AI to beat at least one of the ten top-ranked humans in Starcraft Brood War by January 1, 2019

Created by JoshuaZ on 2016-03-23; known on 2019-01-01; judged wrong by JoshuaZ on 2019-01-01.

• JoshuaZ estimated 73% on 2016-03-23
• JoshuaZ said “For this purpose, beat means beat in at least one game. ” on 2016-03-23
• ioannes estimated 30% on 2016-03-23
• splorridge estimated 58% on 2016-03-23
• themusicgod1 estimated 82% and said “Writing is on the wall.” on 2016-03-23
• PlacidPlatypus estimated 60% and said “I’m sure it’ll happen eventually but the timeframe here is awfully short.” on 2016-03-23
• JoshuaZ said “Placid, good point. I’ve made a second prediction with a longer time range: http://predictionbook.com/predictions/177740” on 2016-03-23
• Raahul_Kumar estimated 0% and said “Assuming AlphaGo does not come, and we are restricted to existing bots only. NO ALPHAGO!http://satirist.org/ai/starcraft/blog/Existing bots are crap, and can’t learn no matter what they try. ” on 2016-03-24
• JoshuaZ said “This prediction makes no assumptions about “no AlphaGo”- Google is certainly interested in Brood War and would count fine for this prediction.” on 2016-03-24
• JoshuaZ said “ Note also that in fact existing bots can learn albeit in limited ways: the Berkley Overmind is one of the older of the high-tier AIs and it learned a lot of what it does. (There are a bunch of good Youtube videos about it.) ” on 2016-03-24
• Raahul_Kumar said “Tscmoo and Overkill gained about 3% in win rate between round 20 and the end . LetaBot has a slight upward trend. The others look flat or even trend downward. ” on 2016-03-27
• bitbatbot estimated 40% on 2016-04-03
• JoshuaZ   judged this prediction wrong on 2019-01-01.