Did someone finish this project?

Im making this project "Games of the General with AI"

Im stuck on making the AI..

Thanks!

Did someone finish this project?

Im making this project "Games of the General with AI"

Im stuck on making the AI..

Thanks!

I got something working. This is how my system (that doesn't use bayes) ended up playing:
1. Opening; plays very strong. It's able to bluff (its an emergent behaviour) and does a great job at determine piece strenghts. It worked particularly well at isolating spies. It was good a positioning and in laying traps.

2. Mid game; weak. Plays very well defensively, but doesn't push ahead. This gives the opponent time too much time to maneuver. Gains made in the opening get lost.

3. End game; very bad. So bad it's embarassing. There usually isn't enough information gathered to even remotely determine what types of pieces are left on the board.

Most of the logical deduction code I wrote was a waste of time; the simple dedection worked quite well, but all of the complex stuff almost never got utilizied in-game!



If I were to do it over again I'd try the bayes approach just to see how it does, and focus on getting the gametree to look as far ahead as possible. I'd also add in a model of the opponents knowledge of the computers pieces; it's not something I did, and the AI suffered for it.

Thank you willh. Im having a hard time writing my AI for this Game.. Do you have any guides where i could start? And what is the best AI to implement this game?