An engine tells you what was best. It cannot tell you what you were thinking, and that is what you need to change. Coaches have long recommended the same routine: your own notes first, the engine second.
Straight after the game
The best time is the same evening, before the game blurs. If you played a club opponent, the few minutes after the game, going through it together, are worth more than any engine.
Mark the moments you thought hard
Play through the game and stop wherever you spent a long time on the clock, or felt unsure. Those are the moments to learn from. Bookmark them.
Write what you feared
At each marked moment write, in a sentence or two, what you were afraid of, what you hoped for and which moves you considered. Be honest; nobody else needs to read it.
Try it on the Opera Game from Black's side. After 9. Bg5, Black is behind in development and pinned on f6. What would you consider?
A good note here might say: "b5 attacks the bishop but my king is still in the centre; can White open lines?" That question is the lesson, whatever the engine says.
Then switch the engine on
Now run a review. Look first at the moments you marked: did the evaluation change there? Then look at the key moments the engine found that you did not mark. Those are the ones you did not see.
Compare, and keep one lesson
Do not try to fix everything. Write one lesson per game in the game notes, in words you will understand in a year: "When my king is in the middle, count the open files before pushing pawns." After a season, read them back.
Written 4 November 2026. A routine, not a rule: coaches differ, and some prefer reviewing with a stronger player before any engine.