Fresh Meat game comparison
Fresh Meat was reactivated only a few days ago (see my brief writeup). Its play shows sudden switches: Compare Fresh Meat’s first game versus Prism Cactus with its second game against the same opponent. The first game looks aimless, the second shows a clear plan more-or-less cleanly executed.
Fresh Meat has been updated repeatedly in a short time, so any given play difference might be caused by a code change. But Hao Pan hinted that it records its opponent’s unit mix and plays to counter it. From his comment: “it is built on the in-game AI provided there’s no past games. This time I rewrote the unit composition record system, and am looking forward to the build orders COEP comes up with.”
The two games and the statement both say: Here we have an example of recording the opponent’s habits and figuring out how to counter them. It’s similar in concept to what I’m doing with opening data. This kind of system can learn very fast because it reacts immediately to new data without having to do any statistical averaging or other slow stuff. And it can make drastic adaptations, because the adaptation is done by reasoning—in the case of COEP, the reasoning takes the form of an evolutionary search. Its limits are set by the data it records and the capability of its reasoning system.
Trackbacks
No Trackbacks
Comments
Hao Pan on :
Fresh Meat now has 5 build orders, only the 5th is driven fully by COEP, the other 4 are prespecified builds whose parameters can be learned. For example, if the opponent rushed in the previous game and won, the number of sunks will be increased in the next game. And as you mentioned, I've been frequently updating Fresh Meat, because I have yet to finish its 4th build.
Regarding the 2nd game vs Prism Cactus, the build order was a prespecified one unfortunately. But here's a game where Fresh Meat scored its first(?) win using COEP:
http://www.openbw.com/replay-viewer/?rep=https://data.basil-ladder.net/bots/Fresh%20Meat/Fresh%20Meat%20vs%20Martin%20Rooijackers%20Heartbreak%20Ridge%20CTR_82CC562F.rep
I have to admit it's a sloppy game and it's obvious that Fresh Meat experienced some production freeze (a bug to fix). COEP was working as intended I believe. Fresh Meat just failed to execute (build/train units) for reasons. This adds another type of limit to those you already pointed out.
Jay Scott on :
The game is similar to one of Steamhammer’s favorite plans against LetaBot by Martin Rooijackers: Fast pool to confuse LetaBot into overreacting, then many drones, then lurkers.
Hao Pan on :
Glad that COEP came up with a similar plan against LetaBot, great minds think alike :P
Jay Scott on :
MicroDK on :
I really like that kind of adaptability. Instead of being reactive it will become more proactive, given the opponent plays the same plan next time. Most bots in the lower end do not change their plan much.
MicroDK on :
Jay Scott on :