The use of serious games on virtual learning platforms as a learning support resource is increasingly common. They are especially effective in helping students acquire mainly applied curricular content. However, a process is required to monitor the effectiveness and students’ perceived satisfaction. The objectives of this study were to (1) identify the most significant characteristics; (2) determine the most relevant predictors of learning outcomes; (3) identify groupings with respect to the different serious game activities; and (4) to determine students’ perceptions of the usefulness of the simple and complex serious game activities. We worked with a sample of 130 university students studying health sciences and biomedical engineering. The serious game activities were applied in a Moodle environment, UBUVirtual, and monitored using the UBUMonitor tool. The degree type and the type of serious game explained differing percentages of the variance in the learning results in the assessment tests (34.4%—multiple choice tests [individual assessment]; 11.2%—project performance [group assessment]; 25.6%—project presentation [group assessment]). Different clusters were found depending on the group of students and the algorithm applied. The Adjusted Rang Index was applied to determine the most appropriate algorithm in each case. The student satisfaction was high in all the cases. However, they indicated complex serious games as being more useful than simple serious games as learning resources for the practical content in both health sciences and biomedical engineering degrees.
