[p1necone]

The usage of 'per-mile' stats for aircraft safety irks me a little bit. It's certainly not how I think of safety when I hop on a plane - my internal comparison is more based on time - like, I'm about to spend 30 minutes on a plane, how much safer/less safe is that vs 30 minutes in a car?

The distance comparison also doesn't make sense because it's not like you could drive across the ocean even if you tried.

I guess it makes sense in terms of aggregate safety for a population for transport planning, but on an individual level it just doesn't communicate what I want to know.

Edit: for an analogy - imagine if someone invented faster than light space travel, but 25% of passengers don't survive the trip. The deaths per 100 miles statistic would be amazing compared to both car and air travel, but would you sign up for a ticket?

[cmeacham98]

The distance comparison does make sense, because the point of getting on a plane/car is to travel. People don't say "I'm going to drive for 30 mins", they say "I'm going to drive from Los Angeles to NYC". Comparing how dangerous that is on a plane requires comparing by distance, not time.

[anamexis]

I would guess the majority of time in general aviation is not for the purpose of travel. It's a hobby.

[cmeacham98]

If you're assuming it is a hobby then comparing it to car/motorcycle usage for travel is useless/invalid.

[anamexis]

Yes, exactly.

[hollerith]

A lot of it is to give pilots the experience they need to qualify to fly commercial airliners, also.

[bitdivision]

Agreed.. Though it's an often used metric: fatalities per billion kilometers (f/bnkm).

Someone told me that risk per unit distance was higher when walking than riding a motorcycle, which I thought sounded like it could be possible. Sadly it seems its not true [1].

Interesting how f/bnkm is so low for driving vs walking though.

[1]: https://www.normalizecycling.com/risk-in-cycling/units-of-ri...

[p1necone]

Eh, not really. Distance travelled !== quality of destination.

I can travel X minutes on a plane for Y cost to one set of destinations, or I can travel A minutes in a car for B cost to a different set of destinations. The actual distance between my current location and my destination means nothing to me, although the potential destinations do, which certainly are more varied with plane travel.

But I live in a pretty nice place, so travelling locally is pretty good too.

[cmeacham98]

We're comparing safety, not "quality of destination".

If you were to travel from point A to point B, and wanted to know whether driving or flying was safer, then the correct metric to look at is the "per distance" one.

[p1necone]

Yeah but most travel isn't to a fixed, "necessary" destination. About the only place like that for me is work, and I certainly can't fly there.

Or to put it another way - comparing two different modes of travel to one specific destination doesn't make much sense when the destination is partially fungible. I want to know what the safest way to get to (any sufficiently nice place) is, not to (one specific nice place).

[tormeh]

The point is, being a hobby pilot is extremely dangerous.

[TylerE]

Being a lazy hobby pilot is dangerous. The majority of GA crashes are pretty basic pilot error/"gotta-get-there-itis".

Motorcycles are the same way, actually. An overwhelming amount of fatal motorcycle accidents involve alcohol at night, usually in combination with not wearing proper gear.

[someweirdperson]

Bikers frequently get killed by other forms of traffic (at least as a major contributing factor), pilots typically kill themselves (usually unintentionally, of course), with rare exceptions.

[fijiaarone]

Getting killed by running into a semi because you were weaving in and out of traffic at 90 mph and ran a red light is not really the traffic’s fault.

[anamexis]

Is it?

In US General Aviation, there were 332 deaths in 19,454,467 flight hours in the year 2020.

https://www.ntsb.gov/news/press-releases/Pages/NR20211117.as...

[oldgradstudent]

Assuming all hours were flown at Cessna 172 cruise speed of 140mph, that gives about 2e7 * 140 = 2.8e9, divided by 332 gives about 8.4 million miles per fatality.

Compared to 85 million miles travelled per fatality on the roads in general, and about 4 million miles travelled per fatality on motorcycles.

2X better than motorcycles, 10X worse than road fatalities in general.

And that's being quite generous about the mileage.

[anamexis]

I don't think miles travelled per fatality is a useful point of comparison for general aviation. This puts it about on par with pedestrian deaths per mile travelled, and I don't think most people would call walking "extremely dangerous."

[ineedasername]

>Is it?

Yes, it is.

Very similar to how dangerous motorcycle riding is. Work out approx. hours of operation from miles driven (say avg. 30-50mph) and from there use annual fatalities. [1] Given that, death-per-hour for 332 deaths/19M in flight hours is roughly comparable to the 6000 deaths seen in motorcycle accidents. Much higher than automobiles, much higher than commercial flight.

[1] https://www.iii.org/fact-statistic/facts-statistics-motorcyc...

[anamexis]

Ok, so per hour it is comparable to motorcycles. But consider also that the average private pilot only flies 100-150 hours per year.

I don't disagree that it is more dangerous than automobiles or commercial flight. But I wouldn't characterize it as "extremely dangerous." Nor would I characterize motorcycles as such.

[ineedasername]

I guess we disagree on motorcycle danger then. I consider motorcycles to be extremely dangerous (mostly to their drivers). A friend of mine died about 1.5 years ago on one. I'd had two other people (not as close) in my life die in motorcycle accidents so I used to cringe inside every time he told me he was going riding over the weekend, though I would just wish him well & to be safe. Riding made him happy, was a stress reliever for him. And as far as that goes there are probably worse habits like smoking & drinking to excess, but that doesn't make any of them non-dangerous.

[Dylan16807]

Google says "motorcycles are usually ridden for around 3,000 miles per year on average" so that's less than a hundred hours. Another result says the median is 1000 and 90th percentile is around 5000.

Neither one is "extremely" dangerous but it's a far cry from "all these strict regulations make it extremely safe" like with commercial flight.