Player Engagement Classification in Mobile Games Using MUSE Headband

Mobile games have replaced computer games due to their popularity and accessibility. The increasing trend of mobile and online games has transformed gaming culture. The success of games requires them to have appealing gaming interfaces, optimal player interactions, and engaging playability characteristics. Therefore, analysis of a game player's engagement in any game is important to determine the effectiveness of any game. This study presents an electroencephalography (EEG) based player engagement classification model by utilizing MUSE Headband. EEG data were acquired from 31 players while playing three (Tetris, Subway Surfer, Smash Hit) games. A questionnaire was filled out to label the acquired data for supervised classification. Data was labeled into three classes (bored, engaged, and excited) based on a score of questionnaires. Two classifiers support vector machine and multi-layer perceptron were used to identify the engagement level of players and accuracy of 86.2% and 80.4% were achieved respectively. A comparative analysis of current methodology with the previous research techniques depict that currently proposed methodology is more effective in terms of level of classification accuracy, robustness and flexibility due to its applicability on variety of games. This shows the potential of our system to be used in in-game engagement analysis for a better player experience and thereby the success of the game in the market.