French Defense: Marshall Gambit – Bb5+ for Black

The Marshall Gambit (4.Bb5+) is White's attempt to disrupt your standard French setup before you've even castled. By checking on b5, White invites complications around the d4 and c5 pawns. You've already answered with 4...Nc6, blocking the check and developing a piece. The central tension is huge: White can capture on d5, push d5, or take on c5 — and each choice leads to a very different kind of middlegame. You're slightly cramped and lagging in development, but the position is rich with counterplay if White doesn't know the precise follow-up. Your main job is to survive the opening phase, complete your development, and target White's centre once it becomes overextended.

Stockfish's top choice at depth 16 is 5.exd5. This is the move that gives White their evaluating edge of +1.23. After 5.exd5, the engine's suggested continuation is 5...exd5 6.dxc5 a6. This sequence wins a pawn for White (the c5-pawn) while Black gets the bishop pair and some play. Notice the key detail: Black recaptures on d5 with the e-pawn (5...exd5), not the queen. This keeps the centre solid and avoids exposing your queen to attack. You'll follow up with ...a6, chasing the bishop, and then recapture on c5 with the bishop — you'll have active pieces and a lead in development as compensation for the pawn.

The statistics from 61,408 games show that White has several common alternatives to 5.exd5, and nearly all of them are mistakes that improve your chances. Here are the three most frequent errors you should be ready to exploit: Nf3 (14,695 games, loses ~1.1 pawns), Bxc6+ (7,212 games, loses ~1.8 pawns), and dxc5 (4,137 games, loses ~1.6 pawns). After 5.Nf3, you can play ...cxd4, striking back in the centre and leaving White's bishop on b5 awkwardly placed. After 5.Bxc6+ bxc6, you get the bishop pair and a semi-open b-file for your rook. After 5.dxc5, simply recapture with 5...Nf6, developing with tempo against the e4-pawn. In all these lines, you are the one fighting for the initiative.

The game results from this exact position tell an encouraging story. Even though Stockfish says White is clearly better (+1.23), the real-world results from 61,408 games are: White wins 50.2%, draws 3.6%, Black wins 46.2%. That means you will win almost as often as your opponent if you play well — surprising for a position the engine rates as +1.23. The key takeaway: this is a practical opening where human players make mistakes, and Black's counterplay is easier to handle than the engine evaluation suggests. White's best move (5.exd5) is played in only 26,151 of those games (about 43%). In the other 57% of games, White plays something suboptimal — and you are well-prepared to pounce.