Michael-- it\'s chaos around here, but briefly--
I\'m not saying that all GIs are faster than all GIIIs, or anything like that. I make part of my living finding the anamolies, like Super Frolic, Santana Strings, etc.-- I look for where performance and accomplishment have not gone hand in hand, and try and turn the former into the latter.
What I am talking about is something else. I am talking about figures not making sense-- for example, the Adirondack. I agree that a graded stake for fillies at Saratoga can go in an 18 in theory-- but not with those horses running in it. That\'s the point about the Jockey Club Gold Cup, and some of the other examples as well-- the only way to give them those figures is to ignore the past histories of the very horses whose performance you are measuring-- AND IT IS ONLY BY USING THOSE FIGURE HISTORIES THAT ANYONE IS ABLE TO MAKE FIGURES AT ALL. That\'s what projection figure making is all about.
The difference is that Ragozin combines a lot of apples with a lot of oranges, and uses averages to do it. He combined the races before they sealed the track at Saratoga with those afterward, and the early (and one turn) races on the Belmont card with the Gold Cup. It is a dogmatic, fundamentalist approach that ignores both the physical realities (sealing the track, etc) and THE SPECIFIC DATA ITSELF-- the histories of the horses in the race itself. It bases figures on a series of assumptions, relationships, and conclusions that have no basis in science or logic, and lumps all kinds of things together to come up with an average. Averages are fine for large population studies, like coming up with the average winning number for a 25 claimer-- but they don\'t help you come up with the winning number for THIS 25 claimer.
If you saw all the Ragozin sheets for the horses that ran in those late dirt races on 7/27 at Saratoga-- or the Gold Cup, or the two Lion Tamer races-- you would see immediately what I am talking about.
Which is why you won\'t.