Evaluation functions
	n For chess, typically linear weighted sum of features Eval(s) = w1 f1(s) + w2 f2(s) + … + wn fn(s) n e.g., w1 = 9 with    f1(s) = (number of white queens) –  (number of black queens), etc. n Caveat: assumes independence of the features n Bishops in chess better at endgame n Unmoved king and rook needed for castling n Should model the expected utility value states with the same feature values lead to.