Success is usually seen as the best possible outcome for any team. Whether it is a football match, a business project or a classroom assignment, achieving the goal is generally taken as a sign that everything went well.
But new research suggests that success may sometimes have an unexpected downside.
Researchers have found that when a team succeeds, individuals may learn less from their own mistakes than when the team fails—even if they made exactly the same error in both situations. The findings suggest that positive team outcomes can reduce the adjustments people make after making a mistake, potentially slowing individual improvement.
The research, published in the Journal of Neurophysiology (citation below), was conducted by Naoyoshi Matsuda, Haruki Kayukawa, and Masaki O. Abe. Matsuda is affiliated with Japan’s National Institute of Information and Communications Technology, while Kayukawa and Abe are with Hokkaido University. The researchers investigated how the overall outcome of a team task influences the way individuals respond to their own errors.
A simple question with important implications
In many team activities, one person’s mistake does not necessarily determine the final result.
In football (soccer), for example, a midfielder may make a poor pass, only for a teammate to recover the ball and score. Similarly, in a workplace project, one employee’s mistake may be corrected by colleagues before the project is completed successfully.
The researchers wanted to know whether people continue learning from their own mistakes when the team ultimately succeeds.
As they explain in the paper, teammates can often compensate for individual errors, making the quality of one person’s action only loosely connected to the team’s final outcome. They therefore asked whether individuals can recognise the consequences of their own actions and learn from them independently of the team’s success or failure.
Recreating a football pass and shot
To investigate the question, the researchers recruited 36 healthy volunteers aged between 18 and 25 years. Participants worked in pairs, with one acting as the “passer” and the other as the “shooter.”
The pair completed a computer-based task designed to resemble a final pass and shot in football.
Using handheld game controllers, the passer first directed a virtual ball toward a target. The shooter then attempted to send the ball into a goal target. The researchers measured how the passer changed their movement after making an error.
The key to the experiment was a special type of trial known as an error-clamp trial.
During these trials, the researchers secretly controlled the ball’s path. Regardless of how either participant moved the controller, the pass always contained the same fixed error. The shot then either succeeded or failed, creating either a successful or unsuccessful team outcome. Because the pass error remained identical in every case, the researchers could determine whether the team’s result alone influenced subsequent learning.
Team success reduced learning
The results showed that participants consistently adjusted their movements after making an error, demonstrating that they had learned from it.
However, the size of that adjustment depended on the team’s overall outcome.
When the shot was successful, participants made smaller corrections on the following attempt than when the shot failed, even though the pass error they experienced was exactly the same. Statistical analysis confirmed that this difference was significant.
The researchers also found that participants’ learning was influenced not only by whether the team succeeded or failed but also by the direction of the final shot error relative to the original passing error.
People recognised their mistakes but learned from them differently
The researchers also wanted to find out whether team success changed how people perceived their own mistakes.
To test this, participants were asked to indicate where they believed the virtual ball had crossed a reference line after each attempt. This allowed the researchers to compare participants’ perception of the error with the actual error.
The results revealed an interesting contrast.
Although participants generally underestimated the size of errors, their perception of the pass error did not differ significantly between successful and unsuccessful team outcomes. In other words, people recognised their mistake to a similar extent regardless of whether the team ultimately succeeded or failed.
What changed was not how they saw the error, but how strongly they adapted their behaviour afterwards.
The authors conclude that the team’s overall success or failure influenced the learning response itself rather than the participants’ reported perception of the error.
Why might this happen?
The researchers discuss several possible explanations but emphasise that their experiment was not designed to identify the precise mechanism.
One possibility is that a successful team outcome acts as a form of reward, reducing the learning response that would normally follow an error. They note that previous research has suggested reward signals can influence error-based learning, but stress that their own results do not prove this explanation. Instead, they present it as a hypothesis that future research should investigate further.
What could the findings mean?
Although the experiment used a simplified laboratory task based on a football pass and shot, the findings may have broader implications wherever people work together.
The study suggests that a successful team outcome can sometimes mask opportunities for individual improvement. If the overall result is positive, people may adjust their behaviour less after making a mistake, even though they recognised the error.
The authors note that extending this line of research to other forms of motor learning could improve understanding of teamwork and may eventually contribute to practical applications, including sports coaching.
Citation
Matsuda, N., Kayukawa, H., & Abe, M. O. (2026). Influence of team task outcomes on individuals’ responses to their own errors. Journal of Neurophysiology, 135, 1611–1620. https://doi.org/10.1152/jn.00109.2026
