Before this goes sideways:

I was not and am not endorsing use of Dosage for handicapping, and I'm certainly, definitely, not advocating using it for specific horses.

There's a difference between large population and small population studies (which fortunately for me nobody in Ragozin's office, including Ragozin, ever understood). It involves mostly when it's right to use averages, and when it's not.

My assumption, since the Dosage guys make their decisions based on the results of races, is that the Dosage stuff has some correlation with generally accepted analysis of pedigrees being "sprinty" or "distance", ON AVERAGE. That makes them of some use in the type of large population study the guys did here, since if the Dosage guys get some wrong it will wash out, plus or minus, given the large numbers.

That's a whole different thing than thinking specific ratings they give out are accurate. Even if you accept their premise, there are a whole lot of things wrong with the methodology used to come up with Dosage-- I once did a segment on Post Time about this. I don't have time now, but after the seminar is up next week maybe I'll go into it, if anyone cares.