ABSTRACT
Multiplayer online battle arena (MOBA) games create immersive virtual gaming worlds in which two teams of players compete with each other to win a battle. Players controlling different game characters in a team coordinate to take down the opponent’s base. Game analysis plays a crucial role in identifying and reasoning about gameplay patterns for the purpose of aiding game designers in understanding gaming behaviors and optimizing the game design. Game pattern mining has been partially studied by research communities. M. Veron, O. Marin, and S. Monnet built a database of League of Legends matches to analyze the matchmaking service of the game. A. Drachen et al. modeled spatio-temporal gaming behaviors by using three data-driven measures: zone changes, distribution of team members, and fuzzy clustering of time series. MOBA games involve two opposing teams of a maximum of five players competing against each other with the objective of destroying the base of the opponent’s team to score a win.
