D) 1 ... Rh7
Finally, there is also the possibility of the rook retiring from h5 so that the queen cannot take it after ... g3.
2 f4
If 2 g3, we get the following entertaining variation: 2 ... Nf3+ 3 Qxf3 gxf3 4 Bg5 Nxe4 5 Bh4 Nxg3! 6 Bxg3 Rah8 mating. An example of ‘clearing up’ on the h-file.
2 ... g3 3 Rf3
Or 3 Qe1 Ne2+ 4 Qxe2 Rah8, etc.
3 ... Rah8 4 Rxg3 Rh1+ 5 Kf2 Nxe4#
JN: There is no argument about Vuković’s first three lines, but in variation ‘D’ White can play 2 Bxd4 Rah8 (after 2 ... Bxd4 3 Re1 Rah8 4 Kf1 the king slips out) 3 Qxg4+! Kc7 4 Qh3 Rxh3 5 gxh3 Bxd4 6 Nc3, escaping into an ending in which he is only slightly worse.
In the above diagram doubling rooks was an easy task, but the crux lay in controlling the squares which the king might use for its escape. However, there are cases where the problem lies in actually doubling the rooks on the intended file. As a rule a necessary condition for this is that the attacker’s back rank should be clear of other pieces so that the rooks can be moved at will. When the rooks on a given rank cover each other, they are said to be ‘united’. Naturally, being ‘united’ can also apply to a queen and rook. The continuation from the following diagram is instructive in this context.
The diagram represents a critical stage in the game Krogius-Niemcla, Lovisa 1934.
White is in a precarious position, even though his pawn formation is sound and he has an open h-file.
If he withdraws his bishop to 1 Bc2, Black has the strong reply 1 ... d4, while 1 ... Bxg5 is also worth considering. If 1 Qg4 cxd3 2 Qh4 (2 Qh3 h5 3 g4 Bc8! or 3 gxh6 d4! is in Black’s favour) 2 ... h5 3 g4, Black plays 3 ... Kg7! 4 gxh5 Rh8 5 h6+ Kh7, and White’s attack is beaten back. Black’s defensive plan against the idea of using the white g-pawn to break open the h-file is worth remembering, since there are often opportunities for such play in practice.
Taking into account all the difficulties described, White decided to strike immediately with his rook at h7; Black failed to discover the correct reply.
1 Rxh7! cxd3?
It is true that Black could not take the rook 1 ... Kxh7 because of 2 Qh5+ Kg8 3 Qh6 cxd3 4 Kf2, after which he has no satisfactory defence against 5 Rh1. However, it was also not a good idea to take the bishop; the precious tempo should have been expended on breaking up White’s pawn formation, which is in fact the decisive factor in White’s success on the h-file. Thus 1 ... Bxg5! would have been correct. In that case two of White’s pieces are threatened and if 2 Kf2, then Black can play 2 ... Kxh7; 2 Rh3 can be met by 2 ... Bc8 and then ... Bh4+, while if 2 fxg5, then either 2 ... Qxg5 or 2 ... Kxh7. By sacrificing on g5 Black eliminates White’s control of h6, and in some variations he captures with his queen on e5, thereby covering the square h8 in the manner of a fianchettoed bishop.
2 Kf2!
Uniting the rooks on the back rank above all else! In fact, the chief value of this kind of move derives from the characteristic dual control which it provides. That is, by playing 2 Kf2! White opens up the square h1 both for the queen and the a1-rook, whereas after, for example, 2 Qg4? only the queen would be brought into action and the a1-rook would be left behind. The inadequacy of 2 Qg4? is shown by the following continuation: 2 ... Kxh7 3 Qh4+ Kg8 4 Kf2 f6! 5 e6 (if 5 exf6 Black plays 5 ... Bxf6, enabling his king to wriggle out via f7) 5 ... fxg5 6 Qh6 Bf6 and Black has defended himself; e.g. if 7 Qxg6+ Bg7 8 Rh1, then 8 ... Rf6, while if 7 Rh1, then 7 ... Qe7 8 Qxg6+ Qg7, etc.
2 ... f6
If 2 ... Kxh7, then 3 Qh1+ Kg8 4 Qh6 leaves Black without a defence against 5 Rh1 (for 4 ... f6 is no good on account of 5 Qxg6+). The white queen must be brought to h6, and it makes no difference whether it goes via h1 or h4 (as both Qg4-h4-h6 and Kf2, Qh1-h6 take three tempi), but in the case of the a1-rook, the continuation 2 Kf2! is better than 2 Qg4?, which does nothing to bring it into action.
Finally, it should be mentioned that a delayed sacrifice at g5 is also of no avail, e.g. 2 ... Bxg5 3 Qh1 Bf6 (or 3 ... f5 4 e6! Bf6 5 Qh6 Qe8 6 Rh1 and 7 Rh8+) 4 Qh6! followed by 5 Rh1 and 6 Rh8+.
3 Qh1 1-0
For if 3 ... Qe8 or 3 ... fxe5, then 4 Qh6 is decisive.
The formation with rook on h1 and pawn on h4
Of the various formations on the h-file the most important in practice is undoubtedly that with a rook on h1 and pawn on h4. It is a good representative example of battery attacks with rook and pawn which are created when the pawn captures an opposing piece and thereby opens up the file for the rook. This situation is often engineered by placing another attacking piece on the square covered by the pawn, creating threats which induce or oblige the opponent to take it. There are two other methods of exploiting the formation of rook and pawn – one is to simply advance the pawn, the other involves switching the rook on to another file (e.g. with rook on h1 and pawns at h4 and g2, moving the rook to h3 and then to g3).
Here are a few examples of these ideas.
In this position we see a formation in which preparations have already been made for the opening of the h-file after the capture of the knight at g5. It is true that White, to move, currently has no very sharp threats, but he reckons correctly that he can quickly strengthen his position; in the meantime, he puts his trust in the efficacy of hxg5 if Black takes the knight.
1 c3 hxg5?
We shall examine this weak move in order to get to know the various typical combinations which occur when the h-file is opened in such circumstances. Correct was 1 ... a5!, stopping b4. White would then probably play 2 g4 to exclude the possibility of ... Nf6-h5-g3. Thus, the game could go on for a few more moves with the knight on g5 en prise, until the critical moment came to decide on the exact significance of the knight’s position: either White’s attack would break through, or Black would take the knight without any ill-effects to himself, or thirdly White might simply withdraw the knight from g5.
2 hxg5 Nxd5 3 Bxd5 Be6
If 3 ... g6, then 4 Ke2! (a characteristically quiet move by the king to unite the heavy pieces on the back rank) 4 ... Kg7 5 Rh7+! Kxh7 6 Qh1+ Kg8 7 Qh6 Ne7 8 Bd2 Nxd5 9 Rh1 followed by mate.
4 f4! Re8
This seems to provide the most resistance. If 4 ... g6, then 5 f5 Bxd5 6 f6 Be6 7 Qf3 and White can easily organize a mating attack.
5 Qh5 Kf8 6 Bxe6 Rxe6
Or 6 ... fxe6 7 g6 Ne7 8 f5 exf5 9 Bg5, etc.
JN: In this line, Black can still resist with 7 ... Ke7. so White should prefer 7 Qh8+ Kf7 8 g6+ Kxg6 9 Qh5+ Kf6 10 fxe5+, winning immediately.
7 f5 Re8 8 g6 Ke7 9 Bg5+ and Black either loses his queen or is mated, e.g. 9 ... f6 10 Bxf6+ gxf6 11 Qh7+ Kf8 12 Qf7#.
This position arose in the game
Boden-Bird, London 1869 . White played a move which made clear his ardent wish that Black should take the bishop on g5.
1 Nf4 fxg5?
Black takes the bishop, even though there is no danger at the moment. He could have played, for example, 1 ... a6 2 Ncd5 b5 3 h5, after which there is a real threat of 4 Ng6+. This can be parried by 3 ... h6 and if 4 Ng6+, then 4 ... Kh7.
JN: Although White then wins easily by 5 Ngxe7 Nxe7 6 Bxf6 gxf6 7 Nxf6+ Kh8 8 Qf4 Ng8 9 Nxg8 Kxg8 10 f6.
2 Ng6+!
A typical method of ‘clearing’ the file before the battery opens fire with hxg5.
2 ... hxg6 3 hxg5+ Nh6 4 gxh6 Bg5+ 5 Kb1 Bxh6
If 5 ... gxh6, then 6 Ne4 Kg7 (or 6 ... Kg8 7 Rxh6 Bxh6 8 Qxg6+ Bg7 9 Qf7+ mating) 7 Rxh6! Kxh6 8 Nxg5! Qxg5 (or 8 ... Kg7 9 Ne4) 9 Rh1+ Qh5 10 Qxg6#.
6 Qxg6 Re7 7 f6 Qe8 8 f7 Qf8 9 Rxh6+ gxh6 10 Rh1 Rxe6 11 Qxe6 Kh7 12 Ne4 Rc7 13 Ng5+ and mates.
The diverse character of the formation with rook on h1 and pawn on h4 is well illustrated by the following examples where the rook is switched from the h- to the g-file. In this case the rook is potentially active on two files at the same time and develops great power as a result.
The above position arose in the game Vuković-Dr Vajda, Debrecen 1925. Black had just threatened White’s bishop by ... h6, and White repl
ied:
1 h4! Re8
If Black accepts the sacrifice, then after 1 ... hxg5 2 hxg5 g6 3 gxf6 Nxf6 4 Ne5 White obtains an attacking position without having to surrender any material.
2 Ne5 Bb7
Consistently refusing the gift offered him on g5, Black hopes to persuade the bishop eventually to retreat. However, he is in any case in a poor position, as a result of his premature advance ... c4 earlier in the game. With his pawn on c5 Black would still have some say in the centre, and a central action is usually a valuable antidote against an attack on the castled king. With the pawn on c4, Black is only able to undertake slow operations on the queenside, and they are not sufficient to counterbalance White’s attack.
3 Rh3!
A manoeuvre designed to bring the rook into the game as quickly as possible; it is characteristic of the formation with rook on h1 and pawn on h4.
3 ... Nf8
Black lacks a good answer to the threat of 4 Bxh6; After 3 ... Bxg2 4 Rg3 his situation deteriorates still further, while if 3 ... Kf8, then 4 Bxh6 gxh6 5 Nxf7 Kxf7 6 Qg6+ Kf8 7 Qxh6+ Kf7 8 Bg6+ Kg8 9 Rg3 leads to mate.
4 Bxh6!
4 ... Nh5
If 4 ... gxh6, then 5 Rg3+ soon decides the issue.
5 Qe2 g6
Or 5 ... Bxg2 6 Qxh5 g6 7 Rg3, etc.
6 g4 Ng7 7 h5 Bf6
Black gives up a second pawn, but he can do nothing to save his position. After 8 hxg6 Bxe5 9 gxf7+ Kxf7 10 dxe5 Qa5 White played 11 e4 and won in ten more moves, but the finish might have been even more abrupt with 11 Bxg7 Kxg7 12 g5 followed by Qh5.
The h·file in connection with the classic bishop sacrifice
The analysis of the classic bishop sacrifice, to which we devoted particular attention in an earlier chapter, must now be supplemented by cases where such a sacrifice is combined with an attack along the h-file. This arises when White has a rook on h1 and pawn on h4, the pawn acting as a support for the knight on g5. We shall examine two examples from master chess in which the h-file is opened up.
The following position comes from the game Steinitz-Golmayo, Havana 1888, and arose after the moves:
1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 f4 c5 6 dxc5 Bxc5 7 Nf3 0-0 8 Bd3 Nc6?
Steinitz at this point played:
9 h4
White could have played immediately 9 Bxh7+ Kxh7 10 Ng5+ Kg6 11 Qd3+ f5 and then 12 Qg3 Qe8 (against other queen moves, 13 Qh4 is decisive) 13 Nxe6+ Kf7 14 Nc7 Qd8 15 Nxa8 when White has the upper hand.
JN: In this line 14 Nxg7! Rg8 15 Nxe8 Rxg3 16 e6+ is decisive.
Steinitz rejected the idea of the sacrifice because on the twelfth move he had reckoned only on 12 Nxe6? Qa5 13 Nxf8+ Nxf8 14 Qxd5 Be6, which would indeed have been favourable for Black.
9 ... f6
White would have had a harder task after 9 ... h6, although even then he could have gained the advantage by 10 Ng5. Given that this is a typical continuation, let us examine a few variations:
1) 10 ... hxg5 11 Bh7+! Kxh7 12 Qh5+ Kg8 13 hxg5 f5 14 g6 and mate.
2) 10 ... Ndxe5 11 fxe5 Nxe5 12 Bh7+ Kh8 13 Bf4 Ng6 14 Bd2 and White is the better placed.
JN: This contention seems highly dubious. After 13 ... Nc4, attacking b2 and threatening 14 ... e5, White is clearly worse. Black has two central pawns for the piece, White’s king is trapped in the centre and his pieces are hopelessly tangled up on the kingside. Therefore White should prefer 12 Nf3, with a double edged position.
3) 10 ... Nb4 11 Bh7+ Kh8 12 Qh5 Qe7 13 Bd2 f5 (if 13 ... Re8 then 14 Bg6, while against 13 ... Nc6 or other such moves White plays 14 0-0-0) 14 Qg6 Nxc2+ (also in White’s favour is 14 ... Qe8 15 Nxe6, etc) 15 Kd1 Nxa1 16 Bg8! followed by mate with the queen on h7.
JN: There is also a fair degree of wishful thinking in this line, for example after 14 ... Qe8 15 Nxe6 Nxc2+ 16 Kd1 (16 Ke2 Qxe6!) 16 ... Qxg6 17 Bxg6 Nxa1 the queens are exchanged and White is a whole rook down. He can regain the exchange easily enough, but rounding up the a1-knight is far from easy – the assessment must be that Black is better. All these lines illustrate the point that Vuković himself makes: launching an attack without first completing your development is a risky business.
10 Ng5! fxg5
11 Bxh7+!
The sacrifice is both correct and necessary; after 11 hxg5 Black would have the reply 11 ... g6 12 Rxh7 Ndxe5 13 fxe5 Bf2+, etc.
The fact that the knight is sacrificed on g5 first and then the bishop on h7 makes no difference to the actual structure of the combination; the order of moves is conditioned by temporal factors.
11 ... Kxh7 12 hxg5+ Kg8 13 Qh5
13 ... Ndxe5
Otherwise 14 g6 is decisive.
JN: Here Black misses 13 ... Ncxe5! 14 fxe5 Qc7!, when it is hard to see how White can continue his attack.
14 fxe5 Rf5
JN: In this position 14 ... Qc7 is less effective, as 15 Bf4! Rxf4 16 Qh8+ Kf7 17 g6+! Ke7 18 Qxg7+ gives White a very dangerous attack – Black needs a defensive knight on d7.
15 g4 Rxe5+ 16 Kd1 Be3
If 16 ... Bd7, then 17 Qh8+ Kf7 18 Rf1+ Ke7 19 Qxg7+ Kd6 20 Nb5#, while if 16 ... Bf8, then 17 Bf4.
17 Bxe3 Rxe3 18 Nb5!
In the preceding note we gave a line ending with Nb5#; now the same move serves to tighten the mating net, which Black had attempted to cut through by 16 ... Be3 (freeing the c5-square for the king’s escape!).
18 ... Rf3
The alternatives were 18 ... Kf8 19 Qh8+ Kf7 20 Rf1+ followed by mate and 18 ... Re5 19 Qh7+ Kf8 20 Rf1+, etc.
19 g6 1-0
It is the formation with pawn on g6 against pawn on g7 (which has already been discussed) that is finally decisive. The part played by the knight on b5 is evident in several variations, as well as in the final position, i.e. if 19 ... Kf8, then 20 Qh8+ Ke7 and 21 Qxg7+, etc.
Alekhine - Asgeirsson
Reykjavik Simultaneous Display, 1931
French Defence
1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Be7 5 Bxf6 Bxf6 6 Nf3 0-0 7 Bd3 Re8?!
An unnecessary loss of time; correct is 7 ... c5 at once.
8 e5 Be7 9 h4 c5?!
White has prepared for the sacrifice of his bishop on h7 by 9 h4, and the reply 9 ... c5 does nothing to stop it. Relatively best was 9 ... f5! 10 exf6 (after 10 Ne2 c5 11 c3 White is the better placed, but his attack is delayed and Black is not without counterplay) 10 ... Bxf6. The situation is then easier for Black. for as a result of the exchange on f6 an extremely important attacking unit has disappeared, namely White’s pawn on e5. Sacrifices in this position are therefore not worthwhile; after 11 Bxh7+ Kxh7 12 Ng5+ Kg8 13 Qh5 Black has an adequate defence in 13 ... Qd7, while if 11 Ng5, he can play 11 ... Bxg5 12 hxg5 Qxg5! 13 Bxh7+ Kf8, and White’s attack is not worth much owing to his weakness in the centre and the activity of Black’s queen, especially on the dark squares.
JN: This is not so. After 14 Qf3+ Ke7 (14 ... Qf6 15 Bg6) 15 Rh5 Qf6 16 Qg3 Qxd4 (or else 17 0-0-0, with a massive lead in development without any sacrifice by White) 17 Rd1 Qf6 18 Bg6 Rf8 19 Rf5 Black’s queen is trapped.
The last continuation can also be brought about by a different series of moves: 11 Ng5 Bxg5 12 Bxh7+, when Black should not play 12 ... Kxh7 (in which case 13 hxg5+ and 14 Qh5 is extremely strong) but 12 ... Kf8! 13 hxg5 Qxg5, ending with the position described above. These possibilities are worth getting to know well, since we are dealing with a typical position. Particularly important is the significance of Black’s ... f5 in this kind of position: if White does not take the pawn at once en passant, the sacrifice on h7 is prevented directly; while if he takes it, then it is indirectly thwarted or at any rate made weaker, since White’s pawn has disappeared from e5.
It is also worth noting that to avert the sacrifice entirely by ... h6 or ... g6 is not a good thing in this position. If 9 ... h6, White plays 10 Ne2 followed by c3 and attacks in a different way (e.g. by means of Bb1 and Qd3 he forces a weakening of the square g6 or, alternatively, he plays g4, etc); in this connection it is important to note that ... f5, if played later on, would create a new weakn
ess on g6. If 9 ... g6, White can immediately play 10 h5, but still better is first 10 Ne2, etc.
10 Bxh7+!
The sacrifice is correct, even though it demands extremely precise and high quality play from the attacker.
10 ... Kxh7 11 Ng5+ Kg8
11 ... Kg6 would be hopeless here; the formation comprising rook on h1 and pawns on h4 and e5 is a formidable weapon against it, e.g. 12 Qd3+ f5 13 h5+ Kxg5 14 Qg3+ Kh6 and 15 Qg6#.
12 Qh5 Bxg5
He must!
13 hxg5 Kf8 14 g6!
This move puts a spoke in Black’s defensive plans. He would lose quickly after 14 ... fxg6 15 Qxg6, while if 14 ... cxd4 White can win either by 15 gxf7 or 15 Qh8+ Ke7 16 Qxg7 Rf8 17 Qf6+ Kd7 18 g7 Rg8 19 Rh8 Qe8 20 Nxd5 Nc6 21 Qh6.
14 ... Ke7 15 gxf7 Rf8
16 0-0-0
Alekhine the player and Alekhine the commentator failed to see that Black could have saved himself after this move. 16 dxc5! was both necessary and good in order to prevent Black from taking the initiative in the centre. This is underlined by the following variations:
1) 16 ... Nd7 17 0-0-0 Qa5 (if 17 ... Rxf7 White wins by 18 Nxd5+ and 19 e6, while if 17 ... Nxc5, then 18 b4) 18 Qg5+ Kxf7 19 Rh7 Rg8 20 Nxd5! Qxa2 (if 20 ... exd5, then 21 e6+ Kxe6 22 Qxd5+ and 23 Qxg8 wins, while if 20 ... Qxc5, then 21 Qh5+ Kf8 22 Nf4 wins) 21 Qe7+ Kg6 22 Qxe6+ Nf6 23 Rxg7+ mating.
2) 16 ... Nc6 17 0-0-0 Qa5 (or 17 ... Rxf7 18 Qg5+ Kd7 19 Qg6 Nxe5 20 Rxd5+ exd5 21 Qd6+ Ke8 and 22 Rh8+, etc, while 19 ... Qe7 transposes to line 3) 18 Qg5+ Kxf7 19 Rh7 Rg8 20 Qf6+ Ke8 21 Qg6+ Ke7 (if 21 ... Kf8, then 22 Rh4 wins, while if 21 ... Kd8 then 22 Nxd5) 22 Rxg7+ Rxg7 23 Qxg7+ Ke8 24 Rh1! Ne7 25 Rh7 Qxc5 26 Na4! Qc7 (or 26 ... Qb4 27 Qf7+ Kd8 28 c3!) 27 Qf7+ Kd8 28 Qf8+ mating.
3) 16 ... Rxf7 17 Qg5+ Kd7 18 Qg6 Qe7 (defending the rook and parrying the threat of Nxd5) 19 0-0-0 Nc6 (if 19 ... Kc7, then 20 Nb5+ and 21 Nd6, while if 19 ... Kc6, then 20 Rxd5) 20 Qg4! Kc7 21 Nb5+ Kb8 22 Nd6 Rf8 23 Rh7 Nxe5 (if 23 ... Rg8, then 24 Rdh1 followed by Rh8, etc) 24 Qg3 Qf6 25 Rxg7 Qf4+ (should Black attempt to free himself with 25 ... Bd7, White wins by 26 Qxe5 and 27 Rxd7) 26 Qxf4 Rxf4 27 Rh1 Rf8 28 Rhh7 and mate on b7.
Art of Attack in Chess Page 14
