University of Twente Student Theses
As of Friday, 8 August 2025, the current Student Theses repository is no longer available for thesis uploads. A new Student Theses repository will be available starting Friday, 15 August 2025.
Predicting Counter-Strike tactics using Graph Neural Network based Machine Learning
Csapo, Szabolcs (2025) Predicting Counter-Strike tactics using Graph Neural Network based Machine Learning.
|
PDF
699kB |
| Abstract: | Counter-Strike has been one of the largest esports scenes for over a decade, with gameplay increasingly defined by complex team tactics. As tactics have evolved, distinguishing and analyzing them has become more challenging. This research automates tactic recognition in each round using a Graph Neural Network (GNN). The developed GNN model achieved an accuracy of 81.17% and an F1-score of 0.6945, demonstrating strong performance in identifying both common and ambiguous tactics. Results show that a simple, two-layered Graph Convolutional Network, combined with all available node features (health, armor, player position, and utility), provides the most effective way of predicting Counter-Strike tactics. The model’s ability to generate reliable predictions from partial in-game data offers significant potential for improving coaching strategies and automating data labeling. These findings advance the state of automated tactic recognition in esports and contribute to the broader development of the esports scene. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Computer Science BSc (56964) |
| Link to this item: | https://purl.utwente.nl/essays/107599 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page