At 10:12 AM 3/16/99 -0800, Craig Blome wrote:
>I'd tend to agree with Jay here. Looking at the performance of the
>folks who trophy at Topeka, who are the statistical outliers of Solo
>II _drivers_, is IMHO a less than reliable gauge of the potential of
>the _cars_.
Certainly. All I was trying to say was that it doesn't necessarily take
the newest model/make of car to be competitive. I'd even go farther
and say (as I did later) that some drivers would be fast even if driving
a shoebox.
>If you got a population of drivers big enough to draw
>statistical inferences from, and could figure out how to correct for
>weather, course design, etc. then you might be able to conclude
>something. But anyone willing to crunch those numbers is a better man
>than I am (and I'm including the ladies in that statement! :)
Even if you could control for a variety of things such as you mention,
you'd still have the selection effect to deal with (as somebody pointed out to
me AFTER I analysed ALL the data from ALL the National Tours last year).
The ideal way to do it would be to have all drivers drive all cars on all
courses
in a counterbalanced order--and not play favorites with their driving AND be
equally familiar or unfamiliar with the cars. Anybody up for that study? And
whose buying the tires? (Of course, each driver would need fresh ones--I'd be
happy to take the leftovers--especially the 225/50/14's.)
We tried something similar last year in our region with four cars (CS, ES,
FS, & HS)
and four drivers. It was a blast, but we still couldn't balance out the
fact that each
driver was most familiar with his own car, tended to be fastest in his own
car, and
next fastest in cars more similar to his. The other finding was (no
surprise) that
one driver was fastest in all the cars--by quite a bit.
mlc
'91 MR2 NA
'91 GMC Safari tow van
'81 Ford F250 tow truck
'XX driver ( :
|