Evolutionary and social functions of gaming
– by Yago Moraes
Imagine two strangers meeting for the first time. Instead of making small talk, they either play an ancient strategy game or role-play a mild disagreement between friends. Weeks later, their impressions of one another have changed, but not in the way many theories about games would predict.
This puzzle lies at the heart of our paper, “Evolutionary and social functions of gaming”, which combines a lab experiment with computer simulations to ask: what are games for in human evolution?
Humans create games in nearly every culture and keep playing them long after childhood. This is striking because, although play is common across many mammals, gaming is different. Games have rules, goals, and winners-losers. They are also especially common among post-pubertal males and often reflect hunting, combat, and competition.
These features have inspired several explanations. Evolutionary researchers argue that games function as low-cost practice for hunting or warfare, and as arenas for status competition, helping players display skill, gain prestige, and attract mates. Social scientists, meanwhile, often emphasize a different role: games as social glue, capable of easing tension, breaking the ice, and fostering group cohesion.
My colleagues and I propose a framework that integrates these perspectives by extending Deaner and colleagues’ Allying with Coalitions Hypothesis. We suggest that games may have evolved not just as practice or entertainment, but as tools for finding valuable allies. In ancestral settings, success in activities such as big-game hunting or warfare depended on choosing reliable partners. A poor ally could be costly; a skilled and trustworthy one could be lifesaving. Games, with clear rules and measurable results, may have offered a low-risk way to judge others: Who is skilled? Who is strategic? Who cooperates well under pressure?
This idea leads to a clear prediction: if games help people judge potential allies, then playing together should speed up judgments of one another’s value and strengthen social bonds more efficiently than other activities.
To test this, we ran an experiment with same-sex pairs of strangers assigned to one of two conditions. In the gaming condition, participants played Nine Men’s Morris, a centuries-old strategy board game. In the control condition, they completed a role-play task in which two friends negotiated different preferences.
Across three sessions spanning roughly one month, participants repeatedly rated one another on two dimensions: perceived peer value (how desirable the other seemed as a potential friend or ally, including traits such as intelligence, charisma, attractiveness, and status) and relational proximity (how close and connected they felt to the other person). If gaming functions as an ally-selection mechanism, then the gaming condition should have produced faster or stronger shifts in these ratings. It did not.
Both activities changed how participants saw one another over time. Repeated interaction mattered, as expected. But gaming did not outperform the control condition. Playing a strategy game did not make strangers feel closer or judge each other more positively any faster than a structured, mildly conflictual conversation. This pattern appeared in both male and female pairs.
At first glance, this seems like a challenge to the hypothesis. But null results can still be informative. Rather than disproving the theory, the findings suggest that gaming’s social effects depend strongly on context. Not all games are equally informative. A low-stakes, one-on-one abstract board game played in a lab may be a poor model for the kinds of games that mattered in human evolution.
To explore this, we turned to simulations. In our model, virtual individuals could adopt either a gaming or non-gaming strategy. They formed alliances, completed risky cooperative tasks, survived or died based on performance, and reproduced across generations. This allowed us to test when gaming might become adaptive.
The simulations showed that gaming spread only under specific conditions. First, the initial gamer had to be above average in skill. Second, environmental risks had to remain meaningful by changing alongside population skill; otherwise, risks became too easy to overcome.
When low-skill individuals adopted gaming, the strategy usually disappeared. This is what we would expect if gaming functions as an honest signal: it benefits only those who can truly display valuable traits. The model also showed that gaming loses value when stakes are low. In safe settings where performance differences matter little, there is less reason for gaming-based assessment to evolve or persist.
This offers a plausible explanation for the experimental null results. Laboratory settings are not ancestral battlefields. No one’s survival depended on winning NMM or choosing the right lab partner.
Together, the experiment and simulations suggest a more nuanced view of gaming. Games do not reliably speed up friendship formation in low-risk, short-term interactions with strangers, at least not more than other structured social activities. But in conditions resembling ancestral life, where success in risky cooperative tasks affects survival, gaming may work as a stable strategy for judging and competing for valuable allies. This view helps explain the widespread presence of games, their male-biased participation in many cultures, and the modest social effects often observed in modern lab studies.
It also suggests we may be focusing on the wrong social phase. Rather than helping primarily to form friendships, games may be especially useful for maintaining them, for instance, by providing a controlled, nonviolent outlet for rivalry or conflict resolution. We refer to this speculative possibility as the Conflict Gambling Hypothesis.
Finally, not all games are created equal. NMM is an abstract, one-on-one board game. Many of the strongest sex differences and coalitional dynamics emerge instead in team sports and real-time multiplayer games such as FPSs or MOBAs, which more closely resemble the coordination, competition, and group dynamics of hunting and warfare.
By combining experiments with evolutionary modeling, this research offers a broader way to understand when games matter socially. It suggests that the evolutionary role of gaming is closely tied to risk, honest signaling, and the long process of building and maintaining alliances when stakes are high.
