Dynamic Difficulty Adjustment of Video Games Using Biofeedback

Abstract

Video games are typically designed to challenge the player, aiming to increase their engagement and enjoyment. However, this sense of challenge must be carefully balanced to provide a successful experience. For instance, if the player is unable to progress due to a high difficulty, they will probably feel frustrated and will possibly leave the game session; conversely, if the perceived difficulty is low, the player will feel bored and will also likely leave the session. In this paper we report the design of an algorithm, based in biofeedback, that dynamically adjusts the difficulty of a video game: if the heart rate of the player decreases, then the difficulty of the game increases; if the heart rate of the player increases, then the difficulty of the game decreases. Through a controlled user study, we evaluated the effectiveness of the algorithm and its implementation in a prototype video game, in terms of performance, perceived gaming experience, and player satisfaction. The obtained results show that users who were affected by the algorithm effectively reported an improved player experience, completed the game levels in a lower time with a fewer number of tries, and displayed a better subjective impression. The proposed algorithm highlights the feasibility of dynamically adjusting the difficulty of a video game using biofeedback, hence providing an improved and personalized player experience. These results can be useful to designers and researchers in the video game industry, as a way to conceive novel—and more engaging—experiences.