The “replication crisis” is a pressing issue faced by many fields within social sciences today. This crisis highlights concerns about whether scientific studies, especially in areas like psychology and team research, can be reproduced with consistent results. These concerns arise due to various factors, including small study sizes and the potential for findings not to hold true outside of controlled research environments. This is why, in our paper, we make a compelling case for using video games as an innovative solution to these challenges. 

With over 215 million Americans playing video games weekly and a majority engaging in multiplayer modes that require teamwork, video games present a vast and untapped resource for studying human behavior and teamwork dynamics on a large scale. To demonstrate the potential of video games for social science research, we rely on a huge dataset of 140,000 PUBG matches played by more than 500,000 players.

PUBG, a popular battle royale game, pits teams against each other on an island where the goal is to be the last team standing. The game's setup creates a complex environment for players, since they must work together to gather resources like weapons and armor while deciding how to move and fight as a cohesive unit. Moreover, what makes PUBG an ideal context is how it mirrors real-life teamwork scenarios under pressure. Teams must balance individual skills with the group's overall strategy, all while adapting to a constantly changing situation, making this video game a rich source of data for understanding how teams function. 

In simple terms, playing PUBG is not just about being good at shooting or finding gear; it's about how well players can work together under pressure, make strategic decisions, and adapt to the challenges thrown at them by the game and other players. This setting offers a unique lens to explore teamwork, strategy, and performance, contributing valuable insights to research on team dynamics and behavior.

Within this empirical context, we focus on a specific concept from team management called “behavioral interdependence,” which is essentially how much team members rely on each other to complete tasks, to explore how this interconnectedness impacts team performance. However, this relationship isn’t straightforward; it forms a curve, meaning there’s an optimal point of collaboration that leads to the best team performance.

Interestingly, our results indicate that a player's experience could change this curve, suggesting that more experienced gamers might handle collaboration differently from novices. In other words, we found that when team members coordinate their actions closely, their teams generally do better. However, this benefit changes depending on how much experience the players have, introducing a fresh perspective on how teams can balance their members' interactions for optimal performance, depending on their familiarity and expertise. More in detail, with new teams, a tight approach to working together is linked to worst outcomes, but as teams get more experienced, the relationship between tight teamwork and success straightens out, improving performance. In summary, the study highlights the importance of finding the right balance in teamwork, especially considering the experience level of the team members.

Besides the specificities of the results, we believe that the broader implication of this research is quite significant. By leveraging data from video games, researchers can access a wealth of information that can help address the replication crisis in social sciences. This approach not only allows for studies with larger and more diverse participant pools but also offers a cost-effective method to explore complex team dynamics and human behavior in a controlled yet flexible environment.

Over the years, social science and managerial research have struggled with making sure their studies can be repeated and trusted, partly because it's been hard to study large enough groups of people to get reliable results. E-sports, where millions of players make countless decisions while gaming, provide a goldmine of data for researchers. This data is not just massive but also very detailed, offering insights into human behavior that are hard to get from traditional studies. This richness and detail mean that researchers can study very subtle effects and understand more about how people behave in teams.

Another big plus is that e-sports research can be a lot cheaper. Traditional studies need a lot of people to participate, which can be very expensive. But with e-sports, the data is already there, waiting to be analyzed. This can make research more accessible, especially for scientists who don't have a lot of funding, helping bring in diverse perspectives and ideas into the field. The study of e-sports is also unique because it lets researchers see how teams work together in real-time in a controlled environment, something that's tough to do in the real world. Therefore, this controlled environment helps us understand better how teams succeed or fail, without worrying about potential alternative explanations or confounding factors.

In short, video games offer a treasure trove of data from millions of players engaging in complex team dynamics and decision-making, providing a larger and more varied pool of information than traditional studies. This not only opens the door to investigating subtle behaviors and team interactions in a way that's both affordable and accessible, allowing a broader range of voices to contribute to the management field, but also could help us address the replication crisis in management research by offering a reliable, scalable, and insightful research context.