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

Agentic AI
Published: arXiv: 2508.13121v1

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.

Paper Summary

Problem
Automated game testing is a challenging task, especially when it comes to detecting potential bugs within a game level. Traditional methods can be slow, inefficient, or even miss important issues.
Key Innovation
This research introduces a novel approach that combines Bayesian Optimization (BO) with a game testing-specific model built on top of a grid map. This allows for efficient search and exploration of the game level, while also providing scalability and uncertainty estimation required by BO.
Practical Impact
The proposed system has significant potential to improve automated game testing by increasing map coverage capabilities in both time efficiency and exploration distribution. This could lead to faster bug detection, reduced development costs, and improved overall gaming experience.
Analogy / Intuitive Explanation
Imagine you're trying to find a specific location within a large maze. Traditional methods would be like searching the entire maze one step at a time, which can be slow and inefficient. The proposed system is like having a map of the maze that helps you navigate and prioritize your search, allowing you to find the location more quickly and effectively. By combining BO with a game testing-specific model, this research has developed a powerful tool for automating game testing, making it easier to detect bugs and improve overall gaming experience.
Paper Information
Categories:
cs.AI
Published Date:

arXiv ID:

2508.13121v1

Quick Actions