[Index][Thread]

Kenny fails basic statistics and sampling theory yet again.



In article <404o89$81s@news-e1a.megaweb.com>, kennystar@megaweb.com 
(KENNY MORSE) writes:
> In article <404cou$l46@clark.net> Kevin Atkinson wrote:
> 
>>They DO show what really happing.  Ok if a LA freeway speed limit went up
>>to 85 mph (well above the 85th percentile speed on almost any road) do you 
>>honesly think most people will do 95-100? (Challenge ONE)
> 
> Nope. I don't. Because Americans are NOT trained to be able to
> handle these speeds safely. A small % WOULD go 95-100.  That's
> not the point.  There will be NO 85 speed limit in yours or my lifetime.
> The issue is if raised to 65 or 75, will they exceed.  I say ABSOLUTELY.

Sure, a few will exceed any posted limit.  But how many?  

>>(Challenge TWO) Ask you students if the limit was raised 10 to 15 mph if 
>>they would exceed it automatically.
> 
> Kev...I DID ask. In TWO classes now (since Jaime asked me to) and
> the answer was "probably".  And there it is.  

That wasn't the question I asked you to ask.  I asked you to ask them,
approximately, "what if the posted limit was above the max speed at
which you feel comfortable on that same road today?  Would you STILL 
drive 10 or 15 mph over the limit?"  That's a different question 
entirely. 

> These are average drivers
> who absolutey reflect a cross section of California drivers who were
      ^^^^^^^^^

This is called "proof by vigorous assertion".  You'll have to do 
better than that.  

> the ones who got caught.  If there were more cops, more would be
> caught.

Two problems here.  First, I gave you a list of at least five 
different groups of drivers who will be under-represented in your 
classes.  If you gave a rebuttal to that, I didn't see it.  

Second, the ones who are "caught" more often (and hence are more
likely to choose to go to traffic school) tend to be the more
dangerous drivers... the ones whose behavior makes them stand out on
the road, even if they're not speeding.  I drive faster than posted
limits all the time... when it's safe to do so... (and well below
posted limits, when that's the safest behavior)... but I don't drive
"conspicuously", so I don't get tickets any more. There are many
drivers out there like me.  We're underrepresented in your classes 
too. 

(It's not because no one sees me, either.  Last weekend I was driving
south on I-5, doing 75-80 in the second-from-the-right car lane. 
About a mile up the hill, I noticed a Kern County Sheriff's car
peeking out of my right rear blind spot. I maintained speed.  After a
few moments he pulled left two lanes and cruised on past.) 

In other words, your classes are biased toward those who get tickets 
more often than most.  Of COURSE many of them would say "Sure, I'd 
exceed a posted limit of 65 or 70".  That's why they're there in your 
class in the first place!

(I'll grant you that you have a good knowledge of traffic laws.  Your 
knowledge of sampling theory, of the ways that bias can get into a 
sample, and of logic, however, leave much to be desired.)

Re 85:  Kenny, we *have* a _de facto_ limit of 85 today -- on I-5 from
the 99 junction (just south of Bakersfield) to, roughly the 152
junction.  Slow drivers do 65-70 on that road, fast drivers do 75-80.
It isn't that there's no enforcement, either; drivers doing 90
get ticketed -- I've yet to be passed by anyone doing much more than
85 on that road and NOT see them pulled over a few or even tens of
miles later.  It works; the accident rate per vehicle-mile on that
road is very low. 

As I see it, the real problem with posted limits that reflect the
actual maximum safe speed under ideal conditions is indeed one of
driver education -- but not for those conditions; they do just fine
under ideal conditions.  I think that what we need is better awareness
of when the posted speed is *too high*.  There are very few CHP
officers who will ticket you for 65 or even 70 on the 210 when the
road is almost deserted, but the posted limit of 55 lets them easily,
without an argument, write you up for 65 or even 60 when you are
obviously unsafe at that speed (due to heavy traffic, rain, whatever).
The trouble is that the posted limit of 55 also lets them write you
for doing 57 under conditions that are safe for 70. 

	--- Jamie Hanrahan, Kernel Mode Systems, San Diego CA
Internet: jeh@cmkrnl.com (JH645)  CompuServe: 74140,2055