Recommender algorithms in activity motivating games

Physical activity motivating game design encourages players to perform real physical activity in order to gain virtual game rewards. Previous research into activity motivating games showed that they have the potential to motivate players to perform physical activity, while retaining the enjoyment of playing. However, it was discovered that a uniform motivating approach resulted in different levels of activity performed by players of varying gaming skills. In this work we present and evaluate two adaptive recommendation-based techniques, which aim to balance the amount of physical activity performed by players by adapting the level of motivation to their observed gaming skills. Experimental evaluation showed that the adaptive techniques not only increase the amount of activity performed and retain the enjoyment of playing, but also balance the amount of activity performed by players of varying gaming skills and allow for game difficulty to be set in a player-dependent manner.