Modelling Physical Activity in Virtual Reality Games

This thesis was inspired by the possibility that virtual reality (VR) games, which are designed primarily to be fun, could also provide exercise. It aimed to gain insights about this by exploring whether people can gain beneficial levels of exercise while playing VR games and how they might use VR games for exercise over several weeks. Furthermore, this work also focuses on how the level of physical activity that can be captured during gameplay and how a long-term user model can be created for individual players, as a foundation for supporting the user in gaining personal informatics insights about their exertion as well as being used for personalisation and external recommendation for VR games. The key contributions of this research are: • The first study of a diverse set of commercial VR games to gain insights about the level of actual and perceived exertion players have. • The first long-term study of VR games in a sedentary workplace to gain insights about the ways people utilise it and the levels of exertion they gain. • Based on reflections on the above studies, this thesis presents a framework and guidelines for designing physical activity VR games. • The systematic creation of a user model for representing a person’s long-term fitness and their VR gameplay, exertion and preferences. • A study of the ways that people can scrutinise their long-term personal informatics user model of exertion from VR game play and incidental walking. These contributions provide a foundation for future researchers and industry practitioners to design VR games that provide beneficial levels of exertion and allow people to gain insights into the relative contribution of the exercise from gameplay