Using Chess to Study Overconfidence
Research with chess players supports key findings about people's overconfidence.
Posted September 24, 2025 | Reviewed by Davia Sills
It is well-known that people tend to be overconfident in their abilities. You may overestimate your chances of getting an interview for a job, answering a question correctly, or succeeding at a New Year’s resolution. There are times when this overconfidence can be a bonus, because it leads you to try harder at tasks, which can make you more likely to succeed. At the same time, you also make decisions about options to pursue or investments to make based on your confidence , so this overconfidence can have career or financial implications.
On top of the general finding of overconfidence, there is also the well-known Dunning-Kruger effect , which finds that the less competent a person is in a domain, the more overconfident they tend to be. There are many explanations for why the worst performers might be the most overconfident, including a lack of awareness of what truly expert performance really entails.
There have been some concerns about the validity of both overconfidence in general and the Dunning-Kruger effect in particular. In many cases, for example, it is hard to tell whether people are truly overconfident, because it is hard to know what scale to use to judge their performance. In addition, in many domains, people don’t get great feedback about how well they’re doing. Finally, some authors have argued that the Dunning-Kruger effect is observed because of the way it is measured rather than because it reflects a psychological tendency.
A fascinating study by Patrick Heck, Daniel Benjamin, Daniel Simons, and Christopher Chabris, published in 2025 in Psychological Science, addresses these concerns about overconfidence and the Dunning-Kruger effect using chess as a domain.
The advantage of using chess as a domain is that tournament chess players are rated based on their performance playing against other players whose rating is known. Analyses of chess play suggest that a player’s rating does a pretty good job of predicting how well they will perform in tournaments. Of course, a player’s rating may go up or down over time based on factors like their time spent studying the game, but at any given moment, a player’s rating is the best estimate of how they will perform. In addition, players who are active in tournaments get a lot of feedback about their performance against players with higher or lower ratings, so they should be getting the feedback they need to judge their performance accurately.
In this study, the data from over 2,500 participants were analyzed. These individuals responded to a survey and had an active rating from either the United States Chess Federation or the International Chess Federation. The participants had an age range from 5 to 88, with an average age of 45.
Participants were asked to state their chess rating as well as whether they thought their actual rating was an accurate reflection of their ability. If they felt that it was not an accurate reflection, they were asked whether their true ability was higher or lower and were asked what they thought their actual rating should be if it were truly accurate. Another important measure in this study was that participants were asked to predict how they would perform across 10 games played against players with ratings higher or lower than their own.
Participants were pretty good at reporting their current rating, so there is no tendency for people to inflate their judgments of their current ranking. However, people generally believed that their current rating was too low compared to their actual ability. On average, participants felt that their rating should be substantially higher than it was. Consistent with the Dunning-Kruger effect, the gap between people’s beliefs about their actual rating and what they believed their rating should be was largest for the people with the lowest ratings and smallest for those with the best ratings. In addition, participants overestimated the likelihood that they would beat other hypothetical players, given the well-known statistics from tournaments about the outcomes of matches between players whose ratings differ.
The researchers also collected the participants’ tournament ratings over the next year. One explanation for the data from this study is that participants were truly improving in their ability, and that their estimates were a good reflection of where they were going to end up six months or a year later. It was the case that the chess ratings for players in this study tended to be higher in the year after they participated in the study. However, the increase in their rating was substantially smaller than their estimate in the study.
This study is interesting because it shows that people remain overconfident in their abilities even when there is a really good scale for measuring their performance that they are familiar with, and even when they get regular feedback about their performance. In addition, there was a robust Dunning-Kruger effect in the data.
Of course, there is generally little harm in being overconfident. Walking into a chess match, job interview, or client engagement feeling confident can make your overall performance better than it would be if you were not confident. Where you really need to be careful about overconfidence is in situations in which you are making investments in which your confidence is unlikely to affect the outcome. When you’re betting on a sporting event or playing the stock market, the result is not going to be changed by your confidence. In those situations, you’re better off trying to find an objective measure of success and sticking with that.
Heck, P. R., Benjamin, D. J., Simons, D. J., & Chabris, C. F. (2025). Overconfidence Persists Despite Years of Accurate, Precise, Public, and Continuous Feedback: Two Studies of Tournament Chess Players. Psychological Science , 36 (9), 732-745. https://doi.org/10.1177/09567976251360747
Share this post Facebook Bluesky Linkedin Email
There was a problem adding your email address. Please try again.
By submitting your information you agree to the Psychology Today Terms & Conditions and Privacy Policy
Art Markman, Ph.D. , is a cognitive scientist at the University of Texas whose research spans a range of topics in the way people think.
Get the help you need from a therapist near you–a FREE service from Psychology Today.
This article is part of the Bringwise Psychology Journal — daily insights on human behavior, mental health, and personal growth.