Week 19: hanging a bishop with check

From May 13 to May 19, 2024 I played 16 rapid games, won 8, lost 8, and climbed from 1022 to 1038. The engine review shows three material-losing decisions: a rook capture, hanging a bishop with check, and a pawn push with an unsafe recapture. These are the three games behind this week's lesson.

Hanging a bishop with check: the week in numbers

The split was perfectly even: 4–4 with White and 4–4 with Black. What wasn't even was where the mistakes happened. Stockfish found 60 mistakes and blunders in my games, and 40 of them came in the middlegame, against 16 in the opening and only 4 in the endgame. Only 75 of my 446 moves were played in an endgame at all.

The most frustrating number: in six of my eight losses (#449, #451, #453, #455, #457 and #464) I had, at some point, at least an 80 % chance of winning. I also lost two games that ended nine minutes apart to the same opponent, NeverGonnaCheckYouUp, once against the Englund Gambit with White and once in a Danish Gambit with Black. Gambits caught me out this week. The King's Gambit was the exception: I won that one with Black, in a game where Stockfish counted 10 blunders for me and 12 for my opponent.

Hanging a bishop with check, and two other errors

A hanging piece is one your opponent can take for free, or for less than it's worth. Here are mine.

The rook: game #449 against Peshuntiy (1051)

I was Black in the Cozio Defense and I was winning, -4.48. One thing needed attention: White's pawn on h5 was attacking my knight on g6.

r . . . r . k .
p p p . . p p p
. . . p . . n .
. . . . . . . P
. . N R . P b .
. . P . B . . .
P P P . . . . .
. . K . . . . R

Stockfish just wanted the knight to step away, 17...Ne7. I took the bishop on e3 with my rook instead, and my rook could be taken right back: 18.Nxe3. Now I had given a rook for a bishop, White's knight on e3 attacked my bishop on g4, which had no defender, and the h5 pawn was still attacking my knight. The evaluation swung from -4.48 to +5.74, from winning for me to winning for White, in one move. I resigned one move later (game on Chess.com).

The bishop: game #450 against phoenixPGI (997)

Same day, now with White in the Zukertort Variation. I was clearly better, +3.02, and I spotted a pawn on h5 that I could take with check.

r . . . k b n r
. q . n . . . .
. p p . . p . .
p . . p . p p p
P . P P . B . .
. Q . . P . . .
. P . N B P P P
R . . . . R K .

15.Bxh5+ is the kind of move I play because it gives check. Checks feel forcing. But the rook on h8 was looking straight down the h-file, and 15...Rxh5 took my bishop for a single pawn. The evaluation went from +3.02 to -3.32. The engine's move was the plain 15.cxd5, taking a pawn in the centre. I still won this game, by checkmate after 85 moves of my own and with 44 seconds left on my clock, but only because phoenixPGI blundered even more than I did (11 blunders to my 7) (game on Chess.com).

The pawn: game #451 against fdsetretresttgfdgsd0 (1177)

The sneakiest one. I was White and completely winning, +6.34, against a higher-rated opponent.

q . r . . r k .
. . . . . p p p
p . . . p . . .
N p b p N . . .
. . . . . P . .
. . P . . . . .
P P Q . . P P P
R . . . . R K .

19.b4 attacks the bishop on c5, and the pawn looks protected by my pawn on c3. It isn't, really. Black can answer 19...Bxb4, and if I take back with 20.cxb4, the c-file opens and Black's rook plays 20...Rxc2, taking my queen. I checked that line with the engine afterwards. So the bishop could take my pawn for free and attack my knight on a5, which has no defender. The evaluation fell from +6.34 to +2.73. The engine's 19.Nb3 attacks the same bishop without giving anything away. This game then ended in a strange way: Chess.com logs it as a loss on time, although my clock still showed 763 seconds after my 20th move. The clock ran out before I made my 21st (game on Chess.com).

The lesson of the week

All three moves did something active: a capture, a check, a pawn attacking a piece. And all three failed the same test. The lesson I took from this week is a single question I want to ask before every capture, check or pawn push: after this move, what can take the piece I just moved, and can I really take back? The last part matters. In game #451 I could "take back" only by opening a file onto my own queen. For me, that question is the most practical answer to how to stop hanging pieces in chess. It doesn't need any opening knowledge, just one second before I let go of the piece.

All games this week

← Week 18 · Week 20 → · All weeks of the journey · Chess Coach finds these moments in your games too: review your own games like this.

For a general method beyond this week's games, see the piece-safety routine.