Hacker News Reader

Comment by dhx | Hacker News Reader

I like the simulation video at [1] which shows how a low head dam works in various conditions and how the risk can change dramatically depending on water flow, velocity, sediment levels or obstructions changing in a river. Note the hydraulic jump at 1:05.

Quite surprisingly, the incidence of deaths from whitewater kayaking and rafting is around 6 deaths per million participant days, compared to around 150 deaths per million participant days driving in cars[2].

[1] https://serc.carleton.edu/details/files/19075.html

[2] https://w.wiki/3Sma

View on HN

That can’t be right. That would mean someone who drives a car every day would have a 5.4% chance of death yearly.

It references this[1] which references this[2] which references this Kayaking paper[3] which says driving has a 152 fatality rate per million participant days if you assume 1 day = 100 miles. That references this[4] which says "The fatality rate per 100 million vehicle miles traveled (VMT)" is 1.52. If you assume 1 day = 100 miles, that means the fatality rate per 1 million days is 1.52. So there was a math error made in the Kayaking paper when they were doing unit conversion. They overestimated by 100x. That error then propagated to Wikidata.

[1] https://www.wikidata.org/wiki/Q999646

[2] https://www.wikidata.org/wiki/Q35627590

[3] https://www.wemjournal.org/article/S1080-6032(03)70564-X/pdf

[4] https://web.archive.org/web/20030323041203/https://www.nhtsa...

Thanks for the investigate work! No wonder I was surprised with the figures from the kayaking paper.

I've updated Wikidata to deprecate the incorrect incidence rate from the kayaking paper, and replaced the incidence rate with more up-to-date data from the US BTS which puts it at 0.8 deaths per 100 million vehicle-miles in the US. Using the assumption from the kayaking paper of 100 miles/day being the average distance traveled by car, whitewater sports are therefore 7-8 times more likely to lead to death than traveling in a car.

New query showing more accurate incidence of death rates: https://w.wiki/3Src

>if you assume 1 day = 100 miles.

Which is one hell of an assumption in case that wasn't obvious.

Even people with absurd commutes tend to have less than 100mi/day weekly or annual averages unless they're doing the same thing on the weekends.

USDOT says average motorist drives 13,500 miles/year, or roughly 37 miles per day[0]. So divide by another factor of 3 to get a more realistic comparison of death rates between whitewater kayaking and driving.

[0] https://www.fhwa.dot.gov/ohim/onh00/bar8.htm

Cars are horrible, is it actually a surprise?

What are the associated important ideas with a statistic like that?

Because the participants are removed from the population, it seems different from the case where they aren't.

Those number's aren't that related to low head dams (most low head dams fatalities are swimmers, not whitewater kayakers or rafters or rec kayakers). While the mechanism behind low head dams (technical term hydrolic, informally called a 'hole') can directly or indirectly kill you, they're typically a natural feature in whitewater fatalities.

I know those numbers b/c it lets me tell my parents that whitewater kayaking isn't as dangerous as they think.

Never, ever, go over a low head dam in a kayak. The danger with them is that they do not kick you out like a normal hole. You go over, and if you don’t clear the hydraulic, you get pulled back in. Then it can hold your boat indefinitely as it drifts back and forth across the river. Then you pull your skirt and you are now a swimmer being recycled in a low head dam. Low head dams make swimmers out of boaters, and kill the swimmers. Do not ever go over one.

And rafting / kayaking is definitely a calculated risk. Skill is involved. Freak accidents can occur that just wouldn’t if you weren’t on the river, but if one is smart about it, it can be about as safe as any other recreational activity (skiing, boating, mountain biking). With friends trained in rescue and experience the safety margin goes up.

For others reading this, the danger depends on what kind of river you’re running, the temperature, and a ton of other factors. In a boat, you’re likely to die if you can’t roll the boat back up and you hit your head / get knocked out, or your boat gets trapped upside down. After you bail from the boat, you could get pinned or caught up in trees, pulled under big rocks that are undercut, trap your ankle in a rock (in shallow water), which then pushes your body underwater downstream, die from exposure / drowning (you wash down the river faster than anyone can get you and just give up the will to live because it’s too cold / you can’t stay above it anymore).

I did a similar calculation when deciding to learn to fly small planes (vs buying a motorcycle) - the accidental death rate is higher for both compared to driving a car - but pilot error is the cause of something like 90% of GA fatalities. But 40% or something of motorcycle fatalities are not the fault of the biker - he did everything right but someone else (or something else) got him.

But the overall rate of GA fatalities is much higher than for motorcycling.

Yes - and you have to be exceptionally discipled, but a huge part of the GA stuff you can control (by not flying in marginal conditions, not trying to extend your range, always using checklists, etc).

But you can do all that on a motorcycle and still get creamed.

It does require you to honestly assess your ability to say "no" when conditions aren't right.

Basically what I'm getting at is there are activities that are dangerous but some the danger comes from you and others the danger comes from outside forces/other people. (Sure, I guess you could argue that you could just not ride your bike when there's other people on the road, but that's not practical).

As a motorcyclist you are 4% less likely to be involved in an accident as compared to an automobile, but 27 TIMES more likely to die.

A friend is a Navy pilot, and a crash investigator, trust you instruments all the time.

Reply on news.ycombinator.com