Representing and Generating Levels Over Time through Playtrace Reconstructive Partitioning

2026-07-13

AI Summary

This work addresses the common oversight in existing procedural content generation (PCG) methods—their neglect of the temporal dynamics inherent in level content. To overcome this limitation, the authors propose Cake, a domain-agnostic temporal level representation that explicitly models the structural evolution of levels over time. Complementing this representation, they introduce Playtrace Reconstructive Partitioning (PRP), an algorithm designed to efficiently reconstruct and segment player trajectories for guided generation. Evaluated on the Sokoban benchmark, the proposed approach significantly outperforms six state-of-the-art PCG methods. The generated levels are not only solvable but also exhibit rich solution diversity, demonstrating the method’s effectiveness in preserving dynamic information while enhancing overall generation quality.

Abstract

Video games are a dynamic medium experienced over time. While there are many Procedural Content Generation (PCG) approaches for generating video game levels, they often use representations that abstract away this dynamic nature. In this paper, we introduce a novel, domain-independent ``cake'' representation for game levels over time which implicitly encodes dynamic information. We present a novel level generation approach Playtrace Reconstructive Partitioning (PRP) specifically developed for this cake representation. We compare against six state-of-the-art PCG approaches in the game domain of \textit{Sokoban}, and find that our approach can generate valid levels without sacrificing solution diversity. We believe our cake representation more neatly encodes the implicit dynamic nature of games compared to existing representations, which allows for our domain-agnostic level generation algorithm PRP.

Problem

• Procedural Content Generation
• dynamic representation
• level generation
• playtrace
• Sokoban

Innovation

• cake representation
• Playtrace Reconstructive Partitioning
• procedural content generation
• dynamic game levels
• domain-independent