Gamification has been explored recently as a way to promote content delivery in education, yielding promising results. However, little is known regarding how it helps different students experience learning and acquire knowledge. In this paper we study and analyze data from a gamified engineering course, to search for distinct behavior patterns. We examined data collected from two gamified years, between which game changes took place. By clustering students according to their performance, we identified three distinct student types, common to both years: Achievers, Disheartened, and Underachievers. Interestingly, in the second year a new type of student emerged: the Late Awakeners. In this paper we carefully describe each student type, and explain how gamification can provide for smarter learning by catering to students with different profiles. Furthermore, we discuss how our findings, both in gamification and cluster analysis can be used to develop adaptive and smart learning environments.