Mechanic Maker 2.0: Reinforcement Learning for Evaluating Generated Rules

TL;DR

This paper explores using Reinforcement Learning to evaluate and generate game rules, creating a new framework that produces more human-like and potentially more usable game mechanics compared to traditional static methods.

Contribution

Introduces RL as a dynamic approximator for human play in automated game design, recreating Mechanic Maker in Unity as an open-source rule generation framework.

Findings

RL produces distinct, potentially more human-like rules

Recreated Mechanic Maker environment in Unity

RL-based rules may be more usable by humans

Abstract

Automated game design (AGD), the study of automatically generating game rules, has a long history in technical games research. AGD approaches generally rely on approximations of human play, either objective functions or AI agents. Despite this, the majority of these approximators are static, meaning they do not reflect human player's ability to learn and improve in a game. In this paper, we investigate the application of Reinforcement Learning (RL) as an approximator for human play for rule generation. We recreate the classic AGD environment Mechanic Maker in Unity as a new, open-source rule generation framework. Our results demonstrate that RL produces distinct sets of rules from an A* agent baseline, which may be more usable by humans.