Machine Learning Algorithms for Player Satisfaction Optimization

Abstract

There are several state-of-the-art algorithms currently used for optimization of various aspects of games affecting player satisfaction. In this paper we give a survey of these methods in order to present the platform of research for modeling player satisfaction for a generic player. We focus on the systems for optimization of overall player experience possible applicable on more genres of games. The algorithms are used for optimization of Non-Player Characters (NPC) behavior, Content Generation, Dynamic Difficulty Adjustment (DDA) etc.