Your brain has a measurable fingerprint for how well you read other people

How quickly can you figure out whether the person across from you is bluffing, calculating, or just guessing? Some people size up an opponent in a few interactions. Others take much longer to read the same signals. A new study published in Nature Neuroscience shows that this variation is not just behavioral. It has a distinct neural signature that can be measured, mapped, and used to predict how well a person adapts to others in real time.

Rock-paper-scissors as a window into social cognition

The research team, led by Christian Ruff, a professor of neuroeconomics and decision neuroscience at the University of Zurich, recruited more than 570 participants to play repeated rounds of rock-paper-scissors against both human and artificial opponents. The game may sound trivial, but it creates exactly the conditions researchers need: a dynamic interaction where players must continuously infer what their opponent is thinking and adjust their strategy accordingly.

Using a novel computational model, the team quantified two things for each participant: how strategically they assessed their opponent, and how quickly they updated that assessment when the opponent changed behavior. The range was wide. Some participants recognized shifts in strategy within a few rounds. Others took far longer to recalibrate.

Four brain regions that fire during reassessment

A subset of participants underwent functional magnetic resonance imaging (fMRI) while playing. The scans revealed a distributed network of brain regions that showed increased activity specifically at moments when participants revised their mental model of their opponent.

The temporoparietal cortex, long known to be involved in thinking about other people's thoughts and intentions, was particularly active during these moments. So was the dorsomedial prefrontal cortex, a region associated with evaluating social information. The anterior insula and adjacent areas of the ventrolateral prefrontal cortex spiked especially when expectations proved wrong and a reassessment became necessary.

The researchers describe this as a neural fingerprint of adaptive mentalization, the brain's ongoing process of modeling and updating its understanding of another person's mind.

Predicting behavior from brain activity

The critical test was whether these neural activity patterns could predict how much a person would adjust their strategy. They could. The prediction model worked for participants whose brain data had not been included in training the model, a standard cross-validation test that guards against overfitting.

The prediction was successful with nearly 90% of participants. That level of accuracy is high for fMRI-based prediction of social behavior, a domain where noise, individual variability, and task complexity typically make prediction difficult.

Beyond static snapshots of social cognition

Previous research on social cognition has relied heavily on static tasks: reading short stories about characters' intentions, making one-shot decisions in trust games, or viewing facial expressions. These approaches capture snapshots of social reasoning but miss the dynamic, ongoing nature of real social interaction, where people are constantly updating their understanding of each other.

The rock-paper-scissors paradigm, combined with the computational modeling approach, captures something closer to what happens in an actual conversation or negotiation. Social cognition is not a trait that either works or does not. It is a process of continuous adaptation, and individuals differ substantially in how quickly and effectively they carry it out.

Implications for clinical assessment

Ruff sees potential applications for neurological and psychiatric conditions that impair social interaction, including autism spectrum disorder and borderline personality disorder. Current clinical assessments of social cognition rely heavily on behavioral observation and self-report, both of which have significant limitations.

A neural marker of adaptive mentalization could provide a more objective measurement, one that tracks an ongoing brain process rather than relying on a patient's description of their social difficulties. In the longer term, such markers could help evaluate whether therapeutic interventions are actually changing the neural processes that underlie social cognition, rather than just changing self-reported symptoms.

That clinical translation is still distant. The study used implanted-level fMRI in a controlled laboratory setting, conditions far removed from a clinic. But the proof of concept is strong: the brain's social adaptation system produces measurable, predictive signals. The challenge now is figuring out how to use them.