Multivariate and Categorical Analysis of Gaming Statistics

This paper provides exploratory analysis on gaming statistics via various multivariate and categorical data analysis approaches. The clustering results show that the principal components associated with the gaming data are related to player expertise and game camp characteristics. More importantly, the player level possesses only limited discriminant power for reflecting player expertise, and hence it indicates that gaming expertise classification is beyond player level. This paper therefore sheds new light on gaming statistic analysis, player expertise evaluation, and player level design.