5 dead in Bakersfield crash of PA32

acrophile said:
Sadly, that made-up statistic is exaggerated by more than an order of magnitude.

Roughly 1% of the US GA pilot population is killed in crashes every two decades (~300 per year out of ~600,000 pilots). So for each pilot killed in a 20-year period, there are merely dozens--not hundreds, and certainly not thousands--who survive.
Click to expand...

I'm not sure I'm following? So you are saying there is a .05% per year chance you will be the pilot that crashes? :dunno:
 
MotoFlier said:
I'm not sure I'm following? So you are saying there is a .05% per year chance you will be the pilot that crashes? :dunno:
Click to expand...

I think that stat is mixing total fatalities including pilots and non-pilots with the pilot population. That is apples:eek:ranges.
 
Bonchie said:
Going by your stats, it'd actually be around 100 survive for every one that dies
Click to expand...

Right. So like I said, it's dozens, not hundreds or thousands.

I'm not so sure 300 pilots are dying a year though. A sizable percent of the ~340ish fatals we see a year are going to be passengers.
Click to expand...

The average number of fatal crashes is well above 300, with more than 600 associated fatalities. Yes, sometimes the pilot survives a fatal crash, but not usually. And some crashes kill multiple pilots.

and another sizable percent of those will be in crashes where 6+ were killed at once (i.e. when a twin or biz jet goes down).
Click to expand...

That factor does not undermine my calculation. If a bizjet crashes and kills all aboard, that's still just one fatal crash (out of the annual ~300) with at least one pilot killed.

And about 1 in 7 of those total fatalities will occur in Alaska.
Click to expand...

Even if we disregard those, the order of magnitude is unchanged.

If anything, given the type of flying he's doing (VFR, lower 48, single engine), he's probably right that for ever death, hundreds do what he does with no crashes.
Click to expand...

Even hundreds would still be an order of magnitude less than the claimed thousands.
 
Last edited:
acrophile said:
That factor does not undermine my calculation. If a bizjet crashes and kills all aboard, that's still just one fatal crash (out of the annual ~300) with at least one pilot killed.
Click to expand...

It would if you used all those fatalities and then used the total *pilot* population as your denominator.

Another factor is that not all pilots even fly for all that long. They may have a cert but don't use it. That would militate for using a smaller number as the denominator which would raise the percentage.
 
Last edited:
petrolero said:
It would if you used all those fatalities and then used the total *pilot* population as your denominator.
Click to expand...

But that would not give you a correct estimate of the fraction of pilots killed, which is what I'm addressing. (Nor, for that matter, would it give a correct estimate of the fraction of passengers killed.)
 
acrophile said:
Hundreds would still be an order of magnitude less than the claimed thousands. But even hundreds is an exaggeration.
Click to expand...

Hundreds of *pilots* as the poster stated is not right, but hundreds or maybe thousands of *flights* may have been what he was thinking of and that order of magnitude is plausible.
 
acrophile said:
Roughly 1% of the US GA pilot population is killed in crashes every two decades (~300 per year out of ~600,000 pilots). So for each pilot killed in a 20-year period, there are merely dozens--not hundreds, and certainly not thousands--who survive.
Click to expand...
Ah, might want to check your math on that. That's 0.05%, not 1%.
 
petrolero said:
Another factor is that not all pilots even fly for all that long. They may have a cert but don't use it. That would militate for using a smaller number as the denominator which would raise the percentage.
Click to expand...

Right, so then the "thousands" estimate would be even further off--a higher percentage of (active) pilots killed means a lower number of survivors per fatality.
 
acrophile said:
Right. So like I said, it's dozens, not hundreds or thousands.
Click to expand...

You and I both know that just saying dozens conjures up images of much less then 100 even if you technically meant 100.

The average number of fatal crashes is well above 300, with more than 600 associated fatalities. Yes, sometimes the pilot survives a fatal crash, but not usually. And some crashes kill multiple pilots.
Click to expand...

600? No. 2013 saw 387 related fatalies. Prior years were more around 440.

That factor does not undermine my calculation. If a bizjet crashes and kills all aboard, that's still just one fatal crash (out of the annual ~300) with at least one pilot killed.
Click to expand...

You are rounding way up to try to prove your point. There aren't ~300 fatal crashes a year. It's more like ~250. There were only 221 in 2013. Can't find 2014 numbers.

I'm sure your retort will be that the numbers have decreased in comparison to a 20 year average. Ok, but who's to say the lower numbers of the past half-decade aren't the new normal given safety emphasis areas?

Even if we disregard those, the order of magnitude is unchanged.
Click to expand...

You have to take into account what's being done and where it's being done to have an actual useful conclusion. Lumping in guys flying hard mountain IMC in Alaska with guys flying VFR day in Texas is pointless. It doesn't tell us anything about risk factors.

Even hundreds would still be an order of magnitude less than the claimed thousands.
Click to expand...

He was pretty specific about what he was doing. I don't know if it's thousands, but it's significantly safer flying single engine VFR in the lower 48 then a lot of other types of flying that are part of the overall numbers.
 
Last edited:
Bonchie said:
600? No. 2013 saw 387 related fatalies. Prior years were more around 440.

You are rounding way up to try to prove your point. There aren't ~300 fatal crashes a year. It's more like ~250. There were only 221 in 2013. Can't find 2014 numbers.

I'm sure your retort will be that the numbers have decreased given a 20 year average. Ok, but who's to say the lower numbers of the past decade aren't the new normal?
Click to expand...

I was looking at the tabulation below, which shows only a few years with numbers as low as you mention. It would be nice if there were a new normal with a lower rate, caused by other than a decline in the pilot population. But that's an unsupported speculation at this point.

http://www.aopa.org/About-AOPA/Gene...l-Aviation-Safety-Record-Current-and-Historic
 
^ No kidding.

This is so terribly awful. My family is the same size and similar ages. Very sobering.
 
Cpt_Kirk said:
^ is any of this really that important?
Click to expand...

Yes. If you're figuring out what level of risk to expose your family to, you don't want to underestimate the risk by an order of magnitude.
 
acrophile said:
Yes. If you're figuring out what level of risk to expose your family to, you don't want to underestimate the risk by an order of magnitude.
Click to expand...
Ha! As if there's a straight mathematical formula which will give you a defined amount of risk for anything.

Honestly, just stop. Your math is wrong and your logic is just silly.
 
mryan75 said:
Ha! As if there's a straight mathematical formula which will give you a defined amount of risk for anything.
Click to expand...

The math is important to consider--especially if the alternative is to just make up some numbers. The ratio of survivors to fatalities certainly matters.

Honestly, just stop. Your math is wrong and your logic is just silly.
Click to expand...

The math and logic are correct. If you think otherwise, please explain.
 
acrophile said:
I was looking at the tabulation below, which shows only a few years with numbers as low as you mention. It would be nice if there were a new normal with a lower rate, caused by other than a decline in the pilot population. But that's an unsupported speculation at this point.

http://www.aopa.org/About-AOPA/Gene...l-Aviation-Safety-Record-Current-and-Historic
Click to expand...

Notice your table stops at 2010.

http://www.rita.dot.gov/bts/sites/r...ansportation_statistics/html/table_02_14.html

That's up to 2013. There's definitely more then just an outlier year here. The lower rate of crashes and fatalities has become a trend.
 
acrophile said:
Yes. If you're figuring out what level of risk to expose your family to, you don't want to underestimate the risk by an order of magnitude.
Click to expand...
It doesn't matter if the chances are one in a million if you're the one, does it? You can't calculate your own risk by looking at statistics. Depending on your situation, you might have less or you might have more.
 
Make another thread for this conversation it doesn't feel right to argue about stupid **** in this one.
 
Bonchie said:
Notice your table stops at 2010.

http://www.rita.dot.gov/bts/sites/r...ansportation_statistics/html/table_02_14.html

That's up to 2013. There's definitely more then just an outlier year here. The lower rate of crashes and fatalities has become a trend.
Click to expand...

I agree, but look too at the accident rate per hours of flight time. Compared to, say, 15 years ago, the lower recent fatality numbers (except for 2013) are due more to a reduction in total flight hours than to greater safety per hour during the flights.
 
Everskyward said:
It doesn't matter if the chances are one in a million if you're the one, does it?
Click to expand...

Well, no. But that's not very useful on the front end, now is it?

Everskyward said:
You can't calculate your own risk by looking at statistics. Depending on your situation, you might have less or you might have more.
Click to expand...

I don't know how you calculate your risk without starting with statistics, even if some adjustments need to be made.
 
PPC1052 said:
I don't know how you calculate your risk without starting with statistics, even if some adjustments need to be made.
Click to expand...
I don't know how you calculate it down to some numerical value, period. Your situation is unique. Statistics reflect what has happened previously. They may or may not reflect the future. Isn't that what is says in your investment's statement? :D
 
petrolero said:
Over Thanksgiving I got stuck in Dallas when that cold front moved through and also there was a tropical system that was pumping moisture up over the top of it. I wasn't going anywhere. I knew by Friday that Sunday wasn't looking good so I bought my family tix on Southwest and sent them home. Monday looked better so I planned to stay until then and check it out. I had a refundable ticket on SW that I didn't release until I had made my decision to fly my own plane back - and the prop was spinning.

I decided to launch through non-freezing clouds around Dallas and fly back at 8000 since winds aloft were ****ty. But I knew I'd be at freezing level by the time I reached Childress. I knew it cleared up there also and that cloud tops were pretty low. My plan was west is best so that when I did land for fuel (lovely 20 to 40kt headwind) it would not be through freezing clouds. I also would not be in freezing clouds at altitude. But if I was then I'd fly further west to get away from the deck - I could see the clear line on satellite before I left. Turned out the deck ended sooner than I expected.

The plan worked and I was above a non-freezing deck for most of the flight then, just before the OAT reached 32 F the deck was gone. I landed later in Guymon for fuel (bucket list item, that :D ).

The point is that I was absolutely NOT going to fly myself or my family into ice. I left for Dallas a week before I was to return so I knew a front was coming but wasn't really sure how long it would hang around. So I had to leave with the thought in mind that I might be sending the family back via airline. That's just the price you have to be willing to pay to do this thing sometimes. I say that if you absolutely have to be someplace at a certain time, don't fly your own. At least that way when you're late you can blame the airline. :D
Click to expand...

Outstanding, and the way I approach it as well.

On a trip to HH last summer, my wife asked me what happens if we get to the fuel stop or hit weather, and can't fly the rest of the way? I said, wait it out, overnight, or if within a couple of hours, drive. Will still get there quicker and more refreshed, than driving the whole way (12 hr drive versus 4 hr flying).

Another old pilot once told me, "never make a guaranteed reservation at a hotel on a trip where you will have to make an overnight, this way you don't have gettheritis, and can stop short if you need to.

I tend to be very conservative and have cancelled many a flight do to weather when I had to be somewhere, making use of backup airline reservations, or even do to predicted turbulence after a long day at the office and not wanting to deal with it when leaving on business.
 
Everskyward said:
I don't know how you calculate it down to some numerical value, period. Your situation is unique. Statistics reflect what has happened previously. They may or may not reflect the future. Isn't that what is says in your investment's statement? :D
Click to expand...

The question is how risk adverse are you and more importantly, how do you choose to manage that risk.
 
Rcmutz said:
The question is how risk adverse are you and more importantly, how do you choose to manage that risk.
Click to expand...
Maybe so, but it isn't something that can be calculated to some numerical value, as much as people want to try to do so. You can guess X is riskier than Y but that's about all.
 
"The average number of fatal crashes is well above 300, with more than 600 associated fatalities. Yes, sometimes the pilot survives a fatal crash, but not usually. And some crashes kill multiple pilots"

So, another way to look at this versus the units of people, is to look at number of GA flights/sorties. How many sorties per year versus the 300 sorties that resulted in a fatal crashes? :popcorn:
 
Everskyward said:
Maybe so, but it isn't something that can be calculated to some numerical value, as much as people want to try to do so. You can guess X is riskier than Y but that's about all.
Click to expand...

You can for yourself, but not for me.

What is risky to you may not be to me. Risk is a state of mind as is "safety". I don't think riding a motorcycle is safe, but my friends who do, think flying a small airplane is not safe. :dunno:

BTW - what is the old saying. "Statistics lie, and liars use statistics". I suspect you as well as I can take any position, and use the stats to support our position, just like anyone else! :yes:
 
Everskyward said:
I don't know why I would even try to determine which is riskier for you...
Click to expand...

That is my point, and we agree.

So the statistics on accidents only give us a generalization, a starting point for us to examine how we approach this business of GA flying. It is up to each pilot to determine the risk, how to manage that risk, and figure out if the residual risk is worth the next flight, the next hour of flight, the next approach, etc.

:thumbsup:
 
Rcmutz said:
That is my point, and we agree.

So the statistics on accidents only give us a generalization, a starting point for us to examine how we approach this business of GA flying. It is up to each pilot to determine the risk, how to manage that risk, and figure out if the residual risk is worth the next flight, the next hour of flight, the next approach, etc.

:thumbsup:
Click to expand...
We do agree. I was disagreeing with the posters who were trying to come up with some mathematical formulas using past statistics, which may or may not have been relevant.
 
Rcmutz said:
"The average number of fatal crashes is well above 300, with more than 600 associated fatalities. Yes, sometimes the pilot survives a fatal crash, but not usually. And some crashes kill multiple pilots"

So, another way to look at this versus the units of people, is to look at number of GA flights/sorties. How many sorties per year versus the 300 sorties that resulted in a fatal crashes? :popcorn:
Click to expand...

Well that's certainly true but the post that was originally made suggested that there were thousands of pilots who fly for decades crash-free for every one who dies in a crash. So you aggregate the number of crashes over 20 years and divide by 600,000 (arguably it could be lower if you use *active* pilots). The order of magnitude is perhaps 100, but not thousands.

Now if you switch the discussion to *flights* then yes it is probably thousands if not tens of thousands of flights per fatality but the stats are probably a bit harder to find.
 
And as others have said... probabilities are descriptive of a population (or a distribution for math nerds like me) retrospectively, not predictive of an individual outcome. But if you aggregate all the future outcomes for the population then it is somewhat predictive. Got that? :D

It's one of those little understood sort of esoteric concepts.
 
MAKG1 said:
Don't jump to conclusions. We have no idea if icing was a factor. It might have been, but there are several other possibilities.
Click to expand...
Fair enough, I hadn't looked into the available info yet, I was reacting to the little I'd read which implied icing.

My point stands that if it WAS icing, it's worth knowing that there are plenty of tools to forecast such conditions and that flying into those conditions without a valid exit strategy can lead to issues.
 
Holy F, they're dropping like flies.

Another PA32 is down, this time on approach to Oakland in IMC. No other details. It is well above freezing, and there is an AIRMET out for moderate turbulence. Ceiling is 2000 with good visibility below.

Not a good week.
 
Last edited:
You must log in or register to reply here.
Back
Top