The Future Possibilities of Mimicking EEG Signals To Generate Flow State in Immersive Educational Games

Abstract

Player engagement in educational games is most effective when players enter the psychological state of flow, defined as the balance between challenge and skill that fosters deep concentration, intrinsic motivation, and sustained enjoyment. Traditional approaches to adaptive gameplay in immersive virtual reality and extended reality environments have relied heavily on electroencephalography (EEG) or other biosensors to monitor cognitive states in real time. While these methods provide accurate insights into player engagement, their reliance on costly, intrusive, and impractical hardware restricts their scalability in mainstream gaming, education, and health application. This paper proposes an EEG free meta learned player behaviour model designed to predict EEG like engagement markers directly from gameplay telemetry, such as input timing, retries, pacing, and decision making patterns. The proposed framework integrates teacher–student knowledge distillation to mimic EEG based models while applying meta learning algorithms (MAML/iMAML) to enable rapid personalisation across diverse players and genres. By validating against EEG calibration datasets and psychological flow measures, the model aims to deliver scalable, sensor free engagement prediction and flow induction in immersive VR/XR educational games. This work represents a preposed framework toexploring, a hybrid mimicry–personalisation framework, and an ethically grounded approach to adaptive game design. Applications extend beyond education to entertainment and healthcare, offering a pathway for large scale, personalised, and responsible affect adaptive systems. Immersive VR educational games offer a unique opportunity, as existing HMDs provide the necessary sensors using this model to allow for a state of Flow to be discovered without any EEG input.