TAG: Pandemic Competition

Gaina, Raluca D.; Balla, Martin

doi:10.1109/cog51982.2022.9893668

Cited by 2 publications

(4 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Tabletop Games framework (TAG) [4] is designed for research on modern TTGs. It provides a common API for SFP agents, various tools for automatic game optimisation, and metrics for evaluating both agents and games.…”

Section: A Tabletop Games Frameworkmentioning

confidence: 99%

“…To this extent, TAG provides a set of components, templates and interfaces to quickly implement new games, and a standard API for creating AI agents to play them. At the time of writing, TAG includes a collection of over 20 modern TTGs, including "Settlers of Catan", "Pandemic" [5], and "Terraforming Mars" [6]. All agents have access to the Forward Model of the game, which can be used to simulate potential future states in the game, when provided with a previous game state and action.…”

Section: A Tabletop Games Frameworkmentioning

confidence: 99%

“…As RL agents require a large amount of experience, slow environments are unsuitable for large-scale experiments. TAG has already shown high running speeds [4], but Java has little support for libraries that allow efficient implementation of RL agents, which leads to the requirement to implement a python API. So far, implementing new agents was only available in Java.…”

Section: Pytagmentioning

confidence: 99%

“…This work introduces PyTAG, a Python interface for interaction with TAG [4]. The communication between TAG and PyTAG is done by sharing the memory locations between Java and Python, which keeps the games running fast and efficiently.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

PyTAG: Challenges and Opportunities for Reinforcement Learning in Tabletop Games

Balla,

Long,

Jeurissen

et al. 2023

2023 IEEE Conference on Games (CoG)

Self Cite

View full text Add to dashboard Cite

In recent years, Game AI research has made important breakthroughs using Reinforcement Learning (RL). Despite this, RL for modern tabletop games has gained little to no attention, even when they offer a range of unique challenges compared to video games. To bridge this gap, we introduce PyTAG, a Python API for interacting with the Tabletop Games framework (TAG). TAG contains a growing set of more than 20 modern tabletop games, with a common API for AI agents. We present techniques for training RL agents in these games and introduce baseline results after training Proximal Policy Optimisation algorithms on a subset of games. Finally, we discuss the unique challenges complex modern tabletop games provide, now open to RL research through PyTAG.

show abstract

Section: A Tabletop Games Frameworkmentioning

confidence: 99%

Section: A Tabletop Games Frameworkmentioning

confidence: 99%

Section: Pytagmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations