In a message dated 7/11/00 12:41:56 PM Eastern Daylight Time,
bthatch@juno.com writes:
> I guess you took too quick a look at my post. There are BIG changes
> proposed in there. Shrinking the number of classes in Stock, Street
> Touring and Street Prepared from 34 to 16 by putting ladies into the next
> lower class. Putting all 3 categories (S, ST and SP) in the same
> continuum isn't simply sorting by class PAX. What this does is set up the
> template by which the SEB could start moving cars up and down classes
> in the future.
You're going to move cars not just between classes but essentially from one
category to another, thus from one set of preparation rules to another? So
then you have cars prepared to different sets of rules running in the same
class? Let's see, people complain now that newcomers to the sport don't
understand the current class structure or the preparation rules...this will
help?
> Going in and moving cars around from the current class structure would be
> a waste of time. Any kind of adjustments like that should be done on a
> committee basis. The SEB is very well suited for that task.
>
> My main concern was to get the classing to be ranked fastest to slowest.
> Now, if class bumping is necessary in the local region or at a Pro, it
> isn't complicated. Ladies know exactly which class they have to compete
> in but they know they will be pitted against everyone in that class and
> won't just have the opportunity to win a meaningless trophy because they
> didn't have any competition.
I don't understand why ranking the_classes_from fastest to slowest is the
highest priority, except for the fact that it lets you move cars from one
class
to another based purely on their performance against an arbitrary index. The
SEB attempts to class cars by type whenever possible, primarily to minimize
course dependencies over the variety of sites and courses around the country.
If your system eventually gets Miatas and Camaros in the same class as a
result of their times at Topeka, for instance, would they still be a good
match
somewhere else in the country in a postage-stamp parking lot?
The fallacy in using the PAX index for classing cars is that the index is
weighted toward Nationals results and consequently toward a particular type
of course design. The further you get away from the size, type, and relative
speed of those courses, the less meaningful the index becomes. Similarly,
comparing times between cars/drivers at a single event by using the index
alone completely ignores course issues that may have favored one car over
another, not to mention the fact that you're_assuming_that each of the
drivers in question drove equally well. Some have compared relative PAX
numbers for the same cars/drivers at different events and said we should
make classing decisions based on that data. So in addition to ignoring any
course design issues, we should also assume that all of the drivers drove at
the exact same proficiency at all the events being compared? Does anyone
ever feel they drove better at one event than another? How do you adjust
the numbers for that fact?
GH
|