Can the use of statistics correct the errors in expert judgment?In the previous introduction, I described how Michael Lewis, ...