King's Gambit Accepted: Fischer Defense — Bc4

The Fischer Defense (4...h6) is a waiting-and-probing move. You are not trying to mate White on the kingside — you are asking White to commit. The pawn on f4 is already captured, and your h6 pawn prevents White from playing Ng5 followed by Bxf7+, a classic sacrificial threat. Your long-term plan is to build a pawn wall with ...g5 and ...Bg7, securing your extra pawn while White struggles to generate real attacking chances. This is a positional approach to the King's Gambit: you take the material, then dare White to prove they have compensation. The engine's -0.68 evaluation confirms that you are slightly better if you follow the correct defensive plan.

Stockfish recommends d4 as White's top move, continuing with the plan d4 g5 h4 Bg7. After 5.d4, you should respond with 5...g5, reinforcing the f4 pawn and preventing White from easily opening lines. Then 6.h4 is White's best try to break down your kingside — the idea is to force ...g4, after which White's knight can retreat to g1 or d4. You answer with 6...Bg7, developing the bishop to its ideal diagonal and completing the core of the Fischer setup. Your king will usually castle queenside or stay in the centre for a while, depending on how White continues. This is the line you want to aim for: you keep the extra pawn and leave White searching for an attack that may not exist.

Across over 627,000 games at this exact position, the results are tighter than you might expect for a line that favours Black: White wins 52.1%, draws 3.2%, and Black wins 44.7%. That 52.1% White win rate is a danger signal — it tells you that many Black players mishandle the position despite the engine saying Black is better. The most popular move, O-O (castling kingside), is played in 283,341 games but only scores 51.7% for White — not bad, but not the engine's first choice either. The engine's recommended d4 (225,768 games) actually scores slightly better for White at 53.3%, which suggests that many Black players don't know the correct response. The moral: the Fischer Defense works if you know the follow-up. If you just play random moves, the statistics will punish you.

The FACTS list two inaccuracies that Black should be ready to punish. First, d3. If White plays 5.d3, the engine says it loses about 0.7 pawns — a clear mistake. In that case, you should simply develop with ...Bg7 and keep your extra pawn; White has weakened the d3 square and given up on central pressure. Second, c3. This move also loses about 0.7 pawns (the engine says d4 was better than both). After 5.c3, White plans d4 in two moves, but you have time to play ...g5, ...Bg7, and possibly ...Nc6 or ...Be6. The key takeaway: if White plays anything other than d4 or O-O, you are already doing even better than the base evaluation. Stay alert, keep your pawn chain intact, and don't rush to return the material.