Predicting Round Result in Counter-Strike: Global Offensive Using Machine Learning
Weison Xu Huang, Jincheng Wang, Yizhi Xu · 2022 7th International Conference on Intelligent Computing and Signal Processing (ICSP) · 2022
This paper first analyzes the research of Xenopoulos et al. and Makarov et al. about using machine learning to predict the result of CS: GO matches. We then carry out our experiment to compare the performance of different machine learning algorithms in predicting. Inspired by the previous research, TrueSkill value was introduced to 5 algorithms: Decision Tree, Gradient Boosted Decision Tree, XGBoost, Logistic Regression, and Neural Network. After comparing each algorithm and implementation of each algorithm with TrueSkill values, the result shows that XGBoost and Neural Network perform better than others, and the difference between these two algorithms is not significant. Our experimental result also reflects that TrueSkill values can slightly improve the accuracy of each algorithm by increasing the size of the data set.