by Ignacio Marin

Nowadays, it is difficult to recommend the King's gambit. The main reason is the intrinsic strength of the alternative 2. Nf3!. However, the King's gambit is always an interesting surprise weapon How strong is this gambit objectively is still an open question. However, it is obvious that white has to be prepared against the most ambitious defense 2...ef4 3. Nf3 g5. One of the most common answers is the Kieseritsky gambit: 4. h4 g4 5 Ne5. A few ideas in the so called “Berlin defense” 5...Nf6 have attracted certain attention in the last years. I will consider here an old idea that was not tested in grandmaster games until recently.

1. e4 e5 2. f4 ef4 3. Nf3 g5 4. h4 g4 5 Ne5 Nf6 6 d4 d6 7 Nd3 d6 8 Bf4 Qe7

Obtaining the main position of the Berlin defense. There are several important alternatives to be considered carefully by both players before reaching this position (3. Bc4, the different alternatives for black in the third move, 4. Bc4 and 4. d4, 7 Ng4, etc). The play in the King's gambit is very tense in the first moves and a mistake is enough to tip the balance, so a thorough knowledge is required to play this opening with both white and black.

9 Be2!?

Classically, 9 Qe2 is considered the main line. However, Black apparently does not have problems in this line, so any playable alternative deserves attention. Although 9. Be2 can be found in ECO, it has been tested only in a few games. The main alternatives here are 9...Bg7 and 9...Nc6, although 9...g3!? looks interesting to me.


Very logical, black is planning to castle queenside as soon as possible.

10. Nc3!?

This is a recommendation of GM Amador Rodriguez in New in Chess 30. The alternative is 10. c3 but after 10...Bf5 black looks OK. In the game Sanchez Almeyra - Amador Rodriguez, Matanzas 1993 white played 11 d5 (11 Nd2 Nc3) and now Rodriguez suggests 11... Ne5 12 Be5 de5 13 Bg4 Ng3 14 Rh3 (14 Bf5 Nh1 15 Qg4 e4 16 Nf4 Qc5) h5! 15 Bf5 Nf5 with black advantage. That is probably why he suggested 10 Nc3 as an improvement. Now he gives two variations: 10 Nc3 Nd4? 11. Nd5 Qd8 12 Nf2! +- and 10 Nc3 Nc3 11 bc3 Bf5 12 g3 with compensation. In fact after 12 g3 and 13 0-0, it looks to me that white has an excellent position. I suggested a few weeks ago the alternative:

10...Nc3 11 bc3 g3!? N

This move occurred to me while reading the comments in New in chess, without even a chessboard. In fact it is very simple, because its only idea is to interfere with the basic white plan based on 0-0. Obviously if now 12 0-0?? Qh4 is pretty bad. I gave some rapid ideas in my post in that were not totally correct. Moreover, Matt Guthrie found a new and totally unexpected way of answering 11...g3 that I didn't notice in my first analyses. I show here a revised version, that I hope it will stand your tests. Let me know if you find a hole!!.

A) 12 Bg3 Rg8 looks OK for black, white is practically forced to play 13 Kf2 that is visually unexciting, but in fact is not so dangerous. Probably best for both sides is then 13...Bh6 14 Bf4 Qf6 15 Qd2 (15 g3 Be6 16 Re1 0-0-0 with a small black advantage) Be6 (15...Ne7 and 15...Ne5!? are interesting) 16. Bf3 Bf4 17 Nf4 Bc4! 18 Rae1 Kd7 with some black advantage.

B) 12 d5?! is only helping black. After 12...Ne5 13 Bg3 (or 13 Ne5 de5 14 Bg3 Rg8 and white is in trouble because after 15 Bf2 Rg2 the king has no good refuge, while 15 Kf2 is now impossible due to 15...Qc5. 15 Bb5+ Kd8! does not help neither). Rg8 14 Kf2 Bh6 15 Nf4 (15 Bf4?? Nd3) Bg4! 16 Bf3 0-0-0 with advantage.

C) 12 Qd2 is very interesting. In my post, I suggested now 12...Rg8 or 12...Bg4 "with a slight black advantage" this is incorrect in fact

C1) 12...Rg8? 13 0-0-0! leaves black in trouble. For example 13...Be6 14 Rde1 0-0-0 15 Bf3! (threatening d5) 15...d5 16 Nc5 Qe8! (trying to eliminate the beast in c5, but insufficient) 17 Ne6! fe6 18 Qe2 and white has a clear advantage (18...Bd6 18 Bg5 Ba3+ 19 Kd2! Rd6 20 Bf4)

C2) 12...Bg4! is correct. Now best is probably 13 0-0-0! anyway. 13 Bg3 Rg8! is dangerous: white is forced to 14. Kf2 and black can then play 14...Bh6 15 Bf4 Bf4! 16 Nf4 Be2 17 Rae1 0-0-0 with some advantage. However, after 13 0-0-0 white get a satisfactory position: 13...Be2 14 Rde1 0-0-0 15 Re2 Qd7 16 Rhe1 Bg7 17 Qe3! (17 Bg3 Rhe8 =) and white is better. It is doubtful if this is significant after something like 17...b6 18 Bg3 Kb7. It is not easy to see how white is going to use its command of the e column. However this is maybe the best line for white against 11...g3 and at least shows that this move does not refute the King's gambit!.

D) 12. Bg5 I superficially considered this move but I thought that after 12...f6 black was OK. Matt Guthrie found however a very interesting continuation showing that nothing is clear in this position

D1) 13 Bf4 was my idea. By closing the h4 - d8 diagonal white prepares 0-0. However, the lost of a move gives black good chances: 13....Bg4! 14 Nc1!? (14 d5 Be2; 14 Bg3 Rg8 are good for black) 0-0-0 15 0-0 Be2 16 Ne2 Rg8 17 Ng3 Bh6 and black has a splendid position.

D2) 13 0-0!? Matt Guthrie's great idea that nearly refutes 11...g3. Now the next moves are forced: 13...fg 14 Bh5+ Kd8 15 Re1 Qd7 16 Re8 Qe8 17 Be8 Ke8 and after 18 Qh5 Kd8 (Kd7 losses quickly) we obtain a critical position. Now 19 Qg5 Ne7! (but no 19...Be7 20 Qg7 and after the capture of the h7 pawn black is in trouble) looks good for black after 20 Re1 Rg8 21 Qf6 Bg4! (better than 21...Bd7 with idea of 22 Qf7?! Be8! 23 Qh7?? Bg6 because 22 Nf4 is strong). Keeping the bishop in the h3 - c8 diagonal black looks safe and can complete development. Although still unclear, this position is probably better for black. The alternative is 19 Qf7 Be7 20 hg!? (20 Qg7 Re8 21 hg Bf5 or 21 Qh7 gh looks good for black) with a confuse situation although looks also defensible after 20...Rf8 21 Qh7 Bg5 22 Re1 Be7. Long term prospects favor black but there is still some way to go before the last black rook enters the fight.

The conclusions of this analysis is that Amador Rodriguez's idea 10 Nc3!? is interesting and that 11...g3 does not refute it. However, black can defend successfully knowing the main ideas and probably best play leads to a roughly equal ending. 11...g3 can be a very good weapon against an unprepared opponent and its better than Rodriguez's suggestion 11...Bf5. After all these analysis I think however than maybe 9...g3!?! is worth to look at carefully. Maybe next time!.

Send me your comments!!.