I made a game simulator in python with a tcl/tk GUI. My AI can be set to 3 levels: offense, medium, defense.
I calculate the best possible move for the level (o. always places ringer, m. does all, and d. places/removes spoilers if it is worth it)
So if you play the game with a computer, you get nice results. But if you play CPU vs. CPU, and they are not both offense, they always end up with a 5-or-so chain and one CPU places a spoiler, the other removes, CPU places it... I still have to fix that
