Lasker -- Capablanca, Game 6

Part 11 of a series recognizing the Centennial of Capablanca -- Lasker, World Chess Championship, Havana 1921.

A notable feature of this World Championship Match is that both players adopted close to the same openings. Through the first five games, we had four Queen's Gambits Declined. Lasker's defensive position differs from Capablanca's, at least so far, because the elder champion has adopted a dubious advance of his b-pawn.

Now, in the sixth game, we see a repeat of the first thirteen moves of game three, but with colors reversed. For commentary on these first thirteen moves, see "Capablanca -- Lasker, Game 3". That post offers a discussion of some of my sources for annotations. These have improved since it was written. I now have Victor Ciobanu's useful English translation of Lasker's book on the match, which I reviewed on Amazon. In addition, I have a PDF copy of Lasker, Mein Wettkampf mit Capablanca (1926), a second printing of the original 1922 text. Not only do these two texts give me a check on the accuracy of the annotations on two ChessBase DVDs, but Lasker's articles for Telegraaf, a newspaper in Amsterdam, have become accessible because he included them in Mein Wettkampf.

Historical Confusion

Databases all seem to have this game played on Wednesday, 30 March 1921 and Friday, 1 April. These dates seem unlikely. Game five concluded on 30 March (see Capablanca -- Lasker, Game 5 [continued]). Are we to believe that after Lasker's blunder and resignation in game five, they would reset the pieces and the clocks and begin the sixth game? And yet, the summary table of dates, openings, and elapsed time that appears in Capablanca's book on the match shows such overlap.

Could the date in all the databases be due to a clerical error, perhaps by Hartwig Cassel, who presumably created the table? I think that is the most reasonable explanation. American Chess Bulletin (April 1921), which also offers stories of the match that were provided by Cassel, states that game six was played 31 March and 1 April. ACB printed information as they received it during and immediately after the match, while Capablanca's book was put together the month after the match. The information in ACB accords with expectations.

Lasker's book does not contain the table, but Ciobanu's edition, which I believe was translated from the Russian edition (it includes an introduction by Peter Romanovsky), includes a table similar to that in Capablanca's text. Some differences are notable: ECO codes rather than opening names are provided, and the dates given for game six agree with American Chess Bulletin.

Because this game was drawn, perhaps, far fewer annotations have been published over the past century than for the decisive games.

Lasker,Emanuel -- Capablanca,José Raúl [C66]
World Championship 12th Havana (6), 31.03.1921

1.e4 e5 2.Nf3 Nc6 3.Bb5

Capablanca played 3.Nc3, so games three and six diverged until Black's sixth move.

3...Nf6 4.0-0 d6 5.d4 Bd7 6.Nc3 Be7 7.Re1 exd4 8.Nxd4 0-0 9.Bxc6 bxc6 10.Bg5 Re8 11.Qd3 h6 12.Bh4 Nh7 13.Bxe7 Rxe7

White to move

14.Qc4

"Up to this point the game was identical with the third. Here Lasker changed the course of the game" (Capablanca).

In the table, game three is presented as Four Knights and game six as Ruy Lopez. Ciobanu's version notes that both games are C66, offering a footnote mentioning the transposition from the Four Knights in game three.

"This drives the black queen to e8 and keeps her away from b6. Meanwhile, the white lady stands exposed here" (Lasker).

Ciobanu's translation has a certain elegance worth noting: "This move kicks the Black queen to e8 and cuts off from b6. However, the White queen is now too advanced" (31).

Was Capablanca's 14.Re3 stronger, weaker, or about the same? Perhaps all that matters is that Lasker's choice is different and that changed the game. Stockfish 13 seems to favor 14.Re3, but the difference among more than half a dozen choices is two or three hundredths of a pawn. Several of those other moves were played over the next few moves.

14...Qe8 15.Re2

"This is artificial. Correct is 14.Re3" (Lasker). 

15...Rb8 16.b3 c5 17.Nf3

Black to move

17...Bb5

"Not the best. Ng5 was the right move. The text move leaves Black with an exceedingly difficult ending" (Capablanca).

18.Nxb5 Qxb5 19.Qxb5 Rxb5

"With this, a draw is marked" (Lasker).

20.Kf1 Ng5

"Black uses the knight very skillfully. Of course, nothing comes out of all of this. You can do almost any maneuver with the minor pieces in balanced positions, as long as they do not change the situation significantly" (Lasker).

21.Nd2 Ne6

"The maneuvers of this Knight are of much greater importance than it might appear on the surface. It is essential to force White to play c2-c3 in order to weaken somewhat the defensive strength of his b-pawn" (Capablanca).

White to move

22.c3 f6 23.Nc4 Nf4

"Again the moves of the Knight have a definite meaning. The student would do well to carefully study this ending" (Capablanca).

24.Re3 Ng6 25.Nd2 Rb8 26.g3 a5 27.a4

"It is now seen why Black had to compel White to play c2-c3. With the White pawn at c2 Black's game would be practically hopeless, since White's b-pawn would not have to be protected by a piece, as is the case now" (Capablanca).

Black to move

27...Ne5 28.f4 Nd7 29.Ke2 Nb6 30.Kd3 c6 31.Rae1 Kf7

White to move

32.Nc4

"An attempt to get the king between enemy pawns, but with insufficient strength" (Lasker).

32.e5 "would have lead to a much more complicated and difficult ending, but Black seems to have an adequate defense by simply playing 32...fxe5, followed by d5 when White retakes the pawn" (Capablanca).

32...Nxc4 33.Kxc4 Re6

"This is the best move, and not Ke6, which would be met by Rd3" (Capablanca).

34.e5 fxe5 35.fxe5 d5+ 36.Kxc5 Rxb3

"The white king has fought his way through. Admittedly, it is not enough for a forced win" (Lasker). 

White to move

37.c4?

"White makes a serious mistake that costs a pawn. After all, 37.Rf1+ would have set more problems for Black. If 37...e7 38.h4 Rg6 39.Rf4 is threatening for White. Black would then probably move 39...h5 40.Rf3. White has the initiative" (Lasker).

"Not the best, but at any rate the game would have been a draw. The best move would have been Rf1+" (Capablanca).

37...dxc4

"Now White must seek a draw" (Lasker).

38.Re4

"Probably the only way to obtain a sure draw" (Capablanca).

Black to move

38...c3 39.Rc4 h5 40.Re3 Rb2 41.Rcxc3 Rxh2 42.Kb6 Rb2+ 43.Kxa5 g5 ½-½

Lasker 2:30 - Capablanca 2:30

"There was not any object for either player to attempt to win such a game" (Capablanca).

Capablanca -- Lasker, Game 5 (continued)

Part 10 of a series recognizing the Centennial of Capablanca -- Lasker, World Chess Championship, Havana 1921. Part 9 has the game up to the diagram position below.

On 30 March 1921, José Raúl Capablanca took his position across a historic table from Emanuel Lasker in a room of the Casino de la Playa. Lasker's sealed move from the previous night's play was revealed on the board and game resumed.

Capablanca,José Raúl -- Lasker,Emanuel [D63]
World Championship 12th Havana (5), 29.03.1921

Black to move

After 31.h4

31...gxh4

I have sequenced the comments to reveal the changing assessments of the sealed move over time.

"[A]pparently the best." (Janowski)

"This was not good. Better 31...Kg6 32.hxg5 Ne4 33.Qd3 Qg4+ 34.Rg2 Qh4 35.Qb1 Kg7. The g5 pawn falls and Black is secure." (Lasker)

In British Chess Magazine (July 1921), George A. Thomas mentioned Lasker's suggestion, adding, "threatening perpetual check" after 34...Qh4.

"This was Lasker's sealed move. It was not the best. His chance to draw was to play Kg6. Any other continuation should lose." (Capablanca)

Myers: "This move (Kg6), Capablanca and Lasker agree, is Black's best chance. Not so. It loses just as readily as the move played in the game, which plainly does not solve Black's problems either."

The ChessBase DVD: Master Class, vol. 04: José Raúl Capablanca (2015), my source for Myers' annotation, also has a question mark after Kg6. Myers, I believe, is NM Hugh Myers (1930-2008), who is best known for innovative opening analysis. The DVD is an excellent resource, but it lists the annotators only by surname, and nothing by Myers is listed in the bibliography.

Garry Kasparov, My Great Predecessors, Part 1 (a book wholly lacking a bibliography) offers more:

It is usual to attach a question mark to [the text move--Kasparov then presents Lasker's analysis, and then:] At first sight here it is indeed impossible to convert the exchange advantage: the White king is exposed, and Black's queen and knight dominate. And yet White has a way to gain an advantage: 36.Qd1! Kg6 37.Qf3! (threatening Qf4) 37...Nxg5 38.Qg3 with good winning chances. So that 31...Kg6 was by no means better than the move in the game. (266)

32.Qxh4 Ng4 33.Qg5+ Kf8

White to move

34.Rf5

Capablanca and Lasker both stated in their books on the match that 34.Rd2 was strong: "quite strong" (Lasker) and "would have won" (Capablanca). Stockfish 13 prefers the text until it has longer to think, when it finds both moves equally strong.

34...h5

"Making use of the slight respite, Black unexpectedly creates counterplay." (Kasparov)

Not 34...Qxe3 35.Qxe3 Nxe3 36.Rf2 and 37.Re2 and White wins. (Lasker)

I. Linder and V. Linder, José Raúl Capablanca: 3rd World Chess Champion add the explanation that the rook, "would cut off the Black king from the queenside and help advance his own king towards the center" (82).

35.Qd8+ Kg7 36.Qg5+ Kf8 37.Qd8+ Kg7 38.Qg5+ Kf8

The triple occurrence of position would be a draw by today's rules. Linder and Linder state, "Black could have claimed a draw". I am not certain that he could. What was the rule in 1921? You will not find the rule in the FIDE Handbook. FIDE was created three years later. 

The rules for the London International Tournament of 1883 specified three-fold repetition of moves, and after the event, it was suggested to modify this to three-fold repetition of position (the modern rule). However, the Fifth American Chess Congress (1889) mentioned six-fold repetition, and William Steinitz's is not explicit about a draw by repetition in The Modern Chess Instructor, published that year. He mentions both perpetual check and repetition of moves.

Edward Winter, "Repetition of Position or Moves in Chess", Chess Notes (updated 30 July 2020) offers some of the critical detail concerning development of the rule, including references to rules governing other World Championship Matches that reference the German Handbuch. The rules for the match in Havana do not reference the Handbuch, nor any other set of general rules.

White to move

39.b3

Stronger was 39.Qxh5 (Kasparov).

The theme of Kasparov's annotations on this game is that Capablanca failed to put forth enough effort to make "detailed calculation of 'dangerous' variations", and that "slight inaccuracies harboured the germ" of his defeat by Alekhine (267).  

39...Qd6 40.Qf4 Qd1+ 41.Qf1 Qd7

White to move

Due to a notation error in Linder and Linder (they have 41...Qd2), they see a mate in nine that Capablanca overlooked. Capablanca's game score in his book on the match has 41...Q-Q2, which matches Lasker's Dd1-d7. 41...Qd7, that is.

42.Rxh5 Nxe3 43.Qf3 Qd4 44.Qa8+ Ke7 45.Qb7+

Black to move

45...Kf8??

What was the status of Lasker's clock?

Capablanca: "A blunder, which loses what would otherwise have been a drawn game. It will be noticed that it was Dr Lasker's forty-fifth move. He had very little time to think and, furthermore, by his own admission, he entirely misjudged the value of the position, believing that he had chances of winning, when, in fact, all he could hope for was a draw."

Lasker: "A terrible mistake. The assumption has been made on various occasions that my gross mistake was the result of a lack of time. But that was by no means the case. I had fifteen minutes to think about it, but I was unable to."

46.Qb8+ 1-0

Following this game, the Cuban Tourism Commission sweetened the prize fund for the match by 5,000 pesos (equal to $5,000 because President Mario García Menocal set the exchange rate of dollars to pesos at 1:1). This detail courtesy of the additional source that arrived in yesterday's mail. Miguel A. Sánchez, José Raúl Capablanca: A Chess Biography (Jefferson, NC: McFarland & Company, 2015), 247.

Capablanca -- Lasker, Game 5

Part 9 of a series recognizing the Centennial of Capablanca -- Lasker, World Chess Championship, Havana 1921.

One hundred years ago today, José R. Capablanca and Emanuel Lasker played the first 31 1/2 moves of a game that would extend to 46 moves over two nights and give Capablanca the first win of the match. Although the move order differed, the game reached the same position at move seven that was seen in the opening game of the match. Capablanca then deviated at move nine, playing the move he suggested, "may be best" in his annotations to that game (see "Capablanca -- Lasker, Game 1").

The game was played in a room of the Casino de la Playa de Marianao, about which more will be said next week. Play began at 9:00 pm and continued until 1:00 am.

Because the game was decisive, it has been annotated in some depth by many chess writers, beginning with Amos Burn in The Field and David Janowski in American Chess Bulletin, and a few other journalists around the world. Garry Kasparov claims that the agreement between the players was the first time exclusive publishing rights for the competitors themselves had been part of a world championship match, but that, "one of the American journalists distributed the game scores right round the world" (My Great Predecessors, Part I, 264). Lasker also distributed the scores during the match with his Monday dispatches to Amsterdam's Telegraaf. These dispatches became the core of Lasker's book on the match, Mein Wettkampf mit Capablanca (1922).

The game was a terrific battle that merits extended study, as "both players displayed brilliant resourcefulness in an unusual position" (Kasparov). Lasker's annotations praise Capablanca's play, while Capablanca notes the depth of understanding displayed in Lasker's performance.

My annotations are supplemented with a small fraction of the extensive comments that have been made by others.

Capablanca,José Raúl -- Lasker,Emanuel [D63]
World Championship 12th Havana (5), 29.03.1921

1.d4 d5 2.Nf3 Nf6 3.c4 e6 4.Bg5 Nbd7

4...Be7 was the move order Lasker employed in game 1, and is the more popular choice. Even so, transpositions are common.

5.e3 Be7 6.Nc3 0-0 7.Rc1

At this point, the position is the same as in game one.

7...b6?!

Lasker criticized this move, calling it of "questionable value" (fraglichem Werte).* It has indeed never become popular at the top levels, The strongest players since Lasker who have played it have been Ljubomir Ljubojevic and Nigel Short. Latvian GM Viesturs Meijers, ranked number 1751 in the world, plays it with some frequency.

8.cxd5

The idea is to make this move before Bc8-b7.

8...exd5

White to move

9.Qa4

In annotations to game one, Capablanca mentioned that this move is stronger than 9.Bb5 (see game 1). 9.Bd3 is more commonly played.

9...c5 10.Qc6

10.Bb5 appears in more games and White's score is better. With this move and the text, the number of games have been too few for statistics to mean anything. 

After 9...c5, Capablanca wrote, "Considered up to now the best answer for Black, but I believe to have had the pleasure of finding over the board in this game the one way to knock it out."

10...Rb8 11.Nxd5

Black to move

11...Bb7

The ChessBase DVD** I have referred to several times in this series presents detailed analysis and about half of the games that continued with 11...Nxd5. The only games with 11...Bb7 are this one, and a simul later in 1921 when Lasker had the White side. That game was drawn. The results of the games beginning with 11...Nxd5 would seem to indicate that Capablanca's claim to have refuted Lasker's idea may have been premature.

These annotations continue 12.Qxd5 Bb7

a) 13.Bxe7 Qxe7 (Bxd5 is no good 14.Bxd8 Bxf3 15.Bxb6 Bxg2 16.Bxg2 axb6 and White is clearly better) 14.Qg5 (14.Qc4 has also been played)...

b) 13.Qf5 Bxf3 14.Bxe7 Qxe7 15.Qxf3 cxd4 16.Rc7 Isakov, K/Tselikov,N -- Alekhine, A/Pertzweig, Moscow, 1914. Won by White in 26 moves.

12.Nxe7+

The intermezzo is forced

12...Qxe7 13.Qa4 Rbc8

13...Bxf3 was played against Lasker in Madrid, and appears to be an improvement.

14.Qa3

Black to move

Capablanca: "This move might be said to be the key of White's whole plan.  The main point is to be able to play Ba6."

This position deserves some detailed analysis. White has a material advantage, but his king could be in a bit of danger. There is long chain of pawns in contact along the a7-f2 diagonal, where c5-d4 threatens to blast open the position while White's king remains in the center. The plan Capablanca says is the key to his whole plan could result in an exchange of bishops, while the bishop pair might otherwise be considered a slight advantage.***

There are several shielding contacts (pins, or opportunities for discovery) that must be observed: the two rooks on the c-file. White's is unprotected. White's e-pawn is pinned by the queen, which itself is a target. White has Black's c-pawn pinned. Two knights--one for each side--are threatened by a bishop. White's knight is defended only by the g-pawn, while Black's knight has three defenders, but cannot move.

14...Qe6

Stepping out of two pins.

15.Bxf6 Qxf6 16.Ba6

Black to move

16...Bxf3

Capablanca: "Dr Lasker thought for over half an hour before deciding upon this continuation. It is not only the best, but it shows at the same time the fine hand of the master. An ordinary player would never have thought of giving up the exchange in order to keep the initiative in this position, which was really the only reasonable way in which he could hope to draw the game."

Lasker noted that after 16...cxd4 17.Rxc8 Rxc8, White had O-O! 

Myers gave 16...Bxa6 17.Qxa6 Qe7 18.O-O Nf6 19.b4! with a clear upper hand for White.

With Lasker's decision to give up the exchange, he seized the initiative and Capablanca was defending a difficult position for many moves. The variations offered in annotations to this game in the several sources seem staggering. 

17.Bxc8 Rxc8 18.gxf3 Qxf3

Matters that I identified after White's move fourteen have been clarifies a bit. White is a clear exchange to the good with poorly coordinated pieces and a vulnerable king.

19.Rg1

If 19.O-O, Black can force a draw should he choose. But, perhaps at that point he would see reason to play for a win. White's king would have little shelter. For instance, 19...Rc6.

19...Re8 20.Qd3

20.Qxa7? Rxe3+! (Euwe)

20...g6

White to move

21.Kf1

Capablanca: "The play here was extremely difficult. I probably did not find the best system of defense. I can not yet tell which was the best defense here, but it is my belief that with the best play White should win."

21.dxc5?! Ne5 (Lasker)

Myers suggests 21.Rg3!?

21...Re4 22.Qd1 Qh3+ 23.Rg2 Nf6 24.Kg1

White has castled, in a certain manner of speaking, but with a rook replacing one of the pawns.

Black to move

24...cxd4 

Only move

25.Rc4

"Ausgezeichnet!" (Lasker) Excellent!

25...dxe3

25...Rg4? 26.Rc8+ Kg7 27.Rxg4 (Lasker)

26.Rxe4 Nxe4 27.Qd8+ Kg7 28.Qd4+ Nf6

White to move

29.fxe3

29.Qxe3 seems tempting to me, but Stockfish 13 assures me that Capablanca's move is better, and that Black would choose to keep queens on the board.

29...Qe6 30.Rf2 g5 31.h4

Black to move

Lasker sealed his 31st move and the game was resumed at 9:00 pm on 30 March.

See "Capablanca -- Lasker, Game 5 (continued)" for the conclusion of this game.

Notes

*Emanuel Lasker, Mein Wettkampf mit Capablanca (Berlin and Leipzig: Walter de Gruyter and Co., 1926 [1922]), 15.

**Additional sources include José Raúl Capablanca, World's Championship Matches, 1921 and 1927 (New York: Dover, 1977); and the ChessBase DVD: Master Class, vol. 04: José Raúl Capablanca (2015). The DVD reproduce most of the annotations by the players in all the match games, and for this game, additional notes by Hugh Myers, Max Euwe, and Dražen Marović. An additional DVD, Master Class, vol. 05: Emanuel Lasker (2015), is a great resource on Lasker, but the annotations on the 1921 match are less extensive than on the Capablanca DVD.

***Capablanca took credit for showing older masters that the bishop is a better piece than the knight.

Lasker -- Capablanca, Game 4

Part 8 of a series recognizing the Centennial of Capablanca -- Lasker, World Chess Championship, Havana 1921.

Game four was played in a single sitting at the Casino de la Playa de Marianao.

Lasker,Emanuel -- Capablanca,Jose Raul [D60]
World Championship 12th Havana (4), 23.03.1921
[Capablanca/Lasker/Stripes]

1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.Bg5

Deviating from game two.

4...Be7 5.e3 0-0 6.Nf3 Nbd7 7.Qc2

7.Rc1 is played about twice as often. Both it and the text score about 57% in master play.

Black to move

7...c6

A development in which black has little freedom of movement. (Lasker)

7...c5 is the proper move (Capablanca).
The theoretical move is 7...c5 (Lasker).

8.Bd3

But that's probably not the right thing to do. Either 8.Rc1; or 8.a3 seem better as long as the c4-pawn is still hanging. For example 8...b6 9.cxd5 cxd5 10.Bb5 Bb7 11.0-0 and white applies a little pressure (Lasker).

8.0-0-0 would have been a much more energetic way of continuing, but probably White did not want to take the risk of exposing himself to a Queenside attack, having then his King on that side of the board (Capablanca).

Four years later, Hans Kmoch tried Capablanca's suggestion and won a long game. 8...dxc4 9.Bxc4 Nd5 10.h4 Re8 11.e4 Nxc3 12.Qxc3 b5 13.Bb3 Bb7 14.Kb1 Qb6 15.Be3 c5 16.d5 exd5 17.Bxd5 Nf6 18.Bxb7 Qxb7 19.e5 Qe4+ 20.Qd3 Qxd3+ 21.Rxd3 Ng4 22.Rhd1 c4 23.Rd7 Nxe3 24.fxe3 Bc5 25.R1d5 Rac8 26.Rb7 Red8 27.e4 a6 28.Ng5 Bf2 29.Nxf7 Rf8 30.Rdd7 Be3 31.Kc2 b4 32.Nd6 Rf2+ 33.Kd1 c3 34.bxc3 bxc3 35.Rxg7+ Kf8 36.Rbf7+ Rxf7 37.Rxf7+ Kg8 38.Kc2 Rc6 39.Rf3 Bd2 40.e6 Rc7 41.Rf7 Rxf7 42.Nxf7 Be1 43.Nh6+ Kf8 44.Nf5 h5 45.e7+ Kf7 46.Ng7 Kxe7 47.Nxh5 Bxh4 48.Kxc3 Kd6 49.Nf4 Bd8 50.Kd3 Ke5 51.Nd5 a5 52.a4 Bh4 53.Ne3 Be1 54.Nc4+ Ke6 55.Nd2 Bh4 56.Nf3 Bd8 57.Kc4 Bb6 58.Kb5 Bc7 59.Kc6 Bg3 60.Kb6 Kd6 61.Kxa5 Kc5 62.e5 1-0 (62) Kmoch,H -- Matisons,H Debrecen 1925

8...dxc4 9.Bxc4 Nd5 10.Bxe7 Qxe7

White to move

This position had been reached in at least four previous games with two White wins and two draws. Black managed to attain a lasting advantage in Maroczy,G -- Alapin,S, Barmen 1905. The last 27 moves were a queen ending with pawns, Black maintaining a one pawn advantage, but not able to deliver the full point. 

11.0-0 Nxc3 12.bxc3 b6 13.Bd3 g6

White won in 64 moves in the other game to reach this position in the 2012 Polish youth championship.

14.a4 Bb7 15.a5 c5

White to move

16.Nd2

This may not have been White's best move. Yet it is extremely difficult to point out anything better (Capablanca).

Too little aggressive! With 16.e4 White could develop the position. For example 16...f5 (or16...Rac8 17.Qb3) 17.Nd2 (Lasker).

16...e5

Probably the only move to save the game. It was essential to break up White's center and to create a weakness in White's game that would compensate Black for his own weakness on the Queen's side of the board (Capablanca).

With this, Black completely balances the game (Lasker).

17.Be4 Bxe4 18.Qxe4 Rae8 19.axb6 axb6 20.Ra7 exd4

White to move

21.Qc6

And this is a useless entanglement. It was just a matter of exchanging queens (Lasker).

21.Qxe7 was slightly better, but Black had, in that case, an adequate defense (Capablanca).

21...Rd8 22.cxd4 cxd4

White to move

23.exd4 

Not 23.Ne4 because of 23...Nb8! (Capablanca)

23...Qf6 24.Qxf6 Nxf6 25.Nf3 Nd5 26.Rb1 f6 27.Kf1 Rf7 28.Rba1 Rdd7 29.Rxd7 Rxd7 30.g3 ½-½

Lasker 2:04 - Capablanca 2:16

There was no reasonable motive to continue such a game, as there was not very much to be done by either player (Capablanca).

The total time used by both players was published in Capablanca's book on the match. However, the time control was 15 moves per hour. Having agreed to a draw after White 30th move, both players should have less than two hours.

Sources

Annotations and game times are from José Raúl Capablanca, World's Championship Matches, 1921 and 1927 (New York: Dover, 1977); and Emanuel Lasker, Mein Wettkampf mit Capablanca (Berlin: Verleger W. de Gruyter & Company, 1922). These are reproduced in two ChessBase DVDs: Master Class, vol. 04: José Raúl Capablanca (2015) and Master Class, vol. 05: Emanuel Lasker (2015). Lasker's annotations are in German.

The Tarrasch Trap

I mentioned the "Tarrasch Trap" in annotations to Capablanca -- Lasker, game 3 in the World Championship match in Havana, March-April 1921. This possibility was noted by American Chess Bulletin as something that Lasker had to avoid in that game. There are, in fact, two Tarrasch Traps in the Spanish opening. The one that is our concern here is sometimes called the Dresden Trap because Siegbert Tarrasch played it against Georg Marco there in a tournament in summer 1892.

In Tarrasch's annotations to this game in Dreihundert Schachpartien (1896), he noted, "it is an exact copy of the analysis given by me in the February edition of the Schachzeitung" (248).* The referenced analysis was in the form of an annotation on the second game of the 1891 World Championship match between Isidor Gunsberg and Wilhelm Steinitz. This was published as "Partieen aus dem Wettkampf Gunsberg-Steinitz", Deutsches Schachzeitung 46, no. 2 (February 1891), 41-56.

Gunsberg,Isidor -- Steinitz,William [C60]
World Championship, New York (2), 11.12.1890
[Tarrasch]

1.e4 e5 2.Nf3 Nc6 3.Bb5 d6

The Spanish game is, theoretically at least, difficult to defend, because in most variants White has the advantage that his king's bishop is better than Black's. Steinitz currently considers the present defense to be the best; However, it gives the player a rather cramped and pressed game, which is why I prefer the usual move Nf6.

4.c3

A stronger attack is 4.d2-d4; if Black takes the pawn, attractive play develops after Nd4: with Nc3, O-O and Bb2!, which gives White an excellent game (cf. the game Tarrasch - Blackburne, Schachzeitung, November 1890). It is best to omit the exchange on c6 completely in order not to remove the tension, but can also (exceptionally) take advantage of the king's bishop, because in this type of game the knights will prove to be very strong--occasionally one gets to f5 and from there already threatens the position of the Black king, which is also threatened by the open line of the queen's bishop.

But if Black avoids the pawn swap in the fourth move, then, according to Steinitz, the following swap itself leads to a position that is advantageous for White: 4...Bd7 5.Nc3 Nf6 6.0-0 Be7 7.Bxc6.

This variant can be made even more compelling 7.Re1! 0-0 8.Bxc6 Bxc6 9.dxe5 dxe5 10.Qxd8 Raxd8 11.Nxe5 Bxe4 12.Nxe4 Nxe4 13.Nd3 f5 14.f3 Bc5+ 15.Nxc5! Nxc5 16.Bg5 and White wins the exchange through the threat Be7.

Also through 16...Rd5 nothing is to be changed because of 17.Be7 -- with the idea 18.c4. Tarrasch offers detailed analysis after 15.Kf1 instead of 15.Nxc5: 15...Bb6 16.fxe4 (16.Nf4 Nd2+) 16...fxe4+ 17.Nf4 g5 18.Rxe4 gxf4 19.Ke2 (19.Bxf4 Rd2; 19.Rxf4 Rd1+ 20.Ke2 Rxf4 21.Kxd1 Rf2 and Black is better) 19...Rfe8 20.Rxe8+ Rxe8+ 21.Kf3 Re1 22. -- threatening Be3.

The rest of Gunsberg -- Steinitz is not our present interest. Nor are Tarrasch's annotations to the rest of this game, which are detailed.

7...Bxc6 8.dxe5 dxe5 9.Qxd8+ Rxd8 10.Nxe5 Bxe4 11.Nxe4 Nxe4 12.Re1] 4...Bd7 5.0-0 Nge7 6.d4 Ng6 7.d5 Nb8 8.Bxd7+ Nxd7 9.Na3 Be7 10.Nc2 Nc5 11.Qe2 Qd7 12.b4 Na4 13.Bd2 0-0 14.c4 f5 15.exf5 Qxf5 16.Rac1 Rae8 17.Nfe1 Bg5 18.g3 Nc3 19.Bxc3 Bxc1 20.Ng2 Qf3 21.Qxf3 Rxf3 22.Nge3 Bxe3 23.Nxe3 Ref8 24.Kg2 c6 25.Bb2 cxd5 26.Nxd5 Rd3 27.Bc1 b5 28.Ne3 bxc4 29.Nxc4 Rd4 30.Ne3 Rxb4 31.Rd1 Rb1 32.Ba3 Rxd1 33.Nxd1 Rd8 34.f3 d5 35.Nc3 d4 36.Ne4 Rb8 37.h4 h5 38.Kf2 Rb1 39.Bd6 Rb2+ 0-1

We now look more closely at the annotations in Deutsches Schachzeitung through Tarrasch -- Marco, Dresden 1892, a game played eighteen months after the analysis was published. This game followed the main line of Tarrasch's 1891 annotations, and he later reproduced the variations mentioned in 1891 in Dreihundert Schachpartien.

Tarrasch,Siegbert -- Marco,Georg [C66]
DSB Kongress-07 Meisterturnier Dresden (7.3), 22.07.1892
[Tarrasch]

Annotations are from Dreihundert Schachpartien, except two references to Chess Informant's Encyclopedia of Chess Miniatures.

1.e4 e5 2.Nf3 Nc6 3.Bb5 d6 4.d4 Bd7 5.Nc3 Be7 6.0-0 Nf6 7.Re1

Black to move

7...0-0

After this move Black is lost. Better is 7...exd4 but it is not good enough to equalize. Annotations of this game in Encyclopedia of Chess Miniatures (2015) continues: 8.Nxd4 0-0 9.Nxc6 bxc6 10.Bf1with a slight advantage for White.

8.Bxc6 Bxc6

8...bxc6 Black loses a pawn.

9.dxe5 dxe5 10.Qxd8 Raxd8

10...Rfxd8 11.Nxe5 Bxe4 12.Nxe4 Nxe4 13.Nd3 f5 14.f3 Bc5+ 15.Kf1 White would have a plus (see below at annotations to move 15). Encyclopedia of Chess Miniatures offers this line as concluding with a decisive advantage for White.

11.Nxe5 Bxe4 12.Nxe4 Nxe4

White to move

13.Nd3 f5 14.f3 Bc5+

White to move

15.Nxc5

15.Kf1 would yield no advantage because of 15...Bb6 16.fxe4 (Or 16.Nf4 Nd2+) 16...fxe4+ 17.Nf4 g5 18.Rxe4 gxf4

Analysis diagram

19.Ke2 (19.Bxf4 Rd2; 19.Rxf4 Rd1+ 20.Ke2 Rxf4 21.Kxd1 Rf2) 19...Rfe8 20.Rxe8+ Rxe8+.

15...Nxc5

White to move

16.Bg5 Rd5 17.Be7

 Not 17.c4 at once because of 17...Rd7 18.Be7 Nd3

1-0 Black resigned here as on 17...Re8 (or 17...Rf7) White wins the exchange by 18.c4. This game is an important contribution to the refutation of Steinitz's 3...d6. It is an exact copy of the analysis given by me in the February edition of the Schachzeitung.

As noted in Friday's post, other books offer annotations of this game. There was also a lively discussion several years ago at chessgames.com.

*Siegbert Tarrasch, Three Hundred Chess Games (Dreihundert Schachpartien), trans. Sol Schwarz (1999).

Capablanca -- Lasker, Game 3

Part 7 of a series recognizing the Centennial of Capablanca -- Lasker, World Chess Championship, Havana 1921.

At first glance, I was not certain what to make of this game. The Four Knights Opening transposed into a Steinitz Defense in the Spanish and it was hard for me to see much opportunity for either player. The middlegame seemed to show a lot of fruitless maneuvering. But, as I spent more time playing through the game and reading annotations, tactical threats and positional nuances became evident.

I started with annotations by both players, and then looked at those in British Chess Magazine by George A. Thomas and the unsigned annotations in American Chess Bulletin. I have the Dover reprint of Capablanca's book on the match, but rely on two ChessBase training DVDs (Master Class vols, 4 and 5) for Lasker's notes. These are in German, which I do not read well.

At one time, I had a decent shelf of German language books, including an excellent dictionary and a couple of grammars. I had two years of high school German, and another year in graduate school. These books were in a box in a closet where a cat's water dish overflowed, ruining the books. Hence, I must rely on Google Translate for unfamiliar words, of which there are plenty.

Game three was played over three days at the Marianao Casino in Havana, mostly on Saturday and Sunday, March 19-20, and then concluded Tuesday, March 22. Monday was a rest day that Lasker ordinarily used to fulfill some journalistic commitments in the Netherlands.

Capablanca,Jose Raul -- Lasker,Emanuel [C66]
World Championship 12th Havana (3), 19.03.1921
[Capablanca/Lasker/Stripes]

1.e4 e5 2.Nf3 Nc6 3.Nc3

In both the Spanish and the Italian, I'd rather put a pawn on c3. As a consequence, I don't think I've played the White side of the Four Knights in this century, except perhaps in a training game with a student. Youth players seem to favor the opening, especially the Italian Four Knights, which FM Jim Maki calls the Spokane Scholastic Opening.

3...Nf6 4.Bb5 d6

4...Nd4 According to I.A. Horowitz, Chess Openings: Theory and Practice (1964), the Rubinstein variation "drove the Four Knights into disrepute" (20). Nonetheless, "there are some finesses to know and some problems not to be underestimated" (13).

Spielmann had credited the Rubinstein variation as, "the main reason why the Four Knights is now played so rarely" (as quoted in Donaldson and Minev, Akiba Rubinstein: Uncrowned King [1994], 195). Donaldson and Minev go on to credit Nigel Short and John Nunn with reviving the opening with new ideas for White.

5.d4 Bd7 6.O-O Be7 7.Re1

Black to move

7...exd4

American Chess Bulletin points out that 7...0-0 allows White to spring the "Tarrasch Trap", an instructive sequence of exchanges that appeared in Tarrasch -- Marco, Dresden 1892. The game is annotated in Tarrasch, 300 Games of Chess; Tartakower and DuMont, 500 Master Games of Chess; and Reti, Masters of the Chessboard. All of these books are on my shelf, and the annotations are reproduced in ChessBase Mega 2020. Perhaps, I'll create a separate post on the game.

That game continued 8.Bxc6 Bxc6 9.dxe5 dxe5 10.Qxd8 Raxd8 11.Nxe5 Bxe4 12.Nxe4 Nxe4 13.Nd3 f5 14.f3 Bc5+ 15.Nxc5 Nxc5 16.Bg5 Rd5 17.Be7 and Marco resigned. The entire sequence has appeared several times in master play since with a scoring percentage for White near 90%.

8.Nxd4 

Lasker had this position twice in his World Championship match with Tarrasch in 1908. He also had it against Janowski in a match a year later that was not an official World Championship match.

8...O-O

8...Nxd4 9.Qxd4 Bxb5 10.Nxb5 0-0 11.Bg5 h6 12.Bh4 Re8 13.Rad1 0-1 (41) Tarrasch,S -- Lasker,E, Duesseldorf & Munich 1908.

9.Bxc6

9.Nxc6 Bxc6 10.Bxc6 bxc6 11.Ne2 Qd7 12.Ng3 Rfe8 13.b3 Rad8 14.Bb2 Ng4 0-1 (41) Tarrasch,S -- Lasker,E, Duesseldorf & Munich 1908.

9...bxc6

White to move

10.Bg5

More popular today are 10.Qf3 and 10.Bf4.

10...h6 11.Bh4 Re8 12.Qd3 Nh7 13.Bxe7

An old move, generally played by all the masters. I believe, however, that 13.Bg3 is the best continuation (Capablanca).

13.Bg3 was played against Capablanca two years prior to this match. 13...Nf8 14.Rad1 Bf6 15.Rd2 Qb8 16.f4 Bxd4+ 17.Qxd4 Qxb2 18.Rb1 Qa3 19.Rb3 Qa5 0-1 (32) Cole,H -- Capablanca,J Hastings 1919.

13...Rxe7

White to move

14.Re3

14.Qc4 will come up in game six.

14...Qb8

This move gives the game its character. Qxb2 is not threatened directly because Ra1-b1, Qb2-a3, Nc3-d5 would then lead to material superiority, but the move provokes b2-b3 and loosens White's position (Lasker).

15.b3

Unnecessary at this point, since Black cannot take the pawn (Capablanca).

15.Rae1 Qxb2? 16.Rb1 Qa3 17.Nd5 White wins material.

15...Qb6 16.Rae1 Rae8 17.Nf3

White wants to proceed with e4-e5 in order to weaken the point c5 (Lasker).

A nice point about the two ChessBase training DVDs is effective use of the software's capabilities for highlighting squares.

Black to move

17...Qa5 18.Qd2

Now Nc3-d5 is threatened (Lasker).

Black to move

18...Ng5

A very good move, which gives Black the better position (Capablanca).

The mutual respect that Capablanca and Lasker had for one another often reveals itself in their annotations.

19.Nxg5

19.Nd5 no longer works 19...Nxf3+ 20.gxf3 Qxd2; 19.Qd3 Nxf3+ 20.Rxf3.

19.Nd4 might have been worth considering.

19...hxg5 20.h3

I would have preferred to try to use the h-file.

This move is necessary because otherwise Black would paralyze the white pawns with g5-g4 (Lasker).

20...Re5 21.Rd1

Black to move

21...Bc8

Admittedly the energetic 21...f5 is bad, not because of 22.exf5 (But White would not do the Black the favor of exchanging. And after 22.f3 the g5 pawn would be weak. For example 22...d5 [or 22...f4 23.Rd3 and now White occasionally threatens h4, and Black is more on the defensive] 23.Rd3) 22...Rxe3 23.fxe3 Bxf5 where the weakness of c2 and e3, even after 24.e4 Bg6 Black gives good prospects (Lasker).

Now 21...g4 was the stronger continuation. For example 22.hxg4 Bxg4 23.f3 Bc8 24.Qd4 Qb6 25.Red3 f5 (Lasker).

21...f5 22.f3 (22.exf5 Rxe3 23.fxe3 Bxf5) 22...g4 (22...d5 23.Rd3; 22...f4 23.Rd3) (Lasker).

21...g4 22.hxg4 Bxg4 23.f3 Bc8 24.Qd4 Qb6 25.Red3 f5 (Lasker).

22.Rd3 Qb6 23.Kh2 R8e6 24.Rg3

White wants to open lines for the rooks with f2-f4, Black prevents the maneuver by counterattacking f2 (Lasker).

24...Rf6

Artifice and no end! (Lasker).

Why not 24...f6 If White then answers 25.f4, he is at a disadvantage after 25...gxf4 26.Qxf4 Rh5. (Lasker).

White to move

25.Kg1

25.f3 would have been answered by 25...Qc5 (Lasker).

25...Kf8 26.Na4 Qa5 27.Qxa5 Rxa5

After exchanging queens, Black has no chance of winning, provided that White goes his straight way without fear or illusion. There is no question that Capablanca is the right man to do this (Lasker).

28.Rc3

Played under the impression that Black would have to play 28...Bd7. Since Black can play the text move, it would have been better for White to have played 28.c4 (Capablanca).

28...Bb7 29.f3 Re6 30.Rcd3 Ba6 31.Rd4 f6

White to move

32.Rc1

Perhaps this was the adjourned position. My sources do not indicate the point where the game was adjourned at 1:00 am Sunday morning to be resumed at 9:00 pm.

32.c4 would be an error, as Thomas noted in British Chess Magazine. 32...c5, rook moves, and Bxc4.

32...c5 33.Rd2 Bb5 34.Nc3 Bc6 35.a4

White now threatens to lock up Black's rook with Rc1-a1. If then Ra5-a6 then a4-a5 (Lasker).

35...Ra6 36.Kf2 Rb6 37.Nd1 Kf7

White to move

38.Ne3 Rb8 39.Rh1 Ree8 40.Rdd1 Rh8 41.g4 Bd7 42.Nd5 Rb7 43.Kg3 Rh4 44.Rd3 Be6 45.c4

Black to move

White purposely makes weak pawns on b3 and h3 in order to prevent the advance of c5 and g5 just in case. With that he blocks the position and reaches a draw (Lasker).

The rest of the game seems almost futile. However, perhaps such maneuvering against lesser players could provoke frustration and error.

45...Rh8 46.Rc1 Ke8

White to move

47.Ne3

47.f4 would be a blunder 47...c6 48.Re3 (48.Ne3 gxf4+ 49.Kxf4 Rxh3 50.g5 (50.Rxd6 g5#) 50...fxg5+ 51.Kxg5) 48...Kd7 (48...cxd5 49.exd5) 49.fxg5 (49.Nc3 gxf4+ 50.Kxf4 Rxb3) 49...fxg5 50.Nc3 Rxb3.

47...Kd7

White has slightly more space, but the backward pawns are targets. Black's bishop guards against intrusions by the knight.

48.Ng2 Rbb8 49.Re1 Kc6 50.Ne3 Rbe8 51.Rb1 Rh7 52.Rd2 Rb8 53.Rd3 Rbh8

White to move

54.Rh1 Kb6 55.Rh2 Kc6 56.Rh1 Rb8 57.Rh2 Rf8 58.Rh1 Kd7 59.Rh2 Bf7 60.Nf5 Rfh8 61.Ne3 Ke6 62.Nd5 Rc8

White to move

Here, the game was adjourned a second time.

63.Ne3 ½-½

Lasker had the initiative for much of the game, but could not provoke a decisive error by Capablanca.

Lasker -- Capablanca, Game 2

Part 6 of a series recognizing the Centennial of Capablanca -- Lasker, World Chess Championship, Havana 1921.

Although Emanuel Lasker and José R. Capablanca enjoyed a quiet audience while playing in a large room at the Union Club, starting with the second game they played in a separate room at the Marianao Casino. Only the players, Judge Ponce, and each player's second were permitted to remain in the room. One additional person came and went for the purpose of ferrying the moves as they were played so the audience elsewhere in the casino could follow the game.

In addition to notes by the players, I have added some that appear in Eugene Znosko-Borovsky, The Middle Game in Chess (1938). Referring to an earlier edition of this text, Capablanca wrote, "It is about the only book on [the middle game] that I consider worth reading" (A Chess Primer [1935]).


Lasker,Emanuel -- Capablanca,Jose Raul [D37]
World Championship 12th Havana (2), 17.03.1921
[Capablanca/Lasker/Znosko-Borovsky]

1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.Nf3 Nbd7 5.e3

"On general principles it is better to bring the Queen's Bishop out first" (Capablanca).

5...Be7 6.Bd3 0-0 7.0-0 dxc4 8.Bxc4 c5 9.Qe2 a6

White to move

10.Rd1

10.a4 is most often played today.

10...b5 11.Bd3 Bb7 12.e4

"Played in order to develop the Queen's Bishop and thereby condemning his whole plan of development, since he could have done that before, as indicated in the previous note, and the only reason he could have had for playing e3 on the fifth move would have been to develop this Bishop via b2" (Capablanca).

12...cxd4 13.Nxd4 Ne5

White to move

14.Nb3 

"Combinations beginning with 14.Bxb5 are wrong, viz., 14...axb5 15.Nxe6 fxe6 16.Rxd8 Raxd8 17.Qxb5 Nxe4 and Black has a won game" (Capablanca).

"The text move is daring. The normal move would be 14.Bc2" (Lasker).

14...Nxd3 15.Rxd3 Qc7

White to move

16.e5

"This move by White releases d5 and therefore makes Black's game easier. Meanwhile, White has no choice" (Lasker).

"White could not play 16.Bg5 because of the rejoinder 16...Nxe4" (Capablanca).

"This opens the long diagonal, and now the White's g-pawn will be under fire from Black's Q and Q B. The threat of a bishop at c5 will prevent the defensive f3 by White. How is this future menace to be parried?" (Znosko-Borovsky, Middle-Game, pp.64-65)

"Should White play 16.f3? Then 16...a5 is strong, because of the possibility of check from b6 gives the pawn on b5 enough cover. As a result, White's game will certainly get into disarray" (Lasker).

After 16.f3, Znosko-Borovsky gives 16...b4 17.Nd1 a5 "At present nearly all the white pieces are badly placed."

16...Nd5 17.Rg3 Nxc3 18.Rxc3

"18.bxc3 was seductive. Black's best answer is probably 18...Kh8 19.Be3 Be4 (19...Qxe5? 20.Bd4) 20.Bd4 Bg6 and Black has an attack on the queenside while his king is protected" (Lasker).

18...Qd7

"It was my impression that, after this move, Black had a very superior game" (Capablanca).

19.Rg3 Rfd8 20.Bh6 g6

White to move

"White, who has achieved his first object, has not yet obtained a satisfactory position, for his knight is still badly placed and the a1 rook is not yet in play" (Znosko-Borovsky).

21.Be3

To provoke Nb3-c5. (Lasker)

21...Qd5

"This leads to the exchange of one of the two Bishops, but it would be very difficult to find a better move" (Capablanca).

22.Na5 Rac8 23.Nxb7 Qxb7

"Black threatened Qb7-e4" (Lasker).

"The position is now simplified, but White still has to cope with some difficulties" (Znosko-Borovsky).

24.Bh6 Qd5 25.b3

Black to move

25...Qd4

"It was probably here where Black failed to make the best move. Instead 25...Bb4 was the better move" (Capablanca).

26.Rf1 Rd5 27.Re3 Ba3

"27...Bf8 was better, as White could not very well afford to take the Bishop; he would be compelled to play first 28.Re4 to be followed later on by Bf4" (Capablanca).

28.g3 Qb2 29.Re1 Rc2

"29...Qxe2 followed by ...Bb4 was the proper course to follow" (Capablanca).

30.Qf3 Be7

"This was my thirtieth move. I was very much pressed for time and I could not make the necessary analysis to find out whether 30...Bf8 would have been a winning or a losing move. If 31.Bxf8 Kxf8 32.Qf6 Kg8 33.h4 and Black would have a very difficult position to defend" (Capablanca).

There must have been a sealed move and an adjournment here or within a couple of moves, but the sources I have examined do not indicate when the game was adjourned.

White to move

31.R3e2

"Covers a2" (Lasker).

31...Rxe2

"31...Qxa2 now would lose because of 32.Rxc2 Qxc2 33.Rc1 followed by Rc8+, etc." (Capablanca).

"Because following 31...Qxa2 is 32.Rxc2 Qxc2 33.Rc1" (Lasker).

32.Rxe2 Qb1+ 33.Kg2 Bf8 34.Bf4 h6 35.h4 b4

White to move

36.Qe4

"36.Qe3 would lead to 36...Rd1" (Lasker).

36...Qxe4+ 37.Rxe4 Kg7 38.Rc4 Bc5 39.Kf3

Black to move

39...g5

"With 39...a5 40.Rc2 Bd4 41.Rd2 Bc3 42.Rxd5 exd5 43.Be3 would favor White" (Lasker).

40.hxg5 hxg5 41.Bxg5  ½-½

Lasker 2:36 - Capablanca 2:37

George A. Thomas wrote at the end of his annotations in British Chess Magazine, "A most interesting game. ... Nevertheless, the impression remains that the Lasker of ten years ago would have made more of the mid-game position."