Making a Checkers game for an academic project and struggling to produce effective evaluation methods to push certain scenarios. My game's logic appears to work fine and everything functions in terms of two players playing but my AI's moves seem sub-par at best.
I'm just looking for a little insight into what factors I should take into account as a brief overview in order to progress my AI to a more challenging opponent, yet still beatable.
I have tried generic methods such as evaluating piece safety, promotion distances, assigning values to certain board positions, counting pieces, etc.
With this I expected a somewhat sophisticated AI with a hard-to-guess move pattern and resembling somewhat of a novice/intermediate Draughts player.
However, from testing, the AI appears to move one piece at a time until it no longer can, then just moves on to the next piece it can move, which in reality is a very simple and counterable sequence of moves.