Mobile Game Evaluation Method Based on Data Mining of Affective Time Series.

Area of Science:

• Game Data Science
• Affective Gaming
• Multimodal Sensor Analysis

Background:

• Understanding factors differentiating highly profitable (MP) from less profitable (LP) video games is crucial for developers.
• A common hypothesis suggests MP games elicit more positive player emotions, leading to sustained engagement.

Purpose of the Study:

• To introduce an original method for evaluating video game quality based on players' emotional reactions.
• To validate the hypothesis that players capable of distinguishing MP from LP games provide reliable emotional feedback.

Main Methods:

• Utilizing multimodal body sensors to analyze players' emotional responses during gameplay.
• Selecting test players based on their ability to differentiate between MP and LP games in short sessions.
• Analyzing the affective spectra of selected players to compare emotional responses to MP and LP games.

Main Results:

• Selected players exhibited significantly more positive and fewer negative emotions when playing MP games compared to LP games.
• This emotional differentiation suggests selected players are more attuned to game quality and engagement factors.
• The findings provide a clear confirmation of the proposed hypothesis regarding player emotional response and game profitability.

Conclusions:

• The developed method, leveraging affective computing and player emotional data, offers a reliable approach to assessing video game quality.
• Identifying and analyzing players with high discrimination ability is key to obtaining meaningful emotional feedback for game development.
• This research supports the link between positive emotional engagement and video game commercial success, validated through objective player data.