Bayesian Optimization-based Search for Agent Control in Automated Game Testing

2025-08-18

AI Summary

This paper addresses the poor scalability and low defect-detection efficiency of conventional approaches in automated game level testing. We propose an intelligent agent testing framework that integrates Bayesian optimization with a lightweight grid-map modeling technique tailored for game testing. Our method introduces a spatially smoothed, uncertainty-aware grid-map representation to enable efficient large-scale level modeling, and couples it with agent control policies guided by Bayesian optimization for adaptive, risk-aware exploration of high-risk regions. Experimental results demonstrate significant improvements: a 23.6% increase in map coverage and a 41.2% reduction in the Kolmogorov–Smirnov statistic—indicating markedly improved exploration uniformity—while maintaining high detection accuracy. The framework achieves substantial gains in system scalability without compromising reliability, offering a production-ready, automated solution for game quality assurance.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Reasoning under Uncertainty: Stochastic Optimization
• Intelligent Robots: Learning & Optimization for ROB

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

This work introduces an automated testing approach that employs agents controlling game characters to detect potential bugs within a game level. Harnessing the power of Bayesian Optimization (BO) to execute sample-efficient search, the method determines the next sampling point by analyzing the data collected so far and calculates the data point that will maximize information acquisition. To support the BO process, we introduce a game testing-specific model built on top of a grid map, that features the smoothness and uncertainty estimation required by BO, however and most importantly, it does not suffer the scalability issues that traditional models carry. The experiments demonstrate that the approach significantly improves map coverage capabilities in both time efficiency and exploration distribution.

Problem

• Automated bug detection in game levels using agent control
• Bayesian Optimization for efficient search and information maximization
• Scalable game testing model overcoming traditional limitations

Innovation

• Bayesian Optimization for efficient game testing
• Grid map model with smoothness and uncertainty
• Scalable solution for automated bug detection