Exploring Aggressors’ In‑Match Cognitive and Emotional Formation and Toxic Behavior Trajectories in MOBA Games

Paper Title

Exploring Aggressors’ In‑Match Cognitive and Emotional Formation and Toxic Behavior Trajectories in MOBA Games

Publication Info

• Topic area: Toxic behavior dynamics in multiplayer online games.
• Keywords: MOBA games, toxic behavior, cognition, emotion, intervention strategies, League of Legends, multiplayer gaming, aggression, player experience, game design.

Background and Problem

• Problem / challenge: Existing research on toxic behavior in online games often lacks focus on aggressors’ cognitive and emotional states during toxicity emergence and its evolution throughout matches. Current interventions primarily target victims and post-hoc actions, leaving gaps in understanding and preventing toxicity at its onset.
• Significance: Toxic behavior undermines player well-being, team cohesion, and game community sustainability, leading to player attrition and reduced engagement.
• Motivation and related work: Prior studies have examined toxicity through aggregated data, surveys, and victim perspectives but rarely capture real-time aggressor narratives or dynamic in-game contexts. The paper builds on the General Aggression Model (GAM) and extends prior frameworks to explore aggressors’ internal states and behavioral trajectories.

Solution

• Proposed approach: A qualitative study using retrospective think-aloud protocols and emotional journey maps to analyze aggressors’ cognitive and emotional states during toxic behavior emergence and evolution in League of Legends matches.
• Novelty:
  1. Development of a multidimensional criterion for evaluating toxicity severity.
  2. Identification of a hierarchical three-layer cognitive structure (perception, orientation, intention) influencing toxic behavior.
  3. Analysis of recurring patterns in toxic behavior trajectories (persistent vs. single-instance).
  4. Translation of findings into preventive intervention strategies for MOBA game design.
• Procedure and key techniques:
  • Recruitment of 18 League of Legends players who initiated toxic behaviors.
  • Collection of gameplay recordings and annotations of toxic events.
  • Retrospective think-aloud interviews to reconstruct players’ thoughts, emotions, and motivations.
  • Emotional journey visualization to map valence and arousal dynamics across matches.
  • Thematic analysis of transcripts and journey maps to identify cognitive and emotional patterns.

Results

• Concrete findings:
  • Aggressors evaluate toxicity severity using five dimensions: aggressive features, negative emotional transmission, timing and frequency, scope and game consequences, and rationality.
  • Toxic behavior is driven by a hierarchical cognitive structure: perception (situation, event, player), orientation (offensive, defensive, cooperative, indifferent), and intention (offensive, defensive, cooperative, indifferent, emotional ventilation).
  • Emotional valence and arousal strongly influence toxicity levels, with higher arousal linked to more severe toxicity.
  • Two toxic behavior trajectories were identified: persistent (sustained within a context) and single-instance (isolated, subsequent, or repetitive).
• Advantage over baselines: Provides a nuanced, process-oriented understanding of toxicity evolution, moving beyond static definitions and victim-focused interventions.
• Experiments / evaluation:
  • 39.5 hours of interview data analyzed using inductive thematic analysis.
  • Emotional journey maps visualized valence (-4 to 4) and arousal (1 to 9) dynamics.
  • Participants represented diverse ranks, roles, and experiences in League of Legends.
• Limitations and future work:
  • Small sample size (18 players) from Mainland China limits generalizability.
  • Focus on aggressors’ perspectives excludes victim narratives.
  • Retrospective self-reports may introduce memory bias.
  • Future research should include cross-cultural samples, real-time data capture, physiological measures, and prototype testing of intervention designs.

Summary

This study investigates how toxic behaviors emerge and evolve in League of Legends matches by analyzing aggressors’ cognition and emotion. It identifies a hierarchical cognitive structure and recurring behavioral patterns, providing insights into the dynamic interplay of perception, orientation, and intention. Findings highlight the need for preventive interventions targeting early-stage toxicity, such as team identity reinforcement, contextual detection, and emotional self-regulation tools. While limited to a small sample, the research offers a foundation for designing healthier gaming ecosystems in MOBA games.