Large language models as RPG game masters : A comparative analysis on performance and player experience

Details

Supervisors

Verhaeghe, Hélène

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] en sciences informatiques, à finalité spécialisée
Master [120] : ingénieur civil en science des données, à finalité spécialisée: cybersécurité

Abstract

enThis master’s thesis investigates the use of large language models (LLMs) as game masters in role-playing games (RPGs). It presents a custom-built 2D application that integrates multiple LLMs into a modular game system, enabling dynamic storytelling, combat, memory handling, and player interaction. The study evaluates the models’ creativity, coherence, responsiveness, and susceptibility to player influence. While the results show that LLMs can generate rich and engaging narratives, they also reveal key limitations such as logical inconsistencies and memory issues. The research highlights both the potential and the current challenges of using LLMs for interactive storytelling in games.