I cannot understand several things in the Monte-Carlo method: What to do at the very beginning, when there is only one node? What fields should the node children have? (I think that the Turn, the number of victories, the number of games) First, two bots have empty trees. They begin to make random moves, thereby complementing the tree. Until what point should this go on? Or does he always need to make random moves? Can enter the probability to make a random move? Do I have to record the opponent's move to the tree?
Objective: to teach bots to play tic-tac-toe using the Monte Carlo method. (You can, of course, sort through all 9! Options, which is not much, but I need this to understand the algorithm)
