Telling Eyes: Linking Eye-Tracking Indicators to Affective Variables

Virtual reality gaming environments offer a unique opportunity to implement adaptive gameplay systems that respond to a player’s emotional state in real time. By utilizing a custom riddle-based game titled Pertubatio, researchers examined the efficacy of eye-tracking metrics—specifically pupil size, blink rate, fixation duration, and saccade patterns—as reliable indicators of confusion and frustration. The study involved 39 participants navigating varying levels of difficulty, ranging from basic to impossible, to determine if physiological eye responses could accurately map to these specific affective states.

The findings confirm that game design successfully induces measurable levels of confusion and frustration, with confusion acting as a significant predictor of subsequent frustration. However, the correlation between these emotions and specific eye-tracking indicators remains inconsistent. Blink rate and gaze-related metrics, such as the ratio of fixations to saccades, emerged as more dependable predictors than pupil size. Pupil dilation proved particularly difficult to isolate as a reliable metric due to frequent luminance fluctuations inherent in virtual reality environments, which often obscure the physiological signals associated with emotional arousal.

Ultimately, while eye-tracking holds significant potential for enhancing player experience through adaptive mechanics, current technical and methodological limitations hinder its immediate application. The research highlights that individual demographic differences, headset calibration, and the environmental impact of game design significantly influence data accuracy. Future advancements in this field must prioritize refining these metrics to account for external variables, ensuring that eye-tracking can serve as a robust tool for real-time affective assessment in immersive gaming contexts.