Dynamic Difficulty Adjustment for Racing Multiplayer Games Using Reinforcement Learning Algorithm

Multiplayer video games are well considered among the players, since players can test their abilities with other players, proof themselves, and enjoy playing with real players. Whenever players with different level of skills play a game with other players, adjusting the difficulty level of the game will be crucial. The excitement of a game would decrease, if the game appears to be easy for some players, while it is very difficult for the others. Since in multiplayer games players interact with each other, dynamic difficulty adjustment is a challenging problem. Generally, difficulty adjustment techniques are proposed for single player games. This paper proposes an automatic system for difficulty adjustment of a multiplayer car racing game using Reinforcement Learning (RL). The results of the user study conducted on this system show the effectiveness of using this module

کلیدواژه ها:

Multiplayer Computer Game ، Dynamic Difficulty Adjustment ، Reinforcement Learning