characterizing our risk

[prkaye]

I know this is a grim topic, but after the recent loss of an RV friend I've been obsessing over the risk factor. As a trained mathematician, I know statistics are very difficult things, and even finding appropriate data can sometimes be impossible. But I wonder how much we can find, and what conclusions we can draw. Lately I feel a very strong need to gain an understanding the risk -to quantify and characterize it.

When I look for aviation stats that I can try to apply to my own operations, I personally don't like the stats "per hour" of operation, for many reasons. I think the risks "per aircraft" make more sense, to me at least. This is the way we think instinctively anyway. When we lose an RV friend, we think this might happen to us, and then take comfort by pointing to our other RV friends who continue to fly without incident.

Anyway, regarding data, here are a few questions:

1) There are over 6600 (?) RVs that have been completed and flown. Probably most of these are actively flying. What would be a good guess, 6000 RVs actively flying in a given year?

2) Do we know how many catastrophic RV accidents occur in a typical year? What about breakdown by model?

3) Of the above, is are there any clear causes that stand-out as leading risk factors? For GA in general, I've heard fuel exhaustion is a leading cause... does that hold true for RVs in particular?

The reason we may tend to obsess about these things sometimes is clear... if we can understand, quantify and characterize the risk, we can better learn to manage and reduce it, and we can also derive some piece of mind. Accidents like the recent one near my home leave a very unsettling feeling that flying is gambling with ones' life, and I want to find a way to reduce that feeling.

[Toobuilder]

My statistics instructors used to say that use of statistics was the ability to "lie with numbers"... In other words, the same data set could be manipulated to "prove" any side of an argument. The statistics per flight hour are appropriate, I think, because it allows a much more broad sample size. But if you really want to accurately evaluate risk, you would want to do it by individual pilot because pilot error is the predominate cause of fatal accidents. This is an impossible task of course, but the fact remains that some of us are at a much higher risk of killing ourselves. I think that because of our highly variable individual processes (in contrast to military or airlines)statistics on hours flown, aircraft type, etc, are not a real relevant bit of information when you simply want to know if you're going to safely complete your next cross country.

Perhaps the best thing to do is accept that flying is a pretty dangerous pastime, and we are putting ourselves at some elevated level of unnecessary risk every time we do it.

[Geico266]

Originally Posted by prkaye

Accidents like the recent one near my home leave a very unsettling feeling that flying is gambling with ones' life, and I want to find a way to reduce that feeling.

Rule #1 Flying is dangerous. You can look at statistics all day long, study them until the cows come home, apply different quantifying formulas and massage the number for decades, but they are not going to change Rule #1. The only stats that make make sense to me are the ones that quantify mechanical failure over pilot error.

One statistic I always find interesting is we are all gonna die someday, and we all take risks, everyday. Eliminating all risk is not possible without eliminating the joys of living, and being alive. "Better to have loved and lost, than never to have loved at all. "

I'm sorry you lost an RV friend. You know the NTSB report and the probably cause of his accident. Apply it to the way you fly and move on. That is all we can do.

(See below for further instruction )

[DarylT]

Phil

There is an article in this months EAA Sport Aviation (Jan. 2010) on page 66.

This is the first month of the combined issue of both EAA magazines in one. It turned out really nice.

The writer took the NTSB database from 1998 to 2007 in which there were 2100 Homebuilt accidents and catagorized them by Human Error(pilot) (builder). Mechanical Error and Other. Each of those groups break down into other catagories. (General Pilot Error, Fuel Exhaustion, IFR to VFR transisition, etc.)

[Alan Carroll]

Originally Posted by prkaye

2) Do we know how many catastrophic RV accidents occur in a typical year? What about breakdown by model?

3) Of the above, is are there any clear causes that stand-out as leading risk factors? For GA in general, I've heard fuel exhaustion is a leading cause... does that hold true for RVs in particular?

Phil - these questions are answerable using the NTSB data search, although it takes some work to find and appropriately organize the data:

http://www.ntsb.gov/NTSB/Query.asp

It would be very interesting to have someone with a deeper understanding of probability look at these data systematically. Personally I think that such an analysis could be very useful for risk management and informed decision making.

[Jetj01]

Phil Kaye,

Your post is very interesting and a feeling shared by most your fellow General Aviation (GA) 'pleasure' fliers! After a 22 year career as a USAF fighter pilot retiring just last year I have had far too many friends killed in military aviation, the vast majority during non-combat activities. Every one of them has a story, and in general most could have been avoided.

After retiring last year I was hired by the FAA and flew as a GA inspector. After my recent heart event I resigned from the FAA...for many reasons, a story in itself. Suffice to say the job did not 'fit' me. However in my short tenure with the FAA I was quite shocked at the GA accident stats.

Unkown to most casual fliers, in our nation of aviation there are LOTS of accidents, everyday. As an FAA employee I got to see the raw, unscrubbed data daily and was shocked how really enept most of us are in accident situations! In general, and to summerize, most accidents where pilot error, and not the tough 'engine failed, tried to glide, hit a tree' type pilot error but rather:

1) Forgot to put the gear down (Over 1/4 of the daily accidents in a given month were gear related!)
2) Flying into bad weather ie. from VFR conditions into IFR and not avoiding/getting out of it!
3) Exceeding limitations of your aircraft ie. flying into known icing ( a case last month where a Bonanza crashed with 4 inches of ice on the wings!)
4) Out of CG, out of weight limits
5) Low level aerobatics, untrained usually
6) Aircraft out of annual, or completely unsafe, and the operator knew it!
7) Not knowing basic FARs and getting violated (Not an accident but potentially career ending!)

An on and on! I was disheartened by the stats, realizing that if pilots would just do that "pilot s--t" like the guy says in Tog Gun, most accidents wouldn't happen!

in general, if we could avoid the dumb stuff and the only accidents that ever happened were Act-Of-God type stuff, this aviation pastime would actually statistically be quite safe. The FAA knows it, and now I know it.

People are going to get hurt in this activity. We are going to lose friends. But we need to minimize the risks by doing the best we can to fly smart. It will make a difference and besides, the alternative is to not fly....that one does not work for me.

Finally, be careful of reading 'stats'. The real story can get lost in the numbers and lead you to conclusions that are just not true. One of my buddys quit flyig last year because a friend crashed his Glasair into a hillside. it wasn't the Glasair's fault.....

Tailwinds!

[SvingenB]

Originally Posted by Toobuilder

But if you really want to accurately evaluate risk, you would want to do it by individual pilot because pilot error is the predominate cause of fatal accidents.

I agree. Statistics doesn't necessarily mean anything to each individual. It is like saying - because one person in a million is a terrorist, then everyone has one in a million chance of being that terrorist. This is of course completely wrong, but statistically correct for a sample of unknown persons.

It is the same with accidents. The whole point is not contributing to the statistics, and there are lots of ways to minimize the risk.

[dmaib]

These are some numbers that I posted a few weeks ago on a thread in the Safety forum. These are for fatal accidents in GA "personal" flying.

I finished my CFI renewal a couple of months ago and one of the topics in the course was the fact that "maneuvering" flying accounts for the highest (25%) of fatal GA accidents. Followed by descent/approach, weather, and takeoff/climb. A significant percentage of the fatal "maneuvering" accidents are identified as having these primary causes:

• buzzing (a whopping 1/3 of "maneuvering" accidents
• traffic pattern distractions
• inadvertent stall/spin
• lack of piloting skills

additionally, aerobatics, formation flying, canyon flying, and other "special operations" are identified as frequent factors. The report notes that a lack of training and proficiency in these operations is a common thread in these fatal accidents.

These numbers are from the Nall Report and I believe they are from 2006.

[KPmarc]

If you would like a substantially less condensed version of that safety article, go here: http://www.kitplanes.com/magazine/mi...s/8485-1.phtml

The data aren't quite as current as Ron's recent short column, but you'll be able to grab the PDF version of the story and see all the charts and figures.

[terrykohler]

The reason we may tend to obsess about these things sometimes is clear... if we can understand, quantify and characterize the risk, we can better learn to manage and reduce it, and we can also derive some piece of mind. Accidents like the recent one near my home leave a very unsettling feeling that flying is gambling with ones' life, and I want to find a way to reduce that feeling.
__________________
Phil Kaye
RV9A (SB), fairings + wheel pants (then assembly!)

It's tough to quantify risk, especially given the pilot population in the experimental arena. Lots of pilots with limited experience and even less recent practice stepping into high performance aircraft (forget about power, it's the speed). Even more pilots who stop training and figure that repeating bad habits qualifies as practice. Kind of like a golfer with a 20 handicap who figures playing more or more time on the range by himself will actually improve his game (I know this won't work, from personal experience). On top of this, throw in an aversion to checklists, FARs, and ADM and it's no small wonder that our stats are poor. All this in spite of the FAA trying to bang away on this problem by making items such as these a major focus of every flight test you'll ever take. These 'Hot Button" topics aren't pulled from thin air - they're there because of accident history.
Forums such as these are a great source of information, both good and bad.
You don't have to search far to discover that it's OK to fly aerobatics in a 9, engines bigger than those recommended by the designer are OK, every pilot should do his own test flying, an EFIS with synthetic vision will keep you out of trouble, etc., etc., etc. That the very people who build and fly experimental airplanes are willing to somehow accept or downplay additional risk shouldn't surprise anyone.
As to your own situation, I think you're already on the right road. Get good transition training and after you've started flying your own plane, get some more. If that training is all mechanical and doesn't include decision making; find another instructor. Most importantly, when you've reached a point where you think you can't learn any more, it's time to find a new hobby.
Terry