Authors:
Matthew Stephenson, Eric Piette, Dennis J. N. J. Soemers, Cameron Browne

Venue:
Advances in Computer Games (ACG), 2021

Topics:
explainable AI, procedural content generation, game description, Ludii, human-AI interaction

Links: PDF · Springer · ACM · arXiv

This paper presents a framework for automatically generating manuals for board games described in the Ludii general game system.

The approach combines multiple components, including translation of game descriptions into English, move visualisation, detection of game endings, and strategy explanations derived from AI heuristics.

The resulting manuals provide intuitive and structured explanations of game rules and mechanics, enabling players to understand and play games even without prior familiarity with the underlying description language.

This work lies at the intersection of general game playing, explainable AI, and procedural content generation.

It extends the Ludii framework beyond game representation and AI agents, by making game descriptions accessible to human users through automatically generated explanations.

The approach also opens new directions for human-AI interaction, including automated tutorials, coaching systems, and adaptive learning environments.

Stephenson, M., Piette, E., Soemers, D. J. N. J., Browne, C. (2021). Automatic Generation of Board Game Manuals. In Advances in Computer Games (ACG).

@inproceedings{stephenson2021manuals,
  author    = {Stephenson, Matthew and Piette, Eric and Soemers, Dennis J. N. J. and Browne, Cameron},
  title     = {Automatic Generation of Board Game Manuals},
  booktitle = {Advances in Computer Games (ACG)},
  year      = {2021},
  url       = {https://arxiv.org/abs/2109.09507}
}