Week 65: why do I blunder so much in chess? 68 blunders in 24 games

Why do I blunder so much in chess? Week 65 put a number on that question. Between April 7 and 13, 2025, I played 24 rapid games, went from 1267 to 1265 (−2) and finished with 11 wins, 12 losses and 1 draw. Stockfish flagged 68 of my moves as blunders, almost three per game. My opponents, rated 1273 on average, made 71. That is why my rating stood still: it was a week of mutual mistakes, and whoever made the last one usually lost.

The week in numbers

With White I went 7–5. With Black it was 4 wins, 7 losses and the draw. The Zukertort Chigorin, my main White setup that spring, split exactly 3–3.

Where 68 blunders come from

The total sounds dramatic, but it wasn't spread evenly. One game alone, game 1440 against tommyr5000 (1294), produced 15 of them. We traded blunders back and forth for 143 half-moves (he also made 15), and it ended as a draw by insufficient material with 2 seconds left on my clock. Game 1447 against howcaniknnow1 (1295) added seven more, and that was my lowest accuracy of the week, 45.6%.

At the other end, four games had zero blunders: 1428, 1429, 1438 and 1441. All four were wins. The best was game 1438 against Darwin-08 (1282): 98.8% accuracy and checkmate on move 59. So the same account that blundered 15 times on Friday had played a clean game on Thursday. Blunders are not a fixed personality trait. They show up in specific situations, and the data shows which ones.

The expensive ones were the seven games where I reached a winning position and didn't win. Game 1432 against iffe_a (1257) hurts the most: at one point the engine gave me a 96.8% winning chance, and I lost on time with my clock at zero. Then came four losses in a row (games 1445 to 1448) that took me from my high of 1296 down to 1265. In two of those four, 1445 and 1447, I'd had a winning position first.

Moment 1: two pawns up, and I let the rook in

Game 1430 against Yea4Jona (1267), with Black (game). By move 34 I was two pawns up in a rook endgame, six pawns against four, and the engine had me at −4.46, which means clearly winning for Black. I pushed my d-pawn one square.

. . . . . . . .
p p . . . . . p
. . p p . . . .
. . . . p . . k
. . . . . . . r
P . . P . . . .
. P P K . . R .
. . . . . . . .

The better move was Rg4: my rook attacks White's rook, which nobody defends, so White has to move it or trade. If the rooks come off, only kings and pawns are left, and I'm the one with two extra pawns. After my pawn move, White's rook gets to g7 instead, deep in my position, attacking my b7 pawn and my h7 pawn at the same time (a double attack). The evaluation dropped from −4.46 to −0.24, roughly equal, with a single quiet pawn move. The game ended in a loss on move 50.

The beginner version of the lesson: in a rook endgame, the rooks matter more than the pawns. Before you push a pawn, ask where the opponent's rook can go next.

Moment 2: a pawn instead of a bishop

Game 1431 against L0R3nZQ (1272), Black again, a Modern Defense (game). My knight on a5 was already attacking White's bishop on b3. Instead of taking it, I used my other knight to grab the pawn on e4.

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

Nxb3 takes a whole bishop. White can take my knight back, but then my other knight can still go after the e4 pawn, as the engine's line shows. After my pawn grab, White had Bd5: the bishop lands in the center and attacks two things at once, my knight on e4 (no defender) and my rook in the corner. The evaluation went from −2.05 (clearly better for me) to +0.26, about equal. I lost the game.

The lesson of the week

Both moments are the same mistake in different clothes. I made a move that looked free, a pawn grab or a pawn push, without asking what it allowed the opponent's pieces to do next. The habit that would have saved both games takes five seconds: after choosing a move, look at the opponent's checks, captures and attacks in the new position. If a bigger capture is on the board, take that one first.

The second lesson is about when it happens. Only 11 of the 68 blunders came in the opening; 57 came in the middlegame and endgame, and only 10 with less than two minutes on my clock. So most of them weren't panic moves. They were moves made with time to spare and no final check. The clock only finished the job in the two worst outcomes, the 15-blunder draw and the loss on time from a winning position.

All games this week

Previous week: week 64 · Next week: week 66 · Back to the journey hub · Want the blunders in your own games explained like this? Try AI chess analysis with Chess Coach.