Topic modeling identifies themes of toxicity in player reports from competitive online games, suggesting areas for intervention.
Toxic behavior is a pervasive issue in multiplayer online games, affecting both player and community wellbeing. Prior research has focused largely on communicative abuse, overlooking the situated nature of toxicity in gameplay. The current study addresses this gap by analyzing 1137 publicly available Reddit posts from four team based competitive games: League of Legends, Marvel Rivals, Counter Strike: Global Offensive, and Rocket League. Using Latent Dirichlet Allocation and manual review, nine themes were identified, including smurfing, AFK behavior, role conflicts, targeted harassment, misogyny, and ineffective moderation. Overall, findings show that competitive and social structures interact with game design to shape toxic behaviors in online games. We also demonstrate that topic modeling can extract meaningful themes from player-generated text, while highlighting challenges such as game-specific terminology, and overlapping topics. These results emphasize the need to consider game design, player behavior, and community norms when developing interventions to reduce toxicity.
No takes yet. Share an insight, caveat, or question.
Hu et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: