Week 98: my rook was captured with check

From December 10–14, 2025 I played five games, with one win and four losses, and my rating moved from 1337 to 1313 (−24). The engine counted 16 blunders and an average accuracy of 60.7%. A rook capture with check after I attacked a queen is the first case below, followed by other material-losing decisions.

My rook was captured with check: the week in numbers

Five games, five different openings. My only win came against the Englund Gambit (1.d4 e5), and it wasn't pretty: 50.3 % accuracy, three blunders of my own, and my opponent abandoned on move 17. Twelve of my sixteen blunders came in the middlegame.

A rook that attacked the queen and died: game 1883

On Wednesday I played the London System as White against emicra1973 (1378). Black's queen was on c6 and their rook on c8, both on the same file. I saw that and jumped.

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

The position.py check describes 15.Rc1 honestly: it attacks the queen on c6 and the rook behind it, "but the rook can now be captured". Black's queen simply took it with check, 15...Qxc1+, and the engine's line goes on with the queen also taking my other rook on h1 (−9.49). I lost a rook and the game was abandoned on that same move. The engine's quiet options — castling, h3 or Kf2 — all kept me only slightly worse (around −0.8). The game is here.

The lesson is boring and true: an attacking move is worth nothing if the attacker can be taken. Before I put a piece on a square, I need to count who defends that square and who attacks it. On c1 the rook was simply taken, with check, so my next move had to be a king move and nothing was ever recaptured.

Attacking the queen just because I can: game 1885

Later that Saturday I had Black against 27aripin (1332), who opened with 1.b3. On move 5 White's queen was on f3, so I attacked it.

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

The problem with 5...Bg4 is where the queen goes after being attacked. The engine's first answer is 6.Qxb7: the queen takes a pawn and attacks my rook in the corner, which has no defender (+1.63). By moving my c8 bishop away I had also taken away the piece that protected b7. With 5...Bg7 I was slightly better (−0.82). My opponent played 6.Qg3 instead, but the b7 pawn fell anyway on move 9, and the game got wild: this was my worst game of the week, 45.1 % accuracy and nine blunders. I even had a winning position at one point (−5.51 before 28...Nd7 turned it into +5.20) and was mated on move 39 (game).

Two more pieces on Sunday

Sunday was two more losses. Against HypeR129714 (1329) in a Pirc, my 7...Be6 walked into 8.d5, a pawn move that attacks my knight on c6 and my bishop on e6 at once — a fork, a single move that threatens two things (+6.40). Taking the knight on f3 instead kept the game about equal. That game lasted 12 moves. In the evening, against dan1el2rtyu (1349), I was winning by more than six pawns on move 32 and played 32.dxe5. Black's pawn had just moved from e6 to e5 and opened the diagonal of their bishop on c8 towards h3, where my queen stood. The engine's line starts with 32...Bxh3, and that is what happened.

The lesson of the week

Hanging pieces isn't one mistake, it's a missing habit. This week I lost material in three ways: a piece moved to a square where it could be taken (15.Rc1), a move that removed a defender (5...Bg4 left b7 alone), and ignoring what my opponent's last move uncovered (31...e5 opened a line to my queen). The habit I'm building now is a two-question check before I touch a piece: what does my opponent's last move attack? and after my move, which of my pieces can be taken for free?

All games this week

← Week 97 · Week 99 → · Back to the journey hub · How Chess Coach reviews games like these

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