Hikaru Nakamura was in a sole lead three rounds before the end of a double-round grandmaster tournament in Biel, Switzerland, last week. Then something went terribly wrong and the 17-year-old U.S. champion lost his last three games. His misfortune helped the veteran Israeli grandmaster Boris Gelfand and the Ukrainian champion Andrei Volokitin, 19, to share first place with six points in 10 games. Yannick Pelletier of Switzerland had five points. Nakamura and Christian Bauer of France ended with 41/2 points. Norway's Magnus Carlsen, 14, finished last with four points.
Hitting the Wall
At age 37, Gelfand is back in the Top 10 on the latest FIDE rating list. With active positional play and an extensive home preparation worked out to the slightest details, he reminds me of the late Soviet grandmaster Lev Polugaevsky. In the last round in Biel, Nakamura decided to challenge Gelfand in one particular variation of the Najdorf Sicilian that the Israeli inherited from his legendary predecessor. It is possible that Gelfand analyzed this line even before Nakamura was born. The American got a lesson he asked for.
1.e4 c5 2.Nf3 (Jumping ahead, Nakamura might have been better off with his almost patented queen-raid 2.Qh5 that he played at this year's World Open. Dutch GM Genna Sosonko recently told me about a St. Petersburg player who attempted to analyze 2.Qh5 seriously. One of his lines continued 2...Nf6 3.Qh4! Nc6 4.Be2 Nd4 5.Bd1! resembling Bobby Fischer's random chess.) 2...d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Bg5 e6 7.f4 Nbd7 (Polugaevsky became known for his adventurous advance 7...b5 that bears his name, but he also knew the line with 7...Nbd7 inside out. Gelfand believes that no one has come close to refuting it.) 8.Qf3 Qc7 9.0-0-0 b5 10.Bd3 (Both 10.Bxb5 and 10. e5 were tried against Gelfand without much success.) 10...Bb7 11.Rhe1 Qb6!? (Initiated by Polugaevsky against his compatriot Efim Geller in Portoroz, Yugoslavia, in 1973. Black attacks the knight on d4, improves his queen and softens the impact of a possible white's knight sacrifice on d5. Geller thought for 90 minutes and sacrificed his knight with 12.Nxe6?!, but after 12...fxe6 13. Qh3 e5! Polugaevsky beat the attack and won in 38 moves.)
12.Nd5!? (The first step of a stunning double knight sacrifice that continues after 12...exd5? by the spectacular 13.Nc6!!, discovered by Geller and his second, Eduard Gufeld, still in Portoroz. The idea is clear after 13...Bxc6 14.exd5+ Be7 15.dxc6 Nc5 16.Bxf6 gxf6 17.Bf5 Qc7 18.b4 Ne6 19.Qh5 Ng7 20.Bd7+ Kf8 21.Qh6! with tremendous pressure as played in the game Chiburdanidze-Dvoirys, Tallinn 1980.) 12...Qxd4! (Fortunately, black can take the other knight.) 13.Bxf6 gxf6 14.Bxb5 Qc5!? (Eliminating one pair of rooks 14...Qxd1+ 15.Qxd1 axb5 16.Nc7+ Ke717.Nxa8 is according to Gelfand in white's favor.)
15.Nxf6+ (Gelfand believes that this move is more challenging than 15.b4, played against him by Alexander Shabalov in Bermuda last year.) 15...Kd8! (The most precise way. The results after 15...Ke7 are disastrous for black. For example, the game Kleiman-Vasquez, Foxwoods Open 2005, finished: 16.Bxd7 Bg7 17.Nh5 Bxb2+ 18.Kxb2 Rhb8 19.Ka1 Kxd7 20.Rd3 Bc6 21.Red1 Qxc2 22.Rxd6+ Kc7 23.Qa3 Rb5 24.Rxc6+ Qxc6 25.Rc1 Rxh5 26.Rxc6+ Kxc6 27.e5 and because of the threat 28.Qf3, black resigned.) 16.Nxd7 (White's attack dies after 16.Bxd7 Be7 17.Qb3 Bxf6 18.Bxe6 fxe6 19.Qxb7 Ra7.) 16...Qxb5 17.Nxf8 Rxf8 18.Qa3 (In 1996 English GM John Nunn assessed this position as slightly better for white. Gelfand has a different opinion.) 18...Rc8!? 19.Qxd6+ Ke8 20.c3 (White should have considered a rook lift to the third rank 20.Re3.) 20...Qc6 21.Qb4 a5! (Deflecting the white queen from the square d6.) 22.Qxa5 (After 22.Qa3 black has an interesting way to exchange queens: 22...Qb6 23.g3 Qb4!) 22...Ra8 23.Qg5 (For the time being preventing 23...Rxa2 because of 24.Rd8 mate.) 23...f6! 24.Qd5?! (A fancy move that results in an inferior endgame. The black pieces will strive on the board even without the queens. The more cautious 24.Qh5+ Rf7 25.a3 gives black the opportunity to get a perpetual check after 25...Rxa3! 26.bxa3 Qxc3+, but there could be more.)
24...Qxc3+! 25.bxc3 exd5 26.exd5+ Kd7 (The black king hides well here, and white has problems defending his pawns.) 27.Kb1 (On 27.d6 comes 27...Rfe8!) 27...Ra4! (Activating his pieces with threats.) 28.g3 Rfa8 29.Rd2 R8a5 30.d6 Be4+ 31.Ka1 h5! (Moving out of danger and threatening 32...Bd5.) 32.h3?! Bd5! 33.g4 (A desperation, but 33.Ree2 h4! destroys the kingside.) 33...Rxa2+ 34.Rxa2 Rxa2+ 35.Kb1 Rh2 36.Re3 (Unfortunately, after 36.gxh5 Rxh3 black keeps one important pawn on the board and white's scattered pawns are an easy prey for the black pieces.) 36...h4! (Fixing the weak pawn on h3.) 37.Kc1 Kxd6 38.f5 Rf2 39.Kd1 Rf3 (The pawn on h3 falls soon.) White resigns.
Solution to today's problem by F. Discart ( White: Ke1,Rf2,Rf6,Bf4,P:g2; Black: Kg1,Rg3,Rh1,Bd5,P:h2): 1.Re2! Bxg2 2.Rg6! Rxg6 (If 2...Rb3 then 3.Rgxg2 mate.) 3.Be3 mate; or 1...Rxg2 2.Rg6! Rxg6 3.Be3 mate; or 1...Rf3 2.gxf3 Bxf3 3.Be3 mate; or 1...Rd3 2.Be3+ Rxe3 3.Rf1 mate.