The Englund Gambit Bg3+ trick I walked into, and the checks I didn't play
The Englund Gambit Bg3+ trick decided nothing in game 1497, but only because my opponent didn't see it. On move 6 I pushed my h-pawn and allowed a knight sacrifice on f2 followed by a bishop check on g3 that would have won my queen. Andrei1990kuz (1265) retreated instead. I then got a winning position three times, missed the checks that finished it, and lost by checkmate with 23 seconds left on my clock.
The recorded game
I played this game on May 24, 2025 as White, on my Chess.com account nacsol3. The rating recorded for this game was 1252. It was a ten-minute rapid game without increment, one game in the From 631 to 1385 journey.
| Detail | Saved record |
|---|---|
| Game | 1497 |
| Opening | Englund Gambit (A40) |
| Result | Loss (0-1), by checkmate |
| My accuracy | 60.6% (Black 62.5%) |
| My blunders / mistakes / inaccuracies | 8 / 2 / 8 |
| Accuracy by phase | Opening 50.1%, middlegame 67.7% |
| My best winning chance in the game | 97.5% |
| Original score | Chess.com game |
Three games earlier, in game 1494, I had already lost an Englund Gambit game to Manoj_2303 (1276), with 40.8% accuracy. Game 1497 was the next time 1...e5 appeared against my 1.d4. The weekly diary for that week covers the two games I played that week, this one and a Pirc win.
How the opening went
The Englund Gambit is 1.d4 e5: Black offers the e-pawn on move one. I didn't take it. I played 2.c3, and the engine marks that as a mistake: Stockfish wanted 2.dxe5, and my evaluation dropped from +1.30 to -0.02. Black answered 2...d6, so it was no longer much of a gambit. After 3.Nf3 Nf6 I took on e5 after all, 4.dxe5 Ng4 5.exd6 Bxd6, and the d-file was empty from Black's queen on d8 to my queen on d1. That open file is what made the trick possible.
Position 1: move 6, the pawn move that allowed the trick
Black's knight on g4 was annoying me, so I kicked it with 6.h3. The saved analysis rates that as a blunder: from +1.29 to -5.92. Stockfish's choice was the quiet 6.Nbd2, which the checker rates as "slightly better for white".
The reason is a trick in three moves. Today's check gives Black 6...Nxf2: the knight takes the pawn next to my king and attacks my queen on d1 and my rook on h1 at the same time. If my king takes the knight, 7.Kxf2, Black plays 7...Bg3+. The bishop gives check, and by leaving d6 it opens the d-file, so Black's queen on d8 now attacks my queen on d1. That is a discovered attack: one piece moves and uncovers an attack by the piece behind it. The line ends 8.Kxg3 Qxd1, and the checker rates the position as "black is winning".
My h3 didn't create the trick, but it spent the move I needed to stop it. Stockfish's 6.Nbd2 puts a knight on d2, between the two queens, so a bishop check on g3 would no longer uncover anything. The saved record also marks the knight retreat 6...Nf6 as a blunder by Black: it went from -5.92 back to +1.64.
Position 2: move 7, the knight that walked into a capture
I didn't learn from it on the spot. One move later I made the same kind of error in a different shape.
I put my knight on e5, in the centre and in front of Black's bishop on d6. The checker's note on my move is short: the knight can now be captured. Today's line for Black is 7...Bxe5 8.Qxd8+ Kxd8: the bishop takes the knight, and because the bishop came off d6, the queens face each other on the d-file again. The queens come off, and I am simply a knight down. The saved evaluation after my move was -6.10. Stockfish's choice, 7.Na3, kept +1.54 in today's check.
Black played 7...Nc6 instead, attacking my knight, and I traded it. The saved analysis marks that as Black's second blunder in a row, back to +1.08 for me. I had been lost twice in two moves and was better again. I didn't notice either time.
Moves 13 to 29: lost, then winning
The middlegame followed the same pattern. On move 13 I grabbed the pawn on c6 with my queen (a mistake, -1.26), and on move 15 I played b3 (a blunder, -2.61). Black's queen took on c3, and by move 17 I was at -5.22. Then Black played 17...Bb4 when the engine wanted 17...Bc2, and the evaluation flipped to +1.80 for me. A few moves later, after Black took my rook on d1 with 20...Rxd1, I had +4.35.
From there I didn't throw anything away for a while. Moves 21 to 29 have no blunders of mine in the saved record, only two inaccuracies. But the clock was moving. On move 30 I had 120 seconds left. Black had 269.
Position 3: move 30, the first check I didn't give
Black's queen had just gone to d6, and the saved analysis marks that as a blunder: +7.19 for me. The checker describes Stockfish's move as a double attack: 30.Qe8+ gives check and hits the pawn on f7, which has no defender. Today's line goes 30...Kh7 31.Bxf7, and my bishop and queen then threaten checkmate. I played 30.Qb3, a quiet queen move, and the evaluation dropped to +1.78.
Position 4: move 36, mate on the board
The second chance was bigger. On move 35 Black's bishop went back to b6, and the saved evaluation jumped to +15.08.
Today's check rates 36.Qc8+ as "white can force checkmate". The line is 36...Bd8 37.Qxd8+ Kh7 38.Bxf7: Black has to block the check with the bishop, my queen takes it, and the bishop on c4 joins the attack on f7. I had 76 seconds. I moved the bishop to d3, and the evaluation fell to +2.12. On the very next move Stockfish wanted the same check, 37.Qc8+, and I played 37.f4 instead. That one went to -0.11.
Position 5: move 42, the pawn I shouldn't have taken
After a few queen checks the position was level. Black's 40...c4 was a mistake that gave me 0.00, and I gave a check on c8. Then this:
I took the pawn on c4. My clock said 23 seconds; Black's said 175. The saved evaluation after my capture was mate in two for Black, and it happened: 42...Qf2+ 43.Kh1 Qg1#. My king had only the corner, and the queen came to g1 next to it, protected by the bishop on b6 along the diagonal. Stockfish's choice was another check, 42.Qf5+, which today's check rates as "about equal": a draw by repeated checks was still available.
The lesson of game 1497
I see two lessons, and they belong to different parts of the game.
The opening lesson is about the Englund Gambit with an early d-pawn trade. Once the d-file is open between the two queens, any bishop or knight standing on that file is a curtain. When it moves, especially with check, my queen is attacked. The Bg3+ trick and the Bxe5 capture on move 7 are the same idea. In game 1497 Black missed both, and I got away with it. My new habit before any move in that structure: look at what happens to my queen if Black's d6 piece moves.
The second lesson is the one that cost me the game. Three times (moves 30, 36 and 37) Stockfish's choice was a queen check, and three times I played a quieter move. On move 42 a check was again the only move that kept the balance. Checks are the first moves to look at, and they matter more with little time, because they limit what the opponent can do. I wrote more about the clock side in why I keep losing on time, and about lost wins in why winning positions get lost.
Where this sits in the journey
The recorded ratings go 1260 for game 1496, 1252 for this game, and then 1260 and 1252 for games 1498 and 1499. Those numbers only say what came next; they don't prove this review changed anything. I have two earlier Englund lessons: the missed en passant capture in game 955 and the queen chase in game 850. If you want to check your own games for missed checks, the beginner game-review guide explains the method, and Chess Coach in Claude or ChatGPT draws these same boards and arrows for a finished game. The rest of the series is on the journey hub.
Frequently asked questions
What is the Bg3+ trick in the Englund Gambit?
In my game, after the d-pawns were traded, Black's knight could take on f2 attacking my queen and rook. If my king took the knight, the bishop on d6 jumped to g3 with check and Black's queen on d8 could take my queen on d1, because the bishop no longer stood between them. Stockfish rates the knight capture as 'black is winning' (-5.54 in today's check).
Did Black play the trick in game 1497?
No. Black retreated the knight to f6 instead, and on the next move missed another chance when my knight went to e5 and could be taken. The saved analysis marks both of Black's moves as blunders.
Why did I lose game 1497?
I missed winning checks on moves 30, 36 and 37, and on move 42, with 23 seconds on my clock, I took a pawn on c4 that allowed mate in two. Black had 175 seconds at that point.
What opening was game 1497?
Chess.com filed it as the Englund Gambit (A40) because Black answered 1.d4 with 1...e5. I played 2.c3 instead of taking the pawn, and the game went 2...d6 3.Nf3 Nf6 4.dxe5 Ng4 5.exd6 Bxd6.