University of Vigo

Explainable E-sports Win Prediction through Machine Learning Classification

The integration of real-time machine learning into competitive gaming addresses the critical need for predictive accuracy and transparency in high-stakes environments.

• A new stream-based machine learning framework for Counter-Strike: Global Offensive achieves over 90% predictive accuracy for match outcomes.
• The system utilizes a sliding window data management approach to integrate pre-game statistics with live in-game performance metrics in real-time.
• Adaptive Random Forest classifiers are used to prioritize computational efficiency while enabling the extraction of clear, interpretable decision paths.