The paper demonstrates that the Self and Peer Assessment (SPA) method used in higher education, to map a group work mark to individual marks, can easily be modelled as a strategic form game. This modelling predicts students to report SPA following their dominant strategies. However, data of a real time SPA indicate that students may not report SPA by adopting dominant strategies. The findings are indicative of a mismatch between the game (SPA) designed by the tutor, and then played by students. The paper concludes that the interpretation of SPA is not possible, and statistics of SPA should not be relied upon to map a group mark to individual marks.