Training notebook · Review your games
Analyse your chess game before turning on the engine
Review your own thoughts first, identify uncertain decisions and bring specific questions to the engine instead of just reading its answers.
Preserve what you were thinking
Immediately after a game, write down the moves where you hesitated, changed your plan or felt surprised. Your memory of the decision is useful evidence and becomes harder to recover once an engine shows a convincing answer. Include your intended reply to the opponent’s best move, even if it now looks wrong.
You do not need to annotate every move. Select a few decision points and record why you chose your move, what you expected, and what you considered instead. If you cannot remember, say so. Inventing a sophisticated explanation afterwards makes it harder to diagnose the actual mistake.
Replay the game with a question
Go through the score without an evaluation bar. At each selected moment, describe material, king safety and the immediate threat. Then choose a candidate independently. Ask whether you would make the same decision now, with more time and without the pressure of the game result.
Focus on the first position you did not understand rather than only the final losing move. A late queen loss may be obvious, while an earlier trade left you defending a position you never evaluated properly. Both matter, but they require different training. Personal analysis helps you distinguish them before a computer labels individual moves.
Test your own variation
Play through the line you thought was possible. Check that every move is legal and that you allowed the opponent an active defence. If your idea only works when the opponent cooperates, find the point where you stopped searching for their best reply.
Write a small number of unresolved questions. For example: does the rook exchange lead to a drawn pawn ending, can the king escape the attack, or is the apparently loose piece tactically defended? These questions make the later engine pass efficient because you have a concrete claim to test.
Bring in feedback and keep the distinction
Now use an engine or discuss the position with a stronger player. Compare their answer with your notes, not just with the move you played. An engine may refute the game move and your proposed improvement for different reasons. Follow enough of the line to explain the decisive feature.
Finish with one practical lesson and one position to revisit. Preserve unresolved questions if the line remains beyond your understanding; a human explanation may be more helpful than additional depth. This process does not make engine analysis unnecessary. It makes the feedback informative by showing what you could and could not find without it. The aim is an accurate account of your decision-making, not an impressive annotation or a guaranteed improvement in your next result.
Put the practice into a routine
Where Brass Pawn fits
Brass Pawn computer feedback can help test practice decisions after your own attempt. This workflow does not assume arbitrary external game import or annotation tools; use the game record and board your chosen platform actually provides.