wphamilton

I refer to a one-to-one comparison between activities. You compare risk for all of the hours driving against risk for all of the hours cycling. That is a broadening of the scope.

Obviously, but that does not invalidate statistical measures of risk.

Semantics, using the term in a different context than intended.


Yes, I have made that argument (that impact with the horizontal ground is independent of horizontal velocity), and it is a correct one. I realize that you disagree.

wphamilton

With that, we're back on track. I agree that the support of "risk compensation is too weak to warrant presenting, even speculatively.

There are actually two ways that the "risk compensation" argument is made. One, that drivers tend to be more careful around non-helmeted cyclists and two, that cyclists tend to be more reckless when wearing helmets. I said that there have been no studies which validated this argument - I'll say either argument. There have been several studies which examined the respective hypotheses but to my knowledge none of them, including the one under discussion, provide much support for them.

mr_bill

From another study published later by one of the same authors.

"...the cyclist population in Norway can be divided into two sub-populations: one speed-happy group that cycle fast and have lots of cycle equipment including helmets, and one traditional kind of cyclist without much equipment, cycling slowly."

"Helmets are not subject to risk compensation, but part of an equipment package."

Apropos nothing at all, I wonder why nobody has mentioned that the study was of THIRTY-FIVE participants? (I actually don't wonder.)

-mr. bill

Six jours

Not my job to do your reading for you. Nor is it really my job to correct you every time you create a strawman - I don't have that kind of time.

Six jours

I suspect it's because your shtick has already gotten old and everyone's ignoring you.

I don't have any studies to back that up, though. It's strictly anecdotal.

njkayaker

Your's is the strawman.

njkayaker

It's no more "broadening" than what you are doing.

For most people, they spend many more hours driving than riding a bicycle. At an average of 12,000 miles per year at 45 mph, that's 266 hours per year. If the average person rides 10 miles once a week at 10 mph (both of these is likely higher than the average), that's 52 hours per year. On a per-mile basis (using the "three fold" number), the average person would have be 7.6 more likely to die driving than riding.

If people road instead of driving (most people), they would end-up travelling many fewer miles. They likely wouldn't ride an average of 12,000 miles per year.

If that's the case, that would mean even with the "three times as many deaths per mile" (which I suspect is too-high an estimate), they would actually have a reduced risk of death in practice.

No, it's the same problem. Most people wouldn't replace driving with riding for the same mileage (it's not generally even possible to do!). They would ride many fewer miles than they would have driven.

No. But the average risk to a population doesn't apply to an individual. That's the mistake you keep making.

For most people, the risk-per-mile exaggerates the practical risk of riding versus driving.

No, it isn't correct because it's an serious oversimplification of what can happen in a high-speed collision. Either because the horizontal drop isn't the only thing the head is impacted by or because other things happen to increase the force of impact.

wphamilton

Frankly you're objecting to the question: what is the relative risk of driving or cycling to the drug store? As an objection the answer, it's frankly silly.

Tell it to your insurance company. The "average risk" applies to each individual in the population. His individual risk may be modified by other factors.



"high speed collision" does not occur in my statement. For convenience, in its entirety: "that impact with the horizontal ground is independent of horizontal velocity".

You're evidently making a host of assumptions about what that statement means. It does not mean, for example, that injury will not occur when a car hits you. It does not mean that nothing else can happen when you crash.

njkayaker

There really isn't a good way to answer it. For a particular individual, especially. That should be obvious.

No, that's wrong. Very, very wrong.

An average is a property of a population. An average does not "apply" to an individual.

The actual risk to any particular individual will, amost certainly, be something other than the average.

For example, if the average size of a population is 6 feet, that doesn't mean any individual is 6 feet. Indeed, the probablilty that any random person is six feet is very low. (There is no requirement that any individual's size matches the average.)

That is, as a way of predicting the size of any particular individual, the average is useless.

Indeed, the proper thing is to express the prediction is as a range based on a confidence interval.

It's even worse predicting the risk to an individual because people often can do things that can increase or decrease their risk compared to the overall-poplulation.

And it's even worse using risk-per-mile because most people (we are talking about populations) will travel fewer miles when they choose to ride over driving because they mostly don't have the time to ride as many miles as they can travel driving.

No one (including you) is being that clear about speed. I'll change "high speed" to "any nontrivial speed" since it could matter for relatively low speed collisions too.

You are the one making a host of assumptions. You don't really know whether the horizontal force is no more than a simple fall in falls/collisions that are not simple.

wphamilton

@njkayaker I think you're a little unclear about probability and statistics.

Let's illustrate with a simple game. The player draws a single card from the deck, shuffling between tries, and wins if he draws an ace.

Let's say a thousand people play this game and , on average, the ace was drawn once every 10 games. The probability of winning is 10%. Every person has a 10% chance of winning, every time he plays. The probability "applies" to every single person, even though it comes from an "average".

There are things a player might do to change his probability. Recognizing the back of aces, nicking the side or warping an ace. Yet still, we'll consider his chances to be 10% unless something is known to modify them.

Risk, in a question this general, works exactly the same way.

wphamilton

Nope. I know how to disregard the more trivial possibilities, and the more likely possibilities that make a more trivial difference.

A black hole may whiz by at exactly the wrong time for example, yanking the cyclist into the ground. I don't "know" that this won't happen. But I DO know that it isn't important. Similarly, a vehicle might strike the cyclist in just such a way that he's somersaulting, head driven to the ground at significantly greater speed. I don't "know" that it won't happen. But I know that it's a more trivial possibility.

rydabent

In another thread there is a question wether you fly a flag on your trike. I replied that I actually fly two since the seat back on my Terratrike has provisions for two.

The point I am making here is not only do I wear my helmet, but I fly two flags, all in the name of my safety. Granted cycling is not all that dangerous. But some here use the numbers of probability as an excuse not to wear a helmet. Some here also would have us believe they are the worlds greatest cyclist, and will NEVER be involved in an accident. The fact is probability may bite you in the butt the very next time you ride, or not for 5 years. But------------why take the chance.

LesterOfPuppets

Per mile is the best way to calculate risk for commuting.

For pleasure riding or riding for exercise, you'll want to go with time spent doing the activity.

Assuming you're trying to zero in on your personal risk.

I-Like-To-Bike

Lester, you are of course correct.

Njkayaker's peculiar construct of measuring travel risk of an activity by total time spent on the activity would lead to a deduction that depriving oneself of proper fitting shoes, or any shoes, would make walking a safer, less risky activity because the shoeless pedestrian will likely not walk as far, or spend as much time walking, or even walk at all to previously traveled destinations.

Using njkayaker's risk determination method, riding a bicycle with flat tires is a safer, less risky method than using a well maintained bicycle since the user would ride many fewer miles than he would have if he pumped up the tires.

Six jours

Yet you still have never explained why you don't wear a cycling-specific neck guard when you ride, nor why you don't wear any safety gear for all the other "not all that dangerous" activities in which you engage which could foreseeably result in an accident.

So again: Why do you pick and choose your on-bike safety gear in such random fashion? And why do you take steps to protect yourself during only that one specific activity?

Six jours

Cyclist suffers broken neck after collision with car door.

Hit and run leaves cyclist with broken neck.

Cyclist breaks neck in collision with deer.

There are a surprising number of such stories. I myself have known several local cyclists to suffer broken necks. And as I've noted, there is a pretty good selection of mountain bike neck protection out there. Sure, you'd look kind of dorky on a road bike, and the more serious ones look like they might be uncomfortable and restrictive, but those arguments have always been shouted down when the bareheaders make them against helmets. So what's your excuse, you barenecked Darwin candidates?

daihard 

If probably is not to be considered, why take the chance while walking? Or taking a shower? Or even sleeping? You might roll off the bed and hit your head on the floor.

rydabent

diahard

This is a cycling forum. It is under the safety and advocacy heading.

Riding off half cocked thinking that probabiliy will keep you safe is not very smart.

daihard 

So it's okay for you to walk with "half cooked thinking that probablity will keep you safe" is okay, but riding a bike with the same thinking is not very smart. That's convincing.

And how about the other body armours for cyclists, such as the neck guard @Six jours mentioned above?

Mark Stone

That's not what he said. The logic you're using, saying that the lack of PPE in other activities makes PPE irrelevant in cycling, is silly.

njkayaker

The player's probability isn't 10%. You just said so!

If you know that player is changing his odds, you'd be stupid to "consider" his chances to be 10%.

If you can take actions that are known to significantly change your risk, the average risk doesn't apply to you.

This is, of course, why "safer" drivers get reduced insurance rates. That subpopulation gets the reduced rates because the overall average does not apply!

It's possible that a significant number of cycling deaths are associated with riders doing unsafe things (riding at night without lights, etc). If that's true, and you don't do those things, then the risk to you riding is going to be less than the average risk.

You, on the other hand, are very unclear about these.

njkayaker

This is silly. You are over simpifying what happens in collisions.

njkayaker

Not necessarily.

De Clarke's Personal Opinion (isn't bisque beautiful?)

Anyway, wphamilton's statistics that worry him so are also not just commuting miles. They include all riding/driving. The driving also includes longer, highway trips that might be safer on a per-mile than short trips.

wphamilton

Not at all. I said relatively little about what happens in collisions, therefore I did not over simplify it.

However, if you DO want to go into more detail about what happens, and if you postulate that some kind of momentum transfer increasing the rider's (head's) vertical velocity, it's up to you to produce a reasonable mechanism and some data to demonstrate that it occurs. I can't be expected to address in advance everything that you might imagine.

Yes, it IS how those rates are determined. After it's determined that in a given population, fewer accidents occur after a certain period of "clean" driving record, members of that population get reduced rates. It's still what you're calling "an average". (And they start with the more general population risk and modify it according to the additional parameters, which is how I've explained it to you)

To see the difference, suppose you feel that the special "Brand X" tires that you purchased reduce your risk of accident. Call up your agent, tell him I have Brand X tires so my rates should be lower.

njkayaker

Yes, it's an average (that's why I used the term).

You still don't get that the average of the overall population does not apply to the subpopulation (when the subpopulation has different properties).

It is wrong to say that the overall average applies to the subpopulation.

If you are a member of a safer subpopulation, the average risk to the overall population isn't your risk.

If you are a safer rider, your risk might not be "three times" (and that's likely an inaccurate measurement anyway using all riders). It seems silly to base much on it.

And the supposed increase in risk of cycling over driving might be compensated by reducing risk elsewhere (due to it being exercise).