Week 73: a rapid chess rating ceiling at 1277

From June 16–22, 2025 I played 34 rapid games: 16 wins, 17 losses and 1 draw, with my rating moving from 1277 to 1270 (−7). All were ten-minute games without increment. I reached 1277 four times and never exceeded it, so this entry follows that rapid chess rating ceiling at 1277 and two threats I failed to answer.

A rapid chess rating ceiling at 1277: the week in numbers

By colour it was almost perfectly flat: 8 wins and 8 losses plus the draw with White, 8 wins and 9 losses with Black.

The ceiling at 1277

The rating graph for this week looks like a saw. I started at 1277, dropped to 1245 by Tuesday morning after four straight losses, climbed back to 1277 with four straight wins the same day, and then kept bouncing: 1277 again on Wednesday, again on Thursday morning, and each time the next game was a loss. All four times. By Friday morning I was at 1246, and a three-game winning streak on Sunday brought me back to 1270.

The day-by-day split shows no magic day either: 1–2 on Monday, 4–3 on Tuesday, 3–3 on Wednesday, 1–4 on Thursday, 3–1 on Friday, 1–4 with a draw on Saturday and 3–0 on Sunday. After a loss I won 8 and lost 9; after a win I won 8 and lost 6. The games weren't getting better as the week went on. They were the same games, played again.

One more number I don't like: 12 of my 17 losses ended with the game abandoned, meaning I left instead of resigning or playing on. That's not a chess skill, but it is a habit, and it shows up week after week in my records.

Moment 1: the rook that was already attacked

Game 1535 against cole00000 (1265), Black, the Pirc's Lion's Jaw (game). On move 10 I had traded my g7 bishop for White's knight on c3, and the queens came off soon after. That left White's bishop on h6 staring at my rook on f8.

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

The threat was simple: the bishop takes the rook. A rook is worth more than a bishop, so even though I could take back, I'd lose material, "the exchange". I played a developing move with my knight as if the threat weren't there, and 13.Bxf8 Kxf8 followed. The evaluation went from +0.33 to +3.22. Re8 just moves the rook. Later in the game the engine actually gave me a winning chance of 92.7%, and I still lost.

Moment 2: attacking while my queen hangs

Game 1538 against Dynopro (1278), White, an East Indian London (game). Black's rook had come all the way to b2, and my knight could jump to c4 and attack it.

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

Nc4 attacks two pieces, so it looks active. But Black doesn't have to answer my threats: 16...Bf5 attacks my queen on d3, and Black is suddenly clearly better (−3.08). Bxe5 first removes the bishop and keeps the position equal. In the game I then put my queen on d5 (−8.52) and f3 (−9.96), and was checkmated on move 21.

The lesson of the week

Both key moments are the same mistake, and it's a rapid-chess mistake. With ten minutes on the clock you want to play your own ideas quickly, a developing move, an attacking move. But the first question in every position is: what is my opponent threatening right now? In game 1535 the threat was a bishop taking my rook. In game 1538 it was a bishop move that would hit my queen. Neither needed deep calculation, just looking at the last move before playing mine.

So, how to get better at rapid chess, according to this week: fewer games, with a review after each one, rather than 34 games that repeat the same error. The review is where the 1277 ceiling shows its cause.

All games this week

Previous week: week 72 · Next week: week 74 · The journey hub · To find the threats you missed in your own rapid games, review them like this.

For a general method beyond this week's games, see the rapid improvement guide.