// Tic tac toe game the initial version #random_tictactoe been implemented to encourage any possible winning in the next move avoid losing there. {it becomes still beatable when there comes two positions for the losing situation}
the next version #normal_tictactoe is a more enhanced version of the random version
-
weights allocation : instead of choosing in random from all the empty places we will find the preferable empty places and call it the best empty places
-
Weights allocation {reverse biasing} : this update in weights leads to prefer the places where even oponent would want to land in the next move {from our best_empty places} calling them the best places
-
lookout weights : this leads to one step ahead biasing { from best places we will shortlist those which will minimize the max(weights) allocated which will avoid possible winning in next to next move}
the pattern building is faster and effecient than any brute force method/
Reviewed and tested the patterns for bot's moves , Penalty of 30ms while reducing the loss of the beginner by 100%.
optimal_version :
