Capablanca's Sources

As I am rereading J. R. Capablanca's classic Chess Fundamentals (1921), I marvel at how much content he packs into such a short book. With respect to pawn endgames, in particular, he presents a small number of examples that together constitute a substantial amount endgame knowledge.

When Nate Fewel showed me this position at the Spokane Chess Club in the mid-1990s, I had a copy of Chess Fundamentals, but had not spent much time studying it. I failed to solve this elementary position. When I was glancing through Capablanca's book a short time after Nate showed the position to me and saw it there, I began to to read the text with some diligence. Over the next few years, my endgame skills improved dramatically.

White to move

Capablanca did not compose the exercise. The breakthrough idea with pawns lined up as they are here dates back at least to Carlo Cozio, Il Giuoco degli Scacchi (1766). Endgame Study Database VI by Harold van der Heiden has this position credited to Cozio (see "Endgame Study Database").

White to move

Although I do not know the contents of Capablanca's library, it seems reasonable to believe that he would have had access to Johann Berger, Theorie und Praxis der Endspiele (1891), as it was the standard endgame work of the day. Position number 539 of Berger's book credits J. H. Sarratt, A Treatise on the Game of Chess (1808).

White to move

Sarratt has the White king on h1 and the Black king on g3, which is closer to what Nate showed me. In all cases, the moves are the same. White achieves the breakthrough by advancing the b-pawn, which Black captures. Then, depending on how Black captures, White sacrifices either the a-pawn or the c-pawn, so the other may promote.

Capablanca also notes that Black on the move draws by advancing the b-pawn.

Why Forks Matter

Inherent in all chess tactics are relationships between pieces. A White rook on e1 can attack a Black pawn on e6 if there are no pieces in the way. This relationship is one of offensive contact.* If there is a king shielded behind the pawn on e7 or e8 (a defensive contact), the pawn is unable to capture a White piece standing on f5 or d5. This restrictive contact is called a pin.

If two Black pieces sit on the sixth rank with three squares between them, they are vulnerable to attack from a White rook that can reach the sixth rank. These two enemy pieces can be attacked in a manner we call a pin, a skewer, or even a fork. The rook can attack both because they are on the same line. These attacks are the relationship I created in the first three exercises for my Essential Tactics worksheets.

Such tactical ideas as pins, skewers, and discoveries take place along a single line—rank, file, or diagonal. A fork also can take place along such a line, but forks are more commonly understood to take place along multiple lines. In this exercise from Essential Tactics, for instance, the correct move places the White king in offensive contact with the Black knight along a diagonal and with the Black bishop along a file.

White to move

Consider this game from the forgotten book, A Treatise on the Game of Chess (1808) by J. H. Sarratt (12-15). The notes are by Sarratt, and the book may be found at Google Books. At three critical moments in this illustrative game, White employs a fork. In the third instance, the fork attacks a king and a target square that must be occupied to put a stop to Black's counterplay.

1.e4 e5 2.Bc4 Bc5 3.c3 Nf6

This move, though very generally played, even by good players, is certainly a bad move.

4.d4 exd4

If 4...Bd6 White would win a piece, as will be shown in a Back-game.

5.e5 Ne4

Instead of this move the Black might play the Queen to his King's second square; or his Queen's Pawn two moves; but you would nevertheless have the best of the game, as will be demonstrated in two Back-games beginning at the fifth move of the Black.

5...Qe7;
5...d5

White to move

6.Bxf7+

Instead of this move, you might play your King's Bishop to the adversary's Queen's fourth square; but he would then sacrifice his knight for three Pawns, as will be shown in the second book.

6...Kxf7 7.Qf3+ Nf6 8.exf6 Qxf6

White to move

9.Qh5+

It seems that if you were to give check at the adv. Queen's fourth square, you would likewise win the Bishop; but he would cover the check with his Queen at his king's third square, and giving you check at the same time, you would be compelled to exchange queens.

9.Qd5+ Qe6+ 10.Qxe6+

9...Qg6 10.Qxc5 Qxg2

White to move

11.Qf5+ Ke8 12.Qf3+- and you will win the game, whether the Black exchange Queens or not.

Modern Grandmaster Play

On the way to winning the US Championship, which concluded Sunday, Hikaru Nakamura had this position with the Black pieces against Ray Robson.

Black to move

29...Qc4

Nakamura forks both rooks, although they easily protect each other. Nonetheless, this move brings the queen to a better square for the assault against the White king.

30.Ref6 a3 31.bxa3 bxa3 32.Qxa3

Black to move

Nakamura must have had this position in mind when he played 29...Qc4. White's move 29 that led to a simple fork of rook was to capture a bishop on e5. Nakumura now initiates a combination that recovers the lost minor piece with interest.

32...Rxc1+ 33.Rxc1 Qxe4+

Black is currently down a rook, but White's moves are forced.

34.Kb2 Qe5+

Another fork.

White to move

35.Rc3 Rb8+ 36.Kc2 Qxf6

With an extra pawn and a less vulnerable king, Black had the better endgame. He used a series of checks to win another pawn and force the rooks off the board.


*See Yuri Averbakh, Chess Tactics for Advanced Players (1992) for definitions of contacts as the foundation of chess tactics.

Lesson of the Week

This week, young players were presented a position from a game played by Jacob Henry Sarratt in 1818. Few of Sarratt's games are recorded, and in most cases his opponent is not named. Sarratt wrote chess books and frequented a coffee house in London where he played and taught chess for pay. He has been called England's first chess professional. He called himself the chess professor.

White to move

Here is the whole game with a few comments. Bold type is the game; normal type are comments and possible variations.

Sarratt,Jacob Henry -- NN [C38]
London, 1818

1.e4 e5 2.f4 exf4 3.Nf3 g5 4.Bc4 Bg7

4...g4 5.0–0 d5 (5...gxf3 6.Qxf3 was played in several Anderssen -- Zuckertort battles) 6.Bxd5 gxf3 7.Qxf3 Qg5 8.d4 Bh6 9.Nc3 Ne7 10.e5 c6 11.Ne4 Bg4 12.Qf2 Qg7 13.Nd6+ Kd7 14.Bxf7 Be6 15.Bxf4 Bxf4 16.Qxf4 Na6 17.Bxe6+ Kxe6 18.Nxb7 Raf8 19.Qe4 Rhg8 20.Rf6+ Kd7 21.Rd6+ Kc7 22.Na5 Nb4 23.a3 Rf4 24.Qe2 Rg4 25.g3 Rxg3+ 0–1 Zukertort,J - Richter Berlin 1871

5.h4

5.d4 was played in a pair of McDonnell -- Labourdonnais games.

5...h6 6.d4 d6 7.c3?! 

7.Qd3 Nc6 8.c3 Qe7 9.hxg5 hxg5 10.Rxh8 Bxh8 11.e5 f5 12.Bxg8 dxe5 13.dxe5 g4 14.Nd4 Nxe5 15.Qe2 Bd7 16.Kd1 0–0–0 17.Bb3 Qg5 18.Bd2 c5 19.Ne6 Bxe6 20.Bxe6+ Kb8 21.Kc2 a6 22.Na3 b5 23.Nxb5 axb5 24.Qxb5+ Kc7 25.Qxc5+ Nc6 26.Qxf5 1–0 Marshall,F -- Gunsberg,I Vienna 1903

7...c6

7...Qe7 8.0–0 0–1 Medunova,V (2115) -- Stanek,S (2101) Czechia 2009 (57 moves).

8.Qb3 Qe7 9.0–0 b5? 

9...Nd7

10.Bd3 Bb7

10...a5

11.a4 a6 12.axb5 axb5?

 12...cxb5

13.Rxa8 Bxa8 14.Qa3 Bb7 15.Qa7 Na6 16.hxg5 hxg5 17.Nxg5 Qxg5??

17...Qd7 would have been better 18.Rxf4 Bf6 19.Rf2 Bc8 20.Qxd7+ Bxd7±

18.Qxb7+- Qh4

18...Nb4 19.Bxf4 Qe7 20.Qa8+ Qd8 21.Qxd8+ Kxd8

19.Qxc6+ Ke7

19...Kf8

20.Bxf4 Qh1+ 21.Kf2 Qh4+ 22.g3 Qh2+ 

22...Bxd4+ 23.cxd4 Qf6

23.Ke1 Qxb2 24.Bxd6+ Kd8 25.Qb6+

25.Qa8+ Kd7 26.Rxf7+ Ne7 27.Qb7+ Kxd6 28.Qxe7+ Kc6 29.Qe6#

25...Kd7 26.Bxb5+ Ke6 

26...Qxb5 27.Qxb5+ Kd8 (27...Kxd6 28.Qd5+ Kc7 29.Rxf7+)

27.d5# 1–0

Italian Opening: Developing Theory

For several hundred years, the Italian Opening was among the most heavily analyzed chess opening. Only the King's Gambit had a comparable number of complete games and opening variations in chess books published prior to Handbuch des Schachspiels (1843), the first comprehensive opening encyclopedia. In this volume, as well, the Italian and the King's Gambit occupied more space than other openings.

In Gioachino Greco's well-known and influential texts, the King's Gambit comprises the largest number of games, followed closely by the Italian. Later Italian chess authors--Ercole del Rio, Alessandro Salvio, and Giambattista Lolli--followed suit. Even François-André Danican Philidor (1726-1795), who emphasized positional play over the tactics of the Italian masters, employed the Italian opening in his most important illustrative games.

The second illustrative game in Jacob Henry Sarratt, A Treatise on the Game of Chess (1808) explicates some of the leading attacking ideas in the Italian Opening in the early nineteenth century. I offer here Sarratt's comments interspersed with a few of my own (designated J.S.). Sarratt's main variations are presented as "Back-Games". I have retained his archaic language--"the Black"--and punctuation in most of the comments.

Although Sarratt misses a few key moves, one of which is almost obvious to class players today, his tactical analysis mostly hits the mark.

Second Game [C54]
A Treatise on the Game of Chess 1808
[Sarratt, J.H.]

1.e4 e5 2.Nf3 Nc6

This is the best method of defending the King's Pawn. It will be proved in several subsequent games, that if Black support his King's Pawn in any other manner, he must lose the game.

3.Bc4 Bc5 4.c3 Nf6 5.d4

5.d3 is more popular today. J.S.

Black to move

5...exd4

5...Bb6 6.dxe5 Nxe4 7.Qd5 Bxf2+ 8.Kf1 Black loses a piece; being obliged to castle, or to play his Q. to his K. second square, to avoid check-mate; and in either case, you will take his K.Kt. with your Q. First Back-Game

5...Bd6 6.dxe5 Nxe5 a (6...Bxe5 7.Ng5 0–0 8.f4) 7.Nxe5 Bxe5 8.f4 Bd6 9.e5 and you will gain a piece; but if he were to play his Q. to his K. second square, you must also play your Q. to your K. second square, for if you were to castle, he would give check with his K.B. and then remove his K.Kt. Second Back-Game.

6.cxd4 Bb6

The Black loses the game by this move; he ought to have given you check with that Bishop. This will be analyzed in the second book. J.H.S.

This error appears in Greco's analysis as well. Both Howard Staunton and Wilhelm Steinitz played this move with success, calling into question Sarratt's assessment. Nonetheless, 6...Bb4+ is vastly more popular, and scores better for Black than 6...Bb6. J.S.

7.e5

Black to move

7...Ne4

7...Ng4 8.Bxf7+ (8.h3 is mentioned as an alternative by Sarratt, and is better than his main line (J.S.). 8...Nh6 9.Bxh6 gxh6 10.Qd2) 8...Kxf7 9.Ng5+ Ke8 10.Qxg4 and the White has a good game. Third Back-Game.

7...Nh5 8.Ng5 0–0 (8...g6 9.Nxf7) 9.Qxh5 h6 10.Nxf7 and you will very easily win the game. Fourth Back-Game.

7...Qe7 8.0–0 your situation would have been very advantageous.

Sarratt does not examine 7...d5, which is almost automatic among players today. Staunton played this move in 1853, and the game continued 8.exf6 dxc4 9.fxg7? (9.d5 and White has a better game) 9...Rg8 10.d5 Ne7 and Black won in 37 moves 0–1 Rives -- Staunton,H Brussels, 1853.  Staunton likely realized after this game that Sarratt was correct, and that he prevailed in this game merely because his opponent played badly. J.S.

8.Bd5 f5

8...Ba5+ 9.Kf1.

9.Bxe4

9.Nc3 is superior to Sarratt's recommendation. J.S.

9...fxe4 10.Bg5 Ne7 11.Nh4

Black to move

11...g6

11...d6 12.Qh5+ g6 (12...Kd7 13.Qg4+ Kc6 14.Qxe4+ d5 15.Qc2+ Kd7 [15...Kb5 16.Na3+ Ka5 17.Bd2+ Ka6 18.Qd3#] 16.Nf5 White is clearly winning, but Sarratt's next moves lose quickly. [J.S.] 16...Re8 17.Nxg7 Rg8 18.e6+ Kd6 19.Bf4#) 13.Nxg6 hxg6 14.Qxh8+ Kd7 15.e6+ Kxe6 16.Qxd8. Sixth Back-Game.

11...h6 12.Qh5+ Kf8 (12...g6 13.Nxg6 Rh7 14.Nxe7+ Kf8 [14...Rf7 15.Qxh6 Rxe7 (15...d5 16.Ng6 Qd7 17.Qh8+ Rf8 18.Qxf8#) 16.Qg6+ Rf7 17.Bxd8] 15.Ng6+ Ke8 16.Bxh6 Rxh6 17.Qxh6 d6 18.d5 dxe5 19.Qf8+ Kd7 20.Nxe5#) 13.Ng6+ Kg8 b 14.Nxe7+ Kf8 15.Ng6+ and you will take his Q. with your Q.B. the next move, &c. Fifth Back-Game.

It is hard to understand why Sarratt missed 11...0–0, unless due to his purpose was to showcase Black's errors. J.S.

12.Nf5 gxf5 13.Qh5+ Kf8 14.Bh6+ Kg8 15.Qg5+

Black to move

15...Kf7

15...Ng6 16.Qxd8+ Kf7 17.Qf6+.

16.Qf6+ Ke8 17.Qxh8+ Kf7

17...Ng8 18.Qxg8+ Ke7 19.Bg5#.

18.Qf6+ Ke8

18...Kg8 19.Qg7#.

19.Qf8# 1–0

Although the ChessBase database and other large collections contain analysis games from Greco and Philidor, this gem of attacking play is not in these collections. The two volumes of Sarratt's Treatise, as well as several other books by him, are readily available via Google Books and the Internet Archive. 

Assessing the King's Gambit

François-André Danican Philidor (1726-1795) asserted that with correct play on both sides, the King's Gambit led to a draw. Jacob Henry Sarratt (1772-1819) contested Philidor's assessment. According to Sarratt, Philidor analyzed the art of defense with less care than his labors analyzing the art of attack. Consequently, Philidor believed the player who has the first move should win, while most other authorities asserted that a game played equally well on both sides should be drawn.

For Philidor, the King's Gambit represented inaccurate play, and hence led to a draw. J. H. Sarratt, in A Treatise on the Game of Chess (London, 1808) concurs that the King's Gambit is not to be recommended. But he disagrees that it leads to a draw. He believes the first player should lose.

PHILIDOR says, the the King's gambit, when properly attacked and defended, ends in a drawn game. The members who composed the celebrated Academy of Chess at Naples, after a most careful analysis, gave it as their opinion, that he who plays the gambit ought to lose the game. Experience tends to confirm their decision: the King's gambit is an instructive game, replete with critical and remarkably striking situations, and very few players know how to defend it; but when the defense is correct, he who attacks has indisputably the disadvantage: Salvio's proverb is well known; "Gambitto a giocator: non farsi lice." (xix-xx)

Sarratt's Treatise marks the beginnings of a revival of interest in the playing style of the Italian masters from Gioachino Greco to Alessandro Salvio, Ercole del Rio, and Giambattista Lolli. The principles of the Modenese School favored active piece play. This emphasis differed from the positional ideas of Philidor focused upon pawn play.

Petroff Defense: Early History

The Petroff Defense (or Russian Defense) has an ancient lineage. It is found in several of the oldest books on modern chess, including the works of Luis Ramírez de Lucena, Pedro Damiano, Rodrigo (Ruy) López de Segura, and Giaochino Greco. These early works present model games that are more a record of analysis than a record of play. In these model games, the Petroff often appears as an opening that cannot be recommended. However, there may have been some veneration of this opening much earlier than commonly thought.

Lucena's Repetición de Amores y Arte de Ajedrez con ci Iuegos de Partido (1497) does not seem to have influenced subsequent writers. According to H. J. R. Murray, A History of Chess (1913), "W. Lewis was the first writer to give an account of the work from the point of view of chess, in his Letters on Chess from C.F. Vogt, translated by U. Ewell, 1848" (787). This work was no translation, but a work that the author refused to acknowledge as his own (see Edward Winter, "A chess Watergate" C.N. 4337).

In contrast to Lucena's lack of influence, Damiano's Questo libro e da imparare giocare a scachi et de li partiti (1512) was printed in several countries in the sixteenth century, and had a clear influence upon the work of Ruy Lopez. Several of the games in Libro de la invencion liberal y arte del juego del axedrez (1561) follow those in Damiano's, but diverge in the last few moves. Games from these texts then appear in Greco's work, although carried forward a move or two further.

Such is the case for the oldest Petroff Defense game found in the ChessBase database. In Big Database 2011, game 57 is attributed to Greco, but matches one copied from Damiano by Joseph Henry Sarratt. One needs access to the archives of the world's best libraries just to see copies of the texts of Lucena, Damiano, and Lopez. However, Sarratt, The Works of Damiano, Ruy-Lopez, and Salvio, on the Game of Chess (1813) is widely available even as a free ebook because Google Books scanned a copy from the New York Public Library. Sarratt asserts that he "has frequently and attentively played and examined" the games in the texts of these authors, and he, "is strongly impressed with the belief that they are calculated to assist in a material degree unpracticed players" (xv). Sarratt's reputation has suffered due to his reputation for errors in his texts. Murray notes:

[Sarratt] introduced his generation to the work of the older masters, Damiano, Lopez, and Salvio, in a series of translations. That, as we now know to be the case, these translations were careless, inaccurate, and incomplete, did not rob them of their value at the time they were made, though this discovery has had a very damaging effect on his reputation as a writer. It is unfortunate that the badness of this portion of Sarratt's literary work should have prevented his successors from recognizing the importance and real merit of his other services to chess.
Murray, A History of Chess, 874.

Among Sarratt's contributions cited by Murray was his advocacy that stalemate should be a draw. Through this advocacy, the London Chess Club adopted a rule that was already standard in other countries.


The Games

Sarratt's notation reflects the state of chess notation in English in the early nineteenth century. It is awkward, but readable to the modern reader. The first game in Sarratt's The Works is presented as two variations (1-5). It is Damiano's record of the Petroff as it may have been played in his day. Damiano's two games show hazards that may befall a careless player of the Black pieces. Greco's sole Petroff carries the first of these games two moves further. Lopez's games 35-37 present three variations that are equal or better for Black (Sarratt, 136-141). Lopez's game scores would seem to suggest that the common belief that Black's 2...Nf6 was held in low regard until the mid-nineteenth century may not be fully accurate.

Below are the game scores from these five early studies of tactics in the Petroff (a name the opening would take on in the nineteenth century when it was revived as a viable alternative for Black). The comments are attributed by Sarratt to Lopez.

(1) Damiano,Pedro [C42]
1512

1.e4 e5 2.Nf3 Nf6 3.Nxe5 Nxe4 4.Qe2 Qe7 5.Qxe4 d6 6.d4 f6 7.f4 dxe5 8.dxe5 Nd7 9.Nc3 fxe5 10.Nd5 Qd6 11.fxe5 Qc6 12.Bb5 Qc5 13.Be3 +-

(2) Damiano,Pedro [C42]
1512

1.e4 e5 2.Nf3 Nf6 3.Nxe5 Nxe4 4.Qe2 Qe7 5.Qxe4 d6 6.d4 f6 7.f4 dxe5 8.dxe5 Nd7 9.Nc3 fxe5 10.Nd5 Qd6 11.fxe5 Qc5 12.Be3 Qa5+ 13.Bd2 Qc5 14.b4 Qc6 15.Bb5 Qg6 16.Qxg6+ hxg6 17.Nxc7+ +-

(3) López de Segura,Rodrigo (Ruy)  [C42]
1561

1.e4 e5 2.Nf3 Nf6 3.Nxe5 Nxe4 4.Qe2 Qe7 5.Qxe4 d6 6.d4 f6 7.f4 dxe5 8.dxe5 Nd7 9.Nc3 fxe5

White to move

10.Nb5 Nf6 Black has the best of the Game

(4) López de Segura,Rodrigo (Ruy)  [C42]
1561

1.e4 e5 2.Nf3 Nf6 3.Nxe5 Nxe4 4.Qe2 Qe7 5.Qxe4 d6 6.d4 f6 7.f4 dxe5 8.dxe5 fxe5 9.fxe5 Nd7 10.Bf4 g5 11.Bg3 Bg7 Black will regain his Pawn

(5) López de Segura,Rodrigo (Ruy)  [C42]
1561

1.e4 e5 2.Nf3 Nf6 3.Nxe5 Qe7 4.d4 d6 5.Nf3 Qxe4+ 6.Be2 Bf5 7.c3 Nbd7 8.Nbd2 Qc2 Black has the best of the Game.

It is curious that in Lopez's first game, he reaches the same position as in Damiano's games. But, in Lopez White plays differently. Inasmuch as the move given by Damiano appears stronger, it raises a question that can be answered through examination of the original Lopez text. Did Lopez include also the variation that is favorable to White, and Sarratt exclude it to avoid duplication?