Predicting Quality of Video Gaming Experience Using Global-Scale Telemetry Data and Federated Learning

TL;DR

This paper develops a federated learning model to predict video game FPS performance across diverse global users, balancing accuracy with privacy, and demonstrates its effectiveness on a large-scale telemetry dataset.

Contribution

It introduces a novel federated learning approach with dynamic kernels for accurate, privacy-preserving FPS prediction across global users and games.

Findings

Achieved a mean Wasserstein distance of 0.469, outperforming baselines.

Identified key factors influencing game FPS on a global scale.

Proposed a plug-and-play kernel scheme to address cold start issues.

Abstract

Frames Per Second (FPS) significantly affects the gaming experience. Providing players with accurate FPS estimates prior to purchase benefits both players and game developers. However, we have a limited understanding of how to predict a game's FPS performance on a specific device. In this paper, we first conduct a comprehensive analysis of a wide range of factors that may affect game FPS on a global-scale dataset to identify the determinants of FPS. This includes player-side and game-side characteristics, as well as country-level socio-economic statistics. Furthermore, recognizing that accurate FPS predictions require extensive user data, which raises privacy concerns, we propose a federated learning-based model to ensure user privacy. Each player and game is assigned a unique learnable knowledge kernel that gradually extracts latent features for improved accuracy. We also introduce a novel training and prediction scheme that allows these kernels to be dynamically plug-and-play, effectively addressing cold start issues. To train this model with minimal bias, we collected a large telemetry dataset from 224 countries and regions, 100,000 users, and 835 games. Our model achieved a mean Wasserstein distance of 0.469 between predicted and ground truth FPS distributions, outperforming all baseline methods.