Evaluation func in a game

Started by
10 comments, last by darkawakenings 20 years, 6 months ago
Ok, players do move simultaeneously, to answer your question. and in practice, player 1 has the advantage. Because its an odd by odd board of course, there is a center square. If player 1 takes this, he usually has a good game. Unless, (lately) if player two moves to the corner. I think that this anti splitting up the board technique forces player 1 to the split up the board more on his next move and gives player 2 easy analysis of the position, identifying safe squares and such.

I dont know if you were thinking that player 2 has the advantage because they can move symmetrically, but thats not the case.

A Mathematical game developed and played excessively by friends and I
Advertisement
A 5 x 5 board isn''t too big that you couldn''t try this:

take a "snap shot" of every board position during the game (one after each turn). Save these snap shots after every game along with an evalutation. If the AI won the match, have it increase the evaluation of the snap shot. If it lost, have it decrease the evaluation.

This provides a weak kind of "learning" for the AI...

Jason

This topic is closed to new replies.

Advertisement