Asking about using Sustained Peaks vs Absolute Peaks and
Adjusted Ratings vs Unadjusted Ratings, Arthur Emerson
asks...
Arthur Emerson wrote:
> Which combo can
> we all agree on gives the best representation in ratings?????
Most drivers use Sustained Peak, and Adjusted Ratings. And
we spend the bulk of our time calibrating the performance
ratings using known drivers and these settings, so in
general I would suggest that these are the best ratings to
use.
Keep in mind that the ratings can easily be course
dependent. It's easier to score high on a short course than
on a long one. It's easier to score high on a course that's
all the same (all sweepers OR all transitions) than on one
that has equal parts of each. But with those caveats in
mind, very few of us score 99's on a regular basis on your
typical 45 - 60 second autocross course using a balance of
features. The shorter and simpler the course, the higher
the score will be. Longer and more difficult courses tend
to be more difficult to score well.
Okay, with all of that said, if you are scoring 99's on
typical to complex courses, of decent length (say 45 seconds
or more), then you are driving very well. There's just no
doubt about it. And at that point, you should then make the
ratings tougher, by switching to Raw Unadjusted Ratings.
As you show, this will lower your ratings considerably,
giving you a lot more room to improve before you hit the
99's again. In fact, while we have a few drivers who
generally run with Raw Unadjusted Ratings to give them
better ratings, I don't think any of them have had trouble
running up against too many 99's once they've made that
change.
In general, then here's the progression:
Sustained Peaks, using 0.5 sustain, Adjusted Ratings:
This is the default value and best for most
autocrossers. In this mode few autocrossers will
score too many 99's to be useful on typical
courses, though most autocrossers will see the
occasional 99. Easy courses and short courses make
it easier to see 99's but they shouldn't make you
get too cocky. This setting is the most carefully
calibrated of our settings, with the values being
set empirically from viewing a variety of drivers
on Nationals type courses. Essentially the
Adjusted Ratings means you are graded on "a curve",
and therefore the ratings refer to the percentile
of driver, rather than the percentage of available
performance. For example, a 90% Usage rating means
that you are a 90th percentile driver among
autocrossers. This is an easier and friendlier
scale than the others.
Sustained Peaks, 0.5 sustain, Unadjusted Ratings:
This mode is the one I recommend for advanced drivers
who are running up against 99's in one or more
categories too often to be useful in the prior mode.
In this mode the only difference is that the ratings
are no longer scaled on a curve, and therefore they
show the particular rating as a percentage of the
car's capabilities. For instance, 90% Usage here
means that your average Usage at each point
around the course is 90% of what GEEZ thinks the car
can do. This is a much more demanding scale than the
above. We have several drivers who run in this
setting most of the time because they get too many
99's with Adjusted Ratings to be useful. But we have
no drivers that I know of who have found this level to
be inadequate for their skill level. Again, typical
national level courses being the expected difficulty.
You can use other settings for different purposes. A few
examples:
Stretch the sustained peaks out to 1 second as a diagnostic
tool to see that the g-levels you are hitting are true
levels and
not the result of transient spikes or bumps. However, this
will make your ratings misleadingly high. Very useful for
vehicle setup and diagnostics, not very useful for ratings.
Using absolute peaks compares your ratings against the
absolute highest tick you recorded in each direction,
without regard to finding whether that peak was caused by a
bump or otherwise. This is more pure, but decidedly less
useful because a transitional, or bump-aided peak is pretty
meaningless to use as your target. It's impossible to score
over 100% at any point of the course on this setting, and
all scores drop a lot. In fact, these scores very quickly
become arbitrary. I wouldn't use Absolute Peaks for much of
anything. Using 1/2 second sustains provides about the right
balance, IMHO.
--Byron Short
888-909-0818
///
/// geez@autox.team.net mailing list
///
|