The four ways to throw away a winning position and what to do about each. An advantage does not defend itself.
It hurts more than losing from the start. You were a piece up, the position was clear, and thirty moves later you are signing a draw or shaking hands. It is not bad luck and it is not "I got nervous": winning an advantage and converting it are two different skills, and only the first one gets trained.
These are the four ways to throw away a won game, in order of frequency.
Number one by a distance. While the game is balanced, you calculate every move. Once you win the piece, your brain switches modes: from solving to managing. And managing, in practice, means moving fast, "not complicating things" and no longer looking at what your opponent is doing.
The problem is that the exact opposite is happening on the other side. The player who is losing has nothing to protect: they start hunting for checks, sacrifices and traps, because it is all they have left. One side eases off precisely as the other presses.
What to do: the same routine as always, and all the more so. Before every move: your opponent's checks, captures and threats. When you are winning, the check does not relax — it gets stricter.
A piece up, three ways to win appear, and the temptation is to pick the flashiest one. The exchange sacrifice leading to mate in seven, the combination that looks good.
The problem is that the flashy one is usually also the hardest to calculate, and a large material advantage does not need anything hard. It needs something simple.
What to do: with a clear advantage, always take the line that is easiest to calculate, even if it is slower. Winning in fifty moves with no risk beats winning in fifteen down a line you cannot verify.
There is one rule that almost solves conversion on its own:
When you are up material, trade pieces, not pawns.
With an extra piece, every piece trade makes your extra piece weigh more in what remains. With few pieces on the board, an extra rook is decisive; with many, it gets lost in the noise.
Pawns are the opposite: they are what lets you queen at the end. Trading them takes you towards drawn endings by insufficient material.
With an extra pawn the rule partly inverts: you still want to trade pieces, but carefully, so as not to reach a king-and-pawn ending that is drawn by opposition. At that point you have to know the endgame.
What to do: before every trade, ask whether it brings the won endgame closer or pushes it further away.
When the position is winning but two minutes are left, the game stops being decided by chess. It is a legitimate way to lose, and also the most avoidable: the time nearly always went earlier, on middlegame moves that did not need it.
What to do: manage the clock while the game is balanced, not once it is already won. And with extra material, simplifying also saves time: fewer pieces is less to calculate.
If you look at your thrown-away games, nearly all of them share a pattern: there is one move where you stopped having a plan and started making moves. It is usually the move immediately after winning the advantage.
That is the point to train. When you win material, stop for a minute and answer three questions before continuing:
A minute spent there is worth more than fifteen fast moves afterwards.
It is the least discussed cause and probably the most frequent below 1600: the game is won and you do not know what would have to be done to convert it. So you improvise, and every improvised move in a winning position hands your opponent a chance back.
Winning a won position is a technique and it can be learned, and it comes down to four cases:
A piece up: trade everything you can. Every pair of pieces that leaves the board increases your relative advantage. With rook against knight and level pawns, trading queens is practically the game.
A pawn up in an endgame: bring the king in. In an endgame the king is a strong piece, and activating it is often worth more than the extra pawn.
Attacking the king: add one more piece to the attack before sacrificing anything. The classic error is sacrificing with three pieces when the fourth was two moves away.
A large positional advantage: do not hurry. Improve your worst piece, one move at a time, and let your opponent — who has nothing to do — go wrong.
If you can name which of the four you are facing, conversion stops being improvisation.
There is one specific instant where more won games are lost than at any other, and it is the move immediately after capturing something important.
What happens there is psychological and very consistent: you take the piece, your body registers that the game is settled, and the level of attention drops sharply. Right then your opponent — who knows they are lost — plays the most awkward thing they can find, because they have nothing to lose. And that awkward move arrives against your worst moment of concentration in the whole game.
The countermeasure is single and mechanical: after winning material, spend twice as long on the next move. Not less, twice. It is counterintuitive and it works, because it corrects exactly the moment where error is most likely.
The same holds, by the same mechanism, after your opponent makes a gross blunder: the danger is not their mistake, it is your relaxation.
It is trainable and almost nobody trains it, because puzzles do not cover it: a puzzle ends where conversion begins.
Play won endgames against the engine. Set up a position a piece up and play it to mate, with the engine on a high level. You will discover that "a piece up" does not win by itself, and that is precisely the lesson.
Review the games you won, not only the ones you lost. It is the habit that fixes this fastest. In almost every game you won there are three or four moves where you handed your opponent a chance they did not take. Those moves are your study material.
Count how many of your losses were won games. It is a depressing and very useful number. For a lot of people between 1200 and 1700 it is above a third, which means that fixing this alone is worth more than any new opening.
And set yourself a clock rule for the ending. Many conversions collapse not from not knowing, but from trying to convert with under three minutes. Saving time for the part where you have to be precise is part of the technique.
Find five games in your history where you had a clear advantage and did not win. For each one, find the move where it turned and write down what you were thinking.
You will find the same thought repeated almost every time. It might be "this is done", it might be "I do not want to complicate things", it might be "I will just defend this". That thought is your pattern, and recognising it in the moment — inside the game — is what makes the next one end differently.
Converting is trainable. And of all the things you can train in chess, it is the one that returns rating fastest, because it does not require learning anything new: it requires not letting go of what you already won.
Is losing this many won games normal? Completely. Between 1200 and 1700, more than a third of losses are positions that were winning at some point. It is not a personal defect: it is the stage where conversion technique does not exist yet.
What is the most useful rule when I am winning? Trade pieces. With extra material, every pair that leaves the board shrinks your opponent's chances and brings closer an endgame where your advantage decides. It is the rule that converts the most games and the hardest to follow, because trading feels passive.
Should I go for mate or take material? Material, unless the mate is calculated to the end. An attack that does not arrive leaves your position dismantled; an extra piece never unwinds.
And if I run out of time exactly when I am winning? It is the commonest case of all, and it gets solved earlier: time is managed in the middlegame, saving for the precise part. Under two minutes, the only policy that works is simplifying as much as possible.
Is it better to offer a draw if I am getting nervous? Almost never. A winning position offered as a draw is half a point given away for a reason that will pass in five minutes, and it also tells your opponent you are uncomfortable. What does help is getting out of the chair, breathing, and looking at the position again from scratch.
How many of my losses should be won games? Ideally none, and in practice always a few. What to watch is the trend: if the number falls from a third to under a tenth over six months, conversion technique is arriving.
If you want to put this into practice with structured material:
Four classes and eight hours of training to master modern evaluation, read the engine, handle complex positions and improve your endgame technique.
ARBy GM Andrés Rodríguez
A six-stage course to improve position evaluation, understand pawn structures and develop practical resources in openings, middlegames and endgames.
ARDFBy GM Andrés Rodríguez and GM Diego Flores
A collection of six courses by GM Andrés Rodríguez to enrich your repertoire, understand complex positions, and improve your game from opening to endgame.
ARBy GM Andrés Rodríguez
The five-step method for getting everything out of a game, why the engine goes last, and what to write down so the analysis becomes a study plan.
Read the articleWhat each time control trains, which one is good for what, and the ratio to play them in if your goal is to climb.
Read the articleWhat actually separates those two bands, how long the crossing takes, and what has to change in your training for it to happen.
Read the articleARDFBy GM Andrés Rodríguez and GM Diego Flores
