rec.autos.simulators

'ring time getting there...

Jan Verschuere

'ring time getting there...

by Jan Verschuere » Mon, 13 Aug 2001 07:11:59

It's true some of them (even those ranked lower than you) have one or two
really low ranks and therefore can afford to fluff the other cars somewhat.
And it's true they could have been higher up had they applied themselves
more. The point is: they didn't and they're not.

Regardless, MonsterRank is ranked according to the total for all the chassis
and all the tracks. Therefore it's valid to place even someone who's -100
for the Lotus and +450 in total for the other chassis in your "ability
brackett" as the yardstick is his total of +350. It's also valid to draw
conclusions based on averages on datasets including such extremes as long as
standard deviation is considered.

Also, logic would have it that the cars you've driven the most *should*,
relatively speaking (i.e. compared to an average), also be those with the
"best" chassis ranks.

Time to get some hard data on this, rather than my gut feeling. Americans
love stats, right? -And you're American, so..... <g>

Here's what I did: I made an Excel file listing the Chassis Ranks for all
drivers within about 1 minute of you. As some drivers had apparently
submitted better times without updating their MonsterRank, I extended the
range to 70 seconds on the plus side, so as to not bias it too heavily (and
even that included a guy who'd recently found some 12 seconds). Now, it's
important to remember you're more or less smack in the middle of this
population, i.e. the numbers will show which of your chassis ranks are above
and which are below "average".

                    LOTUS    EAGLE    FERRARI    HONDA    BRABHAM    COOPER
BRM

            Eldred: 19.67    57.94      25.62    59.30      49.50     55.51
99.05

           Average: 21.11    40.41      34.97    73.15      49.70     56.29
  82.56

Standard Deviation: 18.22    14.36      16.30    15.99      15.17     11.60
14.79
    (on population)

Ok, so I was wrong... you're no weaker than average in the Cooper, no better
than average in the Brabham and whereas I thought your BRM chassis rank was
pretty good, your appreciation was more on target.

However, no matter how illogical it may seem to you, the Honda *is* your
best chassis. You're just within "average minus st.dev", so you're not
exceptionally good (*) at it, but certainly way above average for this
population. Meaning it would take an almost superhuman effort on your part
to improve your MonsterRank through lowering your Honda chassis rank. On the
other hand someone of average ability within the population used for this
statistic (i.e. you, you're smack in the middle, remember?), there's 34secs
waiting to be picked up in the Eagle and the BRM. If I were to choose one
over the other, I'd go for the intrinsically faster chassis, wouldn't you?
;-)

Ok, I'm now going to cane myself for being sad enough to work this out and
then I'm gonna apply this analysys to myself!! <g>

Jan.
=---
(*) Rem.: unfortunately, being outside "average plus st.dev" conversely does
mean
          you're exceptionally bad at the Eagle and the BRM (at least
statistic-
          ally <g>).

Eldre

'ring time getting there...

by Eldre » Mon, 13 Aug 2001 14:40:38



>Ok, so I was wrong... you're no weaker than average in the Cooper, no better
>than average in the Brabham and whereas I thought your BRM chassis rank was
>pretty good, your appreciation was more on target.

Hmm...I don't know if that's a good thing or not<g>

Oh MY...too many numbers!  AUGH! :-)

Well, since I don't really understand all the statistics-speak, I'll take your
word for it...  Guess I need to dust off the Eagle again...

Eldred
--
Dale Earnhardt, Sr. R.I.P. 1951-2001
Homepage - http://www.umich.edu/~epickett
F1 hcp. +16.36...Monster +366.59...

Never argue with an idiot.  He brings you down to his level, then beats you
with experience...
Remove SPAM-OFF to reply.

Jan Verschuere

'ring time getting there...

by Jan Verschuere » Tue, 14 Aug 2001 04:28:43

It's a good thing to know what one's weaknesses are. Not a pleasant thing,
but a good thing.

Well, if I had to do it by hand, I'd think twice too, but copying the
numbers off GplRank and have Excel calculate the averages and stuff only
took about 10 minutes. This sort of number crunching is what computers do
best. In fact it's what they were originally designed for!!

A quick search didn't reveal a good web-based explanation and I'm not sure I
can really make it clear for you, but here goes.

Standard deviation is the average deviation of the values in the dataset to
the average of the dataset. This has two important implications: 1) the
majority of values in the set are between "the average minus the standard
deviation" and "the average plus the standard deviation" and, therefore
values outside this range are exceptional. 2) if the population is large
enough (which, to be honest, it isn't in this case) we can be fairly certain
any new value added to the set will be within this range.

Example: suppose Randy Cassidy completed his MonsterRank and it came out at,
say, +357. Based on the data I collected we can then be fairly certain his
Honda chassis rank would be between 57.16 and 89.14 (73.15-15.99 and
73.15+15.99). If not, his Honda rank is exceptional for drivers having a
MonsterRank between +306 and +426.

Indeed. I need to drive the Lotus more... turns out I'm exceptionally bad at
it. ;-((

Jan.
=---


rec.autos.simulators is a usenet newsgroup formed in December, 1993. As this group was always unmoderated there may be some spam or off topic articles included. Some links do point back to racesimcentral.net as we could not validate the original address. Please report any pages that you believe warrant deletion from this archive (include the link in your email). RaceSimCentral.net is in no way responsible and does not endorse any of the content herein.