Detecting challenge from physiological signals: A primary study with a typical game scenario

BSTRACT Challenge is the core element of digital games. Game challenge with an appropriate type and level that matches with players’ skill, experience and motivation would lead players to achieve the optimal player experience. With a wider spectrum of challenge types such as physical, cognitive, and emotional challenge provided by modern digital games, a questionnaire tool of CORGIS has recently been developed to evaluate the whole range of challenge experiences subjectively. However, such challenge experiences still lack measures to evaluate them objectively ”in real time”. To explore the possibility to detect different challenge types based on physiological signals, we conducted an experiment where 12 players’ physiological signals (EDA, ECG, EMG, RSP and TEM) of overcoming different types of game challenges were recorded. With 80 extracted physiological features, two methods (ANOVA-based and Regression-based) were adopted to select challenge-related physiological features. Results of logistic regression models showed that both methods obtained detection accuracy over 60%, which suggest potential for further development of a real-time challenge measurement instrument.