Re: quick everybody burn your helmets...

On Thu, 14 Sep 2006 10:13:46 -0600, cline <ken.cline@xxxxxxxxxx> wrote:
Ian Smith <ian@xxxxxxxxxxxxxxxxx> writes:

The fact that [the paper] is as yet unpublished does not alter the
validity of my question. Nor does it make it OK to criticise the
research from a position of ignorance.

I have read the extended abstract ("Professional Briefing Document")
and I really wonder about the validity of the results as they apply to
cyclists who know what they are doing. The problem is that the study
measures vehicles passing distances as the cyclist rides absurdly far
from the side of the road. If a cyclist has to ride in traffic 1.25m
from the edge of the lane for any significant distance to avoid
"debris, car doors, and driversâ?? attention patterns at junctions",
then that rider should simply take up the whole lane.

But as you are aware, the 1.25m is merely one limit - the data is for
distances from 0.25 to 1.25m. Those look like reasonable limits for
the band of 'normal' cycling - I'd rarely be that close or that far
from the kerb, but I don't see anything fundamentally wrong with them
as limits for study. Looking at the data in the briefing, the
conclusions hold even if you eliminate the extreme points.

Notable, the study shows that, when riding at what I beleive to be
a reasonable distance from th edge of the road (0.25m),

That's an extremely dangerous and unreasonable location - road
markings are further out than that, and it's extremely dangerous to
try and ride along the thermoplastic used for them. Standard drain
gratings extend to 450mm from the edge, and it is similarly dangerous
to ride over them, since they are normally inclined and always
slippery (especially when wet).

In other words, this data suggests to me that passing behavior is not
influenced by helmet-wearing when a cyclist rides "as close as
practicable to the ... curb or edge of the roadway" as directed by the
Uniform Vehicle Code [http://www.swcp.com/~nmts/laws/UVCBicycles.htm].

The data does not purport to say anything whatsoever about behaviour
in a jurisdiction covered by the Uniform Vehicle Code.

The guidance for the jurisdiction the data _does_ represent says
something very different. For example, 'Cyclecraft', which is the TSO
book with detailed recommendations on cycling describes two positions
- 'primary' and 'secondary'. Primary is dominating the lane - occupy
the centre of the path taken by motor traffic (you've already
referred to this technique). More relevant to this discussion is
secondary, which is the position to take up to allow other vehicles
to pass more easily, where it is safe to do so. This is "not closer
than 0.5m to the edge of the road". The figure of 0.5m as an
absolute minimum is restated elsewhere, and is perfectly sensible -
you should not ride across drain gratings - that's simply asking for
trouble.

Your example here is completely invalid, for several reasons:

1: You are assuming a finite bounded error. This is extremely
unlikely to be true. Assume a normal distribution (or indeed any
other distribution with a tail) and your whole argument collapses.

Nonsense. U-Turn's argument easily extends to infinite distributions.

No it doesn't - in essence he said if your margin is greater than your
scatter then increasing margin does not affect chance of collision.
That's true, but only if you have finite scatter. If you have
infinite scatter (as for a normal distribution), your margin would
need to be infinite for his argument to work.

Even with a finite scatter, the research seems to be well within the
range of likely scatter - and the briefing note makes no claim that
the conclusion can be extrapolated to any clearance.

I agree that there is no reduction in risk passing at infinity + 10cm
compared to passing with infinite clearance. I'm not aware of any
vehicle ever having passed any cycle with infinite clearance.

We then simply consider the distance required to achieve n standard
deviations of confidence that a collision will be avoided, for
whatever n you prefer. Thus we can determine the distance needed to
reduce collisions to whatever small probability you are willing to
accept.

In this context I am prepared to accept any probability, provided the
probability at greater clearance is identical.

For the statement U-Turn says is false to be false, the probability at
distance s must be identical to (or less than) the probability at
distance s+interval. For an infinite distribution this occurs only
when s is infinite. S is never infinite.

regards, Ian SMith
--
|\ /| no .sig
|o o|
|/ \|
.

Relevant Pages

  • Re: Calculus XOR Probability
    ... If one of the elements is to be chosen from the set, then the probability of one being chosen is 1. ... then that 1 representing the fact that one of those will be chosen is defined to be the sum of the probabilities of each. ... If you can establish a uniform probability distribution, then you can say each possibility has the same probability. ... If we are discussing an infinite set of possiblities, and I define infinitesimal in terms of that infinity, it's not circular. ...
    (sci.math)
  • Re: Calculus XOR Probability
    ... the sum of a countable number of 0's; he simply "wants" it to be 1. ... then the probability of one being chosen is 1. ... the question is why the standard system can't accomodate the infinite case, ... The probability of each is 0 in standard analysis, ...
    (sci.math)
  • Re: Calculus XOR Probability
    ... therefore there must exist a uniform distribution on the naturals. ... If you use a 2 element set, you still get a total probability of 1. ... the sum of a countable number of 0's; he simply "wants" it to be 1. ... the question is why the standard system can't accomodate the infinite case, ...
    (sci.math)
  • Re: Calculus XOR Probability
    ... If one of the elements is to be chosen from the set, then the probability of one being chosen is 1. ... then that 1 representing the fact that one of those will be chosen is defined to be the sum of the probabilities of each. ... If you can establish a uniform probability distribution, then you can say each possibility has the same probability. ... If we are discussing an infinite set of possiblities, and I define infinitesimal in terms of that infinity, it's not circular. ...
    (sci.math)
  • Re: Calculus XOR Probability
    ... If one of the elements is to be chosen from the set, then the probability of one being chosen is 1. ... then that 1 representing the fact that one of those will be chosen is defined to be the sum of the probabilities of each. ... If you can establish a uniform probability distribution, then you can say each possibility has the same probability. ... the question is why the standard system can't accomodate the infinite case, with an infinite number of equally likely outcomes, one of which will definitely happen. ...
    (sci.math)
Loading