The basic idea is that if you use pars (as Beyer and Ragozin did for a very long time and may still), you can\'t compare horses from different generations. If you start out by saying a winning 10 claimer will average 7 (or 80 or whatever), and use pars to bring horses to that average, at the end of the year (or decade) they will average 7. And your other claiming levels will do the same.
That means two things. The first is, you won\'t be able to tell if as a group race horses are getting better (or worse, theoretically), which makes your figures useless for generational comparisons. The second is, horses will receive huge figures if they are far superior to the others in THEIR generation.
Andy and I started to get into this at the 2004 DRF Expo, but didn\'t get to finish it. He was saying that Secretariat etc. were obviously better than the horses today, but he was basing that on dominance of their peers, without knowing it.