BRGUARD: Revolutionizing Cheat Detection in Battle Royale Games with Player Behavior Features

The scarcity of large-scale, real-world datasets for Battle Royale (BR) games has significantly hindered the advancement of server-side cheat detection research. To bridge this gap, we present the Farlight-84 Cheat Detection Dataset, a comprehensive collection of gameplay telemetry sourced from the popular BR game Farlight-84. This dataset distinguishes itself from existing First-Person Shooter (FPS) datasets by covering the unique macro-strategic complexities of the BR genre, including looting efficiency, zone rotation logic, and long-range engagement patterns. It comprises over 37,000 matches and 24,000 unique players, labeled with ground-truth verdicts from official anti-cheat records. A key feature of this dataset is its cross-platform coverage, providing distinct behavioral distributions for both Mobile (touchscreen) and PC (mouse and keyboard) environments. This repository aims to facilitate research in behavioral anomaly detection, multi-view ensemble learning, and robust game security systems.