Nutrition, not calories: Findings from the largest open-source dataset of video game behaviour and health

Governments, clinicians, game developers, and parents demand better answers about what constitutes healthy vs unhealthy gaming, but answering that question requires detailed behavioural data largely held in walled industry gardens. In a study of 3,000 US and UK early adults (aged 18–30), we collected 3+ million hours of detailed video game play across five major platforms, using both open-source methods and industry collaboration. Drawing from 50k+ linked survey responses on wellbeing, stress, social play, cognitive performance, and more, I will share how games are used to compensate for a bad day (adaptive in moderation, risky when habitual), how late-night play relates to sleep, and how play habits changes as we age. I will also talk candidly about the process of collaborating with the games industry: trade-offs we made, governance practices future researchers can adopt, and the tension between partnering with platforms and building open tools to work around them. Data and codebooks are openly available.