jsd comments on Some heuristics I use for deciding how much I trust scientific results