Art of Attack in Chess

by Vladimir Vukovic

  The basic elements of the attack on the castled king

  The attack on the castled king is a particular type of mating attack, but in practice it is the most important, since castling takes place in the majority of games.

  In considering an attack of this kind, we must begin with a general survey of the action involved, acquainting ourselves, on the one hand, with the basic characteristics of the actual castled position itself and, on the other, with the good and bad aspects of an attack on that position.

  The castled position consists of the castled king and the protective pawns and also to a certain extent the pieces which are in the vicinity or in some other way connected with the formation.

  This structure should be regarded as a positional formation, i.e. a group of pieces having one or more fairly permanent features. One could also go further and say that castled positions are, as a rule, passive rather than active positional formations, i.e. their more permanent feature stems ultimately from a certain immobility of the pieces, and not in their mobility, as is the case when the pieces are in strong positions. (This is, of course, the general rule, but in practice the castled position may also contain a piece in a strong position, for example a fianchettoed bishop.)

  The passive nature of the castled position results from the lack of mobility of the king, and this is indirectly transferred to the other pieces in the position as well; the first to be affected are the pawns, which ‘may not be moved, if the castled position is not to be weakened’, and which are as a result made indirectly immobile. If they in fact move from their initial squares and so create weaknesses in the castled position, these weaknesses result from the reduced options entailed by the advance. (A pawn at h2 controls the square g3 and also has it in its power, by advancing, to govern g4, g5, g6, etc. If it moves to h3, then, out of its total number of actual or potential areas of control, it loses that over g3, and so loses part of its active potential.) Even the pieces associated with the castled position are affected to a certain extent by the general immobility of the formation, whether functionally because of the need for a certain square to be defended or spatially because of the difficulty involved in their centralization.

  The structure and the degree of weakness of the opponent’s castled position also influence the way in which the attacker employs his pieces in a mating attack. The variety here is extremely wide, ranging from lightning assaults to purely positional manoeuvres and from uncompromising charges to cautious preparations, in which the player really only flirts with an attack on the king in order to restrict his opponent’s options. More will be said on these and similar points later on.

  However, the basic weaknesses of the castled position which I have described also contain one positive feature which becomes apparent when an attempt is made to exploit them. This positive feature derives from the fact that the castled position is always a decentralized one, so that as a rule any attack on it demands a greater or lesser decentralization of the attacker’s pieces. This factor is particularly important in that the main positional rule for normal play is that pieces should be centralized. There is, then, a basic contradiction between consistent centralization and an attack on the castled king, and it is here that the principal risk of such an attack lies. If the attack does not succeed, the fact that some of the attacker’s pieces are out of position as a rule tells against the attacker.

  The reader will come to understand the deeper significance of this argument in the final chapters of this book.

  To understand the attack on the castled king we shall start with the basic factors involved, spatial, material and temporal (the method I applied in my Introduction to Chess and School of Combinations). It is true that in every action which takes place on the chessboard all these three factors can be observed together, since each move has its own spatial, material and temporal significance; however, the emphasis of one’s actual observation may be in one case on the spatial, in another on the material, and in a third on the temporal, aspect. For this reason, the separation of the different factors is the most effective and the clearest method for teaching purposes.

  The attack on the castled king in the past

  At the time when modern chess was in its infancy, i.e. after the reforms of the rules in the last quarter of the fifteenth century, the attack on the castled king did not yet exist, or at any rate there was no trace of it in the games recorded. The castling move itself was at that time only in the process of being created (some were still using the mediaeval king’s jump, while others were castling but in two moves – i.e. first Rh1-f1, and then on the next move Ke1-g1, etc) and combinative technique – especially with the ‘new pieces’, the queen and the knight – was still fairly primitive, so that the decision in a game was usually reached by cruder methods than by attacking the castled king, e.g. by capturing on f7 before castling had taken place or simply by winning material. The first recorded instance of an attack on the castled king occurs in the so-called Paris MS., which originated about 1500, a more mature and fuller development of Lucena’s work of 1497, the oldest printed book on modern chess. In the course of analysing an opening (one later given the name of Philidor Defence) the author arrives at the position:

  Black, to move, plays:

  1 ... g5! 2 Nh2 h5

  He uses his pawns to carry out an attack on the king, a plan suited to this position as it takes advantage of White’s weakness on h3.

  3 f3?

  A mistake, which Lucena makes ‘in accordance with the thought of his times’. White could have defended himself satisfactorily if he had immediately played 3 b5!.

  3 ... Nh6 4 b5 Rg8 5 Qe2?

  A further mistake. White should at once have played 5 bxc6 Nxc6 6 Qd5 Qd7 7 Ba3, when Black, pinned down at c6 and g8, would have had to refrain from going straight ahead with ... g4 and consolidate his position first.

  5 ... g4 6 fxg4 hxg4 7 Nxg4?

  Poor defensive technique! After 7 Kh1 Black would have had a difficult task, for 7 ... gxh3 would not do any good because of 8 Qh5+ followed by Qxh3.

  7 ... Nxg4 8 hxg4 Rxg4 9 bxc6 Nxc6 10 Nd5 Bd8

  11 c4?

  White could have put up more resistance by 11 Ra3. No real advantage would come of 11 Qc4 because of 11 ... Kd7. Naturally, 11 c4? is very weak and quite inappropriate, since it does nothing to help the defence and only obstructs the diagonal for the white queen.

  11 ... Ra7 12 Rf2 Rh7 13 Qd1? Qh6?!

  Black’s technique is poor too. After 13 ... Rgh4 14 Rf1 Qg6!, White would be caught in a mating net.

  14 Kf1 Qh5 15 Qf3 Qg6 16 Qd3 Rh1+?

  The manuscript follows on from here with 17 Ke2 Rxa1 18 Bxa1 Rxg2 and concludes: ‘Black has the better prospects’. To all appearances the author was mainly concerned with the win of a pawn, forgetting that the primary aim of an attack on the castled king is mate! By 16 ... Rg3 17 Qc2 Qg4 18 Ke1 Ba5+ the matter could have been brought to an end considerably more quickly.

  The advance in technique in attacking the castled king was extremely slow throughout the sixteenth century, and it was only at the beginning of the seventeenth that Greco in his manual indicated some combinations which, though new then, now strike modern minds as nothing more than elementary tricks. As an example of the technical level reached by Greco, let us look at a continuation from his book.

  1 e4 e5 2 Nf3 Nc6 3 Bc4 Bc5 4 0-0 Nf6 5 Re1 0-0 6 c3 Qe7?

  Greco bad not yet realized the importance of keeping control of the centre and gives Black this weak move instead of the correct 6 ... d6.

  7 d4 exd4 8 e5!

  This is even stronger than 8 cxd4.

  JN: As we shall see, this comment is incorrect. After 8 cxd4 While has a clear advantage, since Black has no compensation for White’s central control. The move 8 e5 is a mistake which gives Black at least equality.

  8 ... Ng4

  9 cxd4?

  Weakly played. By 9 Bg5 (an intermediate move!) 9 ... Qe8 10 cxd4 White would have obtained a winning position. (Greco and other authors right u
p to the middle of the nineteenth century did not see the intermediate move 9 Bg5, so that 8 e5 came to be regarded in the textbooks as a mistake!

  JN: Black should meet 9 Bg5 by 9 ... dxc3!. After 10 Bxe7 Bxf2+ 11 Kf1 cxb2 12 Nc3 (12 Na3 Bxe1 also favours Black) 12 ... Bxe1 (12 ... bxa1Q 13 Qxa1 Bxe1 14 Nd5 is unclear) 13 Nd5 Nxe7 14 Nxe7+ Kh8 Black has a clear advantage in view of the threats of 15 ... bxa1Q (and 15 ... Ne3+. White’s strongest move is 9 b4! but an analysis of the resulting complications (which lead to a roughly equal position) would take us too far afield. Perhaps those old analysts knew a thing or two after all!

  9 ... Nxd4! 10 Nxd4 Qh4

  This is Greco’s trick. White has two exposed spots, at h2 and f1.

  11 Be3

  Greco naturally also gives the continuation 11 Nf3 Qxf2+ followed by 12 Kh1 Qg1+ and a smothered mate by the knight at f2.

  11 ... Qxh2+ 12 Kf1 Qh1+?!

  In carrying his attack further Greco does not notice all the finesses. 12 ... Nxe3+ followed by ... d6 would have been better here.

  JN: A baffling comment, as after 13 fxe3 d6 14 Nf3 White starts to consolidate his extra material. Greco’s continuation appears best to me.

  13 Ke2 Qxg2 14 Rg1 Nxe3

  15 Kxe3?

  This way White loses two valuable pawns, and Greco concludes his analysis with 15 ... Bxd4+ 16 Kxd4 Qxf2+ 17 Kc3 Qe3+ 18 Bd3 Qxe5+, with advantage to Black.

  Instead of 15 Kxe3?, the correct continuation is 15 Rxg2 Nxd1 16 Nf5, when Black must find the reply 16 ... d5! 17 Nxg7 Bh3!, if he is to retain the better prospects (though even here there is still some uncertainty after 18 Rg3 Bxf2 19 Rxh3 dxc4 20 Nd2, etc). If he played 16 ... Nxb2 17 Nh6+ Kh8 18 Nxf7+, he would have to consent to a draw, since 18 ... Rxf7 is probably not a good idea.

  JN: After 19 Bxf7 b6! followed by ... Ba6+ and developing the rook, Black is virtually winning – three pawns for the exchange and a pair of active bishops is surely too much.

  In step with the general advance in chess skill there can also be seen an advance in the technique of attacking the castled king, first with Philidor and then, particularly, with Morphy. The romantic reaction which followed led to the development of a number of particular types of sacrificial attacks, but without in general reaching Morphy’s level. Steinitz was concerned with perfecting defensive technique, and his influence on the style of the period was slowly beginning to make the attack on the castled king fade from master chess. Lasker, too, the new World Champion, was even in his youth already turning away from forcing such attacks, preferring to aim at perfection in the endgame. At the end of the century, however, there were three great masters who defended the thesis of attacking chess and who in doing so clearly showed themselves to be supporters of the attack on the castled king. These were Chigorin, Tarrasch and Pillsbury, each of whom in his own way helped to enrich the technique of attack. Chigorin illustrated particularly the power of the positional sacrifice in his games; Tarrasch’s thesis was that the rapid development of the pieces and a strong centre were the basic requisites for an attack on the castled king; while Pillsbury created a kind of synthesis of the attacking style and the positional and endgame play of Steinitz and Lasker. In other words, the American grandmaster showed how a rapid switch can be made from attack to a simplification which gives the better ending. In this way all the basic elements of a perfected technique for attacking the castled king were set out, and after a short period of stagnation in the first two decades of the present century there followed the real blossoming of the dynamic style with the arrival of Capablanca and Alekhine. This development is still relevant to the theme of the attack on the castled king today, and more will be said about it at the end of the book.

  4 Mating patterns

  Special relationships in a mating attack can be observed by examining mating patterns, focal-points, and the roles played by the ranks, files, and diagonals. In this chapter we shall be concerned with mating patterns. This choice has a practical justification in that mating patters can be more easily remembered than the ways leading up to mate. For this reason mating patterns, even though they occur only at the very end of a mating attack, are fundamental for practical purposes because of the excellent mnemonic assistance which they provide to the player.

  One first concern is for the most common mating patterns; these hardly ever occur in the middle of the board, but they are very frequent in the castling area, i.e. near the king’s position on the edge of the board.

  Typical mates without enemy pieces

  There are a very large number of typical mates of this kind and most of them are elementary, so we shall consider only a few examples. (For the sake of greater clarity the diagrams are shown without the white king.)

  In the first position White, to move, has a choice between two typical mates.

  1) 1 Bg6+ Kg8 2 Qh7+ Kf8 3 Qf7#

  2) 1 Bf5 Kg8 2 Be6#

  In this position there are again two possible mates, depending on how Black replies.

  If 1 Nf6+ Kh8, then 2 Rh7# (the so-called Arab mate, because it is recorded in Arabic manuscripts of the ninth century). If 1 Nf6+ Kf8, then there follows 2 Rf7#, a mating pattern which is worth remembering.

  Another mate which should be known is this one with two knights, but if there are no other pieces on the board then it cannot be forced.

  Typical mates with enemy pieces

  The so-called back-row mate by the rook (or queen) is one of the most typical mating patterns.

  Of the fairly large number of typical mates with queen and pawn this one is given first. Black is to move; he cannot escape mate, since he is threatened by 1 Qh6 and 2 Qg7#. If he plays 1 ... Kh7, then 2 Qh4+ and 3 Qh6, etc. If Black had a pawn on h7 he would have the defence 1 ... Kh8 2 Qh6 Rg8.

  Damiano’s Mate

  This is a position from the book (1512) by Damiano, a Portuguese apothecary, one of the first authors in the literature of European chess. White mates in five moves.

  1 Rh8+ Kxh8 2 Rh1+ Kg8 3 Rh8+ Kxh8 4 Qh1+ Kg8 5 Qh7#

  The rook sacrifices were naturally introduced to make the essential point clear: the queen must reach the h-file with gain of tempo.

  The pattern of Damiano’s mate can be seen more clearly in the following examples.

  1 Qh5 Be7 2 Qh7+ Kf8 3 Qh8#

  In this case a bishop has been placed on f8 and also a pawn on d6, with the result that the bishop can only move to e7, where it blocks the king’s escape. The general theme of Damiano’s mate, the pattern of which is characterized by a black pawn on g7 and a white pawn or bishop on g6, can be subjected to numerous variations.

  The following diagram serves as a memory aid for mates with queen and rook:

  Black, to move, cannot avoid mate; if he plays 1 ... Kg7, then 2 Qh6+ and 3 Qh7# or 3 Qh8#, while if 1 ... f6, then 2 Qh7#. If Black has another pawn on g7 he can save himself from immediate mate by 1 ... f6, while if he has one on g6 he can extricate himself by 1 ... Kg7.

  Of the mates with queen and bishop this position is worth remembering.

  Black cannot save himself from mate, but if he plays 1 ... g6, White does not reply 2 Bxg6+ (which is answered by 2 ... Kg7, and Black’s king slips away) but 2 Qh6, in which case there is no remedy against mate.

  The smothered mate with queen and knight is a celebrated one: 1 Nf7+ Kg8 2 Nh6+ Kh8 (or 2 ... Kf8 3 Qf7#) 3 Qg8+! Rxg8 (or 3 ... Nxg8) 4 Nf7#

  This position falls into category of typical mates, and it is useful to know the mating patterns which arise if Black tries to parry the threat Qh7#, e.g. with 1 ... Rb8. If there is a black pawn on e7 or if the square is otherwise occupied by Black, then there follows 2 Qh7+ Kf8 3 Qh8#; if this is not so, mate can be achieved, for example, by 2 Qxf7+ Kh8 3 Ne6 Rg8 4 Qh5#

  Of the various attacking positions with queen and knight this scheme is also well worth remembering. On 1 Nf6+, Black replies 1 ... Kh8 and if 2 Nh5+, then 2 ... f6.

  If 1 ... Kg7, White has a double check with 2 Ne8+ (in some cases, depending on the arrangement of the other pieces. 2 Nh5+ may be stronger)
and after 2 ... Kh6 3 Qf4+ g5 4 Qf6+ Kh5 5 Ng7+ White only needs a pawn on g2 or h2 to force mate.

  This position is also worth knowing; Black is forced to play 1 ... h6, after which he is mated by 2 Bxf7+ Kh8 3 Qg6! hxg5 4 Qh5#

  In the case of mating patterns with rook and bishop one should first of all get to know the above position. It is Black’s move and he must try to parry Rg5#; this cannot be achieved by 1 ... h6, because of 2 Rxh6, which leaves him with no way out. So 1 ... Rc8 2 Rg5+ Kf8 3 Rg7!, and now too mate is unavoidable, for Black is unable to free the square d7 in time to provide the king with an escape square. If he did not have, for instance, the pawn on c6, he could play 3 ... Bc6 4 Rxh7 Ke8 and then 5 Rh8+ could be met by 5 ... Kd7. One could replace Black’s bishop and c6-pawn by a white pawn on c6 and a black one on c7; mate would still be unavoidable.

  JN: In the diagram position, Black can avoid mate by 1 ... Ra8 2 Rg5+ Kf8 3 Rg7 Ra4 4 Rxh7 Rg4. It is true to say that this defence is not often available in practical situations, but it is worth bearing in mind the idea of ‘defending from behind’ because it is often overlooked.

  Morphy’s Mate

  The pattern of this mate with rook and bishop becomes clear in its final stage.

  1 Qxf6 gxf6 2 Rg1+ Kh8 3 Bxf6#

  Morphy’s Concealed Mate

  A similar mating pattern occurs in this diagram: 1 Rxg7+ Kh8 2 Rxf7+ (not 2 Rg1+ because of 2 ... f6) 2 ... Kg8 3 Rg7+ Kh8 4 Rg1+ Rf6 5 Bxf6#