after 21...xf6 - 22 xhS. 20 eat Diverting the knight both from the e-file, and away from g6 square beside the king, from where in certain var- iations the queen will continue the attack. 20...£)xe4 21 We8+ G8 Or 21...f8 22 We6+ Hf? 23 2yd5 Wyd6 24 fxe7 Wxd5 25 f6. 22 2)xh5 Again threatening a blow from the flank. 22...4)d6 23 Hael+ @d8 24 He! This essentially concludes the game. 24.NYDS 25 Hfel ¥d5 26 |fd — Wxa2 27 Heb+ Wxeb : After 27..dxe6 28 Axf7 2)xf7 29 WxfB+ Bc7 30 Wxi7+ bb 31 fxe6 - in the near future White will obtain as many queens as he wants. 28 fxe6 Hxf6 29 resigns, In the next diagram Black is inten- ding to complete his queenside dev- elopment with gain of tempi, after ” which he will have at least equal — chances. The hopes for White, who up HI7 BlackKing in the Centre 9 till now has been conducting an attack through the centre, involve a flank continuation of it. Tal-Gufeld Gori 1968 20 £)xb5! cxb5 21 @.xb5+ 4)d7 22 Hdl We7 The best defence was 22...Q¢7, af- ter which White appears to have noth- ing better than to go into an equal end- game: 23 Wyh8+ £8 24 Se5 wyxgs 25 Gxd7+ Qxd? 26 Axd7 we4+ 27 @xf2 wad) 28 Wxf8+ Sxd7 29 ¥yxa8, But now the black king is drawn out into the firing line. 23 Hxd7! Qxd7 24 Qxd74+ &xd7 25 Wd5+ Qe7 26 wyxa8 wyc5 27 3, and White's pawn preponderance de- cided the outcome. The same idea is embodied in the illustrative game Tal-Larsen (p.24), but for the moment here is another typical flank operation by White. There is a second theme - the intrusion of the heavy pieces onto the 7th rank, but it is only extremely rarely that any procedures occur in pure form. Botvinnik-Euwe The Hague/Moscow 1948 The position is very sharp. The white rook has occupied the open file, and for the moment — but not for ever — the black king is stuck in the centre. ‘On the other hand, Black has an excel- lent bishop and a phalanx of central pawns. In short, after the retreat of the knight each side will have his chances. But, as it turns out, the retreat is not obligatory, since White has a different way of proceeding. 22 Wye3! fxe5 23 Wyg7 BIB 24 He7 The invasion has taken place, mates at d7 and e7 are threatened, and the bishop is "hanging" — things are bad for Black. He fails to save the game by 24...84d6 25 Bxb7 d3 26 Ha7 tyd8 27 ¥yxh7, with unavoidable mating threats. Therefore he gave up his queen — 24...%yxe7, but was soon forced to resign. In all these cases the path for the attacking queen to land a "right hook” was already open. The same goal may be achieved, if first the way has to be10 Auack with Mikhail Tal cleared by an advance of the pawns. The subsequent invasion of the pieces can then usually be supported by the far-advanced infantry. Nezhmetdinov-Paoli Bucharest 1953 This position also illustrates to an equal degree the next theme of our dis- cussion: how incautious it is to vol- untarily leave the king in the centre. True, kingside castling is not possible here — White is well prepared for an attack on this area of the board, but the "normal" 12...4)e5 (essential now or ‘on the following move to neutralise the queen's X-ray pressure on f7) 13 eye2 0-0-0 would have given the game a “normal” course. But Black, hoping for a counterattack, leaves his king in the centre, and White lands an energetic knock-out blow from the flank. 12...c8? 13 g5 hxgs 14 hxg5 Hes 15 Wg2! Ags (15...Hfg4 16 WF4) 16 f4 2)c4 17 Qxed Wxed 18 £5 bS 19 BHb1 b4 20 g6! eS After 20...bxc3 21 gxf7+ @xf7 White wins most simply by the direct 22 Wyxg7+ Be8 23 wyxh8, although he also has the more elegant 22 b3! ¥¥c7 : 23 Wyg6+ GPS 24 fxe6, mating. Bur now f7 is defended by the queen, and the white knights are attacked... 21 b3t Now, in contrast to 21 4\d5, Black does not have time to block the posi- tion by 21.,.f6. 21...8yxc3 22 gxt7+ Ged8 23 Wyxe7 exd4 24 4 xd4! The quickest, most accurate (weav- ing a mating net) and prettiest solution. © 24..Wyxe2+ 25 gal Bh2 26 Ab6+ | Be7 27 wyxg8+ Black resigns, since he is mated i two moves, White's play was rewarded - by the first brilliancy prize. 2 Uh, Vi, BS RAE Z DG g ROW OT, He Hay 2 7 Ei Larsen-Spassky "Match of the Century" Belgrade 1970 Black has the initiative, but at first ” sight it appears difficult to approach the white king. Of course, it can be forced to remain for ever in the centre, but the price for this (11...Bxd2 12. )xd2 4xe3 13 yyc3 Bd8) is perhaps -King in the Centre 11 | too great. And so Black begins an attack from the flank. 11...h5 12 h3. Practically forced. Firstly, because the knight, which has been "launched into the penalty area" (we will be talk- ing about this attacking procedure later, in Chapter 3) cannot be tolerated tor long, and secondly, because the concrete 12 4c3 is refuted by the equally concrete 12...8xd2!!, when after 13 Wyxd2 Qxe3 14 we2 Gf2+ White has to part with his queen. 12...h4! 13 hxgd The interposition of 13 Qxg4 Gxg4 and now 14 hxg4 would have led to variations, similar to those that oc- curred in the game: 14.,.hxg3 15 Bgl Bht!! 16 Bxhl g2 17 Kgl what 18 We2 Wxged+ 19 Wel we3+ 20 He2 (or 20 dl Wf2 21 Wxed wyrxgl+ 22 c2 wyf2, winning) 20...¥9f3+ 21 Bel {1e7, when the outcome is decided. 13...hxg3 14 Bgl Bhi!! 15 Rxht 2.16 BEL There is no choice: if 16 Rgi Black has the decisive 16...xyh4+ 17 @di Yh1 18 ¥yc3 Wyxgl+ 19 Bc2 w2. 16...Wyh4+ 17 Sdl gxfl=wy+ White resigns. After 18 xfl ALxg4+ mate is inevitable. LD. This confirms what you said in an interview, given after the "Match of the Century” in Belgrade: "Spassky... won a game, that will undoubtedly find its way into the books”... S M.T. In my opinion, only Nostra- damus has made faultless forecasts, decades and centuries beforehand. Besides, the attack from the flank was carried out brilliantly by Boris, but there was no sacrifice on f2, the theme of this section... LD. Well, what of it? All the same, examples cannot be clinically pure, procedures and themes are interwoven and, incidentally, we should now make a slight digression and jump ahead. The sacrifice on f7, drawing out the king, can also be made after it has moved away from the centre into its shelter. The double defence of this point makes it less vulnerable, of course, but even so the pawn linkage “e6 + f7" continues to draw the atten- tion of the attacking side, and nearly always requires additional defence by other pieces. Otherwise disaster can strike here. This eternal truth is sometimes for- gotten by even the most experienced grandmasters. Smirin-Smysloy 55th USSR Ch., Moscow 1988 Of course, Black stands worse, and the powerful piece outpost at e5 simply obliges him to keep an eye not only on his queenside pawn weaknesses, but also on the "e6 + f7" link of his pawn12 Attack with Mikhail Tal chain. Therefore 21...d7 was essen- tial, and if 22 2g5 4fd5, at least blocking the dangerous diagonal. 21....a67 This activity leads to disaster. 22 4)xf7! Bxf7 23 xe6 A\edS 24 &e5 Baa? 25 Bxc6 25 &xc6 wyb6 26 4)xa7 wyxa7 27 g5 would also have won. 25...4)f4 26 Bxa6! Bxa6 27 &xf7+ @h8 28 Wf Black resigns (28... Wyxd4 29 26). Kaplan-Radulescu Hungary 1980 Black is cramped over the entire front. Especially unpleasant is the rook at c7, preventing him from completing his queenside development. It would be good to send it home by 18,..2\d5, but then for an instant the black king re~ mains completely undefended, and 19 Wyh5S leads to a quick mate. Hence his natural desire to first drive away White's “central striker", also remov- ing his h7 pawn from attack. 18..h6 19 2\xf7! It turns out that now this is possible. After 19...xf7 20 .xf6 wyxe7 (20... gxf6 or 20...@xf6 allows a quick - mate) 21 wh5+ @f8 22 Qxg7+! xg] 23 wg6+ HIB 24 wyxho+ G7 25 26+ 6 from the several possi. ble mates White chose 26 Qh5+ @f: 27 Wy g6+ id 28 Wed mate. meee # ot cone ae eae nell Adorjan-Tarjan Riga Interzonal 1979 Not one of White's pieces is. “looking at the kingside. But for th moment Black's main forces too are blocked in on the left corner of the board, and so White can calmly pre- pare an attack on the classic point f7. 17 &e5t An essential prelude, planned lon; beforehand, otherwise the black knig) leaps via d7 toc5. 17...a7 18 G&h3 {a8 19 2)xf7! @xi7 Alas, forced: WB 21 sxb8. 20 Axb8 Yyxb8 21 Yyxe6+ GHe8 22° Hd? Bxd7 23 tyxd7+ @f7 24 Web+ White contents himself with perpet-: ual check. The alternative was 24 19.,.8xf7 20 Qxe6King in the Centre 13 Qe6+ Sf6 25 5 (threatening mate in one move) 25...¥¥c8 26 e5+ &g6 27 Wxe7 .xd5 28 Yyd6+ Qe6 29 Wxb6 h6 30 xa @h7 when, although in the endgame (31 ¥¥c6) Black has nothing to fear, and with the queens on the white king might come under a coun- terattack, even so White's chances look preferable. 2A... He8 25 Yyd7+ Draw agreed. And, finally, another motif may sug- gest the preparation of this typical combination: the deployment on the a2-g8 diagonal of the black rook and king, or in the following instance, of even the queen and king, Keres-Geller Budapest 1952 18 )xf7! Yyxf7 Capturing with the king would have allowed White the same possibilities, plus one more: the inclusion in the attack at the required moment of the queen from h5. 19 Qb3 Hed 20 Axed bxed 21 xed 2d5 22 Qxe7 Yxe7 (or 22... Bxe7 23 WWd2) 23 exdS tyxe2 24 Hxe2 Hxe2 25 Qxe2 Mxd5 26 a4, and in the endgame White was a pawn up. oy iy YQ, ate mits e. a Botvinnik-Vidmar Nottingham 1936 In this example the same motif dom- inates, the only difference being that the black pawn has already gone from e6, and White has acquired an addit- ional factor: along the h3-c8 diagonal his queen has its "sights" set on the took. 20 £)xf7! Eixf7 Things are very simple after 20... Bxf7 21 Axd5+. 21 Axf6 Axf6 (or 21...Hxf6 22 Hxf6 and 23 Yyxc8) 22 Bxd5 tYyc6 23 dé! Avoiding a little trap: 23 eS O.xd4+, when White has to agree to a draw. 23...¥e8 24 Bd7 Whereas now Black resigned. But let us return to the main theme of this chapter - the attack on the king, caught in the centre. In all primers, even those for beginners, this factor is14 Attack with Mikhail Tal mentioned as a sure indicator of a possible attack. There is probably no sense in once again listing all those difficulties that a king, remaining on its initial square, causes its troops. They are weli known. But also known is this paradoxical fact: for centuries, in one toumament after another, their partict- pants forget to concern themselves about their king, and in such instances the response of the opponent can and should be appropriate. In short, it must be said again and again: the king in the centre is the signal for an attack. Because a delay in evacuating the king can cost a player very dearly. RA] Ue YA a) Y), Ay Nezhmetdinov-A.Zaitsev Kazan 1964 Here, for the first time, Black should have castled, maintaining equa- lity: 11...0-0 12 @c4 Bd8 13 0-0 @h8, or 12 c4 4c7, when White's light- square bishop is not too well placed. But Black was tempted by something different. 11...e5 12 0-0 He7 13 Ac4 Now, even after the exchange of queens, it becomes uncomfortable in the centre for the king: 13..uxd2 14 - 4)bxd2, with the threat of e4-d6+. 13...b5 14 Qb3 Ea6 15 wye2 Hed 16 Hdl Yyc6 17 Ae2 He6 Castling would now be bad: 17. 0-0? 18 b4 4)b7 19 Ged ee8 20 a4 bxad 21 Yye4+ BF7 (21..49f7 loses a piece after 22 Wyc7) 22 a3, and there is literally nothing that Black can — move. 18 Hbd2 Bd8? This second voluntary rejection of castling, for the sake of simplification, leaves Black on the verge of disaster. 19 b3 Bxdl+ 20 Wxdl 2yb7? This third rejection of castling leads to a crushing defeat, although for his © previous errors Black would have had to suffer mounting difficulties in the variation 20...0-0 21 AxcS Wxe5 22 © Ae3 We7 23 4b3. 21 a4 a6 22 axbS axbS Black is not afraid of the check at a8: there would appear to be no way of | approaching his king at f7. But... : 23 4ybd4t exd4 24 4)xd4 wWd7 25 Syxe6 Wxe6 26 Ha8+ 4)c8 Alas, the f7 square has become - "mined" on account of the weakening - of the a2-g8 diagonal. : 27 Qb3 wd7 28 We2+ Hds 29 | Le6 Bes Is everything in order? ; 30 Hxc8+! eyxc8 31 Wydl+ We7 32 - Axc8 Hxc8 33 Ye2+ Black resigns. (see diagram next page) After playing the opening without any particular pretensions, White should have concemed himself about his king - 10 Ge2 and 11 0-0.King in the Centre 15 Gromek-Veresov Poland y. Belorussia 1954 But the premature sortie 10 4a5 placed him on the verge of disaster, especially after 10...c8 11 Qe2. Now came a piece sacrifice. 11...2)xd5! 12 exdS S\b4 13 dxeé The game would have lasted longer after 13 c4 @f5 14 Qd4 Qxd4 15 Qxd4 42+, but with the same result. 13..£)xe2+ 14 BZ Hxe3 15 Wxe3 Yyb6+ 16 Hdd Or 16 @d3 Qh6 17 ext?+ Bxf7 18 4\d4 Wab+ 19 Bed d5+! 20 Sxds Wd6+ 21 ed Aras, 16..8xe6 17 td3 Ee2! Not the only continuation, but the most efficient. Now 18...Qxd4+ is threatened, and if 19 tyxd4 Hixe2+. 18 Hhdi Bxb2 19 £4 e5 20 fxeS dxe5 21 wyce4+ @h8 22 263 Hb4 23 WadS Bxd4 24 ¥yxb7 He4+! White resigns. But, as a tule, a king does not re- © main in the centre voluntarily: it has to be forced to do so, and not “for free". The price may vary: from sacrifices to positional concessions, which may be very diverse. zy ea iw tet spcole Nezhmetdinov-Tal 29th USSR Championship, Baku 1961 Here such a concession was 15 4)h6, which seems to place the white knight into an offside position, This was the reaction of one of the authors, who decided at this point to “punish” the opponent for his anti-positional decision. In principle this would have been right, had not the audacious knight covered all these drawbacks with one trump: it blocks the king's path to a secure shelter, 15...2)€6 16 222 e717 Sxf6t To the positional "sacrifice" is added a normal one, but now the inva- sion at dS becomes possible. 17..8xf6 18 dS wyd8 19 wiz sf? White would have encountered greater difficulties had Black first played 19....xd5 (now or on the next move), when there appears to be no clear-cut win. For example: 20 exd516 Attack with Mikhail Tal SD f4 21 Axf4 exf4 22 wWxf4 wb6+ 23 Bhi 0-0-0 24 c3!! (the direct ex- change of blows 24 xf7 #xb2 25 Afl AhfS 26 ¥yb4 {97 27 Dxd8 Axfl+ 28 Qxfl wf2 29 Gh3 leaves Black with the advantage after 29... Wf+ 30 gl wxd5, but not 29... 047 30 g5+ Hxd8 31 wd6+ etc., when White wins) 24...Ahf&8 25 g5 Re7 26 Bh3+ Gb7 27 Axf7 Ba7! 28 $26, and although White has a serious advantage, there is still a struggle in prospect. 20 {xf4 exf4 21 eS! AxeS Alas, Black is obliged to open the central files, along which he comes under attack, 21...Qh4 is very strongly met by 22 Wed4!t, with the threat of e5-e6, and after 22...8f8 23 Hdl xd5 24 Axd5 Ac8 25 wa7 We7 26 #xa6 White must win, since 26...¥9xc2 is bad on account of 27 Qxf7+ Hxf7 28 wye6+ He7 29 wWe8+. If 21...4g5 White has time for 22 \fo+ Sf 23 Axb7 Hbs 24 Of3 Axh6 25 Hdl we7 26 wh4 497 27 Dd7+ BHe8& 28 wW2 HcB 29 wd, when Black cannot parry the numerous threats, the chief of which is 30 1b7 Ac7 31 axa6. And after 21...9xd5 all the same the e-file is opened, and White wins the queen: 22 exf6! xg2 23 Hel+. 22 Hel £6 23 Sxfé+! wxf6 24 Wad! @f8 25 Axe5 wd8 Or 25...d8 26 He8+! Qg7 27 Be7+. 26 Bf5+ gxf5 27 wxh8+ we7 28 We27+ Seb 29 gxf5+ Black resigns. This ability — to prevent the oppo- nent from removing his king to safety — was also possessed by the top play- 2 ers in the earlier era of romantic chess. P.Morphy-E.Morphy USA 1850 12 e5! dxe5 13 Bfel Ad7 14 Babi © The black king has not been allowed : across to the kingside, but on the queenside too it will evidently be un-- comfortable. But there is no alterna- tive. : 14...0-0-0 15 a6! Had 16 Hect Ac6 17 eyxa5 bxa6 18 eyxa6+ Hd7 19 &ixc6 and White won (19...2xc6" 20 Hdl +). 3 LD. However, let us leave in peace the shadows of our great predecessors. « Their opponents played rather naive. chess, and on the whole they defended: weakly (here, for example, more tena- cious was 15...bxa6 16 ¥yb3 Se4 17. ¥yb7+ Bd7 18 Hbdl+ e8 19 Hxd8+ Bxd8! 20 wa8+ Ac8 21 Bdl+ sda, although White still retains an attack -- 22 Yyd5+). More instructive, in my op- inion, are modern games even at the le- vel of masters, to say nothing of grand- masters, candidates and champions.King in the Centre 17 M.T. Alas, I am forced to agree. In my youth I gained enormous pleasure from examining the games of the old masters, but when the so-called “Tal School" opened in Riga, at the very first meeting £ said to my young friends: “For pleasure you can read the games collections of Anderssen and Chigorin, but for benefit you should study Tarrasch, Keres and Bronstein." In modem chess the art of defence has grown greatly, and it is becoming increasingly difficult to win a won pos- ition. But the basic principles of attack are eternal. Moreover, the assault need not necessarily resemble a meteorite, be instantaneous and be calculated from beginning to end: the king in the centre can be the target of a lengthy, multi-stage offensive, with one phase logically developing into another. An example — and a brilliant one! is provided by the following game. Chiburdanidze-Dvoiris Tallinn 1980 White is fully mobilised, and essen- tially cannot even strengthen the placing of her pieces: they are ex- tremely actively placed. Black, by con- trast, has made several “superfluous” pawn moves, his queen has gone to b6 in two moves, and as a result not all his forces are in play. And yet he needs only one tempo for queenside cast- ling... The player with the advantage is obliged to attack! 12 4d5!! exd5 The other knight is taboo: 12... Wyxd4? 13 Q.xf6 exf6 14 2xb5 and 15 xf6+. 13 Het Before opening the c-file, White cuts off the black king's escape to the saving shelter at c7, 13...4xc6 The attacking file cannot be kept closed: 13...d4 14 e5! 14 exdS+ Ge? (14.08 15 dxc6 4c5 16 c7+) 15 dxc6 AcS 16 A xf6 gxf6 17 £51 And once again the black king has nowhere to go. 17..89¢7 Forced, since Black has to defend both his bishop, and his d6 pawn, and also the advanced white pawn has to be blocked. 18 b4 Ae6 19 Wh5 427 20 Ad7+ Erecting a barrier in front of Black's queen, which is obliged to look after his bishop, 20...@f8 21 Wh6 aS The main point is to bring the queen out to d6, but in passing Black sets a little trap. 22 Hxe7! @xe7 23 Hel+! The trap —~ 23 Qyxg7? wyxf4+ 24 bl Hag’, when Black wins — fails to operate.18 Attack with Mikhail Tal 23...Q18 24 wxf6 Gg8 25 He7 BS 26 Aec6! It's all over! 26..89xe7 27 tyxe7 fxe6 Or 27...2)xe6 28 £5 \g7 29 c7. 28 c7 h5 29 ¥yxf8+ Black resigns. This theme received a quite amazing interpretation in the following world- famous game, M.T. I am happy to admit that 1 watched this game with envy, and would have readily exchanged my silver medal of that championship for the right to sign the scoresheet — the white half, of course. LD. And when White's next move was announced in the press centre, we decided that the impregnable Ratmir Kholmoy, the "central defender" as he was then called, had for a moment taken leave of his senses. Kholmov-Keres 26th USSR Championship Thilisi 1959 12 Heé(!!) The following variation is not too complicated: 12..%9xd1?2_ 13 Hixdl G&b7 (or 13....d7) 14 45, when: 14.....xc6 is bad in view of 15 &\c7+. Sf8 16 )xa8 Qxa8 17 Hd8 mate, and | otherwise after 14...0-0 White simply | picks up the e7 pawn with check. But» Black has another possibility. 4 12...%d7()) 4 ‘This is what was played, and how: can White now save his piece? 13 Axe7t! z Once again two exclamation marks, , although this is merely a continuation: of White's brilliant conception. White: forces the black king to remain in the centre, since in the event of 13..%yxe7 14 AdS Wd8 15 |fo+ Oxf (or, 15..e7 16 WE3 Ge6 17 25, and Black loses material) 16 exf6+ 9e6. 17 Axh6 Wxf6 18 Yye4! without sac-. sificing anything White has a powerful: initiative, which is merely strengthened by the opposite-colour bishops. And: totally bad is 13..¥yxd1 14 Bxdl. @xe7 15 Be5+ Beb 16 Hd6+ BS (16..<@xe5 17 Bd5+, mating) 17 £4, with various mating threats. 13..xe7 14 Axho! 3 Again a paradoxical decision. The. bishop, which was predestined for an’ attack along the weakened dark squares, is exchanged for the "off- side" knight. s M.T. But on the other hand, a major: attacking principle is maintained — "a. tempo". The time that we don't have i: more precious than the piece that we do have. . 14... xh6 15 ¥f3 Ag7! 3 The best defence. Now after 16. Wyxa8 Qb7 17 Wyxa7 cb 18 £3 Bad | 19 A\d5+ Wyxd5 20 wyxb6 the outcome is unclear.King in the Centre 19 16 HdS+ Gas Forced, since Black loses quickly after both 16..@e8 17 f6+ Axf6 18 exf6+ and 19 ¥yxa8, and 16...@f8 17 6 Wb7 18 27+ Hes 19 wi! 17 Badl Qb7 To find the strongest continuation 17...4¥b7 at the board would be practi- cally impossible, in view of its para- doxical nature. True, even in this case after 18 e6! White has a very strong attack, e.g. 18...fxe6 19 b4+ @c7 20 Yye3+ e5 21 BxeS! exb4 22 HcS+, mating. 18 ¥yb3 c6 19 4)xb6 axb6 20 Wt? QxeS 21 Hxd7+ Qxd7 22 Zixe5, and White won. More complicated are instances when both sides appear to have the right to play actively, and when it is ore difficult to decide: should one ‘ontinue attacking or now switch to lefence? Here the deciding role is layed by an accurate appraisal of all ne factors "for" and “against”. sanans ae ow aca Karpov-Yusupov S5th USSR Championship Moscow 1988 Both kings have lost the right to castle, and, in addition, the positions of each side have their pluses and mi- nuses. White, for example, is for the moment playing without his king's rook, and at least three tempi are re- quired to bring it into the game. Black, for his part, is clearly weak on the dark squares, but he only needs to bring his king to £6, when he could think of seizing the initiative. This could have happened, for example, after 16 Yya3+ fo 17 Qd3 yd5 18 Wyxa7 Bed!, or 16 Qxe6 fxe6 17 Wa3+ G6 18 ¥yxa7?? Wyb5+ 19 @gl Ba winning, or 16 Sg5 QB! (here 16...@f6 is weaker in view of the concrete 17 &xe6 fxe6 18 Se4+, when White wins the exchange) [7 4)xe6+ fxe6 18 Qxe6 Wyb5+ 19 Sel He7! with coun- terplay; in this last variation 17 gxe6 fxe6 18 Dxe6+ Wc8B 19 wyb3 He7! also favours Black. ‘The correct evaluation of the posi- tion is only established by the typical attacking device carried out by the ex- World Champion. 16 d51 At the cost of the isolated pawn the black queen's path to bS is cut, and more important -- the weak dark squares become glaringly weak! After all, now White can also attack along the long dark-square diagonal frorn c3. And another achievement for White becomes apparent on his next move. 16...cxd5 17 AbS! Revealing another target — the knight at d7, which is insufficiently protected, as best illustrated in the variation 17... @f8 18 ¥yc3 Hg8 19 44, 20 4)xe6+ and 21 @xd7.20 Attack with Mikhail Tal 17...a6 18 Wa3+ @d8 Alas, forced: 18...@f6 19 \xd7 SLxd7 20 Hye3+. 19 WyaS+ Ge7 On the queenside the king would have had altogether no defenders, and no chance of surviving: 19...ec8 20 Hicl+ Gb8 21 We7+ a8 22 Hd4 Wed 23 9 xa6! Bb8 24 wya5! wyxd4 25 $ixb7+ @xb7 26 Hc7 mate. And al- though this variation is not forced (for example, 22...4/f6 is possible), all the same the sacrifice on a6 destroys the king's shelter and wins, 20 Wh4+ BE Exploiting a tactical nuance, Black tries to save his king (21 (Qxd7 ¥d3+), but the al-h8 diagonal is on the point of collapse. 21 Wd4+ Be7 22 Gd3 whs 22..9f6 23 eyb4+ is even worse, but now in this variation the black king can try to slip away to g7 via f6. 23 h4 With the terrible threat of 24 g4. 23.808 24 £25 Renewing the same threat of 25 g4, strengthened by the possibility of 25... Wh6 26 Bxe6 fxe6 27 Af7+. 2A...BnfB 25 {e2! wyh6 26 Af3 White is effectively a queen up! 26..He7 27 tyb4 26 28 wd6+ Hd7 29 wyf4 28 (29...4727 30 Bxe6) 30 Aged HcB (30...dxg4 31 wWb8 mate) 31 §.xe6 fxe6 32 Hel+ 8 33 )xe6+ e7 34 wyxfS+ wyxfB 35 4)xf8, and White had a winning mate- rial advantage. The basic solution to this position was provided by a breakthrough in the centre, about which we will have more to say later, But the theme itself — “the attack on the king in the centre". — is practically inexhaustible, and al of the above examples are merely milestones on the way. However, sooner or later we have t draw the line, and we do this with the. following position, which as though: encapsulates the entire chapter. Ivkov-Kagan Rio Interzonal 1979 The queens are no longer on the board, neither is there any pawn ten- sion, and all the white pieces are based on their own territory, on the first three ranks. Yet with a clear conscience Blac! can resign, He did this, a little later, after 17. QNbS! cxbS 18 A xbS+ SF 19 Ah6+,. forcing mate or crushing gain of. material, 5 Here there was everything about which we talked earlier: a lead in de-~ velopment, open files and diagonals, : but most important — the opponent's king, caught in the centre.King in the Centre 2t Tal-Portisch Candidates Quarter-Final, Bled 1965 Caro-Kann Defence 1 ed 6 Strangely enough, au unexpected reply, since the Hungarian champion adopts this defence very rarely, and we for more precisely, D had somehow not thought up any plans against the Caro~ “Kann. When, during our pre-match preparations, my trainer asked me what I was intending to play against 1.06, simply replied that this de- fence would not occur, It immediately became clear that the Hungarians had carefully studied the records of my matches with Botvinnik, ‘nd therefore, without any prepared ‘mines", it would have been inex- ‘pedient to employ the variations with hQ-h4 (2 d4 dS 3 Syc3 dxed 4 |yxe4 £55 4)p3 96 6 le2 ALG 7 hd, or d4.d5 3 eS Of5 4 hd), After Portisch's first move I began rantically trying to recall which of the variations I had not hitherto employed in my tournament games. Perhaps, 2 ‘@)c3 d5 3 4)f37 It was this plan that 1 Wecided to try, especially since the ‘Hungarian grandmaster had never played against it. 2 Hc3 a5 3 fs dxed ". This reply shows that White's choice was a happy one. 3...@.24 has long ‘been considered best here, not yielding White an inch in the centre. Who knows, perhaps Portisch did not like the variation 4 83 Qxf3 5 gxf3!?, as I played in the third game of my first thatch with Botvinnik? But now White makes the advance d2-d4 unhindered, obtaining a slight but enduring advan- tage. 4 axed gt 5 3 xfs 6 Wx QdaT 7 a4 Deis 8 Ad3 Axed S wrxed 6 10 0-0 16 ¢3 is perhaps more accurate, in order to retreat the queen to e2 after it is attacked. White, however, has some- thing different in mind... 10... 11 3 eT ate ‘The first critical position. Portisch’s reasoning is clear: Black has a slightly passive position, but one that White will be hard-pushed to crack. He handles such pesitions with great precision. New the follawing natural plan sug- gests itself for White: t¥e2, then @d2 or sf4, when after the obligatory ...c5 and the exchange on c5 he obtains a queenside pawn majority plus the two bishops —- Black faces a gruelling22 Attack with Mikhail Tal defence. And I, of course, realised that 12 We2 was a natural and strong move. But nevertheless J was attracted by another idea. 12 wh4 This provokes Black's reply: WW... ds In a number of variations Black strengthens his defences by the ma- noeuvre ...2)f6-d7-f8, but now from d5 it is difficult for the knight to reach f8. However, this is achieved at the cost of several tempi, and Black has time to stabilise the position. 13° wed Af Not, of course, 13...0-0 14 gh6 S16 15 te4, winning the exchange. 14 Hel Initially 1 had been intending to complete the queen manoeuvre by 14 wed, keeping the black king in the centre, but then I noticed that after 14...2)e7 followed by ...2f5 White's initiative would quickly peter out. Allowing Black to castle is by no means evidence of White's good nature — after 14...0-0 he succeeds in advan- tageously regrouping his pieces: 15 Qh6 Be8 16 Had] ¥yb6 17 fic}, when the bishop returns to what is perhaps its best position, and the rook is actively placed in the centre. So Portisch prevents the development of White's queenside: 14... Wb6 The second critical position. Up till now Portisch had played quickly and confidently. White has clearly not achieved what he could have expected from the opening: the queen is pressing on the b2 pawn, pinning down the bishop at cl. Therefore “correct” (one. can also write it without the inverted commas) was the reserved 15 a3, with the hope, in a protracted struggle, of. exploiting the illustrious bishop pair. All this is true. But for 14 moves the black king has remained with impunity. in the centre of the board, Is there no. way of exploiting this and at the same time abruptly changing the course of. the game? This was how the idea arose of the: rook sacrifice at e6. An amusing variation flashed though my mind — ‘one that was by no means forced, but. highly camouflaged, which afforded a® certain aesthetic pleasure. And I con- cluded that, even if the sacrifice might. lead only to a draw, it was correct, and. in addition the change in the character of the play was bound to be unpleasant for Portisch. 15 cA?! On 15...4e7 [ was intending to play. eyb4 16 d5! cxd5 17 cxd5 2xd5 18 wads, at least preventing Black from castling. 16 Bxe6+ fxe6 17) Wyxe6+Here Black was faced with the problem of how to punish his opponent for his "recklessness". It is clear that this aim is least of all answered by ~ 17...d8: White has 18 ¥yd6+ Bes 19 ¥Ye6+ etc. I did not see anything better during the game, nor do I now. My hopes were mainly associated with the "refutation" 17...¢.e7, when White does indeed lose after 18 125 We7 19 Bel 2xd3! 20 xe? wd7! - Le can, however, obtain a very attract- ve position by interposing 18 G.g6+!, when Black cannot play 18...hxg6 19 $125 We7 20 Bel with the threat of 21 Yxg6+! Therefore he must reply 18... dB 19 FS eYyxd4 (not 19...¥¥c7 20 G4 We8 21 Wed) 20 f4. This position greatly appealed to me, and a ‘ontinuation of the variation (not ex- actly forced, it is true) revealed a most itiractive idea: 20...Ge8 21 Hel g6? 2 ge3 Wd6 23 Qxa7! wyrxed 24 Ab6+ Bc8 (not 24...d7 25 Axe6+ (d6 26 c5 mate) 25 G@xe6+ and 26 ‘Sid7. However, in this variation too lack could have defended. : After the game Portisch admitted hat he had seen the rook sacrifice, but ind not considered it dangerous; and When it in fact took place, he became It at ease. Only this nervousness can xplain why he instantly, without hinking, replied 17... fs was even upset: so much time and fort had been spent on the analysis of lhe most complex and attractive vari- tions, and not one of them was to cur On the board... 18 afa Bas The only good move. Insufficient is King in the Centre 23 18...Be8 19 Qd6+ Ge7 20 Bel wd 21 He3!, or 18..4¥d8 19 Hel gS! 20 £.d6+ $g7 21 Qxb4, when White has a very strong attack with level mate- rial, By parting with his queen, Black can face the future with confidence. 19 5 &)xd3! 20° cxb6 Still out for blood... My first inten- tion here had been at last to force a draw by 20 {1h6, to which Black's only good reply is 20...¥9xb2! (less good is 20..%¥c7 21 wyxf6+ Sg8 22 §ixg7!). But not every c-pawn can reach a7! Therefore White decided on the again rather risky, but not losing capture of the queen, 20... ayxfa 21 wed @d5 22° bxa7 The third critical position. Initially it appeared to me to favour White, but when, awaiting my opponent's reply, I began examining the plan with ...26 and ...g7, I had to change my evaluation: chances were roughly equal. Of course, were Black to suc- ceed in coordinating his forces, his24 Attack with Mikhail Tal position would become clearly prefer- able. But there are two factors that hinder this: the audacious pawn at a7, which at the very least frightens Black, and, although only temporary, the dif- ficulties in developing his king's rook. With his next move Portisch tries simultaneously to solve both of these problems, but the further course of the game shows that the plan chosen is incorrect. 22 see Ber "Suicide?", was my first thought. But then I realised that Black's ma- noeuvre was not without its logic: the king heads for the a7 pawn, and any checks will only act as a "favourable wind”, for example 23 Bel+ @d6 24 We3+ Bd7 25 wed+ Bc7. But Portisch did not see my next move, which, incidentally, would have occurred to a player with an attacking style. Therefore it would have been stronger, without further ado, to play 22...6. How the game would have developed in this case, I honestly do not know. But at the board it seemed to me that to a considerable extent the a7 pawn “insures” White against defeat. Ina number of variations he is able, by giving it up with a8=¥¥, to pick up both of Black's queenside pawns in com- pensation. bat Preparing a "warm reception” for the enemy king. It is bad to take the pawn because of 24 Eb1, but after b4- bS White either exchanges on c6, opening tines, or advances it further, when the lone a7 pawn is transformed into a pawn tandem a7/b6. Here Black could have held the position by 23...4\c7, suggested b: Aronin, But the picture had been unexpectedly transformed, that Por-. tisch succeeded in losing a defensibl position in a matter of moves. 23 Has If not 23...4)c7, then Black shoul at least have chosen 23...d6 24 b: c7, not abandoning, for the moment, the pursuit of the pawn. In any event Black will not have time io take it. 24 Bel+ @d6 25 bs Hixa7 This loses instantly, but by no' Black's defence was very difficul White has at his disposal twe threat: the attacking 26 bxc6 bxc6 27 wye6+: Sc? 28 Hel, and the constricting 263 b6!, after which Black must eith agree to the a7 pawn being a constant threat, or nevertheless go in for the extremely dangerous opening of the file (after 26...2)xb6 27 Hib1). 26 He6+ &ec7 27 Bxf6 Black resign: Tal-Larsen 3 Candidates Semi-Final, Bled 1965 3 Alekhine's Defence a 1 e arte 2 & a d5 3 d4 d6 4 HS dxeS S &xe5 6 So, once again this dubious (there is no other word for it) variation, Directly after the Larsen-Ivkov mate! the Yugoslav grandmaster showed me ™ the final game of that encounter, and “scolded” himself for making the pseudo-active move 6 wh5. At tlsame time, he offered the following opinion: that 6 4)d2 is the quietest reply, and 6 Wf3, although committing, is the strongest, with which, after a cursory analysis, 1 agreed. Of course, in the existing match situation, the committing nature of White's move was by no means a hindrance. we wit 6...€Af6 is perhaps stronger, agree- ing to a slightly passive po: White would have replied 7 4e3, pre- paring queenside castling, After »» Larsen's move the black queen proves to be badly placed, and he has to waste precious time defending it. 7 we3 ho 8 4.25 was threatened. 8 Ac3 White must play energetically, in ‘order to exploit the better placing of his pieces. Since exchanging on c3 would only strengthen White's centre, and his queen's knight is threatening to move with great effect to e4, Black continues manoeuvring with his knight, hoping to divert White from his inten- tions, 8... |b4 9 Qb5+ 6 10 fad a7 . The transference of the bishop to the father inactive post at a4 does not signify that White has given up ideas of an attack. After e4 and c2-c3 the ishop can be advantageously switched {0 the important bl-h7 diagonal, . Here White thought for quite a long time. The advantages of his position are perfectly clear: a lead in devel- opment, Black's difficulty in evac- ating his king, but all this must be King in the Centre 25 effectively exploited. In such a position victory is doubly necessary: if 1 could not win from such a position, the psychological shock would be too great — it would mean that I had altogether forgotten how to win. Here White has several very tempt- ing ways to develop his initiative. The first that suggests itself is the tactical possibility of 11 a3 Qd5 12 &xc6. Then 12...bxc6 loses quickly to 13 Axc6 H5b6 14 DHbS! But the idea proves insufficient after 12...6)xc3 13 bxc3 (13 ¥yxc3 Hb6 14 |b8+ Hd8) 13.,..€)b6 14 @bS Q.d7 15 &xa7 Wd8. The positional 11 .f4 does not give anything decisive after 11...04d5. White gains good attacking chances by 11 Se4 wfS 12 £3, and objectively this was perhaps the strongest con- tinuation, although after 12...4)xe5 13 dxeS d7 14 a3 Sd5 15 c4 Qv6 16 Sic2 Wh it is not so easy to continue the attack. I was intending to continue my analysis of this variation, when suddenly my attention was drawn to the possibility of the piece sacrifice that occurred in the game within a few moves, This idea seemed very tempt- ing. ll 0-0 Exes This knight cannot be tolerated for long. 12) dxe5 13° YB White could have obtained the bet- ter ending by exchanging queens, but this would have been a small achieve- ment. WB vee wis Not altogether successful. Black drives the queen to a more comfortable wee26 Attack with Mikhail Tal position. Stronger was 13...4.d7, defending c6, on which I was intending to play 14 ¥ye2, with the idea of Se4 and ¢2-c3, bringing the light-square bishop into play. Of course, 13...9xc2 14 Axc6+ was out of the question. 14 We2 Qe7 Black appears not to have seen the intended sacrifice. White's intention was that if 14...2)d5 15 4b5!, an idea that it is still possible. However, if 14....d7 (which would have ruled out the knight sacrifice) the manoeuvre Se4, c2-c3 and gc2 would have gained in strength. Now, however, 15 £4 would be met by 15...0-0 16 4e4 bS! 17 Qb3 cS! and if 18 c3 c4, not allowing the bishop onto the bI-h7 diagonal. 1S a3 Had the following move not been available, this would have been a seri- ous mistake, leading to the loss of the initiative. A eo 15...2)a6 can also be answered by 16 4bS, although with the slightly different idea of 16...cxb5? 17 {2.xbS: and 18 4.43, winning the queen. 16 yb5! A move which came as a surprise my opponent. White is intending play his knight to the ideal square After, for example, 16...0-0 17 4d: ¥Wh7 18 c4 46 19 Bc? Black's posi tion is extremely difficult, and so the: acceptance of the sacrifice is forced. 16... exb5 17 WyxbS+ : But here 17 Qxb5+ @f8 18 d3.: fails to 18...4)f4! 4 7 ... Sas ' 18 cd : The critical position. Where should. the knight move to? Larsen follows the: path of least resistance: he gives bac the piece, but the position of his king the centre allows White easily to, mount a decisive attack without an significant material sacrifice. White saw that after 18...4)b6 1 Wa5 he would also regain his piece : since 19...Q.d7 20 S.e3 @c7 2 c5 is: bad for Black. Stronger is the imme-: diate 19..c7 20 c5 @b8 21 cxb6 axb6 22 wb5 Ha5 23 wyb3, although. here too White has very good attacking chances, The main variation of White's coi bination was 18..0f4 19 Bdl+ @c7- 20 Bd7+ Qxd7 21 wyxd7+ bs 22. Wyxe7 Yyxe5 23 Me3 Agé (24 HB was threatened) 24 ¥yxf7? wWf6 Se8! This is the point: 25...uy (25...He5 26 Af4) 26 wyxg6 Hxe8 Qc5! Wd7 28 Gd6+ Sc8 29 c5, and: the bishop at d6 is much stronger than a rook. WW... wyxedKing in the Centre 7 19 cxd5 Ado 20 3 x5 21 we2 It becomes clear that for the pawn White has much more than sufficient compensation. The black king is a long way from any sort of peaceful refuge, and White only needs to develop his queen's bishop, for all his pieces to in in a decisive attack, 21 es He7 22 dt was 23° wed wis . This clever possibility does nothing all to help Black. White considered ere 23...e5, after which neither 24 h44+ £6 nor 24 95+ hxgs 25 xg5+ G8 26 Hxd6 Me6 27 Hadi ! is altogether convincing. But by ntinuing 24 Wxg7 Me6 (or 24... xa4) 25 Qxh6 White gains an resistible attack. 2d hed 24 &yxg72? eS! would have been terrible blunder. ‘The queen manoeuvre emphasises ist how hopeless Black’s position is, ince after 24.,.8.d8 25 (e3 he has no bod move, IA sae wes 25 wd3 wd5 26 Wye3 Jt is for this square that the white wen has been aiming. Now 26...Wye5 {8 not possible on account of 27 Qf4. lack is forced to worsen still further é placing of his pieces. ae fed : 27° Wel 27 we3 wyaS 28 .d2 was also per- tly adequate, but there were already any ways to win. 27... Wes 28 Ad2 Pi6 29 Hacl Now everything is clear. The on- slaught of the white pieces is quite impossible to withstand. DD cows 30 fed 31 wh4 Threatening, incidentally, 32 4S. 3 bs Wb6 Wao 32 Axbs 33° fa 34 Qcé Black resigns, Out of his 33 moves, 13 were made by his queen. ¥¥b7 Abs Tal-Hecht Varna Olympiad 1962 Queen's Indian Defence Although the cascade of sacrifices in this game was linked by the single general aim of attacking the king caught in the centre, at the same time it was, as it were, divided into stages. ‘Thus, when he gave up a pawn, White ‘was not counting on further sacrifices, but was prepared to content himself with the initiative, And since a “primary” sacrifice does not exist — there is only the first, the second, and so on, so as the attack developed the aim also changed, 1 afte 2 e4 26 3 |B b6 4 2:3 Qb4 5 Ae5 ab7 6 3 ho 7 gh Axc3+ 8 bxc3 a628 Attack with Mikhail Tal 9 Hd2 eS 10 £3 we? ll e4 Sbd7 12 43 |p Black stubbornly avoids breaking the pin at the cost of any weakening, Earlier he could have chosen 7...25 8 $g3 e4, and now he could have followed Keres in his game with Botvinnik, 12th USSR Championship, Moscow 1940; 12...g5 13 £2 4yh5 14 g3 4)g7 15 ¥ye2 hS, gaining coun- terplay sufficient for equality. Now, however, by getting his blow in first, White initiates unfathomable complications, However, the move played is not necessarily the strongest continuation in this position. He could have obtained a favourable position by the "Spanish" manoeuvre 13 4f1 fol- lowed by 4)e3 and 4445 (or 4)f5). 13 Wya4+ was also good, and if 13..4d7 14 We2, And yet the move played is the one most in the spirit of the position: White is able to open up the game, for which he happily sacri- fices a pawn. 13. 5 dxeS On 13...bxc5 there could have fol- lowed 14 d5, when White gains control of c4 and b5, as well as the open b-file. 14 dxeS Wyxed Now White has a pawn majority in the centre, i.e. on the part of the board where he is intending to begin active play. But Black's defensive resources are, of course, quite adequate. 15 Wad+ 06? A mistake, after which White also gains the use of the d6 square for his attack, 15...€)6d7 was much — stronger; although here too after 16 ¥c2 White's initiative would have been quite ap-. preciable. 16 0-0 e6 After 16..4xc3 both 17 @a6 and 17 404 b5 18 Qyd6+ Gd7 19 Axb5 exb5 20 xb5+ are unpleasant for Biack. But now White has to hurry, since a position has been reached, where every tempo may be decisive: Black only needs to "hide" his king, and for the pawn White will have no compensation. If 17 (.g3 the strong is 17..¥¥e7, when 18 4)c4 fails to 18...b5. 17 |c4 wee Not 17...b5 18 4)xe5 bxa4 19 2)xg6 fxg6 20 e5, when Black can resign. 18 eS! As already mentioned, White's chief aim is to keep the king at e8. With his 18th move he begins a combination, in anticipation of... a favourable end- game. The position does not promise anything better. 18... bs 18...4)xh4 was interesting. After 19 &d6+ &f8 it is unfavourable for White to take any of the three pieces that are en prise: on 20 eyxh4 or 20 )xb7 there follows 20...8xe5. The eS pawn is more valuable than any of the black minor pieces, and therefore White had decided to play the posi: tional 20 Hael, since with one move the threats to the three pieces cannot be eliminated. (see diagram next page) 19 exf6l!King in the Centre 29 This move brings to mind the well fown game Lilienthal-Capablanca, ‘Hastings 1934/35, in which White, len a young master, quickly forced ¢ ex-World Champion's capitulation. Hecht made his reply without think- . Of course, anyone with the black ieces would be very happy to play 0-0. Then three of White's pieces re en prise, and if he moves his queen, 4)xh4 leads to a position where k is under no threat. But against 19.,.0-0 White has the ry strong rejoinder 20 Kael! If Yd5 there can follow 21 Wc2 xh4 22 Se5 with a powerful attack, While after 20..u¥xel 21 Bxel bxa4 22 Qxg6 fxg6 23 He7 Black cannot jay 23...F7 on account of 24 46. At this point the effusive Miguel jajdorf, who was watching the game, ime up and... kissed me (M.T.). Wo... bxad 20 fxe7 -g8 21 ast! ‘The culmination of the combination! Fter 21,..8/xc4 Black will be a whole een up(!), but he loses: 22 Hfel+ Web 23 Hxe6+ fxe6 24 Axg6+ Od7 25 Bdi+ Sc? 26 Ag3+ bb 27 Hbt+ Bab 28 4d3+ Had 29 Ac7+ mate. Meanwhile, 21...29xf5 leads to a hopeless ending: 22 4d6+ @d7 23 4\xf5 Oxh4 24 Badl+ Bc7 25 Axh4 Hixg7 26 Hifel, and the variation 21..4)xh4 22 Qxe6 fxe6 23 Qd6+ demonstrates the "agility" of the white knight. Black chooses the best defence, having in mind a counterblow on the next move. win ee Y ape | a 6 oe te | i 21 wes ayxhd 22 Axe6 fa6 How is White to save his piece? 23 d6+ = Be? 24 Acd! This concludes the combination, which has brought White a marked advantage in the endgame. 24... xg? 25 3 Bxd6 A mistake, since in positions of this type a bishop is stronger than a knight. Black would have retained some drawing chances after 25...Q.xc4 26 4yxc4 Has.30 Attack with Mikhail Tal 26 Axa6 = |FS 27° Babi £6 28 Hidl+ e7 29 Hel+ dé 30 SP ce Creating a shelter for the king at c5. Nevertheless 30...n5 was more tena- cious, for the moment maintaining the knight at its important post. 31 4 De? 32 Bb? Hags 33° xed = dS 34 Sxd5 — cxd5 35° Bb4 Ac8 Black should at least have ex- changed a pair of kingside pawns by 35.85 36 43 hxg4 37 hxg4 £5, al- though even this would hardly have saved him. After the move played, which leads to a "mutual elimination" of pawns, everything is clear, 36 Bxad— Axe3 37 Ha6+ Oecd 38 = Axf6 nS 39 1B yg 40 hxg4 Bh7 41 5 hs 42 BES Be2+ 43 O38 Bed 44° Blee5 d4 45 96 ahi 46 BcS+ d3 47 Bxe2 Bxc2 48 $f = figt 49 gs Black resigns: 49...&xg5 50 @xg5 d3 51 g7 d2 52 g8=W dl= 53 Wyb3+. WHAT WOULD YOU HAVE PLAYED? No.1 Black is attacking the knight that has crossed the demarcation line, To — where should it retreat? No.2 Here too the white knight has the right to choose...King in the Centre 31 But here there would not appear to White plans to get rid of the enemy be any choice: the knight must be outpost at d4, and then, according to taken and the black king will finally circumstances, either queenside cas- castle... tling with chances of an attack (({h6, h2-hé4 etc.), or else the quieter kingside castling. But it is Black to move.., No.4 No.6 gS 7 iB “Age aa, 4 “ a A A Black has just played 12...c5?, evi- After 15 Wxf3 Wxd4 Black gains a dently expecting the theoretical 13 tempo by attacking the white knight, Axf6r... castles, and then sets about realising his advantage. Correct?2 Breakthrough in the Centre A CERTAIN FRENCH GENERAL went down in history for the following dispatch to his commander: “My centre is broken, my flanks are retreating, I attack"...(I.D.) M.T. Mmm... This might do — in a tournament for beginners. But at our and your level of chess - hardly, True, a "retreating" or even absent queenside does not prevent a desperate assault on the kingside, but the domination of one player in the centre completely rules out activity by the other, Deviations from this rule are more likely to be found in studies, than in practical play. On the other hand, an advantage in the centre almost always allows an attack to be obtained. Whether in the centre itself, or on one of the flanks — this depends on the concrete features of the position. With a mobile pawn centre the strength of the attack grows greatly after a breakthrough in the centre, which usually achieves several aims: — the opening of lines, along which rooks and bishops situated on the queenside can instantly join the attack; often their scope is restricted by their own pawns in the centre; — the vacating of a central square, via which, for example, a slow-moving queen's knight can join the attack, or which, like a spring-board, can be used by a rook for a subsequent switching along the rank; — finally, the disruption, if only for a certain time, of the coordination of the enemy pieces. In short, a breakthrough in the centre is always the start of an attack, which, however, is by no means bound to be swift and explosive, but can grow by degrees, like the seventh wave in the sea. This is why voluntarily conceding the opponent a mobile pawn centre, or perhaps only a single pawn in the centre that is capable of advancing, is equivalent to voluntarily entering the lion's cage. This comparison is perhaps evoked by the following game, in which the player with White was a grandmaster by the name of Lev. {A play on words: dev is the Russian word for dion - Translator's note). Polugayevsky-C.Hansen European Team Ch., Plovdiv 1983 In the opening Black has clearly underestimated the possible pawnBreakthrough in the Centre 33 offensive by his opponent on the king- side, and stands worse. But never- theless, even with the two bishops, he should not have opened the centre and given White a dynamic pawn at d4, The correct 15...4.b7 (15...f6 16 g5, and the king's pawn shelter collapses) 16 exdS exdS would yet have given him hope of gradually neutralising the attack. But... 15...dxe4? 16 {&xe4 (gaining a tempo) 16...4b8 17 g5 h5 18 d5 The central files and the long dark- square diagonal are opened (18...e5 fails to 19 d6) and White's lead in development quickly decides. 18...exd5 19 4)xd5 Ac5 20 We3+ @h7 21 c2 Geb 21...4 is no better, being well met by the quiet 22 4e3. 22 4yf4 Yye8 23 Ayxh5, and without waiting for the completion of the attack after 23...Hg8 24 f6+ Axf6 25 exf6 followed by 26 4g5 and h4-h5, Black resigned. Of course, such gifts of fate do not occur in every round. For a break- through in the centre one usually has to battle — and pay. In the next diagram, the game has hardly passed out of the opening stage, but the pawn formation is already largely determined, and the two sides are considering their strategic plans. If the game were to take a quiet course, Black, after completing his develop- ment, could hope to use the strength of his dark-square bishop, which has no opponent. But.., White makes excellent use of the trumps of his position, including the advanced position of the black queen, to develop a powerful initiative by means of a central breakthrough. 5 mit = Chernin-Miles Tunis Interzonal 1985 12 e5! dxe5 13 2ye4 Wi4 14 Afd2! The white offensive is not "in a hurry", but now two of Black's pieces become targets for attack, and he is forced to lose time. 14.05 15 d3 (xed 16 Axed 407 17 g3 Wed 18 h3 wh5 19 dé The development of Black's kingside is hindered, his queen has been driven to the edge of the board, and his king is still in the centre (as is White's, but Black's is under attack). The pawn sacrificed is a miserly price for all this. 19...¥4g6 20 Hdl! The outpost at d6 is supported, and now 21 4)xc5 is a real threat. Black's tactical attempts to confuse matters are ignored by White. 20...b3 21 We2 Not allowing the black rook to move off a light square with gain of tempo (21 ¥xb3 Hb8).34 Attack with Mikhail Tal 21...f5 22 g4! cd 23 Gobi fxgd 24 ¥yxc4 Wi7 25 Yyc6 Hd8 26 hxgd The immediate 26 2c5 was also possible, since 26...44f3 fails to the bishop check at 26. 26...g6 27 A)c5 Hg8 28 Bed O27 29 &)xa6 Wyf4 30 Dc7+ Hi7 31 wed+ Sf6 32 45+ Black resigns. In this case the breakthrough in the centre was merely strengthened by the fact that the stranded black king was pursued, After a breakthrough, a piece offen- sive against the castled position is more common, as one of the authors experienced in the following game. Polugayevsky-Tal 37th USSR Championship Moscow 1969 This position had already occurred in previous games, and who could have known that, in the quiet of his study, White had worked out a variation giving him a decisive advantage, or... a forced win. 16 a5! exdS 17 eS! By breaking through in the centre, White has imparted a great deal of energy to his e-pawn, but more important — he has opened a way for his pieces to the kingside. 4 17...2)¢4 18 yyf4 Ab2 After 18...h6 White would have won| by 19 Yf5 26 20 Wh3 Se7 21 e6 fxeé - 22 4\d4, while if 18...—c6 he has 19 4)g5 h6 20 Qh7+ Bh8 21 Axh7+ Bixf7 (or 21...Qxh7 22 xd8 Bxf4 23 6!) 22 Wyxf7 @xh7 23 wyxb7. 19 Q.xh7+ @xh7 20 4\g5+ g6 21 hai! With the unequivocal intention of giving mate in the variation 22 h5+! HxhS 23 g4+ He6 24 wFS+ Pho 25. 4xf7+, and 21...f5 fails to save Black on account of 22 Bd4! (with the same. idea of h4-h5) 22..%¥e7 23 hS+ @h6— 24 \f7+ Bh7 25 wWxf5+ Hg8 26 e6 Exf? 27 exf7+ wyxf7 28 vyxf7+ Oxf? - 29 Bi4+ Begs 30 He7. 21..Bc4 22 hS+ Qh6 23 2xf7+ Bh7 24 wiS+ gs 25 e6 iG (there is no other defence against 26 h6) 26 Wyxf6 gxf6 And now, after inaccuracies by both sides, White won easily: 27 Hd2 (27 46 was more energetic) 27...He6 28 Hixb2 He8 (and here 28...9.c8 was more tenacious) 29 Sh6+ @h7 30 4S Hicxe6 31 Exe6 Hxe6 32 Be2 Hc6 33 He2 Qc8 34 Be7+ @h8 35 ah4 £5 36 Dg6+ Hg 37 Bxa7 Black resigns. In the next diagram White has achieved a strategically won position. There is nothing to support Black's counterplay on the queenside, whereas around his king there are many weaknesses, and his pieces (queen and knight) are far from ideally placed.Breakthrough in the Centre 35 Nevertheless, it is only by a break- through that White can capitalise on all the advantages of his position. after 39...c3 40 bxc3 wed 41 Of3 h4 42 wxhd vyxc3 43 Add wed 44 xaé. wey Late en J, iy ol a US Ne 3 is ae ca Junk S ( a Ribli-Gheorghiu Riga Interzonal 1979 27 eS £5 Otherwise absolutely all the lines are opened. 28 )c6 Also good enough to win was 28 e6 Se8 29 AxfS exf5 30 AxfS ws, when 31 4)xe7+ and 31 4xd6 are equally strong. But White prefers, firstly, to clear the e-file, and secondly, to include in the attack the only piece that is not participating — the knight atc3. 28..AxC6 29 dxc6 WS 30 Dd5 Hxc6 The exchange is lost in any case, and that is not all. 31 exd6 Hxel 32 Hxel Yyxd6 33 Del+ Bl 34 Dxch wyxc6 35 wed DIB 36 We7+ eB 37 Bdl Ace6 38 Hd6 we8 39 Gdl The bishop heads for d5, and Black can resign, which he did in fact do Szabo-van Seters Hilversum 1947 Although nominally Black has com- pleted his development, his position gives serious cause for alarm. Ideally aimed against the black king is the bishop at a2, which together with the queen can also operate on the b1-h7 diagonal, and the e-file is open for the white rook, Nevertheless, a rapid suc- cess can be gained only by increasing the number of targets to attack, and this is ideally achieved by a break- through in the centre. 17 d5!! Dxd5 The alternative capture 17...exd5 is refuted by the simple 18 bl g6 19 Hxe7 (this demonstrates the superiority of the 17 d5 breakthrough over the immediate 17 @b1). After 17...2a5 18 4e5 Hdd 19 Qb!I g6 20 Hb5 ¥yb8 21 d6 Black loses material, while the attempt to exploit the pin on the audacious white pawn, 17...Hcd8, is36 Attack with Mikhail Tal refuted by 18 Qxf6! Qxf6 19 we2!, exploiting the weakness of the back rank to create an irresistible threat against e6. 18 Gxd5! Wd8 The same idea, but with 18...Bcd8, is refuted by the thematic 19 Hxe6! But now the second white bishop is also attacked. 19 wea! Renewing the attack on e6 (19... fixg5 20 Qxe6!). 19..exd5 20 Axd5 Axgs 21 x5! Here we can sum up the outcome of the central breakthrough. The white pieces have closed in, and a weakening of the black king's position is in- evitable. In addition its main defenders have been exchanged, and the final assault now commences. 21...g6 22 wh4 h5 23 Df6+ wxt6 24 Bxd7 4d8 25 He8+ Gg7 26 &xf7+, and Black, who is losing his queen, resigned (26...£)xf7 27 £)e6+). LD. Every player has a "game of his life", a victory in which he seemingly moves onto a higher chess plane. You first became champion of your country, and at the same time a grandmaster, by defeating the highly experienced Alexander Tolush in the final round of the 1957 USSR Championship. M.T. Surprisingly, it was also the first time in my serious chess career that I made such an effective break~ through in the centre (see diagram next column), White has of course made more pro- gress in the preceding complications, and for the pawn he has a strong attack, But for 30 {1d2 followed by g4-g5 he has neither the time, nor the right: firstly, Black will establish a piece at e5, and secondly, counterplay on the b-file is already imminent. Be- sides, White must include his reserves — his knight and light-square bishop. All this is achieved with a single thematic move. Tal-Tolush 24th USSR Championship Moscow 1957 30 eS! BxeS j 30...fxg5 also opens the f-file for White, and after 31 (ixg6 hxg6 32 ¥Wyh8+ Sf7 33 Bh7+! he gives mate in afew moves. 31 Axg6 Hb7 For the moment both bishops are immune: the light-square one on ac- count of the same variation, and the dark-square one in view of 31..8xg5 32 (xh7+ @f7 33 ded. 32 Hed! : Another gain from the breakthrough. 32...fxg5 33 Bfl Bxe4 There is no other defence against 34 Dior.Breakthrough in the Centre 37 34 fixed Bg7 35 BE Axed 36 Bhft 4d7 37 Bxd6 we7 38 Axaé Wh 39 Axh7! Abs 40 165+ Hes 41 fe6+ Axe6 42 Axe6 Black resigns, VR fe ty 2a 7 eS ate Be } One 4 Vi, a ee Geller-Spassky 25th USSR Championship Riga 1958 Here too White has gained a clear Positional advantage, with the two bi- shops, the better pawn formation, and - the possibility at the necessary moment of a breakthrough in the centre, 23 Bee3! S\e6 (23...2\fxe4 can be met by 24 Qxe7 Bxe7 25 Ag5!) 24 eS! It is in this way, by ridding Black of his pawn weakness, that White imparts additional dynamic strength to his tooks, which dominate on the e-file, overloaded with black pieces. 24.524 24...dxe5 25 Bxe5 would also have Jed to the loss of a pawn, 25 hxg4 4)xg5 26 exd6 4\xf3+ 27 gxf3 O¢5 28 a4 The ending with two extra pawns is easily won after 28 9.04 Yyd7 29 Bxe8 Bixe8 30 Bxe8+ wWyxe8 31 d7 and 32 £.xa6. But for the moment White decides to stay in the middiegame. 28...%yd7 29 Bxe8 Axe8 30 xe8+ ¥yxe8 31 d7 Wis. If 3t..4¥e7 White has the decisive 32 f4, 32 tyez White is in no hurry to go into the pawn ending after 32 xf? ¥yxf7 33 d8=¥+, but creates the simple threat of 33 Wed. 32...We7 33 Yyd3 h6 34 Ged! Enticing the black a-pawn into an attack and effectively deciding the game, 34...a5 35 Axt7 Qha Or 35..x¥xf7 36 d8=t¥+ Qxd8 37 ¥yxd8+ Bh7 38 wWxa5, and the f3 pawn is immune in view of the exchange of queens. 36 Gg2 A £6 37 Acd wad 38 WAS Ags 39 wWe7 Black resigns. Against the threat of 40 wWe8+ @h7 41 4.d3+ there is no defence. Spassky-O'Kelly San Juan 196938 Attack with Mikhail Tal This game can be considered a mo- del example of this type of operation. 27 AxfS! gxfS The h5 pawn has been weakened, and a path towards the king from g5 has been opened, but... Black's centre has been strengthened, he has gained control of e4, and there is the prospect of his knight reaching there. 28 d5!! This breakthrough is the key to White's plan. At the cost of a pawn he occupies the centre with his pieces, and the next stage of the game commences. 28...Axc3 29 Yyxc3 cxd5 30 Ad4 wa7 The threat of 31 xf5 forces Black first to concede the queenside. 31 cS! 4yh7 32 b4 a6 33 ad! Hc8 Alas, 33...Wyxa4 fails to 34 c6! bxc6 35 Bal, trapping the queen. 34 bS axbS 35 axbS Als The black knight has hopes of occu- pying e4, but at the point when this is achieved disaster is all ready to strike. 36 c6 bxc6 37 bxc6 Wa8 38 Hcl £6 39 c7 Wa7 40 west White has tied down the black forces with his c7 pawn, and now, with his control of the centre, he switches his sights to the kingside, where his attacking queen is unopposed. 40.54 The dream is realised, but a longer resistance could have been offered by 40..4)h7 41 Bec2 @e7 42 Wes+ 28, although after 43 Ac6 White has the irresistible threat of 4)d4-bS- a7(d6); he can also play for a mating attack with 44 £3 followed by g3-p4. 41 £3 e5 42 fxed £4 43 gxf4 exd4 44 Bg2+ Black resigns: 44..@h7 45 ¥f3 Wd6 46 wyxhS+ wh6, and now the — simplest is 47 yxh6+ @xh6 48 exd5. Spassky-Tal Tallinn 1973 Nimzo-Indian Defence 1 a4 are 2 ¢4 06 3 Hc3 I took this to be the first sign of Spassky's aggressive intentions. When in a more peaceable mood he plays 3 £3, which usually leads to the rela- - tively quiet Queen's Indian Defence. 3 see Abs 4 aes Throwing down the gauntlet. This, the Leningrad Variation, which is quite often employed by Spassky, usually leads to a sharp and complicated battle. awe n6 5 aha 5 6 a bS The challenge is accepted. This sharp continuation seems to me quite logical, although Black also has other, quieter moves, such as 6...d6, 6...e5 or 6...xc3+. The move 6...b5 is akin to the Volga (Benko) Gambit, which has recently been played quite frequently. 7 dxe6 Very complicated play results from 7 e4 95 8 Qp3 Axe4 9 Qe5, which occurred in the game Levenfish-Estrin (USSR Championship Semi-Final, 1951). White also suffered a reverse in the game Korchnoi-Durasevic from the USSR-Yugoslavia Match (Belgrade 1956}. It was then that grandmasterBreakthrough in the Centre 39 ‘Tolush suggested for White a different plan, involving the win of a pawn, and subsequently he successfully employed it. This is the course that Spassky chooses. True, to me it seems more correct to say not that White wins 2 pawn, but that Black sacrifices it, 7 fxe6 8 cxbS d5 Now Black has an active pawn centre. 9 3 10 2f3 Better, perhaps, was 10 $4d3 d4 11 exd4 cxd4 12 a3, or the immediate 10 a3. It is on this possibility that the eva- luation of the entire variation depends. 10... wad Lo axf6 This exchange is forced in view of the threatened 11...dye4, Ho... Axf6 12 wd2 12 Ycl is more accurate, so that after a subsequent a2-a3 the black bishop should be immediately attack- ed, since the rook at al is defended, True, Black can reply 12...c4. 12... a6! 13 bxa6 A risky move. 13 b6 or 13 (e2 axbS 14 0-0 was more circumspect. 130... Dc6 Black, of course, does not take on a6, but intensifies the threat of ...d4. 0-0 4 fe2 Here too 14 cl came into consideration, in order to answer 14...d4 with 15 a3 {.xc34 16 bxc3. In this case White would have been in no immediate danger. ao... d4! 15 exd4 i... Eixf3 This combination leads by force to an advantage for Black. 16 Qxf3 17 00 Interesting variations arise after 17 Hcl. In this case I was intending to play 16...s¢xa6 (17...dxc3 18 bxc3 is unpromising) 18 {{xc6 d8, when Black must win, ¢.g. 19 wyc2 dxc3 20 bxc3 WyeS+ 21 fLe4. Here 21...9d3 wins easily, but at the board I worked out another variation that appealed to me: 21...Q.xc3+ 22 Wyxc3 Wyxe4+ 23 We vyxg2 24 ¥yxe6+ (the only move) 24..BHh8 25 web wYxc6 26 Hxc6 {4b7!, and Black wins a rook. More- over, it is amusing that, if White tries to give up his rook as dearly as possible, after 27 Hxh6+ gxh6 he is mated: 28 0-0 Hg8 mate, or 28 Bgl S03 and then ...d1 mate, 17... dxc3 18 bxe3 xed 19 wd6 Exa6é Of course, not 19....xal 20 ¥yxc6. 20° Axcé exd440 Attack with Mikhail Tal If the rook had moved from al, Black would have played 20...2d4. 20... Aba le YY Vy “of The concluding move of the combination. White loses his bishop at c6. 20..fe5 was insufficient on account of 21 ge7. 21) Wyb8 Hxcé 22 Gacl Acs 23° He2 Spassky aims for pressure on the c-file, but in White's position an Achilles’ heel comes to light — at £2. Possibly he should have looked for counter-chances by moving his rook off the c-file, for example to dt. 23... wad 24 wb3 24 fcl was not possible on account of 24... Q.xf2+. 4... wid Here I considered two moves — 24..84e4 and 24.4914. I rejected 24...W¥e4 in view of 25 Bifcl Qb7 26 wxb7 Axt2+ 27 Gl! (not 27 Bhi? Bxc2) 27..eyd3+ 28 x2 Bxc2+ 29 Bxc2 Yxc2+, when it is clear that the resulting queen ending with an extra pawn for Black is an extremely minor achievement. 4 Therefore ] switched my choice to 24..%9f4. Against 25 ¥yb5 I was intending 25...Wyd6, and if 26 fcl | 126, when 27 w¥ya5 is not possible: 27...Axf2+. 25 Wed 25 ¥yf3 was better, and if 25...eyxf3 26 gxf3 e5 27 @hi! b7 28 Bbl Bb6 29 Bxb6 Qxf3+ 30 Bg! 4.xb6 31 a4, when the ending is still unclear. Black, however, would not have exchanged queens, but could have played 25...29d6 or 25...¥c7, retaining the advantage. 25... wis 26 Hifcl Qb7 27° Wf 27 Sb8+ Ph7! (but not 27...2c8 28 wyxc8+ xc8 29 fixc5, when it is White who wins) 28 ¥yxb7 was not. possible on account of 28....¢.xf2+, but 27 h3 was more tenacious, 27 ase Wes 28 wb3 If 28 ¥yg3 the check at f2 is again’) decisive: 28....xf2+ 29 @xf2 (or 29 waxf2 eyxcl+) 29,.Bxc2+ 30 Hxc2. wise. 2B... eT 29 3 On 29 ¥yxe6+ Black would have won by 29..2f7, while if 29 ¥yg3 he again has the decisive 29....xf2+. The only move not to immediately was 29 ¥yh3. 29... Sixf2+ 30 Sxf2 ey f6+ I played this in accordance with my preliminary calculations. 30...4yf5+ 31 eg! we4 wins more quickly. loseBreakthrough in the Centre 41 31 Gel 32 «@fl 32 @dl wWd4+ 33 Wel wWel+ and 32 @f2 Bf7+ 33 Bgl wd4+ were wWes+ equally bad. 32... ab6+ 33° @gl Yydd+ 34 @e2 wed+ 35 el If 35 @h3 Bxc2 and then 36... ffl. 35... Sb7 36 hd wyhi+ 37) @f2 Efy+ 38 Ge2 wWed+ White resigns: after 39 Yye3 06+ 40 Sd2 d7+ he loses his queen. WHAT WOULD YOU HAVE PLAYED? No.7 White stands more actively, but how can he build up his initiative? Although there are signs that the position is double-edged (in particular ‘on account of the g-file), it is neverthe- less clearly in White's favour. Black's king prevents his rooks from uniting, but what particular plan should White aim for? And how should it be carried out? Nod The h5 and g6 pawns are weak, as well as the whole complex of dark squares around the black king. Where should the first blow be struck?3 The Assault Ratio THE PLAYER WITH A LEAD IN development is obliged to attack, or risk losing this advantage. This is a chess truth, formulated back in the last century by Steinitz, repeated by other classic writers, and in the intervening century merely confirmed on thousands of occasions in practice. Here is one single example: Geller-Korchnoi Candidates Quarter-Finat Moscow 1971 White has every justification for taking decisive measures, and indeed 24 QxhS gxh5S 25 Bf6! would have given him a very strong attack, for example: (a) 25..44d8 26 ¥yxh5 £)f8 27 Bh6 {5 28 Bh8+ @g7 29 who+ F7 30 Hh4, or (b) 25...2¥c5 26 Heat hxg4 27 Wxe4+ @f8 28 Bh6 Af 29 exf6 2x6 30 Hxf6 Ye7 31 Hf4! True, a collective analysis later found an improvement here — 27...@h71!!, and there appears to be nothing more thi perpetual check (28 Hed 4xf6 2! exf6 0.8). M.T. But firstly, I am sure that even. here White can strengthen his attack: and should win, and secondly, by dec. lining to attack he altogether concedes * the initiative... Indeed, after 24 h3 QfS 25 Qh? g7 26 He3 4c5 27 Wel Oc6 28 Axcé Yxc6 29 wh4 Hd7 Black ob- tained a counter-attack. White's cautious decision here was’ undoubtedly influenced by the fact that defence has changed fundamentally during this intervening century: the degree of resilience is commensurate. with the previously unknown rein- forced concrete, and defensive and counterattacking procedures have been found and systemised. Therefore a successful attack on the king is guaran- teed only when an especially strong piece grouping can be created in the attack zone, when the power of such an assault exceeds for certain the defen- sive possibilities. (see diagram next page) The pieces of both sides are in a way polarised. But although “phy- sically" White's pieces occupy squares on the kingside, for the moment they are not "looking" directly at the black / king. In short, White cannot manage without strengthening his attacking force.The Assault Ratio 43 Kasparov-Marjanovic Malta Olympiad 1980 17 eds! Axb2 18 4g5! White is already two pawns down, but now the concentration of his pieces clearly exceeds the defensive forces. 18...87c6 Black cannot eliminate the most dangerous attacker: 18...2e6 19 d6, with numerous threats such as 20 4)dxf7+, 20 eyc2 and 20 Yyhs. 19 Sye7 6 20 2)xh7! wd4 21 Wh5 26 22 wh4 There is no way of saving the game, and the players each make an “unnecessary” move: Black, in the sense that he might well have not both- ered to make one, and White, in that he can win "as he pleases". 22....xal 23 Hf6+ Black resigns: 23...@g7 24 eyh6+ @xf6 25 25 mate. In the next diagram, Black's consis- tent pressure on c4 has proved success- ful. In order to defend his pawn, White would have to give up his outpost pawn at eS, by capturing 14 exd6 en passant, but he can also sacrifice it, continuing his plan of concentrating his forces against the black king. i Kotov-Keres Candidates Tournament Budapest 1950 14 Qbl! White chooses the second alterna- tive, the more crucial and only correct one. Besides, in the given instance his strategic idea is supported by concrete calculation, 14.25 Of course, this is not a counterattack in the attack zone, but merely a de- fence, especially since 14...<.xc4 15 We2 26 16 G16! followed by ed2 or W£2-h4 cannot be tolerated. 15 ¥yc2 \g6 Black has eliminated the pin, and the attacking diagonal is blocked... 16 2f4t! Now all White's forces (not includ- ing his rooks) are in the attack, aimed at one point. 16...exh4 The alternative was to try and hold44 Attack with Mikhail Tal the g6 square: 16..%¥e8 17 4hS! Web (or 17...f5 18 exf6 gxh4 19 cxd5! exd5+ 20 $f2, and after regaining his piece — 20...@h7 21 Bei, with the threat of 22 Be7 — White obtains a material advantage with an over- whelming position) 18 cxd5 exdS 19 $1g3, with a decisive strengthening of the attack after the undermining move h2-h4, 17 &)xg6 He8 18 4yh8! While opening the way for the queen, the knight itself also remains in the attack. 18...e7 19 Yyh7+ BFS 20 f4 Axed It is not possible to evacuate the king: 20..e8 21 xf?! Bxf7 22 fig6 We7 23 5. 21 £5 exf5 22 0-0 The f-file is opened for the rooks, and the outcome is settled. 22.4808 23 QxfS Oxf 24 BxfS Be8 25 Axi? Had7 26 w5+ Heb 27 Wy6+ Hd7 28 e6+ Scé Or 28...6d6 29 Bxe7, and 29... ¥8xe7 is bad on account of 30 Af7+. 29 Bxe7 Wyxh8 30 Bxc7+! wb (30...8xc7 31 We7+ Sb8 32 Hfl) 31 We7 a5 32 tyd7+ a6 33 bl Black resigns. The concentration of force before the final phase of the attack is often made by outwardly imperceptible piece movements far from the epicentre of the coming events (see diagram next column). The material balance has been dis- turbed in favour of White. In the end- game this would be of decisive impor- tance, but for the moment, exploiting in the full sense of the word his piece advantage, Black launches an assault on the kingside. In a certain sense he even has two “extra” pieces: after all, the bishops are of opposite colour. Bata at soit oT] 7 ag ay A Y Stein-Geller Kislovodsk 1966 22...$.h4 23 He2 He7! z Black's jast fighting unit, not ready | for the assault, takes up an attacking position. Maximum concentration! 24 g3 Sixg3t 25 fxp3 WES 26 g4 The weakness of the king's pawn screen, the remoteness of the rook at a6, and, as a consequence, Black's decisive superiority in force leave | White without any chances of saving. the game. In the main variation 26 Bg2 Be7 27 Wd4 Bel+ 28 Ph2 wel he | also comes under a mating attack. 26..He7 27 Wd4 wyhd 28 Bal | Hel+ 29 Hxel ¥yxel+ 30 Gh2 we3+ | 31 Bhi Wxh3+ 32 gl (or mate in | three moves: 32 Bh2 Sg3+ 33 Gel Bfl+) 32...2)g3 White resigns. Here the rook was switched to the | decisive part of the battlefield along the 7th rank (or, using the old descrip- tive notation, along the 2nd rank).The Assault Ratio 45 White most often uses the 3rd rank for the same aim, prudently avoiding pawn advances such as h2-h3, but sometimes making use of the move a2-a4. In this case the concentration of force in the altack zone can grow very quickly. Kuzmin-Kondevsky Krasnodar 1970 White has already transferred one of his rooks into a close attacking posi- tion along that 3rd rank. But with only two heavy pieces, the black king's de- fences cannot be smashed, and mean- while his knight, splendidly placed in the centre and suppressing Black's counterplay on the c-file, is attacked. Tf turns out that the knight can be ignored, if the mobilisation is contin- ued! 21 Bdd3!! After thinking for an hour and a half (!), Black... overstepped the time limit. But neither after 21...Rxd4 22 Hdg3 followed by the sacrifice on g7, nor in other variations such as 21...f5 22 Hdg3 Af? 23 eyxh6 OFS 24 eyh8+ Be7 25 xe? Hxg7 26 wxg7+ Sd8 27 &xe6+ is there any way of saving the game. By coincidence, it was on the same move that White's rook made a deci- sive entrance in the following game. Yusupov-Kengis Moscow 1983 24 Ha3! @h8 22 Bg3 An accurate choice of target, al- though White could also have won by 22 Bh3 2.8 23 Bg3 (with the threat of 24 xh6 exh6 25 Wed) 23...2yhd (23...h5 24 SF6 g6 25 ¥yxh5+) 24 Hf and the black queen is trapped. 22...g8 23 2yc4 iming, in passing, at the d6 square, which the same dark-square bishop is obliged to cover. 23....d5 Or 23.....a6 24 Wed. 24 Bh3 1.8 25 b3 Threatening simply 26 exd6. 25....x04 This hastens the end, but what else can be suggested? 26 bxed dxe5 27 Wxe5, and Black resigned, since after 27...f6 (otherwise46 Attack with Mikhail Tal 28 &xh6 mate} his position makes a pitiful impression. LD. There is also another procedure for preparing a decisive assault, which has never previously been considered in chess literature. My co-author has suggested calling it "launching"... M.T. ... and in doing so I have no claims to authorship. This is pure pla- giarism, since in ice-hockey this con- cept has existed since the game was born.” The point of it is that the puck ends up close to the goal, but no one knows what will happen with it next. Who will gain possession of the puck, who they will pass it to, in which di- rection it will fly. At any event, when I “faunch” a piece close to the enemy king, I never aim it only at one point. LD. In your opinion, is such a stra- tagem objectively strong, or is it more a means of psychological pressure? M.T. One cannot, of course, deny the effect of surprise, and often after such an escapade a mistake follows, "It is a popular practice, particularly in Canadian ice-hockey, for a player to pass the puck not to a team member, but simply close to the opponents' goal, in the hope that one of his team will get to it first and will be able to shoot at goal. Similarly in chess, a piece may be moved close to the enemy king, but without any concrete aim, much depending on future developments, including the defence chosen, In the absence of any generally accepted English name for this practice, the term “launching” has een adopted, to convey its speculative nature (Translator's note), But that alse happens in any situation: when there is a sharp turn in the play. But the main thing is that it is "normal" move, and the "launching" is in no way. disproved. After all, this is one version of concentrating force, only without a clear and single aim planned before~ hand. Tal-Vasyukov 32nd USSR Championship Kiev 1964/5 One senses a certain remoteness of the black pieces from the kingside, and White airns to exploit this factor. 15 \hSt Typical “launching”. Approaching the black king at the length of a drawn sword, the Knight is looking both at £6, and at g7 — according to circum- stances. For example, 15...ae8 can be met by 16 c4 Ab4 (16.4516 17 )xf6+ )xf6 allows White to set up a strong piece outpost in the centre with 18 45, when 19 .xf6 and 20 47 is already threatened) 17 §.xh7+ @xh7 18 Bxd7 eyxd7 19 HeS wyd4 20 Hor gxf6 21 wyhS+ with a quick mate. InThe Assault Ratio 47 the game itself the knight goes the other way. M.T. Only one should not harbour any illusions: at the moment of “launching” the knight had already been allotted the role of a condemned man. 15..@h8! A useful move from all points of view. The threat of a possible sacrifice at h7 is not so terrible (without check!), and in the variation 16 04 4576 17 S\xf6 4xf6 Black is ready after a further exchange to play his rook to g8. 16 Ged £6! Again the best, and possibly the only move. If White has time to play 17 4, his advantage over the entire board will become evident. 16...f5 was « weaker in view of 17 Qxd5 Axd5 18 £4, when 18....04 fails to 19 Bxd7! 17 ghd White might have been tempted to win a pawn by 17 Qf4, but after 17...Qyxf4 18 Syxt4 eyxf4t 19 xb7 -Had8 20 Wyxe6 He5! any real advan- - tage would have evaporated instantly — 21 Axd8 Axd8 22 4xe5? is not possible on account of 22...uyxf2+! 17..6.d6 Very interestingly played. Black only needs to follow up with 18...4\c5, sand he will seize the initiative. _ Therefore White's move is forced, but in fact it had already been planned in outline at the moment of “launching”. 18 c4 Qa6 How should White — continue? - Defending the pawn with the rook is = inconsistent: after all, the pin is main- “\ tained. On 19 .d3 Black replies 18... SES 20 xls Qxf4 21 Wreb acd. Drastic measures are required, but 19 &xh7 is insufficient in view of 19...@xh7 20 Wed+ h8! 21 exes &xc4, Only one move remains: 19 &)xg7! xg7 (bad is 19.4 20 ¥d2, or 19...Axc4 20 Axe6) 20 4 d4 HcS 21 wyed+ Gh8 22 Zyxe6 axeG& 23 wyxe6 HaeS 24 wyxdS Axh2+ 25 @hi, and now the lesser evil for Black was 25...%yxc4! 26 Wyxe4 (ixc4 27 Hfel, although the ending favours White, Bot after 25....9f4 26 yh! White retains a winning advantage in the event of 26..4xe4 27 Hd7, and on 26...¢yxe4 (the only move) he could have won immediately by 27 Hdet. M.T. We once talked about “launching" at one of the sessions of the school run by grandmaster Gipslis and myself for young Latvian players, and fairly soon I was gladdened by 17- year-old Alexei Shirov, who by that time was already World Junior Under- 16 Champion. Shirov-Stangl World Junior Ch,, Tunja 198948 Attack with Mikhail Tal White needs to get rid of the one black piece that is preventing him from beginning an offensive on the kingside — the bishop at g4. With this aim, 1! 3 followed by g2-g4 and f4-f5 would have done, but a "launching" is even better, 11 HgS Axe? 12 wxe2 Now the white knight supports the possible breakthrough 13 e6, and it is hard for Black to endure it close to his king. 12...86 13 Syxf7t ‘The goal has arisen of its own ac- cord! 13...@x17 14 £5 gxfS 15 wyh5+ Ses 16 Raf e6 17 g4 wWd8 18 exis exf5 19 Hhel we7 Only White's knight is not partici- pating in the attack, and his advantage in force suggests a solution. 20 Axg7+! @xe7 Had Black given up his queen (20... Wxg7 21 Bgl) he would then have lost, much more. 21 Hgl+ {8 22 Axh6+ Axh6 23 Wxh6+ Ge8 24 Be7 wht Forced: 24..¥9f8 25 Web+ d8 26 Begs. 25 a3! Elegantly removing the defence from f&8 and forcing rnate after 25..%yxd4 26 Beg8+ @f7 27 wWh7+ Heb 28 He’. Black resigns. In the next diagram too, White makes concrete gains by a “launching”. 17 gS! The obvious aim is 18 ge4, but there is also a reserve one that is not yet evident. 17.85 Practically atter forced, since 17...ad8 18 g4! Black altogether has a problem: what to move? 3 Balashov-Stein 39th USSR Championship Leningrad 1971 18 Qh2 £5 19 fel Qes!? ‘The best chance -- to remove his: concerns over the d6 pawn. But... 20 gA! fxg4 21 2)xh7! Goal! Of course, talking in ice- hockey terms, the match is not yet. won, but even after the most fogical. 24...$215 (bad is 21...Qxh7 22 wWe5, and wins) 22 4)g5 White has a clear advantage in view of the weakness of: e6, to where his agile knight will soon: find its way. 5 21...g3?! 22 fxg3 4£5 Here too the capture 22... xh7 23 Wed Qd4+ 24 Bhi DL 25 wxe6+ @h8 26 AFL Axh3 27 Ax Qxf6 23” Wxfo+ Bg8 29 wha d7 (or; 29,..9.08) 30 2e4 would have given: White an irresistible attack, But in the game too, after the correct 23° @eg5! he should have won® “lawfully” -~ 23....4xg3 24 Hxed+The Assault Ratio 49 Hixe8 25 GxfS gxfS 26 Qxg3 Axg3 27 Bh2 HhS 28 Bgl, or 23...Axb1 24 Eixbl Sxg3 25 Qxg3 2xg3 26 Heb with a decisive advantage. In fact, he went wrong with 23 &.xf5? (giving Black counterplay by opening the g-file), and won only after inaccuracies by both sides, during which he was close to defeat. M.T. Which in no way affects the evaluation of the entire idea, begun on the 17th move with the "launching". Tal-Averbakh USSR Team Ch. Semi-Final Riga 1961 Ruy Lopez 1 e4 e5 2 | 4Nc6 3° abs a6 4 Sat |G 5 0-0 Ae? 6 Hel bs 7 Sb3 6 8 3 0-0 9 h3 Had 10 Sc2 cd il a we7 12 Sbd2 Ac6 13 dxe5 dxe5 14 }fl Rds 15 Wed £6 16 Se3 bs Black's 14...8d8 is clearly rather provocative, since his f7 square is ~ deprived of its central defender. But if White develops unhurriedly with @.d2, Bad1 etc., as in many well known games, Black's position will prove quite satisfactory. The knight does not threaten any- thing for the moment, but it clearly intends to sacrifice itself at £7, or — in some cases — at h7. 17... ALS 18 wf3 As though confirming _ his “sacrificial mood" — now there is no way back for the knight. 18... he7 19 4d5! _wWd6 After the capture on d5 the weak- ness of f7 would already have been felt — 20,..4)e7 is not possible. 20 Se3 h6 It was difficult to tolerate the knight any longer since 21 Had! was threat- ened, but now White begins a com- bination that he has calculated right to the end. 21 Dxf6+ = Axf6 Tt would have been better to part with a pawn after 21..4yxf6 22 Wxf6 SLxf6 23 HF3. 22 Hadi = wWe750 Attack with Mikhail Tat 23 QxcS! Axdl 24 Bxdl x5 25 wxf6 hes 26 Qb3! This quiet move is the point of the combination. The black king loses its pawn protection and comes under the fire of ali three remaining white pieces. 26... 8b7 27 tyxg6+ «Of 28 ¥yh6+ And in view of the variation 28...e7 29 Bd5! eyb6 30 vyxg5+ Os 31 who+ Be7 (Gl..Gg8 32 Hd3) 32 wWh4+, Black resigned. WHAT WOULD YOU HAVE PLAYED? No.10 Compared with familiar positions, in the opening Black has lost two tempi — this is too great a handicap. To what does it give White the right? No.1 All five white pieces are aimed at the biack king. Is some further strengthening required, or is it time to seek a combinational blow? .4 Invasion Trajectories INVASION TRAJECTORIES ARE the diagonals, ranks and files that in fact make up the entire black and white arena of the chess battle. But initially they are blocked by pawn barriers, of both sides, and increasing the mobility of his pieces is essentially one of the main problems for each player. For this the pieces need clear trajectories, For such privileges a player has to battle, sacrifice pawns and pieces, or make positional concessions. But let us nevertheless begin with some particular instances: the conclu- sions of attacks along trajectories that have already been cleared. The idea of an attack along a weak- ened long dark-square diagonal is il- _. lustrated in pure form by the following example, froma ladies' competition. Archakova-Andreyeva Kaliningrad 1970 The position of the black king is quite badly compromised. White can choose between a piece attack on h6 and the advance of her h-pawn, which, however, does not completely rule out Black's defensive chances, associated in particular with 28...%b4, But the latter preferred the mercenary 28... &.xc5??, after which the main line of attack became the al-h8 diagonal. 29 Wf6 He8 30 Hd8 tye7 So, has the diversion of the defender not succeeded? Has the attack been parried? 31 Yyh8+! Black resigns, BEE wie ifiie a i ig me £ Kasparov-Korchnoi Candidates Semi-Final, London 1983 In this, the first game of the match, the dark squares in the vicinity of the black king have been weakened, and although White no longer has his dark- square bishop, this weakening could and should have been exploited with the help of his knight, in spite of the$2 Attack with Mikhail Tal piece sacrifice involved. After 21 Sse5! the threat of 22 2)g4 (and, in passing, 22 cA) obliges Black to swim with the tide: 21..Bd8 22 Wh4! Hd5 23 @)g4 BxbS, and now the interposition of 24 Badl!, cutting off the king's escape to the queenside, forces him to find a way of neutralising the oppo- nent's powerful initiative, Thus, for example, 24...4)c6 is bad on account of 25 8d7!, while on 24... 4.d5 White has 25 04 Wxc4 26 Bxd5! exd5 (otherwise the queen is lost) 27 Wd8+ He7 28 wis SFB 29 Hel wes 30 Wh8+ He7 31 Hxed+ dxe4 32 h4, retaining the initiative. Perhaps the only way to save the game is in the variation 24...d5! 25 c4 Hd6! {the open file cannot be con- ceded, since 25...4xd1, in the hope of 26 EXxdl A\c6 27 Ad7 tyxd7! 28 Af6+ @g7 29 4)xd7 Bd8!, is refuted by the interposition of 26 4)f6+!! @f8 — bad is 26.497 27 2)e8+, when Black can resign — 27 Bxdl with excellent at- tacking chances; in addition White has in reserve the drawing mechanism 27 Wh6+ Ge7 28 Ags+) 26 Afo+ SB 27 WF Bc6 28 Wh6+ Be7 29 g8+ Ge8 30 |Hf6+, with a draw, which from the. initial position is not a bad result for White. After all, his queen- side pawns are organically weak, and in the game this factor led to his defeat after 21 )g5 h6 22 Hed Axedt In the next diagram the only open file has been firmly seized by White, and now his problem is to invade "deeper" and “wider”, breaking up the defences around the black king. 34 Hd6 Ded 35 Heb! yb] 36 Ah 46 37 Bdd6 (the first stage) 37... Ql? 38 Wal (the second stage) 38. Hg8. Ragozin-Lisitsin 12th USSR Ch., Moscow 1940 After 38...Afe8 White could have: changed the invasion trajectory with 39 wal. 39 Qxf6 Qxf6 40 wd5, and in view of inevitable loss of material (4G...Bge8 41 Bxe6+ Bxe6 42 Bxe6+) soon Black resigned. But it happens much more often that attacking diagonals and files have first to be opened — and in a very fierce struggle! Perhaps the most daring — and bril- liant! —- attempt to exploit the al-h8 diagonal for attack, after first weak- ening it, was made in the following © game (see diagram next page). At that time this was one of the stan- dard positions in the Sicilian Defence, where it was considered virtually ob- | ligatory for White to repeat moves: 12. Wh6 {97 13 eyh4 Qf6. But here he made a move that initially seemed fantastic, but then quite logical.Invasion Trajectories 53 g as , ERD 8 “o ae A Ree ay ene Nezhmetdinov-Chernikov Rostov-on-Don 1962 12 Wyxf6r?t The idea of the combination imme- diately becomes clear: with .d4, 4)d5 and Ral-d1-d3-f3 to set up powerful pressure on the f6 pawn, and after its fall — an attack along the diagonal. 12...2)e2+! (by diverting the knight, Black gains a tempo) 13 4)xe2 exf6 14 c3 He8 14...d5 seems better, in order after 15 4)xd5 to continue 15...d8 or 15... 46, intending to get rid of the aggres- sive knight. But White too can try and save a tempo by continuing 15 .d4!? 15 Sd5 He6 16 Qd4 g7 17 Had1 dé On 17...xe4 White can interpose 18 {4c3 and then 19 4)xf6, while if 17.,.b5 18 4.3 Yyd8 19 &xf6 Bxf6 20 Bd6. 18 Ad3 Qd7 19 Bf3 Ob5 20 Ac3 Yyd8 21 4)xf6! Ge2 Black loses after 21...Qxfl 22 2gd+ Hg8 23 hor GLB 24 Axi7 ~ We 25 He5+ Be8 26 Hxe6, when the only way to prolong his king's life is by giving up his queen, 22 Exh7+ He8 Or 22..Qxh7 23 Bxf?+ @h6 24 Qd2+ g5 25 GQxe6 Axfl 26 Af5 wh8 27 h4 Qe2 28 Axg5+ BhS 29 3 xf 30 gxf3 wyd4+ 31 Bg? Ah8 32 &f6, and the threat of 33 424 mate forces Black to part with his queen. 23 Gh3 Bes Alas, blocking the fatal diagonal (24 4g5 was threatened) does not prove possible. 244! Qxfl 25 Sxfl Ac8 26 Bd The hasty 26 fxe5 dxe5 27 Gxe5S would have allowed Black a draw by perpetual check: 27...44d1+. But now the knight again moves into a striking position, 26..b5 27 )g5 Be7 Black cannot also plug the neigh- bouring diagonal with 27...8.c4. White first captures both black rooks with his bishops(!), and then carries out the same "little combination" as in the game, And if 27..49f6 he wins by 28 fixf7+ @g7 29 Bh7+ BB 30 Ae6+ Be? 31 Axg6+! HxeG 32 (5+, 28 Qxf7+! Bxf7 29 Bh8+ &xh8 30 &)xf7+ @h7 31 )xd8 Hxed 32 Ac6 xf4+ 33 de2 Black resigns. This is another example from the games of a player, who is not too well known in Europe, international master Rashid Nezhmetdinov... MT. ... Against whom, incidentally, I lost three of our official games, al- though I should have lost all four. And how [ lost thera! LD. Don't worry, we have already given one of them...54 Attack with Mikhail Tal Nezhmetdinov-Kotkov Russian Ch., Krasnodar 1957 The position of Black's king has cer- tainly been weakened by the advance of his pawn to £5, but how can this factor be exploited? 17 €yf4 is met by the simple 17..Qe5, and under the cover of this knight Black gradually completes his development. 17 &xc7!! Wxe7 18 WydS+ hs 19 Bes! Of course, not 19 Qxd6? 4f6!, when the picture changes sharply. 19...4)f6 26 HxfS+ Axf8 21 @b2! White plans — and carries out! — an invasion of the opponent's position along the a2-g8 diagonal, but in this he is helped by pressure and latent threats along the adjacent diagonal. 21.287 After 21..g7 White realises his plan in pure form — 22 04, with the invasion at 28, Black has a seemingly more tena- cious, and certainly more cunning defence in 21...49g7, when if White is tempted by the plausible and showy 22 Be8 Axe8 23 WE7!!, he runs into 22.. Ac6! 23 Axf8+ Hxf8 24 wxeé d3 2. 24 fxg4 26 hxg4 h6 27 f4 95, wher Black gradually frees himself. But instead of this, White had pre- pared in advance the essentially forced variation 22 ¥yd4 4)e4 (22...6)28 is met by 23 He8 wyxd4 24 Qxd4+ Op" 25 c4, while if 22..Qe7 23 wed S118 24 Wye5) 23 £3 dd 24 fxed fixed” 25 WF2!, which Biack can avoid only at the cost of a hopeless ending. Q 22 Seat Qd7 23 Axf6 Axf6 24 WIT ds 24...0.95 is also insufficient: 25 g3 Hc8 26 h4 Qc6 27 He8+ Axed 2: ¥yxc7, and White must merely avoit blundering into the mate after 28. Hel+ 29 @h27? (29 Qfl). The move in the game allows a pretty finish. 25 He8+!! Black resigns. He cannot simulta- neously defend f6 and 8. 8 So that the timely opening of a di- agonal brought the attacking side rich — and pretty, if they can be such dividends. 8 It would be naive to think that a good friend of both authors, five-times= Women's World Champion Nona Gaprindashvili, did not know about. this attacking procedure... ° ID. The point is probably that knowledge comes in different forms. Some, that was once assimilated, is stored somewhere in the depths of the brain, and is soundly asleep... M.T. Of course, no grandmaster, on’ sitting down at the board, repeats in his: mind the entire range of chess knowl: edge, but for me a sign of good form is when it is all "spontaneously", as itInvasion Trajectories 55 were, available, and what is needed surfaces of its own accord at the required moment... LD. But how can this be attained? M.T- Nothing, apart from constant training, has yet been thought up by mankind. Neither in science, nor in art, nor in chess... Gaprindashyili-Vereczi Belgrade 1974 Here White is attacking, Black is counterattacking, and it is not clear which is the more dangerous. It is White's move, and she has a draw in the rook ending (36 w¥f7+), but she has also discovered a mating motif. 36 h3t! Axfl+ 37 @h2 Here there is no point in giving any exclamation marks: White is merely continuing what has been begun. The threat is 38 hxg4+, with mate to fol- low, and 37...gxh3 allows 38 g4+ h4 39 Wyxf6+, mating. 37...g3+ 38 fxg3 We3 Black has taken control of the inva- sion square hG and is herself threaten- ing mate in one move. Therefore White gave perpetual check — 39 wye4+ @h6 40 ¥yg7+, although she could have opened one more diagonal for a decisive attack. After 39 Hxe5+ fxe5S 40 g4+ @h4 the game would have been concluded by 41 ¥e7+, and if 41..¥¥25 42 g3 mate. A finish, worthy of any textbook, in any era! The opening of files is virtually the main motif in the majority of games where the target of one side is the enemy king. Lines can be cleared by exchanges, but more often — by sac- tifices. °y wae DAR cane wa ee al Geller-Vasyukov USSR Spartakiad, Riga 1975 White has an advantage in mobili- sation, and in the placing of his king. In addition Black's position has many weaknesses (the d6 pawn, the d5 and f5 squares, and the kingside pawns), but it is very difficult to approach them. Apart from one way... 16 £4! exf4. 17 Qxf4! gxf4 18 Rxf4 &n7 At the cost of a piece White has opened fines on the kingside, exposed a56 Attack with Mikhail Tal glaring weakness at f7, acquired an outpost at d5, and his attack on the king becomes threatening. 19 AdS Yds 20 b4 Ga7 21 eS Me6 (21...dxe5 is decisively met by 22 Hxf7!) 22 exd6 He6 23 Bed (again aiming at f7) 23..8xd6 24 Qxe6 Bxe6 25 |l6+ Ye? 26 Had+ GiB 27 Radfl Hg5 28 h4 Ad8 29 Bed BeS 30 Hdd 4)e6 (30...2)e4!? came into con- sideration, creating counterplay) 31 Wee wye8 32 Ad3 Axd5 33 Bxd5 Af4 34 Wf He8 35 Bd8, and Black played on for a further ten moves, only because it was a team event. Kupreichik-Belyavsky Chelyabinsk 1974 White has an enormous spatial ad- vantage, and in addition virtually all the black pieces are practically stale- mated. Truc, had it been Black to move, with a pawn sacrifice he could have gained a “gulp of air" — 23...04 24 Axed Qf6 25 Md4 Axd4+ 26 ¥yxd4 (6 27 d3 fe’, and after White's "normal" move — 23 e4, he could have tried to set up some defensive lines with 23...f6, which, however, would hardly be sufficient. But White (and it is in fact him to = move) is not inclined to defer his as- — sault, and he resorts to a typical attack~ ing device, opening the f-file and the bl-h7 diagonal. 23 £6! Qxf6 24 fed Bhs Roughly the same variations would have occurred after the alternative 24,.,.ce8, 25 Axf6! gxf6 26 Gxh7 Ags 27 fl Be7 28 9.e4 £5 29 Q xf5 4 xf5 30 HxfS gs Here Black evidently breathed a sigh of relief... 31 Qh6 Hg6 32 Hes! Exchanging one of the three attack~ ing pieces for the only defender! After this the black king is completely ex- posed, 32..8h7 Or 32...xg5 33 LxgS Hc 34 06 f8 35 Yh7 Bek 36 wese. 33 Hxg6 fxg6 34 wh3 gg8 35 ¥yd7 Black resigns. Kostina-Sammul Tbilisi 1974Invasion Trajectories 57 Every single white piece is ready, either immediately or within a couple of moves, to take part in the attack on the enemy king. But the latter has enough — or nearly enough — de- fenders, there are no serious weak- nesses, and, most important, for the moment the position is a closed one. Hence White's immediate task —- to open lines! 24 4)g6! hxg6 It would have been preferable to decline this Greek gift and to prevent the immediate opening of lines by 24....x26 25 hxg6 h6, although even here after 26 )f7 Hxf? (evidently forced) 27 gxf7+ @xf7 28 Axf6 Axf6 29 £3 exf3+ 30 Wxf3 followed by e3- e4 White stands clearly better. 25 hxg6 (ixg6 26 Hahl 4yh5 27 BxhS Qxh5 28 YxhS Axgs 29 fxg! Not the only solution, but the most consistent: the pawn screen is com- pletely destroyed, Black's king is “un- dressed" and, strictly speaking, she could already have resigned. 29.,.e5 30 ¥yh7+ (other moves were also possible) 30...@f7 31 @xf8+ eS 32 Ac5, and White won. It is usually another gift that is of- fered at g6, when the g-pawn, suppor- ted by its neighbour on the h-file or sometimes even unsupported, moves into a double attack. However, this procedure is exam- ined in detail in one of the illustrative games in this chapter (Tal-Mohrlok). For the moment here is another exam- ple of the opening of lines, and of how the pieces can exploit this. There is nothing to herald a storm, except perhaps that one senses a rather abstract weakness of the a7-g1 diago- nal. And yet... as oH |Z. ae fe Oe emonae vs Oe Fa % & Mititelu-Tal World Student Team Ch, Varna 1958 17...yg4'! This is not a queen sacrifice, since 18 {xd8 af2+ 19 Bgl Axd1+ 20 @hil |Af2+ 21 Hel Hexd8 gives Black both material compensation, and an attack. His idea was a different one, based, incidentally, on White's accu- rate reply. 18 fxgd Wxg5 19 exdS 24 20 d6 Qxe2 21 2xe2 )xd6 22 Hxd6 4b5 23 Wyd2 Wed 24 S\c3 Axd6 25 Wxd6 Hied8 Forcing play has led to Black's de- sired result. The disturbing of the ma- terial balance is nominally not in his favour, but on the other hand all the central files are open, and the black rooks are established on them. Strictly speaking, there is no attack as yet, but with every justification Black can count on one, as soon as his rooks advance, ideally onto the 2nd rank.58 Attack with Mikhail Tat 26 We? hS 27 Bfl Bd7 28 wxaS h4 29 gxh4 Rd2 Black's idea is realised in pure form, although in passing he has to parry some counterthreats by the opponent. 30 2.5 PhS 31 eyb4 wh3 32 wea Bxh2+ 33 Bel Be2! Pawn-grabbing (33...xb2?) would haye given White an important tempo for defence. 34 dl Reb 35 hS Af6 36 wb8+ Bc8 37 492 Bxfl+ After this Black can “sadisticaily" remain the exchange up, or (after in- terposing a check) with an extra queen. White resigns. Minasian-Miles Moscow 1990 One senses that the black king has already endured some anxious mo- ments, but... Now the tripled g-pawns hinder the approach to it, for example, 20 £3 gxf3 21 Wxf3+ 4yf5, and the king is clearly destined for a tong life. But White nevertheless finds a way of opening both files and diagonals for the attack. 20 3! gxh3 21 Qxg6+!! Sxg6 22 ta+ Sf7 23 whs+ Wg8 24 Axes We8 25 Hdgl! Gaining a tempo for the decisive mobilisation of reserves. 25... 18 26 ¥yxh3 There is no need te resort to drastic measures such as 26 g6: Black has practically no useful moves. 26...£)f7 27 WiS Bc8 28 Bxh8+ ° £)xh8 29 ]xe7! Qxg7 30 Qhé There are many black pieces around the king, but none to defend it. 30...Acl+ 31 @xcl! Black resigns. - Diagonals often have to be cleared not only of enemy pieces, but also of one's own. Here, as in play along the files, it can be useful to vacate a square for the piece that can occupy it to greatest effect. Time in attack is precious, and lines are normally cleared with gain of tempo, i.e. with a sacrifice. Schmid-Kinzel Siegen Olympiad 1970 A rook on the 7th rank is always an achievement, but in the given instanceinvasion Trajectories 59 it has no particular prospects, and White also has to concern himself about his knight. He could, of course, try to set up a queen and rook battery along the h2-b8 diagonal, but is it not simpler (of course, knowing this procedure) to clear this diagonal. 24 Bxb7+! wxb7 25 wWes+ Gas Black has little choice: 25...%c8 26 Hicl+ Qd7 27 td6+ Ge8 28 Ac7+, and he loses his queen. 26 Sc7+ Sb8 27 4)xe6+ Has 28 &)xd8, and White easily realised his material advantage. White was also able to extend his attacking trajectories in the following game, Uhimann-Uitumen Palma interzonal 1970 White has the initiative for the pawn, but no direct strengthening of the attack is apparent. In addition the white queen is tied to the back rank. Of course, the black king could be ap- proached by giving check on the al-h8 diagonal, but the 3 square is under » control. 27 bat Whether or not Black accepts the sacrifice, all the same one square, al or b2, is opened for the white queen, 27.89 xb4? The only defence was 27...¥9d8, in order to oppose the white queen from f6, But now Black loses. 28 Axe6 And it transpires that after 28...4.xe6 29 wWal+ the black king is unable simultaneously to defend the rook, the bishop and itself. Hence — 28..Wb2 29 Yxbl Axbl 30 Ad2, and a gradual loss for Black. Vacating a file before one's strong- est piece (usually the queen) moves onto it — this theme is well known from the following textbook pattern, which, however, occurs from time to time in practice with various modifica- tions. Both rooks are sacrificed at h8, and then comes yh1+ and wh7 mate. Therefore we will restrict ourselves to one elegant example. White's queen very much needs to reach g7, but both his own rooks are in60 Attack with Mikhail Tal the way. In addition he has to worry about the c2 square, where he may himself be mated. Z Heemsoth-Heisenbutter Germany 1958 White found a splendid solution. 1 Best! Interrupting Black's counterattack for an instant, with gain of tempo White prepares also to dispose of his second rook — 2 Exh7+. In addition Black's queen is attacked, and he merely deferred his capitulation for the moment by defending his king — Lu.Aje8 2 Hxed Bxed 3 Bxb2. It should be borne in mind that -~ mainly in the games of top players — one also sees another, rather rare form of vacating lines for attack — in this case ranks. This happens when the attack on the king is being mounted from the flank, or when an attacking piece has to be switched from flank to flank. One has to look hard for such a possibility: it is usuaily well camou- flaged and... effective, as in the fol- lowing example, Karpov-Gik Moscow University Ch. 1968/9 Both sides have been playing ac- tively, but now it is only White who i: attacking. His offensive, however, would appear to have come to a halt, since after 24 ¥yxh7+ Qf8 the black king escapes from the firing line, and the knight cannot be included in th attack: 24 &)\d4 Wydl+ and 25...¥yxd4. Black's position proved to be hope: less only after the following far from obvious decision, 24 got! / Usually files are opened by such an’. advance, but here the white rook's scope is expanded, not only forward, but also to the side, along the rank. Its. increased power proves decisive. 24... )xg6 25 tyxh7+ SEB 26 HES! Winning the queen and the game. 26...%yxb34+ 27 axb3 exfS 28 Df4 Hd8 29 Wh6+ eS 30 Axg6 fxg6 3 tyxgé+ He? 32 vyg5+! Not allowing the rook across to th kingside. 32 Be 33 exlS He8 34 wyg8+ HeInvasion Trajectories 6L 35 ¥yg7+. Black resigns: the f-pawn will now advance. a ¥ ws, ES ] ARGS te Stein-Tukmakov USSR 1972 This illustrates a similar idea, but much simpler. White gives up the pride of his position — his passed d-pawn, in order to open up the 7th rank. 38 d8=ty-+! Bbxd8 39 Wye7+ The white queen has joined the at- tack, and Black resigned. Larsen-Kavalek Lugano 1970 The position is sharp and full of life. White is pressing on the half-open d- file, and the h-file has also been opened "in his favour". But Black too has his play: he is threatening to drive away the knight by ...25 and establish a piece outpost at d3, attacking the important dark-square bishop. The scales tip in favour of White only after his far from obvious clearing of the 2nd rank. 20 g4! 2xe4 21 £3 exf3 22 4 xf3 At the cost of a pawn the way for the queen to h2 has been opened. 22...2)e5 Switching to defence — 22...2)f6 23 Wh2 Bfe8 — is insufficient: 24 4)b5 8 25 QL xf6 Axf6 26 Wh7+ HP 27 {dS and White wins (he threatens, in particular, 28 9.xf7 Wyxf7 29 2)xg6+). 23 Wh2 A xcdt? An interesting counterattack, for which, however, Black does not have enough resources. But pure defence would have led to variations similar to those given earlier. 24 bxed 2)xf3 25 wyh7+ Sf7 26 4cdS Bg8 27 4xe7 Hb’ 28 Hal Wxe7 29 Wyx26+ HIS 30 Heb+ eyxe6 31 Sixg7+! Ge7 32 Ai8+! BhxfS 33 Bh7+ And in view of the variation 33...8f7 34 Gixf7+ fyxf7? 35 wyxd6+ He8 36 Wd mate, Black resigned. Holes in the Fortress Walls Thematically linked to the attack along a weak diagonal is the attack along a whole complex of weak squares in the vicinity of the enemy king. This can arise with numerous pawn formations:62 Attack with Mikhail Tal And in various other cases, But in principle there is a single attacking mechanism; pieces are established on the unprotected squares, and from close range they pursue the king, completely destroying its protection. In the next diagram, the white bishop, which has just captured a rook, is clearly out of play, and there is no point in spending time on its recapture. It is much more effective to begin an attack on the dark squares, which for the moment have been left undefended, Botterill-Tal European Team Ch,, Bath 1973 41...Qe3! 42 Hed If 42 Hat Black interposes 42.. Agl+, and only then plays 43...4yxa1. 42.885! This bishop has a great future: if 43 W2, defending against the mate at £4, then 43...Q.h4+, and taking the bishop allows mate on gS. 43 Wed : 43 Wed h5 44 wyd4 is convincingly met by 44..Wel+, and if 45 @h2 Se3, all on those weakened dark _ squares! 43..2e3+ 44 Sed ahs Simultaneously creating two mating threats. 45 Qe7 x 45 4)xf7 is nicely refuted: 45. We3+ 46 @hS Bes! 45...S.Ke7 46 2)xf7 Or 46 Wxc6 hS+. 46..05+ 47 Gxhds Le8 48 Pet. exf5+ 49 @xf5 g6+ M.T. Here Botterill smiled: evi- dently the finish appealed to him too...Invasion Trajectories 63 50 ga Gd7+ White resigns; this attack along the weak squares won Black the brilliancy prize at the European Team Championship, Liebert-Tal Skopje Olympiad 1972 In principle the theme is the same here, with the not especially important difference that the role of the g2 pawn is played by a bishop. 22..4)de5! 23 fxeS 4 xe5+ 24 Ol Wes 25 HES Hh4 26 Ayxh4 wh2+ 27 PA 23+ 28 BF Axhd ‘The first phase of the attack is com- plete. The white king has been drawn out into the open, and is “ready” to ~ perish after the rook invasion at 23. Covering this square by 29 4f4 does not help, since all the same there fol- lows 29...Rg3+. 29 S.dd+ O66 30 wir 30 G.xf6+ Hxf6 31 ¥yf2 would have given a respite, but only for an instant —31...f4. 30...e5! 31 Bhd All the same the weak dark squares cannot be defended. 31...f4+ 32 ed wyxd4 33 wyxd4d ixd4 34 9 £3 He3 Alas for White, the theme of Black's entire play in this game remains un- changed to the end. 35 b3 cS 36 Hefl He7+ 37 Sd2 He3 38 Qdl Hg2+ 39 Yel Bc3+ 40 @b1 {a3 White resigns: dark-square strategy! Damsky-Stefanov Kazan 1962 White's superiority is determined by his spatial advantage and his invulner- able pawn outpost at e5, but most importantly ~~ by the weakness of a whole complex of dark squares on the kingside, of which the white knight has already made "convenient" use, On the kingside, castling is practically impos- sible, on the queenside it cannot be done without losing the f7 pawn, and it only remains for Black to castle artifi- cially under the cover of the blocked pawn centre.64 Attack with Mikhail Tat 16.7 17 pat When lines are opened, the weak- ness of the dark squares will become a decisive factor. 17...nxg4 18 Qxg4 Sc6 19 Baht Ad7 20 bS gxhS 21 AxhS HxhS 22 Sixh5 Hic8 23 Hh7 Des Are attack and defence balanced? if 24 QhS Black bas the sufficient de- fence 24.,.4c7, then ...b6 and ...9b7. True, in reserve White has the advance of his pawn to £5, but the position is already ripe for a “little combination". 24 QxeG! DxeG 25 Hxeb wyxeb 26 Hh6 Q.d7 27 Hxe6+ fxe6 28 wf6, and in view of the weakness of the e6 pawn, Black was unable to set up a "fortress". Kalinkin-Nezhmetdinov Vologda 1962 All the indications suggest that it is White who is attacking. His minor pieces are actively placed, the dark squares around the black king are weakened, and, for example, in the event of 17...xh6 18 S)xh6+ @f8 19 £4 2\c6 20 g4 Hf6 21 g5 this will soon tell. But, it tums out, Biack too has the preconditions for a counterattack — or attack? —- and they too are based or exploiting the weak squares, which in the vicinity of the white king... do not” yet exist! 17...c6! 18 4yc7 |)f4t Of course, 18..4xc7 19 QxgS would have surrendered the complex of dark squares to White. But the ex- change sacrifice enables Black to seize _ the initiative, which, however, would not have led to anything without th idea mentioned above. By threatening the knight at £5, Black begins forcing play. : 19 Axes Wx~5 20 Hg3 hS 2) ®xa8 Hxa8 22 tyxd6 hd 23 ad White does not even suspect any” danger, especially since after 23.,.hxg 24 hxg3 4)h5 he, of course, was not intending 25 £49? ¥yxg3 26 fxeS Ayfd 27 Wd2 4)h3+, when he has to agree to perpetual check, since 28 @f1 lose: to 28...@.e6 29 b3 ¥yh2 with the threat of 30...4%¥g1+ or immediately 30. g4. Instead, the simple 25 axb5 cx 26 9.b3, with the threat of 27 Hxa6, would have given him a material advantage and good prospects. 23.030 It transpires that it is not the whi knight that is attacked, but the securel; defended g2 square, and — all # suddenly weakened light squares on the kingside. Surprisingly, White no longer has any satisfactory defence. 24 axbS cxb5 25 Hxaé Desperation, but 25 Qd! hxg2 2 F3, not allowing the black bishop to gt to g4, is refuted by the simple 26. S)fd3 27 He2 h3!25...H{xa6 26 tyxa6 hxg2 27 Adi ed 28 wal Gf3 29 wa8+ Bh7 30 Yye8 Axdl White resigns. The weakness of squares around the king is especially acute if their natural defender —~ the bishop — is no longer on the board. The attacking side may exchange it, but even more effective is an attack with the participation of an “opposite-colour" bishop, since in this case it does not have any worthy oppo- nent. ED. In short, in the endgame oppo- ite-colour bishops are a saving fac- tor... MLT, ... but it is no accident that between the opening and the endgame the Gods have created the middlegame. "mas a wre BRD om WADE ite YABB Sax-Hebert Rio Interzonal 1979 The plans for the two sides are de- termined: after castling on opposite ides, mutual pawn storms are in pros- pect. To the usual principle in such ‘cases of "who is quicker" is added here — and this is decisive — the factor of opposite-colour bishops. PHT Eee eee Se Invasion Trajectories 65 Incidentally, White took this into account beforehand, when he aimed for this position. 14 Qg5! b4 15 Qxf6 Qxf6 16 &d5 Qxd5 17 Qxd5 Hac8 Contrary to the usual state of affairs, on this occasion simplification has favoured the attacking side: he has an “extra” piece. For the moment the counterattacking side does not have an "extra" piece — the bishop at f6 is rather passive. And the interposition of 17.,,.g5+ does not help: 18 gb! Hac8 19 h4 §h6 20 f6 g6 21 h5, and in principle Black can resign. 18 hd Wh8 19 Hb wye7 20 Bb3 aS 21 g4 Qe7 22 25 £6 And immediately the idea takes shape of a mate at g8... 23 ¥yh5 Q.d8 24 Hhel We7 25 Hg3 ADd6 26 Hd2 Qd4 27 Hdg2 Now the black bishop too can par- ticipate in the attack, but White has achieved more in the concentration of force. 27.084 28 Le6 No deviation! 28..H1b8 29 exf6 vyxf6 The light-square bishop would aiso have triumphed after 29..gxf6 30 wed. 30 Fxg? Wxg7 31 Bxg7 dxg7 32 WeS+ Wh8 33 hS Ab7 34 h6 AcS 35 £6 EET Against 36 ¥g7+ there is no other defence. 36 d5 Hbc7 37 eyed Black re- signs. In this last example we have seen the role that pawns can play in the offensive against a king. Apart from the fact that pawn advances open66 Attack with Mikhail Tal attacking lines and break up the walls of the king's fortress (about which more in detail in a later chapter), they also secure space for their pieces, and cramp the opponent's forces and the enemy king itself. Normally all this is at the cost of their own lives. Geller-Gligoric “Match of the Century" Belgrade 1970 At first sight White appears to have no grounds for a pawn storm, and in addition Black is just one move away from being perfectly safe. Yet with the help of a sacrifice this becomes pos- sible. 23 4)xe5! HxeS 24 Axes wyxeS 25 f4 White's idea, in launching this op- eration, is that the rapid advance of his central pawn pair will drive back the black forces to unfavourable positions. 25...04e6 Here Black's queen comes under a further attack by the white f-pawn, but what can he do? After the "active" 25...We3 26 W2 the queen would simply be lost: 26...2\c6 27 e5 He8 28. Be3 Wyb2 29 2\xh6+ gxh6 30 Ah7+. 26 e5 Hes Trying to hold the position by tac cal means, 26...\e4, does not succeed: 27 Hixe4! WxfS 28 Heel tye6 (or 2! WyhS 29 ¥yd3) 29 £5, and the pawns. continue their formidable advance. a 27 Hh4at The most energetic. Now the weak- ness of Black's h7 becomes fatal. 27..4)¢6 28 Wd 26 29 £5 The culmination of White's entire plan. Now the black king is opened up, and the attack continues as though of its own accord. 29...8x5 30 4)xf5 Threatening mate: 31 e7+ etc. Black's reply is forced, since after 3 e7 31 &xg7 xg? (31..2xg7 Wh7+ @f8 33 wyh8+) 32 wh7+ SFB 33 {2f5! the queen has no retreat 33...8¥e7 34 Yyh8 mate. 30...uyg6 31 wye2! 31 Wd7 would also have won, bul White wishes to deny his opponent any chances at all. 31.025 ' This leads to the loss of the queen, but Black's position was already inde- fensible. Thus after 31..¥¥e6 32 vye4 tel White has a pleasant choice betwee: 33 He3 |d8 34 Bd5 Qxd5 35 cxd5, and the sharper 33 Hd7! Q@c8 34 e6. Qxd7 B4..2yxe6 35 He7+) 35 exd7-A gi (or 35...\c7) 36 Abd! 32 h4 ¥yf4 33 23 The queen is trapped: after: 33..ayxeS 34 wade We7 35 |xg7 and a few further unnecessary moves, Black resigned.Invasion Trajectories 67 Far-advanced pawns can also play a part in the endgame, if the pursuit of the king continues. They may take away squares from the enemy king and create a mating net. Tseshkovsky-Polugayevsky Riga Interzonal 1979 Here the struggle has largely lost its sharpness. White has the advantage, but it becomes decisive only after Black underestimates the strength of the £5/g5 pawn pair and... the always potential weakness of his back rank. 25...Be2+? 26 Bf2 Bxf2+ 27 Sxf2 4)xb2 28 xb2 Axc2+ 29 e3 Bxb2 And so, Black is two pawns up, but the very next move is essentially the last one in the game. 30 Bel! M.T. Strictly speaking, the excla- mation mark should have been at- tached much earlier, to White's entire plan. At the least, to his move 25 €d5. LD. What was the mechanism of Black's mistake? M.T. I think, the absence of the queens from the board. With the queens on, even a player of a much more modest standard than grandmas- ter Polugayevsky would have kept his eyes on the white pawn pair, the strength of which is evident. But after the exchange of queens, even a top- class grandmaster can exclude this factor from his attention. 30...b3+ 31 e2 Hb2+ 32 Gel £6 33 g6 On 32...g6 there would of course have followed 33 £6. 33...Ha2 34 8e8 Hal+ 35 Od2 Ha2+ 36 @d3 8a3+ 37 Sc2 Ba2+ 38 Bcd Now the rook can even be allowed onto the e-file: 38...8e2 39 2f4 (with gain of tempo) and then 4)e6. 38...hxg6 39 De7+ @h7 It would be good to play 39..f7 40 x6 $te7, and if 41 He7 Be2, but alas —- 41 \h8 mate! 40 fxg6+ @h6 41 Bxf8 Be2 Black would have lost even more quickly after 41...xa4 42 hd @h5 43 Bh8+ Se4 44 hS and then h5-h6. However, in the game itself White required only 10 more moves to realise his advantage. In the next diagram there appears to be no sign of an impending tragedy. But when Black went in for this position, he overlooked the fact that the f5 pawn was too close to his king. True, for the moment it is alone... 28 Sxc6! Axc6 29 Axb6 wyxb6 30 We4+ @h8 31 wf? The pawn has received "reinforce- ment" in the form of the queen, and in addition 32 f8 is threatened. Blocking the bishop's diagonal (by 31...2)b4) does not succeed: 32 £6 Bg868 Attack with Mikhail Tal 33 fxg7+ Hxg? 34 Wf8+ and 35 wyxb4, Yusupoy-A.Sokolov 55th USSR Ch., Moscow 1988 31...Hg8 32 £6 wWd8 33 Ge7 Sxe7 34 fxe7 ‘The white pawn has moved away from the king, but only in order to become a queen. This cannot be pre- vented. 34..24d7 35 Hd3 h6 36 Hf3! cS 37 ef8+ Wh7 38 Af? Hel+ 39 Ser We6+ 40 Bh3 Web+ 41 Wh4 Black resigns. Tal-Mohrlok Varna Olympiad 1962 Sicilian Defence 1 e4 5 2 Af Deb 3 ad exd4 4 &xd4 ALG 5 &c3 a6 6 Aes 6 7 @d2 fe? $ 0-0-0 0-0 9 2b3 whe 10 8 a6 gd Has 12 Qe3 We? 13 g5 a7 14 bd bs 15 This pawn sacrifice is a fairly rou- tine one, since it is obvious that White needs open lines for an attack on the black king, which for the moment has been deserted by its "subjects", in particular the knight from f6 and the rook from f8. M.T, Incidentally, I made this sac- Tifice on two earlier occasions, but the game with Mohrlok is taken as the main one, for the reason that after it no one else granted me the same pleas- Z “i, Li, lll, ens 1S... fxg6 This was how the problem of the "impudent" pawn was solved by Boles- lavsky in a game with Spassky (25th USSR Championship, Riga 1958). Ina training game Tal-Koblenz {also in Riga, but a year earlier) Black elimi ated the presumptuous infantryman in ainvasion Trajectories 69 different way: 15...hxg6. M.T. I cannot deny myself the pleasure of giving the attack and com- bination that followed: 16 hS gxh5 17 BxhS 2f6 18 Bhi d5 19 e5! Axe (19...84xe5? 20 a f4 wF5 21 Qd3 de4 22 fxe4 and 23 Yyh2) 20 Af4 Qd6 21 Wh2 SEB 22 wh8+ Ags 23 Wh7 f5 24 4ih6 Hd7 25 AxbS! Hf7 (if 25...axbS 26 &)xb5, 27 4)xd6 and 28 Axg7+) 26 Bgl Ba7 27 |d4 Ded 28 fxg4 Med 29 Sc6! Axc3 30 Ge3! (with the threat of 31 (ic5+) 30...d4 31 Highl! (threatening 32 wyxg8+) 31..Hd7 32 gs axbS 33 Hth6!, and against the concluding stroke 34 Hf6+ there is no defence. Finally, in a telegraph game Tal- Stoltz, 1959, in the diagram position Black declined the gift, and attempted in search of counterplay to strengthen his piece pressure on the queenside: 15...4)¢5. But he too came under an attack — 16 gxf7+ &xf7 17 Qh3! (aiming at the weak e6 pawn) 17... 4)a4 18 £4 b4 19 £5 eS (it tums out that 19...4)xc3 20 bxe3 4)xa2+ 21 @b2 is not possible) 20 4\xa4! (White is not tempted by the win of the ex- change — 20 Ad5 \xd5 21 wyxd5+ BS 22 kyxa8 @b7 23 wa7 Ha8 24 )d4 exd4 25 wyxd4, since then Black gains counterplay: 25... f6 26 wyxd6+ Wxd6 27 Hxd6 Axb2+ 28 bl Axe4) 20..4)xa2+ 21 Bbl bxa4 22 4)a5 Hb8 (the sharpening of the play seems to be not unfavourable for Black, since if 23 @xa2 he has 23... AbS, but...) 23 eydS+! Hf 24 Sxa2 Yyxc2 (or 24...HbS 25 Yyc6 vyxaS 26 £6) 25 Hd2! Bxb2+ 26 Wal wc3 27 ¥wd3!, and Black resigned. 16 nS gxhS 17 ExhS 6 18 gs eS 19 wye2 ALS 20 Mer ca 21° Axcd — bxed 22 d4 Hbs 23° Bhi Bb7 24 Bh6! A difficult move. On 24...6 White had prepared 25 Hhxg6+ hxg6 26 Hxg6+ £7 27 Wes Dh7 28 whs Af6 29 Hixf6+ @xf6 30 4)f5!!, and against the threat of §.g5+ there is no defence. wa... on 25 Bh4 whe 26 dl We7 27° «4 h6é 28 Be6 Hes 29° £5 05 30 2\c3! Threatening 31 Hxf6+. 30... Wd8 31 Aycé Black resigns Torbergsson-Tal Reykjavik 1964 King's Indian Defence 1 d4 a6 2 ¢4 26 3 Ac3 ez 4 e 0-0 5 f4 d6 6 AR es 7 4 e6 8 fe2 exd5 9 exd5 bs!? 10 AxbS Tf 10 cxbS, Black would have con- tinued 10...a6, in analogy with the Benko Gambit.70 Attack with Mikhail Tal 1... Het H 60 a6 12 a3 This then is the idea of Black's sac- rifice! The knight at a3 occupies a most unfavourable position, and Black has time to concentrate his forces in the centre. Perhaps White should have returned the pawn with 12 4\c3. 12... a7l Black finds a “clear road“ to trans- fer his rook te the centre. 13° Qd3 He? 14 ed Gfes 15 Bel It is now clear that White has diffi- culties in developing his queenside. i... Qd7 16 Hed nae 17 wWye2 DHS White has not found the best plan, and is alveady obliged to parry con- crete threats, 18 g3 This move is a success... for Black. But 18 4fl would have been very strongly answered by 18...4)g5! 18 . fide 19 Bxd4 exd4 20 Dg? gs! 21 Bxe7 h3+ 22 @fl If 22 @hi Black has 22...4xe7 23 Sid2 f6 24 Bel De. 22... Gxe7 ‘The natural 22...0/xe7 after 23 Qd2 would have given White a tempo for the defence - 24 Hel. 23° Adz 24 ghd 24 Hel is decisively met by 24... &)g4 25 Hxe7 wyxe7! 26 Ah4 He3+. D6 24... aed 25 |f3 On 25 @g2 Black had prepared the following variation: 25..Wye8 26 £5 He2+ 27 Qxe2 eyxe2+ 28 Hxh3 hS 29 Bhi ape 30 Sg? Axhl +. : as He3 56 @g2 We? 27 «Hel M.T. Here Black could have carried out two interesting combinations. I spent a long time considering 27. Hxel 28 Sxel Af5 29 HF3 tye3 30 Sixe3 Hxe3+ 31 Shi Mxd3! 32 Yyd2 = Set 33 tye2 g5 34 94 h5, and wins. Unfortunately, the combination is not - forced. After 29 ¥ydi I could not see | any way te win, and so } chose the second pessibility. feo :e fo Rk Ba UA) img VEE 27 aes &xf4a4t 28 gxf4 Axel 29 Sxel = he 30° fed Obviously the only move. 30... wxel 31 3 This loses very quickly, but ever after the superior 31 £5 4ye5 Black ha:invasion Trajectories 71 an irresistible attack. SI ow. yh6! 32 £5 ayxf5 33° Af4 Now comes the concluding combi- nation. 33 oes éyha+ 34 Bhd B+ 35 @g2 9 @xh34! 36 @xf3 36 @xh3 loses to 36...4yh4+ and 37...Qelt+. 36... well With the threat of 37...4g4+ 38 Sed f5 mate. If 37 Qh6, then 37... Asd+ 38 Sf4 Qhs is decisive. 37 Gxe6 wed+ 38 «Of yyxfa+ 39 @gl hxgé White resigns Tat-Donner Beverwijk 1968 French Defence 1 e4 e6 2 d4 a5 3 Bcd Abd 4 05 5 5 a3 Axc3+ 6 bxe3 wye7 7 ats b6 8 a4 fae 9 Qxa6 2xab 10 We2 bs In solving the problem of his "French" bishop, Black is forced to pay for it in terms of tempi. Thus the knight is obliged to return home, since on 10...cxd4 there would have fol- lowed 11 Wyb5+ @d8 (or 11...@f8 12 Sa3+ Se7 13 eyxa6 wyxc3+ 14 wer Yyxc2+ 15 4d2 and 16 Bhel) 12 Wyxa6 eyxc3+ 13 Be2, and 13...¥yxal is not possible in view of the bishop check at g5. But since the black pieces have retreated to their initial positions, White needs to open as many files and diagonals as possible for the attack. 11 ast bxa5 If U1...cxd4 12 0-0!, and White achieves his aim. Nevertheless 11... 4)d7 was more circumspect, but Black had not anticipated his opponent's 14th move, 12) Qa3 4d7 Perhaps Black should have given up a pawn to gain a tempo — 12...2\e7 13 Axed Ad7. 13 dxcS 14 c6! A consistent continuation of the same tactics. 14... Wxe6 15 0-0 Wxe3 16 Hfdi D6 First 16...4¥c4 was rather more ac- curate. But Black wants with gain of tempo (the e5 pawn!) to reposition his knights, such that they block the open files on the queenside. Sye7 17° Ad6 Wed 18 wye3 wed 19 wyb3 Ebvo 20 edt M.T. The white rooks need the c-file! Hence this new intuitive sacri- fice of what is now a third pawn. Moreover, in my calculations [ estab- lished that neither 20...2)xc4 21 ¥¥b7, nor 20,..dxe4 21 Wb5 Bc8 22 Hxas is possible. The counterattacking attempt 20...a4 is parried by 21 8xa4! (21 WbS? wyxe4) 22..dxc4 22 wyxb6!72 Attack with Mikhail Tal axb6 23 xa8+ 4d8 24 Hc8, when there is no defence against 25 {4.c7 or 25 Ke7. 20 «6 wxe4 21 wad Wab 22 Hach = eS 23° d2! Aiming for ¢5. 23 ae £6 After the game 23...¢)d4 was sug- gested, and if 24 BxcB+ ¥yxcB 25 wyxaS ¥d7, but White san play more elegantly and strongly: 24 @hi! |f5 25 ExcB+ vyxc8 26 Bel wyd8 27 gc5 f6 28 Yyxa5 fxeS 29 Yyxa7 Hd7 30 {a3 47 31 He7, when the threats of 32 4)f3 and 32 4)b3 are irresistible. 24 exf6 25 WB @d7 25...f7 is insufficient on account of 26 wyhS+ @g7 27 He3, while if 25,..€)07 the following variation, while somewhat "cooperative", is not without its logic: 26 4¥g4 (with the threat of 27 ¥¥e7) 26..@0B 27 eyxe6 He8 28 Wrxd5 We2 29 Act web 30 Axa5 Waxd5 31 Hb7 mate. 26 YYxf6 = Ahe8 Or 26...xd6 27 He4+ Bc7 28: Bes fye2 29 wye7+ Hdb 30 Hb7. mate. 27 Sed Der 27...dxe4 28 eyg7+ De7 29 Qxe7+ G5 30 A f6+ leads to mate. 28 AcS+ Axed 29 QxeS Aes 30 Axe? Black resigns: 30..8xe7 34 xdS+. WHAT WOULD YOU HAVE PLAYED? No.12 In this position from our classical heritage, reckoning that after 18 h6 ~ Black would block the position by 18...g6, and after 18 g6 by 18...fxg6 19 hxg6 h6, White began the conclud ing attack with 18 xh7+ @xh7 19 - g6+ G28 20 Dxed dxe4 21 h6! £6 (or 21...exf3 22 gxf7+ Yxf7 23 hxg7) 22 © hhxg7 exf3 23 Hh8+ @xg7 24 Bh7+, and soon won, We will say straight = away — a possible solution. But is it obligatory?Invasion Trajectories 73 Now13 sate ie ah ‘ates 8 LS White carried out a similar plan in this position: 13 hS! 2c4 14 Qxed Hxed 15 g5 £)d7 16 del We7 17 g6 He8. How can Black's defences be further broken up? (After all, he him- self has prepared counterplay with 18...2xc3.) The impression is that on the kingside Black has been playing not chess, but draughts (checkers). At any event, the pawn formation here re- sembles the jaw of an old man, who not once in his life has been to the dentist. Naturally, the light squares have already been occupied by the white pieces, But what next? No.i5 For the moment Black has managed to forestall the opponent's activity on the queenside and to gain the initiative on the kingside, as, however, often happens in the fianchetto variation of the King's Indian Defence. But now White is threatening by 21 h4 to block all the lines here, and then to begin play on the opposite side of the board, where he is stronger. And after 20...h5 21 gxhS gxh5 22 hi there is no possibility of Black creating an attack on the g-file...5 Lines of Communication Ir IS ON HOW THE PIECES behave on the communication lines — files and diagonals — that the success or failure of an attack depends. There is nothing surprising about this: for 200 years no one has cast doubts on Napoleon's aphorism: "War is com- munications”. How to clear and use them was the subject of the previous chapter. But for victory this is some- times not enough. Not only to give your own pieces scope, but to close lines to the opponent's men: this is the task of any player who sits down at the chess board. Intuitively everyone senses this, but strangely enough, in practice such attacking procedures and plans are not regularly encountered. Evidently our subconsciousness needs to be reinforced by consciousness, based on knowledge --- and then your opponent's pieces will often be too late in reaching the decisive part of the battlefield. Too late, because of you! The Barrier In the theory of chess this is a new concept. The authors define it in the first instance as a non-material barrier, through which either one piece, or a whole group of pieces, is unable to pass. In short, it is a communication line under close-range fire, and the creation of one is usually a basis for future success. It was this motif that prompted White's decision in the following game, the final one of the match. Tal-Larsen Candidates Semi-Final, Bled 1965 16 25! exd5 17 exd5 ‘ The piece sacrifice is a positional one, since it has been used to erect an invisible barrier on the e-file. A num- ber of squares on it (e5, e6) are con- © trolled by white pawns, and a white © rook will soon be moved to el. By con- trast, Black's pieces (queen, rook, bishop and to some extent his knight) are bunched together on the queenside and cannot easily come to the aid of their king. Now White is threatening the routine combination with bishop sacrifices at h? and g7, against which Black cannot defend without making positional concessions. If 17...g6 White can continue his attack either by 18 b4, or the more active 18 ¥yh3! So Larsen aims to cover his h7 with the other pawn.Lines of Communication 75 17.5 But now White's dark-square bishop is too strong. 18 Rdel Here Black has an unpleasant choice: he can either defend the bishop with his rook from f7, but this will then allow the white pawns to open up the kingside with gain of tempo (g5-g6!), or else move yet another piece away from the kingside. On 18...4.d8 the following very curious variation was possible: 19 wh5 4)c5 20 xg7! 2)xd3+ 21 Bb! (not 21 cxd3? we7+) 21...2)xel (21...2)xf4 22 wh6) 22 26 Bxg7 23 Wxh7+ Sf6 24 7 BET 25 g8=4) mate! 18...8£7 19 hd &.b7 20 4 xf5 M.T. But this can be attributed to competitive considerations. Had this position not been reached in the last game of the match, 1 would undoubt- edly have played more sharply: 20 26 hxg6 21 hS g5 22 Q.xf5 (weaker is 22 h6 g4 23 hxg7 Af6, or 23.06 24 Hxe7 gxf3 25 @xf6 Bxe7) with very dangerous threats. E.g., 22...8xf5 23 Hxe7 eS fails to 24 h6! Hxf3 25 h7+ @f8 26 Hxe7 with inevitable mate. But at this point [ wanted to make absolutely sure, and at the board I was unable to find a forced win after 22....f6 23 (.e6 WF! (not 23...a.xd4 24 fxg5 and g5-g6). That there was a win, I was sure (just as I am sure now), but the experience of the preceding game wamed me against spending time on the calculation of long, complicated variations — I might end up in time trouble, particularly since after the move made my position remains very favourable. 20...8xf5 If 20...2)f8 White can simply inten- sify the pressure, by defending the bishop with his queen and renewing the still-present threat of 4 xh7+. 21 Bxe7 eS After the passive 21...8f7 Black is crushed by 22 Bxf7? &xf7 23 g6+ hxg6 24 h5, when bis king is com- pletely exposed, The Danish grand- master tries to seize the initiative by tactical means, but White is prepared for this, 22 Wed WIS! 23 fxeS! Hid 24 wed BB After this move White wins without any great difficulty. The main variation of the combination begun with 20 SuxfS was 24....xd5 25 exd6 Bxd4 (after 25...Q.xhl 26 &xg7+ Black's scattered pieces are helpless) 26 Yyxd4! (weaker is 26 Bel ¥yf4!) 26...4.xh1 27 b3, Here Black probably does best to retum the piece immedi- ately by 27...0.f3 28 Wed+ @h8 29 BE7 tyxd6 30 Bxf3, with some chances of saving the game. The attempt to maintain his material advantage is hopeless —- the white h-pawn, on reaching the 6th rank, lands the deci- sive blow. The exchange of rooks after 27...Be8 also loses to 28 Wye5 Bxe7 29 dxe7 we8 30 Web+ Ph8 31 h5 Af3 32 h6, or 31...c6 32 g6 with the ir- resistible threat of 33 wf7. 25 wWe2 wxe7 No better is 25...Wf4+ 26 wd2 Hfl+ 27 Bxfl eyxfl+ 28 wal, or 25... Sixd5 26 exd6. 26 Wxf3 dxe5 27 Bel Rd8 The ending after 27...8f8 28 Bxe5 ¥yxe5 29 Wyxf8+ Hxf8 30 Axes gives76 Attack with Mikhail Tal White an easy win — Black has no time to take on d5 in view of 31 .d6+. 28 AxeS Wyd6 29 wyfa! With this simple bit of tactics (29... Sxd5 30 He8+) White keeps his two extra pawns. The finish is straight- forward, 29...18 30 wed b3 There is nothing better. 31 axb3 Afl+ 32 d2 Wh4+ 33 3 Wd6 34 QcS! Not altogether necessary (there were many ways to win), but an amusing concluding combination. 34..¥9xe5 35 He8+ Bf8 36 wed+ @h8 37 wi? Black resigns. A similar idea was carried out by White in the following game, with the "slight" difference that a real sacrifice proved unnecessary, and the barrier, cleaving Black's position, was erected not even for free, but with the win of a pawn... TR Ff, ae ‘4 gy i BE Shishkin-Karasev Leningrad 1989 13 QaS! Yyd8 14 FS! exfS 15 ex- £5 @hs ‘The threat of He4-h4 does not even leave Black time for 15...He8. 16 &xe7 Qxp2 17 who! ‘The quickest way to win. IT..A)0T 18 E6+ Bes Alas, after (8...fxg6 19 fxg6 not one of Black's pieces is able to come to the aid of his king - partly on account of the barrier erected by White. 19 4)xf8 WyxfB 20 YyxfB+ OxiB 2: Sxg2, and Black resigned. Geller-Velimirovic Havana 1971 In this position the role of the bar- © rier is partly played by the pawn chain a2/b3/c4/d5. It is practically keeping one black knight out of the game, but that is only part of the matter. White's plans include active play on the king- = side, from which will be excluded the | remaining black pieces, in particular = the queen's rook. This is possible only if he can establish control over the. “transit” square e7, from which the © black queen could carry out the great- © est number of functions — of defence and counterattack. 4Pee creer cee Sas Lines of Communication 77 Hence the fantastic and unusually audacious pian — to sacrifice a rook! Instead of 14 1b2 e3 15 fxe3 2yg4, for which Black was hoping, there came a clap of thunder. 14 |dxed!! |xed 15 |xed A xal 16 Ags Ate Black cannot avoid the coming pin, since, if he moves his queen, the bishop at al is lost, and with material almost equal (a pawn for the ex- change) White gains a powerful attack. 17 &)xf6+ Hxf6 18 wal &f7 19 Hel Despite being a rook down, in the main thrust of the attack White has a superiority in force. It is difficult for Black to bring his queenside pieces into play, since simple developing moves do not work, For example, on 19.,.4.£5 or 19...Q.d7 there follows 20 ¥yc3! (with the threat of 21 Axf6 Wxf6 22 Wyxa5) 20...b6 (the only de- fence) 21 He6! and White wins, since the forced 21...{1xe6 leads to the opening of the long light-square diago- nal — 22 dxe6+, and loss of materiat for Black. 19...Hb8 The alternative defence was 19...h6, but it would not have saved the game after the simple reply 20 {xh6. For example: 20...4.f5 21 9.95 He8 22 24 Axg4 23 Be4 AS 24 Bhd, or 20... Wh8 21 Qg5 Ad7 (if 21..9f5 22 ¥yc3 66 23 He6, whereas now in this variation Black has 23...8f5) 22 Bed! BPS (otherwise 23 4. xf6 wyxf6 24 Bf4, while on 22...8f8 there would have followed 23 Bh4 tyg7 24 {h6 Wyg8 25 Axf8 Wxf8 26 Bh7+ Yes 27 wWel+ Gd8 28 wWxa5+) 23 He7+ Hg& 24 Wyxh8+ Sxh8 25 .d2, with a decisive advantage. 20 He3 b6 Defending the knight, which, as we have seen, was “hanging” in several variations. No better was 20...f5 21 h3! 5 22 p4 hxg4 23 hxg4 Axed 24 He4 Qf5 (24...Qh5 25 He6) 25 Bh4. 21 BES SES 22 4 Wh8 23 Axf6 The simplest solution: White trans- poses into a won ending. 23 Qh3 is not altogether clear in view of 23...828. 23..89xf6 24 wyxf6+ Oxf6 25 gxf5 gxf5 26 He3 An important link in the plan for realising White's advantage. The fron- tal attack on the f5 pawn, 26 (h3, would have allowed Black to maintain equality: 26...2g8+ 27 @f1 Hg5. But now the white rook penetrates into the heart of Black's position. 26...4\D7 27 He6+ Gt7 28 Af3 The threat of a bishop check at hS finally disrupts the coordination of the black pieces, 28... 28+ 29 Sil PES Black cannot hold the h5 square: 29...395 30 h4. 30 QhS HgS 31 Be8+ wg7 32 He7+ @h6 33 Axb7 HxhS 34 Hxa7 §xh2 35 8d7, and White won easily in the rook ending. Interference Closely approaching the “barrier” theme, first formulated here, as an effective attacking procedure, is so- called "interference" — a purely tacti- cal blow, which also has the aim of cutting off the opponent's forces from the defence of some key or simply78 Attack with Mikhail Tal important point. But whereas the bar- rier is lowered for a long time and acts rather as part of a strategic plan, inter- ference is instantaneous and operates immediately. Klovans-Tolush Leningrad 1962 Only the black queen is guarding the mating point g7, and the interference move 37 8e7! is immediately decisive: if 37...Q.xc7 38 ¥yd44, forcing mate. Simagin-Bronstein Moscow 1947 ssaanddiiisiattiy Must White force perpetual check? ‘The answer would have been yes, had he not noticed the thunderous $1 Ags!, when Black must either go into - a lost queen ending after 52..uyxg5 (if 5h...fxg5 52 f6, forcing mate) 52 Wyd8+ Hp7 53 Wo7+ and 54 wyxh2, or else accept the fact that both(!) his queens are cut off from the king's res- idence, He preferred the latter: 51. hi=, and after 52 we8+ bg7 53 We6+ SES 54 wWxf6+ Sg8 55 wd8+ Sye7 56 Wel+ Yg8 $7 We8+ Black resigned in view of mate in three, A.Petrosian-Panchenko Odessa 1973 After incorrectly accepting a pawn. sacrifice in the opening, White is mar- kedly behind in development. True, he has no weaknesses, and at first sight i is not clear how Black can approach: the white king still in the centre. The solution to the problem lies in a far fromm obvious interference move, 14..Qb411 The threat of 15...Qxc3 leave: White no time for [5 &g3, and afterLines of Communication 79 his forced reply his queen foses control of the 4th rank, from which the assault on his king intensifies. 15 exb4 Q.c4 16 Hg3 After 16 Se3 Qxe2 17 Wxe2 4d4+ White loses the right to castle, and he is left with simply nothing to move (18 @el She8, with the threats of 19..8yh5 20 Wdl wWxdi+ or 19... ‘yed). But now he hopes, by returning the material, to take his king to safety. 16...Q.b5! 17 tya3 Otherwise the black knight comes into play with gain of tempo. 17...Hd3! 18 £4 we6 19 b3 And once again the queen's path to the centre of the battle is blocked. 19...53xg3 20 0-0 Hxg2+ 21 dxg2 xe? 22 £3 In parrying the deadly check at g4, White is prepared to part with his gains, but for Black this is not enough. 22..818 23 93 we3 24 tel (24 #2 404) 24... xf3+ 25 Bxf3 tyxf3+ 26 &gl He8, and with a continuing attack Black now has a material ad- vantage. He soon won. It was this attacking procedure that was at the basis of a deeply-calculated variation in the following game. M.T. To be honest, I recalled it with pleasure during the illustrative game that we selected for this chapter. Botvinnik himself admitted that he went in for this position, assuming that after 22 axb3 @b7 Black would not have any serious difficulties. 22 Bal! In this somewhat paradoxical way White ensures the invasion of his rook onto the 8th rank. But it would not have brought any dividends, had he not had "in reserve" his murderous 25th move... Ragozin-Botvinnik Moscow 1945 22... a8 Bad is 22...8b7 23 Hd8+ @h7 24 Ad3+ f5 25 Bxc8 Dd4 26 a6, when Black can resign; 22.,.4)c1 loses to 23 Hd8+ @h7 24 .d3+ 4xd3 25 Qxb8. 23 Hd8+ Of8 This leads to a mortal pin on the 8th rank, but there is no choice, If 23... @h7 24 9d3+ g6 25 cxb3! (after 25 Ae5 Qb7 26 Bd7 4)c5 27 Bxf7+ gs 28 Qc4 a6! or 28 Axe6 Bd8! 29 h4 Hdl+ 30 @h2 d7 Black holds on) 25....b7 26 Qb8 aS 27 a4, with the threat of 28 Ac4. If Black tries to free himself by tactics: 27...Q.c3 (threat- ening 28.95, and if 28 He8 ada, threatening 29...4.a7), then 28 {c7, and after 28...xd8 29 Qxd& @g7 30 £3 White would appear to have a deci- sive advantage in the endgame, in view of Black's weak queenside pawns, 24 axb3(?) Here this “automatic” capture80 Attack with Mikhail Tat "towards the centre" is a serious inaccuracy, spoiling a splendid idea. 24.0 Ab7 So, is the activity of the white rook curtailed? After all, it cannot be kept on the 7th rank after 25 Bd7 Ac8. 25 Abst The crux of White's plan. On the main part of the battlefield he retains an advantage in force. 25...a5 26 [3 Alas, after the correct 24 cxb3 White could now have stalemated the black pieces by 26 a4!, when equally bad for Biack are 26...a6 27 Qxa6 Hxa6 28 Qd6, and 26..cS 27 Ac7, winning the aS pawn, and 26...96 27 Me4 Bg? 28 He5+ and 29 Hd7. True, he still retains an advantage, 26..a4 27 bxad Hxad 28 Wf2 (28 Qd6 Ba8) 28..c5 29 HeI Acé 30 Ad3 Ha8, and now, by continuing 31 - Se4!, White could well have hoped to win, e.g. 31..a3+ 32 @f4 g5+ 33 eS Axed 34 Bxed Has 35 dS, threatening the thematic 36 .d6. Kasparov-Pribyl Skara 1980 After the quiet 20 4 f4 White would have retained sufficient compensation for the pawn. But with a piece sacrifice he excludes the enemy queen from the defence of the weakened kingside, thereby gaining the opportunity to mount an attack with superior forces. 20 d7i fxgS 21 Wed+ Bh8 22 }xg5 Af6 A weighty argument in favour of White's entire plan would have been 22....d4 23 Hxd4! cxd4 24 Uyxd4+ Sg8 25 Heb. 23 S)e6 4\c7 There is nothing else: if 23...4)b4 24 Wid Deb 25 HxfB BxfS 26 d8=¥y Axd8 27 Exd8. 24 4xf8 Oxf8 25 Hd6 e7 26 as=yy! Seemingly paradoxical, but consis- tent and strong. Now d7 is needed for the attack along the 7th rank, and the black queen is still out of play. 26...4x08 Or 26...xd8 27 Hxd84+ Qxd8 28 f7 Wd5 (defending against the mate at £8) 29 ¥xd5 xd5 30 Bd), and White wins, 27 Wyc3+ Se8 28 Ad7 Af6 29 weds Bhs 30 wed The piece is regained, and the end- ing after 30....97 31 ¢yxc7 Yyxc7 32 Hixc? Qd4 33 fl must be won for White. But... 30...4¥a6? 31 Wh6 Black resigns. Obstruction There is also another typical “mine", or more precisely, a typical barricade, that can be erected in the path of the enemy forces: the well known obstruc-Lines of Communication 81 tive sacrifice. If a game takes a nor- mal, even course, to gain a lead in development, giving the basis for an attack, is difficult, if at all possible. One has to resort to radical measures, which nevertheless follow a common pattern. This is normally a pawn sacri- fice, that in one way or another dis- rupts the quiet course of the game. The effectiveness of such a proce- dure — given an inadequate counter to it — is seen in the following example, Spiefmann-Landau Amsterdam 1933 1 e4 S)f6 2 43 d5?! (not the best reaction to White's move) 3 e5 4)fd7 {after 3...2)g8 Black is very far behind in development) 4 e6! fxe6 5 d4 2)f6? 6 4f3 c5 7 dxc5 4\c6 8 AbS Ad7 9 0-0 Yye7 10 Bel h6 11 Axc6 bxc6 12 4)e5, and White won quite easily. An excelfent illustration is provided ~ by the following game. Kasparov-Andersson Tilburg 1981 At this point Black was happy with his position. He plans ...d5, after which his backward c-pawn will be weak only nominally, rather than in actual fact. In addition, the wretched white bishop at b2 will provide compensa- tion, The evaluation is changed by an obstructive sacrifice planned in ad- vance by White. 14. d5! exd5 15 Q22 Now it is very difficult for Black to develop his kingside, and strategically the outcome of the game is already decided. A rare case of such an instant return from this type of sacrifice. 15...06 16 0-0 £6 Biack begins artificial castling, but this entails weakening a whole com- plex of light squares on the king- side,which spells disaster. White mere- ly has to open lines for his rooks with e3-e4, 17 Ret! A useful preparation. The immediate 17 e4 dxe4 18 Qxe4 G17! would have enabled Black to "plug" the light- square holes for the time being, al- though even so after 19 Wee it is hard to offer any good advice. 17..e7 18 tgs Gf7 19 hS Ah? 20 e4 dxe4 21 Qxe4 Qxed 22 E)xe4 DeB The aim of this seemingly strange manoeuvre is to defend the f5 square after ...d5 and ...6)d6. The need for this is apparent in the variation 22...e8 23 We6+ @f8 24 gd, when against the threat of 4)g3-f5 there is no defence. And the natural continuation of the artificial castling manoeuvre, 22...2f8, runs into 23 Rad! (provoking a weak- ening of e6) 23...d5 24 @)xf6!, mating.82 Attack with Mikhail Tal 23 Had Ha7 24 xf6! Another typical sacrifice with this type of pawn structure: the king is deprived of its shelter. 24...gxt6 It all ends even more quickly after 24...Qxf6 25 We6t+ BFS 26 Axf6 exf6 27 Heb. 25 We6+ Si 26 Hel d5 27 Bdat This is more accurate than the im- mediate 27 {.xh6+ Hxh6 28 ¥yxh6+ @g8 29 Hdd Qi 30 Hed+ Be7 33 Expg7+ &xg7 32 wyg6 De7, when for the moment Black can hold on, 27.446 28 Bed DET 29 Q.xh6+ He8 30 Ag? Black resigns. For the h-pawn he has to give up his rook, and, what's more, immediately. Here we see a similar idea: Keres-Schmid Bamberg 1968 Seemingly, Black's bishop pair and sound pawn centre should allow him to face the future with confidence, but White finds an interesting possibility of increasing his lead in development. 12 e6! An excellent sacrifice, obstructing | the development of Black's kingside. 12...89xe6 Totally cheerless is 12...fxe6 13°: )e5, and wherever the queen moves to (13...89d6 14 8.4), White, relying o his powerful piece deployment, creates very strong pressure, 13 Ags Wes After 13..4yd7 14 Qxed dxed 1 ¥yxe4 Black's queenside is broken up, but even so this fooks the lesser evil. 14 &xe4 dxed 15 £4 £5 16 wed! Rather than the tempting 16 ¥yb3, with the threat of 17 ¥yb7, White pre- fers the consistent obstruction of th enemy forces: again 16..e6 is not possible, meanwhile, his plans includ doubling rooks on the open ¢-file, adjacent to the black king. 16.05 Opening a retreat for the queen Black has no objection to 17 ¥xc. Web followed by ...@f7 and ...g6. 17 Bad web 18 {3 An extra file for the attack will do no harm, although 18 Hd2 is also suf- ficient, threatening after the doublin of rooks to invade Black's position. 18...h6 Alas, here too the “normal” 18.. tuns into 19 Hd6! Qxd6 20 Yyxeb+ @d8 21 Hdl, and Black can resign. 19 fxe4 19 Bfel looks even stronger, and i 19...exf3 20 He5 with decisive threats. 19...4yxed 20 YyxcS e6 21 We’ SLe7 22 Bd7! Qd8 23 Rxd8+ Axdd™ 2A Wyxg7 Black's position has collapsed, and: since 24..Wyd4+ 25 tyxd4 Hxd4 fail: to 26 (2e5, it is essentially all over.Lines of Communication 83, 24...Ff8 25 Axh6 wydd+ There is nothing else: 25...f7 26 Wg8+ We7 27 es+. 26 vyxdd Axdd 27 3! A precise tittle move: for the mo- ment the d2 square is under control, 27...d6 28 Axf8 xf 29 Hel, and White realised his great material advantage, Kasparov-Lutikov Minsk 1978 White has a lead in development, but under the cover of his central pawn chain Black hopes gradually to neutral- ise it. Therefore White's obstructive © sacrifice is more than justified, 9 e6! fxe6 10 Qd3 Agee 11 we2 26 12 wyxe6 227 13 0-0 “The opponent's defences will not be ~ cracked by crude pressure down the e- file; operations on one of the flanks are also required, I already had in mind the general idea of a pawn offensive on the queenside, and so I decided to hide my king away on the opposite wing" (Kasparov). But queenside castling foliowed by the advance of the g- and f-pawns was also possible. 13...49h5 14 .g5 468 Preparing to take the king to its “usual place. After 14...Qxd4 15 Bfel Qf6 16 Qxf6 Ahxf6 17 Hat intending c2-c4 White has good play. 15 Wed f6 16 We2 Wd6 17 Rael Planning to open up the game by f2- £4-£5, 17...26 18 ad G7 19 bd 6 White has too many threats after 19...¥9xb4 20 4yc5. 20 Wyd2 He8 21 AFA wWe7 22 bs Wad 23 Ae3 cS 24 Ab! wad 25 dxeS bxc5 26 c4!, and since after 26... d4 the black queen remains out of play, the opening of the centre is inevitable and White's unhurried attack con- tinues. It was begun, as you will recall, by the obstructive pawn sacrifice. In order to seal the opponent's forces in their initial positions, more generous offers can be made, as in the following game. True, here the obstructive sacri- fice also plays the role of a powerful pawn outpost. Beylin-Lipnitsky Yurmata 195084 Attack with Mikhail Tal 15...e4} And immediately it becomes clear that White's extra pawn, his passed pawn in the centre, and his attack on the black bishop do not compensate for his lack of development. For the mo- ment, the sacrifice is not a real one: the threat of 16...Hac8 does not leave White time for anything, apart from 16 Bel Hac8! Nevertheless! 17 Wyxd7 3! The apotheosis of Black's idea. First, the opponent is now playing without his kingside pieces, and sec- ond, the white king is almost encircled, 18 wad Further “pawn-grabbing" — 18 ¥yxb7 — is punished by 18...6xc3 19 Wyb2 Hfc8 20 Axc3 Hxc3 21 g3 Hb3!! 18...8xe3 19 Hdl Afc8 20 g3 Hel 21 Qh3 Hxdl+ 22 Yxdl eyc3+ 23 Sil yyd2! 24 Ser 1D. I asked Mikhail Beylin, who was playing White, whether he felt reassured at this point, and received a reply in the affirmative... M.T. Really? Black's concluding blow is not so complicated, although this does not make it any less pretty. Besides, Black clearly must have an- ticipated it six or seven moves earlier. 24...Hc1! White resigns, since after 25 wyxcl Wrxe2+ 26 Bgl he is mated by 26...¥yf2. Even so, the obstructive sacrifice is usually of a more local nature, and usually prevents the king from being defended by the advance of one of the pawns sheltering it. The pawn is blockaded on the square where it stands, and for this aim a long-range © piece is usually used. Q 1.D. The first combination of my co- author to appear in print was of this type. Did you keep the cutting from the. magazine? M.T. No, but 1 liked it. 1D. What exactly? A M.T. Both the combination, and its publication. : Tal-Leonov Riga 1948 28 BEG? Now Black does not have the de~ fensive advance 28...f5, and "in pass-: ing" the threat of 29 Bxh6 is created. At the same time, the rook is taboo: 28...gxf6 29 Sxh7+ @xh7 30 Hed WR 31 4)xf6+ Hh8 32 Bl, and. against the manoeuvre Hif4-g4 there is no satisfactory defence. : 28...49f8 29 fd Od7 (here 29...f5 no longer works on account of the immediate 30 g4) 30 Aga! (the pre: sure intensifies) 30..Q¢8 31 26+. &)xf6 32 exf6 He7 33 fxg7 xg7 34, Yye5+ Black resigns.Lines of Communication 85 A similar and even more concrete situation developed in the next game. the threat of 4 {.xd8 and 5 4sf6+ is more than unpleasant. Fischer-Benko USA Championship 1963/4 Black's h7 is weak, but the direct attack on it, 19 e5, is parried by 19... £5. Hence the natural — and only good — continuation of the attack. 19 H£6! des After 19...Q.xf6 20 e5 there is no defence against the mate. 20 e5 h6 21 4ye2 A quiet reply, emphasising the hopelessness of Black's position. Against mate in a few moves there is no defence, and so Black resigned. White had to conduct a much more complicated attack in the following game, but at its basis was the same device of obstruction. Black is threatening most of the at- tacking pieces, but at the decisive moment he is unable to make a saving pawn move. 1 &)xg7! @xg7 The gift cannot be declined: L...A\cxd3 2 cxd3 4)xd3 3 4yh5, and Ravinsky-llivitsky Riga 1952 2 Exfd! exf4 3 who+ g8 4 9 fol! The black bishop is lured to f6, where it blocks the f7 pawn. 4...Qxf6 5 e5 £)xd3 6 exf6 Black has avoided the mate at h7, but now there is a threat of mate at 27! 6.224 7 OE1 Hh3+ 8 Sf! Capturing the knight would have allowed Black perpetual check. 8....c44 9 Se2 Axe2+ 10 Hel! The checks are at an end, and so Black resigned. Aronson-Tal 24th USSR Ch., Moscow 1957 Dutch Defence LD. Strictly speaking, the procedure examined in this chapter was employed here by Black in a counterattack, since at one point he simply overstepped the mark, and in general the famous "sense of danger" of the young, even very86 Attack with Mikhail Tal young Tal was still virtually in its em- bryo state... M.T. Well, of what significance is this? Interference ~— it is the same, both in attack, and in counterattack. 1 d4 e6 2 «4 5 3 |f3 |t6 4 c3 Me7 5 gs 0-0 6 fe2 6 7 0-0 Wes 8 Hel wee 9 e4 fxe4 10 xed xed 11 xed Heb The unprotected state of the white rook is clearly illusory: after 11... Wyxe4?? 12 €yh4 the centralised black queen is trapped. 12 Be2 Af 13° Qd2 eS At just the right time: had he al- lowed the white bishop to go to c3, Black would have been cramped. 14 dxe5 dxeS?! Black deliberately weakens his po- sition for the sake of complicating the play. The “normal” 14...2)xe5 15 43 S)xf3+ 16 Wxf3 Qf5 17 He3 and Bb8 would have led to approximate equal- ity and a more than probable draw. M.T. But it wasn't for the sake of such tedium that Black moved onto “Dutch rails" from those of, say, the Orthodox Defence. IS acs GMS 16 h4 fixhd 17 Bxht- Bae’ 18 bd! A natural and powerful reaction to Black's antipositional play. His pieces have abandoned the queenside, and it is here that White begins his offensive. fe 19 Wed h6 In search of tactical chances, asso- ciated with the one current drawback of White's position — the restricted position of his king's rook, 20 b5 4d8 2 Md5+? The simple 21 cS would have seized a mass of space, opened a path for the rook along the 4th rank, and White's advantage would have grown mark- edly. But now the greater part of his advantage is lost, since his bishop at d5 is not only an attacker, but is also the target for a counterblow, 21... Bhs - Of course, not 21...@h7?, when af- ter the exchange of bishops (22 4) Black's chances of counterplay would have been close to nil. “ 22 £47 : Now the character of the position changes sharply, and the main role « begins to be played by the dynamics of the pieces — and Black's are devel- © oped at least as well, and are better coordinated. Of course, trying to win the ex- © change by 22 @b4? would also not have succeeded: 22...25 23 Qxf8 Bxf8 -: 24 He4 c6 (here it is, the vulnerable position of the bishop at d5) 25 Bxe5 exdS, and the threat of 26...4\f7 leaves White no time for 26 cxd5. But even so, 22 cS would have re- tained for White slightly the more pleasant game after 22...c6 23 bxc6 &xc6 24 Bel. 22... exf4!Lines of Communication 87 23 wyd2 Wb6t! By luring White's bishop to d4 and thereby weakening his control of el, Black already has in mind the start of his counterattack. 24 Odd 25 wxf4 On 25 &xf4 Black would have con- tinued simply 25...e6, but now this does not work in view of 26 Bxh6+. In addition, the c7 pawn is attacked. 25... @h7 26 Wyxe7 This looks the logical continuation of his preceding play. Besides, 26 Bfl runs into 26...Qh3!, when the white king faces an unpleasant journey, But now too a surprise - the theme of this chapter - is awaiting him. e6 re ‘ 26... Abit! By shutting out the white rook from the defence of the back rank, Black ensures the invasion of his heavy pieces into the opponent's position. 27 Bes Deb 28 wd6 wes 29° fd White is even prepared to give up material — 29...g5 wins a piece, but Black prefers to continue his attack. 2... Ags 30 wh4 fet At last White's rook can come to the aid of his king, but it is too late: he loses one of his strong bishops, and all Black's pieces have already taken up attacking positions. 31 &xed = Axed 32 Bf Ber 33° Wd6 With a last hope of catching Black in a trap: 33...¢)h3+2? 34 Bxh3 eyxh3 35 tyd3+ and 36 eyxe2. 330... Bxa2d Of course, not in order to win a pawn, but to allow the queen onto the 2nd rank in front of the rook. 34 WS wer 35 cS Bds!! A very fine move, but more impor- tant, an instructive one. The attacking side does not have the right even for a second to forget about the opponent's potential threats. Thus if here, instead of the text move, Black had played the "natural" 35...8¢8, with the threats of -Ee2 or ...el, White would have had available the winning 36 Bxh6+!!, e.g. 36...gxh6 37 wd7+ g6 38 wyxe8+ Bh7 39 we7+! go (or 39...Hg8) 40 Adz! 36 Ad6 Bes Now, when the bishop has stopped “looking at" the black king, this inva- sion of Black's third major piece de- cides the game in his favour. White was "saved" from capitula- tion only by the flag on his clock fall- ing.88 Attack with Mikhail Tai WHAT WOULD YOU HAVE PLAYED? No.16 For the moment f7 and h7 are defen- ded. How can the attack be intensified? No.17 After 23 wh6 Black resigned. Of course, his bishop and rook are attack- ed, and if 23... §e7 there follows mate at g7. But, after all, there is the simple 23...83xg3, and if 24 eyxf8+ Bp8. So was his capitulation was premature’? Tf 1..4yh3+ 2 gl Be5 3 Hed, or 1 Dh} 2 3, while if 1... He5 2 He Bi 3 &\f6+. So does he have to be content with perpetual check? ED, Uh Wty Y Pas Y ) ; 2 - Ab de An attack on 7 is imminent, and th white bishop has its "X-ray" sights set on the black king. But where is that last straw that, so to speak, is capabl of breaking the camel's back? After ail, if 21 g3 Yc7!, defending everything. No.i9 How can Black continue his attack6 Outposts MILITARY DICTIONARIES give a single interpretation of the concept of an outpost: "a well defended fortifi- cation in an advanced position". In the Middle Ages they could have been bul- warks in front of the bridge, not allow- ing an enemy to begin an immediate assault on some nobleman's castle. Later these were fléches and redoubts at Borodino, Waterloo, Port Arthur... Chess has added another meaning to the word "outpost". On the board too it is a securely defended fighting unit, only not with the aim of defence, but for attack. Being advanced, beyond the demarcation line, even within the heart of the opponent's position, the outpost cramps the enemy forces, hinders man- oeuvres aimed at bringing up pieces to the main part of the battlefield, and finally, controls various squares. Relying on the outpost, the attacking side assembles his forces and prepares a decisive assault, in which the outpost piece also usually plays a part. This is why not every advanced and securely defended piece has the right to ‘call” itself an outpost. In the next diagram it is clear that White's "interests" are by no means limited to the queenside, and therefore the role played by the white knight can be called anything you like, only not that of an outpost. Essentially it is merely covering the black passed pawn, and this is too modest, too thankless a role for it. After all, White is a pawn down, and to achieve any kind of dynamic balance he must re- vive his dying initiative. The move sealed by him was an ex- cellent one! 7 at mise Gs nets i: OS Makarychev-Kholmov Frunze 1989 41 Nd8! At the cost of two tempi White transfers his knight to e6, creating a genuine outpost, and its role in the attack on the king very soon becomes apparent. 41...b4 42 Hxad Yyxad 43 Se6 was After the seemingly more consistent 43,..03 White has the prosaic defence 44 @xc7 Wa2 45 Qc. 44 Wi2 wad 45 Gd2 b3 46 we5 Wxd2 47 tyxe7 De8 48 eyxcd! b2 49 ¥¥c6, and instead of allowing perpetual check after 49...b1=¥¥ 50 ¥yxe8 fol- lowed by 4)g5-f7, Black preferred to give it himself: 49..2f6 50 ¥yb790 Autack with Mikhail Tal )e4+ and then ...o)f2-g4. A far from obvious plan for setting up an outpost was carried out by White in the following game. Bareza-Panno Amsterdam Olympiad 1954 Naturally, White can increase his activity only by the break d4-d5, fol- Jowed by occupying the central file that is then opened. But he discovers a potential weakness in Black's position and begins creating it. 26 4e2 Hic? 27 03! b6 The c6 point has been weakened, and the breakthrough now has the aim not of opening files, but of establishing an outpost. 28 d5! cxd5 29 exdS Syc5 30 Gxe7 xe7 31 Hed wWh7 32 HeS bS 33 406 Has ‘The black rook has been denied the d8 square. But even more important is the fact that the knight outpost on the queenside is aimed in particular at e7, and this is close to the black king. 34 Se2 Wh6 35 We2 Hb7 36 Wed+ £6 This creates fresh weaknesses in the king's pawn screen, but there is nothing to be done; 36..,.g8 37 WeS He® 38 Wid Bg7 39 Dd! is no better. 37 S)f4 4)d6 38 Wyd3 He8 39 Hel 40 Heb+ Sxe6 41 Axes is threat- ened, when the pressure on e7 (with: the active participation of the outpost!): becomes unbearable. Black's desperate: attempt to free himself runs into a little 2) combination, : 39..e5 40 dxe6! Bxc6 41 e7 £5. (otherwise White invades at 6) 42: exiS=ty+ Axf8 43 Yyd7+ Bi7 The outpost has done its work. White has broken into the opponent’ position and the simplest now would to give mate by 44 eyxi7+ oxi? Bd7+. In fact he satisfied himself with. } the exchange — 44 4e6+ Hxe6 45 Wyxe6, and won on the 53rd move, For a model example of a piece owt post in the centre, the authors hat chosen (though not unanimously, one: being “for", and one abstaining} @ following game. Tal-Speelman Subotica Interzonal 1987Outposts o1 It is sufficient to say that the black knight reached c7 not in two moves (...4)b8-a6-c7), but in four (...6)b8-d7- b6-d5-c7)! Therefore White is obliged energetically to exploit his lead in de- velopment. 14 He5t The price for this eternal knight in the centre is very modest. 14...tyxd4 15 4.3 Wd8 16 Sets The alternative was 16 Hdl wWe8 17 ¥d3, with the threat of Ag4. 16...d)ce8 17 pdt b5 18 g5! This is not stronger, but it is more pleasant than 18 4\xc6 tyc7 19 cxbS. 18...hxg5 19 2)xg5 Ha6 20 wf3 b4 21 Wh3 g6 Otherwise mate. 22 Axg6! bxc3 Or 22...fxg6 23 xg6 Be? 24 Wh7 mate. 23 Sext7 wd2+ 24 Ofl Axf7 25 Qxf7+! Og7 26 Hel Aiming not at Black's queen, but at his king — 27 Sf3+ @f8 28 wh8+ Qxf7 29 se5 mate, 26...%9xg5 27 Exgs+ @xi7 28 bx- 3! The last precise move: after 28 Wxc3 es 29 Wi3+ A8i6 White would lose the exchange on account of the threatened check at e2. 28...e5, and after dejectedly making this move, Black resigned. Here we saw nearly all the functions of a piece outpost: the support and organisation of a flank offensive, and the landing of a decisive blow. That these actions are typical is confirmed by numerous practical examples, such as the following. 14 p41 4b7 15 g5 405 Keres-Petrov Estonia v. Latvia 1939 The rook is immune: 15...Qxh1 16 gxf6 Axf6 17 Ad7 Web 18 Axf6+ gxf6 19 tyg4+ Bh8 20 wyh4 fS 21 Wf6+, when White mates as he pleases. 16 Bhgl! This is stronger than 16 4\d7 Wyc7 17 &)xf8 4)f4 with counterplay. 16.447 17 Bed With the help of his outpost White prepares a powerful attack. He is al- ready threatening 18 xh7+, with mate in a few moves. 17...26 18 Bh4t Renewing the threat against h7, but now with the rook. 18.....d6 19 Ded A weakening around the black king has been provoked, and the knight outpost itself joins the attack. 19... fc8 20 @b1 Though not without its uses, this prophylaxis against a check at f4 is nevertheless unnecessary. However, White has no reason to hurry.92 Attack with Mikhail Tal 20...b5 21 Bgl ae7? White was threatening 22 |f6+ xf6 23 gxf6 followed by the un- avoidable 24 Hxh7 @xh7 25 gyh5+, but why lose immediately? 21..4¥e7 was more tenacious, when White con- tinues 22 9.e4, threatening 23 §.xd5 and 24 4)f6+. 22 Hhh6+ SB 23 Dxf7 Ac 24 &xh7 9.44 25 tyxe6, and soon Black resigned. Tal-G.Garcia Yurmala 1983 With his last move (14...9e7-f6) Black has prevented the advance of the white pawn centre, which was possi- ble, for example, in the event of 14... Ag6 15 c4 tyd7 16 d5 (Karpoy-Por- tisch, Luceme Olympiad 1982). But now White gains the opportunity to put into operation another plan: to create a Piece outpost at e5 and with the help of a flank pawn attack to drive back the black pieces to unfavourable positions. 15 Qf4 Bac8 16 Be3 a5 17 Hes 518 ga! 0.86 19 2.2 wd8 20 dxcs! The threat of moving the white out- post a further step forward, to d6, and. the fact that 20...Axc5 is not possible on account of 21 447, force Black to spend a tempo on capturing the cS pawn and restoring material equality: But during this time the attack on his: king comes to a head. 20...Axe5 21 BxeS Wi6 22 423 &)c4 23 g5! Yya6 24 He7 The first achievement! 24...Hixc5S 25 wyd4! The tempo for mobilising White's queen's rook is more precious than pawns — his at bS and Black's at b7. 25...HixeS 26 h4 HaS 27 Hael h6 28 fl bS 29 He8! Against the mate on the 8th rank. there is no defence (29..4¥b6 30 Bxf8+ Gxf8 31 wd7 Ges — 31...f5 32 Wyc8+ with mate in three moves —. 32 He8+ Bh7 33 wc8 Ac2 34 hS), and so Black resigned. When a player has an outpost in the centre, with its help he can plan and. carry out flank attacks not only with pawns, but also with pieces. Keres-Kurajica European Team Ch, Kapfenberg 1970Outposts 93 White exploits the strength of his advanced knight to the full. 23 wyxh6+! gs If Black takes the queen, he remains two pawns down with a bad position. White could again have forced this by continuing 24 gyh8+, but he now has a better alternative, 24 Qh4 &)8h7 25 s\xe6t And in view of the mate after 25... fxg6 26 Wyxp6+ f8 27 g3, Black resigned. Ivkov-Kolarov Wageningen 1957 From this position the game did not last long. White played 1 a4, and after 1..bxa4 2 Hxad Wd7!! he resigned, since whatever he does the rook at f7 is reed" and gives mate at fl. The weakness of the back rank certainly told here, but firstly, Black's occupa- tion of the f-file was assisted by his outpost, and secondly, the white rooks were so cramped by the black bishop that the attempt to free them can be considered obligatory. White's position in the following diagram is strategically superior thanks to his occupation of the c-file and the two bishops. Kasparov-Butnoryus USSR Spartakiad, Moscow 1979 Therefore Black's tactical attempt 15..84g5 16 £4 Yyg6 17 fxe5 Ac5 18 fig3 &xa6 would have been quite justified, had it not allowed White to create a powerful piece-pawn outpost. 19 4)f5! Bae’ Naturally, 19...exfS 20 ¥yxd5 would have been practically equivalent to capitulation, the white bishop being so much stronger than the black knight. 20 £\d6 He7 21 Hf4t Threatening in passing 22 Sg4 wyh6 23 4 eyhS 24 Hxg7+, White creates decisive pressure on f7, begun by the knight manoeuvre to d6. 21...n5 22 e4 a8 23 And Hd7 24 Hic3 wWh6 25 eyfl Ac7 26 Hef3! £5 There is nothing else, since on 26... ¥g6 there would have followed 27 $Lf6, with the threat of 28 823. 27 exf6! “The Moor has done his duty..."94 Attack with Mikhail Tal Sacrificing both his outpost, and a piece, White begins a concrete vari- ation. 27...xd6 28 17+ @h7 29 Qe7 e5 30 Axf8 exfd 31 Axd6 eyxd6 32 Wad3 we, and here 33 Wd7 wyxd7 34 f8=4\4+ would have won for White, as would 33 Wed @h6 34 f8=vy wyxf8 35 Wxc7 25 36 83. He in fact chose the more compli- cated 34 Bxf4 Ne6 35 Wyc8 Yyd6 36 Wh8+ Og6 37 (H=)+ HxfB 38 wyxfB Wadl+ 39 Of2 wd2+ 40 Ygs Wels 41 Qh3, and since 41...A.xe4 fails to 42 Wye8+, Black resigned. As a rule, the deeper the attacking side is able to establish his outpost in the opponent's territory, the greater the difficulties it causes the opponent. The defending pieces become even more cramped, the proximity to the king becomes even more dangerous... Bakinskaite-Voronova Vilnius 1985 The role of outpost is being equally shared by the black knight and the f3 pawn. They are taking away squares from the white king, and this coul have been exploited by the forcin; manoeuvre L.wWhS5 2 2\xe4 (other. replies do not change anything) ¥Wh3 3 e3 wyxh2+ and 4...8h5 mate. Suba-Tal Tallinn 1983 M.T. In the opening for some reasot my opponent allowed me to make knight move, about which there is waming in all books on the King‘ Indian Defence. Black's position ha: immediately become the more pleas- ant, but it is too early to exchange: mind for matter: 20,..Qxf2 21 @g2: Abb 22 Ge2 He5 23 Axed Axed 2: Hd7, and White's active rook com pensates for all his losses, Besides, it was tempting to continue the pursuit of the king, which even without th queens is perfectly possible: along th b-file! 20...n6! 21 gxh6 Qxc3! Essential, otherwise after 21...25 22 Bxd4! Qxd4 23 Qxg5 it is altogether unclear who stands better. 22 bxe3 g5 23 Qc7 Hac8 24 4d6Outposts 95 Keeping this square for the rook — 24 Bxb7 Be6 25 Bd6 does not suc- ceed: 25...Exc7. 24..He6 25 h7+ Qxh7 26 Ge2 Bh6 27 243 There is no other square for the bishop. 27...xd6 Not 27...4h2+ 28 @fl Bhi+ 29 Gg2t Bxdi 30 xd] Ayh4+ 31 gxh4 _ &xdl 32 hxgs, when it is White who has the attacking prospects. 28 e5+ Hxd3 29 Hxd3 Afs 30 Hxf3 (30 Ghi+ 24+) 30..axbi 3t Axf7+ &g6 32 Hxb7 Axaz 33 £4 (33 Bxa7 Qd5+ 34 @fl g4, and White can resign) 33...gxf4 34 gxf4 Ba8, and me 5G moves later Black eventually converted his advantage into a win. For the creation of such a powerful outpost one must sometimes be pre- pared to pay an appropriate price. ‘Tal-Gligoric Moscow 1963 This position kad occurred fre- quently in practice, and had been han- dled in different ways. The Yugoslav grandmaster himself, for example, against Fischer in the 1959 Candidates Tournament, tried here to exploit the positioning of the white bishop and black king on the same h3-c8 diagonal with 15 £5, which is perfectly feasible. LD. Why did you reject a well-tried and promising idea? M.T. Least of all because the author of it was Gligoric, and not at all for the reason that at the start White makes a very small sacrifice — a piece for two pawns. I was attracted by the power of the white knight at e6 and realised that with the help of such a knight the remaining white pieces would be ca- pable of anything. 15 Qxe6! fxe6 16 xe6 wes 16..%b6 must be considered the lesser evil, although even then after 17 ®d5 Axd5 18 exdS White has suffi- cient compensation for the piece. 17 d5t ‘This is the whole point. On his 15th move White had no intention of ex- changing two pieces for a rook and two pawns, but was intending to demon- strate that the splendid knight at e6 is worth more than the fat and unwieldy took at 8, 17..xd5 18 exdS @b7 Forced, since 19 b3 was threatened, and 18...4)c5 is bad on account of 19 3 We4 20 Wyc3, when Black is unable to parry both 2] b4 and 21 Bhel. 19 b3 We8 20 Bd3 Hb 22 Hed WaT 22 Be7+ The white knight has after all done its duty. 22...8xe7 23 |XcT Oxe7 24 We3+ Sb3 25 wyxe7 Ac8 26 Hel White has a material advantage. In96 Attack with Mikhail Tat addition it is very difficult for Black to coordinate his pieces and bring them out into play. The outcome of the game is decided, and an illustration of this is the possible variation 26..Rhg8 27 ¥xh7 Bh8 28 Bxe7! 26..Hdg8 27 Yyd4 Od8 28 Heb BES 29 h4 h6 30 26 Ahg8 31 nS BES In the event of 31...8xf4 32 wyxf4 gS 33 exes hxgs 34 h6, in place of one queen White quickly acquires another, 32 wed KxhS 33 He8 Hxe8 34 Wxe8 {£6 35 cd! bxc4 36 bxc4 Bh3 37 Sd2 Gc3+ 38 He2 Add 39 £5 Hxa3 40 eS! dxeS 41 d6, and after giving a few checks, Black resigned. Until now the diagram positions have mainly shown piece outposts. The other type — pawn outposts, are even more common, Their existence is dete- mined even by opening theory: for ex- ample, in the French Defence a white pawn is established at e5, and the eval- uation of many variations depends on whether Black can destroy this outpost or will be forced to endure its pressure, Moreover, it should be bore in mind that a classic piece outpost, sup- ported by a pawn, has the property that it can be transformed into a pawn outpost, with all the attendant conse- quences, In the following diagram Black has no weaknesses, but he also has little breathing-space, mainly on account of the powerful piece outpost in the centre. With his next move White further strengthens it, 15 £4 We7 After the exchange 15..2)xe5 16 dxeS the piece outpost is transformed into a pawn outpost, which in tum secures White control of d6, and after 16...e¥e7 17 Hxd8 Wxd8 (alas, forced, otherwise the a7 pawn is lost) 18 dl White, after seizing the only open file, creates a new outpost at d6, Spassky-Donner Leiden 1970 With the move in the game Black. begins a battle against the e5 outpost, planning ...f6. But this creates a weak: ness at e6, which means that White's outpost has done its job. . 16 h4! £6 17 DB Qh5 18 gh3 Axf3 19 Wxf3 HS 20 Bhel wt? 21 Af] Hd6 22 Qc4 Had8 23 £5! Bxd4 24 fxe6 Bxdi+ The blockade of the white pawn does not succeed: 24..8¥e7 25 Wa} c5 26 tyxa7 Exdl+ 27 Bxdl Bxdl+ 28 Wxd! Dg6 29 Wyxb7! 25 Exdl Bxdl+ 26 wxdl Black resigns (if 26...eve7 27 ¥yd8). The next diagram is another typical transformation of an outpost — and its fruits,Outposts o7 cee RA S faa Larsen-Tal Leningrad Interzonal 1973 After further strengthening his out- post by 19..e5, and at the same time removing the pin on the knight, Black invites his opponent to give a reply to the etemal question - "what to do?" 20 Dxd4 After 20 442 White would have had to reckon with the flank attack 20...g5, again under cover of the out- post in the centre, 20...exd4 21 dyed Wh5 22 wad The lesser evil was 22 e3 dxe3 23 Bxe3, and although Black would have immediately restored his outpost with 23,..Hd4, nevertheless the unpleasant pressure of the black d4 pawn would have disappeared. But now this outpost also plays the role of a barrier, cutting off the white queen from the kingside. 22... fe8 23 Yyxa7 Hdé Not so much to defend the b6 pawn, as to attack f2. 24 b4 Bf6 24...cxb4 was also good, but the position is ripe for a forcing variation. 25 bxe5 wWf5! Inducing a serious weakening, since 26 Hfl loses immediately to 26...8xe2. 26 £3 Yyh3 27 Yyc7 HFS! A quiet move with an irresistible threat. As soon as Black places one of his rooks at e5, the white king will be absolutely defenceless. 28 exb6 BfeS Or 28...Hee5 29 g4 ¥yg3+ 30 Whi 0 @fl 8xf3+) 30..8h5!, which would also have won. 29 e4 The main variation of Black's attack was 29 tyxe5+ Hxe5 30 g4, hoping with rook against queen to exploit the strength of the b-pawn.Then he can continue 30...47¢3+ 31 @hl (after 31 fl wyh2! White cannot do anything, since on 32 b7 there follows 32...b5, when 33 Hb1 is not possible, while if 32 a4 the attack can be continued by 32...n5!, e.g. 33 b7 Be7 34 Habl hxgd 34 b8=W g3) 31...h5! 32 b7 HbS 33 Bebl bxg4! 34 Hxb5 wh3+ 35 Sel 33. 29...8yxg3+ 30 @hi wyxhd+ 3t Bg2 Hes+ 32 Sl wh3+ 33 we2 He2+ 34 Odi wxf3+ 35 cl WR White resigns, Even in the concluding attack the pawn outpost at d4 played its part, tak- ing a square away from the white king. A pawn outpost, even more than a piece outpost, gains in strength the more advanced it is. On the 6th rank the pawn is not only close to its queening square, but — in the context of our attacking theme — it usually becomes virtually the main component of the mating net around the enemy king. This is so obvious that it will be98 Attack with Mikhail Tal sufficient to give one single example from “girls™ chess... A, 4@ a # a A emma Ziedinya-Auzinya Riga 1980 A part is undoubtedly played by the inadequately defended back rank, but the chief credit in the mate: 1 Wg6+! Axg6 2 hxg6+ Yg8 3 Hb8+ neverthe- less goes to the pawn outpost at 26. # is ea aa Mae # Levenfish-Ryumin Moscow 1936 And now here is one example from "men's" chess, where such an oppor- tunity was missed. White has a clear advantage: better” development, outpost at e5, occupatior of the open file, and a knight that has.:: been "launched" into the vicinity of th black king. All this could have beet realised instantly — 27 4)f6+! exf6 2! exf6, when the f6 pawn “guarantees’ one of the two mates — 29 gyg3+ ot 29 Wxf8+. But after the lame 27 4)g3' Black was able to save the game: 27...%b5 28 Hfet {.b7 etc. And now another slightly exotic ex. ample of how dangerous it is to allow. the opponent to set up a bind with pawn outpost in the centre... Kaila-Kivi Helsinki 1949 1 e4 c6 2: d4 d5 3 exd5 exd5 4 c4 dxe4 5 Axc4 e6 6 c3 Hd7? 7 d5 5: &xe5+ OhS 14 He2 (bringing up th assault cavalry) 14..W%e8 15 wf3: Sigd 16 tyxgd+ Dxgd 17 3+ Ghd 18 £)f3 mate. This game should not be taken too™ seriously, but as the saying goes, “i every joke there is a grain of truth". A pawn outpost contains an addi- tional, latent strength, when it is not blockaded and retains its mobility. In the next example, in full accor- dance with opening theory, back ot move 1] Black advanced his pawn to f4, and with energetic play he has: made more progress on the kingside, than White on the queenside. But t "get at" the king is not so simple: if 39,..@h1 40 eed, and he has to reckon =Outposts 99 with the thematic Hc8. Ame VY "GY pene aa te ae mee a Najdorf-Gligoric Mar del Plata 1953 ft Everything is decided by the energy of the pawn outpost, whose main function has already been fulfilled. 39...£3! 40 tye3 (all three captures _ lead to the loss of the queen!) 40...8f4 41 gxf3 But now the king is completely ex- posed, and the attack on it proceeds from both the flank and the rear. 41..h2+ 42 Bel whl+ 43 Ge2 QS 44 Gd? Bxf3! 45 WygS+ Ge7 46 @e2 (another check would not have changed anything - 46 Hc8+ @h7) 46...Hf24+ 47 Qd2 Wdl+ 48 Yc3 Wal+, and in view of the inevitable mate (49 @d3 e2+ 50 Be - or 50 Bc2 Adl+ 51 Sd3 wol+ - 50.,.4f3+ 51 @xe2 Wil mate) White resigned. The above ideas were as though synthesised in the following example. Black would appear to have achieved a great deal, but... While still on the way to this position, White had planned to eliminate the enemy out- post, realising that the basis for creat- ing his own had already been laid. Chekhov-Pahtz Lvov 1983 23 Hfxe3! fxe3 24 Yyxe3 wyeg6 25 Hed! No deviations from the general course, such as 25 )xe5... 25..Haes The knight could have been pre- vented from going any further by 25... h6, but then 26 ¥h3 with a very strong attack on the light squares (h5). 26 gS Ph8 27 Le6 HES Returning the exchange by 27... Bxe6 would have simply meant losing the eS pawn. 28 Qf3 Bg8 29 3 BET Evidently 29...49f7 would have been slightly more tenacious, 30 Yye2! wWh6 31 AhS HES 32 Aged Hf6 33 4)xe5! Ac8 The pawn has been captured at just the right time, since on 33.,.6)xe5 34 Wrxe5 Hxe4 35 Hfl Hes6 36 Hxf6 Hxf6 White wins by 37 Wyb8+. Here too the e6 outpost plays a decisive role.100 34 4)xd7 Axd7 35 WeS! Axeé The white bishop is still immune, but now the piece outpost is trans- formed into a pawn outpost -—~ or a far- advanced passed pawn! 36 dxe6 Hxgd 37 Hfl WeS If 37...g7 or 37...Agg6 the simple 38 e7 wins. 38 Yyxf6+ Wyxf6 39 Hxf6 Hed 40 @f2 Sg7 41 Hf3 Hel, and simulta- neously with this move Black re- signed, Polugayevsky-Tal Riga Interzonal 1979 English Opening M.T. This game was played exactly ten years after, in September 1969, in the 2nd round of the 37th USSR Championship, Polugayevsky had opened the score in ovr personal encounters. That game received wide recognition, and found its way into the theory books. And now it was September 1979, again the 2nd round (the events, it is true, were different); naturally, this sort of association did nothing for my frame of mind, and besides, in the intervening decade Polugayevsky had increased the score to 5-1. But... that is the way the pairings came out! 1 5} Of course, an exclamation mark is attached to this move not because it is the strongest, the best. Simply, I did not want to allow my opponent the possibility of repeating the opening of the first game won by him, an extract from which is given in the chapter Breakthrough in the Centre, Attack with Mikhail Tal 2 4 ate 3 &c3 a5 4 exdS ayxd5 5 e The move 5 e4 has enjoyed its “ebbs" and "flows". According to thé latest theoretical conclusions, it casts doubts on the 3...d5 variation. Incidentally, the game Timman-Tal (Montreal 1979) developed in the same: way. Black played 5...)xc3, and after: 6 dxc3 he ended up in a difficult position. Only then did I realise that. the offer to exchange queens here is by: no means synonymous with an offer of: adraw. Possibly 5 e4 is not such a terrible. move for Black. In particular, three. months later this was demonstrated by. Vaganian in a game with me in the. USSR Championship in Minsk. After 5...4)xc3 6 dxc3 Wxdl+ 7 @xd! the Yerevan grandmaster played the much. stronger 7...4)c6 8 e3 b6 9 a4 Ab7 10 Ad2 0-0-0 11 @c2 a5, ar achieved a perfectly acceptable po: tion. But at any event, in September I di not yet know this game, and neither dic Polugayevsky. nee abd ‘This is considered the most critical move. As it transpired only after the toumament, my opponent had specially prepared this variation for his game. with me, And it so happened that my: trainer and I had also specially exam- ined this very continuation before the game with Polugayevsky. 6 ged The bishop check at b5, which has 4 recently been fashionable, achievesOutposts 101 little after 6...4)8c6 7 d4 (this is much stronger than 7 a3, as Poutiainen played against me in Tallinn 1977, or 7 0-0, the continuation employed against me by Tukmakov, 45th USSR Cham- pionship, Leningrad 1977) 7...cxd4 8 a3, and here the simple 8...42d7, if there is nothing better, secures Black a - practically equal game. eee Ae6 The whims of chess fashion! — I suspect that 50 years ago this con- . tinuation would have shocked its legis- lators. But now it is considered the vongest, and virtually the only move. Black allows the doubling of his pawns on the e-file, obtaining in return only ope compensating factor, but an at- tractive one: his knight ensconces itself at d3, but whether its residency will be temporary or permanent, depends on So far this is a standard variation. Until quite recently it was thought that Black could achieve an excellent game here by continuing 9...45c6 10 4)xe6 Wd7 11 Hd5 Bc8 ete. But in the game Timman-Stean from the Amsterdam Zonal Toumament (1978) the Dutch grandmaster prepared a significant im- provement: instead of 1t 4d5 came the simple, but nevertheless unexpec- ted 11 4)xc5!, and it transpired that after 11...2)xc5 12 wyh5+ g6 13 YxeS Black has an initiative that might com- pensate for one pawn, but two pawns in this position is too high a price for it. In short, mind gave way to matter. This game was widely reported, and searches for an improvement in Black's play began. Players reverted to the plausible 9..4¥d7, and 9...2)c6 was tried. And in the game Sekey-Palatnik (Frunze 1979) Black employed the experimental 9.,.4)a6. True, from this game it is not possible to give a real assessment of the move, since imme- diately after 10 Yya4d+ ¥d7 the players agreed a draw. But at any event the position had become problematic, and one of the attempts to solve this prob- lem was the move employed in the present game. Dd we Wbel? I am not inclined to overrate the strength of this move, but in the given instance the experiment fully justified itself. During a training session three of us analysed this position: Kapengut, Koblenz and I. And the following move, made by Polugayevsky, was suggested by Koblenz. 10 tye2 Other continuations would seem to be more dangerous for Black. In par- ticular, 10 ¥yc2, 10 wf3, or 10 4)a4. However, analysis in one's study is one102 thing, playing in a toumament is an- other. It seems to me that only the latter can serve as a true criterion. 1... cd lL b3 This move is the idea behind White's plan. His primary task is to deprive the advanced knight of support, since it is perfectly evident that nearly all Black's remaining pieces are still occupying their "god-given" positions. But, strangely enough, Black never- theless succeeds in consolidating his grip on the “enemy beach-head". Jumping ahead I should mention that the knight holds its ground at d3 almost to the end of the game. Wl... h6t The point is that White's forces too are not well placed. The knight at g5 is his only active piece, With his last move Black drives it back, and at the same time prepares to develop his queenside. He does not fear the check at h5, since after 12 tyh5+ @d7 his king moves to safety, while in certain cases the position of the enemy queen at hS even gives him a tempo for the developing move ...g6. 12 £3 kt is important that the attempt at an immediate attack does not succeed, since after 12 a4 Wd4 (12...49c6 is also possible) 13 4)xe6 Wxal 14 WhS+ gO! (14.7? 15 BYd5+ Bc 16 Wd8 mate) 15 eyxg6+ Bd7 no decisive continuation is apparent. It is probable that 12 2yh3, covering £2, was slightly more accurate, but in this case, I think, there can be no question of the variation being refuted. 12 0. D6 Attack with Mikhail Tal 13° bxc4 0-0-0 White's first sortie has not succee ed, The black knight has not only he on to d3, but has even, it could be sai established itself more comfortably. « The evaluation of the plan chosen by Black depends largely on the sharp variation 14 4)d5!? But during the game both players established that. after 14..exd5 15 tyxd3 Black can’ interpose the strong move 15...4)b4!, when White's plan is essentially futed. For example: 16 ¥yb! dxc4, 16: Wyd4 eyxd4 17 &xd4 dxe4, or 16 Wei dxe4 17 He5 g5 18 Hf7 Ag7 19 Wxe7 43, and the other black knigh settles in at the sacred d3 square, 14 9143 was also examined, In the event of 14...¥a5 15 4)b5 a6 16 Ad6+ exd6 17 ¥yxd3 Sye5 18 Wb3 Syxf3 19: gxf3 Yyxd2 20 Hb White has a good. game, but Black was planning to an swer 14 9.23 with 14...g5. Perhaps some consideration should be given to the positional piece sac fice 14 )d5 exd5 15 cxdS 4)ce5: although White does not appear to gi a great deal in compensation. Be that as it may, but White's rather phlegmatic development leads to very difficult position. Black onl needs to bring into play his bishop and. then to occupy the half-open f-file, and his attack will become very powerful. 4 gs es 15 Bg2 The exchanging operation 15 4d! exd5 16 Wyxd3 Qg7! 17 eS (17 Bb dxc4!) 17...8hf8 holds little promise for White. Now he is ready to "smoke out" the knight from d3 by 4 el, ane therefore, rather than the naturalOutposts | development of the bishop at g7, Black : prefers the following move, which proves very effective. 1S... Wes! On & e208 am tm ys Laing “atecee a AA a Rynoe ae 2 If now 16 Hel, simply 16...kyxc4, and instead of the knight a much "heavier" piece establishes itself at d3. 16 Abi Agi Perhaps only now can one confi- dently state that White's position is very difficult, Sensing this, Polu- gayevsky attempts a counterattack, but it proves insufficiently justified. 17° )b5 wyxe4 Active defence (18...¢)f4+ is threat- ened). Of course, not 17...a6? 18 ga3 Wxe4 19 Ha7+! (19,..)xa7 20 Bhet). 18 tye3 hfs Black's threats are the more imme- diate: 19...8xf3 and 19...g4. Therefore. White is forced to cover f2. 19 Bl 24 Strictly speaking, now 19...a6 was not at all bad, but simply Black had worked out a fairly clear-cut variation. 20 h4 4xf2t 21 26 21 Axf2 Axf2+ 22 @xf2 HfS+ or 103 22 Wxf2 Yxe4+ is quite unacceptable for White. Polugayevsky was probably pinning some hopes on the interposi- tion 21 )g6, although after it the white king is too alone. 21... Bd3! In this game the d3 square has pro- ved to be something of a staging post for Black, 21..8f3 would have been much weaker, since after 22 4)xa7+ Q)xa7 23 Wyxa7 wyxe4 White has 24 Qxe7+ QGe7 25 wWo6+ wd7 26 ¥yxb7+. Here it is important that the e4 pawn is not captured with check. Thus now this variation does not work, since after ...4yxe4+ Black gives mate. 22° ad Another interposition. 22 wel Hdf3 would have led positions similar to those in the game. For Black there is naturally no point in considering variations where the queens are exchanged — his main aim is to maintain the attack on e4, There- fore his next move suggests itself. 22... Wad 23 wel Adf3 Now there is nothing to prevent the104 black knight from returning to the fatal (for White) square, and what's more, with much greater effect than earlier. To complete the picture it should be added that by this time White was already significantly short of time. 24 Axf$ = =Ad3 25 wdl Not without its piquancy is the var- jation 25 Wye2 4\d4 26 Wd! wyxdi 27 Hixdl Hf2+ with mate after 28...6sf3. Here too 25...54f2+ was very strong, but for some reason I did not want to allow White the possibility of sacrific- ing his queen, although, of course, this would not have saved him. I chose the continuation in the game, only because I had planned it beforehand. 25... wyxe4 26 8xf3 exf3+ 27) Sfl wis 28 gl Ada White resigns WHAT WOULD YOU HAVE PLAYED? No.20 Attack with Mikhail Tal The attempt to set up a piece outpost at d6 does not succeed: if 19 c3 or 19° §ic3 Black immediately replies 19, &c5, exchanging knights. But White. already has a pawn outpost at eS... No.21 Taking account of what has been said, there are at least three reasons that allow Black to play... what? No.22, This position, where it is White to move, should not cause any problems.7 Eliminating Defenders DEFEND TO THE LAST MAN! — such was the order usually given by commanders of fortresses under siege, after they themselves had received similar dispatches from their superiors. It is another matter that attitudes have changed to such exalted and threat- ening words, thereby saving precious human lives. And besides, what could objectively be done by a dozen or so remaining defenders against a much greater enemy force? In chess — a wonderful model of our everyday life — things are differ- ent. The dimensions are different, and the role of the single lone "soldier" is iso different. Even on his own, he is sometimes capable of holding a dubi- ous game, or saving a position that, as they say, is at death's door. Therefore the most important task in any attack is to eliminate those defenders that stand in the way to the king, even if a con- siderable price has to be paid. However, there are several ways of eliminating a particular piece, that is the mainstay of the enemy defences. The most convenient for the attacking side is to do this, so to speak, for free — by forcing its exchange. As usual in positions with the type of pawn formation in the following diagram, the main defender of the dark squares is the bishop of the corre- sponding colour, and it is this piece that White begins pursuing. 21 Qg5! a4 22 wes! White is prepared to tolerate the knight at d4 for a time, and to give up the exchange: 22...¢\c2 23 Wyf4 &xg5 24 &\xg5 Hf8 25 De4!, so long as the dark squares become a gaping wound in the opponent's position. Black can no longer avoid the exchange. Fischer-Feuerstein USA Ch., New York 1957/8 22...Axg5 23 wyxg5 S\e8 (the func- tions of the bishop are taken on by the knight) 24 2yg4 4y£5 25 Bacl The position is not the sort where you have to burn your boats in making a swift attack. Besides, in passing the weakness at b3 can also be harassed. 25..A¥C7 26 4)d2 Bd4 27 4xb3 Hixe4 28 Hed! Leaving the black bishop unem- ployed for the time being. 28...Ha4 29 Hed Qb5 30 Hel! ¥yb6 31 Hd2 The white pieces retum to the main part of the battlefield.106 31...Hxe4 32 £\xe4 .d3 33 Def6+ Bhs 34 gd Now both black knights lose control over g7, and White wins easily: 34...xed 35 Axed Add 36 Axe Wd8 37 Wxd8 Axd8 38 |d6, and soon Black resigned, However, more often the exchange of a defender is achieved by sacrificial means, the material losses being more than compensated by the fact that the squares being defended are fully or partly taken over by the attacker. By giving up a more valuable piece for the defender, the attacking side greatly increases the energetic potential of his offensive. There is no opponent, and hence there is no counter to the attack on a particular part of the battlefield. Ubilava-Timoshchenko Chelyabinsk 1974 Black has just attacked the white queen, but for this he has incautiously used the sole defender of his weak dark squares. Its disappearance from the board is equivalent to defeat for Black, especially since a new master of the Attack with Mikhail Tal dark squares appears. 14 WyxeS!! dxe5 15 Hxd8+ xd: 16 h.g5+ Ges 16...d7 is pointless, while after 16...c7 17 Qf6 and 18 4.xe5+ Blacl is a pawn down in a hopeless ending. 17 Qf6 He8 18 Hdl Ad? 19 Had Hb4 It is easy to see that there is no other defence. For example, 19...c5 20 €)xc5. SibS 21 a4, and after taking on a6 th white knight gives check at c7. 20 4\c5 Hdd 21 Hxdd exdd 22 ha! The domination of the dark square: must be consolidated: if now 22...g5: 23 WS, and the black rook is stilt shui in, 22,.e5 23 £3 fe6 24 b3 aS 25 4. HS 26 g5 Qh3 27 Ol2 (the white kin; heads for the queenside, and it is all Tal-Suetin Tbilisi 1969/70 Jn the opening Black has taken too. many liberties. He has left his king ithe centre and exchanged his dark- square bishop, and his knight at g6 is badly placed. In addition, without hin- drance White has “launched" (cf. Chapter 3) his queen into Black's “goal area", He now begins a decisive at- tack, by helping the opponent to obtain an excellent "central" defender, and — by planning in advance its elimination. 16 £5! exf5 17 exfS He5 18 Ae6! Mxe6 19 fxe6 26 Clearly, 19...0-0-0 20 exf7 Hdf8 21 a4 with a crushing attack is not a real- istic alternative. But now after 20 exf7+ @\xf7 21 wyh4 0-0 22 Hdf2 2\c6 Black is somehow able to hold on, thanks in particular to his knight at e5. 20 wxe5! (immediately resolving the problem of the defender and of the defence in general) 20..dxe5 21 exf7+, and Black resigned, rather than face the piquant variation 21... d7 22 Qf5+ cb 23 Ge4+ Hd5 24 QAxdS+ Gd7 25 Axa8+ We7 26 Aes+ SFB 27 Gh6+ We? 28 =wW+ Gereben-Geller Budapest 1952 Eliminating Defenders 107 After castling on opposite sides, pawn storms on the principle of “who is quicker" are a standard procedure, and therefore instead of the possible 17...$,d7 Black preferred not to waste a tempo, since for a start he had a quite specific aim, after, of course, the open- ing of lines. 17...a4! 18 4 For the moment it is bad to accept the gift: 18 fxe6 fxe6 19 Wg3 ghd 20 Wyh2 Hf2 21 hel .g5+, and Black wins. 18...axb3 19 axb3 a2! 20 fxe6 fxe6 21 Wye3 Yad 22 c4 Bxb2! Removing one defender... 23 Qxb2 Wa3+ 24 Hbl Bas 25 cl tyal+ 26 e2 Ba2+! And now a second, which had come to the aid of the king. After this the gathering in of the "harvest" begins. 27 &xad Wxa2+ 28 Bcl Axb3+ 29 wyxb3 Yyxb3 Black now has a material advan- tage, and his attack continues, When the bishop joined in, White resigned. Very often the main defensive func- tions are entrusted to a_ knight (remember the aphorism: “bishops for attack, knights for defence"), but on open or half-open files they can typ- ically become a target for the attacker's long-range rooks (see diagram next page). Contrary to the aphorism just men- tioned, White's attack is supported by his knight, but in addition the black knight is the only (for the moment, at any event, until ...f6 is played) de- fender of f6. And it is from this square that the black king can be invited to go for a walk.108 Nezhmetdinev-Romanoy Arkhangelsk 1950 25 Bxd7! Qxd7 26 wi6+ @hé 27 BES! 27 Bf3 was not dangerous in view of 27....xg4, but now against mate by BhS there is no defence. Unless Black prefers 27...xf5 28 £)xf5 mate. 27..8a7+ 28 Og2 Black resigns, Keres-Szabo USSR v. Hungary 1955 Here we have a position of so-called Attack with Mikhail Tal dynamic equilibrium. White's play is. on the kingside, whereas Black's is on the b-file and against the e5 outpost, which is insufficiently supported, Therefore everything depends on the energy of the two players and — an accurate choice of goal. 18 Bxd7!! Realising that the black knight was ready to act as the sole defender of the: kingside {from £8}, White gains time: for his attack and acquires an advan-, tage in force on the required part of the. board. 18..,.2.xd7 19 Gd3 b6 Parrying the main threat of 20 Wyg5 g6 21 5 with a crusking attack and, perhaps, secretly hoping for 20 ¥yxh6?, WxeS 21 Wh7+ SFB 22 Wh8+ Be7 23 wyxg7 wxg7, which would suit Black perfectly well. 20 ¥yf4 Hi (defending against 21 Wf6) 21 Hxg7! Now the white pawn outpost will have its say, . A.g6, and in view of 23...e7 24 Wh8 mate Black resigned. The next diagram shows a slightly: more complicated implementation of a similar theme. White's standard threats on the h-file are strong, But he does not have time to double heavy pieces by 1 Yyh4 on account of 1...¥¥e3+ and 2...¥h6, when on 3 Bh5 Black can interposé 3...2)xg6. And the typical defence elimination 1 Rxf8+ is parried by xf! (1..Axf8 2 Bh8+ SxhB F Wyh5+ and 4 Wh7 mate) 2 Bh8+ wee, when the game is still alive.Eliminating Defenders Kosten-Sanz Hastings Challengers 1978/9 But from the defending knight White can first remove the defending king! 1 Bh8+!! @xh8 2 Axf8+ Bxf8 3 wWh5+ Black resigns. Vasyukov-Duraseyic Belgrade 1961 White already has a material advan- tage, which he could increase with the simple 1 §ixb7, when Black cannot even escape from the pin by 1...¥c6 on 109 account of 2 Hb6. But the ideal position of White's rook (on the 7th rank) and bishop (on the long dark-square diagonal) de- mands that he look closely at the point where their “glances” intersect — the g7 square. True, it is defended three times(!) and very securely. Securely? 1 &xe6! (the first defender is elimi- nated) 1...Wyxe6 2 Wxf8! (and now the second) 2...ixf8 3 &xg7+ (the third is powerless) 3..@?h8 4 xg6+! (the most precise), and Black resigned, as 4...14f6 fails to save him on account of 5 Bixf6, when the queen has nowhere to hide on a completely open board. ‘The elimination of defenders around the king can be the theme of the most unexpected operations. Saigin-Tal Match, Riga 1954 M.T. It can be considered unex- pected only as regards its ultimate aim, since in general the black pieces are ideally placed, the rook invasion at e2 is threatened, and there are also other motifs in the air...110 25...€)e3! 26 fxe3 Qxe3+ 27 Ohi Mxd2 28 Wxd2 Re2 29 ¥yc3 Hxg2! This is the point of Black's idea. By leaving the white king without its last defender, Black intends to combine the advance of his passed pawn with an attack on the light squares, that are no longer controlled. For example, 30 xg? d2 31 Bdl 94 32 263 wd3! Therefore White resigned. oBae a ae Spassky-Petrosian World Championship, Moscow 1969 Any player with the slightest experi- ence will immediately judge Black's position to be very dubious. Indeed: White has open files for his attack, the number of pawns sheltering the black king has been reduced, all the light squares around it are weak — and all this for one single pawn, But... Firstly, the key point g7 is securely defended. Secondly, the light-square bishop cannot be switched to the bI-h7 diagonal, where together with the queen it could begin an attack on the light squares. Thirdly, there is no dark- square bishop to be sacrificed on the Attack with Mikhail Tal traditional h6 square, and it is not evi- dent how one of the knights can be moved across with this aim. As for the manoeuvre Bf1-f3-h3xh6, it is both lengthy, and it leaves open the ques: tion: what next? Finally, Black has the: possibility of defending his weakness at e6 a third time and of including hi: queen's rook in the defence along th Sth rank. Therefore the means by whic! White eliminates the main defenders of. the black king must be considere: typical. 21 eS!! dxe5 22 Hed! HhS The essence of White's manoeuvre. lies in the variations 22...¢)xe4 23. Bxf8+ Bxf8 24 wyxg7 mate, and 22. exd4 23 4)xf6 g5 24 wWh3 Be7 2. Bxg5 Qe7 26 Bxg7, mating. But now too the additional defence of the g’ squares proves insufficient. 23 We6! exd4 After 23...4)f4 another motif woul have worked: 24 Hixf4! exf4 25 3) and —— an attack on the light square: after (9.02 and 4f6. 24 4)g5t The most energetic, and in view the forced variation 24...hxg5 2. wWxhs+ Yes 26 w+ Gh8 27 BFS Black resigned. Sometimes a defender cannot be: eliminated even at a high price — i too securely defended, But in suc! cases one must consider: can it perhaps be diverted?! In the next example Black appear: to have been the more successful: ht has a material advantage, and the white queen, that has broken into hi position, is restricted by the bishojEliminating Defenders which cannot be advantageously elim- inated. Spassky-Geller Candidates -Final, Sukhumi 1968 36 Gh6! Gxh6 An exchange? But what to do after 37 wrxh6+ Ge? 38 Axf6 wxf6 39 Hxf6 @xf6? — the black rooks are clearly stronger than the white queen! 37 Wh8+! Ge7 38 Hxf6t After luring away the defender of he £6 pawn, White continues his at- ack, although he is now a rook down. 38.04? Black keeps his material advantage, but loses the game. By 38...4f8, bloc- King White's main line of attack, he ould have held the balance: 39 Bel+ wWe5 40 Bxe5+ dxe5 41 5+ Beco. 39 g5 Heb 40 Web+ HFS 41 wxi7 He7 (41...xg5 42 Hed+) 42 wyxc7 Sxgs The defence planned by Black on his 38rd move — 42...Hal 43 ¥yc8+ Hed (43..Hxg5 44 Hedt wrest 45 dxe4 Exfl+ 46 @g2 leads to the loss of the bishop) 44 @\g4+ @d5 would 1 have lost to the problem-like 45 c4+! 43 We7! (now the black king is unable to escape from the firing line) 43..0e3 Tf 43...49xf6 White wins the queen, or gives mate: 44 Hgl+ @f5 45 Wed. 44 edt HhS 45 wh7+ Ah6 46 wa7 In order to answer 46...4.25 with 47 Wed+ @h6 48 h4 Ge7 49 Bgl, when g6 cannot be defended. 46.44 47 f6+ Sg5 48 Ads Black resigns. Gurieli-Wu Ming Qien Women's Interzonal Zheleznovodsk 1985 Both kings are rather unfavourably, if not badly placed. True, neither can be approached directly, and therefore at first sight Black's next move does not appear particularly good, 23.0527 Certainly, the h3 pawn is now at- tacked, but the important d6 square is also weakened, and it is immediately occupied by White. 24 edb112 Threatening the deadly 25 (ic6+. If 24...4d7 White has the extremely unpleasant 25 45, while after 24...e4 25 Wyxf6 G&xh3 26 xe4 the bishop check at c6 is again on the agenda. But not only White is attacking, and the bishop at g2 is not only a striker, but also the central defender. 24.4! Black diverts the bishop from the defence of the h3 pawn, simultaneously opening the diagonal for the queen. White no longer has any attack, nor indeed any defence. 25 @gl Qha4 (25...h2+ would have won more quickly) 26 cS Axg2 (and here 26..Bxg2+ forces mate), and soon White resigned. Gurgenidze-Sergievsky USSR 1962 Strictly speaking, here too neither king feels altogether comfortable. One has been deprived of its pawn shelter, while the other is still in the centre. But whereas Black's king is not threat- ened by a single enemy piece, all his forces without exception are aimed Attack with Mikhail Tal directly or indirectly at the white king: there is pressure both along the g- an h-files, and along the long light-square. diagonal. Even so, at first sight it appears that White can hold on. As yet there is no immediate threat against h2, and e4, the key point of his position, is quit securely defended. But suddenly... 27...9.021 28 wyxd2 Forced to worry about his knight, White's queen is diverted from th defence of e4, and everything promptl collapses. 28...)xe4 29 We3 West! Paradoxically, this is stronger thai winning the queen after 29...4)g5+ 31 Sg3 Bh3+. 30 Wxg5 xg5+ 31 &g3 Hh3I+ 3: Sed 213+ 33 Yxgs BhS mate. Tal-Dvoretsky 42nd USSR Ch., Leningrad 1972 King's Indian Defence 1 d4 St6 2 c4 26 3 He3 ae? 4 e4 d6 5 Mer 0-0 6 |B eS 7 Sed 6 M.T. I cannot claim to have hat much experience of this variation, bu here either 7...4)c6 or 7...6)bd7 hai been played against me. For a time: 7..%¥e7 was also popular. As Dvoretsky admitted after the game, at literally the last minute he glanced i Boleslavsky’s book, and discoveret that the author recommends the modest pawn move 7...c6. For the momentEliminating Defenders Black refrains from developing his queen's knight. After some thought, I could find nothing better than castling, but perhaps White should play 8 d5!? 8 0-0 exd4 9 Qxd4d The position after 9 Qxd4 He8 10 13 d5 13 cxd5 cxdS 12 b5 hd7 or 10 s£f3 4ybd7 does not promise White. any advantage, but the capture with the bishop is also fairly harmless, vee Bes 10 we2 ybd7 Possibly Black should not have been in hurry to make this move. The direct 10...ue7 11 fel cS came into consid- eration, when White has either to give up his bishop, or a pawn: 12 {¢e3 &)xe4 13 xed wrxed 14 Wd2, for which he gains some compensation. Yi Sadi = We7 12 &fel &e5 If 12...29¢5, which looks more pur- poseful, White had prepared 13 b4, ‘when it transpires that the threat to the e4 pawn is illusory: 13...Q)cxed 14 Ad3 d5 15 cxd5 exd5 16 &xf6! Black could have continued 13,..e6 14 @¢3 4)g4 with a complicated game, After lhe move played, White at least no longer has to worry about his e4 pawn. 13° 3 he An interesting move, essentially for- cing White to go in for complications, which, however, turn out to be in his favour. After other moves he would have had time for the regrouping {3 and ¥d2, with positional pressure. 14 ba! ‘The threat of 14...c5 was rather un- pleasant, M4... b6 13 15 cS! As we have already indicated, this is forced, but... also good. 15 we bxeS M.T. After the game the dejected Dvoretsky said that he had simply forgotten about the transposition of moves that occurred in the game, For my part, I thought for a long time before making my next move. 16 2)xe5 M.T. Since I was quite happy about the main variation (after 15...2)xf3+ 16 &xf3) 16...bxc5 17 bxeS dxcS 18 Sie3 xe3 19 Kxe3: for the pawn White has more than sufficient com- pensation, I hesitated, wondering whe- ther I shouldn't be content with this. But then I decided that White could gain a more appreciable advantage. 16... dxeS Better chances were probably of- fered by 16...cxd4 17 2yxc6 wb7. 17 &xcS)Wb7 White has a clear positional advan- tage, but this is only the first stage. 18 a4 Aiming at the blockading square ¢5.id 18... Reo 18... 8f8 was more tenacious, 19 Ad6 ad7 20 Acs &yxcS, 21° bxeS! The point is that now a transfor- mation of advantages has taken place: instead of pressure on the weak c6 pawn and occupation of c5, White has acquired another, significantly more important advantage -—- undisputed (as long as the bishop at d6 is “alive") possession of the b-file. 21... afs 21...%¥b4 was perhaps more active, in order to answer 22 Bbi with 22... ¥ya3. True, White bas 22 a6, forcing the black queen to guard the b7 square. M.T. The move in the game sets a clever trap. At first I began examining interesting variations such as 22 &xe5 fxa2 23 fal wb3 24 Wd2 AxcS (bad is 24..Hxe4d 25 Qf3 Exel+ 26 Hxel &g7 27 Axg7 xg7 28 Bat!) 25 we5. The variations appeared con- vincing enough, but, after some hesita- tion, I decided against opening the floodgates for the black pieces, And I was right: on 22 A.xe5 Black had pre- pared 22...4.b3!, not only securing opposite-colour bishops, but also re- gaining his pawn! 22 Rb1 White sticks to his overafl plan. 22 vee wd7 23° Bedi For the moment all White's pieces are directed against the queenside. 230 wee Axd6 24 cxd6t Another transformation, The point of White's play is his quiet 27th move. Attack with Mikhail Tal wma... Babs 25 ad A new trump comes into play. Th advance of this pawn to a6 followed b: the invasion at b7 is bound to decid the game. 25 we Bxbl 26 wxbl = wyds It only remains for Black to pla 27...¥¥b6, and his position will be ten. able, but... decisive 28 Bb1. The black queen i obliged to revert to its cheerless o cupation of blockading the d6 pawn. 27 oe wd7 28 wycS £6 Clearly, 28...4.b3 29 Bbl Axa4 3 Bal is bad for Black. 29 a5 S27 30° Hbi Has 31 a6 Behe Now White could have played 3: HBb7 ¥yxd6 33 Yyxa7, when the a-pawH: must decide the game, but it was timi to think also about the black king! 32 We3+ More accurate than 32 Bb7 tyxdt 33 we3+ g5 34 Wf3 2.47, when all th same White is "forced" to capture a7 pawn. 32 oe. 25 33° Wf wi 34° Bb7 Bd7 (see diagram next page) 35 Act By diverting the defender of the square, White wins by force. 35... tS Clever, but insufficient. The accep-. tance of the sacrifice leads to mat35,..dixc4 36 Bxd7, and if 36...2yxd7 37 Wxf6+. Black also loses after 35... e7 36 Axe6 wyxe6 37 WS. But now misfortune strikes from the other side. 36 exfS The c6 pawn is still weak! On ..$ixcd there follows 37 ¥yxc6! 36 oe wyxid, 37° &xe6 —_—_ Black resigns WHAT WOULD YOU HAVE PLAYED? ‘No.23 Eliminating Defenders 115 White, to move, is one of the kings of chess. No,24 Again the player with White is a former World Champion, No.25 There is no way to defend the key c2 pawn, whereas Black's bishop and knight are covering the approaches to his king. Does this mean that the scales have tipped in his favour? Even if the turn to move is with his opponent, also a former World Champion?8 At the Royal Court IF IT IS ACCEPTED AS AN AXIOM that the white and black troops are led by their kings, the royal court must be taken as the two extreme ranks at either end of the board, the Ist and 2nd, and the 7th and 8th. As in life, other pieces aim to come here for an audience, but in contrast to life — with evil intentions. After all, an invasion of the opponent's residence is both an aim, and a highly effective means of attack, and by the very nature of chess the pieces most suitable for this are the “straightforward-moving" rook and queen. Their "thoughts" are usually also straightforward: either to give mate on the back rank (if the king has no escape square, or if it is insuffi- ciently open), or to drive the king via the escape square out of his residence into the open field, where it will come under the fire of the other pieces, For the moment we are tatking only about the very back rank, of which a detailed discussion invariably takes place in any iesson for beginners, and which is present in every chess primer. Nevertheless, variations on this theme are encountered from time to time even by the most famous of grandmasters... (see diagram next column) If World Champion Garry Kasparov were asked to find a win for White in this position, he would be unlikely to require more than five seconds of thought. ys Bona # 2 Kasparoy-Ribli World Cup, Skelleftea 1989 Yet he himself offered a draw, wh after the obligatory 26 Hxb5 (xe3 could have played the very pretty anc not especially complicated 27 Bd8 WxbS (27...Axd8 28 Hd5!) 28 wade Axf2+ (28...He8 29 We? Axf2+ 3 He2! wyo6+ 31 HF] whi+ 32 Hxf2 Wxh2+ 33 Hel tyxg3+ 34 Sd Wel+ 36 Hd2, and the checks come an end) 29 xf2 wWf5+ (or 29...Be8 3 ad! Wyxad 32 wWxa6!) 30 Hg2 wed: 31 Bh3 wWis+ 32 et Wii+ 33 & Welt 34 HF3 Wil+ 35 He3 Wh3+ 3 dd! (36 d22? who+ 37 Hdl gt 36...e5+ 37 Bd5 tye2+ 38 HeS wel 39 He6 Welt 40 Sb7 wbi+ 41 Ba’ Welt 42 Sxa6, and Black can resign. M.T. Kasparov overlooked this op-. portunity, but in the following game I. did not miss the chance to allow a mate in two moves.At the Royal Court F.Olafsson-Tal Alekhine Memorial, Moscow 1971 Here one might well ask the ques- tion — "What would you nor have played?", and the overwhelming ma- jority of players would give the correct answer. But... 21...e8?? 22 ¥yxe8+ and, greatly prised at what had happened, Black resigned. LD. In Kasparov's case it was all more or less understandable. The World Champion was "lulled" by the feisurely course of the game, without ‘any particular tension, with exchanges, and, inwardly reconciled to a draw, at the last moment he simply did make the effort once more to check how well the 8th rank was defended. But what happened in your case? What was the mechanism of your blunder? M.T. It was also purely psychologi- cal, as, evidently, is invariably the se, We know so much about trage- dies on the back rank, that we simply forget about them, That is, these threats exist somewhere in theory, set 17 apart, as it were, from our positions on the board. For me this game as though united them, if not for ever, then for a long time. And for probably ten years, in each game that I played, when the type of middlegame had been deter- mined, I would spend a second, in order at almost every step to inspect my back rank. This would seem to have helped, and no more surprises of this type occurred. LD. But grandmaster Samuel Reshevsky was less lucky. It is well known that three times(!}, at intervals of exactly eleven years, he overlooked stalemating combinations by his oppo- nents. But things were also no better with the theme of our present conver- sation. Here are just two examples; there were also others. Unzicker-Resheysky Munich Olympiad 1958 For the moment White's extra pawn does not mean anything, since the f4 and a2 pawns are both "hanging", and, more important, a rook check at el is threatened,18 White is saved by the eternal theme of the back rank. 27 We2l! wWe8 (after the exchange of queens White would have increased his material advantage) 28 Bc?! wd8 29 wed The black queen has been driven back, and the game is decided, 29..d5 30 Qxd5 DHd2 31 Web HES 32 Hxf? Hxf7 33 96 bxg6 34 fxe6 SiS 35 gxf7? Det (mate in three moves was threatened: 36 WeS+ tye7 37 We8+) 36 We8+ Black resigns. d muates “Ok ao its Pn Reshevsky-Fischer Palma de Mallorca Interzonat 1970 White's attack on £7 has reached its peak and would seem bound to be crowned by success, since after 29...4yd4+ 30 @hi1 Black's rook is attacked, and if 30...Be? he loses an important pawn — 31 tyxd6. But is this obligatory? 30...9f21, and White resigned (31 Hg] Hel), because in the heat of the battle he completely forgot about the weakness of his back rank. But things are not limited to these Attack with Mikhail Tal typical cases. In attacks on the back rank the heavy pieces are often helped by minor pieces, and sometimes the. invasion of the back rank is merely a subsidiary factor, strengthening other attacking motifs. Rossolimo-Livingston New York 1961 The black king is in a mating net, woven in particular by the white bish- ops, and White's entire subsequen' play pursues just one aim: that of i vading the 8th rank, This is not easy t0 achieve. 18 c4i! Wxg5+ 19 Ag3 Bxdi 20 Baxdi 505 20...47a5 is less tenacious: 21 b4 Wyb6 22 c5 Wyxa6 23 Hd8+, mating. 21 exd5 c5 22 b4 04 23 Bd All the same, Black is unable to. keep the files closed. 23...e5 24 dxe6 4.xb4 25 Hd7 Abs Defending against 26 4b? mate. 26 Bfd] White renews this threat: 27 4.b7- Hixb7 28 Hd8+, and mates. 26...£.€7 27 ext? c3At the Royal Court On 27...89f6 White would have won by 28 Axb8 @xb8 29 Hbl+. 28 f8=y = The diversion of the black rook or bishop leads to mate, Black resigns, N.N.-Rossolimo Paris 1944 In this game of the future grandmas- r, the "sights" of Black's bishop and took at £8 intersect at £2, which for the moment is defended. For the moment! LB! Exploiting an additional nuance of ne position — the overloading of hite's queen's rook, Black creates irresistible threats both along the first, ad the second (f2) ranks. White could ready resign. 2 Qxb7+ Hb8 3 ed Hxf2! This is much quicker than 3.,.Q.xf2+ 4 Wxf2 Bxf2 5 Bfxdl Wye5 6 d5+ Bc8 7 Ph2, although here too matters -end in a mating attack — 7...Bf6, 4 WxbS Hfxfl+ 5 @h2 Bhi mate < In this last example we had a glimpse of the diversion theme (of the queen's rook from the defence of the 119 queen), which in principle is the main device in the attack on the back rank. Classic examples of this theme (Adams-Torre, New Orleans 1920, and Bernstein-Capablanca, Moscow 1914) are weil known, and so we will restrict ourselves to three others, that create an unusually strong emotional impression. Levitsky-Marshall Breslau 1912 Despite being a piece down, White is clearly intending to take the initia- tive: after placing his rook on the 7th rank next move, he will create mating threats, and then, after the black knight moves, he will capture the rook on h3, True, Black already has a drawing mechanism in place: 23...ge24+ 24 Qh 4)g3+ 25 gl He2+. But he found what was perhaps the most beautiful move in the history of chess, emphasising the vertical and diagonal harmony of all the black pieces, attacking the king. 23..¥¥g3!! White resigns Mate at h2 is threatened, and the120 black queen is immune: 24 hxg3 4e2 mate, 24 fxg3 He2+ 25 Shi Axfl mate, 24 Wyxg3 He2+ 25 Phl Hxg3+ 26 @gi Axfl or 26..He2+, and Black remains a piece up. This has something in common with the next game, where the blow struck by Black was equally spectacular, and as a manifestation of diagonal-vertical interaction was perhaps even more difficult to find. Mikenas-Bronstein 33rd USSR Ch., Tallinn 1965 White appears to stand not at all badly. He has the better pawn forma- tion on the queenside, a passed pawn in the centre, and an active rook on b4, white the possible invasion by Black of his back rank — 24...Wel+ is parried by the simple 25 Wfl. But his shrewd opponent had seen that the black queen on the long dark- square diagonal was aiming not only at the b2 pawn, which is securely de- fended, but also — “X-ray" fashion - at the unprotected rook on al. This means that it would be good to Attack with Mikhail Tal divert the b2 pawn, and if not it, th the queen from the defence of f1, and not it, then the rook from the bac! rank... And this was how Black co: ceived an amazing move, rivalling th concluding stroke in the previous e: ample. 24...Bxa3tt This attack on a three-times(!!) de fended pawn, noticed moreover b: “peripheral vision" on the edge of the board, led to White's capitulation. The two preceding examples ai chess classics, well and widely know But here is one from more recent times. Kengis-Gufeld Moscow 1983 All White's pieces have invaded th opponent's position, the price being thé right for Black to have an extr (second!) queen. But after the imme diate 45...d1=ty 46 Bxf8+ wyxf8 4 We6+ Ph7 48 Axf8+ Ph8, firstly, th number of queens becomes equal, an secondly (and most important) Whit promptly makes a new invasion — 4‘At the Royal Court 121 Wye8!, with the threat of a “discovered Black therefore decided to play the diverting 45...Hxe5, hoping after 46 ExeS d= 47 Bc8 td6 to advance. his a-pawn and thereby divert the white rook or obtain a queen ending with an extra pawn. But here the strength of a rook in the opposing king's residence is fully revealed. 46 tyes! That's it! Black has no choice — 46...49xe6 allows mate in two, and he could have spared himself the last few moves: 47 S\e7+ @h7 48 ¥yxf7 Ae6 49 &xc8 dl=ty 50 4e7 wd8 51 Web. But even if there is no direct mate, pieces invading the 8th rank can attack the opponent's king from the rear, and land unexpected blows in places that are inaccessible from "in front". Geller-Tukmakoy 50th USSR Ch., Moscow 1983 The game had a curious finish: 35 Gh8+ He6 36 yg3+ hs 37 wes mate! This only became possible because the white rook was pinning the h6 pawn. In the following game the then World Champion overlooked a similar possibility. Al 7k YEU) aia Karpov-Hiibner Montreal 1979 After driving the king onto a light square — 39 898+ @h7, White could have vacated the g8 square with gain of tempo for a "discovered check" to the queen and won a pawn — 40 Hixg6 (he could also have won by 40 we3, and if 40...4.d8 41 Bh8+!). Instead he switched his rook to the 7th rank by 39 wed HI6 40 He7, which in this case proved insufficient for a win. But usually, even a lone rook that has broken into the royal residence is a terrible force. The opposing pawns are initially arranged along the 7th rank, and in addition the enemy king is in the immediate vicinity. But when two pieces penetrate, and there is also sup- port at hand...122 Ditman-Balanel Erfurt 1955 If he picks up two pawns, White risks losing the game: 29 Wxh6 wh8 30 Wxg5 (30 Hh7 ¥f6, and White has nothing better than to repeat moves, e.g. 31 Bf2 ¥xc3!) 30...Be5. But after all, he commands the cherished 7th rank! 29 Wd7! |yb6 30 Hf8+! and mate next move. A typical idea: one attack- ing piece makes way for another, that is capable of more. It is interesting that 29..Wh8 30 8h7 Wf6 would have led to a mirror variation — 31 Bh8+! In the next diagram Black's rook has not only already penetrated into the ‘opponent's position, but from the 2nd rank it is also ready to land a blow on the Ist rank — with the help of the other pieces, of course. 41... f3+! 42 Of A "step to the right" — 42 @hl looks more dangerous, and this im- pression is confirmed by fantastically beautiful variations such as 42... Attack with Mikhail Tal Sixe3!! 43 fxe3 Hdxg2!! 44 exe! Gg3, mating, or 43 De6 Axf2 44 ex! Bfl+ 45 @h2 exf3! 46 Axg5 gfe mate. Sunye-Kasparov World Junior Team Ch., Graz 1981 However, the "step to the left" also. leaves the white king under fire. 42..4xe3! Joining the sights of both rooks the g2 pawn. 43 fxe3 Naturally, 43 ¥xe3?? would hav allowed mate on the Ist rank. 43...Edxg2! A rare instance, when the secon rook does not aim for the 2nd rank. But here Black needs to create the threat of 44...\d2+. 44 Wc3 Hh2 45 4)e2 Ph7! Parrying the threat of perpetu: check from c8 and f5, Black plans thi "normal" invasion ...Bigg2. 46 Wyc8?! 46 Wyb4 was better, but even th Black would have won after 46...f5 4! WS fa! 48 eyb4 Ad2+! 49 wxd2 (49 Sel £3 50 Sxd2 Bxe2+ 51 Hc3 Hixe3+ 52 gd4 £2) 49...Bh1+ 50 @f2 13! 46...h1+ 47 G2 d2! And White resigned in view of the forced 48 4g3 Bh2+ 49 Bel |Hf3+ 50 ft Hxb2, when further resistance is pointless, Spassky-Tal 25th USSR Ch., Riga 1958 Nimzo-Indian Defence 1 d4 a6 2 4 e6 3 3 Ab4 4 a3 Axc3+ 5 bxe3 cS 6 e3 26 7 Ads @5 Even at that time Black knew per- fectly well that the usual 7...0-0 8 4ye2 b6 9 ed He8 leads to a complicated ‘game, pleasant enough for him. But what if it was in this variation that White had prepared something? And ~~ Black deviates! LD. One of the favourite sayings of y co-author has always been: “When tram-driver seeks new paths, the tram goes off the rails"... M.T. But not this time! 8 her On 8 d5 Black had "devised" 8...e4, ‘when he stands well after either 9 @e2 Se5, or 9 dxc6 exd3 10 cxd7+ wxd7. 8 os ed 9 gpl b6 10 5g3 has ll £3 ~ The only real alternative was 11 Gyxe4 Axed 12 Mxed Axc4 13 13. At the Royal Court 123 The line recommended by Keres after the game, 11 #ya4 a5 12 4)xe4, is less convincing, since Black quickly restores material equality by 12... &xe4 13 Qxe4 Hc8, with good play. MW... fixed “Pawn-grabbing" is punished: 11... exf3 12 ¥yxf3 gxc4 13 4f5 0-0 14 e4, and White has a powerful in- itiative. 12 5! In the variation 12 fxe4 d6 13 ¥yf3 0-0 14 eS dxeS 15 ¥xc6 exd4 it is Black who has a strong attack on the enemy king stranded in the centre. ... 0-0 There is nothing else, but nothing else is required. 13° d6 a3 14. Qxd3—exd3 15 ¥yxd3—exd4 16 cxd4 4est 7 £5 d5 White's pawn centre has essentially been blocked, and Black can be happy with the result of his opening experi- ment. 18 ad 19 |xd6 Of course, not 19 a3 4xf5 20 Axf8? We5, with numerous threats. db wo... wWxd6 20 4a3 8 8=b4 21 wb3 a5 22 «(0-0 Hfcs8 23° Hackl we6 Here Black offered a draw, having worked out the “long” variation 24 fixb4 axb4 25 @f2 wWdG 26 @e_l web. 24 Qxb4— axb4 28 @f2 wd6= tt t at 2s . mia Max We a ae ae Be ot ss 26 3 White, naturally, avoids the objec- tively strongest continuation, since the tournament situation demanded that he play only for a win, but now after 26...n5, with the idea of ...n4, the initiative would have passed to Black, M.T. But he played a routine move! 26... PP? 27 Heat Hixe2+ 28 Wxe2 = -g6 29 Hel wd7 30 Yyc6 Wxc6 31 Bxc6 a6 More accurate, of course, was 31...Hxa4 32 Hxb6 Be7 with a quick draw. But now the game reverts... to the middlegame. 3 ad b3 Here 32...8xa5 33 Bxb6 would have left White a pawn up. 33° axb6 bz 34 b7 bl=u 35 He8+ = &g7 36 b8=y The new queens have appeared on the same file, and in addition all the pieces are in the opponent's rearguard. Attack with Mikhail Tal An unusual state of affairs! 36 ee Hate 37 &g3 38 @h2 39 Hg8+ = hfe The draw that Black needed woul have been more simply achieved b 39..@h6 40 Wf8+ BhS 41 wx Eixg2+!, forcing perpetual check. 40 wd6+ web 41 Wh+ wi 42 wWd6+ web 43 We3 The position has radically changed: The white king is now in safeti whereas Black’s is completely expo: to the elements. 43... we 44 ha Bez! ‘The tempting 44..6al 45 wd Web 46 tyi4+ WES 47 wWhé is weaker, since Black's pieces lose their coordi nation, and his king is in jeopardy. 45 Wd6+ wWe6 46 Wide | ys 47 tyh6 With the same idea. a7... Ber 48 Wyh8+ Bio 49 We7+ Be7 50 Had! The threat of a check at a7 forces the black pieces to retreat. 50... wd7 Sl wf8+ Be 52 Ha6+ He6 53° Wh8+ = e7 S4 Has Het 55 &g3 hs 56 BZ He6 (see diagram next page)57 cB! Now Black can only move his rook, ince his queen is tied to the defence of dS and e8. In the event of 57...%4d6 White wins by 58 He8+ and 59 Hd8+, and pawn moves are impossible. And since 57.,.2c6 fails to 58 wf8+ Wf6 59 Hd8 wc7 60 wh8+ Be7 61 e8+ d7 62 HeS, when White car- ‘ties out a favourable regrouping, Black's reply is forced, 57 oes 58 (Wyf8+? An imperceptibic, but significant mistake, which in the heat of the mo- ment went unnoticed by both players. It is not possible to occupy, or more precisely, to take control of all the squares around the black king with the queen and rook alone, and, as Chekh- over later showed, White should have mobilised his pawns with 58 g4! Then 58...He6 loses to 59 gS (taking away the £6 square) 59..,8c6 60 ¥yf8+ Geb Gl He&+ SF5 62 He7 Hc2+ 63 gd Wd6+ 64 He5+. The alternative is 58... hxg4 59 Wf8+ @f6 60 fxg4. Now on 60...He6 there follows 61 8c3 Be4 62 Ad6 Al the Royal Court 125 83+ Se6 63 g5, with an unavoidable invasion at f6, while if 60.6 61 He8 Hc2+ 62 G13 He3+ 63 Of4 We7+ 64 Be5!, with the decisive threat of g4-25 mate. After the move played Black breathed a certain sigh of relief. 5B St6 59 Hes Here 59 g4 is no longer so effective on account of 59...Be6, and White prefers to threaten mate in two moves. 59... Beé 60 Wh8+ = Of5 61 whe Either capture of the rook again leads to mate, and White continues the attack on the dark squares, although it is no longer dangerous. 61... ete 62 Wh8+ Winning a pawn by 62 Ad8 We6 63 Wyg54+ Ge7 64 eyxd5 (64 Bxd5 £6) would have allowed Black to seize the initiative after 64..We2+ 65 @g3 ¥e7+ 66 f2 eyf4. 62... wets 63 a8126 White should now have admitted that he had missed the win, and agreed to a draw, but in the heat of the battle how difficult it can be to soberly assess achange in the chess climate! 63s Wyeé! Threatening for a start to check at c2 and c7. 64 Hc8? White should have concerned him- self over his own king — 64 #g3, when he is still within the safety zone. 64... Wao 65 &g3 wyd6+ 66 Bhs No better is 66 @f2 wb4, when it is the black pieces that "invite them- selves" to a reception with the white king — with evil intentions. Had White left his rook at d8, this would not have been possible on account of the capture on d5, forcing mate. 66... Hel 67 Or 67 g4+ hxg4+ 68 fxe4+ G4 69 Wh6+ Wed! 70 He8+ F3, and the white king is mated. 67... Bgl 6s f4 Bel Black has excluded the counter g3- 4+, which was possible in certain cases, and has taken complete com- mand. He now threatens 69...ue6 with a discovered check and a quick mate, The only way to defend against this was 69 He8, but then after 69..Bxe8 70 tyxe8 web the d4 pawn is soon lost — and with it the game. Rather than suffering without end, White prefers an end without suffering. 69 He2 Web 70 Bf2 Bhl+ Attack with Mikhail Tal 71. Sg2 wed+ 72 Bf Bed 73° We8+ £5 White resigns: after 74 ¥c3 Blac has the simple, but elegant 74...Bf1 and in the pawn ending he can, if wishes, have four extra pawns, WHAT WOULD YOU HAVE PLAYED? White is on the attack, and he woul appear to have achieved a great de: He has broken up the black king pawn shelter, and after the fall of the £5 pawn he will begin a direct assaul on h7 and £7, True, it is Black to move, but hi counterplay with 22...0¢4+ 23 ®I S)f2+ 24 Bxf2 Qxf2 does not work account of 25 4)xe8.What then shoul he do?At the Royal Court 127 No.28 For the pawn White has gained more than sufficient compensation. He as seized the open file in the centre, he black knights have been lured to the edge of the board, and finally, there is the possibility of an attack on g7 (24 ¢5). But from its ideal position the ‘central defender* —~ the black queen is capable of withstanding the ssault... For a long time Black's position in (0.28 was lost. But what is it now? The black pieces have, after all, Ntered the white king's residence... It is true that the white rooks have occupied the 7th rank, on whieh, inci- dentally, there is not a single pawn. But what next?9 Destroying the Fortress Walls AN ENGLISHMAN'S HOME is his castle. This saying is also encountered in various versions among Russians, Germans, Spaniards, and, of course... chess kings. The pawn rank f7/g7/h7 (kingside castling does after all occur much more frequently than queenside, at least for the defending side) is that wall behind which the king seeks shelter. And for a time he finds it until -— or if —- an attack achieves its goal: the destruction of the fortress walls. Moreover, the pawns are most sol- idly placed when they are on their initial squares. Firstly, in this case the attacker must spend much more time ‘on the concentration of force (we will see how a rapid attack is possible from afar against a pawn that has advanced to h6), and secondly, the advance at the required moment of the required pawn is capable to cutting a trajectory of the attacking pieces, But as is known, a pawn is not able to reverse its move, and a step forward may irreparably reduce the further defen- sive resources, as occurred, for exam- ple, in the following game. Black already has his sights set on the enemy king position, his rook is on course for hS, and the simplifying attempt 15 ixg4 Sxe4 16 h3 is fraught with danger: 16...4)xf2! 17 @xf2 wyh4+ with a powerful attack. White could still have maintained ap- proximate equality by the calm 15 Sel, defending f2 and now planning the exchange of bishops. Instead, an incautious advance presented Bla with a clear target. Gilfer-Book Stockholm Olympiad 1937 Now the king's escape is cut off the centre, where it would be cover by the e- and f-pawns. The only way of prolonging resistance (although it would aot ha’ changed the outcome) was by 17 Wxh3 18 fxeS {LxeS 19 Qf3 (other wise mate) 19...v4xf3 20 Mel Ded Hd2, when a possible finish would 21..4)xe3 22 BR Wet 23 Ohl wWh3+ 24 Bgl Qh2+ 25 Bhi O34: 26 Sg) Qxf2+ 27 x2 Be8, and wins. 17..g5+ 18 @hi we6+ 19 4 4yxed 20 Dyxed eyxed+ 21 13 wed 22. f4 Wed+ 23 @h2 Oxf4+ 24 Axia wyxf4+ 25 @hi wyf2 White resigns.Destroying the Fortress Wails Castaldi-Euwe Stockholm Olympiad 1937 In this game from the same ‘Olympiad, for the exchange White has powerful attacking grouping against king, whose pawn shelter is not upported by picces. Now 32 Bd6 followed by @h6 is threatened, and tempts to avert this by the advance of any of the three black pawns merely make things worse. E.g.: 31...h6? 32 Qxh6, or 31,..f6? 2 Gh6 g6 33 Ae7+ He8 34 Hh6+, 1, finally, 31...26 32 4)h6 @e7 (32... Yyxg42? 33 Af6 mate} 33 ¥yxe6 fxe6 4 Hd7+ G8 35 g3!, with the ir- resistible threat of 44, mating. Only in this last variation does Black retain draw after 32...8f8 33 ¥xe6 fxe6 34 8d8 Yg7 35 4d74, although here too White has other possibilities, such as 33 Wf3, with the idea on 33...f5, 33... £6 or 33,..8727 of replying 34 Ag4. Black did not disturb his pawns prematurely, after 31...ac8 32 Qh6 g6 (now is the right time!) 33 wWg5 fid4 34 Qo7+ dg8 35 A116 A207 36 129 h4 cS he retained hopes of a de- fence, and after mistakes by White he | even won, Therefore an assault on the royal fortress often begins with the discovery and creation in it of weak points, which then provide targets for a deci- sive blow. As, for example, in the following game. Salov-Sax World Cup, Skelleftea 1989 White's powerful piece outpost at d5 gives him an appreciable advantage, but even so it is hard to imagine that the walls of the royal fortress, which have no vulnerable points, will col- lapse within a few moves. After all, Black intends ...c5 and ...4e7, elim- inating the main enemy, after which his queen will reach the kingside via c6. 23 Yyf3it By the threat of 22 &)xe7+ White gains time to transfer his knight to £5, and — most important — to force the opponent into a fatal weakening of one of the links in his pawn chain. 21...8fe8130 Forced, since 21...Bae8 22 f5 leaves Black either the exchange down — 22..x8¥c8 23 AQdxe7+ QHxe7 24 446, or simply in a bad position — 22...89b8 23 Hdxe7+ Dxe7 24 Dds Bd8 25 He3. 22 HFS AEB Apparently satisfactory, but the {7 pawn is no longer covered by the rook, 23 |Hh6+! gxh6 24 Af6+ Ph 25 Hxd7 Now 25..Wc8 26 wf5 leads to mate, while the game continuation 25... Wxd7 26 2)xd7 Qd4 27 wWd5 Ae7 28 e3 Se6 merely allowed Black to proiong a hopeless struggle. Murey-Poldauf Podolsk 1991 By combining two opening schemes (Q'S and ...tya5-a4) Black has not equalised fully, although he has crea- ted some counterplay. Thus, in partic- ular, he threatens il..c4 12 Qe2 ¥yxc2, when on the kingside the black queen becomes the central defender, So White begins trying to break up the enemy king's pawn shelter. Attack with Mikhail Tat 11 WhS! h6 Forced, since in the event of 1 exd4 12 Ag5 OHxe5 13 |xh7+ Ex! 14 Wxh7 5g6 the advance of Whit h-pawe is decisive. And after 11... he mounts a swift attack on the dark squares, 12 dxeS Wa5 Simplification by 12.06 Axg6 fxgd 14 Wyxg6 weds 15 Wx dxe4 16 S\h4 Qf7 17 £4! leaves Whi too many pawns ahead — even with opposite-colour bishops. : 13 0-0 eyxc5 14 Afbl 4a8 (15 Hbs was threatened) 15 Dyg5 g6 16 Wh4 White has forced another weakening. advance. The g6 and h6 pawns have now become targets for attack, and there is an invasion point at f6. Is tl sufficient? 16...6)¢4 17 &)xf7! Quite sufficient! The shelter's | support is destroyed, after which # is no defence. 17...xf7 If 17...@)xd2, then 18 Dxh8 Ax! 19 xg6+ |xgs 20 wi6+ is me than adequate. 18 Wf6+ Sz 19 Axg6 Hxes No better is 19...8h7 20 Qxh7+ QBxh7 21 Wh7+ Bh8 22 wyf8+ Ph7 23 Axh6, when the pinned knight is able to defend g7. 20 QhS H7c6 Otherwise (for example, after 20... 4yd6) the white rook joins the attack along the 4th rank. 21 Sf4 BiB 22 Axed wyxf6 23 &xf6, and White won easily and quickly: 23...h7 24 Hel Hc7 25 HeF Qh7 26 £4 a7 27 £5! exfS 28 Bae! £4.29 He7+! 2)xe7 30 Hxe7+ S28 3Destroying the Fortress Walls Te7+ @£8 32 Bh7. The attack on the bare king continued even after the Tal-Rantanen Tallinn 1979 White's advantage comprises sever- | factors (powerful piece outpost in he centre, rook on the 7th rank, occu- ation of the open f-file, and, finally, uperior pawn formation and great iece activity), but is this not all tem- orary? For example, 24 Adfl (h5 25 a7 Hb7, and the pressure of the rook along the 7th rank has already been eutralised, and besides, after each xchange of heavy pieces Black's osition becomes increasingly safe. But White finds another way: he reates weaknesses around the enemy :king. For the moment g7 is securely defended, and there is no way of in- “creasing the pressure on it, while not ‘one of his pieces is aimed towards h7. 24 ESf6t! ext6 This forced (24...Qxd1 25 wyh4 h5S 26 Wg5 Bb7 27 tyg6, and mates) capture of the knight immediately 131 transforms h7 into a real weakness, and the attack proceeds against it. 25 wyh4 0.27 26 Ah6 Qxd1 The only way to resist was by 26... 8g8 27 Hxd6 Hb7 28 Axg7+ Bxp7 29 8xb7 Yxb7 30 Bd8+ Bgs 31 Axes+ Qxg8 32 vyxg44, although with an extra pawn and more secure king, White must win the queen ending. 27 Qxe7+ gb After the game it transpired that, strangely enough, both(!) players had aimed for this position. Alas... 28 Qh8! Gxi7 29 wWxt6+ Hes 30 We7 mate F.Olafsson-Donner Lugano 1970 While Black has been engaged in exchanging White's fianchettoed dark- square bishop, White has "launched" his knight to g5 and now embarks on an attack against the king's fortress. His first task is to weaken the pawn screen, to prepare breaches in the fortress wails. 13 £4 exf4 14 Axf4 he The threat of 15 Q¢4 forces Black132 to make this advance, but this is only the start. 15 AR ‘There was little point in playing 15 &ge4: at the required moment the de- fender could have begun simplifying. 15...We7 16 Bfl Res Black vacates a square for his queen's knight. This weakens his de- fence of f6, but how else is he to com- plete his development? This problem is not solved by 16...c5 17 44! b6 18 {5 Web 19 2£3 Hb 20 45, when he can essentially resign: 21 {124 fol- lowed by @xh6+ is threatened, and 20...4)xd5 allows the crushing 21 Sixd5 tye8 22 4xh6+ gxh6 23 vyg6+. 17 Aad DEB 18 Gd3 He6 19 Axio After this the walls of the king's fortress lie in ruins. 19...gxf6 20 De4 a7 21 wai £5 Otherwise 22 Wh5 and 23 Hf3 with a quick mate, 22 2yxf5 Axf5 23 xfs Now the attack continues with ma- terial practically equal. 23...8e6 24 Wf3 eS 25 wye3+ SB (25..Gh8 26 HS) 26 Afer 4\xd3 27 Qh7+ HeF 28 wye7+ wd7 29 Hxf7+ Qxf7 30 wxd7+ He7 31 WF5+ and Black resigned, since after 31...8g8 White wins most quickly by 32 wye6+. After playing the opening badly (see diagram next column), Black is al- ready obliged to parry the threat of 10 2)g5 26 IL hd. 9...Ae8 10 403 hé Insufficient is 10..4f8 11 4g5, when 11...h6 is bad in view of 12 &h7+. Black immediately has to grant the opponent a "chink" in his defences. Attack with Mikhail Tal Salov-Taulbut Moscow 1983 Hi hd 2yb6 12 Hh3 dxe4 13 fe section of the wall —~ 16 Rixg7! 15 Bxh6 g6 (15...gxh6 16 wd2) 1 gs wd7 17 After Se8 18 A Black resigns. Sacrifices on h7 Yes, Black's h7 pawn (or White's ‘at h2) is always slightly weak, even wit the ideal pawn formation, if, of cout it is defended only by the king. The attacking mechanism has been too well known for 150, if not 200 years, for u: to give a detailed discussion here, bishop sacrifice at h7 draws out king, the knight advances to g5 wit check, clearing the way for the queet (usually to h5) and a rook (usual along the 3rd rank}. It is probable th; thousands of such games throughout: the world have been played — and won!Destroying the Fortress Walls ~ so we have included here some that are complicated by certain additional motifs. After all, for a chess player, as for a musician, it is not enough to know only the rudiments, and to be able to perform only a single song... In the table of the Taxco Interzonal Tournament, 1985, the result of the game Speelman-Tal was quite differ- ent, But it ended rather too quickly in a draw, and the players sat down... to play blitz! It is easy to see that the framework of the combination is in place (13... &xh2+ and 14,.Qg4+), but will not the entire concept collapse? After ail, it appears that White has a defence... 13..Axh24! 14 Sxh2 Hgd+ 15 oes The only way to avoid inevitable de- feat was by 15 Qxgd dyxh4+ 16 Gh3 Axh3 17 gxh3 wyxf2+ 18 Bh) Axd4 19 cxd4 ¥yxd4... MT. ...but in this case I simply = would not have shown this game. Speelman, fortunately, played differ- ently, reckoning after the queen check at d6 to block with the £2 pawn, 133 15...He3+! Diversion of the defender, obstruc- tion — call it what you will: 16 fxe3 Wd6+ 17 @F3 AceS+ 18 Sg3 Ad3+ 19 @f3 Sh2 mate. MT. Quite an attractive concept? LD. In-struc-tive! This same mechanism can work even when the g5 invasion square is securely controlled, Veresov-Dzentelovsky Minsk 1956 Nominally Black has no weak- nesses, and his position wouid be ideal, but - with his knight at f6 (defending h7) and his rook at f8 (defending £7). ‘Without them White's thematic attack is perfectly correct, 16 Qxh7+! @xh7 17 Hest The £7 pawn is attacked, and the queen's invasion at h7 will be supported by the rook, whose path along the 3rd rank has been cleared. The further course of the game is of No interest, since Black made a blun- der. But even the comparatively best 17...@g8 18 Wh5 HFS 19 Hed 6 20134 )g6! (in the variation 20 4xd7 Wxd7 21 ¥yxa5 e5! Black has serious coun- terplay for the pawn, while for the reckless 20 Hh3 fxeS 21 Wh7+ @f7 22 Hig3 Qf6! 23 dxeS Qxe5! 24 wyhS+ eB 25 wyxeS wb there is no need) leaves White with a very strong attack, e.g.: 20...8)c4 21 Bpg3 Qc6 22 Ah6, or 20....e8 21 Dxe7+ Wyxe7 22 wyxaS Bf? 23 Qa3 Wyc7 24 Bhs, and he is simply a “healthy” pawn to the good. Tal-N.N. Simultaneous Display, Berlin 1974 With his last move 14 Qp5 White has removed Black's control of g5, and if 14...Qxp5 the sacrifice 15 .xh7+ works, so to speak, in pure form (e.g. 15..@xh7 16 Hxg5+ Sge6 17 SyxF7!). However... 14...2)xeS M.T. To be completely honest, I overlooked this capture. But in “justif- ication", this was perhaps because I intuitively sensed that you can't play like this in the opening, and I had to find a reason why this move was bad. And J found one. Attack with Mikhail Tat 15 Qxe7 Hxf3 16 Bxf3 tyxe7 1 Axh7+ @xh7 18 Bh3+ Ogs Now the routine 19 ¥yh5 does n achieve anything after the equally rou- tine 19...f6. But in this ancient idea proves possible to include a relativel new motif. 19 4f5 ¥yg5 20 WhS! Black re signs: he is mated after both 20...Wxh. 21 Se7+, and 20...f6 21 He7+. For the successful pursuit of a kin; it is by no means always essential destroy its pawn shelter. It can be quit sufficient to break up the formation, bj doubling the pawns on the f-file. Thi immediately gives rise to a whole cot plex of weak squares, on which the enemy pieces can approach the kin; and, more important, the pawn at hi (h2) is always practically indefensible. Popov-Malevinsky Sevastopol 1969 For a minimal payment Black hi managed to “ruffle the hair" of th white king, which has prudently tak shelter in the corner. Without losing. time, he continues the attack.Destroying the Fortress Walls 14...4)hS! 15 Bgl White has no time to take the bishop: 15 fxg4 wh4, and the h2 pawn s vulnerable (16 f4 yg3+). 18...eyh4 16 Hg? Qh3 17 di If 17 wf Black wins prettily by “ 17..6xe3! 18 fxed Ag3+! 19 hxg3 Sixg24+ 20 Gxg2 Wyxga+ and 21... Yyh2 mate. 17...Q,x82+ 18 @xg2 Wxh2+ 19 Whi+ 22 de2 A\f4+! White resigns. LD. For probably twenty years I have been pressing you to disclose that nechanism by which you seek — and, more important, find, your famous mbinational attacks. But you always point to your pose and say that you rcadily sense something attractive, and not necessarily on the chess board... M.T. I am also happy to repeat all his now! But to be serious, the follow- ing example is indeed very apt, and a Nikhamkin-Gruzdey Klaipeda 1984 That White has prepared an attack 135 — this is evident. The e-pawn has advanced a long way, his bishops are well employed, while Black's hanging pawns, although defended, are never- theless hanging on open files, and in this position the advance of either of them is aot a breakthrough, but a weakening, After the exchange of the knight at f6, the h7 pawn is supported by the king alone, and this may prove insufficient. So that of all the kingside pawns, only the one at g7 feels secure, As soon as I saw this position, two quite old games immediately flashed through my mind, Tal-Spassky Candidates Final, Tbilisi 1965 Here White's light-square bishop is operating with an "X-ray" beam along the pawn-free a2-g8 diagonal, the d5 square is the key one in Black’s de- fences, and it was in the hope of the "ricochet" 28 &\xe6 Bxc6 29 a8! that White made his 28th move. Black, however, saw through White's threat, replied 28...4.xc6, and managed to draw without difficulty.136 Tal v. Brinck-Claussen Havana Olympiad 1966 Here there is a similar pattern, and after 33 4xd5 Q.xd5 34 Ba8! Qxa2 35 Hxd8+ @f7 36 we6! Ae7 37 We8+ Heb 38 Hc8 White successfully concluded his pursuit of the king: 38...2¥d5 39 He3 Wad7 40 He3+ Hd 41 Wb8+ gc6 42 ad |dS 43 Hel Wd6 44 Bel+ Gd7 45 wWe8+ Black resigns, LD. Yet in our example the light- square bishop is not in the same place, and also... M.T. After the capture on f7 the di- agonal wil] be opened, the d5 pawn's defence can be removed, and in gen- eral one senses the "ricochet" theme. I.D. Again intuition, sensing, and all these hunting associations! M.T. Precisely! But with one nu- ance: the position is not difficult to calculate, knowing the idea of the attack, and if there is nothing concrete, then for the moment White stiil has strengthening moves — Had1 and, perhaps, 4)c3. Attack with Mikhail Tal A concrete and forcing variation was found, including the "h7 theme", 18 |bG!! Yxb6 Otherwise simply 19 2xc8 and 20: exf7+, winning the exchange. 19 Qxf6 Qxf6 20 xh7+! 4 The wail of another royal fortress collapses. 20... xh7 21 wyh5+ &g8 22 exf7+. Axf7 23 He8+ HIS 24 vyxd5+ Black resigns. Even the best mov 24...9.e6 does not save him after 25: Hixe6, e.g. 25...Had8 26 Bd6+ Bf7 Wyxf7+ Sxf7 28 Bxb6. Sacrifices on h7 and 7 Of the attacks on these squares, after which the king proves complete! “bare, pride of place goes to the dot ble-bishop sacrifice, which was di covered On the chess board by the then, World Champion in the following game. A rare instance, where the ti that an important tactical device w: conceived is known to the very day! Lasker-Bauer Amsterdam 1889Destroying the Fortress Walls The white knight has just advanced from g3 to h5, and White's subsequent blows are as inevitable as the rising of the sun, 14..Dxh5 There is nothing else. 14...Qe8 is decisively met by 15 Q.xg7 xg7 16 Wed, and 14...h6 by 15 Qxf6 Axf6 16 ®xt6+ gxf6 17 Wee Hh8 18 wyh4 W9g7 19 Bf3 Bfd8 20 Bg3+, winning he f6 and h6 pawns. If 14..d4 15 OAxf6 Qxf6 16 Wed e5 Black mns into 17 e4!, while 16...@h8 17 83 g8 allows 18 ixh7 Hgd8 19 wh3, 15 Qxh7+! "The start of a very deep and elegant mbination” (Steinitz). 15..xh7 16 WxhS+ @g8 17 xg7!! Lasker's contemporaries were very ‘well familiar with the sacrifice on h7, ut the second sacrifice on g7 was a he next example), but then 18 &f3 e8 19 Wh8+ Gf7 20 wh7. 18 Yygd+ Ph7 19 AL e5 26 Bh3+ ‘h6 21 Kixh6+ &xh6 22 Yd7 A finale, without which the combi- ation would be incorrect: a double at- ack on the two bishops. Here the cur- in could already have been lowered... 22.06 23 tyxb7 Sg7 24 Bil ‘ab8 25 Wd7 Hfd8 26 vyg4+ BLS 27 -{xe5 197 28 e6 Hb7 29 We6 £6 30 _ Hxf6+ O xf6 31 Wyxf6+ Be8 32 wyhs+ We7 33 w27+ Black resigns. Since then more than a hundred ears have passed, and although only a mall number of such combinational 137 routs have appeared in print, they have long become textbook examples. Moreover, the difference lies only in the preparation of the attack and the nuances in its implementation. Nimzowitsch-Tarrasch St Petersburg 1914 A brilliant tactician like Tarrasch could easily, had he wished, have found a way of shattering the peace of the white king — 19...Qxe2!, and after 20 &xg2 (20 dxc5 Wg5, and wins) 20...e¥g5+ 21 @f3 (21 @hi W4) 21... Aifes! 22 Hel Wide 23 Ber He2 White could have resigned with a clear conscience. But Tarrasch had no desire to conduct the attack in this way: he already knew of Lasker's combination! Hence: 19...Qxh2+ 20 @xh2 wh4+ 21 Bgl Axe? 22 3 The bishop is immune on account of a similar double attack to that of 25 years earlier, made at the end of the combination: 22 @xg2 wg4+ 23 Bhi d5 24 Yyxc5 Bh5+! 25 Wyxhd wyxhd+ 26 @g2 We5+ and 27...¥yxd2.138 22... fe8! Preparing ...8¥g3, which if played immediately would have been parried by 23 ed. 23 Sed yhi+ 24 G2 Axl 25 dd Naturally, White cannot take the bishop on account of the loss of his queen, but he can try to give mate on the a1-h8 diagonal... 25 G51 Here too Black should not be criti- cised! The choice was to win prosai- cally by 25..%¥g2+ 26 We3 (26 Bel Wyxf3) 26..89xc2 27 Hxe2 £5 28 Bi2 fxe4, or to go in for a forcing and pretty variation calculated to the end. 26 We3 we2+ 27 Ge3 Bxed+! 28 fxed f4+ One imagines that Tarrasch delib- erately avoided the "crude" mate in three moves: 28...¥yg3+ 29 Wd2 wf2+ 30 dl we2 mate. 29 Sxf4 BfS+ 30 wed wh2+ 31 Be6 He8+ 32 Gd7 AbS mate Also a pure mate, but here Black has no "superfluous" inactive pieces, Gaudin-Oskam Bromley 1920 Attack with Mikhail Tal 22 &xf6 Qxad 23 Axh7+ Bxh! 24 WyhS+ @e8 25 Axe7 4 xh2+ So as somehow to include the queen: in the defence. If immediately 25. @xe7 26 wWes+ Ph7 27 |F5, at Black has no adequate defence, e. 27.45 28 whSt+ Hg8 29 HeT+. Be 30 wrxed+ Wh 31 we3+ BI (intending to answer 32 e3 with 32. fic2!) 32 Wh3+ @_g7 33 wWed+ Bhs (33... h7 34 Be3) 34 whS+ He7 35 45+ Hg8 36 wWed+! and 37 He: winning the queen. 26 Yxh2 Hxg7 27 |LS+ Hi6 (0 27...9g8 28 Wes+ we6 29 De7+) 28: Wh6+ Sxf5 29 wyxb6 G.b5 30 He3 f6 31 Yye6+ Black resigns. Alekhine-Drewitt Portsmouth 1923 For the knight Black has three: pawns, but... There was Amsterd: 1889, St Petersburg 1914, and... 20 Gxh7+ "A similar sacrifice of two bishop: had already occurred earlier. Here i interest lies solely in the way in whi this combination was prepared," writes,Destroying the Fortress Walis Alekhine, and he goes on to assert that White's earlier pressure on the queen- side was merely a camouflage for the attack on the king. 20...Qxh7 21 Bh3+ Begs 22 .xg7 If he takes the bishop, Black is mated in three moves, while on 22...f6 White does not need to win the queen for rook, bishop and pawn (23 ¥yh5S Wyxg7 24 Hp3), since he has 23 {Lh6! Wh7 24 wh5! Of8 (or 24...e8 25 _ Hg3+ Bh8 26 27+ Bes 27 gxf6+) 25 Wed+ Bh8(F7) 26 AxfB. It is not clear whether Drewitt saw ail this, or if he took his opponent "at his word", but at any event Black resigned. mv Kuzmin-Sveshnikov 41st USSR Ch., Moscow 1973 In the next diagram, for the pawn White has a lead in development and a huge spatial advantage, but if he at- ~ tacks in the approved manner with 16 © Qxh7+ Gxh7 17 whd+ Ses 18 Lixg? xg? 19 tyed+ Bh7 20 BF3, Black's defences hold after 20,.2sxf4 21 Hxf4 £5. Therefore the combination begins with an overture! 139 16 b6! 2)xb6, and now that given above. In the final position, although three pieces down, White gives mate from h3. Sacrifices on g7 Here everything is more or less clear, if the attacking side succeeds in open- ing a file for his heavy artillery. A.Sokolov-Salov USSR 1983 At the cost of a piece White has op- ened the g-file and the a2-g8 diagonal for his attack, but the success of it de- pends on whether he can further break. up the black king's pawn shelter: other- wise one of White's strikers, his light- square bishop, will be exchanged. 18 Hxg7+! xg? 19 Hgl+ Dge (the only move) 20 exf5 (with the ter- rible threat of 21 wyh5) 20..8h8 21 Ad4+ § £6 22 xg6 fxg6 The defence is not eased by 22... Sixd4 23 gxf7+ Gf 24 we (threat- ening 25 eyg5 mate) 24..@e7 25 ¥yxd4, and although a rook up, Black cannot save the game,140 23 Wed Bh6 24 Qxf6+ Bh7 (24... @xf6 25 Wdd+) 25 Bel AxdS 26 &|xd5 wye8 27 He7+ Gg8 28 Ae7+ GPS 29 Hg8+ Sxg8 30 Ae7+, and Black resigned. LD. Misha, it would seem that we will be unable here to achieve an ideal arrangement of our topics. It should be said that the sacrifice on g7 is espe- cially effective when the pawn rank has already been disturbed, but the vulnerability of the h6 pawn is our next theme, M.T. Well, so what! Remember how they exhorted us in University: “Marxism is not a dogma, but a guide to action". Life has confirmed all this to be the exact opposite, but to chess this is indeed applicable. Hartston-Richardson Westergate 1983 In this position position, where, inci- dentally, White won the brilliancy prize, the centre is blocked by pawns, and the absence of his bishops means that his only chances of an attack lie on the open g-file. But Black has made Attack with Mikhail Tal preparations to parry it, by vacating g8: for his rook, and he could have faced the future with confidence, had not hi h-pawn made a step forward in the: opening. True, for the moment there nothing to attack it with, but it can be. weakened... 18 Exg7! xg7 19 Bgi+ Oh 2 Sye5+! hs The acceptance of this second sacri fice would merely have hastened the end. 21 Deb This does not come into the catego of quiet moves, but such an idea worth "memorising’. (Besides, th “natural” 21 &)xf7+2? simply lose: since it brings in the black rook to th defence of g7 and h7). Now Blac] could have resigned, but he steadfast! played on to the end. 21..4h7 22 wyxh6 Be8 23 8; Bxg7 24 Yyxe7 mate Kasparov-Nikolic Manila Olympiad 1992 ‘Nominally Black has achieved ev rything, or nearly everything. HeDestroying the Fortress Walis 14} developed and he has an ideal pawn formation, but for an instant, before he plays his bishop to f8, his g7 pawn is protected only by the king. And it is this instant that White grasps! 17 Qyxg! Oxg7 18 WS AB Otherwise the threat of 19 Hxd7 cannot be parried: 18...ad8 19 Hdfl 4)f8 20 We5+ Age 21 |Hh4, and the pin on the long dark-square diagonal is decisive, while 18...@8 loses to 19 © Begs h6 20 Bxd7 Wxd7 21 Hh7+, . 19hdt White unhurriedly prepares to take the g6 square away from the black knight, and in doing so not only to re- gain his piece, but also open the h-file. 19...h6 The only move. Just how dangerous Black's position is can be judged from these variations: (a) 19..g8 20 hS h6 21 Axf6 SLxf6 22 Wxf6 Heb 23 Yyc3 Hae8 24 Hd4, with the decisive threat of 25 Bed+; {b} 19...Had8 20 Hdfl @g8 21 Wyes+ Dg6 22 h5 Hed 23 hx Axgs (23...£%g6 24 Wxg6+ hxg6 25 Bh8+ SIT 26 He5 mate) 24 Ax~g5 {6 (24... fxg6 25 Hxh7) 25 gxh7+ @h8 (25... eg? 26 Deb6+) 26 Bxf6 Axf6 27 SLxf6+ tg? 28 DF7 mate. Very at- tractive! 20 94 Probably even stronger was 20 Ved 226 (20...0ph8 21 Sg5 Ads 22 Hxd8) 21 h5 eyb6 22 hxg6 Wyxe3+ 23 @bl fxg6 (23..xyxc3 24 wd, mating or winning the queen) 24 wyh4, with numerous irresistible threats (later Kasparov himself also thought so —1LD.). 20.878! 21 eyxc8 The attack could also have been continued with the queens on — 21 We2 @yg8 22 gS Agd 23 exh6 f5 24 Hhgl tye6 25 4d4 wyxe3+ 26 bl. But even after the exchange the black king has no peace: its residence has been destroyed. 21...Baxc8 22 g5 48h7 23 ed! Hcd8 24 Badfl @f8 25 gxf6 4 xf6 Or 25...€)xf6 26 He5!, when there is simply nothing that Black can move, and Eif5 then Ethfl is threatened. 26 e5 e727 Bhgl White also has an overwhelming position after 27 (.b4+ @g8 28 Hhgl S)f8 29 Hed Deb 30 Ad6. 27...c5 28 Be2 He6 29 Bed Ah8 30 bd (opening the b-file for a decisive attack from the flank) 30...b6 31 bxcS bxc5 32 bl Ba6 33 Ab2 O97 34 &b7! Axa2+ 35 Qb3 Hab 36 e6 Bxe6 37 Sixg7 Black resigns. LD. It is difficult to explain why, but for some reason sacrifices on g7 are very often of an intuitive nature, Perhaps the point is that there is no clear-cut attacking mechanism, such as Axh7+, DHes+ and wyhS+ (ee4), about which we have already talked... M.T. Well, there are some attacking procedures on g7 — the sacrifice followed by an attack along the al-h8 diagonal, for example, but in principle Lagree with you. Here it makes sense to rely on your intuition and... to be- lieve that boldness will win the day. Especially since, in contrast to the patented sacrifice on h7, here a precise calculation is not usually possible — at the board, of course. 1 will never forget, for example, my142 game with grandmaster Yevgeny Vas- yukov from the 32nd USSR Cham- pionship in Kiev, 1964/5. It was given in Chapter 3. There we reached a very complicated position, in which I was intending to sacrifice a knight. It was a not altogether obvious sacrifice, lead- ing to a mass of variations. 1 began trying to calculate them, and realised to my horror that nothing was coming, of this. My thoughts piled up one on top of another. A subtle reply by the opponent, suitable in one case, was suddenly transferred by me to another situation, and there, of course, it proved quite unsuitable. In general, in my mind there arose a completely chaotic accumulation of all kinds of moves, sometimes not even linked to one another, and the notorious “tree of variations", from which trainers recommend cutting off one branch at a time, grew with incredible speed. And suddenly, for some reason, I remembered the classic couplet by Koren Chukovsky: Oh, what a task so harsh To drag a hippo out of a marsh. I don't know by what association this hippopotamus climbed onto the chess board, but although the specta- tors were convinced that { was continu- ing to study the position, in fact I was at that time trying to understand how indeed you would extract a hippopota- mus from a marsh. [ remember that in my thoughts there figured jacks, levers, helicopters and even a rope-ladder. After lengthy reflection I couldn't find a single way of dragging it out of the quagmire, and I maliciously thought: "Well, let it drown!" And suddenly the Attack with Mikhail Tal hippopotamus disappeared. Just as had come onto the chess board, so went away. Of its own accord! An suddenly the position did not seem si complicated. Somehow I immediatel: realised that it was impossible t calculate all the variations, and that knight sacrifice was purely intuitive And since it promised an interest game, of course, I did not restrai myself. And the following day I read wi great pleasure in a paper that Mikhai Tal, after careful thought, made accurately calculated piece sacrifice. The winner of the following gat also admitted that his queen sacrific was absolutely intuitive. Suetin-Bagiroy Leningrad 1963 It appears that White's piece attack on the kingside will be exhausted with. the exchange of dark-square bisho, — 18 Qxd6 Bxd6, and 19 wyxd6 fails to 19...0xg2+. The attempt 1 Hxf6 QxeS 19 Bxc6 is also clear! insufficient, since Black interposes.19..44xb2! After weighing up these straightforward variations and trying in vain to work out an alternative, White took an intuitive risk. 18 Wxg7H!? Oxg7 19 A xf6+ Hho Much later, in a "collective" — but by no means joint! — grandmaster analysis, doubts were cast on the cor- rectness of the sacrifice by 19...@g6, when the tempting 20 443 is parried by the almost paradoxical 20...{9¢7! 21 Qxe7 Hxd3 22 BgS+ @h6 23 cxd3 He8! 24 9 f6 Be6! M.T. I too analysed this position, and after 20 Baf! ¥ye3 21 Qd3 Oho! 1 iggested 22 Ad] wyd2 23 H5f2, but then 23...2yxf2! 24 Axf2 Hde8 25 c3 £5! favours Black, L.D. During the game White consid- ered 22 Qxd8 Bxd8 23 Bxf7? Bd7 24 4)d5, and only then established that his idea would be ruined by the counter 24...xd5 25 Bxd7 Axg2+ 26 Sxg2 We5+ 27 @f3 wia+. M.T. But as they say, a game and analysis are very different things. Indeed, after 20 Hafl! (threatening 21 Hh5+ Qe6 22 Hh4, with a deadly bishop check at d3 or h5) 20... Yye3 21 BhS+ @g6 22 Bhd fa! (the only defence) 23 Hhxf4 hS (there is no other defence against 24 Hg4+ or 24 Bh4) 24 Qxd8 Bxd8 25 Ad3+ Oxd3 (the strongest) 26 exd3 ¥yxd3 27 Bf6+ @g5 28 Hxf7 h4 29 Gel we3+ 30 47f2 objectively the chances were al- ready with White, and the time trouble mistake 30...h3? (it was still pos- ible to save the game by 30...h3! 31 exh3 £3) 31 ez! Bh6 32 |Hf4 ad 33 Adi a4 34 h3 @h7 35 4)d5 allow- ed him to secure his king completely Destroying the Fortress Walls 143 and to create a new phase of the attack. After 35..uc5 36 Df6+ @g7 37 a3 &g6 38 Lg4 Hh7 39 Hel wd6 40 D|e3 yg 41 HLS wyd8 42 Heb+ HT 43 S\d4+ @g7 44 Hed 7 45 |B White won the h4 pawn, and on the 84th move — the game. Sacrifices on £7 The enticing of a king out of his fortress by a sacrifice at f7 has already been examined — and in considerable detail — in Chapter 1. Therefore here we will limit ourselves to a single, but highly instructive game. Me fies Wa, YES Sveshnikov-A.Sokoloy Sochi 1983 In the evaluation of this position, three factors stand out: Black's minor pieces are bunched together on the queenside, his f7 has been weakened by the departure of the rook from f8, and his queen has clearly come out prematurely. All this suggested to White the idea of an unusual combina- tional attack,144 15 Qh5 Wi6 16 Ags! Diverting the last defender (of course, apart from the king) of f7. 16...%xg5 17 xf7+ @xf7 (other- wise 18 &xe6, 19 4xd8 and mate on the 8th rank) 18 4)xe6 We6 19 wWd5 6 (if 19...e8 the simplest is 20 Hd6+ AQxd6 21 Wxd6) 20 syxd8+ BFS (20...e8 21 Hxe7+ and wins) 21 We5 cS (Black must keep control of the dark squares in the centre; after 21...Q.xd8? 22 Wd6+ his king has no- where to hide) 22 4)d4 (22... xf2+ was threatened) 22...@28. Now the energetic 23 He3! .d7 24 &xb7 He8 25 wd Hxe3 26 fxe3 Qxd4 27 exd4 Qc6 28 d5 Axb7 29 Wadd led to a position with an unusual balance of forces, but nevertheless won for White. ‘Until now we have been pointing out various ways of demolishing the initial, untouched wall in front of the castled position. But it only requires one of the pawns to make a step forward, for it immediately to become, if not a target for attack, then at least an object of attention by the other side. Pawns at f7-g6-h7 This pawn formation, if it is “cemen- ted" by the dark-square bishop at 7, is virtually the most secure wall around the king. But Black still has to reckon with the rapid advance of the white h-pawn, even in those cases when he has his “own play" (or counterplay). M.T. Even so, it should once again be mentioned: there is no need to be afraid, but simply to reckon with this possibility. Attack with Mikhail Tal LD. Instructive here are both thi following example, and the notes to by the winner, taken from Shakh- matnye Olympiady (Moscow, 1974), Bronstein-Palmiotti Munich Olympiad 1958 From this advance (6 ¢5) there much to be gained: (1) the knight at is driven from its post; (2) the scope the bishop at g7 is temporarily r stricted; (3) the defence of the square: hS and h7 is weakened. White promptly exploits this a: factor, by beginning a sharp attack the kingside. However, he does ne succeed in achieving any immedia gains, and, perhaps for this reason, Many games where this position w: reached, White played 6 Qe2 or. fic4, but hardly ever h2-h4. I sa “hardly ever", since I am not sure that am familiar with all the games. Inde Black's threat of breaking up the whi centre by ...c5 looks very imposing, bt I have never feared ghosts, and, afi specifically weighing up the vari ations, I came (of course, at the board}Destroying the Fortress Walls to an unexpected conclusion: is White's centre really so necessary? Especially after the weight of the struggle has switched to the h-file! Incidentally, in the given position it is not White's centre that crumbles, but his central pawns. I should like to draw the readers’ attention to a certain chance factor in White's success — the availability to the white queen of the favourable ma- noeuvre ¥yd1-d4-f2-h4. But chance factors of this type are present in any position, and the skill of a player con- sists in finding and exploiting such features. 6...2)£d7 7 h4 5 8 h5 cxdd The plans of the two players are more than transparent: White is play- ing for a direct mate on the b-file, while Black is trustingly following the dogma "play in the centre is the best form of defence against a flank attack”, Precise calculation is now required. For example, can't White rush straight into the mass of variations with the tempting 9 hxg6!? Alas, for the mo- ment such an attack js unjustified: 9... dxc3 10 gxf7+ Bxf7 11 Dg5 Df etc. There is ample scope for analysis, but since a conclusion to the attack is not apparent {which means that no real gain is also apparent), without any great regret White "courageously" cap- tured the d4 pawn with his queen, which is fully in accordance with his strategic plan and is not at all a loss of time. All the same the white queen has to make its way to the h-file, and it makes no difference by which route it goes! 9 wyxd4 dxe5S 145 Black has achieved the maximum — he has broken up the enemy centre and won a pawn. Nevertheless, after White's next move the position is, to say the least, double-edged. 10 Yyf2 exf4 As shown by later analyses, 10...e4! was correct, 11 hxg6 bxg6 12 axf4 |f6 13 wWh4 Wad A clever defence: Black threatens .o¥yh5! Here I understood the opti- mism of my opponent — 14 9.25 or 14 Se5 is pointless, while if 14 Sg5 &g4, and how is White going to cas- tle? Meanwhile threats against c3 are imminent. But even so... 14 Sg5 S24 The impression is that White's plans have been mined — castling on the queenside is difficult, and there is no mate on the h-file. But in fact White had altogether no intention of giving an immediate mate. It was important for him to open the file, in order to have the opportunity of using this threat, and as for queenside castling, well, it is not obligatory: the king will even feel More secure in the right-hand corner, among its own pieces. But won't this weaken White's attack? After all, the rook will move off the open file, Of course not. On the h-file there is al- ready one heavy piece, and, in view of the threatening position of the knight at 85, this is more than sufficient. On the contrary, there was nothing for the rook to do at hl, whereas after Kingside castling it opens fire on the File. 15 §d3 Qbd7 16 0-0 GAhS 17 Bael e5146 This decisively weakens a number of important squares (d5, f7, f6), but with passive defence too White's attack can no longer be stopped, It has to be admitted that, in decid- ing on his main strategic plan of action, White has been able to delve more deeply into the resulting com- plications, 18 Qd2 weS+ 19 Ge3 Web 20 Abs The most accurate solution, but by no means the only one. The manoeuvre 4e4-g3 would have achieved the same aim. 20...¢yc7 21 Qe2! White attacks the opponent's most vulnerable point. If the bastion at h5 falls, Black's entire army will be de- moralised, And the bastion falls! 21....xe2 22 Hixe2 Hfc8 23 Acet Bringing up the last reserves. The black king starts running. 23.08 24 |xf6 Hxf6 25 Bxfé Axf6 26 Hh7+ The white forces have broken into the Promised Land. 26...BeT 27 wyxi6+ Hes 28 Bd2 28 4.44 would possibly have led to Black's resignation, but the rook move, cutting off the king from the queenside, leads to mate. 28..a5 Indecision, or belief in a miracle, 29 eT Black resigns in view of inevitable mate. Jn the following example Black certainly has sufficient compensation for the pawn, but what was it that made White create in his position another target for attack? After all, he Attack with Mikhail Tat could have covered the d3 square, for: which the black queen and knight are clearly aiming, by 18 ¥¥b5, preparing. &d2, and at least completing his queenside development. In view of his pawn mass in the centre, there would still have been ail to play for. But... Stiwa-Stahlberg Géteborg Interzonal 1955 18 g3? HS! 19 H_2 b4 20 He2 £5 21...f4 is threatened, and after 21 f4 with what can White cover all the light squares around his king, and of whal use is such a white bishop? 21 Wye2 4)f6 22 Wed fat With fresh pawn sacrifices Black: opens the attacking files. 23 éxf4 9 xf4 24 exf4 bS! Ensuring an invasion at c2 ore2. = 25 Wyxb5 Yyc2+ 26 Wh3 hxgs 27 wyb3 Tf 27 hxg3 Black wins immediately by 27.97! 27..WES+ 28 Gxg3 He2 29 e3 We6+ 30 Hh4 Wh6+ 31 Hes sd. Two moves before mate, White re= signs.Destroying the Fortress Walls Pawns at f7-26-h6 Here it is not only the g6 pawn that is vulnerable, but also the adjacent pawn at h6, and there is a whole range of "keys" to this fortification, as an exam- ination of the following examples will reveal. 2 U7 43 Uh, WY Y, Wk oe le, Larue a “hs, Yuferoy-Nikitin Moscow 1972 The piece-pawn formation around the black king is typical of many games in which Black employs the kingside fianchetto. Typical too are both White's attack on what is here the most real weakness of this set-up — the h6 pawn, and the defence chosen by Black, who by threatening an ex- change is trying to drive back the white queen. But note should be taken of the typical and spectacular manoeuvre that allows White to continue his attack. By accepting the sacrifice — and there is nothing else! — Black opens the b-file for his opponent. 17 Ags! hxgS 18 4)xg5 Hes 19 Ads! 147 In the given instance it all proceeds very smoothly. Black has no choice. 19...Re7 20 Had 4)c7 21 Q.xf7+! Eixf7 22 eyh7+ iS 23 wyxg6 wxgd Forced, since if 23...Ge7 White's of- fensive resembles an avalanche: 24 f4 ed 25 f5, while after 23...8f6 24 S\h7+ Hes 25 Hxf6+ wWxls 26 Wxl6 Axf6 27 &xd7 further losses are inevitable. 24 WxgS Heb 25 whS Hc4 26 xd7! Bxd7 27 Wed A quiet move, attacking both black knights and — by X-ray — the rook at d7. White increased his material ad- vantage still further, and realised it without any particular problems. iy # aw bd fi O1ee os My “"o 14 “e os a Es ‘Boe. WY, Us Me: — aa ae a Tal-Petrosian Moscow 1974 WS ws i Ra “ aa Black is so far behind in develop- ment that in his kingside assault White can even include quiet moves. It is the same idea: his pieces engage in a “hand-to-hand” fight with the con- troller of the dark squares, 19 Syeg5+! hxgs Even if he declines the sacrifice, Black cannot avoid a decisive attack148 Attack with Mikhail Tal on f7 — queen from c4, rook from e7 and, if necessary, knight from e5. 20 |xg5+ Sg8 21 wid ‘The threat against f7 is incidental, the main one being ¥yh4-h7 mate, 21...4)d7 (heading for f6) 22 Hxd7! &xd7 (the other capture would have allowed mate at h7) 23 &xf7+ (now this too is possible) Black resigns. Turnhuber-Persig correspondence 1991 In principle it is just the same in this position, except that the invasion square is the adjacent one. As soon as the white queen invades at h6, the black king will succumb, But for the moment the transit square f5 is cov- ered, and so White begins by eliminat- ing the defender, 22 Hxd7 Wyxd7 Capturing with either knight is no better, e.g. 22...46xd7 23 Ahf5+ gxf5 24 &\xf5+ Sf6 25 wWxh6+ dg6 26 h4 Hg8 27 tyg5 mate. 23 &\gf5+t, and as a way of capitu- lating Black chose 23...28 24 )xg6 4xg6 25 yxh6, which differs litle from 23...gxf5 24 S)xf5+ @g8 2 we3+. M.T. In general it is rather optimis: lic to regard the h6 pawn as a reliabl guard of the g5 square... LD. ... with which the immediat approaches to the king's fortress can be considered to begin, and an invasiot on which is nearly always either th overture to an attack, or its beginning, However, the following practical ex: amples are already from another act. M.T. ... of the same opera! Pawns at £7-g7-h6 If variations with the fianchetto of king's bishop in the opening are dis- regarded, it is the h-pawn, more oft than its colleagues, that makes a ste| forward, sometimes even in the open- ing, and, as a rute, in order to drive back or at least determine the in: tentions of a white bishop at g5. 7 = qT, Si Et] Bt | Weta a MB, Mg Myre woe Tal-Byrne Biel Interzonal 1976 Black has just invited the whitDestroying the Fortress Walls bishop to "declare its intentions”, and in the event of 15 hd he is threatening the typical counter 15... &xe4, since at h4 the bishop is "unprotected. But here too White is prepared, at the cost of a piece, to effect the opening of the rook’s file. 15 ha! hxg5 (15...8fd8 was never- theless better, freeing in good time a flight path for the king) 16 hxg5 4)xe4 (on this counter Black was pinning all his hopes; after 17 2)xe4 Wxd2 his =. position is quite acceptable) 17 ¥yd3!, and it transpires that Black loses after her 17..€)xc3+ 18 bxc3 g6 19 Wh3+! Gg7 20 Wh6+ &g8 21 Bhi, or 17..2xg5 18 Bhi+ Bg8 19 fed £6 20 g6, or 17..f5 18 wyh3+ gs 19 Qxe4 Qxe4 20 g6 fih4 21 eyxh4d Hfe8 22 &xf5! He chose 17...xg5 18 4xe4 Axed 19 Bxed Ah6 20 g4 £5 21 Bxeé Sixfa 22 2xf5, and in view of the threats of 23 4)xg7, 23 Se7, 23 Bhl+ and 23 Be7, after thinking for 40 min- utes he resigned. In the game Shabalov-Smirin (Manila Olympiad 1992) in the well known variation 1 e4 5 2 Af3 dé 3 d4 exd4 4 &)xd4 Af6 5 2)c3 cb 6 Sigs e6 7 Wd2 s.e7 8 0-0-0 0-0 9 f4 h6 10 h4 White showed that he-has his opinion regarding the position after 10...2)xd4 11 Wyxd4 hxgS 12 hxgs e4 13 fe2 e5 14 wel exf4 15 Sixe4 Sixes 16 wyh2 Qh6 17 Af3 Wf6, which theory evaluates in favour of Black. M.T. Yes, we analysed it in our school (the Latvian School for Young Chess Players under the direction of Mikhail Tal - I-D.), and Shabalov was 149 there. But now the theme is different... 10...€)xd4 11 exd4 a6 12 Mer WaS 13 .f3 Hd8 14 g4 .d7 Thus the h6 pawn has not forced the white bishop to abandon its aggressive post, and it now becomes that brick, a blow at which can break up the entire defensive wall. 15 {Uxh6! gxh6 16 g5 4e8 Alas, the simplifying attempt 16... ¥yc5 leaves Black both a pawn down, and under pressure by the white pieces — 17 Wyxc5 dxc5 18 gxf6 Axf6 19 eS e720 4 xb7 Ba7 21 43. 17 Bdgl! The threat (18 gxh6+ @h7 19 He7+ @xh6 20 Bxf7) is stronger than its immediate execution (17 gxh6 .f6). 17...n5 18 {xh5 $.F8 19 £5 we5 20 Wd2 exf5 21 g6! fxg6 This opens files for the opponent's attack, but after 21...f6 22 g7 4\xg7 23 A.26 (or 23 Wh6 Ae8 24 4)d5) 23...f4 24 &d5 it is hard to find any defence for Black. 22 EBxg6+ @h7 23 Hhgl fxed 24 Bh6+!! Qxh6 25 fe6+ Sg7 26 Qixe8+ Dh7 27 Ap6+ He7 28150 fixed+ Sf7 29 wyxh6 Bh8 30 ad5+ Be8 31 Wd2 Sd8 (32 Bel was threatened) 32 Hel wyhS 33 Ae4 wh6 (otherwise 34 Yya5+) 34 2yg5 Bec7 35 He7 Hae8 36 WaS+ b6 (36...ec8 37 $.xb7+), and although White missed the quickest win by 37 ¥yxa6, and if 37..Axe7 38 Yyb7+ Hd8 39 wyxb6+ @c8 40 Ab7+, mating, he neverthe- less concluded the pursuit of the king: 37 We3+ @d8 38 Axd7+! xd7 39 We6+ He7 40 eyc7+ Wi 41 wyxd6+ BES 42 wyd7+ HeS 43 AM7 etc., that began with the sacrifice on h6. (This game and another extract from a game played in Manila, were ana- lysed for this book by the authors in a Moscow hospital, four days before the death of Mikhail Tal, the eighth World Champion in the history of chess — LD.). The illusory defence of g5 was even more energetically emphasised in the next game, in which White decisively weakened the defence of one of two vulnerable points — h6 or f7. Geller-Portisch Moscow 1967 Attack with Mikhail Tal 18 Ag5!! wd7 It transpires that there is nothin; else: on 18...47xg5 White mates by 1 ¥yxf7+, while if 18...hxg5 19 2g6! 19 Bad1 §.d6 20 Axh6 Now, when the over-protection (a: Nimzowitsch expressed it) has been. eliminated, the wall can be attacked. * 20...gxh6 No better is 20...2)xb3 21 Axe?! Gxe7 22 |f5+, when on 22...eg! there follows 23 ¥yg5+, while if 2: Sf6 23 Whét. 21 Wg6+ G8 22 w6 (threatenin 23 2)g6+ and 24 ¥yh8 mate) 22...8! 23 e3 Black resigns. Sometimes a less mobile piece succeed in breaking through onto seemingly well-defended _ "penalty spot", opposite the king, : Bronstein-Zamikhovsky Leningrad 1970 14 2)g5! BIS If 14...hxgs 15 fxg5 DdS 16 Gh7+, with mate in a few moves. 15 4)g6! He8 Alas, on this knight too there is’Destroying the Fortress Walls taboo: 15...fxg6 16 eyxe6+ Bh8 17 Wxe7 hxgS 18 fxgs Hh7 19 Axfo+ 2 bxfB 20 Be4 Qb7 21 Af4 wyxf4 22 Axb7 Ab8 23 Hf, and it is difficult for Black to coordinate his pieces. 16 Dxe6! WG 17 FS! fxg6 GF 17...fxe6 White would have continued as in the game, but with even greater effect) 18 .f4 tyc6 19 AbS wed 20 Wxed 4)xe4 21 fxg6, and although he was a piece up, Black was unable to free himself. The game concluded 21...6)df6 22 4)c7 QA7 23 Qxd7 A\xd7 24 Bael Haf6 25 2)xa8 Exa8 26 xed! Axed 27 Bel Qd6 28 Exe4 Axf4 29 Bxfs, and White easily realised his material advantage, Since from time to time the de- fender's control over g5 proves simply to be illusory, the move ...h6 should be regarded with particular caution and scepticism. Of course, in chess, as in life, there are no absolute truths: what is bad in one situation may prove good in another, and the advance of the h- pawn may avoid, for example, a mate on the back rank, But much more often it proves ruinous, by presenting a target for a pawn or piece attack. In the following position White's pieces are unusually well coordinated: they are all aimed at the black king, the undoing of which is the h-pawn. Indeed, if it were at h7 with the knight at f8, Black's defensive resources. would not yet be exhausted. As it is, the advance of a white pawn com- pletely breaks up the king's fortress. 22 gat dé In the hope of 23 g5 ¥d5, since 22...4)df8 is clearly insufficient. 15) Alekhine-Weenink Prague 193] 23 Ag6 BES 24 g5 A xd4 25 gxh6 2)df6 26 hxg7+ Sxg7 27 wWh6+ Gh8 28 dyxd4 Yyxd4 29 Ab? Black resigned, since if 29..%yd7 30 d3! ‘The weakness of White's back rank apparently does not allow him to con- tinue his attack in the "natural" way — 30 xh6+ Qxh6 31 Bxf+ HxfB 32 ¥yxh6+ He8, and it is Black who wins. But the assault ratio. reaches its152 maximum after 30 ¥yxh6!! The queen is immune on account of mate in two moves, 30...xf7 loses to 31 Hd8+ and 30...Hf6 to 31 Wh8+!! Qxh8 32 Hho mate, and the only defence 30...25 leaves Black both two pawns down (31 ¥yxg5), and under attack. To the delight of one of the authors, the game Schneider-Tal (Lucerne 1982) took a quite different course... LD. Before this next example it should be said over and over again: to conduct an attack, relying on one sin- gle "pure" stratagem, even a standard one, is practically impossible. MAT. At one time, in my youth, David Bronstein's maxim about the great benefit of knowledge seemed something of an affectation. Now it is simply banal for us to talk about this: a player who is scientifically "equipped", not only in the opening, but also in the endgame, sees the board differently. Gradually this procedure became typical in similar positions, and from being the possession of grandmasters it became "common property". Of course, although the — sacrificial mechanism may be the same, each game has its own features (see diagram next column). 24 Qxh6! gxh6 25 Hf6+ Gh8 26 wed2 Af4 27 wd4t Within two moves we will see why White lured the black bishop to £4, and did not play 26 ¥¥c3 immediately. 27.05 28 Axe6 L.xc6 29 yh5+ £6 30 \xf4 This is the whole point. White regains his piece, he is a pawn up, and, most important, his attack on the broken king's position continues. Altack with Mikhail Tal Ivanchuk-Rakhmanov Junior Tournament, Klaipeda 198. 30..Wf5 31 Heb Hes (31.47 32 &)d8) 32 .d3! Black resigns. : Since if 32..WyhS 33 Wxf6+ @h7. 34 }fS+! HexfB (34...Aaxf8) 35 He7+. Even when it appears that nothi can threaten the h6 pawn, it sometimes. proves possible to exploit this very. important nuance of the position. : Tal-Keres Tallinn 1973Destroying the Fortress Walls 16 4)d5! White was ready to content himself with a positional advantage after 16... wWd8, but Black decided to accept the challenge. 16...%xf3 17 Qdl 4h4 18 gxh4 Wh3 19 46+! This Black did not expect, and he gave up the exchange by 19...@h8 20 &xe8 Bxe8 21 h5, after which he gradually lost. Of course, essential was 19...2xf6 20 Wxh6 exd4 21 @h] He5 22 Bgt+ - 4 (22..426 is insufficient — 23 Eixg6+ fxg6 24 Yyxg6+ Gh8 25 wh6+ - Sg8 26 Gh5) with a sharp game, but » White's entire manoeuvre became pos- sible precisely because of the advance of the h-pawn from its initial post. It is worth studying the possibility of an explosive attack on h6, even when he opponent has obvious counterplay. Varanin-Moskvin USSR 1989 For example, here Black is ready by .2)d4 to set White serious problems, ind at the least to exchange part of the 153 attacking forces, while maintaining control of the open file. And yet White still risked the sacrifice: after all, the advanced pawn at h6 is as though created for this. 31 Qxh6! After 31 )xh6+ gxh6 32 &xh6 4 f8 White has no good continuation. 31...2)04 32 Wyf4 dyc2 Preferable was 32,..Q.xf3 33 Qxf3 @xf3+ 34 wyxf3 Ags 35 Axes Axe5, with some compensation for the pawn. 33 Qxg7! exg7 (33..Qxel 34 Dh6+ Sxgl 35 wyxf7+ Ph8 36 Hh4!, and Black loses too much) 34 4)f6!! The culmination of the attack. Exploiting the fact that his knight is immune, White again creates a threat against f7. Again Black has no time for 34... &)xel on account of 35 4)g5! 4)xg2 36 He8+! Gh 37 Hxf7+ He8 38 wede Sxt7 39 Wye6+ SFB 40 Axc7 Hgs 41 h6. 34...8h8 35 Hed1! (forcing the op- ponent's reply) 35...g)xal 36 2)g5! All White's pieces are now in the attack, whereas after 36 Hd7 ¥yxd7 37 &xd7 Axf3 38 Axf3 HAgs there would have remained insufficient force. 36...Had8 Black loses amusingly in the forcing variation 36...Qxg5 37 Wxg5+ Of8 38 Hd7 Wxd7 (38...xf6 39 exf6 ¥yxd7 40 ¥¥g7+ leads to mate) 39 D)xd7+ Hes 40 Hf6+ Gxf6 41 wyxf6 BxhS 42 4xb7 Hb8 43 4f3 Bh7 44 Se4 BhS 45 94 Bh3 46 @e2. 37 HeB+! And in view of mate in three moves (37...@h6 38 xf7+ SxhS 39 wh6+134 Sed 40 tyh3) Black resigned. In addition, the appearance of a pawn at h6 immediately provokes (or should provoke) in the player with White the seditious thought: is it not possible to carry out an "attack from afar": sacrifice the dark-square bishop from its initial position, and then with a quiet move bring the queen onto the cl-h6 diagonal, 1D. I came across this manoeuvre in some games from the 1920s... M.T. ... but even so it received full recognition only after the following game. Bronstein-Keres Goteborg Interzonal 1955 For the moment White has sacri- ficed two pawns, obtaining in return the £5 square for one of his knights, 46 for the other, and the real possibility of regaining one of the pawns almost by force: 14 4)f5 He8, and apart from the possible sacrifice on h6 he also has 15 2)bd6 (which, however, does not lead to the win of the exchange: 15... Axd6 16 4)xd6 He6! and if 17 4\xb7 Wc7). Attack with Mikhail Tai Bat he made use of a different tacking mechanism. 14 Qxh6!? gxh6 15 Wd2z Hh7 Later analyses showed that by it mediately bringing up his reserves 15...2)c5!, with the aim of blocking b1-h7 diagonal, Black would have hi better chances of defending his king, eg. 16 Haet 2d3 17 Qxd3 cxd3, if 18 DPS Het 19 Hbd4 Bes D|xh6+ GAB 21 Wes Heb 22 Hx Hxe7 23 Wyxf6 Hed! 24 Wh8+ Ge7. @Qhf5+ AxfS 26 HxfS+ Beb 27 wy ¥yh8, when nothing forcing is apy rent, and he remains the exchange up. But the attack can also be continu by 18 Wyxh6!, when 18... Hb7 19 Hf M16 20 He3 He8 21 He3+ Hh8 22 Hg7 is bad for Black, and 18..He8 ¢: be met by 19 f5 Af8 20 wye5+ HI 21 )\bd6 He6 22 He3 gS 23 Wh yhG6 24 Hxh6 Axh6 25 wWxl7+ HI 26 Hxe6 dxe6 27 Wyxb7. In the game Black decided simply to extinguish the first wave of the attack, keeping a favourable material balan 16 Yyxh6 £5 17 2xf5 Exfs He has to give up the exchange; 17..Hf7 18 We6+ SEB 19 Ab ALxd6 20 Wyxd6+ Hes 21 H6+. 18 AxfS Hf 19 Badl Ae5 WhS WEG 21 Ddb Ac 22 wed HY Parrying two threats, 23 ¥xc4+ 23 h4, But the king's fortress has b completely destroyed, and with his very next move White succeeds in ating a further three(!) threats, whi cannot all be parried simultaneously. 23 Qedt On the agenda is 24 ¥yxg5 follow by 25 4\f7+, and 24 @xc4 followed bs 25 &\e4, and — the game continuation.Destroying the Fortress Walls 23.86 24 Axc6 dxc6 25 kyxc4 4yc5 26 b4 4\ce6 27 tyxc6 White now has a material advan- tage, and the pursuit of the king is merely postponed for the moment. 27..Bb8 28 4)e4 Wye6 29 Rd6 997 30 f4 Wed 31 h3 wye2 32 293 weds 33 @h2 Ad4 Or 33...4)xf4 34 Bh6+ 4h7 (if 34... He8 35 Weds 2406 36 Bxf8+) 35 Wd6! Axh6 36 wyxb8+ Af8 37 wd6, and the ending after the exchange of queens should be won for White. 34 wWd5 HeB 35 Hh5 He2 36 Q)xge7 We3+ 37 Sh |xfd 38 Wh 4e2 39 Bh6+ Black resigns. If the king's position is even more weakened, it can be very straightfor- ward to mount an attack of this type, as in the following example. Voitsekh-Morozov USSR Correspondence Ch, 1973-5 The pawn formation around the lack king suggests only one solution. An experienced player would not even bother calculating variations in ad- vance, since this can be done later: the 155 weakness of the b}-h7 diagonal is just too gaping, the f- and h-pawns, which have incautiously stepped forward, are just too vulnerable. 22 &xh6+! gxh6 23 Yyg4+ Gh8 24 Axhe! LD. Misha, how much time would you have spent considering this second sacrifice? M.T. One second at most. Firstly, White's rook at el was undefended, and the exchange at e8 would bring up the black queen to the defence. Secondly, it is a pseudo-sacrifice: after 24...4.xh6 25 ¥¥g6 White, at the least, regains the piece. And thirdly, if you and I are writing a book about attac- king procedures, it means that they are familiar to us... LD. In my opinion, from this ex- ample one can best of all not so much understand, as sense what is meant by the weakness of a single square or a complex of squares. Once David Bron- stein gave a classic formulation of this: by moving to a weak square, a piece is able to attack the neighbouring squares and the forces standing on them. An ideal illustration is the variation 24... Sxh6 25 wg6. Such an invasion would not be possible, for example, with a black pawn at f7, i.e. without the "holes" in Black's position. 24...He7 25 Qxf8 eyxf8 26 wyhS+ @g8 27 Hxe7 wyxe7 In principle, simplification favours the defending side, if he also has a material advantage, But here exchan- ges have eliminated defenders that are difficult to replace, whereas all White's reserves can easily join the attack. 28 AfS Wd7 29 eye6+ Bis156 (29...@h8 30 Wxfo+ Hes 31 Dh6 mate) 30 Wyxf6+ WE7 31 wWh6+ Ge8 (31...g8 32 Hel 4)bdS 33 He6) 32 Hel+ @d7 33 He7+ wxe7 34 2)xe7 @xe7 35 we7+ And in view of the forced variation 35...@d6 36 Wes+ Sd7 37 Qfs+ Hs 38 Wh6+ He8 39 Heb, with a quick mate or major loss of material, Black resigned. Tal-Gurgenidze 36th USSR Ch., Alma-Ata 1969 Caro-Kann Defence 1 e 6 2 44 d5 3 Qed bs 4 a3 dxe4 5 &xed OFS The game has immediately taken an unusual course, but even so Black decides against finally leaving the safety zone, by taking the pawn, either now (6.,.¥xd4 7 €\f3 wyd5 8 Yye2), or two moves later. 7 Gxed 56 8 343 6 9 &f3 Ret 10 we2 éybd7 11 0-0 0-0 12 Hel Bes 13 eS xe5 14 dxeS Qd5 15 wea Not objecting to a few exchanges, White has removed from d4 a target for counterplay —- the pawn, driven away a defender — the knight at f6, and created all the preconditions for an attack on the king. Therefore prophy- jaxis by 15...g6 was indicated for Attack with Mikhail Tat Black, when nothing now is achieved. by the hasty 16 h4 axh4 17 g3 Ae7 18 @g2 on account of the sim 18.,.h5, White would have had to pre- pare it with 16 g3, But Black did not: yet sense the danger. 15... ad 16 hd! : ‘Threatening a direct win after 17 h5: and 18 §h6 Qf8 19 Axe7 Axg7 h6. So the pawn sacrifice, whi however, serves merely as an overture: to a planned rook sacrifice, has to accepted, 16... Axh4 17 33 Sev7 18 @g2 86 Black no longer has time 20 tyh5+ @g8 21 Bhl, and barrica such as 18... Qf8 19 Bh1 £5 cannot maintained: 20 ¥yhS h6 21 eyg6 @ 22 Qxh6 gxh6 23 Hxh6+ Qxh6 Bhi 2)e3+ 25 fxe3 wd5+ 26 e4. 19 Bhi = ge 20 Best An important interposition, not al- lowing the opponent to complicate game to his advantage in the event. 0 the immediate 20 Hxh7 @xh7 21 95 &e3+! 22 fxe3 (or 22 xe3 27:2: Bhi+ @g8 24 Wh3 We7 or 24...d and 25..¥yxeS) 22..u¥d5+ 23 Gs Wyxe5 24 wyhd+ Be8 25 4f6 whs. 20... wWe7 The black queen has to give up th idea at some point of checking at d: because of variations such as 20... 21 xh? Qxg5 (or 21...@xh’ WhS+ @g8 23 Qxg6) 22 Bahl 23 Axe6, or 20..4\e7 21 Axg6Destroying the Fortress Walls (21...nxg6 22 Qf6 G27 23 tyh4 also concludes matters) 22 ¥yxe6+, and of the many mates, this one is rather elegant: 22...@h8 23 Bxh7+ @xh7 24 Bhl+ Qh6 25 Bxh6+ g7 26 Afe+ Gxh6 27 wh3 mate. 21° Gxh7! | Wyxe5 If 21...@xh7 22 Bhi+ gs 23 Af6 @xf6 24 exf6 Wes 25 Axe6 wx 26 &h7+ @h8 27 eyes mate. 220 Oxf?! Persisting with his desire to offer this rook. 22... xi? 23° Axg6+ Sg8 Or 23...xg6 24 Gfd+. 24 gxeB fig? After 24...Gxe8 the black queen would have been attacked from the other side — 25 4 f6+. As it is, White - continues his offensive, now with a material advantage. 25 §d7 Nc 26 Axc6 Hf 27° «Bdl Wed 28 Qf wyxe2 29 Bd7 8m 30 Bd8+ = BS 157 31° Af6 wh7 32 fed whe 33° Ags whs 34 «Bd7 Black resigns: if 34...8f7 35 Bxc7. Spassky-Tal Montreal 1979 Queen's Indian Defence 1 d4 ate 2 4 6 3 AB b6é 4 3 The Queen's Indian Defence is an opening that until recently had a very quiet reputation. Sometimes in news- paper reports one would even read comments such as: "The players used the Queen's Indian Defence for peace- able aims.” However, a certain reas- sessment of values is currently taking place, and in particular, this opening is often and very successfully employed by Karpov. There is no denying that White, if he wishes, can play with a high degree of solidity (the move that best meets this aim is 4 g3). But peaceableness on the part of my opponent was the last thing Texpected in this game. 4... ib7 3 &d3 dS 6 b3 This is no better and no worse than 60-0. Perhaps the only slight nuance is that White does not allow his opponent the opportunity for relieving man- oeuvres such as 6...dxc4 and 7...c5. 6... dé 7 00 0-0 8 gb2 Q\bd7158 9 bd2 — tye7 10 Bel Bads Up to now the moves have been more or less obligatory. Both sides are deploying their forces as well as pos- sible in anticipation of the coming bat- tle. The altemative here was 10... e4, but the position after 11 wWe2 f5 12 4)e5 did not particularly appeal to me. With his last, consolidating move, Black invites his opponent to deter- mine the position of his queen. In reply to 11 wye2 I was now intending 11... $)e4, while after the continuation cho- sen by Spassky, I thought that the advance of the ¢~pawn would gain in strength, Ut wed cS 12 exd5 The other, roughly equivalent pos- sibility was 12 Bfdl, against which I had prepared 12...cxd4 13 exd4 Bc8. In such situations the loss of a tempo is not so significant, o., 13° dxcS But this exchange is, I think, bad. The weakness of the hanging pawns in not apparent at all, and Black's pieces operate together most harmoniously. In my opinion, 13 &b5 or 13 Af5 was stronger. 13... 14 wed And this is simply a challenge... When after the game I asked Spassky why he made this move, he spread his hands; "A beggar isn't afraid of being robbed". The idea of a pawn breakthrough now becomes very possible. I spent some time examining the immediate exd5 bxe5 Attack with Mikhail Tal 14,..¢4, and came to the conclusi that for the moment it did not promi: anything real, and so, guided by eral considerations, I brought into pl my last, “sleeping” piece. 4... Bees 15 Bfdl That day it would seem that Boris was betrayed by his sense of danger. expected 15 Bfel, in reply to which I was intending seriously to consider: 15...04 (16 bxc4 {2b4 17 Wye2 dxed) But now White himself provokes breakthrough in the centre, and Blac pieces are so ideally placed that the is no sense in delaying it. 16 exdd 17° Wad Good or bad, White should have cepted the pawn sacrifice. It is true tt 17 Wyxd4 loses material after 17...2) but after 17 4xd4 I could see no fore: ing way of developing the attack. For the sake of my conscience, so to speak, I had prepared a reserve variatit 17...8¥e5 18 Q4f3 Wh5, with full co pensation for the pawn, but even so. exddwould probably have been unable to resist the temptation — it is not often that opportunity occurs for sacrificing a bishop at h2 in games at grandmaster level. But here things are not com- » pletely clear. After 17... Qxh2+ 18 Bxh2 Ded+ 19 g3 (bad is 19 Bgl < Wyh4 20 2 4f3 Wyxf2+ 21 Bhi Hes! 22 FS He3) 19...¥¥e5+ 20 f4 (after 20 &xg4 |f6+ White is mated) 20... We3+ 21 4f3 |df6 Black's attack is very dangerous, but White is by no means doomed (there is the possibility of 22 Qxh7+, for instance). At any event, the continuation cho- sen by Spassky is bad. Firstly, because the "condemned" pawn is still alive, and in addition the position of the queen on the Sth rank shortly gives Black the opportunity, with gain of tempo, for the decisive inclusion of his rook in the attack, 17... 18 &xe5 18 Hel was perhaps slightly more tenacious, although Black has a very attractive choice between 18...4axf3 and 18...6)xd3!? 1... 19 He4 Black's attack develops slightly more slowly, but probably just as ef- fectively, after 19 4)f1 £)d5 20 4)g3 Hes fixeS 9... ds Here it is, the decisive tempo. For the bishop sacrifice all is ready. 20 wd2 I was expecting 20 {a3 (with the idea of driving the queen off the d8-h4 diagonal), which is decisively met by 20.896 21 ¥yd2 Gxh2+! 22 Sxh2 Destroying the Fortress Walls 159 AbS+ 23 @gl Bhi+!, with mate in two moves, 20... Sxh2+ A familiar theme in a slightly new Setting. 21 @xh2— BhS+! White was evidently counting only on 21..Qg44+ 22 @g3, but now the way forward for the king is blocked. After 22 @g3 Black has the immed- iately decisive 22..6e4+ (however, 22...g5+ is also good) 23 Axed wh4+, 22 @egl AeA White resigns Against the two threats, the prosaic 23...¥h4 and the elegant 23... Hhl+ (in reply, say, to 23 Hel) there is no defence. Tal-Stean Alekhine Memorial, Moscow 1975 Tarrasch Defence 1 Of DI 2 4 5 3 c3 6 4 3 a5160 5 exd5 exdS 6 d4 Deo 7 fez Me 8 0-0 0-0 9 ges exd4 10 4xd4—h6 Ti Ae3 Bes 12 Wb3 From the fact that after this my op- ponent thought for a long time, I con- cluded that I had succeeded in devising something new. 2... QS 13 Wed At a5 the knight stands no better than at c6, but for the moment Black plays logically. 13 |... Qed 14 Ad7 But this is inconsistent. 14...Qh5 was clearly stronger, keeping the e2 pawn under attack, The continuation could have been 15 Had! Bc8 16 4 fS 4ib4, with an unclear game. 15 Bad It transpires that at d7 the bishop is only in the way of its own pieces. Hed 16 |fS Not so much play against the iso- lated pawn (White has in reserve the strategic manoeuvre $.d4 and )e3), as the start of an attack on the king. It was for this reason that on the previous move the dl square was occupied by the queen's rook. The point is that in teply to 16...2)c4 White has a combi- nation: 17 4)xd5 4xd5 (17...2)xe3 18 fxe7+ Hxe7 19 Wxe8!) 18 Axd5 )xe3 19 |xh6+! YB 20 fxe3!, and the rook at fl has its say. Comparatively best was 16...Q.e6, Attack with Mikhail Tal but this would have meant admittin; that all the preceding manoeuvres wit this bishop were pointless. 16... AES 17 &xdS eS Possibly Stean was hoping to gai positional compensation for the paws after 17..Qxd5 18 BxdS tyc7, but then noticed that 19 {3 f4 leads to fur: ther loss of material. Otherwise Blac] would hardly have willingly allow the following piece sacrifice. 18 2xh6+ gxh6 19 Wg6+ Bhs 20 sixf7? — ac The only move. The game woul have concluded instantly after 20. 97 21 Qxh6 Axh6 (21...¥yf8 Wxf6) 22 Hxd7! 21 Hd5 My old illness — trying for too much, Far simpler was 21 .xh6 @.xht 22 wWxh6+ Ah7 23 eyf4, when al Black's pieces are "hanging". But now he should have played 21:. Hixd5 22 4)xd5 27, wher it appeat that White has nothing better than 2: &xh6, which, however, is quitsufficient for a win. 21... we7 It was this very move that White was reckoning on, since the resulting position very much appealed to him, dt = &c4 23 «fa It transpires that if 23..Qf5, as planned, White wins simply with 24 BxeS Hxe5 25 Wxf5 Axf7 26 Ads. 24 SxeS Axes White would have been set greater problems by 24...6\e3, but of course this too could not have saved the game. For example, 25 f3 S\exd5 26 4)xd5, or 25.865 26 Qxf6 Hxf6 27 Be, and several pieces disappear from the board, but they are mainly Black's. 25 HxeS wi8 White has such a large material re- serve that it is by no means essential to hang on to the exchange. 26 Gb3 ~ ded 27° was yxe5 In reply to 27...4xe5 White wins most simply by 28 ¥xd7, 28 fxeS WeS+ Destroying the Fortress Walls 161 29 @h2 30 06 White's extra material does not pre- vent him from continuing his attack. e8 see Bd6 31 Wed Black resigns Tal-Portisch Candidates Quarter Final, Bled 1965 French Defence 1 e4 6 2 a4 as 3 Ac3 26 4 25 dxe4 5 &xed = \bd7 Until 1962 this variation was not especially popular, and was employed only rarely — and on those occasions when Black very much needed a draw. At the Candidates Tournament in Cur- acao, this variation was chosen against me by Petrosian in round eight, and by Benko two rounds later, In the game with Petrosian I played 6 \xf6+ 4)xf6 7 &f3 c5, and after an hour's thought I chose the absolutely unique plan of 8 Wd3 Be7 9 Axf6 Axf6 10 wb5+. Naturally, White lost very quickly. The game with Benko developed more normally, but during the encounter with Portisch I could not recall the pre- cise move order (until now I have en- deavoured fairly successfully to forget all the games that I played in Curacao), 6 axf6+ Perhaps the most accurate. After 6 £3 §e7 Black succeeds in simplify- ing the position, since 7 4)xf6+ can be answered by 7...axf6. 6... xt6 7 AB cS162 Attack with Mikhail Tal 8 ged I think that it is in this way, without trying to refute Black's opening, that White attains the most promising posi- tion. Attempts to force matters by 8 ®e5 or 8 &b5+ do not achieve any- thing real against accurate defence. es exd4 > 0-0 fhe? 10 wWe2 ho Black's desire to get tid of the bishop at g5 is understandable. But in the future the pawn at h6 will draw the attention of the white pieces aimed at the kingside, The h6 square is a very convenient one on which to sacrifice. Perhaps more in the spirit of the position was 10...0-0 11 Hadi d5 12 &xe7 4)xe7, when Black's knight can begin shadowing its white opponent: in the event of 4)e5 Black replies ...2)26, while with the white knight at d4 Black will prepare ...£)c6. In positions of this type the exchange of knights is to the advantage of the defender. 1k Ata 0-0 12 Badi 4d7 13° Bxd4 White thought for about half an hour over this move. It was difficult to decide which was stronger — the openly aggressive move in the game, or the more reserved 13 4xd4, after which White could have manoeuvred his rook along the third rank without loss of time. In nearly all variations White's attack would have developed completely unhindered. But what did not much appeal to me was that Black could reply 13...2)d5 14 Me5 Af6 15 Axd5 exdS 16 Axfb Wxf6, and although White has an un- disputed positional advantage, it ma prove insufficient for a win. White cai easily obtain three quarters of a poi but after a defeat one wants more.. 3... Webs 14 Wd2! Tt was with this continuation in mi that on his previous move White. cided to capture the pawn with rook. I did not consider any of! moves. Therefore [ was most surpris when after the game Portisch info me that only here had we diver from the Tal-Benke game, where I chosen the, in my (present) opinion, ridiculous continuation 14 d3. It wat not surprising that here, having up against an innovation, Porti thought for some forty minutes, 14... Bcé Black could also have defended bishop, and simultaneously prepare meet the coming bishop sacrifice, 14...Hfc8. After this I was intendi once more to check the sharp vatial 15 Qxh6 gxhG £6 vyxh6 wyxd4! &xd4 Hxc4 18 Bdi with dangerout threats, and if this proved insufficient:Destroying the Fortress Walls had in reserve the unpretentious retreat 15 (b3, maintaining an attractive position, Now White's reply is practi- cally forced, since otherwise 14 ¥d2 is simply a waste of time. 15 Axh6 ed ‘The only move. In the event of 15... gxh6 White would have continued the attack by 16 ¥yxh6, when if 16....xf3 he has the decisive 17 wyg5+ @h8 18 Bh4+ h7 (or 18...Qh5 19 4.03, with the irresistible threat of 20 Rxh5+) 19 Yyxe7, when 19...g8 fails to 20 . WiG+, while 16...4se4 is very strongly met by 17 43. If Black changes the move order by playing 15...4.xf3, if he wishes White can transpose into variations already considered by 16 {1.xg7. Besides this, the simple 16 gxf3 gxh6 17 @h1 is also good. 16 wf gxhé 17) Bxed This gives the game a rather differ- ent direction. In return for the sacri- ficed exchange, White counts on main- taining a persistent initiative. The attempt to force matters by 17 Wxh6 succeeds after 17...Q4c5 18 Qg5!, 17... Hfd8 18 Qxe6!, or 17...e¥e5 18 Ad3, but after the only good move 17... Had8!, keeping {7 defended, White probably has nothing better than to force a draw by perpetual check: 18 Mxe6 Bxd4 19 gf5 4f6. IT ass Axed 18 Wxed Here Portisch again thought for a considerable length of time. He has a minimal advantage in force — the exchange for a pawn. But the exposed position of his king and the presence of Opposite-colour bishops call on him to 163 be extra-careful, Thus 18...4yxb2 loses almost immediately to 19 4e5!, when it is impossible to defend against the various sacrifices on e6 and f7 (19... 4416 20 4.03). I thought that the best defensive resource was 18...0f6 19 Ad3 AfdS 20 Yh7+ SS 21 wxh6+ Bel (21...4.g7 is weaker: 22 wf4 or 22 Wyh5) 22 b3, but here White now has two pawns for the exchange, and his initiative shows no signs of diminishing. Portisch aims to include his rook in the defence. Hadst The prospect of reestablishing ma- terial equality after 19 4.43 Hxd3 and 26...09xb2 appeared insufficient. 19... Acs Now the idea of Black's defence is revealed. First of all, for the moment he is restricting the white rook by the attack on £2, and he plans for the bishop to take part in the defence from d4. The following manoeuvre by White is aimed at further weakening Black's kingside. It involves the calculation of a lengthy variation, a calculation which, unfortunately, is inaccurate. Meanwhile, by continuing simply 20 ¢3, White would have retained all the advantages of his position, and the defence would have involved enormous difficulties. However, White's over- sight i a rather amusing one. 20 wa We7 21 WeS+ This forces the advance of the pawn, since it is hopeless to allow the queen in at f6, a... fe 22 We3+164 Naturally, White did not even con- sider capturing on e6. 22 oe 23° Bel Here Black had a very interesting defensive possibility: 23...(.b4. I had taken this into account, and had pre- pared what I imagined to be a winning variation: 24 xe6 Axel 25 Af5+ Bhs 26 Wed Axf2+ 27 SFl Adi+ 28 Se2 we3t+ 29 Hxdl. But while Portisch was considering his move, I noticed to my horror that, by playing 28...Bel+, Black would give mate first. White would probably have had to move his rook, or else force a draw by 24 c3 Qxc3 25 Hh4 Axel 26 We6+. After the mistake committed by Black, White's attack gains in strength with every move. 23 see Bg8 24 Wyhd Bd6 25 Gf I do not feel inclined to attach a question mark to this move. It looks perfectly logical. White frees his queen from having to defend the £2 pawn, since nothing is gained by capturing it Attack with Mikhail Tal without check. But here he had a very fine opp tunity of instantly gaining a decisive advantage, by playing 25 c3 a5 26 a3: (this possibility was pointed out, im- mediately after the game, by my trainer, Koblenz). Now, in view of the. threat of 27 b4, Black is forced to move either his queen or his rook, b then White captures one of the pawns, e6 or f6, with decisive effect. E, 26.06 27 Wxf6 Axfl+ 28 Sfb Bxc4 29 Bxeb, or 26...£5 27 b4 axb4 28 axb4 Eg4 29 wye7+ Be7 30 bxed: Wrxe5 31 WS! Bg8 32 w+ Ag7 33 5+! hxgs 34 gyhS+ Sg8 35 Bxeb. Fortunately, White's omission does not. alter the general evaluation of the p ition. It is extremely difficult for Bla to defend against the numerous threats, especially when in severe time trouble. WB see £5 26 b3 Preparing for the g-pawn to come into play as a “battering-ram”. : 26... He6 26...4¥c6, preventing White's next move, was perhaps more tenacious. 27° gat This destroys, once and for all, the black king's shelter. Black achieves nothing by 27...fxg4 28 hxg4 Yyc6 on account of the simple 29 4e5, when does not have a single check. Here too. the best defence was probably 2’ We6 28 gxfS Wxf3 29 fxg6+ Hxg6, but then White is material up with an: active position. The move played by Portisch loses very quickly. 27 ee &d7 28 Bxeét It was still possible to fail into a traj— 28 Sxe6? @xf2!, and Black suc- ceeds in simplifying the position. But now it is all over. 28... Bdl+ 29° Sg2 Bxe6 30 xe6 fixed 31 owxgd) = Ad 32 Hes Black resigns WHAT WOULD YOU HAVE PLAYED? No.30 The white outpost at e5 creates the preconditions for an attack, but it is inadequately supported and sooner or later will inevitably become a target for Black's counterattack. Nevertheless White has some initiative, and... _ _ At first sight position No.31 seems ‘airly simple, but in fact it is extremely ifficult for Black. 13 {.g5 is threat- ned, and the simplifying attempt 12... 4)xd5 loses quickly to 13 Q.xd5+ @h8 14 (1 g5 We8 15 wyhs! Destroying the Fortress Walls 165 Black could have defended with 12..4e7 13 &xf6+ Bxf6 14 Bxf6 exf6, although here too White has a great advantage. In fact he chose 12... hG (preventing the pin) 13 Yyd3 Qe6. Now 14 4xf6+ 8xf6 15 d5 does not work in view of 15...4e5, but... Combinations of this type should occur to a player "spontaneously"...166 Attack with Mikhail Tal This applies even more so here... No.34 Although nominally White has material advantage, can bis king safe? After all, the leading bastior his fortress, the g3 pawn, see securely defended.. But in this position a minute or two's thought is in order,., Here the cement from the fort walls has already been washed awayAnswers to: What would you have played? No.1 (p.30) Tal-Simagin 23rd USSR Ch., Leningrad 1956 We do not know, because the knight did not even think of retreating. It offered itself as a sacrifice, depriving the black king of the right to castle. 12 &xf7! @xf7 13 £5 Naturally, lines have to be opened for the attack on the king. 13...dxe5 14 fxe6+ Hxe6 15 Bb! Unexpectedly White's attacking po- tential is increased, for example, in the variation 15..Wa5 16 &xb7, when his threats become irresistible. The rook also comes in useful on the b-file in the event of 15...4a6 16 Wed+ d6 17 dxe5+ @c7 18 @f4, and therefore Black accepts the challenge. 15..xyxbl 16 Yyed+ YHd6 17 Y.a3+ He7 18 Hxbl 4.xa3 19 wb3t For the queen Black has sufficient material compensation, but White's lead in development makes his attack irresistible. 19... L.¢7 20 Wxb7+ Bd6 The king is forced to go back. 21 dxe5+ S)xe5 22 Hdl+ Be6 23 Wyb3+ OFS 24 Afl+ Hed 25 Hel+ OS 26 gd+ OHf6 27 Hfl+ Yes 28 Weot+ Ph7 (28...46 29 wi5+ and 30 ¥yxe5) 29 Yyxe5, and apart from the itiative White now has also a mate- rial advantage. No.2 (p.30) Tal-Tukmakov Sochi 1970 Yet, strangely enough, it practically has no choice. If, with such strong “X-ray” pressure along the a2-g8 di- agonal you do not begin an attack on the black king, then when do you ever attack? Unless perhaps White's extra pawn could have raised its voice after the retreat of the knight and the slow realisation of his extra pawn, But what player listens to the opinion of a pawn?! 22 E)xf7! Gxl7 23 dxe5 QxcS Black decides not to be obstinate, and promptly returns the piece. After 23...2)h7 24 wyf34+ gs 25 Hdl he would nevertheless have had to do this, but in even more unfavourable cir- cumstances, 24 exf6 Hxel+ 25 fyxel Wxf6 26 Sed we7 27 tye2 BdB 28 wxbS Wyxed 29 Wyxb7+ Bl 30 He3 Returning part of the material won, White opens the f-file for the con- cluding attack on the king. Therefore Black resigns, No.3 (p.31) Tal-Zaid Moscow 1972 Nothing of the sort! Chess is not168 draughts (checkers), capturing, as we know, is not obligatory, and important intermediate moves are sometimes available. 19 Axf7+! In this way the black king is kept in the centre, there being no question of White's attack being calculated as far as mate. It is based on the sincere belief (and knowledge!) that a king in such a situation is bound sooner or later to succumb. 19...xf7 20 wdS+ G26 Tf 20...@f8 21 Bxed, and 21..44d6 fails to 22 Hfl+. 21 Wyxed+ Of7 22 Hfl+ Ale Now, after interposing 23 ¥yd5+, White could have brought to the centre his last reserve his knight. But: 23 Dc5S M.T. I did not anticipate the bril- liant defence of my young opponent; the game was played in a clock simul- taneous, and Zaid was then only a candidate master. 23...fle8!! 24 wydS+ If 24 §xf6+ Black would not have replied 24...gxf6 25 Wh7+, with mate in a few moves, but 24...@xf6!, while on 24 Bixe8 WxeS+ 25 Bhi he had prepared 25..¥¥c6 (the showy 25... Af5 does not work, since White again interposes 26 HxfS Bxe8 27 Wxe8+), and after the forced exchange of queens the black bishops are ready to show their strength. 24.8 25 Wyd6+ G7 26 Qxb7 Wad?! Time scramble; 26,...xb7 27 ¥yxa3 Bixe5 was better. 27 Wd5+ BB 28 Dd6 Heb 29 Hixe6 Exe6 30 Wyxe6 Wdd+ 31 Gh Attack with Mikhail Tal Wa7 32 DFS wyi7 33 Wd6+ Ae7 We6 d8 35 ¥yxa6 Black resigns. No.4 (p.31) Tal vy. Gomez-Baillo Termas de Rio Hondo 1987 But the {7 square is already "ripe’ and White immediately exploits this. 13 4)d6! exd4 Or 13..We7 14 Qf4 4h5 15 Bfe WG 16 GeS sxe5 17 dxed with great advantage, since the pressure £7 continues. 14 Hxf7 HcS White also wins the exchange af 14.,.Bxf7 15 Qc4 4c5 16 Axi Shs (16... F8 17 b4) 17 ¥ya3. Ab4) 19 Bfel : And Black resigned in view of variation 19,..b7 20 4)f7+ @e8 4)d6. No.5 (p.31) Barcza-Tal Tallinn 197] And Black found a way of keepinj the white king in the centre! 9... Qb3! 10 2yfxd4 After 10 0-0 xf3+ White loses tt exchange, while 10 4\exd4 Qxg2 1 Bgl exd4 12 4xd4, in the hope 12...h3 13 g4 followed by Hg3, countered by 12...c5! 13 4\b3 £31 24 £5! 10...xg2 11 Hel exdd 12 Axi 5 13 4)b5 9 £3 14 gd For the piece White has on}Answers to: What would you have played? pawn, but the black bishop may be trapped at any moment, and after 14... £5 15 gxf5 gxfS 16 Axg7+! xg? 17 Ah6+ Hg8 18 AxfB wyxfB 19 A xd6 fxed 20 Byg5+ Dg6 21 wyd5+ the situation becomes complicated. And so Black strikes from the other side. 14..d5! 15 Qxe5 Be8 16 fad dxe4 17 dxe4 Now both 17...Wyxd2+ 18 @xd2 Fife8 19 Ge3 and 17...a.xe4 18 Yyxd8 Hfxd8 19 (.xe7 are bad for Black. But on the other hand, the continuation in the game is good. 17...44b6! 18 .xe7 WyxbS 19 9 xf8 Wxb2 20 Axe7 xg? In this position, where he has a quite appreciable material advantage, White resigned. He has no defence. 21 Hd1 is met by 21...4ixc2, mating after 22 Wyf4 He2+ or winning a rook in the variation 22 wyd4+ wxd4 23 Bxd4 Bci+. And after 21 Hel Hd8 22 we3 the game is con- cluded in highly spectacular fashion by 22...8yxc2! No.6 (p.31) Kapengut-Kupreichik Minsk 1969 Not altogether. White found a way of keeping the black king in the centre, even at the cost of weakening his pawn formation. 15 |d6+! exd6 16 Hel+ Of8 17 gxf3 £6 17.,.Ag8 18 {e3 and 19 cxd6 is even worse for Black. 18 Be6 dxe5 19 Qxc5+ After 19 Axf6+ @g7 20 Ad6+ 4)xd4 the attack on the king comes to 169 an end, and for the queen Black ob- tains quite sufficient compensation. 19.27 20 Hd6 we8 21 Bd7+ @h6 22 Wyd6 (22 HbI came into con sideration) 22...WyeS 23 ¢3+ @hS? 23...85 offers more chances, al- though after 24 Wyd2 Black has to reckon with the threat of 25 f4. 24 Wadl! Web 25 Hd6 We7 26 f4+ @h6 27 (5+ g5 28 h4 Aad8 29 hxg5+ Bg7 30 exi6+ wxf6 31 eygd+ Black resigns, No.7 (p.41) Veresov-Osnos Grozny 1969 By breaking through in the centre. This is the only way for White to bring his reserves into play, since otherwise he has no way of intensifying the pres- sure on the kingside. 17 eS! dxeS 18 4\e4 Not so much attacking the f6 square (the exchange of queens will reduce Black's problems), as threatening the dagger-blow 19 d6! 18.807 Now on 19 d6 Black has a retreat square for his bishop at d8, and he is not especially afraid of 19 4)xf6+. But... 19 tyxf6! The defensive lines are breached in a different way. Simplification for Black is unacceptable: after 19...4.xf6 20 2)xf6+ Gh8 21 Dxd7 Axd7 the invasion 22 Bf7 leads to the loss of a Piece. 19....d8 20 Hf2 eye7 21 wyf3 Hd7 Other moves too would not have avoided loss of material. For example,170 21..He8 22 Ses wWeT (22.87 23 Sixd8 and 24 4f6+) 23 d6, and 23.., ¥¥c6 is completely bad on account of Mate in three moves, 22 Ags wxgs 23 AxgsS Axed 24 d6 e4 25 Wi7+ Bh8 26 wWxd7, and White won easily. No.8 (p.41) Capablanca-Alekhine New York 1924 He has to open files, otherwise Black will create counterplay. But 18 exb6 axb6 immediately makes appar- ent White's pawn weaknesses on the queenside, and so he prepared a dif- ferent manoeuvre. 18 Bfdi! bxeS Castling is dangerous (19 a4), and it is not possible to find anything else. 19 d5! Wd6 The only move, since Black loses quickly after 19...exd5 20 Syxd5 ¥d6 21 abl with the threat of 22 Ab6. 20 dxe6 wyxe6 21 WxeS ¥yh6 22 Wf2!, and the five black pawn islands were a more than sufficient basis for a win. White easily won a pawn and... failed to realise his advantage, through his own fault. But this bears no rela- tion to our theme. No.9 (p.41) Gipslis-Borer Berlin 1991 First of all White opened a path for his queen to the kingside. 22.45! Gd7 It is clear that the pawn cannot be captured (22...exd5?? 23 He8+), while Attack with Mikhail Tat matic in such positions: 23 Bxh‘ ons a Bt, with numerous threal xhS!! gxhS 24 wi6 eyxdS Fonds in view of 24...e8 25 Hed. Bxe8+, and hoping for 25 HeS Qes HixdS &xd5, when there is still sot play in the position. Even here White: should win — 27 Se5 followed by eyg5xh5, but he wants more. bxgd 30 hS! Another reserve joins the attack, and: there are hardly any defenders. 30.48 31 Hxgd+ B18 32 WeT+ e7 33 Ais was? A tougher defence was offered by 33.906 34 Bxf?+ Bd6 35 wddt dS 36 Yyb4+ Hc6 37 wxb7+ HAs, when although White continues tacking with material now equal, there: is no forced win. Now, however, game concludes quickly. 34 Wi+ GIB 34...d6 35 Bd44+) 35 h6 Hd8 36 h7 Black resigns. No.10 (p.50) Troinov-Popoy USSR 1962 It gives him not only the right, also the obligation — to attack. White rightly “launches" his queen into vicinity of the black king. 13 Wh Hes Black intends to parry the attack on h? (&f3-b3) by transferring his knight to £8. But the move played creates other weaknesses, and White now strikes from the other side.Answers to: What would you have played? 171 14 Qxd5!! exdS 15 Wyxf7+! @xi7 The Greek gift has to be accepted (15...@7h8 16 4e6). 16 Axd5+ Gg6 Evidently Black considered the variation 16...f8 17 Se6+ gs 18 @xd8+ Bh8 19 Axc6 bxc6 20 Axc6 Hb8 21 6 Df6 22 Axe8 Axe8 23 f5 to be too cheerless (23..xb2 24 Habl), and he sends his king out into the storm, 17 15+ PhS 18 Gt3+ Bhd 19 23+ @Bh3 20 922+ Ged 21 Beat And, two moves before mate, Black resigned. No.11 (p.50) Geller-Polugayeysky 28th USSR Ch., Moscow 1961 Alas, there does not appear to be any combination breaking up the king's shelter: after all, Black has drawn up nearly all his forces for the defence, id his queen, although temporarity lagging behind", is ready to come into play by taking on d5, although he then has to reckon with £)xg7. But White has a possibility of fur- ther increasing his attacking potential. 27 e3! Hd2 28 2)e4 With gain of tempo White has kept the black queen passive, and, more important, he has increased the of- fensive power of his rook at fl and found a new target for his knight to attack: not only the h6 pawn, but also f6. Now all his pieces are operating at full power. 28.4407 29 hat » There is no point in hurrying with 29 wxh6+, since now this threat is intensified: 30 ¥yxh6+ gxh6 31 S)f6+ @h8 32 Axd7+ Ae7 33 Axg7+ xg7 34 h5, winning a piece. 29..We7 30 He3 Wd7 31 wWyxh6+ Se8 Or 31..gxh6 32 4f6+ Qh8 33 }xd7+ Se5 34 Qc3 He2 35 d4. 32 Hxe8 Mate at g7 and the rook at d2 are both "hanging". Black resigns. No.12 (p.72) Rubinstein-Teichmann Vienna 1908 No, because the opening of lines for attack could have been carried out more quickly, and perhaps more effi- ciently: 18 g6 (nevertheless!) 18...fxg6 19 hxg6 h6 20 4)xe4 dxed 21 4)g5, and the game is over: both 22 @e6+ and 22 ¥yxc4+ are threatened, and 21....d5 is met by the decisive 22 Sixed. No.13 (p.73) Adorjan-Ribli Riga Interzonal 1979 By interposing 18 Q.h6! And since counterplay with 18..xc3 is now clearly too late — 19 gxf7+ &xf7 20 Hxp7+ Be8 21 bxc3 wyxc3 22 wyxc3 4xc3 23 Hxh7, and since the bishop is immune — 18..gxh6 19 ¥yxh6, mating, Black has to defend his g7. 18...4f6 19 gxh7+ xh7 20 Sxg7! Qxg7 21 h6 46 22 wer Black resigns, since against the in- vasion 23 Yyg7+ (23...xg7 24 hxg7+ @e8 25 Ahk mate) there is no de- fence.172 No.14 (p.73) Kochiev-F.Wade Moscow 1983 White used his occupation of the light squares to invade the opponent's position. 29 WE |f7? 30 |fe3 Ae7 31 WiS+ Qg7 32 we6 Audi 33 Hxdl Hd8 34 Bxd8 Wxd8 35 @h2 Contrary to the usual state of affairs, the exchanges have not eased Black's position. White's queen is tying its black opponent to the defence of the bishop, and on the light squares the reserves will soon be brought up. 35...818 36 £5 we7 37 ce3 hS 38 &)xe7 tyxe7 39 Wyc8+ 2d8 40 O65 Wes 41 wye7 Mate can be avoided only at the cost of the queenside pawns. Black re- signs. No.15 (p.73) Levenfish-Eliskases Moscow 1936 In fact Black nevertheless played 20...n5!, since if 21 gxhS he has 21... g5! (there is no need to resort to the sharp 21...@h8) followed by 22...8h8 and 23,,..xh5, with a very strong attack down the open h-file. The situation is similar after 21 h3 hxg4 22 hxg4 Hh8, and so White's reply is practically forced. 21 h3 Bhs 22 eyg2 Ah4 23 |bs Axel 24 Hfxel a6 25 gS Necessary, alas, otherwise after 25... hxg4 26 Axed Axed 27 wxed Hh4 White loses slowly, but surely, 25..Axh3 26 eyxh3 Wxgs+ 27 Attack with Mikhail Tal Bhi we7 28 Hgl Ph6 29 c5!? A desperate attempt to break the ut happy course of events. Otherwi , -Baf8, the advance of the ki Especially since the only alternati was the passive 29...Bae8, since 29. White the initiative — 30 cxd6 cx 31 Bye6! 30 xc7 tyxc7 31 Hxg6+ @xg6 WiS+ Ph6 33 wWl6+ Bh7 34 wis: Sez8t In giving up his queen, Black minds his opponent that one piece. its own cannot do much. 35 Hgl+ We7 36 Wi Bh7 Wxd6 2d3! 38 wye6+ Ph8 39 Hx; xg? 40 Wh6+ He8 41 h4 Hel! | Avoiding the last pitfall: 41...Bal intending mate after 42 wWxh5 Hag6, refuted by 42 d6! 42 Ye6+ @h7 43 wh3 : Tt turns out that 43 wxes isn possible on account of 43...Hgi+! 43...ag8 White resigns. No.16 (p.88) Stolberg-Zak USSR 1938 By the most ordinary interferen After 28 Hd7! Black could, of cout have prolonged his resistance in hopeless position — 28...26 29 Hi: gxhS 30 4)xf7, but he "preferred’ allow a smothered rate in the mai variation 28....xd7 29 wxf7+ @ 30 dcd 496 31 we8+ Hxg8 32 2 mate.Answers to: What would you have played? No.17 (p.88) Tal-Platonov Dubna 1973 It fact Black's resignation was very timely! At the last moment he noticed that in reply to his planned 23...8%xg3 there would follow 24 4.26!!, when the rook's retreat is cut off, and mate is inevitable. No.18 (p.88) Polugayevsky-Osnos 36th USSR Ch., Alma-Ata 1969 ‘The last straw for Black was 21 e5! This unusual interference move severs Black's connections on the b8-h2 di- agonal, and his position collapses. 21,..td2 22 a3 HES 23 Axes! g5 If 23...fxe6 24 Wxe6+ Bh8 25 g3, and White's material advantage is enough for a win. 24 g3 fxe6 No better is 24...W¥e2 25 Hxf4! exf4 26 WeS+ Bh 27 who! Hg8 (27... Heg8 28 ALS) 28 wf6+ He? 29 Axi? Wydl+ 30 a2, when the e-pawn cannot be stopped. 25 Wyxe6+ @g7 26 gxfa wyg2 Equally hopeless is 26...Qxf4 27 Baxf4 wyxfd 28 Wyd7+ eg6 29 wxb5 Wxh2 30 wd3+ we7 31 we. 27 Hdl exf4 28 wd7+ HET 29 e6 Black resigns, No.19 (p.88) Fridman-Tornblom Stockholm 1974 A move away from victory, this would have been a strange decision! 173 By 1...He3! Black blocked the e2 pawn, If 2 fxe3, then 2...2)h3 is now decisive, and at the same time 2...xf3 is threatened. White resigned. No.20 (p.104) Tal-Andersson Match, Stockholm 1976 Yes, and relying on this outpost (the {6 square is under control) White be- gan a direct attack on the king. 19 Hg3 wWxc2 20 te3 (with the threat of 21 ¥yh6) 20..We4 21 b3 Wad5 22 6+ Axf6 23 exf6 The white outpost has moved to the 6th rank, and although it is insuffi- ciently defended, the attack has been strengthened, 23.26 24 Hed @h8 25 Gb! He8 Taking the unprotected bishop — 25,..)xb6 — diverts Black's attention from the kingside, and he is mated or loses his queen: 26 ¥yh6 Re8 27 Bh4. 26 wh6 Hes 27 Had Here 27 Kh4 is partied by 27...2f8, when Black himself launches a coun- terattack. 27 )XD6 If 27...%b7 White achieves his aim — 28 Yyxh7+, mating. 28 BxdS }xd5 29 83 He3 30 Hxc3 Hxc3 31 we3 b4 32 wa7 HES 33 WeS Hb8 34 tyd6, and in view of the unavoidable loss of his queenside pawns, Black resigned. No.21 (p.104) Kask-Kamenetsky USSR Corr. Team Ch., 1974-5 Firstly, White's main forces are174 concentrated on the queenside, and the black outpost at e4 greatly restricts their possible rapid regrouping. Sec- ondly, there is a classic target for attack — the advanced pawn at h3. And thirdly, also classic is the deployment of Black's queen and bishops, which are aimed at the kingside, In short, there are more than sufficient grounds for the decision. 23...Qxh3! 24 gxh3 ¥yxh3 25 eS Hd6 (renewing the threat of mate in a few moves) 26 Ag4 Hg6 27 eyxg6 Wxed+ 28 Ol fxg6 29 wyed+ Ph7 30 a2 ‘The threat of 30,..8xf2+ 31 @xf2 Ag3+ with a quick mate has been parried, but the pursuit of the king continues — incidentally, with level material. 30...h5 31 Yel h4 32 whl h3 33 Hd2 Qe3 34 He3 AES 35 Wel wes 36 Raz Compared with the initial diagram, the queen has "castled very long", but has not gained any peace. By destroy- ing the pawn chain around the king, Black drives it back. To its doom. 36...Q,xf2! 37 Bxf2 Bxe3+ 38 Gd2 Even in a non-correspondence game it would not be difficult to calculate the forcing variation 38 @b2 Bb3+ 39 Sc2 (or 39 Gal Bxad+ 40 Hb! Wyb3+ 41 Rb2 wad, mating) 39... Wo4+ 40 @dl texd4+ 41 Hel We3+ 42 Odi Hd3+ 43 Bc2 Ac3+ 44 bi Bxcl+, leading to a queen ending where Black is five(!!) pawns up. 38..uyb3 39 WEL Hd34 40 Sel Wxa3 41 Acc2 Walt 42 Ge2 wyxd4 43 Wel Hg3 44 Hd2 We3+ 45 Gf wes Attack with Mikhail Tal There are too many threats, an White resigned. No.22 (p.104) Menchik-Thomas London 1932 At any event, the first Women’ World Champion concluded matters force. 20 f6+ @hs After 20...@xf6 White gains tempi required for the creation of. mating net: 21 Weg5+ Sg7 22 h6 Begs 23 wis. 21 Wh6 He8 22 hxg6 fxg6 Yyxh7+! And Black resigned, since he mated by the opponent's mini forces: rook and £6 pawn. No.23 (p.115) Botvinnik-Chekhover Moscow 1935 This position was reached after. prolonged sacrificial attack. Whit concluded by eliminating the mail defender: 32 Exf6! Qxf6 33 Yyh7: BiB 34 Hel! Qe5 (what else?) Wh8+ Be7 36 wWxel+ db WxeS+ Ba7, and White, still a rool down, announced mate in six mov 38 WfS+ Geb 39 d5+ He5 40 a3: xed 41 tyed+ c3 42 Ab4+ PI 43 Yyb1 mate. No.24 (p.115) Tal-Najdorf Leipzig Olympiad 1960 White has assembled a_ strilAnswers to: What would you have played? force of pieces on the kingside, But everything turns on the position of Black's knight: it is not attacking, but is primarily defending the approaches to his king, in particular f6. Therefore: 21 Bxed! dxed 22 Of6! wyb6 The “slightest movement" of the ¢7 pawn would have allowed mate in one or two Moves. 23 Qxp7 Hie8 24 Be5 wWe6 25 |\h6+ GES 26 £5! Biack resigns: 26...u¥g5 is met by 27 &xf7!, while if 26..%%c6 White wins by either 27 ¥yhS ¥d7 28 Ad6+!, or 27 wyg3 wWxh6 28 Ad6+ Be? 29 xe7+ Oxe7 30 we5+. No.25 (p.115} Spassky-Geller Candidates '4-Final, Sukhumi 1968 Alas, the defenders of the black king can be eliminated, and Spassky, who was on his way to the chess throne, did this on nearly every move! 23 Bxf6!! exf6 24 wyh7+ Sf8 25 x7! ‘The £7 pawn was also playing a de- fensive role, and now 25...xf7 26 Mh6 Bg8 27 4yf4 is a very convincing variation, The attackers that have per- ished on the decisive part of the battle~ field are replaced by fresh troops, whereas the defenders (of the g6 pawn, at least) cannot be replaced. Therefore: 25...ExcZ 26 Qh6 ixcl+ It is too late for 26..2xd3 27 Wxg7+ He8 28 Hxc2 Hxc2 29 Af4, when White wins. 27 &xel Oxf? Or 27..Qxh6 28 &xh6 eB 29 4)g8!, and the pursuit of the black king 175 continues. 28 tyxg7+ Be’, and now White could have quickly decided matters by 29 eS Sixe2 30 e6, mating. But the game continuation 29 g5 £5 30 tyxg6+ Qd7 31 Wi7+ G6 32 exfS+ also led toa win, No.26 (p.126) Brink-Claussen v. J.Littlewood Varna Olympiad 1962 He should divert the white queen from the defence of {2 and exploit the weakness of the back rank. 22..A9xd6! 23 wyxd6 Add+ 24 Shi H+ White resigns: he is either mated, or is left a piece down: 25 @gl Sed+. To be fair, it should be mentioned that the less spectacular 22...(ie5 would also have won. No.27 (p.127} Makogonov-Smyslov 12th USSR Ch., Moscow 1940 This means that the queen must be driven away, exploiting the barely perceptible weakness of the back rank. 24 HeS!! we8 (the rook is clearly immune) 25 43! One blow after another. If now 25... ¥Wyb8, then 26 Hg5 g6 27 Hb5, and there is no defence against the threat of Ah6 (27..4c5 28 Gh6 es 29 Wxe5, and again the back rank is Black’s downfall). At the same time the immediate 25 Bg5 would not have achieved its aim on account of 25...2c3. So Black decides on a sacrifice.176 25.08 26 AxcS Axdi+ 27 Bez Bxc8 Alas, the attack continues. 28 Kes g6 29 RbS (the way for the bishop to h6 is opened) 29..d6 30 Sxb3 4)xb3 31 eyxb3 Black resigns. No.28 (p.127) Bouaziz-Miles Riga interzonal 1979 The position is now drawn, and after 45...9xh3! White should have rec- onciled himself to this, by playing 46 Wl! Bg3+ 47 B2 Axf3+ 48 Sxf3 Wxfi+s, with perpetual check after 49 ed, But through inertia he chose 46 $xh3??, overlooking the fearful blow 46..e9h1+ 47 Wyh2 wyxf3+ 48 Sxhd Be7+ 49 95 Qxes+t Only here did White realise that he would be prettily mated: 50 @xg5 f6+ 51 @h4 gS mate, or 51 es wet mate. No.29 (p.127) Stein-Sokolsky Odessa 1960 Even without his queen White was able to create decisive threats against the king. 27 Fig7+! Qxg7 28 Bxg7+ Bhs 29 Bev! White's material advantage becomes. decisive. He threatens 30 {g7+ @g8 31 46+, and the black knight is attacked. A few moves later Black resigned (if 29..e¥e5 30 Qf4 wf5 31 Axc6 ete.), Attack with Mikhail Tat No.30 (p.165) Mikhalchishin-Yudasin Nikolayev 1983 Everything depends on whether’ White can break up the pawn screen around the black king. 3 20 Qh6! Bxdl+ 21 Yxdi (other wise the e5 outpost falls) 21...8f8 2: 3 Ha7 The attempt to exchange bishops by. 22...4.d5 does not succeed: 23 .c2,. aims by pressure on the eS pawn to ti down the opponent's forces. ¥ 23 Bel aS 24 Qc2 ad 25 Hed! PhS (26 Hg3 was threatened) 26 Ags {maintaining in certain variations th possibility of Of6) 26..Ha5 27 Bl h6 28 Wyfd 2)xe5? This leaves £6 undefended and loses. by force. White has more problem: after 28..Axe5 29 Qxh6 gxhé 3 Bxh6+ Sg8, when Black's defensive resources are by no means exhausted, 29 Qxh6! gxh6 30 Hxh6+ gs 3 &h7+ Gh8, and without waiting fc White's reply, Black resigned: 32. Wi6+ We7 33 Yyd8+ leads to mate. No.3t (p.165) Tal-Plaskett Sochi 1984 But White can exploit the target created to begin an attack, 14 Axh6 AxdS 15 exdS Ab4 16: Wee We? Has the first wave been repulse: and can Black now pick up the d5 pawn? .Answers to: What would you have played? 17 a3! White even urges him to! After 17... Qbxd5 18 Qxg7 wyxg7 19 Qxd5+ @xd5 20 We6+ White is a pawn up and the black king is "bare" . 17...A)c2 18 Sixg7 Wyxg7 19 Yyxc2, and Black could have spared himself the following five moves. No.32 (p.165) Tal-Gedevanishvili Georgian Championship 1970 Of course, White's concentration of force against the black king is not the maximum, but it is evidently quite sufficient. Queen, rook, bishop and knight — part of them can be sac- tificed to break up Black's defensive lines. 18 4)f6+! gxf6 19 4 xh7+ Ph No better is 19...xh7 20 exf6 Ag6 21 whS+ @g8 22 Hbfl(gl) or 22 Hffl, and against the mate at g7 there is no defence. 20 Hh4 g7 21 wel Ags 22 4x98 Black resigns. No.33 (p. 166) Tal-Taimanoy Blitz Tournament, Leningrad 1977 Were the black pawn still at h7, there would be a long and hard strug- gle in prospect, although White has every chance (pawn majority with an outpost at e5) of an attack on the kingside. But the pawn has "stuck out", and this leads to the rapid collapse of Black's position. 29 Axh6! gxh6 30 Yxh6 Yye6 31 We5+ Dee 177 Or 31...@h8 32 f4, and the way for the rooks to h3 is open. 32 £4 Web 33 £5 We7 34 £6 18 35 Sixg6 fxg6 36 Wxe6+ @h8 37 RFA Black resigns. No.34 (p.166) Tal-Miles Porz 1981/2 But hardly any more, since the in- sufficiently defended £7 pawn may also be attacked through a blow at another Jink of the pawn chain, a device that is as typical as it is simple. 15 Qxh7+! &xh7 After 15...@xh7 the defender of the 3 square is diverted by 16 Hxd7, and if 16...2)xd7 17 WyhS+ Begs 18 Wxf7+ @h7 19 4h5 with a quick mate. Therefore Black prefers immedi- ately to remain a pawn down, which, however, is equivalent to capitulation. 16 Hxd7 26 Tf 16...f6 17 Sg6 Yc6 White wins by 18 4xb7!, while after 16...)f6 he does not content himself with an extra pawn, but continues his decisive at- tack: 17 4\n5! xd7 18 Wyg4 g5 (18... Sf8 19 4)xd7 and 20 4)xf8, or 18.26 19 &)xg6) 19 4)xd7 Wyd8 20 Adl! 17b4 White could have simply set about realising his material advantage, but instead he makes use of a chance pos- sibility: 17..2yxb4 diverts the queen from the defence of e6 and allows 18 xg6, while 17...cxb4 clears the way for the white c-pawn. 17...6.¢8 18 bxeS Wxe5 19 Hed Wh6 20 Wis .D. 20 43 with the idea of 21 Hg3178 is more natural, but the move in the game is prettier: on 20...xd7 White had prepared 21 .e3!, while if 20...f5 21 We3. 20...8yb2 21 Q\xf7 None of White's pieces can be taken, 21.8427 22 2h6+ @h8 23 He7 Winning the queen in the variation 23 fed Qxd7 24 &xp7+ Sxg7 does not satisfy White, 23.-Hf8 24 Hxe7 Black resigned in view of the dis- mal choice between 24...uyxe7 25 Se5+, and 24.. .Axf4 25 Axg7 Axf3 26 Best. No.35 (p.166) Drike-Krantz Correspondence 1989-91 Only at first sight. In fact White's defences collapsed instantly after 17...4yxh2!, and in view of the unat- Attack with Mikhail Tal tractive choice between 18 xh’ Axg3+ 19 fxg3 Wxg3+ 20 Oh Hh8+, and (8 wyh6 wyxh6 19 Qxh Shed 20 Qd2 4)xf2!, White re signed. No.36 (p.166) Tal-Chepukaitis Blitz Tournament, Leningrad 1977 There are obvious possibilities of sacrifices on h7 and g6. 23 |xh7! Axf6 24 4)xf6+ lB 25 Wxg6 WE7 26 wes With the intention of winning th queen after 27 4\h7+ He 28 O26. 26.0727 27 Syh7+ BLT 28 was - Threatening “only” 29 yg5+. 28..%9h6 29 Des+ SeT 30 we7+ 28 31 Ah7+ Black resigns, since White has an unusually pleasant choice between winning the queen and mating in a few: moves,Postscript THIS BOOK did not turn out quite as we had planned. Unfortunately, there is no chapter about the offensive on the queenside: Mikhail Tal had some interesting thoughts about this, and regarded activity on this part of the board both as an independent opera- tion, valuable in its own right, and as a diverting, as it wete secondary opera- tion, undertaken in order at a conven- ient moment to nevertheless assail the place where the enemy King was sheltered. Of this planned chapter there remained only fragments, and two or three of the other sections could per- haps have been a little fuller. But we did not have time. One of the greatest of chess magicians, the former World Champion Mikhail Tal, has passed away. Right to his last day he maintained his interest in chess, and, of course, not nly chess: from a Moscow hospital he travelled to a tournament in Spain, he went out for a day to play in a blitz-tournament, and he analysed and made recordings of the material for this book. To profound regret, all this was to be his last. This is why the title planned earlier Everything about Attack has been changed. Tal was a wonderful master of attack, and in his career of more than a third of a century in top-level chess, what pleased him most was not a point in the tournament table, and not a victory, but a successful attack. And { should like once again to repeat the words of Mikhail Tal, that appear in the introduction to this book: may the readers win their games, by attacking, Attacking brilliantly. It was this that always made Tal happy. lakov DamskyIndex of Players Adorjan 12, 171 Alekhine 138, 151, 170 Andersson 81, 173 Andreyeva 51 Archakova 51 Auzinya 98 Bagirov 142 Bakinskaite 94 Balanel 122 Balashov 48 Bareza 90, 168 Bauer £36 Belyavsky 56 Benko 85 Beylin 83 Bilek 7 Bédk 128 Borer 170 Botterill 62 Botvinnik 9, 13, 79, 174 Bouaziz 176 Brinck-Claussen 136, 175 Bronstein 78, 120, 144, 150, 154 Butnoryus 93 Byme 148 Capablanca 170 Castaldi 129 Chekhov 99 Chekhover 174 Chepukaitis 178 Chernikov 53 Chemin 33 Chiburdanidze 17 Damsky 63 Ditman 122 Donner 96, 131 Drewitt 138 Drike 178 Durasevic 109 Dvoiris 17 Dzentelovsky 133 Eliskases £72 Euwe 9, 129 Feuerstein 105 Fischer 85, 105, 118 Fridman 173 Gaprindashvili 55 Garcia, G. 92 Gaudin 138 Gedevanishvili_ 177 Geller 4, 13,37, 42, 44, 55, 66, 76, 107, 111, 121, 150, 171, 175 Gereben 107 Gheorghiu 35 Gik 60 Gilfer 128 Gipslis 170 Gligoric 66, 95, 99 Gomez-Baillo 168 Grigorian 3 Gromek 15 Gruzdev 135 Gufeld 9, 120 Gurgenidze 112 Gurieli 111 Hansen, C. 32 Hartston 140 Hebert 65 Heemsoth 60 Heisenbutter 60Index of Players 181 Hiibner 121 llivitsky 85 Ivanchuk 152 Ivkov 20, 93 Kagan 20 Kaila 98 Kalinkin 64 Kamenetsky 173 Kamyshov 2 Kapengut 169 Kaplan 12 Karasev 76 Karpov 19, 60, 121 Kask 173 Kavalek 61 Kasparov 43, 51, 80, 81, 83, 93, 116, 122, 140 Kengis 45, 120 Keres 4, 13, 18, 43, 82, 91, 92, 108, 152, 154 Kholmov 18, 89 Kinzel 58 Kivi 98 Klovans 78 Kochiev 172 Kolarov 93 Kondevsky 45 Korehnoi 42, 51 Kosten 109 Kostina 56 Kotkov 54 Kotov 43 Krantz 178 Kupreichik 3, 56, 169 Kurajica 92 Kuamin 45, 139 Landau 81 Larsen 1, 10, 61, 74, 97 Lasker 136 Leonoy 84 Levenfish 98, 172 Levitsky 119 Liebert 63 Lipnitsky 83 Lisitsin 52 Littlewood, J. 175 Livingston 118 Lutikov 83 Makogonov 175 Makarychev 89 Malevinsky 134 Marjanovic 43 Marshall 119 Menchik 174 Mikenas 120 Mikhalchishin 176 Miles 33, 58, 176, 177 Minasian 58 Mititelu 57 Morozov 155 Morphy, BE. 16 Morphy, P. 16 Moskvin 153 Murey 130 Najdort 4, 99, 174 Nezhmetdinoy 2, 5, 10, 14, 15, 53, 54, 64, 108 Nikhamkin 135 Nikitin 147 Nikolic 140 Nimzowitsch 137 NN. 119, 134 Oelly 37 Olafsson, F. 117, 131 Oskam 138 Osnos 169, 173 Pabtz 99182 Attack with Mikhail Tal Palmiotti 144 Panchenko 78 Panno 4,90 Panov 6 Paoli 10 Persig 148 Petrosian, A. 78 Petrosian, T. 110, 147 Petrov 91 Pilnik 4 Plaskett 176 Platonov 173 Poldauf 130 Polugayevsky 32, 34, 67, 171, 173 Popov 134, 170 Portisch 150 Pribyl 80 Radulescu 12 Ragozin 52,79 Rakhmanov 152 Rantanen 131 Ravinsky 6, 85 Reshevsky 117, 118 Ribli 35, 116, 171 Richardson 140 Romanov 108 Rossolimo 118, 119 Rubinstein 171 Ryumin 98 Saigin 109 Salov 129, 132, 139 Sammui 56 Sanz 109 Sax 65, 129 Schmid 58, 82 Schneider 151 Sergievsky 112 Shabaloy 149 Shirov 47 Shishkin 76 Simagin 78, 167 Sliwa 146 Smirin 11, 149 Smyslov 11,175 Sokolov, A. 68, 139, 143 Sokolsky 176 Spassky 4, 10,37, 96, 110, £11, 13: 175 Speelman 90, 133 Spielmann 81 Stahlberg 146 Stangi 47 Stefanov 63 Stein 44, 48, 61, 176 Stolberg 172 Suba 94 Suetin 5, 106, 142 Sunye 122 Sveshnikoy 3, 139, 143 Szabo 35, 108 Taimanov 177 Tal 1,3, 7, 8, 9, 15, 34, 46, 57, 62, 63, 74, 84, 90, 92, 94, 95, 97, 106, 109, 117, 131, 133, 134, 135, 136, 147, 148, 151, 152, 167, 168, 173, 174, 176, 177, 178 : Tarjan 12 Tarrasch 137 Taulbut 132 Teichmann [71 Thomas 174 Timoshchenko 106 Tolush 36,78 Tombiom 173 Troinov 170 Tseshkovsky 67 Tukmakoy 61, 121, 167 Turnhuber 148 Ubilava 106 Uhlmann 59Uitumen 59 Unzicker 117 van Seters 35 Varanin 153 Vasyukov 46, 55, 109 Velimirovic 76 Vereczi 55 Veresov 15, 133, 169 Vidmar 13 Voitsekh 155 Vooremaa 8 Voronova 94 Index of Players Wade, F. 172 Weenink 151 Wu Ming Qien 111 Yudasin 176 Yuferov 147 Yusupov 19, 45, 68 Zaid 167 Zaitsev, A. 14 Zak 172 Zamikhovsky 150 Ziedinya 98 Illustrative Games Tal-Portisch, 1965 Tal-Larsen, 1965 Tal-Hecht, 1962 Spassky-Tal, 1973 Tal-Averbakh, 1961 Tal-Mohrlok, 1962 Torbergsson-Tal, 1964 Tai-Donner, 1968 21 24 27 38 49 68 69 7 Aronson-Tal, 1957 Polugayevsky-Tal, 1979 Tal-Dvoretsky, 1972 Spassky-Tal, 1958 Tal-Gurgenidze, 1969 Spassky-Tal, 1979 Tal-Stean, 1975 Tal-Portisch, 1965 183 85 100 112 123 156 157 159 161$18.95£12.99 ATTACK WITH MIKHAIL TAL MIKHAIL TAL & IAKOV DAMSKY The legendary Mikhail Tal was the finest ever exponent of attacking chess. His all-out sacrificial style took him all the way to the World Championship title. Dogged by ill-health, Tal’s reign was a short one, and he was never able to regain his crown, but Tal remained an outstanding player right up to his death in 1992. His departure has taken away one of the chess world’s most charismatic figures; a player who excited chess passions wherever he went. In this, his last book, written together with Russian sports journalist lakov Damsky, Tal shows how to unleash devastating attacks in all kinds of positions. ISBN 1-85744-043-9