It’s more interesting to play Markov chain mode, the algorithm learns to predict the player, and in the end it becomes very difficult to defeat him.
If only there wasn’t an error in the algorithm that I didn’t have time to fix. (ノ_<。)
It has been fixed in the new version)