Game players almost always experience games through electronic devices, such as computers, cell phones, headphones, and microphones, which allow them to play and communicate. However, players’ emotional changes during games cannot be sensed accurately in real time. Therefore, to evaluate whether players like a particular game, questionnaires are generally required afterwards to collect information on players’ feelings, which do not provide a complete record or real-time information. Advances in AI have increased convenience in everyday life, and emotion sensors based on AI, which employ face recognition techniques and physiological sensors, have been introduced. Thus, through the interpretation of facial expressions, users’ emotional reactions while they are using electronic devices can be recognized. Moreover, voice recognition can also assist in improving the accuracy of AI-based emotional recognition. In this study, we examine how an emotion sensor based on AI can enable a more immediate and complete assessment of game playability, serving as an important reference for game development.