After seeing a thread on tig-source about genetic algorithms, and realizing i had 2 weeks to start and finish my science project for this year, i decided i would take a crack at making my own genetic solver thingy for my project!
<img src="http://dl.dropbox.com/u/1010927/eatman.png">
http://dl.dropbox.com/u/1010927/GeneticFoodMan.exe
Right now its pretty simple, and the amount of food gathered is the sole thing survival is rated on, but more things need to be programmed in. It works by seeding a bunch of random individuals, storing their movements, what they ate etc, waiting for everyone to die, then choosing the best 3 in the population, restarting using their movement paths, but then deviating onto their own random path once they reach the point where they're ancestors died, and i'm going to add a high mutation rate that messes around with the "death point" thats passed on to offspring. once all individuals die a new generation is created choosing the best 3 of the last and so forth. if an individual passes over a foodless space, or a space where they had already eaten for longer than "5 frames" they die. I'm probably gonna change this a little bit so that energy is lost and can only be regained through eating an equivalent number of food pieces.
Red just shows the collective mass of whats been eaten, but each organism only exists in its own specific world, unaffected by the others and their actions, so in short, a hell of a lot of data to track!