Player co-modelling in a strategy board game: discovering how to play fast

TL;DR

This paper explores evolving neural network-based models to play a strategy board game quickly, highlighting the relationship between game speed, winning ability, and human involvement levels.

Contribution

It introduces a method for evolving playing models with reinforcement learning to optimize game speed while analyzing human involvement effects.

Findings

Faster models tend to win more often.

There is a threshold of human involvement beyond which learning stagnates.

Speed and success are positively correlated in evolved models.

Abstract

In this paper we experiment with a 2-player strategy board game where playing models are evolved using reinforcement learning and neural networks. The models are evolved to speed up automatic game development based on human involvement at varying levels of sophistication and density when compared to fully autonomous playing. The experimental results suggest a clear and measurable association between the ability to win games and the ability to do that fast, while at the same time demonstrating that there is a minimum level of human involvement beyond which no learning really occurs.