This study explores an experimental design methodology to circumvent issues with gathering behavioural data through serious games in citizen science. Public transportation research is used as a case, as it offers a complex problem for the game to solve. Four players were interviewed and observed playing a prototype of a serious game intended on gathering data on public transportation use in Gothenburg. Alongside this, a definition of different types of behaviours in games was developed to examine whether the players were playing rationally with regards to the game context (the magic circle) and the real world. The results show promise for the simplified real world mechanics utilized, to enable engaging serious game experiences. However, many issues and challenges were also observed, revolving around the players tendency to fluctuate between different types of behaviours. The implications of these challenges and the usefulness of the experimental design are discussed at the end.