Toward Game Level Generation from Gameplay Videos

TL;DR

This paper introduces a method to automatically learn game design knowledge from gameplay videos and use it to generate new game level sections, demonstrated on Super Mario Bros.

Contribution

It presents a novel approach to extract game design patterns from videos and generate game levels using probabilistic models, reducing reliance on manual design.

Findings

Successfully learned sprite placement patterns from gameplay videos

Generated game level sections with high playability

Achieved stylistic similarity to original levels

Abstract

Algorithms that generate computer game content require game design knowledge. We present an approach to automatically learn game design knowledge for level design from gameplay videos. We further demonstrate how the acquired design knowledge can be used to generate sections of game levels. Our approach involves parsing video of people playing a game to detect the appearance of patterns of sprites and utilizing machine learning to build a probabilistic model of sprite placement. We show how rich game design information can be automatically parsed from gameplay videos and represented as a set of generative probabilistic models. We use Super Mario Bros. as a proof of concept. We evaluate our approach on a measure of playability and stylistic similarity to the original levels as represented in the gameplay videos.