What is cognitive dissonance?

It's the discomfort you feel when two pieces of knowledge are incompatible, for example "I'm a good player" and "I just lost stupidly." Leon Festinger proposed the theory in 1957. His central idea is that this discomfort pushes you to act in order to reduce it, and that the shortest route rarely runs through questioning your self-image.

Why do people say "I was winning" after a loss?

Because it's often true, and that's what makes the sentence so effective. It reduces dissonance without lying. The problem isn't its accuracy, it's what it lets you avoid examining: why you didn't convert. An unconverted winning position is information about your level, not a mitigating circumstance.

Why is it so hard to abandon an opening you've worked on?

Because of effort justification, an effect demonstrated by Aronson and Mills in 1959: the more getting access to something cost you, the more value you assign it. The hours you sank into a repertoire don't make that repertoire better, but they make dropping it psychologically expensive. You're no longer defending the opening, you're defending the investment.

Is cognitive dissonance theory still valid?

It's solid as a description, shakier as an experimental mechanism. In 2024, a replication run across 39 laboratories in 19 countries with nearly 4,900 participants failed to find the central effect of the historic paradigm. The researchers conclude that this paradigm does not provide robust evidence of dissonance. That people reduce discomfort by reinterpreting facts remains well documented by other routes.

How do you avoid cognitive dissonance after losing a chess game?

You don't, it's automatic. What you can do is delay its resolution. The most effective rule is to separate the moment of the loss from the moment of the analysis: note the position and the facts while it's raw, without any conclusion, and analyse cold twenty-four hours later. Dissonance reduction does most of its work in the minutes right after the game.

Cognitive Dissonance in Chess: The Psychology of "I Was Winning"

Outside the playing hall, club tournament, round 3. Three players have just lost. Listen to them.

"I was winning." "I'd calculated everything, I just played the wrong move order." "Anyway, at this time control, it doesn't mean anything."

None of the three is lying. That's what makes this interesting. The first one probably was winning. The second one probably had calculated. The third sincerely believes what he's saying about fast time controls.

And yet something has just happened in their heads, in under three minutes, that has very little to do with chess.

The Discomfort, and What We Do With It

In 1957, Leon Festinger published A Theory of Cognitive Dissonance. The proposal fits in one sentence: when two pieces of knowledge are incompatible, that incompatibility produces a state of discomfort, and this discomfort has motivational force of its own. It pushes you to act, exactly like hunger or thirst.

The canonical example is the smoker who knows smoking kills. Two cognitions: "I smoke" and "smoking kills." They can't sit together peacefully.

Three exits are available. Change the behaviour, quit smoking. Change the cognition, dispute the evidence. Or add a third cognition that reconciles the first two: "my grandfather smoked and lived to 92."

The first solution is the only one that actually solves the problem. It's also, by a wide margin, the most expensive. The brain almost always takes one of the other two.

Apply that to the game you just lost. Cognition 1: "I'm a decent player, I understand what I'm doing." Cognition 2: "I just lost a winning position against someone at my level." Incompatible.

Changing the first is expensive: it touches who you are. Changing the second is impossible: the result is recorded. That leaves the third route, the one with extra cognitions. "I was winning." "I was tired." "Blitz doesn't count."

Three minutes. Discomfort switched off.

The One-Dollar Experiment

In 1959, Festinger and his student James Carlsmith designed the protocol that would make the theory famous, and it remains one of the most cited experiments in social psychology.

Stanford students spent an hour on a task built to be unbearably boring: turning pegs a quarter-turn, endlessly, then filling and emptying a tray of spools. One hour.

On the way out, some were asked to lie to the next person: to tell them the experiment had been interesting and enjoyable. Some were paid one dollar for that lie. Others were paid twenty dollars, a serious sum in 1959. A control group didn't lie at all.

Finally, everyone was asked what they had really thought of the task.

The result is counterintuitive, which is what made its reputation. It was the students paid one dollar who reported finding the task most enjoyable. Sixty usable participants, twenty per group.

Festinger's reading: the man who got twenty dollars has sufficient external justification. He lied, he knows why, there's no dissonance. The man who got one dollar has no such excuse. He lied for almost nothing. The only way to restore coherence is to conclude that it wasn't quite a lie, and therefore that the task wasn't so terrible.

Weak external justification produces more attitude change than strong justification. Hold on to that sentence, it has direct consequences on the board.

Three Dissonances of a Chess Player

"I Was Winning"

The queen of post-game sentences, and a textbook case, because it's true.

That's exactly what makes it so effective. A lie can be taken apart. A partial truth can't. You were winning on move 28, the engine will confirm it, case closed.

Except the sentence works as a full stop where there should be a question. An unconverted winning position isn't a mitigating circumstance: it's information about your level, and some of the most valuable information you'll get. Converting is part of the game, exactly like obtaining the advantage in the first place. A player who reaches ten winning positions and converts four wasn't unlucky ten times. He has a conversion problem, which is an identifiable, trainable skill.

The dissonance has been reduced. The cost of that reduction is the question nobody asked.

The Opening You Can't Let Go Of

This one runs deeper, and it comes from a different branch of the theory.

Also in 1959, Elliot Aronson and Judson Mills tested a derived idea: if effort produces dissonance, then the more painful it is to get access to something, the more you should value it, in order to justify the effort you put in.

Sixty-three female students volunteered to join a discussion group. Some were admitted with no conditions. Others had to pass a mild screening test. Others a severe and embarrassing one: reading sexually explicit passages aloud in front of the experimenter, in 1959 America.

Then all of them listened to the same deliberately dull group discussion.

Among those who went through the severe initiation, 97% rated the discussion highly interesting. That figure was 21% in the mild condition, and 13% with no initiation at all. The content was rigorously identical.

Now transpose. You've spent four months on a repertoire. You bought two books, watched hours of video, memorized lines out to move 15. And your results with that repertoire are mediocre.

Those four months improved nothing about the objective quality of your opening. But they made dropping it very expensive. So you don't drop it. You conclude you haven't studied it enough, and you invest more. Which mechanically makes dropping it even more expensive next month.

You're no longer defending the opening. You're defending the investment. It's the same mechanism that makes it so hard to walk away from a course you paid a lot for, and it's a close cousin of the sunk cost fallacy. Our article on whether you really need to study openings approaches the question from another angle.

The Move You Defend Instead of Examining

The third one is the subtlest, and it happens during the game.

You play a move. Three moves later a doubt appears: maybe that wasn't the right idea. Two options. Acknowledge the error and redirect the plan, which costs you: you have to admit you judged badly, reassess, start again. Or continue in the chosen direction, which has the advantage of preserving coherence.

The dollar principle applies here almost literally. The more voluntary and unconstrained the move was, the less external justification you have, and the more you need it to have been good. A forced move you drop without pain. A move you chose at length, against an alternative you could see, becomes very hard to disown.

This is why the most expensive mistakes are almost never a single move, but a sequence of moves defending the first one.

Now the Bad News

Something has to be said here that most articles on cognitive dissonance leave out: the experimental foundation of the theory has moved, seriously.

Two tremors.

In 2010, Keith Chen and Jane Risen published an analysis in the Journal of Personality and Social Psychology of the "free-choice paradigm," the field's other major protocol. That paradigm has people rank objects, choose between two objects ranked equally, then rank again. The chosen object reliably goes up and the rejected one goes down, which gets interpreted as dissonance reduction. Chen and Risen showed this result appears even if no preference actually changes, because the choice itself reveals information about the subject's true preferences. The protocol produced the expected result by statistical construction. The debate that followed qualified the exact scope of their demonstration, but the alarm landed.

In 2024, heavier still: David Vaidis, Willem Sleegers and colleagues published a multi-lab replication of the induced-compliance paradigm, the direct heir of the 1959 experiment. 39 laboratories, 19 countries, nearly 4,900 participants, preregistered protocol. Participants given a choice about writing an essay against their own opinion did not change attitude any more than those who had it imposed on them. That is precisely the central effect of the theory, and it did not appear. The authors conclude that this paradigm does not provide robust evidence of cognitive dissonance.

What Survives

These results don't mean the phenomenon described in this article doesn't exist. They mark the failure of particular laboratory protocols, designed sixty years ago, without today's standards of statistical power and preregistration.

What remains well documented, by other routes, is that people reinterpret facts to preserve the coherence of their self-image. Effort justification in particular has survived far more testing than the dollar experiment.

The distinction to draw is this: cognitive dissonance is an excellent description of behaviour you can watch outside any tournament hall. It is a much less established causal mechanism than a century of popularization suggests. Distrusting the explanation doesn't oblige you to deny the observation.

And if you find it uncomfortable that an article on cognitive dissonance undercuts its own theory, watch what you're currently doing with that discomfort.

Four Levers

Delay the resolution, don't try to prevent it. You can't decide not to feel the dissonance. You can decide not to conclude right away. Reduction does most of its work in the minutes after the loss, when the discomfort peaks. Note the facts while it's raw, with no interpretation: the move where it turned, your evaluation at that moment, your remaining time. Analyse twenty-four hours later. The conclusion you draw cold isn't the same one.

Turn "I was winning" into a question. The day "I was winning on move 28" becomes "why didn't I convert on move 28," you've changed what the sentence does. The same information stops being a full stop and becomes a starting point. The full method is in our guide on how to analyze your games.

Set the abandonment criterion before you invest. This is the direct remedy for effort justification, and it has to be set in advance, while the investment is still zero. Before starting a repertoire, write down what would make you drop it: a score below 40% over thirty games, for instance. Written afterwards, the criterion gets negotiated. Written beforehand, it holds.

Treat your moves as hypotheses. A move isn't a promise. The question after a dubious move isn't "how do I justify what I played" but "what's the best continuation from this position, this one, right now." That's also the remedy for anchoring bias, and it's no coincidence: both consist of refusing to start again from the real position.

What to Take Away

Losing produces real mental discomfort, and the brain treats it as a problem to be settled fast. The shortest route never runs through questioning your self-image: it runs through adding an explanation that makes the loss compatible with being a good player.

Those explanations are often accurate. You were winning. You were tired. It was blitz. Their problem isn't truth, it's when they arrive: three minutes after the game, when the discomfort is at its peak and the function of an explanation is to make it stop.

The experimental foundation of the theory is shakier than the retellings suggest, and the 2024 replication deserves to be known. The behaviour, though, you can watch tonight on your way out of the club.

After reading: at your next loss, write down the first explanation that comes to you, and do nothing else. Come back to the game the next day. Compare what you wrote to what you find. The gap between the two is the measure of your dissonance.

This article is part of a series on psychology applied to chess. See also the Dunning-Kruger effect in chess. Upcoming articles will cover rationalization and defense mechanisms, the tools through which this reduction operates.

Sources

Festinger, L. (1957). A Theory of Cognitive Dissonance. Stanford University Press.
Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. Journal of Abnormal and Social Psychology, 58(2), 203-210.
Aronson, E., & Mills, J. (1959). The effect of severity of initiation on liking for a group. Journal of Abnormal and Social Psychology, 59, 177-181.
Chen, M. K., & Risen, J. L. (2010). How choice affects and reflects preferences: Revisiting the free-choice paradigm. Journal of Personality and Social Psychology, 99(4), 573-594.
Vaidis, D. C., Sleegers, W. W. A., et al. (2024). A multilab replication of the induced-compliance paradigm of cognitive dissonance. Advances in Methods and Practices in Psychological Science, 7(1).

Cognitive Dissonance in Chess: Why You Say "I Was Winning"