Roman\'s article, which I scanned quickly, seems to make its case based on anecdotal evidence (the opinion of 7 \"experts\", who pick the best horses of all time using criteria of their own choosing), and as such is of no use in any reality-based serious comparison.
I certainly agree with you about the flaws in \"Dosage\", and then some. I don\'t remember any in-depth discussion about this here before, but I could be wrong.
1-- The underlying premise of Dosage, paraphrasing, is this: a small number of horses (stallions) are responsible for \"shaping\" the breed, so they are the only ones worth taking into account when looking at a horse\'s pedigree. The premise is debatable on its own, the conclusion ridiculous. Every stallion in a pedigree contributes, and certainly a sire who is not a chef-de-race (more francophilia) has more genetic impact than a horse 3 generations back who is.
2-- As I said once on Post Time, most of those of us who had mothers would agree they had something to do with how we turned out. Dams are not taken into account in Dosage, and a geneticist will tell you they have impact equal to that of the sires. I could show you that when it comes to horse racing they actually have a lot more impact, but that\'s another conversation.
On that same Post Time we showed a 4 generation pedigree, then showed it again leaving only the horses who were measured for Dosage. There were only a couple of names left out of the 14 in the 3 previous generations.
3-- The choice of which stallions to use is subjective, as is the assignment of which category they belong in.
There is nothing wrong conceptually with the idea of coming up with something that gives you a \"genetic profile\" of a horse, as long as you realize it will only give you a general idea (even if excecuted perfectly it is in effect a broad average of what a lot of identical matings would produce, not a blueprint for an individual horse), and as long as the way you come up with the \"Dosage\" (for lack of a better word) is driven by accurate data, and measured correctly. Meaning, all the horses in a pedigree, in terms of the characteristics they pass on-- speed, ability to go long, etc.
Well, guess who is in position to do this.
I use pedigree to look for tendencies when advising on purchases of horses that have already run, and we took our first steps down the road of quantifying things with the Sire Profiles, which as far as I know were groundbreaking, followed by the Dam profiles. We\'ve been kicking the idea of a Thoro-Graph Dosage around for quite a while, but we might finally get to do it in the next few months-- the company is rolling along pretty well now, I am starting to have a little more free time, and it is just one of the innovations on the way. I will be interested in suggestions from the usual suspects (and others) on how to do the Dosage.
In the next few weeks we\'ll be rolling out a few things. First will be the figure based trainer profiles, which we sneak previewed on BC day. Second, we FINALLY got the rebate thing done-- we will be sending out an e-mail soon, and anyone interested should make sure we have a live e-mail address for them. At some point, hopefully before the winter is over, we also hope to finish off a very big and groundbreaking project-- a study of figure patterns that will enable TG players to use a look up table to see how, for example, a Spring 3 year old does off a pair of tops followed by an off race (%of new tops, return to top, etc., same as the trainer studies).
That oughtta keep everyone (including our programmer) busy for a while.
