How the training loop works

EloNotes is designed for people who would rather learn by playing chess than by turning improvement into another pile of homework.

Step 1: play real games

Your own games are the raw material. The positions matter because you actually reached them, with your clock, your habits, and your level of calculation.

Step 2: find the engine signal

Stockfish helps identify where the evaluation shifted, which positions deserve a closer look, and what line held up better.

Step 3: group the pattern

One game can be noisy. Several games can reveal a habit. EloNotes looks for the repeated decisions that cost the most advantage across games, not just the single worst move.

Step 4: make it rememberable

The output should be one or two things to remember next game. Not twenty-seven annotations. Not a spreadsheet of regret. Just enough signal to make the next game more interesting.

Step 5: add human review when it matters

Guided adds human review on selected moments, so the most painful positions get teaching judgment on top of the app-assisted draft.