Dynamic Difficulty Adjustment in Digital Games Using Genetic Algorithms

Abstract

The difficulty of a game is intrinsically connected with the experience of immersion in it and with its success. One of the main reasons for a player to drop a game is that the game is either too easy or too hard for him/her. In practice, players become either bored or frustrated if playing a game that is not balanced for them. An approach to prevent this kind of behavior is to dynamically adjust the difficulty of a game so that the game adapts to the player's experience by evaluating the difficulty of a game and changing its environment to become easier or harder for the player. In this paper, we propose a real-time solution using a Genetic Algorithm which helps to provide the exact amount of challenge that a player needs to not be bored or frustrated thus balancing the difficulty of a game. We review several other papers that approached this problem, which characteristics an algorithm has to have to approach the problem, and how to balance this in a generic way. The main idea of this paper is to create an approach that can be modified and coupled to any kind of game by using a Genetic Algorithm.