A players clustering method to enhance the players' experience in multi-player games

Online multi player games are a kind of video game where players have to interact together in a same game environment through an Internet connection. In these games, the players are grouped in different sessions where they can only interact with the member of the session. In order to maximize the player's experience, game designers have to solve different issues. The first one is to maximize the number of players in the different game sessions. As these game are designed to provide the better game experience when the maximum number of players are in the session. It is important to avoid session with few players. The second one is to create session where the players have the same skill level. A too large difference between the level skills of the players can create frustration. In this short paper we focus on player's skill and present an approach in development to automatically detect communities of players in order to create game session. Our approach is based on game play component, user profile and player interaction with the different game play component. We describe the experiment scheme that has been designed in order to evaluate the impact of the proposition on player satisfaction.