Chess openings by real win rate
Ranked by the score each opening actually gets for its own side in real Lichess games at 1600-1800. Sample size is shown for every row: a rate over a small sample is not a signal. Want just two side by side? Try the Compare Openings tool →. Want the full move tree for your own rating band instead? Try the Repertoire Explorer →.
Compare all 12 openings
Ranked by score among games between similarly-rated opponents (rating gap ≤50), which removes the biggest confound in a raw comparison like this: players who choose one opening are not the same players who choose another, so a raw score difference partly reflects who tends to play each opening, not just how it performs. This ranking does not control for anything beyond that rating gap; time-control mix and how each player group prepares are still unmeasured. Rows too close to call given their own sample sizes share a rank.
Scroll to see more →
| # | Opening | ECO | First moves | Games (1600-1800) | Score for its side |
|---|---|---|---|---|---|
| 1 | Scotch Game | C44 | 1.e4 e5 2.Nf3 Nc6 3.d4 | 3,538 | 54.0%Rating-gap-controlled score: 95 percent confidence interval 52.1 to 55.9 percent, 2,649 games. (54.6% all games) |
| 1 | Queen’s Gambit | D06 | 1.d4 d5 2.c4 | 7,344 | 53.3%Rating-gap-controlled score: 95 percent confidence interval 52.0 to 54.6 percent, 5,432 games. (54.1% all games) |
| 1 | London System | D02 | 1.d4 d5 2.Bf4 | 4,003 | 53.3%Rating-gap-controlled score: 95 percent confidence interval 51.6 to 55.0 percent, 3,050 games. (52.3% all games) |
| 1 | Ruy Lopez | C60 | 1.e4 e5 2.Nf3 Nc6 3.Bb5 | 2,941 | 51.6%Rating-gap-controlled score: 95 percent confidence interval 49.5 to 53.7 percent, 2,122 games. (52.6% all games) |
| 5 | Italian Game | C50 | 1.e4 e5 2.Nf3 Nc6 3.Bc4 | 6,665 | 50.4%Rating-gap-controlled score: 95 percent confidence interval 49.0 to 51.8 percent, 4,928 games. (51.0% all games) |
| 5 | French Defense | C00 | 1.e4 e6 | 11,261 | 50.2%Rating-gap-controlled score: 95 percent confidence interval 49.2 to 51.2 percent, 8,398 games. (50.0% all games) |
| 5 | Scandinavian Defense | B01 | 1.e4 d5 | 11,721 | 49.6%Rating-gap-controlled score: 95 percent confidence interval 48.6 to 50.6 percent, 8,864 games. (49.6% all games) |
| 5 | Sicilian Defense | B20 | 1.e4 c5 | 19,794 | 49.4%Rating-gap-controlled score: 95 percent confidence interval 48.6 to 50.2 percent, 14,369 games. (50.1% all games) |
| 5 | Caro-Kann Defense | B10 | 1.e4 c6 | 9,647 | 49.4%Rating-gap-controlled score: 95 percent confidence interval 48.3 to 50.5 percent, 7,065 games. (49.8% all games) |
| 5 | Slav Defense | D10 | 1.d4 d5 2.c4 c6 | 2,020 | 48.4%Rating-gap-controlled score: 95 percent confidence interval 45.9 to 50.9 percent, 1,488 games. (48.3% all games) |
| King’s Indian Defense | E60 | 1.d4 Nf6 2.c4 g6 | 979 | n/a | |
| Nimzo-Indian Defense | E20 | 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 | 371 | n/a |
Browse by opening
The same openings, as their win / draw / loss split at 1600-1800. Open a card to see the full move-by-move breakdown.
Scotch Game
Queen’s Gambit
London System
Ruy Lopez
Italian Game
French Defense
Scandinavian Defense
Sicilian Defense
Caro-Kann Defense
Slav Defense
King’s Indian Defense
E60, playing as black
Nimzo-Indian Defense
E20, playing as black
Browse the Chess Opening Encyclopedia
Every one of the 3,810 named lines in the standard Encyclopaedia of Chess Openings classification, grouped into 149 opening families. Browse the Chess Opening Encyclopedia →