Evaluation Algorithm in Game Development

OData support
Supervisor:
Rajacsics Tamás
Department of Automation and Applied Informatics

The reason behind conducting this paper is to show that implementing an evaluation algorithm in game development can produce for us; a smarter version of the AI player that can be competitive enough or even in some cases unbeatable to play against. In exploring different algorithms, we can decide which one is fit for our game based on the type and the set of rules that govern our game from irregular behaviors. For my thesis work, I developed a two-player based board game and implemented an algorithm called Minimax; to improve the decision-making of my AI player where it can see in the future which the best possible next move, thus reducing the chance of losing to the human player. The end result of my game produced a balanced version of the AI player; which is smarter enough to beat the human player and has a better performance to stay away from out of memory problem caused by the limitations of the web browser. All in all, I would say that there is no best algorithm to improve the AI player it all depends on the requirements of your game and to what extent you want to improve, but having it as a part in your game development can result in a more non boring joyful experience for the human player to try.

Downloads

Please sign in to download the files of this thesis.