[kotach]

AlphaGo has a learned evaluation function for each move.

Evaluation function exists but it is not as simple as it can be for chess.