10 October 2026

Do famous chess traps actually work? Eleven of them, checked by Stockfish

Every trap is a bet. We measured eleven famous ones with Stockfish: what each costs you if your opponent knows it, and what it wins if they fall in.

Every trap is a bet

A trap is a move that hopes. It is usually not the best move on the board, and it is played because the opponent might answer it badly. Trap videos show the second half of that sentence, the opponent falling in, and almost never the first: what it costs when they do not.

So we asked Stockfish three questions about each trap, always from the side of the player who sets it. How good was the position if you had simply played the engine's move instead? How good is it after the trap, against someone who answers correctly? And how good is it if they fall in? The answers are given as a chance of winning, the same scale the game review on this site uses, where 50% is level.

All eleven, from free to expensive

Eleven opening traps: what each costs to set, and what it wins
Trap	Costs you	Your chance if they know it	…if they fall in
Légal's Mate	nothing	66%	mate
Lasker Trap	nothing	53%	87%
Noah's Ark Trap	nothing	50%	72%
Fried Liver Attack	nothing	52%	59%
Elephant Trap	nothing	47%	79%
Siberian Trap	nothing	48%	75%
Scholar's Mate	4 points	48%	mate
Blackburne Shilling Gambit	7 points	41%	54%
Stafford Gambit	8 points	38%	79%
Fishing Pole Trap	10 points	39%	73%
Englund Gambit trap	10 points	37%	82%

"Costs you" is how much winning chance the trap gives up against a prepared opponent, compared with the engine's own move. Anything under two points is inside the engine's own margin, and is shown as nothing.

The free ones: set them every time

The best traps are not tricks at all. They are good moves that happen to leave the opponent one natural-looking way to go wrong. If they do not take it, you have lost nothing.

Elephant Trap: White to move, and 6.Nxd5 falls into it — Elephant Trap1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.Bg5 Nbd7 5.cxd5 exd5 6.Nxd5?

The Elephant Trap is the purest example. 4…Nbd7 is an ordinary developing move in the Queen's Gambit Declined. It just happens to make the pawn on d5 look loose, and a player who takes it with the knight loses a piece to a check on b4.

Noah's Ark Trap: White to move, and 8.Qxd4 falls into it — Noah's Ark Trap1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 d6 5.d4 b5 6.Bb3 Nxd4 7.Nxd4 exd4 8.Qxd4?

Siberian Trap: White to move, and 9.h3 falls into it — Siberian Trap1.e4 c5 2.d4 cxd4 3.c3 dxc3 4.Nxc3 Nc6 5.Nf3 e6 6.Bc4 Qc7 7.O-O Nf6 8.Qe2 Ng4 9.h3?

The Siberian Trap catches the gambiteer rather than the defender. White's 9.h3 is the natural answer to the mate threat on h2, and it is the losing one. 9.Rd1, giving the king a square on f1, keeps the game level.

Lasker Trap: White to move, and 6.Bxb4 falls into it — Lasker Trap1.d4 d5 2.c4 e5 3.dxe5 d4 4.e3 Bb4+ 5.Bd2 dxe3 6.Bxb4?

Légal's Mate: Black to move, and 5…Bxd1 falls into it — Légal's Mate1.e4 e5 2.Nf3 d6 3.Bc4 Bg4 4.Nc3 g6 5.Nxe5 5…Bxd1?

Légal's Mate goes one better: by the time Black has played the careless 4…g6, taking on e5 is simply the strongest move, trap or no trap. Taking the queen with 5…Bxd1 walks into mate in two.

The famous ones that pay less than you think

Fried Liver Attack: Black to move, and 5…Nxd5 falls into it — Fried Liver Attack1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 5.exd5 5…Nxd5?

The Fried Liver is the trap everybody learns first, and the engine is unimpressed by it. 4.Ng5 costs nothing, but when Black does recapture on d5 with the knight and White sacrifices on f7, White is about a pawn better: an attack worth playing, not a won game. The defence, 5…Na5, keeps the game level.

Blackburne Shilling Gambit: White to move, and 4.Nxe5 falls into it — Blackburne Shilling Gambit1.e4 e5 2.Nf3 Nc6 3.Bc4 Nd4 4.Nxe5?

The Blackburne Shilling Gambit is famous for a mate on move seven. But taking the pawn on e5 is not what loses: after 4…Qg5, White castles and the position is close to level. The mate only comes if White also grabs on f7. Meanwhile 3…Nd4 itself costs Black real winning chance against anyone who simply takes the knight on d4.

The gambles

These traps work, spectacularly, when they work. Each of them also hands a prepared opponent a clearly better game, so they are a bet on who is sitting across the board.

Englund Gambit trap: White to move, and 6.Bc3 falls into it — Englund Gambit trap1.d4 e5 2.dxe5 Nc6 3.Nf3 Qe7 4.Bf4 Qb4+ 5.Bd2 Qxb2 6.Bc3?

Stafford Gambit: White to move, and 6.Bg5 falls into it — Stafford Gambit1.e4 e5 2.Nf3 Nf6 3.Nxe5 Nc6 4.Nxc6 dxc6 5.d3 Bc5 6.Bg5?

Fishing Pole Trap: White to move, and 6.hxg4 falls into it — Fishing Pole Trap1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.O-O Ng4 5.h3 h5 6.hxg4?

Scholar's Mate: Black to move, and 3…Nf6 falls into it — Scholar's Mate1.e4 e5 2.Bc4 Nc6 3.Qh5 3…Nf6?

Scholar's Mate is the smallest gamble of the four. Bringing the queen out on move three gives away a few points of winning chance against a player who blocks with …g6, and against a beginner who plays 3…Nf6 it ends the game on the spot.

What to do with this

Play the free traps whenever the position offers them: you lose nothing and sometimes win a piece or the game. Treat the gambles as what they are. Against players who have not seen them, they win a lot of games quickly; against players who have, you start from a worse position than you needed to. If you are learning, the free ones are also the ones that teach good chess, because the trap move is a good move.

Each line was measured by Stockfish 18 at depth 22. The numbers are the engine's estimate at that depth, not a proof, except where it found a mate; the page on how we check explains what that means.

Sources

Facts about the news and the games are as reported by the sources above. The engine numbers are this site's own: how we check them.