Hi Ludovico, I took a look at your code.
A README file would have helped navigate your code and also make clearer your idea; eventually, I got your intentions. Personally, I don't enjoy taking a well-defined strategy randomly: in your code, also, the choice of taking one strategy in the adaptive function depends also on the position of the strategies inside the dictionary.
A minor error is selecting the best inside the same generation, even if you have not generated it entirely: the selection should happen at the end, and the winner used in the next run to bring to life a new whole generation.
The thing I liked the most in your fitness was considering both situations in which the starting player is 0 or 1, to create more equity, and also to create games of random dimensions (maybe, adding some other possible values would've been nicer, but still okay).
Hope it was helpful, I wish you the best for the next labs! :)
Hi Ludovico, I took a look at your code. A README file would have helped navigate your code and also make clearer your idea; eventually, I got your intentions. Personally, I don't enjoy taking a well-defined strategy randomly: in your code, also, the choice of taking one strategy in the adaptive function depends also on the position of the strategies inside the dictionary. A minor error is selecting the best inside the same generation, even if you have not generated it entirely: the selection should happen at the end, and the winner used in the next run to bring to life a new whole generation. The thing I liked the most in your fitness was considering both situations in which the starting player is 0 or 1, to create more equity, and also to create games of random dimensions (maybe, adding some other possible values would've been nicer, but still okay). Hope it was helpful, I wish you the best for the next labs! :)