Lab10 peer review

[stefbarcio]

The readme is well-done and makes it easy to follow the code in its different sections. I really liked the idea of unbalancing the reward to address the first player advantage. I noticed the same problem in my simulations but wasn't able to figure out a solution, this could be a smart one. Definitively worth a try!
One small suggestion, you could take advantage of the highly symmetric structure of the game to aggregate the learned features on "different" boards that are actually the same