Chessy
The Four Knights Game: a3 — Black’s Practical Equaliser
After 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 White plays the unassuming 4.a3, a quiet move that prepares Bb5 without allowing the pin-breaking …Bb4. On the surface nothing much has happened — yet the statistics are striking. Across nearly 238,000 games Black scores 50.4%, while White wins only 45.2%. Stockfish calls the resulting position dead level at -0.14, a tiny edge for Black. That means you are already fractionally better in a position the engine considers equal. The key is to meet 4.a3 with the immediate 4…d5, striking in the centre before White can consolidate. The interactive drill below will sharpen your feel for this line.
Play the Four Knights Game: a3 against the engine
Free, no signup — you play black, the engine adapts to your level.
Try the interactive drill below — practise the 4…d5 reply against the Four Knights Game: a3 and learn to handle White’s best responses with confidence. Create a
Create a free account →The Big Idea: Seize the Centre Before White Does
White’s 4.a3 is a waiting move that does nothing for development or central control. Your reply 4…d5 exploits that instantly. You challenge the e4-pawn and open lines for your pieces, especially the c8-bishop and the queen. Black is the one playing for a lead in development here — White’s extra a3 move has already started to look like a tempo loss. The engine’s best continuation (after the most popular reply 5.exd5 Nxd5) is 6.Bb5, when …Nxc3 keeps the pressure on. The resulting structure gives Black easy equality and often a small initiative because White’s pieces are still on their starting squares while yours are active.
The Critical Moment: After 5.exd5
The most common continuation by far is 5.exd5 (played in 158,775 games), and it’s also the engine’s top choice. You reply 5…Nxd5, attacking the knight on c3. Now White usually plays 6.Bb5 — pinning your knight and threatening to double your pawns. The engine suggests 6…Nxc3 immediately, and after 7.bxc3 your structure is slightly compromised but your lead in development and the bishop pair more than compensate. The key takeaway: don’t fear this exchange. White’s doubled c-pawns can become a long-term target, and your pieces, especially the light-squared bishop, gain activity. This is where the engine’s -0.14 evaluation comes from — the position is dead level, and you’re comfortably in the game.
What the Numbers Tell Us
The database numbers for Black are unusually encouraging for a quiet opening line. Black wins 50.4% of games, and White wins only 45.2% (draws 4.4%). That 5% gap in Black’s favour is significant at club level — it means White is under more pressure to find good moves than Black is. Compare that to White’s score after other moves: if White plays 5.Bb5 (28,100 games), White scores 48.8% — slightly better for White than 5.exd5. The worst White can do is 5.h3 (5,882 games, White scores just 39.4%) or 5.d4 (5,626 games, 40.9%). If White plays 5.h3, the engine says it’s a mistake that loses about 1.4 pawns — a serious blunder. 5.Bd3 is also an inaccuracy, losing about 0.7 pawns. So not only is the position good for Black, White has plenty of ways to go wrong.
Two White Mistakes You Can Punish
Watch out for these common White errors in the 4…d5 position. First, 5.Bd3 is an inaccuracy — better was Bb5. Bd3 doesn’t challenge your knight on d5 or your central pressure, and it allows you to develop quickly with …Be7 or …Bb4. The engine says this costs White about 0.7 pawns. Second, 5.h3 is even worse — a full mistake that loses about 1.4 pawns. White wastes a tempo preventing …Bg4, but your …d5 break has already given you the initiative. After 5.h3 you can continue with …dxe4 or …Bb4, keeping the pressure. The simplest approach is to play naturally: develop your pieces, maintain the tension, and let White’s extra a3 and h3 moves become permanent liabilities.
Results across 237,914 Lichess games
| Most-played continuation | Games | White wins |
|---|---|---|
| exd5 | 158,775 | 45.5% |
| Bb5 | 28,100 | 48.8% |
| d3 | 22,422 | 42.3% |
| Bd3 | 7,864 | 46.9% |
| h3 | 5,882 | 39.4% |
| d4 | 5,626 | 40.9% |
Frequently asked questions
Is the Four Knights Game: a3 a good surprise weapon for Black?
Yes, it’s surprisingly effective. Black scores 50.4% in the database, and the engine evaluates the position as dead level (-0.14). Since many White players expect the main lines like 4.Bb5 or 4.d4, the passive 4.a3 can leave White unprepared for the central strike 4…d5.
What is Black’s best reply to 4.a3?
The best and most principled reply is 4…d5, fighting for the centre immediately. It’s the only move that makes statistical sense — Black wins half the games from this position, and the engine confirms equality with a tiny edge for Black.
Should Black be afraid of the doubled pawns after 5.exd5 Nxd5 6.Bb5 Nxc3?
Not at all. After 7.bxc3 your pawns are doubled, but you gain the bishop pair and a lead in development. White’s a3 move becomes an extra weakness, and the c3-c4 push can be a long-term target. The engine evaluates the position as dead level, and the statistics show Black scores over 50%.
What should Black do if White plays 5.h3?
5.h3 is a mistake that loses about 1.4 pawns according to the engine. Black can simply take on e4 (5…dxe4), or develop with 5…Bb4, maintaining the pressure. White has wasted a tempo, and Black’s central break has given Black a clear advantage.
How many games feature the Four Knights Game: a3?
Over 237K Lichess games have reached the Four Knights Game: a3 position. White wins 45.2%, Black wins 50.4%, with 4.4% draws — based on real rated games.