Collaborative learning has been shown to enhance the motivation and enjoyment of learners. However, when trying to utilize these advantages in collaborative educational games, the game has to effectively keep track of the learners’ states of knowledge and adapt the game accordingly. This includes only confronting players with tasks appropriate for their capabilities to avoid frustration or boredom and promote engagement. This requires the consideration of multiple learners’ skills and respecting their potentially varying learning paces. In this paper, a concept is developed for updating the learners’ knowledge states after completing a task and dynamically selecting the next appropriate level based on these states. This concept leverages the Competence-based Knowledge Space Theory (CbKST) to represent the learner’s state of knowledge. The developed concept is implemented into a collaborative learning game, which is subsequently evaluated through a user study. The results of the evaluation indicate that the concept successfully enhanced player motivation and enjoyment within the collaborative learning game. However, it is important to note that these findings are limited in significance and require further research to fully determine the impact of the proposed approach on the players’ motivation and enjoyment.