Persuasive attitude changing games are a growing genre and a powerful tool when wanting to reach out and affect the opinion and attitude of larger groups in society. But how well do persuasive games actually work and can they be improved further? This article studies persuasive computer games and comes up with the proposal of using the Multiple Identification Theory, an established model for attitude change as a foundation for designing persuasive games. By analyzing every step of the MIT and translating it into the field of digital games, key concepts have been identified and proposed as guidelines for future development of persuasive games. In addition an analysis of existing persuasive games have been made, with the aim to study how well they meet the criteria of the MIT and determinate whether they could have improved by using it or not.