I think over estimated are the key words. I'm sure we've all been around long enough to see leaders win/lose, backmarkers win/lose, both in fast/slow races, inside/outside barriers win/lose, odds on favs win/lose, 100/1 shots win/lose, etc etc I'm sure someone has the time to do the barrier study that I suggested, the reult being immaterial to me as I don't do the form, just bet to figures.
Good Punting.