Week 82: why do I keep blundering in chess? 23 blunders in 11 games
Week 82 is the week I'd file under "why do I keep blundering in chess?". I played 11 rapid games between August 18 and 20, 2025, won 3, lost 8, and my rating fell from 1208 to 1171. Stockfish counted 23 blunders in my moves, an average of more than two per game, and the week ended with five losses in a row. It wasn't one disaster. It was the same kind of small oversight, game after game.
The week in numbers
| What | This week |
|---|---|
| Rated rapid games | 11 |
| Score (W–L–D) | 3–8–0 |
| Rating change | -37 (1208 to 1171) |
| Average accuracy (Stockfish) | 78.2% |
| My blunders | 23 |
| Openings I played most | Modern Defense with 1 e4 (2), Indian Game (2), English Opening Four Knights Korchnoi Line (1) |
Seven of the eleven games were on a single day, August 18, and I went 3–4 that day. The other four games, spread over the next two days, were all losses. One of them, game 1641, has no moves in the record at all: it ended by resignation before I played a move, so it has no accuracy and no blunders. Three games were at 15 minutes plus 10 seconds instead of my usual 10 minutes. The slower clock didn't save me: I lost two of those three.
Where the 23 blunders happened
Before looking at single moves, I counted. Of my 23 blunders, 17 came in the middlegame, 3 in the opening and 3 in the endgame. None of them was flagged as played with a low clock. That matters, because "I was short of time" is my favourite excuse, and this week it isn't true. I had time. I used it to look at my own plan instead of at what my move changed.
There's another number that stings. In four of the eight losses (games 1637, 1639, 1642 and 1643), Stockfish says my best winning chance during the game was above 80%. So it isn't that I was outplayed from move one. I had good positions and gave them back.
Three blunders, three versions of the same mistake
Against Joy_Absurd (1234) the position was roughly level (+0.53). My 10...Nd4 attacked White's bishop on f3, which felt like the active choice. But the knight on c6 had a second job I didn't see: it stood on the long diagonal between that bishop and my pawn on b7. Once it moved, the bishop simply took the pawn with 11.Bxb7, also hitting my rook on a8, and I had to spend a move with 11...Rb8. The evaluation jumped to +2.48. The engine's 10...Qd7 isn't spectacular; it develops the queen and eyes the h3 pawn, which had no defender. The point is that it doesn't remove a wall.
As White against gerolxbaros (1226), I was clearly better (+2.00) after a London-style opening. Then 11.Qd3. Black answered 11...d5, a pawn attacking my bishop on c4. The natural retreat for that bishop was d3, and my own queen was standing there. I put the bishop on b5 instead, and the c6 pawn took it. Stockfish's move was the same square for the other piece: 11.Bd3, bishop first. The evaluation went from +2.00 to -4.92 with one queen move. I eventually lost it on time, with 5 seconds left after my last move, but the position was already lost; the clock just made it official.
The third one is from a game I won by checkmate, against FunkeyPandeyy (1193), which is why it's useful. I was winning (-3.90; minus means Black was better). White's queen on c7 was attacking my e7 pawn, so I pushed it to e6. Stockfish called that a blunder (-0.92). Its move was 20...Qh4, attacking the knight on h2 with White's king right behind it. In its line, my bishop joins with ...Bf4 to attack the same knight, and the engine is happy to let the e7 pawn go later. I defended a pawn when I could have kept attacking a king.
The lesson of the week
All three moves have the same root: I moved a piece without asking what it was already doing. The knight on c6 was blocking a diagonal. The square d3 was the bishop's escape route. My queen on h3 was part of an attack that mattered more than a pawn. My checklist from this week is one question long, and I now try to ask it before every non-obvious move: what does this piece stop doing when it moves?
All games this week
| # | Date | Colour | Opponent | Result | Opening | Accuracy | |
|---|---|---|---|---|---|---|---|
| 1633 | 2025-08-18 | Black | FunkeyPandeyy (1193) | Win | Modern Defense with 1 e4 | 86.2% | game |
| 1634 | 2025-08-18 | Black | Joy_Absurd (1234) | Loss | English Opening Four Knights Korchnoi Line | 70.8% | game |
| 1635 | 2025-08-18 | White | gerolxbaros (1226) | Loss | Indian Game | 77.2% | game |
| 1636 | 2025-08-18 | Black | kandu12345 (1325) | Win | Pirc Defense | 94.6% | game |
| 1637 | 2025-08-18 | White | kenay2018 (1222) | Loss | Queen's Pawn Opening Zukertort Variation | 84.0% | game |
| 1638 | 2025-08-18 | White | DRViganj (1188) | Win | Modern Defense with 1 d4 | 89.9% | game |
| 1639 | 2025-08-18 | Black | miladgholami0278 (1218) | Loss | Van 't Kruijs Opening | 69.9% | game |
| 1640 | 2025-08-19 | White | bashkim3467 (1217) | Loss | Indian Game | 79.1% | game |
| 1641 | 2025-08-19 | White | DavidMercurio (1207) | Loss | — | — | game |
| 1642 | 2025-08-19 | Black | Hasanrashi (1189) | Loss | Modern Defense with 1 e4 | 78.3% | game |
| 1643 | 2025-08-20 | Black | shm12334 (1169) | Loss | Pirc Defense Main Line | 51.9% | game |
Previous week: week 81. Next week: week 83. Every week is listed on the journey page. If you want to know whether a chatbot can do this kind of review, I wrote about whether Claude can analyze chess games.
Frequently asked questions
Why do I keep blundering in chess?
In my case this week, mostly in the middlegame: 17 of my 23 blunders came there, and none of them was played with a low clock. They came from moves that changed what a piece was already doing, not from time pressure.
Do longer time controls stop blunders?
Not for me this week. I played three games at 15 minutes plus 10 seconds: I won one at 94.6% accuracy and lost the other two, with 5 and 2 blunders.
How many points did 23 blunders cost me?
My rating went from 1208 to 1171, a drop of 37 points over 11 games, with 3 wins and 8 losses.