Student Starcraft AI Tournament Press Kit

Contents:

Factsheet


What is SSCAIT?

The Student StarCraft AI Tournament is an educational event, first held in 2011. It serves as a challenging competitive environment for students of Artificial Intelligence and Computer Science, as well as other AI and Broodwar enthousiasts. They are submitting the bots, which are programmed in C++ or Java using the custom API BWAPI to play 1v1 StarCraft matches against one another.

This being the 5th year we organise the tournament, we invited back a number of participants that have been in the scene from the very start, and of course numerous newly developed entrants hope to achieve big things, or add a little glamour to their BA or MA theses.

In 2014, Martin Rooijackers from the Netherlands won with LetaBot, this year, 44 participants (so far), from all over the world will try to take the title from him. While Chris Coxe with ZZZKBot won CIG this year, and Vegard Mella with tscmoo won AIIDE, strong contenders beside them, for this year's SSCAIT appear to be Martin Rooijackers (LetaBot), Sijia Xu (Overkill), Marian Devecka (Killerbot) and Dave Churchill (UAlbertaBot).

This being a strictly AI vs AI tournament, humans will only be coding and handling the servers. Russian player Djem5 played a couple of demonstration matches for AIIDE 2015, in which he summarily beat all AI opposition without breaking a sweat. The current state of Broodwar AI is such, that the top 5-10 bots are about equal to casual players. They aren't as adaptive yet as humans, so they can be tricked by abusing situations they are not yet able to handle.

What this year't tournament is all about

This year we're seeing an increase in the amount of actual AI involved in the matches. While most of the older bots have hard-coded plans of what to build when, where to attack and how, newer bots apply a mixture of those methods and adaptive programming. They are able to respond to the adversary's air-units by building anti-air weapons. Another possibility is to delay the production of combat units in favour of economic expansion, based on scouting information. Thus, they will have a stronger economy, and a better chance of victory in the long run. A handful of bots actively engage in learning strategies, like Martin Rooijackers' LetaBot, or Dave Churchill's UAlbertaBot. tscmoo is currently working on learning algorithms that will allow the AI to learn from watching replays of earlier played games. The most successful AI is the one with the optimal balance between a good plan, clever heuristics and fastest learning capabilities.


Videos

Introduction to Broodwar AI and SSCAIT.

2014 finals between XIMP and LetaBot.

tscmoo nuking everything.

krasi0 vs tscmoo zerg.

More recent videos.


Images

These images are from our tournament live stream server. Some contain debug-information from the bots in question.

High fps debug information.

Zergling rush.

Terran vs Protoss with graphical path-finding.

Terran Wraiths destroying the enemy.


Logo

SSCAIT logo

Team

Michal Čertický Founder, Admin, Programming homepage
Marek Šuppa Security, Programming (web) homepage
Björn Persson Mattsson Programming (tournament module & observer) homepage
Tomáš Vajda Programming (torurnament module & observer)
Rafał Poniatowski Bug reporting, Programming (web)
Sören Klett Bug reporting
Paul Paradies PR, Promo homepage

In the news


Press/PR Contact


Banners

SSCAIT banner 1 SSCAIT banner 2

Posters

SSCAIT poster 1 SSCAIT poster 2

return to the top