In this work, we investigate traveller's behaviour based on data collected through a computer-based travel simulator experiment (mobility decision game). The main aspects that we focus on are risk profiles, perception, learning from experience and habit, the impact of advanced information systems (ATIS), and travellers' behavior under competitive games. The understanding of these behavioural aspects will allow us to test the adequacy of different choice models in predicting the travellers' decisions, which is essential for dynamic traffic microscopic simulation platforms.