“Dosage was created in the early part of the 20th century by Lt. Col. J. J. Vuillier, a retired French military officer and pedigree authority who was an active participant in the pedigree debates of his day. One of the more contested issues at the time among racing devotees focused on the relative merits of two of the three stallions from which all of today’s Thoroughbreds descend, Eclipse and Herod. Vuillier contributed to a resolution through an in-depth analysis of the pedigrees of major European winners, classic and otherwise. To the delight of one side and to the dismay of the other, he found that Herod was the dominant influence of the two. More importantly, he determined that the proportion of Herod’s influence in extended pedigrees through 12 generations was essentially the same in all of the horses analyzed. This was indeed a remarkable discovery, laying the groundwork for further research and development of the Dosage system, eventually published in Les Croisements Rationnels (Rational Crossbreeding).”
The quote above is from Racehorse Breeding Theories, page 191. Vuillier had indeed made a remarkable discovery, but he then proceeded to misinterpret it totally. Vuillier and his followers concluded that the best way to breed a good racehorse was to duplicate the successful pedigrees of the past. That in itself was not a bad idea. Specifically, though, they tried to duplicate the “dosages” of the successful racehorses of the past. That was where they made their first mistake, and the one mistake that rendered their entire theory useless in terms of attempting to breed better racehorses.
Evidently it did not even OCCUR to Vuillier and his followers that if they had studied the pedigrees of the least successful racehorses (instead of the most successful racehorses), the results just might have been exactly the same. In other words, they had no control group. They studied the pedigrees of the BEST racehorses only and leaped to erroneous conclusions from those studies. This problem continues to the present day with the modern version of dosage.
In retrospect this was shockingly stupid. But in the context of the times (about a century ago), it was somewhat understandable. It was somewhat understandable because a century ago it was much more difficult to compile pedigrees.
Have you ever tried to compile a pedigree by hand using only available stud books and other reference sources????? I have, and it is not easy. It takes a lot of time. It takes a lot of time to do just a four-cross pedigree. Vuillier was dealing with 12-cross pedigrees.
As recently as 30 years ago the major racing publications were still compiling pedigrees by hand. I know because I worked there, and I spent a lot of time in the research department. It is only within the past 30 years or so that the whole process has been computerized and hence greatly simplified.
So Vuillier and his followers had a good excuse for studying the pedigrees of the BEST horses ONLY. The process was so labor intensive that they did not want to “waste time” on the pedigrees of a control group. But if they had utilized a control group and compared it to the best group, they might have realized that there were ZERO differences between the two groups and dropped that line of study completely. Alas, it did not happen that way.
The modern version of dosage has the exact same problem. All of its statistical “verfication” is based on stakes winners (best horses) ONLY. There is NO control group. Vuillier and his followers had some excuse for this unscientific behavior. Modern dosage has ZERO excuse for this unscientific behavior. With computers now, it is just as easy to compile and study the pedigrees of average nags as it is for good nags. Dosage does NOT do so and arrogantly claims that this does NOT constitute a problem.
The project I have been working on focuses on dosage total points in pedigrees. To review, the minimum is zero, the maximum is 64, the average is a little over 20, and the median is either 18 or 20 (I will tell which more definitively when I am done data mining).
I have examined all 2,413 (at last count) stakes winners among this group of sales foals of 2003-2007. I could give you the entire distribution of those stakes winners by dosage total points, but instead I will summarize the results. The 18 group has the most stakes winners, 270, followed by the 16 group with 232, the 14 group with 213, and the 20 group with 206. Those are the only groups with more than 200 stakes winners apiece. The numbers decrease from the 14 group down to the zero group and from the 20 group to the 40+ group.
So does that mean that the 18 group is best???? Does that mean that the 14-16-18-20 groups are collectively best????? You might conclude so if you have studied BEST HORSES ONLY. And a lot of “research” uses this garbage method of “best horses only,” including dosage.
The missing factor here is how many FOALS are in each group. If you IGNORE that factor (as a lot of garbage “research” does), you will almost certainly come to erroneous conclusions.
Let us think for a minute about these implications for dosage, which bases its conclusions on stakes winners (best horses) ONLY. If dosage proponents were honest, they would tell you something like this:
“Dosage is based on stakes winners only. Therefore, any prediction we make about any nag is valid ONLY if that nag becomes a stakes winner. About 3% of the breed becomes stakes winners. Therefore, by definition, dosage does NOT apply to about 97% of the breed.
So basically we can tell you what type of stakes winner you might expect to have if your nag does indeed become a stakes winner. And, oh, yeah, maybe about a third of all stakes winners win only one stakes in their lives. So dosage does not have much to say about these nags. The prediction becomes valid only after the nag becomes a stakes winner. If the nag does not win more than one stakes, then the prediction refers to a nonevent (stakes won after the first one).”
If dosage proponents were honest, this is what they would tell you. Have you ever heard this analysis before????? Probably not, and certainly not from dosage proponents.
From which I conclude that dosage proponents are essentially dishonest by nature. They are trying to pull the wool over your eyes. Or maybe they are just too stupid to realize the internal contradictions of their own creed. Six of one, half a dozen of the other.