After 1. 2. 3., the Berlin is : you hit the e4 pawn, which has no defender. If your opponent castles with 4., strong players simply take it with . Across 125,720 Lichess games rated 900 to 1700 (September 2026), you score 46.4%, and every rating band lands between 46 and 47.
Lichess games from September 2026 (30,577,162 rated 900 to 1700) and 31,791 over-the-board games rated 2300+ from the last 12 months. Data updated: October 14, 2026.
24 named lines in this family; 1 with their own page.
The opening's share in each Lichess rating band.
At your level, the most common reply is 4., in 31 games out of 100: your opponent gives up the bishop for your knight and hands you the bishop pair. Next come 4. (27), 4. (20) and 4. (18). Among online players rated 2200 and up, 4. barely shows up (2 in 100). They go for 4. (59) or 4. (34.5). Against 4., players at your level prefer (30 in 100) and only 24 take on e4, while strong players play in 88 games out of 100. After 5., 99.5 in 100 strong players retreat with . At your level, 29 do.
| Move | Lichess 900–1700 | White's score | Online, 2200+ | Over the board, 2300+ |
|---|---|---|---|---|
| 31.1% | 52.8% | 1.8% | no data | |
| 26.5% | 57.6% | 59.3% | 49.4% | |
| 20.0% | 52.3% | 34.5% | 47.9% | |
| 18.3% | 53.3% | 1.5% | 1.6% | |
| 1.7% | 53.0% | 1.1% | 1.1% | |
| 1.2% | 51.4% | no data | no data | |
| no data | no data | 1.6% | no data |
| Move | Lichess 900–1700 | White's score | Online, 2200+ | Over the board, 2300+ |
|---|---|---|---|---|
| 30.2% | 57.0% | 5.9% | 2.3% | |
| 23.9% | 59.2% | 87.6% | 97.7% | |
| 16.7% | 55.3% | 3.0% | no data | |
| 13.2% | 59.0% | 1.3% | no data | |
| 4.0% | 61.3% | 0.7% | no data | |
| 3.0% | 53.9% | 1.4% | no data |
| Move | Lichess 900–1700 | White's score | Online, 2200+ | Over the board, 2300+ |
|---|---|---|---|---|
| 68.2% | 60.1% | 52.7% | 52.8% | |
| 16.1% | 59.9% | 46.5% | 47.2% | |
| 9.6% | 55.6% | 0.4% | no data | |
| 2.7% | 53.4% | no data | no data | |
| 1.3% | 61.4% | 0.4% | no data | |
| 0.8% | 51.4% | no data | no data |
| Move | Lichess 900–1700 | White's score | Online, 2200+ | Over the board, 2300+ |
|---|---|---|---|---|
| 29.0% | 48.2% | 99.5% | 100.0% | |
| 26.6% | 61.1% | 0.3% | no data | |
| 21.2% | 66.8% | 0.2% | no data | |
| 9.3% | 69.4% | no data | no data | |
| 6.6% | 64.2% | no data | no data | |
| 2.7% | 62.3% | no data | no data |
| Move | Lichess 900–1700 | White's score | Online, 2200+ | Over the board, 2300+ |
|---|---|---|---|---|
| 53.8% | 44.3% | 4.3% | no data | |
| 39.2% | 54.5% | 84.8% | 75.2% | |
| 3.0% | 43.3% | 0.7% | no data | |
| 1.6% | 35.7% | 0.2% | no data | |
| 1.3% | 69.6% | no data | no data | |
| 1.1% | 47.5% | 5.2% | 5.3% | |
| no data | no data | 4.7% | 19.5% |
The strong-player line is 4. 5. 6. 7. . Where it leads: you take the e4 pawn, your opponent wins back the one on e5, and both sides end up without their e-pawn and with an open e-file. You don't lose material, and your bishop on e7 blocks that file in front of your king.
works because, after castling, nothing your opponent has covers that pawn. 5. attacks the knight, and once it moves, your e5 pawn is hit by the rook and the f3 knight, with only your c6 knight holding it. That's why the retreat square matters. gains time, because from d6 the knight attacks the b5 bishop. From f6 it attacks no enemy piece, and at your level that retreat scores 39 in 100, against 52 for .
6. wins the pawn back. steps onto the e-file between the e1 rook and your king, and it clears f8 so you can castle short. Your d6 knight is still hitting the b5 bishop, so 7. sends it home, out of reach. Then takes the knight that landed on e5. The order makes sense: with your bishop already on e7, the rook that recaptures on e5 doesn't give check.
The other strong-player choice is 5., in 46.5 games out of 100. After 6. 7., the e5 pawn kicks your d6 knight and puts it somewhere safe. With 8. the queens come off. You lose the right to castle, but you keep two bishops against one. That's the Berlin Wall, and you'll rarely see it at your level.
At your level, results split at move 5. After 5., 27 in 100 players drop back to f6 and 21 prop up the knight with , which scores 33. The rule for this line: when the rook hits your e4 knight, bring it back to d6, where it attacks the bishop.
| Move | Strong games | Chosen by |
|---|---|---|
| 13,973 | 59.0% | |
| 8,217 | 87.9% | |
| 7,205 | 52.7% | |
| 3,797 | 99.6% | |
| 3,776 | 84.5% | |
| 3,189 | 82.3% | |
| 2,614 | 92.8% | |
| 2,448 | 93.1% |
Against 4., moves the d7 pawn off the diagonal, so the b5 bishop pins your c6 knight. After 5. 6. you spend undoing the pin, and your opponent keeps the e4 pawn. It's an inaccuracy that rarely gets punished at your level: 23 in 100 play it and score 48, against 50 with .
23 in 100 play it in this position. It gives away 0.7 pawns of advantage. It scores 48.3%; scores 50.0%.
Against 4., opens the bishop's line to your king, so your c6 knight is pinned and can't recapture on e5. 5. grabs the pawn and hits c6, which you hold with , and 6. 7. follows. It's a rare mistake: 3 in 100 play it and score 45, against 50 with .
3 in 100 play it in this position. It gives away 2.5 pawns of advantage. It scores 45.2%; scores 50.0%.
After 5., leaves your c6 knight pinned to your king. 6. hits the bishop, which steps away with , and 7. hits the knight, which can't move. After 8. , the d5 pawn is still attacking c6. It's a serious mistake: 23 in 100 play it and score 36, against 42 with .
23 in 100 play it in this position. It gives away 5.2 pawns of advantage. It scores 36.4%; scores 41.8%.
After 4., attacks the bishop, but 5. removes the defender of e5 and 6. wins the pawn. After 7. you're still a pawn down. It's a mistake, but at your level it doesn't show in the results: 13 in 100 play it and score 41, the same as with .
13 in 100 play it in this position. It gives away 1.5 pawns of advantage. It scores 41.0%; scores 40.8%.
Against 4., allows 5. , and with the c6 knight gone nothing guards e5, so 6. takes it. Since d3 protects e4, you can't win the pawn back there. After 7. you're still a pawn down. It's a mistake: 18 in 100 play it and score 46, against 50 with .
18 in 100 play it in this position. It gives away 1.4 pawns of advantage. It scores 45.6%; scores 50.2%.
After 5., props up the e4 knight, but 6. adds a second attacker. goes after the rook, and after 7. 8. your b4 bishop and e4 knight are both under attack. It's a serious mistake: 21 in 100 play it and score 33, against 52 with .
21 in 100 play it in this position. It gives away 4.5 pawns of advantage. It scores 33.2%; scores 51.8%.
| White | Black | Games |
|---|---|---|
| 75,031 | ||
| Not castled | 18,476 | |
| 11,302 | ||
| Not castled | 7,110 | |
| Not castled | Not castled | 5,618 |
| 3,320 |
| White | Black | Games |
|---|---|---|
| 306 | ||
| Not castled | 50 | |
| 17 | ||
| 6 | ||
| Not castled | 4 | |
| Not castled | Not castled | 1 |
1. 2. 3. 4. · 27,294 amateur games · 153 over the board
4. shores up e4 with a pawn and opens the c1 bishop's diagonal: you score 48 in 100.
1. 2. 3. 4. · 14,738 amateur games · 3 over the board
After 4., aims at f2 instead of taking on e4: you score 46 in 100.
1. 2. 3. 4. · 13,837 amateur games · 1 over the board
After 4., holds e5 but lets the bishop pin your c6 knight: you score 46 in 100.
1. 2. 3. 4. · 8,783 amateur games · 113 over the board
, the strong-player choice, takes the e4 pawn: at your level you score 41 in 100.
1. 2. 3. 4. 5. 6. · C66
750 amateur games · White's score: 54.7%
1. 2. 3. 4. · C65
406 amateur games · White's score: 44.0%
1. 2. 3. 4. 5. 6. · C66
299 amateur games · White's score: 51.7%
1. 2. 3. 4. 5. · C67
267 amateur games · White's score: 50.0%
1. 2. 3. 4. 5. · C65
242 amateur games · White's score: 49.8%
1. 2. 3. 4. 5. 6. 7. · C66
104 amateur games · White's score: 63.0%
1. 2. 3. 4. 5. 6. 7. · C66
98 amateur games · White's score: 58.7%
1. 2. 3. 4. 5. 6. 7. 8. · C67
79 amateur games · White's score: 43.0%
1. 2. 3. 4. 5. · C67
68 amateur games · White's score: 60.3%
1. 2. 3. 4. 5. 6. 7. · C66
49 amateur games · White's score: 49.0%
1. 2. 3. 4. 5. · C67
36 amateur games · White's score: 51.4%
1. 2. 3. 4. 5. 6. 7. 8. 9. · C67
22 amateur games · White's score: 27.3%
1. 2. 3. 4. 5. 6. · C67
19 amateur games · White's score: 52.6%
1. 2. 3. 4. 5. 6. 7. · C67
2 amateur games · White's score: 50.0%
1. 2. 3. 4. 5. 6. · C67
1 amateur game · White's score: 100.0%
1. 2. 3. 4. 5. 6. 7. 8. · C67
1. 2. 3. 4. 5. 6. 7. 8. 9. · C67
1. 2. 3. 4. 5. 6. 7. 8. 9. · C67
1. 2. 3. 4. 5. 6. 7. 8. 9. · C67
It's answering the Ruy Lopez with , after 1. 2. 3.. You develop the knight and hit the e4 pawn. Between 900 and 1700 it comes up in fewer than 1 game in 100, a bit more often at 900 to 1300 (5 in 1000) than at 1500 to 1700 (3 in 1000).
It's a sound choice. Online players rated 2200 and up score about 50 in 100 with it, and you score between 46 and 47 in every band from 900 to 1700. The hard part at your level is move 5: after 4. 5., the retreat scores 52 in 100, while scores 33.
If your opponent castles, take on e4 with , the move 88 in 100 strong players choose. Against 5., bring the knight back to d6, where it attacks the b5 bishop. Against 4., which guards e4, avoid : after 5. and your recapture, your e5 pawn has no defender. scores 50 in 100.