Predicting Player of the Game in Multiplayer Online Battle Arena Games

The multiplayer online battle arena (MOBA) is currently one of the most popular game genres. Among these, League of Legends (LoL) stands out as a particularly noteworthy MOBA game, with its professional league matches attracting high viewership and immense popularity. In this paper, we propose a method to predict the player of the game (PoG), the player contributing the most to victory in MOBA games. To implement the prediction model, we construct a dataset by merging professional league match data with PoG selection results. We select and analyze key factors with consideration for potential applicability to other MOBA games. The experimental results demonstrate that the incorporation of relative metrics enhances the model's performance, allowing for a more accurate consideration of the game's impact. The improved performance of the model is deemed sufficient for practical application in real-world services.