Two traps in the King's Gambit

The King's Gambit, that most romantic of openings, has repeatedly been pronounced dead. Over 80 years ago, Siegbert Tarrasch wrote in The Game of Chess (published in German in 1931, English translation in 1938), "I maintain it to be a decisive mistake." Rudolf Spielmann, "The Last Knight of the King's Gambit," was similarly pessimistic, writing an article at about the same time entitled, "From the Deathbed of the King's Gambit." In 1961, the first article in the inaugural issue of the American Chess Quarterly was Bobby Fischer's famous "A Bust to the King's Gambit," in which he opined, "the King's Gambit is busted. It loses by force."

Yet the King's Gambit, like a monster rising from the crypt, refuses to die. Fischer, a couple of years after pronouncing it dead, started playing it himself, scoring 3/3 against Evans (the editor of American Chess Quarterly!), Wade, and Minić! Today another American chess icon, Hikaru Nakamura, keeps playing it, even against world-class GMs. ChessCafe.com just published a three-part survey on the opening by Tim Harding (Part 1, Part 2, Part 3).

For almost as long as the gambit has existed, the Kieseritzky Variation (1.e4 e5 2.f4 exf4 3.Nf3 g5!? 4.h4! g4 5.Ne5) has been a critical line. Greco used it to thrash poor NN in 1620, over two centuries before its eponym, Lionel Kieseritzky, took up the line. Spassky and Fischer played the Kieseritzky in their famous first encounter at Mar del Plata 1960. (Fischer's experience in that game, in which he stood better out of the opening but lost, inspired him to advocate 3...d6 in his article, aiming for an improved version of the line, since White can no longer play Ne5.)

The Kieseritzky features very principled play by both sides. Black fights to hang onto his extra pawn, even at the cost of weakening his king-side with 3...g5; if unmolested, he will consolidate with ...Bg7 and ...h6. White further compromises his own king-side with 4.h4 in order to thwart this plan. In the olden days White often played 4.Bc4 instead, intending to sac a piece after 4...g4 5.0-0, the Muzio Gambit. This is rarely seen nowadays, since Black need not be so greedy. More solid is 4.Bc4 Bg7! 5.0-0 d6 (5...h6 and 5...Nc6 are also good). One example of this is Van de Wynkele-Semina, 1997, which continued 6.g3!? Nc6! 7.gxf4 gxf4 8.d4 Bg4 9.c3 Qf6 10.Qb3 0-0-0! Black won in 29 with an attack against White's gratuitously weakened king-side.

The game below is a short, sharp encounter between two 19th Century giants. I have altered its move order in order to illustrate a second trap. The game actually began 1.e4 e5 2.f4 exf4 3.Nf3 g5 4.h4 g4 5.Ne5 Bg7 (Paulsen's move, still considered strong today) 6.Nxg4?! (6.d4! is better) d5! 7.exd5?? Qe7+! The changed move order (3...d5 4.exd5 g5? (4...Nf6 is correct)) shows an ill-advised attempt to combine 3...g5 with the Modern Defense 3...d5, a strong and solid defense popular today. (See diagram.) This could have been refuted by 5.Qe2+! (mentioned in Volume 1 of Estrin and Glaskov's 1982 book Play the King's Gambit). Then Black drops a pawn or two after 5...Qe7 6.Nxg5; 5...Be7 6.Nxg5; or 5...Ne7 6.Qe5! Rg8 7.Nxg5; and loses his queen after 5...Kd7?? 6.Ne5+ (6...Ke8/e7 7.Nc6+; 6...Kd6 7.Nxf7+). Amazingly, no one seems to have actually played this very strong move. www.365chess.com has seven games with this line, including one by Alekhine a year before his death. None of the players found 5.Qe2+!

The second trap is seen after 7...Qe7+!, an echo of the first. Once again a check on the King 2 (e2 for White, e7 for Black) spells disaster for the opponent, since an interposition will cause the opponent's piece on King Knight 4 (g5 for Black, g4 for White) to hang. Alternatively, a king move will leave the king fatally exposed. Paulsen had pulled off this same trap in a simul a year before. There he had played 9...Bxg4+ 10.Kxg4 Nf6+ 11.Kh3 Qd7+, when White could still have put up some resistance with 12.g4! fxg3+ 13.Kg2! Qxd5+ 14.Qf3 Qxf3+ 15.Kxf3. Although Black still has a won game after 15...Rg8, surviving to a bad ending is more than White deserves. This time around Paulsen improved with the bone-crushing 9...h5! Then 10.Nf2 Bg4+! quickly induced resignation in Gruzman-Arkanov, Russia (ch) 1978. Mackenzie tried 10.Bb5+, when 10...c6, 10...Kf8 (Fritz's preference), and Paulsen's 10...Kd8 all win with ease. Paulsen checkmated Mackenzie in short order.