Introduction

News

  • [Aug 23 2022] [Result] updated including auxiliary files
  • [Aug 22 2022] Competition result announced at 2022 IEEE CoG
  • [July 23 2022] Machine specs posted.
    – Windows 10 64bits
    – Intel i7-8700 CPU @3.20GHz
    – 8GB of memory
  • [June 23 2022] Entry updated (Metabot deprecated)
  • [June 22 2022] Announce Registration result on [Result]
  • [June 20 2022] Registration closed
  • [June 13 2022] Registration deadline was extended to 19th June
  • [June 7 2022] Map pack updated for BW 1.16.1 compatible
  • [Mar 11 2022] 2022 Competition page was created

Introduction

Welcome to the home of the annual IEEE CoG StarCraft AI competition which is organized by the Cognition & Intelligence Lab (CILab) at GIST, Gwangju, Korea. It is sponsored by the IEEE Conference on Games (CoG).

During this competition, programs (“bots”) will play 1v1 Starcraft Broodwar games against each other using BWAPI, a software library that makes it possible to connect programs to the Starcraft: BroodWar game engine.

The purpose of this competition is to foster the development & evaluation of progress in AI development applied to real-time strategy (RTS) games and solve challenging issues in RTS game AI such as uncertainty, real-time processing, and managing & coordinating agents. Where feasible, the competition strives for openness, transparency, reusability and reproducibility, both in the way the competition rules are defined and evaluated, and in the bots themselves.

RTS games pose a much greater challenge for AI research than chess because of hidden information, vast state and action spaces, and the requirement to act quickly. The best human players still have the upper hand in RTS games, but in the years to come this will likely change, thanks to competitions like this one. IEEE CoG StarCraft competitions have shown significant progress in the development and evolution of new StarCraft bots. For the evolution of the bots, participants have used various approaches to write AI bots and it has enriched game AI and methods such as HMM, Bayesian model, CBR, Potential fields, and reinforcement learning. However, it is still quite challenging to develop AI for the game because it should handle many units and buildings while considering resource management and high-level tactics.

Getting Started

Additional Software