Recent research published in evolution and human behavior We present evidence that the social benefits of playing games are highly dependent on circumstances such as the player’s skill level and the risks present in the environment. Scientists use computer simulations to suggest that while playing games may not immediately lead to new friendships in a low-risk laboratory setting, games may have evolved as a way for early humans to identify highly skilled allies for dangerous missions.
Playing and gaming are universal human activities, but they differ in important ways. Play involves spontaneous activity without a specific purpose, and is common in human infants and many animals. Games consist of attempting to overcome rules-bound challenges and often involve competition in scenarios similar to war or hunting.
Because games emerged late in development and appear to be unique to humans, scientists wanted to understand the selection pressures that caused their origin. The researchers proposed the competition-for-mates hypothesis, which suggests that games evolved as a strategic tool for individuals to compete and form and maintain new relationships.
“I have been interested in gaming for a long time and realized early on that gaming was more prevalent among boys and men than girls and women,” explains lead researcher Yago Luxevičius de Moraes, assistant professor at the Santo André Foundation and author of Psycho-Andrology Substack. “In 2014 I started researching male psychology, and in 2019 during my master’s I looked at whether gaming contributed to men’s mating and parenting success. The evidence suggested that gaming was more related to friendship formation than mating, which motivated me to pursue this topic in my PhD.”
To test this hypothesis, scientists conducted two complementary studies. The first study included 40 people between the ages of 18 and 40, who were divided into unknown same-sex pairs. The scientists used same-sex pairs because the competition-for-mates hypothesis suggests that men may exhibit stronger alliance-oriented game dynamics due to ancestral differences in physical competition.
Before interacting, participants completed a questionnaire designed to measure two specific concepts. The first concept was the perceived value of a coworker, which involved rating a partner on traits such as intelligence, charisma, and cooperativeness. In an evolutionary context, mate value refers to attributes that historically increase a person’s chances of survival or success.
The second concept is relational proximity, which measures the level of closeness and intimacy that occurs between strangers. After an initial survey, half of the pairs played a game of Nine Men’s Morris. Nine Men’s Morris is an ancient abstract board game in which players take turns placing and moving pieces to steal their opponent’s pieces. The other half of the pairs engaged in non-gaming role-playing activities, acting as two friends with opposing desires who needed to come to an agreement.
After participants completed their assigned activities, they filled out the questionnaire again. The pairs returned to the lab 14 and 28 days later, completed the activities in reverse order, and were surveyed again. The scientists hoped that playing board games would facilitate the bonding process, allowing participants to stabilize their values of each other and bring them closer together more quickly than role-playing activities.
However, the data did not support this prediction. Participants updated their impressions of each other over repeated sessions, but these changes were caused simply by the time they spent together, rather than by the specific activities they performed. Gaming did not induce stronger or faster changes in perceived peer value or relationship intimacy compared to role-playing.
Because laboratory settings often lack important elements like the ancestral human environment, the researchers conducted a second study using an agent-based evolutionary model. It’s a computer simulation that allows scientists to program artificial individuals, or agents, with specific behavioral rules and watch how their interactions unfold over many generations. Each simulated population consisted of 100 agents and featured 99 natural non-gamers and one mutant gamer with game-specific characteristics.
In each generation, agents sought a single cooperative partner to undertake tasks representative of their ancestral activities, such as hunting or war. The chance of surviving the mission depended on the combined skills of the two partners and the environmental hazard level, which represented a fatal risk if the mission failed. Simulations revealed that the game spread through the population only under very specific conditions.
Games were only successful as an evolutionary trait if the original gamers had above-average skill. If the initial gamer’s skill was low, game characteristics will consistently disappear from the population. “One of our unexpected findings was that our simulations showed that the game can evolve even without individuals remembering past partners,” de Moraes told PsyPost. “This suggests that the social benefits of gaming do not necessarily depend on long-term relationships, but may result from short-term interactions.”
The simulations also showed that environmental risks need to co-evolve with population skill levels. This means that as your agent’s skills increase, your survival tasks will need to become more dangerous in order to maintain your advantage in the game. When the environment was static and the risks didn’t increase, the game’s behavior tended to drift and disappear.
“People report feeling more connected after a game, but this effect doesn’t happen automatically,” said De Moraes. “It depends on factors such as the type of game, the skill level of the players, and the context of the interaction. The simulations support the idea that games may have evolved to foster social bonds, but only under certain conditions, such as when skilled individuals are relatively rare and the environment involves a meaningful level of risk.”
The researchers noted that their study had several limitations. “This is an early study and requires replication and extension in both experimental and simulation components,” de Moraes explained. “Experimental design would benefit from larger samples, longer interaction durations, different types of games, and more standardized measurements. Also, simulation results should be tested across a wider range of parameters and alternative assumptions.”
Looking ahead, the researchers aim to improve the computer model. “We plan to extend the theoretical model to incorporate mechanisms such as social learning and genetic associations,” De Moraes said. “Empirically, we intend to address current limitations and test additional hypotheses, including whether games function as nonviolent mechanisms for resolving social conflicts.”
“Games are a complex and widespread human behavior, yet research on games remains fragmented across disciplines,” de Moraes added. “Advances in this field will require further comparative research across different types of games, including sports, video games, board games, card games, tabletop RPGs, and games of chance.”
The study, “Evolutionary and social functions of games: Integrating experimental evidence and mathematical modeling,” was authored by Yago Lukševičius de Moraes, Marco Antonio Correa Varella, Leonardo Cezar Silva Costa, Nayara Teles, and Jaroslava Varella Valentova.
