Mooney or Cirrus?

[aviatoreb]

Just now, RogueOne said:

Thank you for posting this.  The comments point out some additional areas of bias in the totals.  This is a MUCH better use of statistics as it shows many makes and signifigant models within their ranks.  Of course the “new Cirrus” that is saying pull early/pull often will no doubt get those numbers back in line...

I bet you are right in your assertion regarding the new philosophy of "pull early"...

The statistical concept that this alludes to is "stationarity."  Which I alluded to when I said don't include statistics from the 1950's in any of this as probably the probabilities of all things danger have changed for many reasons over time so including them only includes biases.  And likely the new pull early philosophy was an abrupt (favorable) improvement of statistics for the cirrus folks so averaging before and after does not reflect properly the current truth.  Any good study includes as much description of what was included and why.

[RogueOne]

1 minute ago, aviatoreb said:

That is right - it is easy to lie with faux-statistics. Take it from me - I really know what I am talking about on this topic as I have taught statistics many times and I am a math professor by trade (but not a stats professor),

But it is also quite important and possible to actually do a good statistical study on a topic like this, one that is designed to faithfully study the problem, and that includes all assertions of which data was included, how it was included, why some might be rejected, etc.  Such as when above I was asserting that perhaps the oldest data from the 1950's would be perhaps less relevant when comparing to another airplane that may not have existed until 2000.  Doing a good statistical study is not something one does in 5 or 10 min.  A I bet a good statistician who is familiar with the data sources already (where. to find all the right data) could put together a good study in about a days work.  A full work-days worth of work.  The sort of thing that an organization such as the Nail report people could and should do.  Consider this as like legal advice when one hires a lawyer.  A real job takes a real pro to actually dive in.  Anything less may be surprisingly wildly wrong and perhaps worse than none.  Eg in the legal advice scenario - suppose you had a real problem for which you needed a lawyer to investigate options, say on an expensive and intricate land deal worth millions and involving interactions of many different entities - would you consider the 5 min knee jerk job worth much - I wouldn't.  (RogueOne this is addresses "you" but I mean I am addressing this to all of us, not you specifically).

Absolutely no offense taken.  I would love to see that analysis.  I hope you devote a good day to it sometime.  I would very much enjoy reading with footnotes/sources and known bias identified.

[aviatoreb]

12 minutes ago, RogueOne said:

Multiple in an aircraft that goes down.  That would be just one example of “cherry picking” numbers.  Figures lie and liars figure is an adage that I live by when it comes to statistics.  Anyone that drops a stat out of the vapor with no supporting information and then disappers back into the ether is SUSPECT NUMBER ONE for an agenda.  Not buying what you are selling.

In any case, it has no relevance unless we know what the other aircraft manufactures did. And also I am not interested in what happened in the 1950s.  For example, no one wore seatbelts in the 1950s. And there was no weather radar.  And so on.

[RogueOne]

10 hours ago, jetdriven said:

Thanks for the factual reply ignore list for you. Don’t gear up any more airplanes please.  

Thank you everyone.  Byron has left the building...Goodnight.

[aviatoreb]

1 minute ago, RogueOne said:

Absolutely no offense taken.  I would love to see that analysis.  I hope you devote a good day to it sometime.  I would very much enjoy reading with footnotes/sources and known bias identified.

...actually I wouldn't do it.  I know stats quite well enough to do a convincing job of it and probably quite good enough.  But I would not do such a thing and post it publicly.  Since I am not qualified to post such a study publicly.  I have a professional reputation to uphold and this is not my subfield but putting such a study forward publicly would be a matter of putting my reputation behind it.  Think of it again in law terms - a patent lawyer would never give advice to anyone publicly regarding divorce law - even if for my own sake I feel I might have an idea of how it all works - to easy to give false advice.

So I apologize but for sake of respecting what it takes to do a good study, I won't do it myself.  

I would have thought the Nail report organization would have already don this though.

[RogueOne]

4 minutes ago, aviatoreb said:

In any case, it has no relevance unless we know what the other aircraft manufactures did. And also I am not interested in what happened in the 1950s.  For example, no one wore seatbelts in the 1950s. And there was no weather radar.  And so on.

So no increase in 1 and 15 stat if you take ALL fatalities in Mooney’s (numbers that in one incident could of been 4) and add that to the total airframes manufactured?

[ArtVandelay]

My opinion is the chute is nice to have but it’s not a panacea.
Cirrus has a reputation for being expensive to maintain, not just chute costs. I also don’t like the fixed gear or the linking the throttle/RPM together. It wasn’t just 1 thing for me that I didn’t consider a Cirrus.


Tom

[aviatoreb]

4 minutes ago, RogueOne said:

So no increase in 1 and 15 stat if you take ALL fatalities in Mooney’s (numbers that in one incident could of been 4) and add that to the total airframes manufactured?

Im not following.

[aviatoreb]

3 minutes ago, ArtVandelay said:

My opinion is the chute is nice to have but it’s not a panacea.
Cirrus has a reputation for being expensive to maintain, not just chute costs. I also don’t like the fixed gear or the linking the throttle/RPM together. It wasn’t just 1 thing for me that I didn’t consider a Cirrus.


Tom

Parachutes utility or success in flying safety aside - they clearly sell airplanes.  That is my main take away on the topic.

But I chose a Mooney and I could have purchased a Cirrus if I had wanted that.

[Jerry 5TJ]

Stochastic data or not — If the take-away is that flying light airplanes is dangerous I’d agree.   

Anecdotally, not statistically speaking—I have known a fair number of fellow pilots who have died flying in their planes.  Can’t say that about skiing or scuba diving or even driving around in cars. 

Be careful out there. 

[RogueOne]

3 minutes ago, aviatoreb said:

Im not following.

You have one number: Fatal Accidents involving Mooney aircraft

You have another number: Lives lost in Mooney aircraft

What number you are using makes a difference in the end ratio.

[aviatoreb]

2 minutes ago, RogueOne said:

You have one number: Fatal Accidents involving Mooney aircraft

You have another number: Lives lost in Mooney aircraft

What number you are using makes a difference in the end ratio.

...oh right - yes of course.  Thanks!

[Tom]

FWIW my point in referencing the above numbers is in attempt to recalibrate perspective.  

Sure lies, damned lies, and statistics.  While it’s most difficult to compare the milieu of the 1950s vs today, it’s reasonable to analyse the 1950s and 1960s statistics and ask “what likely result would XX (modern) countermeasure/technology have on the previous accident rates?”  After all, such questions are what forms our progress in the first place. It’s how we develop countermeasures.  (Otherwise a chute could have helped a lot of early mishaps, including weather-related...while I would not discount much the intellect or skills of the pilots of yesteryear who could actually afford to fly more often and by most accounts were less lizard-brained than modern folks)

I too deal with statistics on a daily and nightly basis at work. The end-all statistic in my work is all-cause mortality; closely subordinate statistics are number needed to treat and number needed to harm.  The first metric is what I posted above.  The second two metrics are what people argue about, whether they know it or not.  

As a very basic idea, when the number needed to treat (to obtain the desired outcome) is lower than the number needed to harm, then the intervention is at least completely sound as a matter of logic.  This is what some people here don’t get.  

Now, we can argue about useful load change and other subordinate factors, but when the intervention overall reduces all-cause mortality and is more cost-effective (if to society, maybe not the individual who can’t afford it) it’s a flat-earther position to not accept the overall cost-effectiveness of the intervention and the lives saved.

I’d love to get a chute STCd for my Mooney.  I’ve spoken to Boris Popov about it and write about it here previously, but there’s too little demand in the Mooney community (at least 4 years ago and it looks like little has changed).  Regarding the overall new GA purchaser demographic, I’d submit that they are more concerned, too, about all-cause mortality than a bunch of nuanced suppositions that neglects to consider the most important metric. 

[Austintatious]

6 hours ago, Tom said:

So the number is now 710 airframes.   People can run this themselves.  Go to https://www.ntsb.gov/_layouts/ntsb.aviation/index.aspx  and set the search for ~1950 to today, Mooney (as manufacturer), fatal as injury severity.

A quick review of the most recent 200 airframe losses shows ~334 lives lost associated with those accidents.  One can generalize the total fatality count up to 710 airframes to far exceed 1,000 souls lost.  Very, very costly.

The above is a very quick/dirty analysis that took about 5 minutes; it is admittedly completely crude without case-by-case read to distinguish mechanism of the fatality # listed on the search result page.

Within the context of a chute discussion...obviously not all the fatal accidents occurred in an operational phase where the chute could have been a benefit (e.g. in failure to launch accidents).  But in those accidents the other intrinsic safety features (e.g. roll-cage, shoulder belt or not) were relied upon, with associated results. And obviously many accidents are Darwin feats by any measure (a phenomenon not unique to any manufacturer's statistics).  Also, odd-ball stuff like maybe a Mooney being landed upon (or something) likely are in the 710 number, but still.

At any rate, arguments against chutes as being more expensive to either the flying community or to the larger society are completely devoid of greater understanding of the real world...at least from my perspective.
 

Im looking at this now...  Your numbers are correct, but your logic is not.   I dont have the patience to sift through every fatal accident but it would be required to get to the real bottom line.

We would need to determine how many of those fatalities would have been preventable with a parachute.   Were they high enough for it to deploy?  Slow enough for it to deploy?  Were there no other options that would have achieved a safe outcome?  Would the pilot have even pulled the chute if equipped (or were they oblivious to the problem?)

 

just throwing out the numbers and suggesting all of those could have been prevented is dishonest.

 

Also, the same search with Cirrus yields 134 fatal accidents.  This is since 1998 ...... Fatal Mooney accidents for the same time period are 169.

So at best we can say that all things being equal, the cirrus fleet is only slightly more safe than the mooney fleet.

However all things are not equal.  Mooneys fleet is much much older and generally way  less equipped avionics wise.  It would take a very time consuming analysis and still some speculation as to what is actually responsible for the slightly lower cirrus numbers for the same timeframe.  It may be that the EFIS is pulling a lot of the load.

 

give Cirrus another 48 years in the air and then we can see how they fair as they age , get cheaper and become more and more available to average Joe ( an M20C can be had for 35 grand, a used cirrus is 130+)

 

See the difficulty?

Edited August 5, 2019 by Austintatious

[HXG]

The statistics are useful, but it’s not all doom and gloom. You can learn a lot about your risks by searching accidents involving your airplane make and model (as well as similar airplanes) on the NTSB or Aviation Safety Network websites. I also review accident reports daily. I found some accidents that could happen to me on a bad or unlucky day,  but there were also many accidents and incidents that I would consider myself at low risk for experiencing. There will always be risks, but we can help mitigate those risks by learning from prior accidents.  

[Tom]

12 minutes ago, Austintatious said:

Im looking at this now...  Your numbers are correct, but your logic is not.   I dont have the patience to sift through every fatal accident but it would be required to get to the real bottom line.

We would need to determine how many of those fatalities would have been preventable with a parachute.   Were they high enough for it to deploy?  Slow enough for it to deploy?  Were there no other options that would have achieved a safe outcome?  Would the pilot have even pulled the chute if equipped (or were they oblivious to the problem?)

 

just throwing out the numbers and suggesting all of those could have been prevented is dishonest.

 

Also, the same search with Cirrus yields 134 fatal accidents.  This is since 1998 ...... give them another 48 years in the air and then we can see how they fair.

I've already stipulated that a chute would have been of no benefit in many accidents.  I specifically wrote:
"The above is a very quick/dirty analysis that took about 5 minutes; it is admittedly completely crude without case-by-case read to distinguish mechanism of the fatality # listed on the search result page.
Within the context of a chute discussion...obviously not all the fatal accidents occurred in an operational phase where the chute could have been a benefit (e.g. in failure to launch accidents).  But in those accidents the other intrinsic safety features (e.g. roll-cage, shoulder belt or not) were relied upon, with associated results. And obviously many accidents are Darwin feats by any measure (a phenomenon not unique to any manufacturer's statistics).  Also, odd-ball stuff like maybe a Mooney being landed upon (or something) likely are in the 710 number, but still."
So where is the dishonesty?

I'd go so far as to agree that maybe a chute would have been of no benefit in 85% of the accidents.  But here's the deal.  Somewhere...at some number... we find that the cost of installing a chute in every plane, maintaining it, etc...reaches parity with whatever dollar figure someone wants to pin onto a human life, plus all the expenditures (family, societal/medical/legal/insurance) associated with accidents.  To me, that's the golden number, whatever % of all accidents (fatal and non-fatal but serious) that occurs.  Insofar as there are multiple serious trauma cases for ever fatal case, I'd suppose/assume/submit/lie/whatever that if chutes prevented only 15% of fatal accidents, they'd also prevent more than 15% of serious accidents, and that that an aggregate cost savings to all parties would occur.  But there's no explaining this to some people.

For sure lots of Cirrus early adopters ended up in fatals.  Same with Mooney.  Same with other high-performance planes, at least to some large degree.  But in analyzing the big picture...in comparing A vs B vs C...for the sake of analysis I'd hold factors like pilot training and currency equal over the expected lifespan of the machine.  I'd submit that this more in line with the thinking style of someone who has 900k to spend on a private plane.  To NOT hold these latter variables equal is, to me...is cherry picking (at best). 

[RogueOne]

1 hour ago, Tom said:

FWIW my point in referencing the above numbers is in attempt to recalibrate perspective.  

Sure lies, damned lies, and statistics.  While it’s most difficult to compare the milieu of the 1950s vs today, it’s reasonable to analyse the 1950s and 1960s statistics and ask “what likely result would XX (modern) countermeasure/technology have on the previous accident rates?”  After all, such questions are what forms our progress in the first place. It’s how we develop countermeasures.  (Otherwise a chute could have helped a lot of early mishaps, including weather-related...while I would not discount much the intellect or skills of the pilots of yesteryear who could actually afford to fly more often and by most accounts were less lizard-brained than modern folks)

I too deal with statistics on a daily and nightly basis at work. The end-all statistic in my work is all-cause mortality; closely subordinate statistics are number needed to treat and number needed to harm.  The first metric is what I posted above.  The second two metrics are what people argue about, whether they know it or not.  

As a very basic idea, when the number needed to treat (to obtain the desired outcome) is lower than the number needed to harm, then the intervention is at least completely sound as a matter of logic.  This is what some people here don’t get.  

Now, we can argue about useful load change and other subordinate factors, but when the intervention overall reduces all-cause mortality and is more cost-effective (if to society, maybe not the individual who can’t afford it) it’s a flat-earther position to not accept the overall cost-effectiveness of the intervention and the lives saved.

I’d love to get a chute STCd for my Mooney.  I’ve spoken to Boris Popov about it and write about it here previously, but there’s too little demand in the Mooney community (at least 4 years ago and it looks like little has changed).  Regarding the overall new GA purchaser demographic, I’d submit that they are more concerned, too, about all-cause mortality than a bunch of nuanced suppositions that neglects to consider the most important metric. 

I have bought my last airplane.  It does NOT have a BRS.  As long as like airbags and seatbelts government doesn’t mandate older unequipped aircraft to be equipped or use discontinued I am good.  I clearly recognize the value of BSS.  I do not pine for one on my aircraft.  That is my lizard brain working.

[Austintatious]

If you are suggesting I dont hold pilot training and currency up as important as other factors you are dead wrong.    I hold them above all else.  The occurrence of an event that is only survivable by way of a parachute is exceedingly rare.  Especially when one considers the reality of what they are doing and trains and plans accordingly.  This is one problem I have with the CAPS parachute mentality.  I would be VERY VERY careful crossing any large area with no suitable landing areas in a SE aircraft.  I am also VERY Leary of night flight in SE aircraft, especially in hilly terrain.  How many people sitting under CAPS just disregard such caution because they have a chute?

Your golden 15% is a number that can never be verified.  Once the chute is pulled, there is almost no way to know if there would have for sure been a fatal accident or not.  In some cases where a wing broke off it would be obvious... but not so much in a case where an engine fails.  At best the parachutes saved every person they lowered to the ground safely... at worse every person that pulled the chute had other acceptable options.  Very difficult to find what the reality is.   The Cirrus sales team likely says every parachute pull was a life saving event, this is unlikely true.

The best we can do is look at the platforms side by side.  This has its difficulties, as I mentioned, when you are comparing two fleets where one has been around for 50 more years in service

I am all for using technology to improve safety.  But we should never allow ourselves to slack in the fundamentals because we have said Technology.  For instance, if My flying necessitated night flights over hilly un-landable terrain, than a SE is NOT a smart choice.  The availability of a parachute should not change that fact.

[Austintatious]

@ tom, you could always just wear a parachute on your back!  Buy 4 for your aircraft.  They would likely be lighter and cheaper than BRS.  Mine only cost 70 dollars every 6 months to repack.  Maybe you can get an STC to have a handle to release the door completely from the aircraft done by someone ( or just a field approval)

[aviatoreb]

1 hour ago, Austintatious said:

Also, the same search with Cirrus yields 134 fatal accidents.  This is since 1998 ...... Fatal Mooney accidents for the same time period are 169.

So at best we can say that all things being equal, the cirrus fleet is only slightly more safe than the mooney fleet.

You didn't even calibrate for fleet size.  Or for hours flown.  All I want to know is risk of injury, (or of fatality) per 100,000 of exposure in the end.

[Austintatious]

5 minutes ago, aviatoreb said:

You didn't even calibrate for fleet size.  Or for hours flown.  All I want to know is risk of injury, (or of fatality) per 100,000 of exposure in the end.

I know, that is why I said " at best"

I should have mentioned the flip side for clarity... the flip side being that the fleet size and hours flown being considered might show that the Mooney is more safe even without the chute..

Edited August 5, 2019 by Austintatious

[M20F]

The only thing that matters statistically is there has never been a fatality in a CAPS deployment within the envelope (and the envelope is reasonably big).  I really don’t know what other statistics matter to the conversation.  

Based on current odds you have 100% survivability in a Cirrus in envelope.  Nobody else can claim that.  

[Austintatious]

1 minute ago, M20F said:

The only thing that matters statistically is there has never been a fatality in a CAPS deployment within the envelope (and the envelope is reasonably big).  I really don’t know what other statistics matter to the conversation.  

Based on current odds you have 100% survivability in a Cirrus in envelope.  Nobody else can claim that.  

And how does that matter if none of the deployments were necessary?

[M20F]

5 minutes ago, Austintatious said:

And how does that matter if none of the deployments were necessary?

I don’t understand your question. 

[Austintatious]

1 minute ago, M20F said:

I don’t understand your question. 

You stating that the only thing that mattes is that if you are in envelope, you have a 100% survival rating is irrelevant to the discussion.  For starters, it is a 100% survival chance as of now.  You can bet that this will change to something other than 100%, and it wont be higher.

Furthermore, for arguments sake, lets say that NONE of the deployments were actually required to achieve a safe outcome.  That would mean that the parachute do not increase safety at all.  I am not saying that this is the case, but you seem to assume that ALL of the deployments were the only option for a safe outcome.  This is unlikely true as well.  So therefore it means that only a portion of the deployments contribute to added safety. To pinpoint where reality sits is extremely difficult and likely in many cases subjective.