Week 1 lesson
Your games, honestly
This course is built from your own games: 200 rated blitz games on chess.com, every one of them run through Stockfish. Every position you solve in the next 84 days happened on one of your boards.
What your rating did
| Month | High | Low | Games | Score |
|---|---|---|---|---|
| 2026-09 | 1385 | 1218 | 107 | 52% |
| 2026-10 | 1309 | 1245 | 46 | 54% |
| 2026-11 | 1294 | 1184 | 47 | 36% |
The upper number is each month's highest rating, the lower its lowest. The gap between them is the hot and cold streak.
| Window | Value |
|---|---|
| Current rating | 1184 |
| Peak | 1385 (2026-09-22) |
| Performance over your last 200 games | 1279 |
| Normal two-week swing at this speed | about ±100 |
Performance rating is the rating your results would earn against the opponents you actually faced: their average rating, plus 400 × (wins − losses) ÷ games. Over 200 games it is a much steadier number than the rating on your profile. Your drop from 1385 to 1184 (201 points) is larger than the usual swing of about ±100 at this speed. Review the patterns below before drawing conclusions.
So the goal of this course is not to chase a number. It is to move the performance underneath, so your current peak becomes a normal week.
What is not your problem
These are worth knowing because they free up study time:
- Tilt. You score 43.1% after a loss and 40.0% after a win. You don't tilt.
- The clock. 77% of your blunders happen with more than a minute left (median 114 seconds). You have the time; you don't spend it checking.
The leaks that are
Ranked by the rating points we estimate they cost you. The formula is simple and shown on your report: performance rating moves 4 points for every extra half-point per 100 games.
- Blunders left unpunished. 44% of their blunders go unpunished. Your opponents blunder 2.7 times a game. 44% of those chances pass, because your very next move gives the advantage back. That is vision, not calculation. (estimated +60 points; weeks 2)
- Won positions not won. 20% of your games. You had at least 80% winning chances in 39 games and did not win them. Converting half of them is worth about 47 rating points, with no new chess knowledge. (estimated +47 points; weeks 3)
- Early blunders. First blunder by move 12 in 48% of games. In 48% of your games the first real blunder comes by move 12, in every opening at about the same rate. That is a checking habit, not opening theory. (estimated +38 points; weeks 4, 5)
The three rules
Everything in this course trains three habits. They come in one at a time.
- Rule 1, their move first. Before you think about your plan, ask what their last move attacks and what it stopped defending.
- Rule 2, the hand-off check. After you have chosen a move and before you let go of the piece, list every check and capture they have after it.
- Rule 3, winning means slow down. Once you are clearly better, spend ten extra seconds a move, ask what their only hope is, and trade pieces, not pawns.
The positions behind the numbers
Your most recent engine-verified positions. The last move is highlighted. Look at it first, then find the move.
Find the best move: click a piece, then its square.
From your game on 12 Nov 2026, move 17 · View game ↗
Find the best move: click a piece, then its square.
From your game on 12 Nov 2026, move 20 · View game ↗
Find the best move: click a piece, then its square.
From your game on 12 Nov 2026, move 11 · View game ↗
Find the best move: click a piece, then its square.
From your game on 11 Nov 2026, move 13 · View game ↗
How each day works
Every day page has two tiers. The core is the day: about ten minutes (twenty on the day you read the deep dive, fifteen on the weekly test) of positions from your games, plus positions you have seen before coming back on a 2/7/21-day schedule. If that is all the time you have, you are done. The rest is optional: a ten-minute themed drill from the public-domain Lichess puzzle database, a capped play block, and a five-minute post-mortem. Solved positions are remembered: green means solved first try, amber means you needed help.