Chessy
Two Knights Defense: d3 – Playing Black with 4...Bc5
After 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.d3, most club players instinctively keep developing with 4...Bc5 — and that's exactly right. You mirror White's bishop, eye the f2 pawn, and keep the tension in the centre. The engine gives +0.34, a small edge for White, so you're slightly worse here, but the statistics tell a different story: across nearly 13 million games, Black scores 46.2% wins against 49.9% for White. That's a razor-thin gap for a line that's supposed to be 'worse.' The drill below puts you in Black's shoes against White's best responses.
Play the Two Knights Defense: d3 against the engine
Free, no signup — you play black, the engine adapts to your level.
Jump into the position and test yourself against the engine. See if you can navigate White's best tries and turn that slight +0.34 edge into a full point.
Create a free account →The Main Idea: Develop and Challenge the Centre
The Two Knights Defense: d3 is a patient, classical set-up. Your 4...Bc5 completes a natural development pattern: knights out, bishops aimed at the kingside, and the e5 pawn solidly defended. Unlike sharper lines like the Fried Liver or the Two Knights with 4.Ng5, White's d3 avoids immediate tactics. Black's job is to finish development with d6, O-O, and later consider breaks like ...d5 or ...Re8 to pressure the centre. The position remains balanced, and small inaccuracies from White can quickly turn the tables in your favour. This is a great opening if you want clear, principled chess without memorising long forcing lines.
White's Best Move: 5.O-O
Stockfish's top recommendation is 5.O-O, and with good reason — it's also the most popular choice, appearing in over 2.6 million games. White castles, prioritising king safety, and scores a modest 49.7% from there. The engine anticipates a continuation like 5...d6 6.c3 a5, where Black gains queenside space and prepares ...Ba6 to trade off the light-squared bishop. Your plan is straightforward: complete development, keep the centre flexible, and look for chances to strike in the middlegame. Since Black scores almost as well as White from this position, 5.O-O is nothing to fear.
The Critical Mistake: 5.Ng5
Here's where you can punish White. Playing 5.Ng5 is a known inaccuracy that loses about 0.7 pawns compared to castling. In over 2.3 million games, White actually scores worse with Ng5 (45.7%) than with any other major move — including the passive-looking h3 (50.8%). Why? Because after 5...d5 (the classic counter), Black gains space, opens lines for the bishops, and White's knight on g5 can become misplaced. If your opponent plays 5.Ng5, you're already in a favourable spot. Recognise this moment and strike in the centre with ...d5.
What the Statistics Tell You
The numbers are surprisingly encouraging for Black. From this position, White wins 49.9%, draws 3.9%, and Black wins 46.2%. That's a 3.7% gap — nearly even in practice, especially at club level. White's other popular moves (5.Bg5, 5.c3, 5.Nc3, 5.h3) all score 50.8%–52.4%, which means White's advantage is never decisive. The +0.34 evaluation is real but slight, and it evaporates quickly if White doesn't play accurately. You can play this line confidently: the positions are sound, the plans are clear, and your practical winning chances are excellent.
Results across 12,994,475 Lichess games
| Most-played continuation | Games | White wins |
|---|---|---|
| O-O | 2,654,217 | 49.7% |
| Ng5 | 2,336,699 | 45.7% |
| Bg5 | 2,078,412 | 51.3% |
| c3 | 1,814,792 | 52.4% |
| Nc3 | 1,439,048 | 51.7% |
| h3 | 1,212,696 | 50.8% |
Frequently asked questions
Is the Two Knights Defense: d3 good for Black?
Yes, it's a solid, principled choice. The engine gives +0.34 (a small edge for White), but Black wins 46.2% of games versus White's 49.9% across nearly 13 million games. At club level, those numbers are essentially even, and many White players know less theory here than in sharper lines.
What is White's best move in the Two Knights Defense: d3?
Stockfish recommends 5.O-O, and it's also the most common move (2.6 million games). Other popular options include 5.Ng5, 5.Bg5, 5.c3, 5.Nc3, and 5.h3. Crucially, 5.Ng5 is considered a mistake that gives Black the edge.
Why is 5.Ng5 bad for White in this line?
The engine says 5.Ng5 loses about 0.7 pawns compared to 5.O-O. In practice, White scores only 45.7% after Ng5 — worse than any other major move. Black can reply with 5...d5, gaining centre space and opening lines for the bishops, leaving White's knight awkwardly placed.
How should Black respond to White playing 5.c3 or 5.Nc3?
Both are solid waiting moves. Against 5.c3 (White scores 52.4%), Black should continue developing with ...d6 and ...O-O, and may later play ...a5 to challenge White's queenside. Against 5.Nc3 (White scores 51.7%), the same classical development works: ...d6, ...O-O, and maintain the central tension.