Carlos Torre - Friedrich Saemisch [A46]
Moscow International Tournament/Moscow (8) 1925
In his annotations to this game in "Masters of the Chessboard," Richard Reti writes that "it is very characteristic of the style employed by Torre, the way he gradually lags behind...yet knows how to maintain the balance by the counter possibilities of combinations." Torre biographer Gabriel Velasco compares him to his "friend and colleague Frank Marshall" since both were prone toward "cunning swindles that often caught their opponents unawares" as the present game illustrates.
We now call this the "Torre Attack" though it is hardly threatening to Black. Reti writes: "Torre plays the opening in the American style. No serious attempt to obtain superiority in the center, but a healthy development of the pieces. This kind of scheme is not especially profound, and it does not promise any advantage, but Black must be on his guard lest it lead him to make a precipitate advance, as White's powerfully developed pieces can easily bring about a caastrophe." Torre's simple mode of development is still popular today likely for the same reasons he chose it himself: it does not require a lot of opening preparation. Meanwhile, the emerging "Soviet style" of the 20's and later placed emphasis on sharp lines where opening preparation could make a significant difference.
The most common way players met Torre's system at Moscow and generally a good way to meet an early Bg5 or Bf4 by White, since it clears the way for the Queen to go to b6 with pressure against the weakened b2 pawn.
In later games at Moscow 1925, Em. Lasker and Gruenfeld chose instead 4... cxd4 5. exd4 (5. Qxd4!? Nc6 6. Qh4 Be7 7. Nbd2=) 5... Be7 6. Nbd2 d6 with a Scheveningen-Sicilian-style development for Black which today's players would turn into a Hedgehog with b6, a6, Bb7, Nbd7, and Qc7.
6. Rb1!? with a sort of reversed Chigorin Defense.
(White can also play a gambit with 6. dxc5 Qxb2 7. Nb5 Qb4+ 8. Nd2 Na6 9. Rb1 Qxc5 as in Marshall-Steiner, Brno 1928, and now better than Marshall's 10.Bd3 was 10. Bxf6!? gxf6 11. c4 with compensation for the pawn. )
Reti notes that this move seems to give White the advantage in the center, but "Saemisch wanted above all to prevent White's advance in the center later on with e4.
"A very good post for the Knight, as Black would weaken his position if he surrendered the square e5 by Pd5" and would weaken the dark squares generally by Pb5.
Reti calls this "An excellent move, which protects the weak points e6 and f7, opens up the long diagonal for the Bb7, and opens the c-file for Black's pieces."
Reti writes: "It can be seen that Saemisch's profound scheme has borne fruit. White's pieces can find no point of attack and Black already threatens to introduce a decisive initiative with ...f5! Torre's sharp eye perceives the danger and he plays for complications on the kingside. "
A risky idea, but Saemisch felt justified to play for a win given White's rather lackluster choice of opening. He did not reckon on White's reply.
Bogoljubow points out that White likely had in mind the following cunning trap: if Black plays the logical-looking 16... Bxh4 17. Bxh4 Nf4 (which would likely be your computer's preferred line even afterlongdeliberati on), there would follow the stunning 18. Qg4!! (18. Ne3 Nxd3=) 18... Nxd3 19. Bf6 g6 (19... g5 20. Qh5 Kh7 (20... Qc6 21. Ne3 Qe4 22. Qxh6 Qh7 23. Qxg5+ Qg6 24. Qh4 Qh7 25. Qg3+ Qg6 26. Ng4+-) 21. Bxg5 f6 "and from here Torre would certainly have found right way" writes Bogoljubow, without offering it himself. It is by no means obvious: 22. Bxh6!! Rh8 23. Nxd6!! Nxe1 24. Rxe1! and Black will need to surrender tremendous material to stave off mate.) 20. Qh4 Kh7 (20... Qc6 21. Ne3 Kh7 22. d5! with attack ) (20... h5 21. Qg5 Kh7 22. Ne3 d5 23. g3 Nxe1 24. Rxe1 Rh8 25. Be5 f6 26. Qxf6 with attack ) 21. Ne3! h5 22. Ng4 Nf4 (22... Nxe1 23. Qg5) 23. Qg5 hxg4 24. Qxf4 etc.
Perhaps Saemisch should have played 16... f5!? anyway: 17. Ng6 (17. Qh5?! f4 18. Qg6 Rf6 19. Qh7+ Kf7 20. Bg6+ Rxg6 21. Qxg6+ Kf8 ) 17... f4! 18. Bh4 (18. Nxf8 Bxf8 19. Bh4 g5 ) 18... Bxh4 19. Nxh4 f3 unclear
Saemisch was unable to re-evaluate the position and therefore still preferred his chances. He therefore thought he was avoiding the draw that he expected would follow from 17... gxh4 18. Qxh6 f5 (not 18... Nf6? 19. Qg5+ Kh8 20. Qxh4+ Kg7 21. Qg5+ Kh8 22. Qh6+ Kg8 23. Bxd6! Bxd6?! 24. Qg5+ Kh8 25. Qxf6+ Kg8 26. Qg5+ Kh8 27. Qh6+ Kg8 28. Qh7#) However, Torre showed immediately following the game that he could also "have chosen a continuation rich in possibilities" with 19. Rxe6!? (19. Qg6+ Kh8 20. Qh6+= is the draw Saemisch sought to avoid) 19... Nxe6 20. Qxe6+ Rf7! 21. Bxd6 Bxd6 22. Nxd6 though Black certainly has chances himself after 22... Nf4! forcing matters into the ending 23. Qxf7+ Qxf7 24. Nxf7 Kxf7 (24... Nxd3 25. Nd6! Rc7 26. Nxb7 Rxb7 27. b3 unclear ) 25. Bxf5 Re8! when Helestad gave a long analysis (extended by Torre biographer Gabriel Velasco) showing that it is likely Black who has the better winning chances in this complex ending.
18. Rxe6!!+- According to Reti, Black had counted on 18. Nf3 Nf6 19. Qh3 g4 but after 20. Qh4 gxf3 21. Bf4 with attack. Black is still not out of the woods, e.g.: 21... Rh8 22. Nxd6 Nd5 (22... Bxd6 23. Qg3+) 23. Be5+ Kf8 24. Qh5 Bxd6 25. Bxh8 and Black is still in trouble.
[Michael Goeller, based on notes by Richard Reti, Efim Bogoljubow, and Gabriel Velasco]
Game in PGN