A Fuzzy Approach to Generating Adaptive Opponents in the Dead End Game

In most of current computer games, computer-controlled opponents are often guided by a limited set of fixed strategies. The behavioral repetition of computer-controlled opponents reduces the human player's enjoyment during gameplay. Furthermore, the predefined fixed difficulty levels of a game are not satisfactory to most human players. Human players prefer challenging games which keep level with them. Hence, in this study we propose a fuzzy approach to adapting opponents' tactics to the behavior of the player such that the player's win/loss rate is kept at their desired rate. The value of the desired rate can be preset by the player according to the player's preference for challenge. Our experiments are conducted on a predator/prey game. The experimental results show the adaptation efficiency and robustness of the proposed method.

並列關鍵字

computer game ； adaptive opponents ； artificial intelligence ； fuzzy inference ； predator/prey game

國際替代計量

全文下載

主題瀏覽