Converting a Winning Position in Chess: Why One in Four Never Gets Converted
October 20, 2026 · ChessPivot · Guide
Episode 5 of the series "The Immortal Mistakes", drawn from the Amateur Chess Observatory, Fall 2026 edition.
You have won a piece. The engine would say +5. All that is left is to "finish". Twenty moves later, the game is a draw — or lost. Every player knows this scene; the data says it is not an exception.
The Amateur Chess Observatory followed 52,005 real games, from 400 to 2399 rating, move by move. Among the indicators re-measured in every edition, one describes precisely what happens to winning positions.
What the data says
A winning position is one where the evaluation has reached a decisive advantage. In the lowest band, one in three does not become a win. In the highest band, still one in four. Sixteen hundred rating points buy only seven points of conversion.
Next to it, other indicators improve much faster: big mistakes per move are nearly halved, the hanging piece divided by three. Converting an advantage is a separate skill, and it improves slowly.
Why we fail to convert
Reading games, the causes fall into four families.
- Attention drops. Piece won, the player moves faster and looks less, and makes the kind of blunder they were not making before. The unconverted advantage page describes this mechanism as the analysis detects it.
- We simplify badly. Trading is the right idea, but we trade pawns rather than pieces, or trade our active piece for a passive one, and the advantage melts away.
- We enter an endgame we do not know. Rook versus pawn, opposite-coloured bishops, a passive king: an endgame entered with a large advantage gets lost or drawn through technical ignorance.
- We lose on time. A won game that drags on is a game where the clock keeps running. Time trouble turns a simple technique into a lottery.
What amateur endgames reveal
The endgame numbers shed a different light on conversion.
In other words, most amateur endgames start already won for one side. The question is not playing a level endgame — that is rare — but not throwing away an endgame already won. And contested endgames, when they exist, decide: in the 400-799 band, 86% of endgames entered level produce a winner; still 70% at 2000-2399. Equality does not survive there, one way or the other.
Conversion is therefore played in two places: before the endgame, to enter it with the advantage intact; and in the endgame, to not give it back.
The method: five rules to finish a won game
- Do not hurry. A winning position stays winning if you do nothing dangerous. The safest move is almost always better than the fastest one. Take the time you took to win the piece.
- Trade pieces, not pawns. With extra material, every enemy piece off the board reduces their counterplay; every pawn off the board reduces your winning chances. Keep your pawns, trade their pieces — above all their most active one.
- Secure the king first. The opponent’s counter-attack is the other side’s only hope. Before advancing anything, check that your king is not the target of a desperate sacrifice. In the endgame it is the reverse: activate the king, it becomes a piece.
- Create a second weakness. One enemy weakness can be defended; two cannot. Before forcing matters, look for where to open a second front — that is the principle of two weaknesses, and it is the most reliable conversion technique there is.
- Keep time for the end. As soon as you are winning, decide on a time reserve you will not go below. A technical endgame played with thirty seconds is no longer a technical endgame.
Before entering the endgame: the checklist
Most failed conversions are decided at the moment the player chooses to enter the endgame. Before accepting the last simplification, check:
- The type of endgame. Opposite-coloured bishops a pawn up is often drawn; rook and pawn versus rook has its known positions; knight versus bishop with pawns on both wings favours the bishop. Choose the endgame, do not suffer it.
- Your king’s activity. In the endgame, a centralised king is worth a pawn or two. If your king is still at the back and your opponent’s is already active, wait.
- Your passed pawns. A protected passed pawn is the safest asset in any endgame; make sure you will have one, or that your opponent will not.
- Your clock. A winning endgame requires precision; without time, keep pieces on and look for simple moves.
Common mistakes
- Attacking the king "to finish fast" a piece up, when a quiet simplification won without risk.
- Trading your last active piece for a passive enemy piece in order to "simplify".
- Pushing passed pawns too early, before the pieces support them: an advanced pawn without support becomes a target.
- Accepting a draw by repetition in a winning position, for fear of "losing the thread". If you cannot see the plan, play a safe move and keep looking.
- Playing fast because the position is won. That is the exact moment your opponent is counting on your speed.
Conclusion
Winning a piece and winning the game are two skills, and the data shows the second improves far more slowly than the first: one winning position in four stays unconverted at 2000 rating. The good news is that converting an advantage rests on simple, stable rules — do not hurry, trade pieces, create a second weakness, keep time. An analysis of your games on ChessPivot spots each of your unconverted winning positions and shows you the move where the advantage started to melt.
Read also
Frequently asked questions
- What counts as a winning position in these numbers?
- A position where the engine's evaluation has reached the decisive-advantage threshold used by the study for one side, in a game still in progress. "Not converted" means that side did not end up winning the game: a draw or a loss. The number does not count positions that are merely "better", where the outcome is genuinely uncertain.
- Should you attack the king or simplify when you are a piece up?
- Simplify, in the vast majority of cases. A piece up, an endgame with few pieces is won almost mechanically, whereas a king attack requires precision and gives the opponent a chance. The attack is only justified if it is clearly winning within a few moves and carries no risk for your own king.
- Why are amateur endgames so often already decided?
- Because amateur games are mostly decided in the middlegame: a hanging piece, a missed combination, a missed mate. By the time the endgame arrives, the material gap is already there — in the 400-799 band, nearly three endgames in four start with more than three pawns of difference. Conversion, among amateurs, is therefore mostly about not throwing away an endgame already won.
- How do you stop losing won games on time?
- By deciding on a time reserve as soon as you are winning, and choosing simple moves rather than optimal ones when that reserve gets close. A won position is also won with the second-best move; it is lost with the best move found too late. If the time control has an increment, play the obvious moves fast to rebuild time before the technical moments.
- Is endgame technique really necessary below 1500 rating?
- Less than people say, and more than people think. Less, because only one game in four or five reaches an endgame at those levels, and most arrive already decided. More, because genuinely contested endgames almost always end with a winner: whoever knows the basic positions wins those. A few elementary endgames — king and pawn versus king, rook and pawn versus rook — are enough to start.