Why do I keep losing winning positions? How to convert an advantage calmly

Being ahead changes the job. You no longer need the sharpest move. You need a move that keeps the advantage and takes away their fun. The clock still matters. Time management is how you arrive in that position with time left.

• Relaxing. The position feels decided, so you stop asking what their last move did.
• Hurrying. You want the game over, so you play a combination you have not checked.
• Greed. You grab a second pawn and open a line for their pieces.
• Time trouble. The conversion happens with seconds left, so safe moves lose to the clock.

A worked moment, as a method rather than a fake game: you are a piece up, their rook is on an open file, and you have a pawn to capture. Write two candidate moves. One captures the pawn. One stops the file or trades the rook. If you cannot say why their rook is harmless, do not take the pawn. That sentence is the whole conversion check.

Kill counterplay first

Name their threat. Meet it before you improve your own position. A winning position with an unanswered check is not winning for long.

Improve your worst piece

Find the piece that does nothing. Give it a square. You do not need a new tactic if their counterplay is already stopped.

Trade pieces, not pawns, when ahead

Trading pieces reduces the army that can attack you. Trading pawns can trade away the material you needed to win the ending. Keep enough pawns to make a passer.

Simplify into endgames you know

Trade toward an ending you have actually practiced. If you do not know the ending, do not head there just because an engine would win it.

Start with endings that show up when you are a pawn or a piece ahead: king and pawn opposition, a rook ending where the rook belongs behind a passer, and the habit of putting the king to work. The band notes at 1200 and 1400 include endgame ideas for those ranges. Practice one ending until you can win it without looking up the moves, then add another.

A drill

1. From your losses, collect positions where you were clearly better and then lost or drew.
2. For each, write the threat you ignored.
3. Replay the position and play only moves that answer the threat or trade a piece.
4. Set up one basic ending you missed and win it three times against a friend or an engine on a low setting.

The fingerprint compares your score in games of 40 full moves or fewer with your score in longer games. A worse score in the long games is a signal, not proof that your endgame is the weakness. Long games also contain time trouble and late blunders.

The deep scan tags a missed win when you were at least 200 centipawns better before the move and under 100 after it. That tag can be replaced if the best reply is a capture (hanging) or if the move took under 5 seconds (impulse). Count the missed-win tags that survive, and look at the phase. The single-game version of this review is how to analyze your own games. The batch version is how to find your weaknesses. Fast moves that throw the win away are also in how to stop blundering.

Common mistakes

• Assuming every long loss is an endgame loss.
• Trading pawns because "simplify" sounded like a rule.
• Looking for a faster win when a boring trade holds the advantage.
• Treating one missed win as a personality. Count them first.