-4
$\begingroup$

I'm compiling a list of essentially "Different ways a human can fly" and then ranking them in order of safest to most dangerous.

This is my final list, is it statistically (and practically) accurate?

  1. Commercial flying internationally (airliners)
  2. Commercial flying domestically
  3. General Aviation (learning or as a passenger in a Cessna and the like)
  4. Gliding
  5. Paragliding / parachuting
  6. Hang gliding

Sources/data - fatality ratios:

  1. Commercial flying safety is well established and easily sourced
  2. GA: 0.84:100,000 hours
  3. Gliding: 0.7:100,000 flights
  4. Paragliding: 2:10,000 annual pilot fatalities
  5. Hang gliding: 1:1000 annual pilot fatalities
$\endgroup$
  • 1
    $\begingroup$ You are comparing apples and oranges. What is your definition of how safe? Per unit time or distance, per flight? In your examples you mix up fatalities per hour, per flight, per year! When doing something recreational (like gliding), you typically care about safety per time. But when flying commercially to a given destination you typically care about safety per distance, since that is fixed! $\endgroup$ – Bianfable Jun 3 at 13:12
  • $\begingroup$ Adding to @Bianfable's comment, don't forget passenger distance. For some purposes, it's not unreasonable to statistically treat a C172 with four people onboard no different from a B747 or A380 each with four people onboard flying the same distance. (Whether such a substitution is reasonable from, say, an economics or operational point of view, is a completely different matter.) $\endgroup$ – a CVn Jun 3 at 13:18
  • 1
    $\begingroup$ I'm voting to close this question as off-topic because the answer is already included in the question and no specific concerns are raised about it. $\endgroup$ – David Richerby Jun 3 at 13:57
  • 1
    $\begingroup$ If you're cruising around in a Cessna 172 with a decently maintained engine with lots of gas on a sunny afternoon, you really have to work at it to kill yourself. It takes a random event of unbelievable bad luck, like the engine falling off, or severe incompetence. $\endgroup$ – John K Jun 3 at 13:57
  • 1
    $\begingroup$ Just because the answer to this question may be "no, this is not a good analysis" that doesn't make it a bad question. The question may be based on mistaken premises and other errors, but it can still be answered in a way that provides information and helps correct those errors. That makes it a good question in my opinion. $\endgroup$ – Daniele Procida Jun 3 at 13:57
4
$\begingroup$

No, this is not a good way of understanding safety in aviation.

The figures you have may be accurate, but they are not figures describing the same things. You have figures describing fatalities per 100'000 hours, 100'000 flights, and per year.

In order to establish relative risk, you first must settle upon a common metric for each of these ways of flying, which is difficult, and if not arbitrary, is at least a choice you must make, excluding other useful choices.

The only reason a question like "Which form of flying is safest" looks simple is because it hide a great deal of complexity. You will need to ask a more sophisticated question, like: *Based on available data, in how many hours is it probable that an individual will experience a fatal incident, in each of these forms of flight?"

$\endgroup$
1
$\begingroup$

You seem to list the fatal accident numbers in the correct descending order, while leaving out a number for commercial flying.

"General Aviation (learning or as a passenger in a Cessna and the like)"

General aviation covers much more than your simplfied comment as noted in the FAA link you provided for a 2017 Nall Report covering data from 2015, while only addressing "The accident rate for GA non-commercial fixed-wing aircraft".

According to the Nall Report "To put these numbers in context, we must look back even farther: In 1950, the total accident rate was 46.68 per 100,000 flight hours; the fatal accident rate was 5.17 per 100,000 hours flown. Fast forward to today: The accident and fatal accident rates have plunged to an estimated 5.32 and 0.84 per 100,000 hours, respectively. Clearly, we’ve come a long way in aviation safety"

https://www.aopa.org/-/media/files/aopa/home/training-and-safety/nall-report/27thnallreport2018.pdf?la=en&hash=C52F88B38FD95CB7C0A43F3B587A12E2692A8502

so I feel pretty safe flying my plane around. Won't catch me jumping off a cliff or similar in a gliding suit, or out of a perfectly good airplane tho :)

General Aviation in the Nall report includes: o Piston single-engine • Piston multiengine • Turboprop single-engine • Turboprop multiengine • Turbojet • Helicopter • Experimental • Light Sport

It does not include: o FAR Part 121 airline operations • Military operations • Fixed-wing aircraft weighing more than 12,500 pounds • Weight-shift control aircraft • Powered parachutes • Gyroplanes • Gliders • Airships • Balloons • Unmanned aerial systems (UAS, or “drones”)

$\endgroup$
  • $\begingroup$ Why wouldn't you jump out of a plane? Skydiving is comparable safe to GA. Thanks for the answer though :) $\endgroup$ – Cloud Jun 3 at 13:54
  • $\begingroup$ I don't see any parachute stats in the list presented. I'd rather fly down to the ground. Only takes one mishap with a parachute to ruin your day, and I'd rather not go out that way. $\endgroup$ – CrossRoads Jun 3 at 13:57
1
$\begingroup$

For assessing GA safety (in all categories really, down to paramotors), you have to evaluate things from the standpoint of a non-idiot pilot, not just the raw statistical numbers, because the raw numbers are totally polluted by stupidity.

If you filtered out GA accidents caused by pilots doing stupid things; running out of gas, running into weather, crashing showing off, flying in conditions they shouldn't be, non bothering to check for water in fuel, and all sorts of other things, in other words, the accident rate for safe, prudent, sensible non-idiot pilots, I'm pretty confident you'll get a statistical prediction of getting killed that is down in transport aircraft territory.

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.