USC Study Finds Game Incentives Cannot Rewire Player Behavior Across Titles
A study tracking 4,830 players across League of Legends and Teamfight Tactics finds individual preferences beat out game design when shaping behavior.
ProTakeaways
- USC and UC Santa Cruz tracked 4,830 players across League of Legends and Teamfight Tactics to test agency vs structure.
- League rewards specialization while TFT rewards flexibility, yet individual players stay consistent across both games.
- Cross-game flexibility predicted a player's behavior better than their competitive win rate inside each game.
- Elite Master to Challenger players adapted partially to game structure but retained baseline dispositional preferences.
- Kernel regression with a Laplacian kernel hit R² of 0.2235 for League and 0.4551 for TFT flexibility.
- Findings suggest cross-platform behavioral signals may outperform within-platform metrics for user modeling and personalization.
Game designers spend enormous energy tuning incentives to steer player behavior. A new study from USC and UC Santa Cruz suggests those levers may be weaker than the industry assumes. When the same people are tracked across two games that reward opposite strategies, they largely keep doing what they already prefer.
The researchers followed 4,830 people who logged at least 50 competitive matches in both League of Legends and Teamfight Tactics, then measured how flexible each person was in their choice of champions or board compositions. Although League rewards specialization and TFT rewards flexibility, individuals exhibited consistent behavior across both.
Two games, opposite rules, same account
Most behavioral research on games struggles with self-selection bias. If you compare Fortnite players to Chess.com players, you cannot tell whether the games shape behavior or whether different types of people simply pick different games. The paper gets around this by exploiting the fact that both titles run on the same Riot Games client and account system, letting the authors track the same PUUID across both environments.
The two games work almost as a purpose-built A/B test for behavioral incentives:
- League requires players to commit to their champion before gameplay begins, with no opportunity to swap mid-match, rewarding deep mastery of a narrow pool.