Peer review lab2 by Alberto Ricatto - s309141

[albyrika]

What i liked

• the code works and the strategy converges
• the code is clean and comprehensible even if there is no readme file and the documentation is not extensive
• good to evaluate over a random number of dimensions, i think that helps with exploration
• the optimal strategy is really an optimal one

What can be improved

• documentation and readme file would help
• the startegy used is, as a matter of facts, a (1+1) hill climber, not an ES
• the code can be expanded, also trying to:
  • tweaking more than 1 element in the genome at a time
  • trying to play 1 entire game with a single strategy and not changing it every move
  • adding some sort of dynamicity, e.g. by making the strategy self-adapting