Week 64: why do I suddenly keep losing in chess? Two weeks, 8–14

At the end of week 62 I was at 1307. Two weeks later I was at 1267. Week 64 itself, March 31 to April 6, 2025, was 13 rapid games, 5 wins and 8 losses, rating 1282 → 1267 (−15), and together with the week before it makes 8 wins and 14 losses in 22 games. It felt sudden, which is why I looked at it the way I'd look at anyone's slump: why do I suddenly keep losing in chess, and what does the engine say changed?

The week in numbers

The week had two losing streaks: four in a row from game 1416 to game 1419 (April 2 and 3), then three in a row from game 1423 to game 1425. My average accuracy was 72.3%, after 70.7% the week before. The clock wasn't the issue: none of my 29 blunders came with less than a minute left. And the wins were fine, including 92.1% in a London System against AlexandrIu1961 (1291) and 94.4% in an 11-move Modern Defense against phtos2 (1245).

What actually changed: winning positions

Here is the number that explains the slump. Five of my eight losses this week came from positions where my winning chance had been 80% or more, and in week 63 it was three of six. Over the two weeks, 8 of my 14 losses were games I had been winning:

• Game 1417 against Virgile45456 (1278): I was at +7.62 on move 11 and +7.41 on move 15, and lost (below).
• Game 1418 against Shabs419 (1293): −4.22 for me on move 15; my queen took a pawn with 23...Qxb2 (−0.03 to +5.89) and I was mated.
• Game 1419 against mz1275mgmailcom (1254): +7.05 on move 18; 18.Nxa6 brought it to −0.27, and 22.Qa1 (+5.57 to −4.74) lost it. I was mated.
• Game 1425 against ASphotografer (1259): −14.55 for me on move 31, and 31...Qxf5+ turned it into +8.50.

I also checked something about streaks across all my games up to this week. After three losses in a row, I won 41.7% of the next games; my overall win rate was 46.8%. A small difference, but it's there.

Game 1417: the double attack I didn't play

With White against Virgile45456 (1278) in a Queen's Pawn Opening, Zukertort Variation (game on Chess.com), Black's rook and knight were still at home on a8 and b8. I was winning, +4.44.

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

Qb7 is a double attack, really a triple one: the undefended rook on a8, the knight on b8 and the bishop on d7 with the knight on e7 behind it. The engine line is 10.Qb7 Qb6 11.Qxa8. I took a pawn on a4 instead, +4.44 became +1.86, and then it got stranger. On move 11 I was winning again (+7.62) and played 11.Bb5; the engine wanted 11.Qxa5. On move 15, at +7.41, I played 15.Qxa5 and lost my queen for a bishop after 15...Nxa5. The evaluation went to −5.13, and the game ended as abandoned after 16...Rxb2.

Game 1416: a retreat that invited a pawn

With Black against tfong2005 (1271) in a Pirc Classical (game link), the position was nearly equal, +0.62 for White, after White pushed a pawn to e5 against my knight on f6.

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

Both moves retreat the knight. After Nfd7, the engine's answer is 8.e6, a pawn move close to my king that attacks the knight on d7. +0.62 became +2.50, and my opponent played exactly 8.e6: after 8...fxe6 9.Bxe6+ my king was in check on move 9. With 7...Ne8 the knight would not have been in the pawn's way. I later fought back to a winning position (−2.87 on move 21), and 21...Ne4 turned it into +7.27.

(The third key moment, game 1414 against AleksRadul (1291), an English Defense, was a knight move in a balanced position: 24.Ne4 attacked a pawn on c5, but after 24...cxd4 the evaluation was −5.17 for me. The engine preferred 24.Nf3.)

The lesson of the week

The slump wasn't about openings, the clock or stronger opponents. It was about what I did when I was winning. In game 1417 alone I had three moves at +4 or more and let all three go. Losing a won position costs the same rating points as losing a bad one, and it costs a lot more confidence. The rule I'm taking from these two weeks: when the position is clearly winning, every move gets a full check of the opponent's checks and captures, even the "obvious" ones, because the obvious moves are where I gave the games back. And after three losses in a row, I stop and review one of them before playing again; my own numbers say the next game goes worse.

All games this week

Read the start of the slide in week 63, or see what came next in week 65. Every week is on the journey hub, and the "winning positions lost" count is something Chess Coach can show for your games too.