A Study on New Gameplay Based on Brain-Computer Interface

This study examines whether brain-computer interface (BCI) devices can serve as a viable new gaming input, approaching the question from a game design perspective rather than purely technical or medical application. The authors note that BCI hardware has become smaller and cheaper, opening it to entertainment uses, yet design-oriented research on BCI games has been scarce. The work proposes game play elements built on consumer-grade BCI signals, presents a PC prototype that implements several of them, and evaluates whether BCI-augmented control is clearer and more efficient than keyboard and mouse alone.

The proposed elements map two brainwave-derived states, attention (concentration) and meditation (relaxation), onto familiar mechanics. In the prototype, attention reduces crosshair jitter to make aiming steadier, meditation speeds the recovery of the player character and affects searching for hidden objects, and sustained states reveal hints or hidden passages in puzzle and adventure contexts. A training mode with four mini-games teaches players to control these states, and the story mode combines BCI input with conventional controls. The design rationale draws on prior work using steady-state visual evoked potentials, which the authors identify as a foundation for real-time control, and on Sid Meier's definition of games as a series of interesting choices.

The evaluation compared two versions of the same game, one using a NeuroSky MindSet alongside traditional input and one without BCI, with 20 participants split evenly between groups. Questionnaire scores on convenience, fun, and intuitiveness were higher for the BCI version (149 versus 102 total, with means of 14.9 and 10.2), and a t-test at the 0.05 significance level indicated a statistically significant advantage (t ≈ 5.605). The authors conclude that BCI elements can make play more intuitive and immersive and offer guidelines for future BCI game design. The small sample size and the subjective, self-reported nature of the measures, however, limit how far these findings generalize.