Dynamic difficulty adjustment in games by using an interactive self-organizing architecture

If difficulty level of a game does not match player´s skills, the game could be frustrating or disappointing. In this paper we propose a self-organizing system (SOS) to adjust difficulty level of games. For this purpose, we use Artificial Neural Network and Interactive Evolutionary Algorithms to evolve Non-Player Characters (NPCs), and focus on player´s hidden responses to determine fitness of the system. Results show that the proposed interactive SOS can adapt itself with different level of skills.