Towards a physiology based difficulty control system for serious games

Serious games can be most efficient if the users are motivated and engaged in the gameplay. Whereas improving core gameplay mechanics can probably yield good results, a dynamic difficulty controlling system might offer additional benefit to a well-designed game. Although performance is the easiest measure to base a control framework on, it is a combination of many different effects. Physiology channels, such as electroencephalography, electrocardiography and pupillometry can reveal the changes in invested mental effort levels and would serve as a more objective base for a difficulty controlling system. To test this assumption, we used Tetris because it is a well-known game and it is simple to manipulate difficulty. In the first half of the experiment, all participants played four pre-set difficulty levels and one customized. In the second part, we tried to elicit certain mental states (boredom, frustration, engagement and fatigue) that are important markers for a controlling system by manipulating difficulty in different patterns. We have found that only pupil diameter data corresponded with expected results. Possible causes and future directions are discussed.