Now we would have here the case of the backward Q B P, which will in no way be able to advance to Q B 4. Such a position may be said to be theoretically lost, and in practice a first-class master will invariably win it from Black. (If I may be excused the reference, I will say that I won the game above referred to.)

After a few moves the position may be easily thus:

The Black pieces can be said to be fixed. If White plays Q - Q B 3, Black must answer Q - Q 2, otherwise he will lose a Pawn, and if White returns with the Queen to Q R 3 Black will have again to return to Q Kt 2 with the Queen or lose a Pawn. Thus Black can only move according to White's lead, and under such conditions White can easily advance with his Pawns to K B 4 and K Kt 4, until Black will be forced to stop P - B 5 by playing P - K B 4, and we might finally have some such position as this:

Example 66.

In this situation the game might go on as follows:

1. P × P, P × P; 2. Q - K B 3, Q - Q 2

White threatened to win a Pawn by Q × P, and Black could not play 2...R - K B 1, because 3 R × B P would also win a Pawn at least.

Now suppose that in the position in the preceding diagram it were Black's move, and he played R - K B 1. White would then simply defend his K B P by some move like Q - K B 3, threatening R × Q B P, and then he would bring his King up to Kt 3, and when the time came, break through, as in the previous case. White might even be able to obtain the following position:

Black would now be forced to play R - B 1, and White could then play Q - B 2, and follow it up with K B 3, and thus force Black to play P × P, which would give White a greater advantage.

A careful examination of all these positions will reveal that, besides the advantage of freedom of manœuvre on White's part, the power of the Pawn at K 5 is enormous, and that it is the commanding position of this Pawn, and the fact that it is free to advance, once all the pieces are exchanged, that constitute the pivot of all White's manœuvres.

I have purposely given positions without the moves which lead to them so that the student may become accustomed to build up in his own mind possible positions that may arise (out of any given situation). Thus he will learn to make strategical plans and be on his way to the master class. The student can derive enormous benefit by further practice of this kind.

33. THE INFLUENCE OF A "HOLE"

The influence of a so-called "hole" in a game has already been illustrated in my game against Blanco (page 81), where has been shown the influence exercised by the different pieces posted in the hole created at White's K 5.

Example 67.—In order to further illustrate this point, I now give a game played in the Havana International Masters Tournament of 1913. (Queen's Gambit Declined.) White: D. Janowski. Black: A. Kupchick.

Of course the idea is to post a Knight at Q 4, but as it is the other Knight which will be posted there this manœuvre does not seem logical. The Knight at Kt 3 does nothing except to prevent the development of his own Q B. The normal course O - O, followed by P - Q B 4, is more reasonable. For a beautiful illustration of how to play White in that variation, see the Janowski-Rubinstein game of the St. Petersburg Tournament of 1914.

B - Kt 3 has some points in its favour in this position, the most important being the possibility of advancing the King's Pawn immediately after 8. ... K Kt - Q 4; 9 B × B, Q × B.

Had White's Bishop been at Q Kt 3 he could now play P - K 4 as indicated in the previous note, a move which he cannot make in the present position, because of Kt - K B 5 threatening, not only the K Kt P, but also Kt × B ch. As White's King's Bishop should never be exchanged in this opening without a very good reason White therefore cannot play P - K 4.

White is perfectly developed, and now threatens to win a Pawn as follows: Kt × Kt, Kt × Kt; P - K 4, followed by R × P.

The fact that Black is practically forced to make this move in order to avoid the loss of a Pawn is sufficient reason in itself to condemn the whole system of development on Black's part. In effect, he plays B - Q 2 and now he has to shut off the action of his own Bishop, which thereby becomes little more than a Pawn for a while. In fact, it is hard to see how this Bishop will ever be able to attack anything. Besides, it can be easily seen that White will soon post his two Knights at K 5 and Q B 5 respectively, and that Black will not be able to dislodge them without seriously weakening his game, if he can do it at all. From all these reasons it can be gathered that it would probably have been better for Black to play Kt × Kt and thus get rid of one of the two White Knights before assuming such a defensive position. In such cases, the less the number of pieces on the board, the better chances there are to escape.

This practically amounts to committing suicide, since it creates a hole at K 5 for White's Knight, from where it will be practically impossible to dislodge him. If Black intended to make such a move he should have done it before, when at least there would have been an object in preventing the White Knight from reaching B 5.

The position of White's Knights, especially the one at K 5, might be said to be ideal, and a single glance shows how they dominate the position. The question henceforth will be how is White going to derive the full benefit from such an advantageous situation, This we shall soon see.

There is no object in this move, unless it is to be followed by Kt - Q 2. As that is not the case, he might have gone with the Rook to B 1, as he does later.

White threatened to win the exchange by playing either Kt - B 7 or Kt - Kt 4.

Perhaps all these precautions are unnecessary, but White feels that he has more than enough time to prepare his attack, and wants to be secure in every way before he begins.

He had better have played Kt - B 3; and tried later on to get rid of White's Knights by means of Kt - Q 2.

with this sacrifice of the Rook for a Knight and Pawn White obtains an overwhelming position.

Kt - Q 2 was better in order to get rid of one of the two White Knights. There were, however, any number of good replies to it, among them the following: Kt (B 5) × Kt, B × Kt; Q × P, Q × Q; Kt - B 7 ch, K - Kt 2; Kt × Q, and with two Pawns for the exchange, and the position so much in his favour, White should have no trouble in winning.

All these moves are practically forced, and as it is easily seen they tie up Black's position more and more. White's manœuvres from move 24 onwards are highly instructive.

This wandering Knight has done nothing throughout the game.

If P × P; K - B 2, and Black would be helpless.

There is nothing to be done. If B - Kt 1; Q - R 2 ch, K - Kt 2; B × P.

The student should notice that, apart from other things, White throughout the game has had control of the Black squares, principally those at K 5 and Q B 5.

From now on to the end of the book I shall give a collection of my games both lost and won, chosen so as to serve as illustrations of the general principles laid down in the foregoing pages.

PART II

GAME 1. QUEEN'S GAMBIT DECLINED

(Match, 1909)

White: F. J. Marshall.          Black: J. R. Capablanca.

I had played this defence twice before in the match with good results, and although I lost this game I still played it until the very last game, when I changed my tactics. The reason was my total lack of knowledge of the different variations in this opening, coupled with the fact that I knew that Dr. E. Lasker had been successful with it against Marshall himself in 1907. I thought that since Dr. Lasker had played it so often, it should be good. The object is to exchange a couple of pieces and at the same time to bring about a position full of possibilities and with promising chances of success once the end-game stage is reached. On general principles it should be wrong, because the same Knight is moved three times in the opening, although it involves the exchange of two pieces. In reality the difficulty in this variation, as well as in nearly all the variations of the Queen's gambit, lies in the slow development of Black's Queen Bishop. However, whether this variation can or cannot be safely played is a question still to be decided, and it is outside the scope of this book. I may add that at present my preference is for a different system of development, but it is not unlikely that I should some time come back to this variation.

P × P is preferable for reasons that we shall soon see.

Now P × P would be a better way to develop the game. The idea is that after 8...P × P; 9 B × B P, P - Q Kt 3, followed by B - Kt 2, would give Black's Bishop a powerful range. For this variation see the eleventh game of the match.

No longer would 9...P × P; 10 B × P, P - Q Kt 3 be good, because 11 B - Kt 5 would prevent B - Kt 2 on account of Kt - K 5.

Played with the intention of obtaining the majority of Pawns on the Queen's side. Yet it is doubtful whether this move is good, since it leaves Black's Queen's-side Pawns disrupted in a way. The safer course would have been to play P - B 3.

This exposes Black to further attack by P - R 5 without any compensation for it. If I had to play this position nowadays I would simply play 13...R - K 1. Then after 14 Q × P, Q × Q would follow, and I believe that Black would regain the Pawn. If, instead, White played 14 P × P then B - Kt 5 would give Black an excellent game.

Black's position was bad and perhaps lost in any case, but the text move makes matters worse. As a matter of fact I never saw White's reply B - B 5. It never even passed through my mind that this was threatened. Black's best move would have been 16...K R - Kt 1. If that loses, then any other move would lose as well.

From bad to worse. Kt - B 3 offered the only hope.

The game was lost. One move was as good as another.

Of course, if 25 Kt × B, R - Kt 8 ch would have drawn. The text move is pretty and finishes quickly. A well-played game on Marshall's part.

GAME 2. QUEEN'S GAMBIT DECLINED

(San Sebastian, 1911)

White: A. K. Rubinstein.          Black: J. R. Capablanca.

Kt - B 3 is the normal move in this variation. White's development was first introduced by Schlechter and elaborated later on by Rubinstein. It aims at the isolation of Black's Q P, against which the White pieces are gradually concentrated. In making the text move I was trying to avoid the beaten track. Being a developing move there should be no objection to it in the way of general principles, except that the Knights ought to come out before the Bishops.

In pursuance of the idea of changing the normal course of this variation, but with very poor success. The move in theory ought to be unsound, since Black's K Kt is yet undeveloped. I had not yet learned of the attack founded on Kt - Kt 5 and the exchange of the B at K 3. Either Kt - B 3 or P - K R 3; to prevent either B or Kt - K Kt 5, was right.

This is a mistake. The right move was R - Q 1 in order to get the Rook away from the line of the Bishop at R 3 and at the same time to support the Q P. Incidentally it shows that White failed to take proper advantage of Black's weak opening moves. Against the text move White makes a very fine combination which I had seen, but which I thought could be defeated.

I considered P × B, which it seemed would give me a playable game, but I thought White's combination unsound and therefore let him play it, to my lasting regret.

This is the move which I had not considered. I thought that Rubinstein would have to play B - Kt 2, when I had in mind the following winning combination: 16 B - Kt 2, Kt - K 4 ! 17 Kt - B 4 (if R - B 1, Q × R !! Q × Q, B × P ch wins), Kt - Kt 5; 18 P - K R 3 (if Kt - R 3, B × P ch wins the exchange), Kt × P; 19 R × Kt, B × R ch; 20 K × B, P - K Kt 4, and Black should win. It is curious that this combination has been overlooked. It has been taken for granted that I did not see the 17th move Q - B 1.

After White's last move there was nothing for me to do but submit to the inevitable.

This gives Black a chance. He should have played K R - K 1. If then Kt - B 7; R × R ch, R × R; R - Q B 1, R - K 7; K - B 1, Kt - Q 5 (if R - Q 7; B - K 6 ch, K - B 1; B × P would win); R - B 8 ch, K - B 2; R - B 7 ch, R - K 2; R - B 5 wins.

A bad move, which gives away any legitimate chance Black had to draw. It loses a very important move. In fact, as the course of the game will show, it loses several moves. The proper way was to play K - Q 3. If then R - Q Kt 5, R × R; B × R, Kt - Q 5; followed by P - Q Kt 4; and White would have an exceedingly difficult game to draw on account of the dominating position of the Knight at Q 5 in conjunction with the extra Pawn on the Queen's side and the awkward position of White's King. (See how this is so.)

With these last three moves White again gives Black a chance. Even before the last move B - B 4 would have won with comparative ease, but the text move is a downright blunder, of which, fortunately for him, Black does not avail himself.