New usage of telemetry for anti-cheating in FPS game

Abstract

First-person shooter games are experiencing a surge in popularity. As more players join, advanced AI-based cheats have emerged. These cheats simulate human gameplay, sending mouse inputs, making them hard to detect and counter. Therefore, this research presents a novel approach that utilizes telemetry data analysis to identify and counteract cheating in FPS games. The main objective of this study is to develop an innovative anti-cheating system that can effectively detect and prevent players from exploiting AI-based cheats to gain unfair advantages. To achieve this, extensive telemetry data is collected during gameplay. The data contains the real-time cursor position when the player is playing the game. Besides, Machine learning and deep algorithms are applied to analyse the telemetry data and distinguish between human player behaviour and AI-driven cheating patterns. Decision Tree, Random Forest, LSTM, and CNN are applied for this research. And in the final evaluation, CNNs accuracy reached around 80% which proves it is a possible mode to be used for this problem. The significance of this research lies in its contribution against cheating in FPS games, particularly those exploiting AI technologies to gain unfair advantage. The proposed telemetry-based approach offers a solution to safeguard competitive gaming and insight into the game company based on this novel way for further experiments.