Flagged in a winning position: game 1344
I got flagged in a winning position in game 1344: before my 57th move the saved evaluation was -6.90, completely winning for me as Black, but I had 14 seconds on my clock. I played the wrong capture, and nine moves later the record says bonjours (1365) won on time.
The recorded game
I played this game on January 22, 2025 as Black, on my Chess.com account nacsol3. My recorded rating was 1302. It was a ten-minute rapid game without increment, part of the sequence used for the From 631 to 1385 journey.
| Detail | Saved record |
|---|---|
| Game | 1344 |
| Opening | Pirc Defense (B07), with 2.Bc4 |
| Result | Loss (1-0), bonjours won on time |
| My accuracy | 63.1% (White 68.6%) |
| My blunders / mistakes / inaccuracies | 11 / 4 / 6 |
| White's blunders / mistakes / inaccuracies | 12 / 3 / 7 |
| Accuracy by phase | Opening 94.0%, middlegame 58.5% over 52 moves |
| My best winning chance in the game | 93.8% |
| Original score | Chess.com game |
I use the word collapse for what happened between move 57 and the flag, not for a claim that I was winning from start to finish. Both of us made a lot of errors. White's count of blunders was even higher than mine. But I had the last big chance and the least time to use it.
Context: just over 1300, on a bad run
This game came one week after my first crossing of 1300. Games 1340 and 1341 were wins and took me to 1324. Then I lost games 1342 and 1343, both with White, and this was the third loss in a row. It wasn't the last: I also lost games 1345 to 1348, seven straight defeats, and my recorded rating went down to 1273. Game 1345 was the first time I was below 1300 since crossing it, and I didn't get back to 1300 until game 1398. The weekly diary puts it alongside the other results.
The opening was fine
After 1.e4 d6 2.Bc4 g6 3.Nc3 Nf6 4.d3 Bg7 I castled and pushed my queenside pawns: c6, b5, a5 and then b4. My twelve opening moves scored 94.0% with no errors. The first saved mistake came on move 15, when I moved my queen to d6 and Stockfish preferred taking White's bishop on b3 with my knight first. I took it on the next move anyway.
Position 1: move 22, the f-pawn waits
By move 22 I had a big advantage. White's 22.Rc1 was a blunder, and the evaluation before my move was -5.31. The numbers are from White's side, so minus is good for me. I moved my bishop from b7 to a6. The evaluation went to -0.75.
Stockfish wanted 22...f4: the pawn goes forward and attacks the knight on g3. Its line keeps pushing that pawn, to f3 and then capturing on g2. On move 23 the engine's first choice was again f4, and I moved the bishop back to b7 (marked as a blunder, -4.35 to +0.08). On move 24 it was f4 a third time, and I put a rook on c8. I did push the pawn to f4 eventually, on move 34, in a completely different position.
I had more than seven minutes on my clock at this point. Those were not time-trouble moves. They were the start of a long game in which I kept choosing slower moves than the position needed.
Position 2: move 41, swapping into equality
After a lot more back and forth, I was winning again. Before my 41st move the evaluation was -6.11. White's queen had just given check from h5. I blocked with my queen on h6.
The checker says my move attacked White's queen on h5, which had no defender. That sounds active, but the engine's evaluation after it was 0.00. Its own line after my move is 42.Qf7+ Qg7 43.Qh5+, with the queens checking back and forth. Stockfish preferred the plain king move to g7, which kept -5.46 in its check. I had 138 seconds left after this move. White had 266.
The record can't tell me why I chose the queen move. It does show that a quiet king step was better, and that I had less than two and a half minutes to find it.
Position 3: move 57, the turning point
Between moves 46 and 53 I blundered five more times while White kept checking my king, with less than a minute on my clock. One of those moves, 51...Kg8, matched the engine. Then White's 57.Qe1 was a blunder, and the evaluation before my reply was -6.90. I had 14 seconds. White had 112.
I took the pawn on a4. Stockfish wanted my queen to take on g3: the checker describes it as a capture of one of the pawns protecting White's king, with check, attacking the queen on e1 and the knight on g2. The saved engine line is 57...Qxg3+ 58.Qxg3+ Bxg3+ 59.Kxg3 b3: the queens come off and I keep a bishop and a passed b-pawn against a knight. That is a much simpler position to play with 14 seconds than one full of queen checks. My capture on a4 moved the evaluation to +1.37, now better for White.
That is the single move the analysis picks as the turning point. In time trouble a capture with check that forces a queen trade is exactly the move I want to look for first, because it makes the game simpler.
Position 4: move 64, three seconds left
The position kept changing and was about equal by move 64. I had 3 seconds on my clock. My pawn on b3 was two steps from promotion, and I pushed it.
The checker sees the problem clearly: the pawn push stops defending my queen on a2, which can now be captured. White's knight took it (65.Nxa2), I promoted with check (65...b1=Q+) and White's queen took my new queen. The evaluation was +8.72 after my push. Then my clock ran out. Stockfish's choice was 64...Qa1+, a check that also attacks the knight on c3 (a double attack), and its evaluation of that line was about equal.
The lesson of game 1344
This is not a game I lost because of one bad move. I had a winning advantage in the middlegame (moves 22 to 24) without finding the forcing pawn push, and by move 41 I was down to 138 seconds. That left me playing a long, sharp queen ending with seconds, against an opponent with more than a minute and a half.
Two things I take from it. First, when I'm clearly better and have time, I should look for the forcing move: here the f-pawn push that attacks a piece. Second, when I'm winning and short of time, I should look for trades that make the position simple, like the queen trade on move 57. The answer on losing on time and the one on why winning positions get lost cover both habits in general.
Where this sits in the journey
The recorded ratings are 1309 for game 1343, 1302 for this game and 1296 for game 1345. They describe a slump, not a forecast. If you want to look at your own lost-on-time games the same way, with every key moment on a board, you can use Chess Coach in Claude or ChatGPT. The guide to reading Stockfish analysis explains the evaluation signs, and the journey hub has the rest of the series.
Frequently asked questions
How did game 1344 end?
I lost on time with Black against bonjours (1365). The record says bonjours won on time after White's 66th move.
Was I really winning?
Several times. The saved evaluation reached -6.90 for Black before my 57th move, and my best winning chance in the game was 93.8%.
How much time did I have left?
My clock read 138 seconds after move 41, 14 seconds after move 57 and 3 seconds after move 64. White had 112 seconds at move 57.
What did this loss do to my rating?
It was the third of seven straight losses. My recorded rating went from 1302 in this game to 1273 in game 1348.