Cardistry
Cardistry explores the design space of machine learning for card games. It aims to develop interesting and enriching game experiences that could only be possible through machine learning and research the effectiveness towards meeting desired emotional outcomes and other psychosocial benefits for its players.
At present, Cardistry is a novel AI-powered card generation engine made with Unity that prompts users to tell personal stories. It then turns the users’ stories into bespoke playing card. Cardistry is powered by OpenAI. The application creates premium plastic game cards that users can keep as a keepsake or use to enhance any card game that benefits from evocative artwork or a suit.
The application is set up to interpret the stories according to parameters that game designers can provide. The engine can easily turn stories into any kind of game card a designer can envision. We have been able to use it to generate regular playing cards, Tarot-cards, Dixit cards and Magic the Gathering-style TCG cards.
Demos
Foundations of Digital Games Conference
Session 5 – Games and Demos
Worcester, MA
May, 2024
Presented By: Brandon Lyman, Ala Ebrahimi
Game Developers Conference
Exposition Floor
San Francisco, CA
March, 2023
Presented By: Games@Northeastern, Brandon Lyman, Bob De Schutter
Videos
Cardistry in action.
Publications
Cardistry: Exploring a GPT Model Workflow as an Adapted Method of Gaminiscing
Lyman, B., Ebrahimi, A., Cox, J., Chan, S., Barney, C., & De Schutter, B. (2024). Cardistry: Exploring a GPT Model Workflow as an Adapted Method of Gaminiscing. Proceedings of the 19th International Conference on the Foundations of Digital Games, 1–4. https://doi.org/10.1145/3649921.3656984
Cardistry is an application that enables users to create their own playing cards for use in evocative storytelling games. It is driven by OpenAI’s Generative Pre-trained Transformer (GPT) models that generate unique card titles, cards suits, imagery, and poetry based on the user’s input. It allows the user to preserve their digital cards in an online repository and print them for tabletop game play use. Cardistry was designed to begin exploring the question of whether widely available GPT models could be used to adapt the process of gaminiscing to make it more accessible to designers and players alike. This short paper details the concept, design, and implementation of Cardistry as a first step in exploring this research question. It explains how the adapted gaminiscing process is different from the original process, discusses the limitations of the implementation, and expresses what future research would be required to answer the research question.