How Riot Games' behavioral testing of its 27M daily League of Legends players is helping it curb toxic behavior
so people play nice. Kosher? https://medium.com/... Kevin Werbach / @kwerb : “Big data tools allow companies to study and manipulate online...behavior with unprecedented scale & speed.” https://medium.com/... Dan Gillmor / @dangillmor : Fascinating piece by @jeremyhsu about world's “Largest Virtual Psychology Lab” — an online game. https://t.co/nPBnv6uZkw on @backchnnl
Context & Ripple Effects
Backchannel's piece frames Riot Games as running what Dan Gillmor calls the world's 'largest virtual psychology lab' — 27 million daily League of Legends players serving as subjects for behavioral experiments aimed at curbing toxicity. Kevin Werbach's quoted warning cuts both ways: big-data tools let companies study and manipulate online behavior with unprecedented scale and speed, so the same instrumentation that polices conduct is a manipulation surface.
What followed in the coverage validates both readings. By 2017, vendors like Silicon Studio were selling AI that profiles player actions to influence not just behavior but spending, showing how quickly lab-style tooling commercialized. A decade later Riot itself is still deep in the game — its head of anti-cheat now discusses discrediting cheat developers outright.
First-order effects
- Riot's 27M daily players are live subjects in ongoing behavioral tests, with the company tuning interventions at population scale faster than any traditional psychology study could.
- The toxic-player problem gets an operational fix inside League of Legends, giving Riot a moderation playbook most studios lacked in 2015.
Second-order effects
- The approach becomes productized: by 2017 Silicon Studio's Yokozuna was marketing AI that profiles player actions to influence behavior AND spending, collapsing the line between civility tooling and monetization levers.
- Rivals face rising expectations for data-driven moderation, a bar later raised further by AI voice-chat tools like Modulate's ToxMod that gaming companies adopted amid accuracy and privacy concerns.
Third-order effects
- If the pattern holds, game platforms consolidate into de facto behavioral-science infrastructure — which invites exactly the scrutiny seen when US lawmakers pressed top game companies on how they assess extremism risks, citing ADL findings.
- Werbach's manipulation-at-scale critique points toward a standing governance question: player populations are experimented on without the consent frameworks that govern academic research, and no regulator yet owns that gap.
The trend: Game studios are evolving into large-scale behavioral laboratories whose data tooling spans moderation, monetization, and now regulatory attention.