Mark Darby wrote:
> Hi Team!
Hi again!
> It is apperent to me that those whom have responded
> to my 2 year results compilation didnt actually read
> my "RAW" dat but read the "Bastardized" version
> created by another member of Team.Net.
Actually, I read both and found them both to be without statistical merit.
Extreme values (fastest times and the like) are very volatile, especially
given the number of uncontrolled variables (driver, different run groups,
course design, etc.). I didn't see anybody quote a variance, a t-statistic,
an alpha risk.
I'm not trying to be a smart-ass here. There are reasons why statisticians
insist on these things. If one is allowed to make a case based on small
sample averages, one can prove just about anything. Lies, damn lies, and
statistics is the saying.
<snip>
> Comparing these
> times creates an Apples vs Oranges situation,since the
> National Series in Pro Solo for 1999 was intended to
> create a "Junior Varsity" and increase attendance at
> Pro Solos.
That's one way to look at it. I think some would debate that. I won't.
Here's another way to slice the data. I'll only use NT results. Roll all the
DS Neons into GS and look at the trophies.
Meridian: 17 Drivers, 6 Trophies
Eclipse, Eclipse, Neon, Neon, Neon, Neon
San Diego: 15 Drivers, 5 Trophies
Type R, Type R, 325, Prelude, Probe
Fort Worth: 22 Drivers, 8 Trophies
Type R, Impreza, Camaro, Type R, Neon, Talon, Probe, 328
Ayer: 21 Drivers, 7 Trophies
Neon, Neon, Neon, Corrado, Type R, Neon, Sunfire
Peru: 22 Drivers, 8 Trophies
Prelude, Corrado, Eclipse, Neon, Neon, Type R, Neon, Eclipse
Bremerton: 17 Drivers, 6 Trophies
Type R, Neon, Neon, TT, Neon, Shelby, Neon
All in all that's Neons taking 14 of 40 trophies and one overall win. I'm
not sure how one concludes from these results that the Neon isn't
competitive. Certainly not a class killer, but not getting killed, either.
Perhaps it's time the Neon drivers woke up to the fact that not every car
gets its own class.
> My "RAW" data
> concluded that,for 1999, the G Stock cars were an
> average of .617 of a second faster. For 2000, the data
> shows a difference of 2.178 sec that the G Stock cars
> enjoy.
No, what your data showed was the margin between the *fastest* G-stock car
and the *fastest* Neon. That's a total of 6 data points. Hardly enough for a
conclusive sample, especially given the variance. Projecting this onto the
entire Neon/GS population is dubious at best. When a larger sample is
considered, the Neon looks a whole lot better.
> The fact that the STS/R rules are derived by
> combinning the stock class rules and the street
> prepared rules,how does one come to the conclusion
> that STS/R classes are not a consideration when
> forming new stock classes since the basic rules are
> the same.
You're mixing class and category (easy enough to do with a hybrid
category/class like ST). STS/R (and SM for that matter) are derived from the
*Category* allowance for Stock and SP. There is absolutely NO mention of
*Class* anywhere in the ST rules except to distinguish STS from STR.
Eric Buckley
7STR: 98 Integra GSR
St Louis Region
|