They call it "the numbers game," and it is the universal spoiler in the world of sleddog breeds. Its effects are far-reaching, affecting issues that have nothing much to do with dogsled racing and people who are not involved in that sport. Dr. Roland Lombard defined it publicly for fanciers of the Siberian Husky breed at a talk he gave in 1982 to the Garden State Siberian Husky Club. He described the kind of breeding and selection processes that produce teams that win the Fairbanks Open North American Championship and the Anchorage Fur Rendezvous races and stated that a single top Alaskan 16-dog team might well be selected from over 500 fine dogs, and that the puppy base of those five hundred candidates would probably be over 1500 animals, and stated "it is a numbers game." On another occasion, Yellowknife mid-distance racer Grant Beck stated that he had to produce over 200 pups to get 5 or 6 replacement team dogs that would "make the cut" for his first-string team.
I call this "culling to the curve," or simply "the numbers game," in order to have a handy label for the concept. It is simply the implementation of the statistics of standard distribution of a variable. The basic idea is that you take advantage of the extreme upper toe-end of the Gaussian Standard Distribution Curve (the "bell curve"), selecting those animals that in terms of speed come out more than two standard deviations above the mean -- in practice, the top two or three percent. Fortunately for ONAC racers you don't need to have an education in statistics to accomplish this. You just run the candidates in a fast team to see which ones can keep up the pace; you keep the "best" and blow away the rest (while claiming that you find nice pet homes for all 197 of this year's culls, if you bred 200 pups like Grant Beck).
Now you might think that this procedure would make those three dogs very expensive animals, and if you could actually buy them, you'd probably be right. But you can't buy them. The breeder keeps those for his own team as they are the goal of the whole tiresome, cruel business. Since the other 197 can't make the cut, THEY are worth very little. Hence the unrealistically low perceptions of sleddog prices and values discussed in the preceding post!
Lest anyone fail to get the point, I'll summarise in words of one syllable. Any dumb shit can breed two hundred pups and pick out the best two to five of them, and can multiply the process or repeat it to produce an entire team of ten, twelve, sixteen or twenty dogs. And call himself a "pro racer." And many dumb shits do just that, encouraged by such books as Winning Strategies for Distance Mushers by Joe Runyan, "Iditarod, Yukon Quest, Alpirod Champion." Many slightly sharper shits will happily tell you that they don't do that, that they are such skilled breeders and trainers that they only need produce two or three stellar litters to create an entire winning team. If you believe that one, then they'll tell you another one. More likely that instead of doing all that breeding, they prefer to try to buy their team dogs from others, on the somewhat dubious premise that a dog that isn't fast enough for the breeder may be fast enough for them. Of course, with 15 to 25 elite-level competitors for these super-sprint events... duhhhh... the only reason that works at all is that ONAC and Rondy aren't the only sprint races around.
A great deal more might be said about the numbers game, but not tonight. It's already past midnight.
Posted by ditkoofseppala at July 10, 2010 10:11 PM