Daniel Kahnaman and Amos Tversky (two psychologists) have collected tons of experimental evidence showing that people struggle to assess probabilities and even expressing preference consistently. Gerd Gigerenzer and his academic allies have more recently identified some kind of decision where the simple rules of thumb deliver consistently better results than analytic style methods.
UC Berkeley economist David Romer found in 2006 that football teams go for points on their fourth down much less often then they would if they didn’t