Authors: Quan Li, Peng Xu, Huamin Qu
Abstract: The rate at which frames are rendered in a computer game directly influences both game playability and enjoyability. Players frequently have to deal with the trade-off between high frame rates and good resolution. Analyzing patterns in frame rate data and their correlation with the overall game performance is important in designing games (e.g., graphic card/display setting suggestion and game performance measurement). However, this task is challenging because game frame rates vary both temporally and spatially. In addition, players may adjust their display settings based on their gaming experience and hardware conditions, which further contributes to the unpredictability of frame rates. In this paper, we present a comprehensive visual analytics system FPSSeer, to help game designers gain insight into frame rate data. Our system consists of four major views: 1) a frame rate view to show the overall distribution in a geographic scale, 2) a grid view to show the frame rate distribution and grid element clusters based on their similarity, 3) a FootRiver view to reveal the temporal patterns in game condition changes and potential spatiotemporal correlations, and 4) a comparison view to evaluate game performance discrepancy under different game tests. The real-world case studies demonstrate the effectiveness of our system. The system has been applied to an online commercial game to monitor its performance and to provide feedbacks to designers and developers.