"Hunt Takes Hare": Theming Games Through Game-Word Vector Translation

TL;DR

This work addresses the challenge of theming games by linking game embeddings learned from play logs with word embeddings that capture real-world semantics. It proposes three retheming models, including a Word-vectors-only baseline and two joint models that map game-space relationships into word-space translations using either a per-token linear transform or guiding vectors derived from specific examples or semantic fields. Empirical results on chess demonstrate that incorporating game embeddings yields more coherent and dynamic wildlife-themed rethemes than word-only methods, illustrating the value of dynamic game knowledge for thematic translation. The approach offers a scalable, data-driven path to theming, tutorialisation, and procedurally generated game content, with potential applicability across a broader set of games and themes.

Abstract

A game's theme is an important part of its design -- it conveys narrative information, rhetorical messages, helps the player intuit strategies, aids in tutorialisation and more. Thematic elements of games are notoriously difficult for AI systems to understand and manipulate, however, and often rely on large amounts of hand-written interpretations and knowledge. In this paper we present a technique which connects game embeddings, a recent method for modelling game dynamics from log data, and word embeddings, which models semantic information about language. We explain two different approaches for using game embeddings in this way, and show evidence that game embeddings enhance the linguistic translations of game concepts from one theme to another, opening up exciting new possibilities for reasoning about the thematic elements of games in the future.