一句话：通过两个游戏项目的协作式自民族志研究，揭示LLM嵌入游戏架构后对玩法、可玩性与玩家体验的重塑及质量挑战。
本文考察将大型语言模型集成到游戏开发中会如何影响玩法（gameplay）、可玩性（playability）与玩家体验。作者采用协作式自民族志方法，对两个将LLM作为架构组件嵌入的游戏项目进行研究，并以反思性叙事与开发制品为材料，围绕上述三个构念展开分析。研究发现，LLM的集成提升了游戏内容的可变性与个性化程度，但同时引入了正确性、难度校准以及跨这些构念的结构一致性方面的挑战。该工作为生成式AI集成如何重塑既有游戏构念、并在游戏工程实践中带来新的架构与质量考量提供了初步的实证洞见。
This paper investigates how the integration of large language models influences gameplay, playability, and player experience in game development. We report a collaborative autoethnographic study of two game projects in which LLMs were embedded as architectural components. Reflective narratives and development artifacts were analyzed using gameplay, playability, and player experience as guiding constructs. The findings suggest that LLM integration increases variability and personalization while introducing challenges related to correctness, difficulty calibration, and structural coherence across these concepts. The study provides preliminary empirical insight into how generative AI integration reshapes established game constructs and introduces new architectural and quality considerations within game engineering practice.
DOI 原文 · arXiv · PDF（开放获取） · DBLP
@inproceedings{abs-2603-27896,
title = {Large Language Models in Game Development: Implications for Gameplay, Playability, and Player Experience},
author = {Keeryn Johnson and Muhammad Ahmed and Charlie Lang and Sahib Thethi and Wilson Zheng and Ronnie de Souza Santos},
booktitle = {FSE},
year = {2026},
doi = {10.1145/3803437.3806368}
}