18

Besides the Queen's Gambit (which isn't really a gambit anyway), which gambits have been most successful as far as winning percentage and usage at high levels of play?

ETD
  • 23,592
  • 4
  • 95
  • 169
Andrew Latham
  • 3,691
  • 2
  • 23
  • 38
  • I'm not sure which one is best, but it is definitely not the king's gambit. Fischer clearly refutes the line with his move `3... d6`: `1. e4 e5 2. f4 exf4 3. Nf3 d6` – Travis J Jun 21 '12 at 18:29
  • @TravisJ: Carslen does not seem to share your view, though. See http://www.365chess.com/game.php?gid=3733824 . – thb Jun 22 '12 at 15:26
  • @thb - Perhaps Yue should review Fischer's games with regards to the king's gambit. Clearly exf4 is winning a pawn and Yue blundered by not accepting in my opinion. A KGD is a good position for white. – Travis J Jun 22 '12 at 18:23

2 Answers2

15

According to these statistics from 4 million games, the Marshall Gambit of the Semi-Slav defense is the most successful with 58.4% wins. I only counted gambit openings that had more than 1000 occurrences in the database.

The least successful gambit appears to be the Latvian Gambit with only 41.1% wins.

Note that since the database does not discriminate against older games, the statistics may not necessarily reflect the possible outcome of an opening if it were played today.

ETD
  • 23,592
  • 4
  • 95
  • 169
sam hocevar
  • 656
  • 6
  • 7
  • So, you are basically saying that the Queen's Gambit is the most successful (as the Marshall is really just Queen's Gambit Declined or equivalent due to transposition). – Travis J Jun 21 '12 at 19:50
  • 1
    @TravisJ, that's not what Sam's answer is saying. As Andrew points out in the question, the Queen's Gambit in and of itself is not truly a gambit; but the Marshall Gambit is one way to proceed in true gambit fashion if Black declines the QG with the Semi-Slav Defense. – ETD Jun 22 '12 at 14:39
  • @EdDean - At that point, you opened with a queen's gambit and then are hoping to transpose to this system. The opening still goes towards a queen's gambit. To note, Nakamura plays 1. d4 almost exclusively so I am not arguing against starting with that move, but if you follow with 2. c4, that is a queens gambit period. Everything else afterwards depends on the defense played and what you transpose to. – Travis J Jun 22 '12 at 18:25
  • 1
    @TravisJ, I agree that, in the case of the Marshall Gambit, the game opens as a QG. Then Black makes it a Semi-Slav Defense, and then White can play the Marshall Gambit. The original question is about how particular gambits fare statistically, and the 58.4% winning statistic that Sam gives *is for the Marshall Gambit specifically*, not for the QG more generally (which Andrew wasn't interested in asking about). That's why Sam's answer isn't "basically saying that the Queen's Gambit is the most successful" as you suggested in your first comment. – ETD Jun 22 '12 at 18:38
  • @EdDean - The first line of the question is "besides the queen's gambit" making this question more about gambits which do not include white starting with d4 c4. – Travis J Jun 22 '12 at 18:49
  • 1
    @TravisJ: I can understand why someone might read it the way you did, but actually, "besides the Queen's Gambit (which isn't really a gambit anyway)" means that the OP wanted to consider lines that include an actual pawn sacrifice (rather than just the pseudo-sacrifice which `2. c4` offers), but does *not* mean to rule out all further (true) gambit lines that might have begun with the moves `1. d4 d5 2. c4`. The Marshall Gambit involves an honest-to-goodness pawn offer, and so fits into the category of opening lines that the OP wanted stats about. – ETD Jun 22 '12 at 19:08
  • @EdDean - Fair enough. – Travis J Jun 22 '12 at 22:08
  • @ajax333221 I was hoping the last paragraph was clear about that; if you think I should add an even larger disclaimer maybe you can help me phrase it properly? – sam hocevar Aug 08 '12 at 20:41
  • 1
    @SamHocevar you are right, sorry I was too tired and missed that. But I +1'ed if that makes you feel better – ajax333221 Aug 08 '12 at 21:09
  • Interestingly I think the most successful gambit for Black has the same name -- the Marshall Gambit _in the Ruy Lopez_ seems to lead to a forced draw, it's basically put the main line of the Ruy out of business at top level. – RemcoGerlich Oct 20 '14 at 07:51
-1

According to Stockfish, the best is the Fried Liver Attack (+1.57), and the worst is the Latvian gambit (+1.61) for black, the worst is not the Halloween Gambit, although it's almost there.

Glorfindel
  • 24,745
  • 6
  • 67
  • 114
Rookie
  • 43
  • 2
  • 1
    The questions used _winning percentage and usage at high levels of play_ as criteria. While computer evaluation (with unknown settings) gives some picture, it does not necessarily reveal the usefulness of the opening to a human player and for practical play. – Anton Menshov May 31 '20 at 01:11