In choosing the pawn avalanche as a means of attacking the castled king at an early stage of the game, White has undertaken a considerable obligation to carry out his attack incisively to the end. For if Black has his say, his counterattack will find a mass of weak points, files, and diagonals in White’s territory; if an endgame is reached, White’s advanced pawns may easily become ‘cannon-fodder’.
The moment of undertaking an obligation is also the moment of crisis, the moment for the thrust in the centre, which in the Sicilian as a whole, and in this position particularly, is unstable. Botvinnik perceived the correct moment and struck with 10 ... d5!.
10 ... d5!
Let us first see why ... d5 is correct at the precise moment when White plays g4. Let us suppose he has played 10 0-0 and examine the difference.
1) 10 0-0 d5? 11 e5 Ne8 (11 ... d4 does not work, since White can simply take the pawn) 12 Bf3 Nc7 13 Nc5, when Black must accept the weakening of his position by Nxe6 fxe6, since 13 ... d4 is defeated by 14 Bxc6.
2) 10 g4 d5! 11 e5 d4! (now this move is good) 12 Nxd4 (or 12 exf6 Bxf6, when Black wins the piece back with the better game) 12 ... Nxd4 13 Bxd4 Nxg4! 14 Bxg4 Bxg4, and Black preserves the material balance while gaining the advantage with regard to space, position, and time. There is undoubtedly a great difference between the first and second instances of ... d5.
Now the question arises why ... d5! is necessary in answer to 10 g4 and what happens if Black does not employ the defensive central thrust. This can be answered by citing the continuation of the game Foltys-Eliskases, Podebrad 1936, where in the same position Black met 10 g4 with 10 ... Na5 aiming at the square c4. The continuation was 11 g5 Ne8 12 Bd4 (in the game Kan-Botvinnik, Moscow, 1936, 12 Qd2 was played, but Foltys’ move is better) 12 ... Rc8 13 h4 Nc4 14 Bxc4 Rxc4 15 Qd3 Rc8 16 0-0-0 Qd7 17 Rd2 Bg4 (to combat h5, but it surrenders d5 to the enemy knight) 18 Nd5 b6 19 f5! (the avalanche begins to move; Black cannot take the pawn, since the eventual Nxe7+ is decisive) 19 ... e6 20 Bxg7 Kxg7 21 f6+ Kh8 22 Ne7 Rd8 23 Nd4 Qa4 24 Kb1 Nc7 25 h5 gxh5 26 Rdh2 Rd7 27 e5 Rxe7 (threatening 28 Qxh7+ Kxh7 29 Rxh5+ mating) 28 Rxh5 Bxh5 29 Rxh5 1-0.
Now to return to the Nottingham game, in which Botvinnik did not allow himself to be smothered in this way.
11 f5
Clearly, this is the only alternative to 11 e5 which has already been analysed. If 11 exd5 Nxd5 12 Nxd5, Black would gladly exchange queens by 12 ... Qxd5, thus winning the b-pawn, to say nothing of the ending.
11 ... Bc8
Though the bishop has to withdraw, it is still aggressively placed as far as White’s weak pawns at f5 and g4 are concerned.
12 exd5 Nb4
There is to be no more talk of an attack on the king; the players are now concerned with a struggle for the centre pawn and the d5-square. If White holds out, he can expect success, but if Black overcomes the advancing pawn, the first player’s prospects are poor.
13 d6?
Whenever Alekhine made an unexpected move the world of commentators dutifully gave it one or two exclamation marks. This happened in the case of this move too, which abandons the struggle for the square d5 and, in fact, also risks losing the game.
A better line here is 13 Bf3! gxf5 14 a3 fxg4 15 Bg2. Suggested long ago by Dr Euwe, this was tried in the 2nd match game between Fischer and Reshevsky, New York 1961, when White secured an advantage after 15 ... Na6 16 Qd3 e6 17 0-0-0 Nxd5 18 h3 g3 19 Rhg1 Qd6 20 Bxd5 exd5 21 Nxd5 Kh8 22 Bf4 Qg6. Now Fischer played 23 Qd2, but 23 Qf3! seems to me to be even stronger.
13 ... Qxd6?
Alekhine’s original assertion that 13 ... exd6! was impossible because of 14 a3 Nc6 15 g5 Ne8? 16 f6 f6 etc., has remained unchallenged in chess literature. It is clear, however, that in this kind of situation it is worth looking to see whether the sacrifice of a piece will work or not Indeed, sound logic argues that it must work, for the whole variation quoted by Alekhine so strengthens Black’s position that it is worth his giving up a knight for two pawns. Above all, White’s position is exposed as a result of his pawn advances: the thrust d6 leaves him without a ‘backbone’ in the centre; the reply ... exd6 opens up the e-file and the diagonal d8-h4, while g5 relinquishes control of the c8-h3 diagonal to Black’s c8-bishop. The sacrifice can take a number of forms, but I shall give the one which seems best to me. It runs 13 ... exd6! 14 a3 Nc6 15 g5 Re8! with the following possibilities:
1) 16 gxf6 Rxe3 17 fxg7 (or 17 Nd5 Re5 18 fxg7 Qh4+ 19 Kf1 Bxf5! and Black has a strong attack) 17 ... Qh4+ 18 Kd2 (or 18 Kf1 Bxf5) 18 ... Qh6 19 Ke1 and Black has excellent chances. He has no need to consent to a draw by 19 ... Qh4+ 20 Kd2 Qh6, but can play for a win with 19 ... Bxf5.
2) 16 Bd4 Nxd4 17 Nxd4 Ng4 18 f6 Bxf6 19 gxf6 Qxf6 20 Rf1 Qh4+ 21 Kd2 Ne3 22 Nf3 Qh6 and Black wins.
3) 16 Bf2 Ne4 17 Nxe4 (or 17 f6 Nxc3 18 bxc3 Bg4 19 Nc1 Qa5 20 Bd4 Nxd4 21 Qxd4 Qxg5 22 fxg7 Bxe2 23 Nxe2 Qg2 and Black wins) 17 ... Rxe4 18 f6 Bg4 19 Nc1 Qa5+ 20 Qd2 Qe5 21 Kf1 Qxb2 22 Bxg4 Rxg4 23 Ra2 Qb5+ 24 Ne2 (alternatives are 24 Nd3 Qd5 and 24 Qd3 Qxg5) 24 ... Re4 25 fxg7 Rxe2 26 Qxe2 Qb1+ with advantage to Black.
Botvinnik must certainly have prepared the move 10 ... d5! in his private analysis and so should have known all the important continuations. He had probably not reckoned with 13 d6?, and when Alekhine confronted him with it, did not venture on 13 ... exd6 but was bluffed into playing the drawing move 13 ... Qxd6?. All great masters, from Homer onwards, have engaged in a little bluff at times.
14 Bc5 Qf4!
14 ... Qxd1+ would have been weaker, e.g. 15 Rxd1 Nc6 (not 15 ... Nxc2+ 16 Kd2!) and then White really can play 16 g5 Nd7 17 f6.
15 Rf1 Qxh2
The breakthrough in the centre has allowed Black’s queen to infiltrate into White’s position, which was weakened by the move g4. Perpetual check is now already in sight.
16 Bxb4 Nxg4! 17 Bxg4
An attempt to defend by 17 Ne4 Qh4+ 18 Nf2? would result in unfortunate consequences, i.e. 18 ... Ne3 19 Qd2 Qxb4!, etc.
17 ... Qg3+ 18 Rf2
Not 18 Kd2 Bh6+!; nor 18 Ke2 Qxg4+ and 19 ... Qxb4.
18 ... Qg1+ 19 Rf1 Qg3+ 20 Rf2 Qg1+ ½-½
This position is from the game Kotov-Keres, Budapest Ct 1950. White’s attack is extremely dangerous, his main threat being Bb1 followed by Qc2. It is Black’s move, and the moment is clearly suitable for a thrust in the centre, where Black has not yet engaged his opponent. Keres decided on 1 ... d5?, but the correct choice was 1 ... d6!, as we shall see later. Let us first follow the course the game actually took.
1 ... d5? 2 Bb1! (he surrenders both the pawn and the square c4 to Black and threatens Qc2) 2 ... g5 (if 2 ... Bxc4, then 3 Qc2 g6 4 Qd2 Kh7 5 Bf6 Nb3 6 Ng5+ Kg8 7 Qf4 Nxa1 8 Qh4 h5 9 Qxh5! followed by mate with the bishop on h7) 3 Qc2 Ng6 4 Nf4! gxh4 5 Nxg6 Re8 6 Nh8! Re7 7 Qh7+ Kf8 8 f4 Nxc4 9 f5 exf5 10 0-0 Bc8 11 Bxf5 Bxf5 12 Rxf5 Ke8 13 Rxf7 Kd7 14 Qf5+ Kc6 15 Qf6+ Kd7 16 e6+ Kc6 (or 16 ... Kd6 17 Rxe7 Qxe7 18 Nf7+, winning the queen) 17 Rxe7 Qxh8 18 Rxc7+ Kb5 (or 18 ... Kxc7 19 Qe7+ Kb8 20 Rf1) 19 Qe7 a5 20 Qd7+ Ka6 21 Rb1 1-0
After 1 ... d5?, what is the outstanding feature of the play? Thanks to this move Black won the pawn on c4, but he neither obtained a proper counterattack nor did he harass White’s ‘captain in the centre’, his pawn on e5. White’s attack flowed on unchecked. So the move 1 ... d5? had only the form of a central thrust and, in essence, did not greatly alter the situation. The true central thrust here, which does in fact parry the attack, is 1 ... d6! – this has White’s e-pawn as its target. Let us look at the variations that arise from this.
1 ... d6!
2 Bb1
White could, admittedly, maintain his wedge in the centre by 2 f4, but Black would profit from this in two ways: f4 is closed to White’s knight, while the continuation 2 ... Qe8 3 Qd1 Bxc4 4 Bb1 Nf5 5 Qc2 Qc6 shows that Black also has a tempo more here, since the white queen has needed two moves to reach c2; the combination of these factors turns things in Black’s favour.
2 ... dxe5 3 dxe5
Now 3 Qc2 can of course be answered by 3 ... f5.
&n
bsp; 3 ... Qe8 4 Qc2 Ng6 5 Nf4
Or 5 0-0 Bxc4 6 Re1 Nb3 7 Ra2 Nd4! 8 cxd4 Bxa2 9 Bxa2 Nxh4, when White is left with neither an attack nor any compensation for the exchange.
5 ... Nxc4
The counterattack in the centre by ... d6 has attained its object: now White cannot sustain both his attack and his pawn on e5. He cannot achieve anything by 6 Nh5 Ncxe5 7 Bf6 because of 7 ... Nd3+ 8 Kd1 Bb5 9 a4 Bxa4 10 Rxa4 Qxa4 and Black wins. If 6 Nxg6 fxg6 7 Bg3, Black plays 7 ... Rd8, after which the threats of ... Rd2 and ... Ne3 guarantee him a material advantage. Should White take the g-pawn, then the exchange of queens leads to a won ending for Black.
The various continuations which result from the thrust 1 ... d6! provide good examples of the following themes:
1) the destruction of the centre when this centre is essential to the attack; in this case Black also gained space for a counterattack via the centre;
2) liquidating the position to bring about a favourable ending on the basis of the defender’s strength in the centre.
Particular attention should be paid to those cases where the central thrust does not lead immediately to a decision but merely serves to maintain the tension in the centre. The next example illustrates the significance of this method of defence.
The diagram below is taken from the game Marshall-Burn, Paris 1900. The opening ran as follows:1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 Bg5 Be7 5 e3 0-0 6 Nf3 b6 (nowadays this is usually preceded by 6 ... h6) 7 Bd3 Bb7 8 cxd5 exd5 9 Bxf6 (the so-called Albin-Marshall Attack in the Queen’s Gambit) 9 ... Bxf6 10 h4
Here Burn timidly played 10 ... g6?; this was followed by 11 h5 Re8 12 hxg6 hxg6 13 Qc2 Bg7? (he should have played 13 ... Kg7 and if 14 0-0-0 then 14 ... Rh8) 14 Bxg6 fxg6 15 Qxg6 Nd7 16 Ng5 Qf6 17 Rh8+ Kxh8 18 Qh7#
In playing 10 ... g6? Burn probably wished to exclude the possibility of the classic bishop sacrifice; 10 ... h6? would not have been correct on account of 11 g4. But is the sacrifice really threatened? And has Black no better move to prevent it than one which weakens his king position? It is an interesting fact here that there is no threat of a sacrifice at all in the original position, yet, if Black plays 10 ... Re8 or 10 ... Nd7, it is already, in practice anyway, rather dangerous, although it would still really be incorrect. Let us now play Black’s most natural move, 10 ... c5!, and see why the sacrifice is incorrect: 11 Bxh7+ Kxh7 12 Ng5+ Kh6! (not 12 ... Kg6 because of 13 Qd3+; nor 12 ... Kg8 on account of 13 Qh5) 13 Ne2 (White’s most incisive move) 13 ... g6 14 Nf4 Bxg5! (otherwise 15 h5 is strong; 14 ... Kg7 fails against 15 Nge6+) 15 hxg5+ Kg7 (if 15 ... Kxg5 then 16 Rh5+ affords good attacking prospects) 16 Rh6 (had Black played ... Nd7 rather than ... c5 on his tenth move, White would now have a chance with 16 Qg4) 16 ... Rh8! (16 ... Qxg5 allows White to get some material back by 17 Rxg6+) 17 Rxg6+ (or 17 Qg4 Rxh6 18 gxh6+ Kh7) 17 ... fxg6 18 Ne6+ Kg8 19 Nxd8 Rh1+ 20 Kd2 Rxd1+ 21 Rxd1 Bc8 22 dxc5 Kf8! 23 c6 Ke8 24 c7 Na6 with advantage to Black.
JN: 16 ... Bc8 may be even stronger. By covering e6, Black eliminates the Rxg6+ trick and can safely take on g5 next move.
Accordingly, there was no threat of a sacrifice and Burn’s 10 ... g6? simply weakened the position without any necessity; this is always a much more dangerous thing than a forced weakening, as Tartakower so finely expressed in his Hypermodern. (For compulsion creates antitheses and hence the possibility of counterplay or compensation, while in the case of a voluntary weakening there is only a naked weak spot without compensation!)
We shall now examine an example of play from the above diagram, where Black correctly maintains the tension in the centre as a means of defence, while avoiding as much as possible any weakening of his king position.
10 ... c5 11 g4
White sees that 11 Qc2 g6 12 h5 can be answered by 12 ... Qe8 protecting g6 and threatening ... cxd4. If he then continues with 13 0-0-0 Black resolves the central tension and starts an attack on the castled king (by 13 ... c4 14 Be2 b5, etc) which is stronger than White’s. 11 Qb1 meets a similar response – and White can no longer castle queenside.
11 ... Re8
Putting pressure on White’s d-pawn; now 12 g5 fails against 12 ... Bxd4.
12 Kf1
The pressure in the centre has made White resort to artificial castling.
12 ... g6
This weakening is necessary and well-founded here, and therefore not an error. Black perceives the prospects that his light-squared bishop has on the long diagonal, and to open this up he needs to maintain the pressure against d4; to achieve this his dark-squared bishop must also remain on its long diagonal, and accordingly the square g7 must be vacated to provide against g5. Black is naturally not afraid of the ‘hanging pawns’ arising from dxc5; indeed he actually wants to induce them. In the presence of the two bishops such pawns are not usually considered a weakness in any case; here there is the additional matter of forcing open the light-squared diagonal.
The move 12 ... Nc6 is worth mentioning too; this also increases the pressure on the square d4, but it allows 13 g5 after which 13 ... cxd4 is insufficient on account of 14 gxf6 dxc3 15 Bxh7+!. The classic bishop sacrifice, incorrect on the eleventh move and replaced by g4, has appeared in a correct form in an altered situation.
JN: Vuković offers no analysis to support his view that the sacrifice is correct in this position. It still looks unsound to me.
13 g5 Bg7 14 h5 Nc6 15 Rh4
White holds his ground in the centre and at the same time prepares for an eventual doubling on the h-file.
15 ... Rc8
Not for the sake of the c-file, but to bring the rook to e7, where it will be equally well placed for attack and defence.
16 Kg2 Nb4
Black gives up the pressure on the centre temporarily in order to drive White’s bishop back to b1, thus making it more difficult for the rook on a1 to go into action; this is important, since without both the rooks it is hard to carry out an attack on Black’s ‘artificially-fianchettoed’ king position.
17 Bb1
If 17 Bb5 Black would be able to play 17 ... Re4! aiming to eliminate the enemy rook, which is very annoying for him at h4. White could not accept the sacrifice, e.g. 18 Nxe4 dxe4 19 Nd2 cxd4!, etc.
17 ... Rc7 18 a3
18 Nb5 Rce7 19 Nxa7? would be incorrect in view of 19 ... Ba6 20 a3 Rxa7 21 axb4 Bf1+.
18 ... Nc6 19 hxg6
White takes this pawn at a moment when 19 ... fxg6 would still be dangerous because of the possibility of 20 Rxh7. If White waits any longer, hxg6 will lose its effect and the answer ... fxg6 will be good, opening up the f-file to Black’s advantage. If 19 dxc5 Black has the powerful reply 19 ... Ne5!.
19 ... hxg6 20 Ne2 Rce7 21 Nf4 cxd4
The tension in the centre served Black well while he was defending. Now he turns to attack with the intention of conquering both the d-pawn and the square on which it stands.
22 exd4
If 22 Nxd4, Black can either play 22 ... Ne5 or 22 ... Nxd4 23 exd4 Re1.
22 ... Re4! 23 Bxe4
If White does not accept the sacrifice, he loses his d-pawn.
23 ... dxe4 24 Nh2
Or 24 Ne5 Nxd4 when Black is attacking the knight on e5 as well.
24 ... Nxd4 25 Kg1 Qxg5+ 26 Rg4 Qf5 27 Qa4 Bc6 28 Qxa7 Nf3+ and Black wins.
11 The phases of the attack on the castled king
The exposition up to this point may have appeared to some readers to have been a series of fragments, in which the various aspects of the attack on the castled king have been closely looked at but which still need to be combined into a whole. However, it was essential first to collect together as much knowledge as possible about the various elements of the attack on the castled king, then to get to know the main principles of defence against this attack, and only after that to begin to bring everything together in this chapter, which is devoted to the temporal aspect or phases of the attack on the castled king. The temporal aspect here is naturally a framework for the points we wish to make, which rest on the following basic theses, already fairly well elaborated in earlie
r chapters.
1) An attack on the castled king cannot be successfully carried out on the basis of the attacker’s wishes or ‘brilliance’ alone, but the essential conditions for it must also be present.
2) These preconditions may be fulfilled to a greater or lesser degree, and the attacker’s game must adapt itself accordingly.
3) The degree and type of these preconditions accurately determines the extent of the commitments which the attacker can undertake without harm; thus there is a logical and reciprocal relationship between preconditions and commitments, and by this the actual phase which the attack has reached can be recognized.
The proposition of these theses, which will be illustrated in later examples, leads us first of all to a brief discussion of preconditions and commitments.
Types of preconditions
Preconditions are naturally contained in the position itself, whether in the attacker’s territory or his opponent’s.
The attacker, for instance, may have his pieces strongly deployed in the vicinity of the enemy king position or aimed in that direction, and it may also be that the general mobility of his pieces or his advantage in space assists him in attacking.
At the same time, there must be some deficiencies in the defender’s camp; the chief of these is the vulnerability of the king position, either through weaknesses in the pawn structure or because pieces are misplaced in relation to the castled position, being overburdened by obligations to cover squares on other parts of the board.
These and similar preconditions should generally be fulfilled at the first or initial phase of the attack, but it may also be that they can be created by force at this stage, e.g. by the provocation of weaknesses.
The second phase of an attack can follow a wide variety of courses. It may be a question of a quick attack leading to mate, usually by means of a sacrifice, and in this case it is naturally unimportant to talk of further preconditions. However, if the attack is slower or does not have an uncompromising character, the second phase is usually concerned with the creation of further conditions as guarantees of success. The following are among the most important operations at this stage: the provocation of further weaknesses; the exclusion of one’s opponent’s counterattack or its suppression; control of the centre; restraint against a central counter-thrust by one’s opponent, and so on.
Art of Attack in Chess Page 25
