Genetic Algorithms Used to Find Optimal Starcraft 2 Zerg Build Order
A poster on the TeamLiquid Starcraft 2 forum named Lomilar has devised a program, called Evolution Chamber, that uses genetic algorithms to find the most optimal build orders for the Zerg race. As indicated by the above video, the build method seems to provide a fairly large winning percentage, making the strategy nearly unstoppable for the time being.
Of course, this sparked debates in which players search for ways to stop the nigh invincible build order, and as explained by Louis Brandy in an excellent in-depth explanation and analysis, the strategy can be utterly unstoppable.
The odd thing about the strategy, aside from being relatively unknown to the Starcraft community, is that it’s extremely counterintuitive to the basic rules of Starcraft that are followed by most of the skilled players, which again, are explained in detail in Louis Brandy’s excellent analysis.
One of the reasons build-order optimization is so important is that you can discover openings that “hard-counter” other openings. If I can get an army of N size into your base when you do opening X, you will always lose.
We gamers could rule the world if we weren’t busy using our powers for gaming purposes.
(via Rock, Paper, Shotgun)