[RC] rider weight studies, part 2 - Susan E. Garlinghouse, DVM
>If I
were looking at this type of result I would further look to see if there was an
>underlying selection factor. For example 1100 lb horses may be more likely
to be >selected by heavyweights than featherweights etc.
Possibly,
but are you hypothesizing that the total mass conclusion is skewed, or the
rider-weight-ration conclusions, or? I might be able to clear it up,
since we did consider some of those points, as much as we were able.
>Please remember, I
am not formally statistically trained. I did however use statistics >to
guide my research for 30 years. During this time I ran into a whole bunch of
>correlations that were not good at predicting anything.
Fair
enough. I think by the time everything was analyzed to death, these
results are pretty good at predicting *Tevis* performance, but you have
to be careful not to try to extrapolate every finding to every other endurance
ride---possibly ANY other endurance rides. The last time I was at Tevis
collecting data for a different study, I also collected rider weight, horse
weight, condition scores and a couple other parameters, and just for fun,
predicted who would finish and who would not, based solely on the predictive
values derived from the above data. I was right for 88% of the horses on
my list. If you removed the horses that had a “bad luck” pull
(ie, sick rider, horse spooked at a snake and impaled themselves on a stick,
that sort of thing), then I was right 94% of the time. Does that mean
it’s all written in stone, hell no. One horse that I swore wouldn’t
make it past Robinson’s won the ride. Several others that I thought
looked great on paper pooped out well before the finish. You can NEVER
statistically remove all the infinite variables involved in endurance riding.
That’s the beauty of endurance, eh?
>Finding
the real >cause and effect is much harder with biological systems and even
worse when actual >controlled experiments can not be done. Figuring out what
really happens on the endurance trail is about as hard as it comes.
Tell me
about it. <g>
Okay,
someone else commented there was no mention made of effects during a fifty mile
endurance ride. That’s because you can fake it and get away with
more at a fifty mile ride than you can at a 100, so if you’re trying to
extract clearer conclusions, you try to do it at a 100. It’s sort
of the statistical version of “let the trail sort it out”.
However, you can still extract some 50-mile predictions from the 100-mile data
since, after all, in order to ride 100 miles, you first have to ride 50.
The most common conclusion, however, is going to be, “gee, you can sure
get away with a lot at a 50 mile ride that you couldn’t at a 100.”
Heidi also
thought there was a study that concluded completion rate drops significantly
when horses carry more than 23% of their own body weight. There might be *opinions*
out there stating that, but there are no peer-reviewed studies that draw that
conclusion. Believe me, I’ve read every last study on the subject
published everywhere in the free world (it’s required for the thesis
defense for the degree). Doesn’t exist.
Hope this
clears a few things up. Probably not. <g>
