In his book Reassess Your Chess renowned author and teacher Jeremy Silman wrote that he struggled with whether or not to include a section on elementary endings at the beginning of the book which is on the middlegame. His conclusion was that everybody needs to know the basics.
Today's game was played in the great New York 1924 tournament. In December 1923, following an unsuccessful attempt to arrange a World Championship match between Capablanca and Alekhine, Hermann Helms, publisher of the American Chess Bulletin, Harry Latz, the General Manager of the Alamac Hotel in New York and Norbert Lederer, the Secretary of the Manhattan Chess Club organized a tournament to rival the 1904 Cambridge Springs tournament.
Capablanca was expected to be the winner but the 55-year-old Emanuel Lasker ran away with the tournament. It also featured Capablanca's first tournament loss in eight years when, as white, Reti played his own opening and won.
When I post games in this blog I rarely include endings because most players are likely to skip over them because they are perceived as boring. Nevertheless, today's game features a Capablanca-Tartakower game that is a classic example of endgame play and one that has been analyzed by the greats and near greats ever since the day it was played.
Irving Chernev wrote that the ending "...may very well be the finest and the best of it's kind, is one to delight the connoisseur, and one to grant the student a priceless lesson in the art of finishing a game with elegance and accuracy."
During my analysis over the
past two days I have consulted notes by Reti, Alekhine, Graham Burgess,
Andrew Soltis, Irving Chernev, US Life Master A.J. Colby, John Nunn and
the engines Stockfish and Komodo.
As often happens in these old games, today's engines find lines that were missed even by world champions, but the important thing is we can learn ideas from watching Capablanca's play.
If Tartakower mishandled the position and all the great annotators in the past missed the importance of 33...Nd1 instead of Tartakower's 33...Nf5 then anybody we ordinary players meet probably would too! Therefore it's likely if we are familiar with the ideas Capablanca used in this great game we can apply them in our own games with the same result...hopefully!
Jose Capablanca - Savielly Tartakower
Result: 1-0
Site: New York
Date: 1924.03.23
Dutch Defense
[...] 1.d4 e6 2.♘f3 f5 The rarely seen fighting Dutch Defense is an exciting opening that offer black a good chance to win because black is able to avoid simplification and can reach unbalanced positions, which allow him to play for more than equality. 3.c4 By far the most common move here is 3.g3 because on g2 white's B is well placed. 3...♘f6 4.♗g5 Different sources have shown different move orders, but they all transpose to this position. More modern are either 4.Nc3 or 4.g3, but Capa prefers classical development where he wants to play e3 without shutting in his B. 4...♗e7
24.cxd5 exd5 25.♕xe8+ ♕xe8 26.♖xe8+ ♔xe8 The exchanges have not eased black's defensive burden, but only increase it because now white's R invades his position. 27.h5 Brilliant! 27...♖f6 28.hxg6 hxg6 Graham Burgess wrote - How would you assess this ending? It may seem that c3-Pawn is a very serious weakness, but it turns out that the g6-pawn is just as easily attacked. Moreover, it is far easier for White to create a passed P on the K-side than it is for black on the Q-side. Thus white should play very actively, rather than trying to defend his Q-side Ps. 29.♖h1 Capablanca does not concern himself with his weakness on the Q-side. Reti observed that now Tartakower makes a defensive move which he could perhaps have omitted. On the other hand, Stockfish agrees with Tartakower. 29...♔f8
4...♗b4+ This was Alekhine's preferred move, but it's not any better than the text. 5.♘bd2 ♘c6 6.e3 O-O
5.♘c3 O-O 6.e3 b6 Black plans to fianchetto his B, an idea often seen in the classical Dutch. 7.♗d3 ♗b7 8.O-O ♕e8 Black plans on transferring his Q to the K-side for attacking purposes, another common idea in the Dutch. This move was Tartakower's improvement on his game against Marshall from an earlier round.
8...♘c6 9.♖c1 h6 10.♗h4 ♘e4 11.♗xe7 ♘xe7 12.♗xe4 fxe4 led nowhere and the game was eventually drawn in Nogueiras,J (2575) -Jussupow,A (2610)/Rotterdam 1989)
8...♘e4 allows white to gain the advantage. 9.♗xe7 ♕xe7 10.♗xe4 fxe4 11.♘d2 ♕h4 12.♘cxe4 ♘c6 13.f4 ♘e7 14.♖f3 ♘f5 15.♖h3 ♕e7 16.♕h5 with a strong attack. Marshall,F-Tartakower,S/ New York 1924. Unfortunately for him, Marshall soon misplayed the position and only drew.
9.♕e2 Threatening to get a space advantage with e4.
9.♗xf6 ♗xf6 10.e4 fxe4 Correct was 10...Qh5 11.♘xe4 ♗e7 Minenko,A (2326) -Yushinov,N/ Berezovskoe 2008. White would have had a slight advantage after 12.Re1
9...♘e4 Capablanca wrote that "Black seeks the safety of simplification" and Alekhine thought black was slightly better after thisd/ 10.♗xe7 ♘xc3 11.bxc3 ♕xe7 Black threatens a Q-side invasion with ...Qa3 which Capa 's next move prevents. 12.a4 ♗xf3 How is this move explained? Graham Burgess suggested that because white was planning to play Rfb1 and c5 undoubling his Ps, black exchanged his B so he can meet c5 with ...bxc5. off his Bishop. It's also been suggested that Tartakower was trying to create a situation where he could use his N to exploit white's doubled Ps and so he created this N vs B imbalance. 13.♕xf3 In the tournament book Alekhine wrote - Now it happens - as usual in mobile P formations - that the B is superior to the N. The rest of the game is a very fine example of the utilization of such an advantage. It's interesting that all the great players and annotators give white the advantage here, but Stockfish and Komodo evaluate the position as equal. Who is right? Life Master AJ Colby said it best - the most impressive thing about this game is the slow, but sure and very careful and methodical, manner in which Capa manages to gradually increase his advantage to the point where it is a real and a tangible influence. 13...♘c6 14.♖fb1 Also worth considering was 14.e4, but Capa is doing the GM thing and playing on both wings. 14...♖ae8 15.♕h3 This prevents ...f4 because of the mate threat. 15...♖f6 This move has been criticized by a number of annotators, but none have offered anything that's really better. Perhaps Tartakower was hoping to use the R to attack on the K-side although with his next move he seems to abandon that whole idea and he concentrates on the Q-side instead.
15...f4 The move he would like to play has been prevented. 16.♗xh7+ ♔h8 (16...♔f7 17.♕h5+ is also mate) 17.♗g6+ mates
16.f4 A dual purpose move: it prevents the advance of black's f-Pawn and the P-break ...e6-e5 is now impossible. 16...♘a5 Other moves have been suggested here, but Stockfish likes this one. 17.♕f3 Now that the Q has fulfilled its purpose on h3 it returns to a more useful position. 17...d6 Flohr and Soltis were critical of this move while Nunn and Burgess weren't. Stockfish is indifferent. (17...c5 Stockfish 18.g3 ♖c8 19.e4 is equal) 18.♖e1 It's work being finished on the b-file, the R moves to the center to support a break by e4 - Chernev 18...♕d7 Eyeing the a-Pawn.
18...e5 Before engines this move was suggested and heavily analyzed, but Stockfish instantly condemns it! 19.e4 fxe4 20.♕xe4 g6 21.fxe5 ♖e6 22.♕g4 dxe5 23.♗f5 Nice! 23...♖f6 24.♗d7 ♖f4 25.♕h3 ♖ef8 26.♖xe5 with the advntage.
19.e4 fxe4 20.♕xe4 g6 21.g3 Highly praised by all the annotators because if stabilizes the f-Pawn and prepares for hsi K-side attack. 21...♔f8 22.♔g2 ♖f7 This has been (correctly it seems) criticized as too cautious.
22...e5 This fully equalizes after 23.fxe5 dxe5 24.♖f1 ♖xf1 25.♖xf1+ ♔g7
23.h4 White is getting a dangerous initiative. 23...d5 But this counterattack in the center is just too risky. Capturing on c4 with the N would have resulted in equality.
23...♘xc4 24.h5
24.♗xc4 also fails to gain any advantage. 24...d5 25.♕d3 dxc4 26.♕xc4 ♖f5
24...d5 25.♕e2 ♖g7 26.h6 ♖f7 27.♗xc4 dxc4 28.♕xc4 ♖f5 with a position almost identical to the one after 24.Bxc4 29...♘c4 This seems to offer black much better defensive chances. For example 30.♗xc4 dxc4 31.♔f3 g5 32.♖h5 gxf4 33.gxf4 ♖f7 34.f5 ♖e7 35.♔f4 ♖e1 and his R has active play.
30.♖h7 ♖c6 31.g4 Here, as Reti put it, "Anxious nature might have moved the King towards the queenside, but Capablanca adheres to the principle of aggression that governs Rook endings." Capa wants to trade his weak c-Pawn for black's g-Pawn. 31...♘c4 Rather than try to defend his Q-side weakness Capablanca gives black the opportunity to win a P because he has confidence in the passed pawn he gets coupled with his control of the absolute 7th rank. 32.g5 This move fixes the g-Pawn and threatens Rh6 followed by f5. 32...♘e3+ Played in an attempt to block the B's diagonal. In fact, this move is good, but it was played with the wrong idea! 33.♔f3 ♘f5 This is the real turning point in the game. Some annotators suggested Stockfish's 33...Nd1, but their analysis was flawed.
33...♘d1 This move saves the game as Stockfish demonstrates. 34.♖h6 ♔f7 35.f5 ♖xc3 36.♔e2 gxf5 37.♗xf5 Stockfish's next move is brilliant as it leads to a position with unbalanced material where black has compensation for the exchange. 37...♘e3 38.♖h7+ ♔g8 39.♖h3 Black's moves don't look right because he loses the exchange, but white won't be able to take advantage of it. 39...♘xf5 40.♖xc3 ♘xd4+ Black's N+2Ps offset the strength of the R! 41.♔d2 c5 42.♖e3 ♔f7 43.♔c3 b5 44.♖e5 bxa4 45.♖xd5 ♔g6 46.♖xc5 ♘f3 47.♔d3 ♘xg5 ...and according to the Lomonosov tablebase this position is drawn.
34.♗xf5 Simple and compelling - Alekhine. Several masters watching this game had divided opinions as to the outcome and Marshall felt that black might even be better! 34...gxf5 35.♔g3 Decisive! White sacrifices material in order to obtain the classical position with King on f6, pawn on g6, and Rook on h7, whereupon the black pawns tumble like ripe apples. -- Alekhine 35...♖xc3+ It is extremely instructive to see how Capablanca is no longer in the least concerned about material equality, but thinks only of supporting his passed pawn. - Reti 36.♔h4 ♖f3 Logical...he wins another P with check, but it's not the best move.
36...a6 While insufficient, the move getting passed P is his best option. 37.♔h5 b5 38.♔g6 Ignoring any "threats" on the Q-side. 38...bxa4 39.♖f7+ ♔g8 40.♖xf5 a3 41.♖xd5 ♖c6+ (41...a2 42.♖d8#) 42.♔h5 ♖d6 43.♖a5 ♖xd4 44.f5 Black's Ps are stymied and white's two connected passed Ps win.
37.g6 ♖xf4+ 38.♔g5 ♖e4
38...♖xd4 is also insufficient. 39.♔f6 ♔e8 40.♖h8+ ♔d7 41.g7 ♖g4 42.g8=♕ ♖xg8 43.♖xg8 ♔d6 44.♔xf5 White has a won ending.
39.♔f6 It is a frequently available finesse in such positions not to capture hostile pawns, but to pass them by in order to be protected in the rear against checks by the rook. -- Reti The P on f6 is useful to white because it shields his K from annoying checks.
39.♔xf5 would be simply awful!! It only draws. 39...♖xd4 40.♖xc7 ♖xa4 This position is drawn.
39...♔g8 40.♖g7+ Driving the black K to a more exposed square. 40...♔h8 41.♖xc7 ♖e8 42.♔xf5 Again the simplest. Kf7 would not yet have been disastrous because of Rd8, etc. - Alekhine. Now that blacks' R has been forced back to the first rank, because of the mating threats white can finally take the P. 42...♖e4 43.♔f6 ♖f4+
43...♖e8 44.♔f7 ♖d8 45.♖xa7 ♖g8 46.♖e7 ♖g7+ 47.♔f6 ♖xe7 48.♔xe7 ♔g7 49.♔d6 ♔xg6 50.♔xd5 ♔f5 51.♔c6 ♔g6 52.d5 ♔g5 53.d6 ♔f6 54.♔c7 b5 55.axb5 ♔e5 56.d7 ♔d4 57.d8=♕+ ♔e4 58.b6 ♔e3 59.♔c6 ♔f4 60.b7 ♔e3 61.b8=♕ ♔e4 62.♕d5+ ♔e3 63.♕g3+ ♔e2 64.♕dd3#
44.♔e5 ♖g4 45.g7+ ♔g8 After exchanging rooks, White would win still more easily. -- Alekhine (45...♖xg7 46.♖xg7 ♔xg7 47.♔xd5 wins) 46.♖xa7 ♖g1 47.♔xd5 ♖c1 Trying to cut off white's K from approaching the Q-side. 48.♔d6 ♖c2 49.d5 ♖c1 50.♖c7 ♖a1 51.♔c6 White need not be concerned about the a-Pawn. 51...♖xa4 52.d6 Black resigned. Capablanca's management of the endgame gives the impression of being so natural that one easily forgets the difficulty of such precise play. The difficulty is chiefly psychological. In chess, as in life, one is so accustomed to place value on the material factors that it is not easy to conceive the idea of indulging in pawn sacrifices when there is so little available material. - Reti
52.d6 b5 53.d7 ♖d4 54.♔b7 b4 55.♔c8 b3 56.d8=♕+ ♖xd8+ 57.♔xd8 b2 58.♖b7 b1=♕ 59.♖xb1 ♔xg7 60.♔e7 ♔g6 61.♖b5 ♔g7 62.♖b6 ♔h8 63.♔f8 ♔h7 64.♖d6 ♔h8 65.♖h6#
Powered by Aquarium
Great performance by Lasker when you consider the field.
ReplyDelete