agrawal-rohit / tic-tac-toe-ai-bots Goto Github PK

from where i can download this file, trained_state _values_0.txt

Hi

I was surprised to see that Minimax loose from RL.
So I checked your code. And I found that the return value (type) of getBestMove(state, player) is not (consistently) correct.
The function returns a "best move" ( return best_move ), but in the beginning of the function it may return a "-1", "1", or "0", which is the best score (instead of a move).
Hopefully you can fix the issue with this information.

Update: I fixed your code, and now Minimax is even better than RL, and it never looses from a human player :)
Please check it out. The trick is to return both the score and the move as a tuple in the function getBestMove.

Regards

Marco Brassé
The Netherlands

UPDATED CODE (with also some small edits of "==" instead of "is")
See
HumanvsAI_Minimax.zip

its showing , invalid syntax at line 211.could you please resolve this issue

how are you getting those values in .txt values. are they random or is there any procedure to get them