I forget if I've asked this before, but what do y'all think about genetic algorithms as a technique for solving large games like, um, various forms of lowball poker?

I ask b/c implementing a genetic algorithm is easy, whereas I have no idea how to solve a large game via standard exact methods. I mean, encoding the strategy in a gene is not necessarily easy, and involves simplifications which introduce error, but simplifying poker into attributes is something I have quite a bit of experience with. Solving only that problem seems tractable.

I've started playing online a bit again lately, and since the last time I played, there seems to have erupted a world of $200-$400 2-7TD on Stars. I suspect that even good players don't play optimally, so perhaps there is some money to be made there.