Reexamining our bias against heuristics