How safe is it, really?

Why am I using the composite automobile numbers when highway numbers
are much higher (more dangerous)? Why am I using *any* set of numbers?
If we can pick and chose the numbers we want, we can 'prove' virtually
anything. It made the most sense to me, when comparing 'travel by car'
to 'travel by GA plane' to use the figures for *all* cars vs. *all* GA
planes.

Please feel free to break out 'self-piloted' GA numbers from the total
number of hours, the total number of deaths and the total number of
injuries if you so wish...but when you analyse the question 'Will you
be safer on a 1000 mile trip if you travel by car or by GA airplane?'
*Even if* you use the '50% higher' figures you want to use, you will
STILL find that

If 'safety' = 'probability of arriving at your destination without
injury or death', then travel by GA plane (personal flying), is
*still* safer than travel by car.

If 'safety' = 'probability of not getting killed before reaching your
destination', then travel by car is safer than travel by GA (personal
flying).

It depends on which definition you want to use. What is 'safe'? Just
for giggles, I asked that question ("Which of these two definitions
would you personally use in determeing if something was safe or not?")
to 8 non-aviator co-workers today. 6 of them said 'Injured or killed'
(which favors GA) and 2 of them said 'killed' (which favors cars).

The numbers don't lie tho...to say that aviation is 'less safe' than
car travel, one has to use a particular definition of 'safe'. You may
feel it is the 'better' definition. I don't.

Cheers,

Cap


"Mike Rapoport" wrote in message link.net...
Let's look at the 'miles per incident' rates for various events:

Event Automobile Plane
--------------------------------------------------------
Deaths 36,837,209 8,029,030
Injuries 495,000 1,742,969
Accidents 251,429 2,614,453
Total Casualties 488,437 1,432,087




Now, from these statistics, it is pretty clear that your chances of
dying in a GA plane are significantly higher (per mile) than in an
automobile. But they are both quite low.

But, your chances of being a 'casualty' (being injured *or* killed) is
*much* greater in a car than in an airplane. There is one casualty for
every 488,000 miles in a car...only one for every 1,432,000 miles in a
GA plane. Additionally, you are *10 times* as likely to be in a car
wreck (again per mile) than in a plane wreck. But again, they are
still pretty low.


Why are you using the composite light GA numbers when personal flying has an
accident rate 50% higher?

Mike
MU-2

I would say safety is a function of surviving the trip! Your last numbers
showed a fatal accident rate for aircraft 4.6 time greater than for autos on
a per mile basis. Looking at another set of numbers for autos, the NTSB
shows a rate of 1.48 fatal accidents/100 million miles. Converting the NTSB
data for GA to miles (assuming 125kts and 1.15 sm/nm) we get 9.46 fatals/100
million miles and as I pointed out earlier, this number understates the risk
for light GA personal flying by a factor of two. The overwhelming majority
of auto injuries are minor, some are not even noticed before the ambulance
chaser suggest them. If you rephrased the question including the fact that
the flying is 12 times as likely to result in death but the auto has a
higher chance of minor injury, I doubt if anyone would consider flying to be
safer.

Mike
MU-2




"Captain Wubba" wrote in message
om...
Why am I using the composite automobile numbers when highway numbers
are much higher (more dangerous)? Why am I using *any* set of numbers?
If we can pick and chose the numbers we want, we can 'prove' virtually
anything. It made the most sense to me, when comparing 'travel by car'
to 'travel by GA plane' to use the figures for *all* cars vs. *all* GA
planes.

Please feel free to break out 'self-piloted' GA numbers from the total
number of hours, the total number of deaths and the total number of
injuries if you so wish...but when you analyse the question 'Will you
be safer on a 1000 mile trip if you travel by car or by GA airplane?'
*Even if* you use the '50% higher' figures you want to use, you will
STILL find that

If 'safety' = 'probability of arriving at your destination without
injury or death', then travel by GA plane (personal flying), is
*still* safer than travel by car.

If 'safety' = 'probability of not getting killed before reaching your
destination', then travel by car is safer than travel by GA (personal
flying).

It depends on which definition you want to use. What is 'safe'? Just
for giggles, I asked that question ("Which of these two definitions
would you personally use in determeing if something was safe or not?")
to 8 non-aviator co-workers today. 6 of them said 'Injured or killed'
(which favors GA) and 2 of them said 'killed' (which favors cars).

The numbers don't lie tho...to say that aviation is 'less safe' than
car travel, one has to use a particular definition of 'safe'. You may
feel it is the 'better' definition. I don't.

Cheers,

Cap


"Mike Rapoport" wrote in message
link.net...
Let's look at the 'miles per incident' rates for various events:

Event Automobile Plane
--------------------------------------------------------
Deaths 36,837,209 8,029,030
Injuries 495,000 1,742,969
Accidents 251,429 2,614,453
Total Casualties 488,437 1,432,087




Now, from these statistics, it is pretty clear that your chances of
dying in a GA plane are significantly higher (per mile) than in an
automobile. But they are both quite low.

But, your chances of being a 'casualty' (being injured *or* killed) is
*much* greater in a car than in an airplane. There is one casualty for
every 488,000 miles in a car...only one for every 1,432,000 miles in a
GA plane. Additionally, you are *10 times* as likely to be in a car
wreck (again per mile) than in a plane wreck. But again, they are
still pretty low.


Why are you using the composite light GA numbers when personal flying has
an
accident rate 50% higher?

Mike
MU-2

Sure. We can rephrase the question to get any answer you are looking
for. Let me know which answer you want to hear...I'll give you the
'right' question to ask. The question I 'choose' to look at is "Which
mode of travel is more likely to get me from point A 100 miles to
point B without injury or death?" Well, the answer to *that specific*
question is PROVABLY 'general aviation flying' over 'automobile'. Even
'owner flown GA' over automobile. You want to ask a different
question? Only considering deaths? Fine...you'll get the answer that
GA travel is less safe.

And by the way, the Nall Report only covers Fixed Wing GA aircraft
weighing under 12,500 lbs. So that takes out many of the profesionally
driven G-IVs, Lears, Citations, Challengers, etc.

You want to break out the numbers to prove various things? Great.
Multiengine flying is *much* more dangerous than single engine flying,
at least in GA, per the Nall Report. Multiengine aircraft flew
something along the lines of 8% of GA hours, but were responsible for
almost 22% of fatalities. Should we tell people that, statistically,
if they only fly in single engine planes they will be much safer than
if they fly in multiengine GA planes? Lets look at hours. According
to the Nall Report, almost 80% of accidents involved pilots with less
than 500 hours in type. So should we break it out and tell the
original poster 'Well, once your husband reaches 500 hours in type, he
becomes *much* safer, statistically?'. Almost 40% of accidents
involved pilots with less than 500 hours total. Should we tell her
that once he hits 500 hours, he's safe to fly with?

So what numbers should we use? I chose to use 'all' GA versus 'all'
auto travel. Which definition of 'safe' should we use?

Either way, there is not much chance of dying in either. In a GA
airplane, I'd have to fly almost 8000 hours before I even had a 10%
chance of dying in a plane. But by then, of course, my risk per hour
would be much lower since high-time pilots are clearly much safer than
low-time pilots. I don't think that would apply to driving.

But either way, I'm not worrying much about it. For that 10%
probability of dying in an airplane to happen I'd have to fly *very*
actively...10 hours per week, every week, month-in and month-out for
over 15 years.

And one of the other issues was about how much pilot 'personality' and
decision making affects safety. Let's just look at single-engine
fixed-wing travel for the moment...that accounted for 412 deaths in
2001. Maneuvering flight accidents are almost *always* avoidable.
Actually pretty easy to avoid...don't buzz, always watch your
airspeed, coordinate your turns...the basic stuff I drill into primary
students all the time. Maneuvering accidents accounted for 38% of
fixed-wing single-engine fatal accidents. Weather-related incidents
accounted for another 10% of fatal accidents.

OK. This isn't rocket science. If a pilot is suffucuently well trained
and disciplined to *never* buzz, to *always* go around when an landing
looks shaky [so they don't have to do erratic maneuvering to get back
to the centerline], *never* go below the sector safe altitue, unless
you know precisely where every obstacle is, and *never* fly unless you
know that the weather is well above marginal VFR conditions, then that
pilot has removed himself from the conditions that cause nearly 50% of
all fatal accidents.

If you fly with/as a pilot who is able to avoid those conditions that
lead to those deaths (actually pretty easy to fix, with sufficient
training and discipline), then you are left with an accident
probability of 1/2 of what it is for all GA pilots taken as a whole.

If you remove those, do you know how many fatal accidents would have
occurred in 2001, in single-engine fixed-wing planes? 65. Total.

Take these numbers with some 'reasonable' assumptions, and now you are
up near one fatality for every 15 million miles, with a pilot
'disciplined' and well-trained enough to not out himself in
circumstances where a manevuvering or weather incident is likely.

So basically, if you fly with a pilot in a fixed-wing single, who is
proficient, who is well trained, and who is disciplined enough to
avoid the almost-entirely avoidable accidents involving weather and
maneuvering, then you are in a situation where, before you would have
even an 0.1 probability of dying in an aircraft accident, you'd have
to fly with him 10 hours per week, every week of every month, for *30
years*. Is that safe enough for you?

As I said before, there are lots of ways to look at the numbers...and
depending on how you want to slice and dice them, and which questions
you choose to ask, you can find anything. But in the end, as a CFI and
as a pilot, I feel *very* comfortable telling people (truthfully) that
general aviation is quite safe. And I believe I have the evidence to
back that up.


Cheers,

Chris



"Mike Rapoport" wrote in message link.net...
I would say safety is a function of surviving the trip! Your last numbers
showed a fatal accident rate for aircraft 4.6 time greater than for autos on
a per mile basis. Looking at another set of numbers for autos, the NTSB
shows a rate of 1.48 fatal accidents/100 million miles. Converting the NTSB
data for GA to miles (assuming 125kts and 1.15 sm/nm) we get 9.46 fatals/100
million miles and as I pointed out earlier, this number understates the risk
for light GA personal flying by a factor of two. The overwhelming majority
of auto injuries are minor, some are not even noticed before the ambulance
chaser suggest them. If you rephrased the question including the fact that
the flying is 12 times as likely to result in death but the auto has a
higher chance of minor injury, I doubt if anyone would consider flying to be
safer.

Mike
MU-2

"Mike Rapoport" wrote in message
ink.net...
I would say safety is a function of surviving the trip! Your last numbers
showed a fatal accident rate for aircraft 4.6 time greater than for autos
on
a per mile basis. Looking at another set of numbers for autos, the NTSB
shows a rate of 1.48 fatal accidents/100 million miles. Converting the
NTSB
data for GA to miles (assuming 125kts and 1.15 sm/nm) we get 9.46
fatals/100
million miles and as I pointed out earlier, this number understates the
risk
for light GA personal flying by a factor of two. The overwhelming
majority
of auto injuries are minor, some are not even noticed before the ambulance
chaser suggest them. If you rephrased the question including the fact
that
the flying is 12 times as likely to result in death but the auto has a
higher chance of minor injury, I doubt if anyone would consider flying to
be
safer.


"Aviation in itself is not inherently dangerous. But to an even greater
degree than the sea, it is terribly unforgiving of any carelessness,
incapacity, or neglect." -- Unknown

On Fri, 3 Dec 2004, Matt Barrow wrote:

"Aviation in itself is not inherently dangerous. But to an even greater
degree than the sea, it is terribly unforgiving of any carelessness,
incapacity, or neglect." -- Unknown

"Captain A. G. Lamplugh, British Aviation Insurance Group, London. Circa
early 1930's..." (http://www.skygod.com/quotes/safety.html - 2nd quote
from the top)

I've also seen it elsewhere with the same author credited, so it seems
legit.

I've got it set up as one of my wallpaper images on my PC... One of these
days I'll have to stick some of my aviation wallpaper things up on my
website...

Brian
PP-ASEL/Night
http://www.warbard.ca/avgas/index.html

On Sat, 4 Dec 2004 20:03:56 -0800, Brian Burger
wrote in c.ca::

On Fri, 3 Dec 2004, Matt Barrow wrote:

"Aviation in itself is not inherently dangerous. But to an even greater
degree than the sea, it is terribly unforgiving of any carelessness,
incapacity, or neglect." -- Unknown

"Captain A. G. Lamplugh, British Aviation Insurance Group, London. Circa
early 1930's..." (http://www.skygod.com/quotes/safety.html - 2nd quote
from the top)

Thank you very much for that link. The quotations there are
remarkable for their original insights and aptly articulated truths.
David English's choice of which to include adds immeasurably the
content. Bravo!

"Captain Wubba" wrote in message
om...

It made the most sense to me, when comparing 'travel by car'
to 'travel by GA plane' to use the figures for *all* cars vs. *all* GA
planes.

Except that it doesn't, really. A 500-hour pilot flying an Arrow and two
ATPs cuing the FMS on a Gulfstream V are about as different as a wheelbarrow
and a submarine. Even owner-flown jets and turboprops rarely match the
safety record of profesionally-crewed flights in the same equipment. The
data are unambiguous on this point.

It depends on which definition you want to use. What is 'safe'? Just
for giggles, I asked that question ("Which of these two definitions
would you personally use in determeing if something was safe or not?")
to 8 non-aviator co-workers today. 6 of them said 'Injured or killed'
(which favors GA) and 2 of them said 'killed' (which favors cars).

As any exit pollster will tell you, how people answer the question is
largely determined by how you ask it.

Try asking the question this way: "Activity A is three times more likely to
cause you an injury than Activity B. Activity B is four and a half times
more likely to kill you than Activity A. Which sounds like the safer
activity?"

Another problem is that you're not weighting for the severity of injury.
Breaking an arm and being paralyzed from the neck down are thus being
counted the same. Without knowing this breakdown we can only guess at what's
going on.

The numbers don't lie tho...to say that aviation is 'less safe' than
car travel, one has to use a particular definition of 'safe'. You may
feel it is the 'better' definition. I don't.

By your own tortured numbers you are 4.5 times as likely to die in a plane
crash as a car crash. QED.

-cwk.

"C Kingsbury" wrote in message link.net...
"Captain Wubba" wrote in message
om...

It made the most sense to me, when comparing 'travel by car'
to 'travel by GA plane' to use the figures for *all* cars vs. *all* GA
planes.

Except that it doesn't, really. A 500-hour pilot flying an Arrow and two
ATPs cuing the FMS on a Gulfstream V are about as different as a wheelbarrow
and a submarine. Even owner-flown jets and turboprops rarely match the
safety record of profesionally-crewed flights in the same equipment. The
data are unambiguous on this point.

Indeed. And since we are comparing 'autos' to 'GA Airplanes', then if
you are going to start teasing out certain components from one side to
make the data appear a certain way, then we need to tease the data out
from the other side as well. Highway travel, for instance, is
significantly more deadly than local driving. So...do you want to
tease out 'car trips over 200 miles' and compare them to 'plane trips
over 200 miles'? How about 'Plane trips over 200 miles flown by pilots
over 25 years of age' versus 'car trips over 200 miles made by....'
You want to talk 'wheelbarrows' and 'submarines'? Then it is equally
unfair to use auto statistics that include 100 drivers driving 2 miles
to the video store each way and back and comparing that to a Mooney
driver flying a 400 mile XC in hard IFR at night. And when you start
teasing out all the possible permutations, the data really becomes
meaningless. Is it *really* useful to know that travelling 300 miles
at night in the Southwest US during October, you are 3.16 times more
likely to be injured in a car than in an airplane?

I'm not arguing that professionally flown aircraft are safer.
Professionally driven cars are safer too. But even increasing the
accident, injury, and death per hour rates by 50%, you *still* find
that by using GA (even owner-flown) you are *still* more likely to
arrive at your destination without a scratch (i.e. without being
injured or killed) than if you take that same trip by car.


It depends on which definition you want to use. What is 'safe'? Just
for giggles, I asked that question ("Which of these two definitions
would you personally use in determeing if something was safe or not?")
to 8 non-aviator co-workers today. 6 of them said 'Injured or killed'
(which favors GA) and 2 of them said 'killed' (which favors cars).

As any exit pollster will tell you, how people answer the question is
largely determined by how you ask it.

Try asking the question this way: "Activity A is three times more likely to
cause you an injury than Activity B. Activity B is four and a half times
more likely to kill you than Activity A. Which sounds like the safer
activity?"

Indeed. And I can ask the exact same question a different way and get
a different response. I understand what and agree with what you are
saying.


Another problem is that you're not weighting for the severity of injury.
Breaking an arm and being paralyzed from the neck down are thus being
counted the same. Without knowing this breakdown we can only guess at what's
going on.

The numbers don't lie tho...to say that aviation is 'less safe' than
car travel, one has to use a particular definition of 'safe'. You may
feel it is the 'better' definition. I don't.

By your own tortured numbers you are 4.5 times as likely to die in a plane
crash as a car crash. QED.

Well, that isn't quite 'true' The liklihood of dying in any event is
proportional to the time spend performing it. But basically you are
right...and you are 4 times more likely to be injured per mile while
driving a car than flying. But to have a 'serious' (i.e. 10%)
probability of dying in *either*, one would have to spent several
*years* doing either as a full time job. And in aviation, it's been
very clearly shown that low-time pilots (under 350 hours) have a
*vastly* higher accident and death rate than more experienced pilots
(See "The Killing Zone", by Paul A. Craig), then the more you fly, the
lower your odds per mile traveled of dying becomes. I doubt that is
the case withd riving, but I don't know.

This issue is harder to get a hold of than some people seem to think.
It is *not* as simple as just saying 'GA aviation is more dangerous
than driving' It is *provable* that if you define 'more dangerous' as
'more likely to experience injury or death', then GA is actually
clearly *safer* than driving. if you define 'more dangerous' as 'more
likely to experience death', then GA travel is clearly *more
dangerous*.

Cheers,

Cap





-cwk.

"Captain Wubba" wrote in message
om...


This issue is harder to get a hold of than some people seem to think.
It is *not* as simple as just saying 'GA aviation is more dangerous
than driving' It is *provable* that if you define 'more dangerous' as
'more likely to experience injury or death', then GA is actually
clearly *safer* than driving. if you define 'more dangerous' as 'more
likely to experience death', then GA travel is clearly *more
dangerous*.

One of my statistics profs in college was fond of saying, "If you torture
the data long enough, eventually it will confess to anything."

It might be that for an IFR pilot to go up and putter around in a 172 for an
hour or two on a nice VFR day is safer than the proverbial drive to the
airport. If he invites a friend to come along, he could reasonably answer
the "how safe is it" question, "safer than the drive to the airport." It is,
on that flight.

But the OP's question was basically, "is my husbnad going to kill himself in
an airplane one of these days." If he does, odds are it isn't going to be on
a sunny Saturday morning. But if his flying contains a mix of conditions,
we'd need to take into account all the types of flying he does. And then you
get into the game of whether a pilot who flies regular IFR is safer because
he's more skilled and able to handle bad conditions, or more likely to get
killed because he "tempts fate" by flying approaches in minimums and such
more often. Guys who fly the bush in Alaska are tremendous airmen but
they're still far more likely to get killed flying a plane than a weekend
hamburger-fetcher in Connecticut.

So rather than falling down the rabbit hole, you look at the gross average,
which by its nature weights for all the possibilities. Imperfectly, to be
sure, as all statistical measures are. But it is by far more valid for
forecasting purposes than picking-and-choosing at every level.

-cwk.

"C Kingsbury" wrote in message link.net...


But the OP's question was basically, "is my husbnad going to kill himself in
an airplane one of these days."

So rather than falling down the rabbit hole, you look at the gross average,
which by its nature weights for all the possibilities. Imperfectly, to be
sure, as all statistical measures are. But it is by far more valid for
forecasting purposes than picking-and-choosing at every level.

-cwk.

Exactly. Which is precisely why I chose to use *all* auto statistics
versus *all* GA Fixed-wing data. The Nall Report doesn't cover the
G-IVs and Citations of the world..it covers planes below 12,500
lbs...which is the kind of plane her husband will be flying.

And without trying to bend the data one way ot the other, taking *all*
of the data for light fixed-wing aircraft, we come to the conclusion
that her husband is more likely to arrive at his destination *without
a scratch* if he flys GA, but more likely to arrive *alive* of he
travels in an auto.

And either way, he is *very* likely to be fine. To have even a 10%
probability of dying in an aircraft accident, one would have to fky 10
hours every week, of every month, of every year for over 15
years...we're not talking about the danger of explosive ordnance
disposal versus sitting in a rocking chair knitting. Both are 'safe',
and their relative safety (as borne out by the statistical data)
depends on whoch question you ask.

Cheers,

Cap