I think what I'm trying to say is: if your fitness function contains the solution, then your problem is entirely contrived, and not a good illustration of genetic algorithms - you're just using a very round-about way of calculating something you already knew, inefficiently.