ENGLISH | JAPANESE

2nd International Aiwolf Contest

2nd International AIWolf Competition

AIWolf – 2nd International Werewolf AI Competition (and 6th Werewolf AI Competition).

The 2nd International Werewolf AI Competition will be held as part of the
2020 Automated Negotiating Agents Competition (ANAC 2020)
, at IJCAI 2020 in Yokohama.

Your objective in this competition is to build an agent that can play the game of Werewolf using artificial intelligence.
This competition has two divisions. Each division will be evaluated separately, so please check the details below.

NEWS:

Protocol Division:

Build an artificial intelligence agent that can play the werewolf game, using a formal grammar to communicate with other agents.
Participants in this competition will be evaluated by their overall winning rate.

  1. Competition Regulation: TBA
  2. How to participate
    • Make a team with one or more participants;
    • Each team must register at the registration site;
    • Registration is free;
    • See this site for the results of the practice contest;
  3. Schedule for the competition:
    The schedule below may change at any time; Please check often!
    Account Registration Opens: TBA
    Account Registration Deadline: May 1st, 2020
    Preliminary Contest Agent Submission Deadline: May 1st, 2020
    Announcement of the Finalists: June 1st, 2020
    Finalist Agent Submission Deadline: June 29th, 2020
    Deadline for submitting the Code Description: TBA (Just before IJCAI 2020)
    Announcement of the results: TBA (During IJCAI 2020)

  4. About the implementation of AIWolf Agents: You can submit an agent using Java, .Net or Python. However, the only programming language officially supported by the Contest organization is Java. If you are planning to make an agent in Java, please refer to the “How to Make an Agent” page. The Java version used in the competition environment is 1.8. Regarding .NET and Python, although we do not provide official support, the following libraries/how-to’s are provided. If you wish to use a particular library or language not listed above, please contact the organizers at gm_at_aiwolf.org (we cannot guarantee to accommodate all requests).

    According to the rules of this contest, we publish the source code for all finalist agents in this website. Please be mindful of the specific license for each project before using code from agents of past competitions.

  5. What to submit
    • Agent File:
      Please submit the agent files in a format that is already built and ready for running. The exact format depends on each language, so please refer to the regulations.
    • Code Description:
      Each selected finalist will need to submit, in addition to the source code for their agent and the documentation for running it (path and options, etc), a document with a simple description of the algorithm used. This document must be at least 2 A4 pages, and will be published after the competition. Please try to write it in a manner accessible for non-specialists. The Code Description must be in English. If you have any concerns about this requirement, please contact us directly.
  6. Awards/Prize: TBA
  7. Related Links

Natural Language Division:

In this division, participants must build an artificial intelligence agent that can play the werewolf game as humans do, using
Natural Language. Participants in this division will be evaluated by a panel of judges, who will grade the subjective quality of the dialog
generated by the agent. Agents must communicate in English (if your agent communicates in another language, use internal machine translation
to convert it to English).

  1. Competition Regulation: TBA
  2. How to participate:
    • Registration:
      A team should send a mail to aiwolf at kanolab.net (replace at by @), describing your team name, a contact e-mail address,
      names and affiliations of its members (please mark a contact person when a team consists of multiple members). Registration is free.
    • Agent Testing
      A participant is required to implement an AI werewolf agent that connects to our AIWolf server at the specified time.
      Before that, we will provide an AIWolf server where participants can try connecting with dummy agents to check their system’s behavior. Make sure that
      your system works correctly on the testing system before the actual competition run.
    • Preliminary Run
      Participants should run five agents connected to our server, running 50 games. Submit your game logs to the organisers via e-mail.
      If there are too many participants for the final run, organisers will select final run teams from these logs. In either case, these logs will be used in the final evaluation.
    • Final Run
      Each participant team is required to connect its agent to our server at a specified time to play games with other participants.
      If your team cannot remotely attend the final run, you should prepare your system so that the organisers can somehow trigger your remote system to connect
      our server and run, for example, via a web browser.
    • System Evaluation
      Participants should submit a system design description document to the organizers. This document and logs of the games might be used for research purposes
      and included and published in an overview paper without any further permission. Participants are encouraged to submit a paper to the ANAC workshop.
      Reviewers will perform subjective evaluations on the game logs, using following criteria:

      • A Natural utterance expressions
      • B Contextually natural conversation
      • C Coherent (not contradictory) conversation
      • D Coherent game actions (vote, attack, divine) with conversation contents
      • E Diverse utterance expressions, including coherent characterization

      Please note that vague utterances that could be used regardless of context are not always natural in the werewolf game.

  3. Schedule for the competition:
    The schedule below may change at any time; Please check often!
    Preliminary run (self-match games): TBA
    Final run (online games between participants): TBP (late-april)
    Announcement of the results: TBA (During IJCAI 2020)

  4. About the implementation of AIWolf Agents

    Agents must communicate in English (either natively or by using machine translation internally). The participat is required to implement an AI werewolf agent that connects to
    our AIWolf server via the network. Therefore, there is no limitation on the client side regarding hardware or software other than the network I/O specifications.

    We recommend to use our sample agent code with runtime environment (java), distributed from our Github repository:
    https://github.com/aiwolfdial/RandomTalkAgent

Contact Information

If you have any questions, please contact us at:

If you have any questions, please contact us at gm [at] aiwolf.org