Effects of Peripheral Visual Information on Performance of Video Game with Hemi-Spherical Immersive Projection Screen

This study examines how players of an arcade action game, "Mobile Suit GUNDAM Senjyo no Kizuna," use peripheral visual and auditory information when playing on a hemi-spherical immersive projection screen inside a Panoramic Optical Display cockpit. The authors aim to estimate the size of the "effective visual space," defined as the minimal visual area that players continuously use during play, since understanding this is essential for designing wide-view immersive displays. The work draws on prior gaze-contingent window research, extending it to a larger display and to auditory conditions.

Three experiments were conducted with three highly trained professional game debuggers in the first two experiments and four in the third, all of whom had prior familiarity with the game. In the first experiment, circular visual windows at four sizes were used, and game scores rose with window size before plateauing between roughly 2 and 2.5 times the criterion size, indicating an effective visual space of a limited extent. Ratings of enjoyment, sensation of presence, and visually induced motion sickness showed no significant differences across window sizes. The second experiment masked upper, lower, or both regions of the visual field, finding that lower-field masking tended to reduce scores, though effects varied considerably across individuals and may reflect game-specific strategies such as using opponents' ground shadows as cues.

The third experiment compared play with and without sound. In the simpler first stage, scores were significantly lower with sound, suggesting auditory input may have distracted players, while ratings for enjoyment and presence were higher with sound. The authors note limitations including the very small, highly experienced sample, which may explain the weak association between subjective ratings and objective scores, and they call for further work on game strategy and prior experience.