Miranda, Nadia MontañaChover, MiguelArgudo, OscarIparraguirre, Olatz2025-05-262025-05-262025978-3-03868-284-4https://doi.org/10.2312/ceig.20251108https://diglib.eg.org/handle/10.2312/ceig20251108Non-player characters (NPCs) play a crucial role in the gaming experience and should be designed to suit the characteristics of each game genre. This study examines the impact of AI-driven NPCs on player experience by comparing their impact in a hospital management simulation game and in a competitive exergame called Fitoon. In the simulation game, a needs-based AI system was implemented, where agents autonomously make decisions based on their internal states, such as hunger or stress. In Fitoon, machine learning techniques, including reinforcement and imitation learning, were applied to train agents capable of competing against players in obstacle races. Through user testing utilizing the In-Game version of the Game Experience Questionnaire, this study examines how the presence of these NPCs influences the players' perception of the gaming experience in both contexts. The results indicate that while NPCs in the simulation game have a greater impact on reducing feelings of tedium and boredom, in the exergame, they significantly increase the sense of challenge.Attribution 4.0 International LicenseCCS Concepts: Computing methodologies → Artificial intelligence; Multi-agent systems; Intelligent agents; Machine learning; Learning from demonstrations; Reinforcement learning; Human-centered computing → HCI design and evaluation methods; Software and its engineering → Interactive games;Computing methodologies → Artificial intelligenceMulti agent systemsIntelligent agentsMachine learningLearning from demonstrationsReinforcement learningHuman centered computing → HCI design and evaluation methodsSoftware and its engineering → Interactive games10.2312/ceig.202511088 pages