California Stallions

Oregon Stallions

Washington Stallions

Bloodstock Agents & Consultants
Farms

Handicapping/Wagering
Organizations
Racehorse Conditioning
& Rehabilitation

Racing Media
Training Centers

Search this website:

Dosage and Its Fatal Flaws

This article appeared in The T I A Newsletter, September, 1997.

     Secretariat has been named a chef-de-race.
     Does anyone care?
     Well, yes. The Daily Racing Form devoted two-thirds of a page to the news, in the form of a "Bloodlines" column by Ed Fontaine. So there must be some handful of people out there who are still laboring to work out dosage calculations, all the way to the second decimal place.
     It's no secret that breeders will wander far into the outer regions of the occult in their search for the Magic Bullet of Mating (note the great popularity of nicking), but this dosage thing is out on the other side of that.
     As it is presently practiced, dosage was developed by Dr. Steven A. Roman, Ph. D., of Richmond, Tex., who otherwise appears to be a sensible and reasonable man.
     The heart of the dosage business is the roster of chefs-de-race, the 139 sires whose distance-siring tendencies provide the foundation of the program. The Dosage Index and Center of Distribution are calculated from the mix of chefs in the first four generations of a horse's pedigree.
     Roman maintains a list of chefs and periodically adds a new name, whenever he has  determined to his satisfaction the distance tendencies of the descendants of an influential sire. The goal of dosage, Roman points out, is not to determine how good a runner a horse will be but to determine his best distance.
     Thus, if dosage works and you get the numbers right, you could wind up with a runner that gets the desired distance, but slowly. (Of course, any horse can get a classic distance if he runs slowly enough.)
     The best use of dosage, in other words, is to check the mating you've already chosen on the basis of quality to make sure you don't breed a four-furlong flash or a two-mile plodder. 
     Fair enough, if you accept Roman's calculation of the probable best distances for the descendants of those chefs de race.
     You can believe that Roman has worked at it. He's a scientist, and he doesn't install a new chef until he's satisfied that the sire stamps his descendents' stamina in a consistent pattern.
     Having made that calculation, Roman then places the chef in one of five categories ranging from "brilliant" (sprinters) to "professional" (routers). Sometimes Roman assigns a chef to two categories. Such is the case with Secretariat, who is classified as both "intermediate" and "classic."
     A great deal depends upon Roman's ability to make accurate category assignments. Since the final calculation is carried out to two decimal places, there's no room for guesswork or estimates.
     So far so good, maybe.
     However, the next step--the first in the actual calculation--is more likely to produce hysterical laughter than respect.
     As in, "You do what?"
     What you do next is ignore (a) all the females and (b) all the non-chef sires in the first four generations of the horse being dosed. That instantly eliminates a minimum of 50 percent of the horses contributing to the subject horse. Since most horses' pedigrees aren't jammed with chefs, it also eliminates a fair number of the sires in the pedigree.
     For example, a dosage calculation for Bertrando, an Eclipse Award winner now at stud in California, would consider just five of the 30 horses in the first four generations of his pedigree. That's 16.7 percent of Bertrando's pedigree. That other 83.3 percent is just ignored.
     On the other hand, a calculation for the royally-bred A. P. Indy would consider 11 of the 30, consigning only 19 (63.3 percent) to a state of non-influence.
     Point values are assigned to the chefs in the pedigree--more to the close-up sires--and the result is calculated to the second decimal place. 
     How can anyone take this seriously?

To The Thoroughbred Showcase of the West home page.