Tales of Tribute

Tales of Tribute Competition

Tales of Tribute Competition

Introduction

The deckbuilding card game Tales of Tribute is a special activity added to the massively popular multiplayer online role-playing video game The Elder Scrolls Online. Although the game remains small by CCG standards (about 200 cards and only a few keywords), it is interesting and challenging for humans, with a significant potential for being a good AI testbed.

The players start with the same base cards and build their decks during the game by buying cards from a shared resource pool called the Tavern (a concept similar to Dominion). Tales of Tributes encourages strategic long-term planning– to ensure the deck we build contains enough strong cards and is not clustered with weak ones. Because of a huge randomness factor influencing, e.g., which cards occur in the Tavern, it is hard for the agents to stick with a pre-made plan successfully. So the game also requires considerable adaptiveness.

Tales of Tribute AI Competition aims to fill the void after the Hearthstone AI Competition while being significantly more challenging than a toy problem covered by Legends of Code and Magic and the Strategy Card Game AI Competition.

The competition will be run using ScriptsOfTribute, an open reimplementation of the original game in the .Net framework designed especially for this event. It features the game manager that allows running AI agents implemented as C# classes (adapters for other languages are possible) against each other and a graphical user interface that support human vs. AI games.

The competition’s winners will be decided upon the global winrate in all vs. all tournament, conducted on a large number of mirror matches using the same seeds.

Logistics

Console game runner and Unity GUI client are available for download from the competition webpage.>/p>
Send source agent and its description to jko@cs.uni.wroc.pl
Time limit: 10 seconds for every turn
Available languages: C# is native to the game, Python „as native” with access to the forward model, other languages supported via adapter
Evaluation: symmetrical games between all pairs of players (may include example agents) until the ranking clearly stabilizes.
We are planning to apply to CIS for prizes (there were prizes previously)
Entry barrier is medium. For reasonable submission, an understanding of the forward model plus some search algorithm or reasonable heuristics is required

Participants

This year, we expect some more submissions, as multiple new people showed interest via the discord server. We also did some work to be more python+NN’s friendly.
Target audience: students with a background in search/ML, competitive AI programmers
Additionally, the contest will be advertised for Competitive AI programmers, e.g. within the CodinGame community.
We plan a number of improvements, including better python integration, improved GUI, new decks available.