Is the King’s Gambit playable at the highest level?


In my book I wrote that “over the board it is clear that the King’s Gambit is effective at all levels up to and including 2800+.” Maybe I should revise that to “all levels up to and including 3100+.”

At the recent clash of the best engines in the world (the “TCEC Supermatch”) Stockfish triumphed over Komodo, but Komodo had the consolation of a magnificent win with the King’s Gambit in the final game of the match.

The game is below with very brief comments.
TCEC Season 6 – Superfinal, Round 64, 2014.05.19

White: “Komodo 7x” Elo 3155

Black: “Stockfish 170514” Elo 3164
1.e4 e5 2.f4 exf4 3.Nf3 Be7

The Cunningham – a Hebden favourite and perfectly respectable, but KG players do not lie awake at night worrying about 3…Be7.

4.Nc3 Bh4+ 5.Ke2 d6

“5…d6 is rather slow, and after 6.d4 Bg4 7.Bxf4 White’s powerful pawn centre and strong pieces outweigh his misplaced king.” So says page 352.

6.d4 Bg4 7.Bxf4

I have to agree with page 352.


7…Nc6 8.Qd3 Nge7 9.Kd2 Bf6 10.e5!N

A novelty. Its merits depend on an assessment of an unbalanced queen-versus-three-minors position.

10…Bf5 11.exf6 Bxd3 12.fxe7 Qd7 13.Bxd3 Nxe7

Apparently Komodo liked the pieces while Stockfish was happy enough with the queen. The game suggests Komodo was right.

14.Raf1 O-O 15.Kc1 a6 16.h4 b5 17.h5 b4 18.Nd1 Nd5 19.Bd2 a5 20.Nh4 c6 21.Rf3 Nc7 22.Nf5 Ne6 23.d5!

The d1-knight is the worst-placed piece, but not for long.

23…cxd5 24.Nde3 Nc5 25.Nxd5 Nxd3+ 26.Rxd3

Relying on a fork on e7 to save the f5-knight – lucky.

26…Rae8 27.Rf1 f6 28.h6 g6 29.Nfe3 f5 30.Nc4



31.Rxf4 Rxf4 32.Bxf4 Re1+ 33.Kd2 Re6 34.b3 Qd8 35.Kc1 g5 36.Rg3 Rg6 37.Be3 Kf7 38.Rf3+ Ke6 39.Ncb6 Rxh6 40.Bd4 g4 41.Rf1 Qh4 42.Nc7+ Ke7 43.Kb2 Qg5 44.a4 bxa3+ 45.Kxa3 Qg6 46.c4 a4 47.Ncd5+ Ke8 48.Re1+ Kd8 49.Re7 Qd3 50.Bc3

A beautifully controlled game.


12 thoughts on “Is the King’s Gambit playable at the highest level?”

  1. Very interesting! One only wonders why Stockfish played a variation which does not keep the KG fans awake at night… I would like to see a computer game with the Quaade piece sacrifice 🙂

  2. Very curious. The version of Stockfish I have installed on my laptop likes many other moves more than the passive …d6, for example, and also disputes other moves played here by Black. Is this the case of a human-specified opening book gone bad, or changes in the evaluation function in the engine for the tournament?

  3. I think there’s a big difference between playable lines for humans and engines though. Engines are known to play all kinds of ridiculous moves in the opening to get out of theory and since they don’t have a preference for types of positions, they play them all very good. On the other hand many objectively playable(like only 0.30 behind) positions don’t look very good to the eye and are very hard to play for humans, even for Carlsen. But a 2400 rated engine wouldn’t have a problem with them. Usually top players prefer to play for a win without taking big risks, which is not possible with the King’s Gambit. They sometimes use it as a surprise weapon though but I don’t think any top player is going to play the King’s Gambit as much as other e4 top players play the Spanish.

  4. @Lukas Warts

    Hi Lukas,

    It would be even easier if you wrote out exactly which lines you mean – I am not going to watch a video in German to figure it out.

    After 4.Nc3 I guess 4…Nf6, which is a serious line, but then 5.d4 and then what? White has some good ideas after 5…d5 if I recall the book correctly.

    As ‘SugarLips’ mentioned, the engines were not choosing on their own at this stage – the match was a test of playing strength not opening prep, so they had to play both sides of the same selected opening positions.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top