Chess.com says the gambit has surged among club players, in part because of IM Levy Rozman's video, and the pitch is simple: give a pawn, set a queen trap, and win fast. Chess.com's opening page confirms the surge as of October 2026 and adds that stronger players mostly avoid it. What the pitch leaves out is the part before the trap. If White declines to cooperate on move two, there is no trap at all. We ran the maia3-5m checkpoint on the position after 1.d4 e5 and again after 1.d4 e5 2.dxe5 Nc6, at seven rating bands from 1000 to 2200, and graded every White reply with Stockfish 18 at depth 20. Maia3 is a human-move prediction model, so its percentages are the probability that a player of that rating picks a move, not a count of real games. It is the same method behind our chess openings by rating dataset; for the background, see how Maia chess actually works.
What the Englund Gambit is
The Englund Gambit starts 1.d4 e5 and is filed under ECO A40. Wikipedia says it is named for the Swedish player Fritz Englund, who sponsored a thematic tournament in late 1932 and early 1933, and that it is also called the Charlick Gambit after the Australian champion Henry Charlick. Boris Avrukh is quoted on Wikipedia saying 1...e5 seems to him the worst possible reply to White's first move, and the opening is rare at top level. It lives in club chess, where it has a reputation as a trap repertoire for fast time controls and a riskier choice in slow games, because White has time to solve the early problems. For a 1.d4 gambit the engine treats far more kindly, see the Benko Gambit by rating, where White keeps under a pawn in the main line. On the 1.e4 side, the Danish Gambit by rating shows a gambit where Black stands better after taking both pawns, yet takes them less often as rating rises.
The idea is a pawn sacrifice for quick development. After 2.dxe5 Black plays 2...Nc6, and the knight on c6 plus the open lines toward White's queenside give Black the targets.
Position A: after 1.d4 e5, what does White play?
This is the only decision that decides whether the gambit gets played. Stockfish ranks 2.dxe5 first at +1.38, then 2.Nc3 at +0.39, 2.e3 at +0.27, 2.c3 at +0.08, 2.c4 at +0.02, 2.e4 at -0.14, 2.g3 at -0.26 and 2.d5 at -0.28. Taking the pawn is not close.
| Rating | 1000 | 1200 | 1400 | 1600 | 1800 | 2000 | 2200 |
|---|---|---|---|---|---|---|---|
| Probability of 2.dxe5 (best, +1.38) | 52.9% | 58.9% | 61.8% | 63.1% | 64.5% | 67.3% | 72.3% |
| Probability of 2.d5 (-0.28) | 7.0% | 8.8% | 10.1% | 10.7% | 10.2% | 7.7% | 4.5% |
| Probability of 2.c4 (+0.02) | 5.8% | 6.9% | 7.6% | 8.1% | 8.5% | 8.0% | 6.5% |
| Probability of 2.e3 (+0.27) | 9.1% | 7.0% | 5.8% | 5.3% | 4.9% | 4.5% | 3.5% |
The engine's best move is also the likeliest human move at every band, so the gap between the crowd and the engine is zero throughout. It climbs steadily, from 52.9 percent at 1000 to 72.3 at 2200. Even at the lowest band, more than half of White players capture. At 1000, the other likeliest moves are 2.Nf3 at 9.9 percent, 2.e3 at 9.1, 2.d5 at 7.0 and 2.c4 at 5.8. At 2200, 2.Nf3 is 7.0 percent, 2.c4 is 6.5, 2.d5 is 4.5 and 2.e3 is 3.5.
For a second view, 365chess's database lists 1,288 games of the Englund, of which 2.dxe5 is played in 1,001, about 78 percent. That is a mixed database of over the board and online games, not a club sample, and it points the same way as ours: most Whites accept. As of October 2026 the same page shows White winning 49.1 percent of the 1,288, drawing 17.9 and losing 33.1, which tells you the gambit is no walkover for either side in practice.
Position B: after 2.dxe5 Nc6, which third move?
After 2...Nc6 White must decide how to hold or return the pawn. Stockfish puts 3.Nf3 first at +1.33, then 3.Bf4 at +0.89, 3.Nc3 at +0.85, 3.e4 at +0.77 and 3.f4 at +0.43.
| Rating | 1000 | 1200 | 1400 | 1600 | 1800 | 2000 | 2200 |
|---|---|---|---|---|---|---|---|
| Probability of 3.Nf3 (best, +1.33) | 63.1% | 68.6% | 70.9% | 72.3% | 74.3% | 77.9% | 82.1% |
| Probability of 3.Bf4 (+0.89) | 13.7% | 12.3% | 10.4% | 8.3% | 6.6% | 5.5% | 4.6% |
| Probability of 3.f4 (+0.43) | 14.6% | 11.2% | 9.6% | 9.0% | 8.5% | 7.0% | 4.3% |
Again the likeliest move is the engine's move at every band, 63.1 percent at 1000 rising to 82.1 at 2200. The pattern in the minor moves is a rating one. Players near 1000 are likelier to push 3.f4 at 14.6 percent, a move that tries to defend the pawn and scores +0.43 against +1.33 for 3.Nf3. By 2200 3.f4 has fallen to 4.3 percent and 3.e4, which scores +0.77, is up to 5.2.
The point for Black is that once White takes and Black plays 2...Nc6, 3.Nf3 is the likeliest reply at every band from 1000 to 2200. That is the position the trap is built for, so the setup is not the hard part. The hard part is move four.
The main line: 3...Qe7
The standard continuation is 3...Qe7, attacking the pawn on e5 and eyeing b4. Stockfish 18 at depth 22 says White's best reply is 4.Bg5 at +1.47, which develops with tempo by attacking the queen. Sources do not agree on one answer. Wikipedia and Chess Doctrine favour the queen move 4.Qd5, which keeps the pawn by centralizing the queen. Ian's Chess Gambits calls 4.Nc3 sensible and not a refutation, and says it gives White a slightly larger than normal advantage. Our dataset stops at White's third move, so we have no frequencies for any of these.
The move people actually search for is 4.Bf4, because it is natural, it develops, it defends e5, and it walks into the trap. Our engine did not rate it above 4.Bg5 and we will not quote a number for it, but it is not a blunder by itself. The trouble starts a few moves later.
The trap and the refutation
The famous line runs 1.d4 e5 2.dxe5 Nc6 3.Nf3 Qe7 4.Bf4 Qb4+ 5.Bd2 Qxb2. Black has won a pawn back and the queen sits on b2, hitting the rook on a1 and the knight on b1. Now comes the test.
6.Bc3?? is the move the trap wants. It attacks the queen and seems to cover the rook. Black plays 6...Bb4!, and the bishop on c3 is pinned to the king on e1, so it cannot take the queen on b2 and ...Qxa1 is coming. White has two ways to go wrong. After 7.Bxb4 Nxb4, our Stockfish 18 check at depth 22 gives -4.57 for White. After 7.Qd2?? Bxc3 8.Qxc3, Black plays 8...Qc1#, and we confirmed the mate with the engine: the c2 pawn blocks White's queen from taking on c1, nothing can interpose on d1, and the king has no escape. The least bad try is 7.Bd2, a retreat that still scores -4.25.
The refutation is 6.Nc3, which protects the rook on a1 and keeps White at +1.74 by the same check. The alternatives are poor: 6.e6 scores -1.87 for White, 6.e4 -2.32 and 6.e3 -3.06. Earlier on, the engine also puts 5.Bd2 at +1.52 after 4...Qb4+, while 5.Nc3 drops to +0.08 and 5.Nbd2 is at -1.54 or worse. So the safe route is 5.Bd2 and 6.Nc3, after which the engine rates White at +1.74 with the black queen offside on b2.
After 6.Nc3 Bb4 Ian's Chess Gambits says the old main line 7.Rb1 Qa3 8.Rb3 is playable for Black, but that 8.Nd5 is close to a theoretical refutation, since it attacks c7 and forces the king to d8. Wikipedia gives the same trap avoidance sequence. One caution about frequency: our data covers only White's second and third moves, so we cannot say how often White plays 4.Bf4 or 6.Bc3 at any rating. Web pages that claim to give that percentage give no denominator, and we do not repeat them.
Other Black ideas
Wikipedia names several variations that branch earlier than 3...Qe7. Each needs a different answer from White.
- Soller Gambit, 3...f6. After 2.dxe5 Nc6 3.Nf3 f6 4.exf6 Nxf6 5.Bg5, Wikipedia says Black can reach partial compensation with ...h6.
- Hartlaub-Charlick Gambit, 2...d6. Also called the Blackburne-Hartlaub Gambit, with 3.exd6 Bxd6 as the usual continuation. Wikipedia says Black aims for compensation for a pawn.
- Felbecker Gambit, 3...Bc5. After 2.dxe5 Nc6 3.Nf3 Bc5, Wikipedia says Black usually follows with ...f6.
- The 4.Qd5 line. After 3...Qe7 4.Qd5, the usual Black reply is 4...f6 and, after 5.exf6 Nxf6 6.Qb3 d5, Chess Doctrine says Black has some but not full compensation. 4...h6 is met by 5.Nc3 g5 6.Qe4.
We did not source refutations of these lines in detail, so treat the one line summaries above as pointers, not verdicts. If you are going to play one of them, check a current database for your own rating before you invest hours.
How White should play by rating
The data gives one clear instruction and a few judgment calls. The clear one: take the pawn and play 3.Nf3. That is the engine's top move at +1.38 and +1.33, it is the likeliest human move at every band, and it needs no preparation beyond move four.
From 1000 to 1400, White's temptation is to protect the extra pawn at once with 3.f4, which Maia3 gives 14.6 percent at 1000 and 9.6 at 1400. It scores +0.43 against +1.33 for the developing move, so it throws away most of the gain for a pawn that White could keep with simple development. Do not do it. 3.Bf4 at +0.89 is better than 3.f4 but still lets Black choose the line, since it invites the Qe7 sequence on Black's terms.
From 1600 up, you will meet a Black player who knows the book. Learn three answers: 4.Bg5 at +1.47 as the engine's top choice, 4.Qd5 if you prefer to keep the pawn with the queen, and 6.Nc3 whenever the queen reaches b2. Do not memorize more than that. If you want a structure that does not depend on theory, forum advice for London system players is to take the pawn, develop with e3, Nbd2 and Be2, or else play 3.e4 whatever Black does. That is community opinion without evidence, so check it against your own games. It fits why learning plans beats memorizing openings and the setups in our London System guide.
Englund Gambit declined
White does not have to take the pawn. Chess.com lists 2.d5 and 2.e4 as ways to decline, and Wikipedia adds 2.e3 and 2.Nf3. Every one of them is worth less on the engine's scale than 2.dxe5 at +1.38.
2.d5 scores -0.28. It peaks at 10.7 percent at 1600 in our data, then falls to 4.5 at 2200. 2.e3 scores +0.27 and is highest at 1000, 9.1 percent, shrinking to 3.5 at 2200. It keeps a pawn structure White knows and sidesteps the gambit, which suits a London player and is the natural fit for someone who dislikes theory. 2.c4 scores +0.02, rises from 5.8 percent at 1000 to 8.5 at 1800 and falls to 6.5 at 2200. 2.Nf3 scores -0.38, with 9.9 percent at 1000 and 7.0 at 2200. 2.e4 is -0.14 and transposes to king pawn territory, which can be a good way to refuse the gambit if you know it. Each decline means White is choosing to play a normal opening and gives up the extra pawn and the engine's plus. That is a fair trade for comfort. For many club players, comfort wins more games than a better engine score they cannot hold.
Should Black play it?
The honest answer is yes, in some games, with a clear picture of the price. With correct play the engine has White clearly better, +1.38 after 2.dxe5 and +1.33 after 2...Nc6 3.Nf3. Chess Doctrine calls the gambit objectively unsound against accurate play, and we agree with the word dubious. It is a sound practical choice in bullet and blitz because the trap is natural, because White has to find 6.Nc3 or a known alternative, and because the structure that results is easy to play. It is a poor choice in slow games where White has time and a book.
Where the opening is bad is easy to say. It is bad when White knows 4.Bg5 or 4.Qd5, because Black has given a pawn for a lead in development that does not turn into an attack. It is bad when White declines with 2.e3 or 2.Nf3 and Black has no pawn to play for, though that costs White some of the engine's plus. It is bad when you reach move six and White has played the correct 6.Nc3, because the sacrifice has bought you a queen on b2 that may get stuck. It is a good fit for a player who plays many short games, enjoys tactics and is happy to lose some of them for the chance of a quick win. Compare it with the Scandinavian Defense data by rating, which is a sounder way to meet a different first move, and for a longer view on how these choices connect, see our guide to the best chess moves by position.
Where other pages are better
This page is narrow on purpose. Wikipedia has the fuller variation tree and the history. Chess.com's page is better for the Levy Rozman angle and the decline options. 365chess gives aggregate database numbers we cannot reproduce. Chess Doctrine and Ian's Chess Gambits go deeper on the 4.Qd5 and 4.Nc3 answers. What none of them does is show what White plays at each rating on moves two and three, or check the trap line with an engine. What we do not measure is also worth saying: any White move after the third, the frequency of 4.Bf4 or 6.Bc3, and any result data.
What to take from it
The Englund Gambit gets a lot of trap talk and not much honesty about what happens before it. At every band from 1000 to 2200 the likeliest White reply is 2.dxe5, and after 2...Nc6 the likeliest is 3.Nf3, so Black usually gets the position it wants. The engine has White at +1.33 there, and the way White cashes in is to answer 3...Qe7 with 4.Bg5 or 4.Qd5 and, if the queen reaches b2, to play 6.Nc3 and not 6.Bc3. If you play White, learn that single move. If you play Black, play it for the traps and know you are paying for them.
Maia3 is a model of human choice built from the maia3-5m checkpoint, not a count of real games, and every percentage above should be read that way.
Frequently asked
- Is the Englund Gambit good?
- Not against accurate play. Stockfish 18 at depth 20 scores the position after 1.d4 e5 2.dxe5 at +1.38 for White, and after 2...Nc6 3.Nf3 at +1.33. That is a clear plus, though not a lost game for Black. As of October 2026, Chess.com says the gambit has surged among club players, partly because of IM Levy Rozman's video, and is not popular among stronger players. It works as a surprise weapon in fast games because the trap is natural to play, not because the engine likes it.
- How do I play against the Englund Gambit?
- Take the pawn with 2.dxe5 and develop with 3.Nf3. Our Maia3 data says that is what White does most often at every band from 1000 to 2200, and it is also Stockfish's top move at +1.33. After 3...Qe7 the engine's first choice is 4.Bg5 at +1.47. Wikipedia and Chess Doctrine favour 4.Qd5, and Ian's Chess Gambits calls 4.Nc3 sensible. No source agrees on one best answer, so pick one and know the follow up.
- What is the Englund Gambit trap?
- It runs 1.d4 e5 2.dxe5 Nc6 3.Nf3 Qe7 4.Bf4 Qb4+ 5.Bd2 Qxb2 6.Bc3?? Bb4!. The bishop on c3 is pinned to the king on e1, and after 7.Qd2 Bxc3 8.Qxc3 Qc1# it is mate. Our Stockfish 18 check at depth 22 puts 6.Bc3 Bb4 7.Bxb4 Nxb4 at -4.57 for White. The right move at move 6 is 6.Nc3, which keeps White at +1.74.
- What happens if White declines the Englund Gambit?
- White has several quiet moves, and each is worth less on the engine's scale than taking the pawn. Stockfish 18 gives 2.e3 +0.27, 2.c4 +0.02, 2.d5 -0.28 and 2.Nf3 -0.38, against +1.38 for 2.dxe5. In our Maia3 data 2.d5 peaks at 10.7 percent at 1600, and 2.Nf3 is the likeliest decline at both 1000 and 2200. Declining is playable, but it hands back most of the edge.
- Is the Englund Gambit the same as the Charlick Gambit?
- Yes. It begins 1.d4 e5, carries ECO code A40, and Wikipedia says it is named for Swedish player Fritz Englund, who sponsored a thematic tournament in 1932 and 1933. It is also called the Charlick Gambit after the Australian champion Henry Charlick.
Want to know whether you would walk into the Qe7 trap?
Chessdrive runs the real maia3-5m checkpoint on-device as your opponent at any rating you set from 600 to 2600, so you can play the Englund Gambit against a human-like White, or defend against it, and find out what you do under a clock instead of what a table says. Finished games are reviewed separately with Stockfish so you can see the exact move where the position turned. Available now on iOS.
Keep reading
Updated