Life Finds a Way Out

What am I looking at?

A 1v4 game where the player earns points for zapping or tagging their opponents and loses points for getting zapped. In order to clear the room, the player has to zap/tag each opponent four times. At any given moment, the player can either move or zap/tag. They move up, down, left or right, or they can zap/tag in one of those directions. So at each moment, they have to choose 1 of 8 options.

The player's choices are determined by a gene with an encoded sequence of instructions. For example, a gene might read “Move Left – Fire Up – Move Left – Move Down – Fire Right.” The simulation program will read those instructions and have the player act accordingly until the room’s goal is satisfied or the gene runs out of instructions. Then the gene will be mutated to look for a better solution.

What do the controls do?

Generations controls how many times the Player tries the exercise. One generation means one round of going through the entire sequence of actions encoded by their gene. After each generation, if they got the same score or better as the last generation, then the new gene is kept and mutated for the next generation. Otherwise, the old gene is mutated for the next generation. The more generations, the more opportunities they have to solve the exercise, but the longer it will take to run.

Genome length controls how many actions the Player can complete in one generation. The longer the genome, the more moves they make and the more time they have to solve the exercise. But again, it will take longer to run.

Mutation rate controls how often mutations occur, expressed as the probability of mutation per genome site per generation. For example, if the genome length is 100 and the mutation rate is 0.01 (1.00e-2), on average there will be one change per generation. If the mutation rate is 0.5, on average half of the genome locations will change per generation.

Under the Team section, you can decide which Player type trying the exercise. A zapper can zap opponents with a laser from afar. A tagger has to get next to the opponent to tag them.

The Begin! button will start a simulation with your chosen team and parameters. So that it doesn't take forever, the generations will evolve in the background. The most recent generation will be animated on the screen; once one animation completes, the next available generation will be animated. When the whole simulation is done, the final best solution will be animated. A movie of this animation will be available to download, in case your Players evolve something cool you want to save for posterity.

The chart below will show the fitness for generations where it stayed the same or increased. Each time you run an experiment, a new series of points will be added. You can right click and save "Save image as..." if you get results you want to keep.