ӣ��Ӱ��

Topic

Representative Features for Game Agnostic Movement Evaluation (ReFGAME): Extending a Trajectory Analysis Framework from Human Mobility to Video Games

Department of Computer Science

Date & location

• Tuesday, November 25, 2025

• 8:30 A.M.

• Engineering Computer Science Building

• Room 468

Reviewers

Supervisory Committee

• Dr. Kevin Stanley, Department of Computer Science, ӣ��Ӱ�� (Supervisor)

• Dr. Regan Mandryk, Department of Computer Science, UVic (Member)

• Dr. Brandon Haworth, Department of Computer Science, UVic (Member) 

External Examiner

• Dr. Sam Liu, School of Exercise Science, Physical and Health Education, ӣ��Ӱ�� 

Chair of Oral Examination

• Dr. Sandra Gibbons, School of Exercise Science, Physical and Health Education, UVic 

Abstract

Understanding how players interact with virtual environments underpins the design of video games and the contextualization of player behaviour. This work extends a framework from human mobility research to the analysis of player movement and camera orientation trajectories in video games. Using features that capture spatiotemporal properties of trajectories, this work demonstrates the framework’s applicability in a gaming context through supervised and unsupervised machine learning tasks conducted on data from professional Counter-Strike: Global Offensive matches. The framework can be used to distinguish between the gameplay environment and side of play of teams with 93% accuracy when using features derived from player movement and 97% accuracy with the addition of camera orientation derived features. The framework reveals design archetypes between environments and potentially between the roles of players. Findings validate the generalizability of spatial mobility features to a gaming context, and highlight their potential for applications in role identification, environment design, anomaly detection, and cross-game analysis.