Prev: If you can't miss a dump truck sitting in the road...
Next: If there is a Japanese God out there...
From: N8N on 9 Apr 2010 13:45
On Apr 9, 1:41 pm, Brent <tetraethylleadREMOVET...(a)yahoo.com> wrote:
> On 2010-04-09, N8N <njna...(a)hotmail.com> wrote:
>
>
>
>
>
> > On Apr 9, 1:28 pm, Brent <tetraethylleadREMOVET...(a)yahoo.com> wrote:
> >> On 2010-04-09, Sir Ray <waterboi5...(a)hotmail.com> wrote:
>
> >> > On Apr 9, 12:42 pm, gpsman <gps...(a)driversmail.com> wrote:
> >> >> On Apr 9, 8:07 am, Dave__67 <spamTHIS...(a)yahoo.com> wrote:
>
> >> >> > On Apr 8, 9:08 pm, gpsman <gps...(a)driversmail.com> wrote:
>
> >> >> > > Mogil said he tested it 15 times with an average of only 3.8 seconds."
>
> >> >> > > "Average" suggests variation and inaccuracy (thanks "news" media!).
>
> >> >> > > An NTSC video of the signal would provide accuracy to within 1/29.97
> >> >> > > seconds,
>
> >> >> > If the average is 3.8 that means there are times when it's even
> >> >> > shorter than that, unless it was 3.8 every single time.
> >> >> Fallacy of the single cause.
> >> >> > If the light was set for 4.5 seconds, the recommended MINIMUM, the
> >> >> > average could be NO LESS THAN 4.5.
>
> >> >> Fallacy of the single cause.
>
> >> >> > Afraid you need a refresher on grade-school math, old boy.
>
> >> >> No math is involved, perfesser.
> >> >>  -----
>
> >> >> - gpsman
>
> >> > You do realize others on this newgroups know, or can look up, 'Fallacy
> >> > of the single cause'*.
> >> > So your lame attempt (and massive fail, of course) to apply it to this
> >> > black and white statement: "If the light was set for 4.5 seconds, the
> >> > recommended MINIMUM, the average could be NO LESS THAN 4.5" - which is
> >> > pretty much a straightfoward definition (minimum in this case is least
> >> > amount of time...not whatever random thing you may pontificate on) -
> >> > is, as usual, absolutely pathetic (in case, gpsboi you, really, really
> >> > don't get it, there is no fallacy here; a minimum definition of time
> >> > is just that).  I really hope you are much smarter than this in real-
> >> > life, for society's sake.
>
> >> > *For those who can't be arzed to look it up (which clearly includes
> >> > gps-sham himself), from wiki:
> >> > "The fallacy of the single cause, also known as joint effect or causal
> >> > oversimplification, is a logical fallacy of causation that occurs when
> >> > it is assumed that there is a single, simple cause of an outcome when
> >> > in reality it may have been caused by a number of only jointly
> >> > sufficient causes."
>
> >> It is even more simple. gpstroll doesn't even understand a normal
> >> distribution. Random error would be centered around the mean. Thus if
> >> there were some random machine error the average would be 4.5 seconds
> >> with some instances falling longer or shorter. For the average to be 3..8
> >> seconds means there is some sort of constant error built into the
> >> system if we are to accept that 4.5 seconds was the target. This would
> >> mean who ever set the timing didn't check his work properly as such an
> >> error could be compensated for by setting the light at 5.2 seconds,
> >> which would achive a 4.5 second average and the normal distribution
> >> about that mean.
>
> > Far more likely is that the person in the article was using a handheld
> > stopwatch, sports watch, or similar device and the need to average
> > several readings comes from simply the differences in time recorded
> > due to differences in reaction time of the operator.  If you take
> > enough readings, discard the first few readings as "practice" and then
> > average the rest, you should come pretty close to the actual value.
> > The yellow interval itself likely doesn't vary significantly if at
> > all.
>
> > If there were a serious error in his findings or methodology that the
> > authorities would have simply responded with a comment to the effect
> > "his measurements are wrong, the yellow interval for that intersection
> > is (x) seconds, which complies with requirements."  The fact that that
> > didn't happen leads one to believe that the intersection was, in fact,
> > mistimed.
>
> Guess I missed something in the trimming. I thought it was about the
> honesty of the light's setting, not the report. A stop watch is more
> than good enough for measurements to a tenth of a second. The raw data
> probably had a spread of 3.7 to 3.9.

I'm sure that you're right, or close enough to it that the point is
still valid.

nate
From: gpsman on 9 Apr 2010 14:16
On Apr 9, 1:35 pm, N8N <njna...(a)hotmail.com> wrote:
>
> Far more likely is that the person in the article was using a handheld
> stopwatch, sports watch, or similar device

Did some other Nate Nagel sneak in here and start this thread?
"Armed with a stopwatch<>"
http://www.nbc-2.com/Global/story.asp?S=12265823

"Her math tutor husband took a stopwatch<>"
http://consumerist.com/2010/04/have-you-ever-suspected-that.html

> and the need to average
> several readings comes from simply the differences in time recorded
> due to differences in reaction time of the operator.  If you take
> enough readings, discard the first few readings as "practice" and then
> average the rest, you should come pretty close to the actual value.
> The yellow interval itself likely doesn't vary significantly if at
> all.

That may be the most sensible string of characters I've ever seen
escape your keyboard.

The fallacy of the single cause is committed when the other source of
possible variation and error, the light, is omitted.
-----

- gpsman
From: N8N on 9 Apr 2010 16:38
On Apr 9, 2:16 pm, gpsman <gps...(a)driversmail.com> wrote:
> On Apr 9, 1:35 pm, N8N <njna...(a)hotmail.com> wrote:
>
>
>
> > Far more likely is that the person in the article was using a handheld
> > stopwatch, sports watch, or similar device
>
> Did some other Nate Nagel sneak in here and start this thread?
> "Armed with a stopwatch<>"http://www.nbc-2.com/Global/story.asp?S=12265823
>
> "Her math tutor husband took a stopwatch<>"http://consumerist.com/2010/04/have-you-ever-suspected-that.html

Right. thanks for looking that up. Nice you see you've actually
started to read what's posted. Now we need to start working on
actually comprehending what you read.

> > and the need to average
> > several readings comes from simply the differences in time recorded
> > due to differences in reaction time of the operator.  If you take
> > enough readings, discard the first few readings as "practice" and then
> > average the rest, you should come pretty close to the actual value.
> > The yellow interval itself likely doesn't vary significantly if at
> > all.
>
> That may be the most sensible string of characters I've ever seen
> escape your keyboard.
>
> The fallacy of the single cause is committed when the other source of
> possible variation and error, the light, is omitted.

Do you have an example of even one signal in use on a public road that
has a variable yellow interval? I suppose it's possible that it might
vary from cycle to cycle by some small fraction of a second, but I'm
not aware of any that intentionally do so.

nate
First  |  Prev  | 
Pages: 1 2
Prev: If you can't miss a dump truck sitting in the road...
Next: If there is a Japanese God out there...