This piece about estimating in IT, and specifically software development, discusses the reasons why we tend to ignore the fruits of experience and produce estimates that are over-confident and over-optimistic.