For Othello, picking the positionally best square (corners very good, adjacent to corners very bad, 2 squares from corner good, etc.) beats lookahead, likely be

For Othello, picking the positionally best square (corners very good, adjacent to corners very bad, 2 squares from corner good, etc.) beats lookahead, likely because having unflippable pieces in good positions leads to large swings in score in the endgame that a short lookahead won't see. I play almost purely positionally with no conscious lookahead and I beat it soundly every time, playing as both black and white, frequently with the AI having to pass because it had no legal moves. A "don't yield a corner" heuristic to never move in the three squares around a corner unless there are no other legal moves would likely significantly strengthen its play.