Chessy
The Bishop's Opening: Horwitz Gambit – A Dangerous Surprise for Black
You're playing Black, and after 1.e4 e5 2.Bc4 Nf6 White tries 3.Nc3, developing naturally. Now it's your turn — and you have a sharp idea: 3...b5, the Horwitz Gambit. You strike at the light-squared bishop immediately, offering a pawn to disrupt White's development and snatch the initiative. It's a bold choice, and the statistics across 7,356 games show it's no joke: Black wins 41.2% of the time here. But there's a catch — the engine has a clear verdict on how White should respond. Let's see what you're up against.
Play the Bishop's Opening: Horwitz Gambit against the engine
Free, no signup — you play black, the engine adapts to your level.
Test your Horwitz Gambit skills now — play the interactive drill and see if you can punish White's mistakes. A free account saves your progress and tracks your
Create a free account →The Engine's Verdict: Honest but Not Fatal
Let's get the bad news out of the way first, because it matters. Stockfish evaluates this position at +1.32, a clear edge for White. That means you are worse from the start — the computer believes White's advantage is real and lasting, not just a tactical flash. The engine's recommended reply is 4.Bxb5, capturing your gambit pawn, and it plans to follow up with 4...c6 5.Ba4 Qa5, when Black has some compensation but the engine still prefers White. So why play this gambit? Because 41.2% of Black wins is a respectable number, and club-level opponents often struggle to find the best continuations. The Horwitz Gambit is a practical weapon, not a theoretical equaliser.
White's Most Common Moves – and Which Ones to Hope For
You have a clear preference about what White does next. The most-played move is 4.Bxb5 (4,011 games), where White scores 55.7% — solid but not crushing. That's the critical test of your gambit. But look at what else appears in the database: 4.Bb3 (1,976 games, White scores 60.0%) is actually a mistake that loses about 1.1 pawns of advantage. 4.Nxb5 (716 games, White scores 56.7%) is an inaccuracy, losing roughly 0.8 pawns. Even more tempting: 4.Bd5 (198 games, White scores just 48.5%) is a clear mistake, losing about 1.8 pawns. Your best-case scenario is facing 4.Bb3, 4.Nxb5, or 4.Bd5 — all suboptimal moves where the statistics or engine analysis favour you. If your opponent knows to take with the bishop, you're in for a tougher fight.
The Critical Reply: 4.Bxb5 and the Most-Played Line
When White finds the best move, 4.Bxb5, you need a plan. The engine's suggested continuation is 4...c6 5.Ba4 Qa5 — you chase the bishop, gain time, and start generating threats against the undefended knight on c3 and the bishop on a4. This is the main line you should study in the interactive drill. The idea is straightforward: you're sacrificing a pawn for rapid development and active piece play. Your queen comes out with tempo, you'll follow with ...d5 or ...Bc5, and White's extra pawn isn't going to matter if you can castle quickly and open the centre. Know this continuation cold — if you don't, your gambit can backfire when White plays accurately.
Three White Mistakes You Can Punish
Your opponents will often make things easier for you. The engine flags four suboptimal moves, and three are worth memorising as opportunities: 4.Bb3 is a mistake — the bishop retreats to a passive square and you've already gained time with your b-pawn push. Your follow-up ideas include ...d5 or ...Bb7, seizing the centre. 4.Nxb5 is an inaccuracy — White grabs a pawn with the knight, but after 4...c6 the knight has to retreat, and you've gained tempi while keeping your bishop pair intact. 4.Bd5 is an even bigger mistake — the bishop looks active on d5 but it's exposed to ...c6 and ...d5 pushes that gain time and space. Each of these moves gives you concrete chances to turn the evaluation around. When you see any of them in the drill, take a moment to find the most active response.
Results across 7,356 Lichess games
| Most-played continuation | Games | White wins |
|---|---|---|
| Bxb5 | 4,011 | 55.7% |
| Bb3 | 1,976 | 60.0% |
| Nxb5 | 716 | 56.7% |
| Bd5 | 198 | 48.5% |
| Nf3 | 117 | 41.9% |
| d3 | 101 | 36.6% |
Frequently asked questions
Is the Horwitz Gambit sound for Black?
Objectively, Stockfish evaluates the position at +1.32 in White's favour after 3...b5, so the gambit is not theoretically sound. White's best reply, 4.Bxb5, maintains a clear advantage. However, in practical play across 7,356 games, Black still wins 41.2% of the time — a respectable figure that makes it a useful surprise weapon, especially against opponents who don't know the best response.
What is White's best move against the Horwitz Gambit?
The engine's top choice is 4.Bxb5, capturing the gambit pawn. The suggested follow-up is 4...c6 5.Ba4 Qa5, when Black has some activity but White keeps the advantage. This is the line you should be most prepared to face as Black.
Which White moves are bad for White in this position?
Multiple moves give you better chances. 4.Bb3 is a mistake (loses ~1.1 pawns), 4.Nxb5 is an inaccuracy (~0.8 pawns), and 4.Bd5 is a clear mistake (~1.8 pawns). If your opponent plays any of these, you have good chances to seize the initiative.
How often does the Horwitz Gambit actually win for Black?
In the Lichess database of 7,356 games at this exact position, Black wins 41.2% of the time, White wins 56.0%, and draws occur 2.8% of the time. Those are surprisingly good odds for a gambit that is objectively not quite sound.