Why We Don't Care About Fatal Accidents

  • E-Mail this Article
  • View Printable Article
  • Text size:

    • A
    • A
    • A

At last week's SAFE symposium, John King banged a drum I've heard him bang before. He points out that in general aviation, we continue to promulgate the idea that flying a GA airplane is "safe." The oft-used comparison is that the drive to the airport is the most dangerous part of the trip. While that's true for an airline trip, it's most definitely not true when GA is part of the equation. I'll get to the numbers in a minute.

But first, when King makes this argument—which I happen to agree with and have made many times in the publications I've edited—the reaction is curious. It's basically no reaction. The eyes glaze over because most people don't want to confront the nature of risk or to wrap their heads around the statistical measures. It's not denial, I don't think, but just utter disinterest. We suffer between 250 and 300 fatal aviation accidents a year (400 to 600 deaths) and few care enough to do anything meaningful about it, other than basic lip service.

Why is this? I have a theory, but first the numbers. The GA fatal rate is 1.2/100,000 hours, although I think it's actually higher because the hours-flown estimate is a dodgy value. I think we fly fewer hours than the estimate. Standing alone, the number is hard to parse without comparing it to something else. So let's compare it to another activity in which a minimally trained and qualified operator is using a vehicle at high speeds. And that would be driving—specifically driving a car which is involved in a single-car, fatal accident. That compares favorably to GA because there aren't that many mid-air collisions so it's as close to apples-to-apples as I can get. (Motorcycles might be better, but the data isn't as good.)

The numbers are grim. On a per-registration basis, general aviation suffers 141 fatals per 100,000 registered vehicles compared to 7.5/100,000 for single-car fatal crashes. The per-hour rate tracks a worse ratio. For light GA aircraft, the fatal rate is 1.2/100,000 hours compared to .028/100,000 for single-car crashes, a 43-fold difference.* If all fatal traffic accidents are considered, the rate is .05/100,000 hours, a 24-fold difference. The risk Delta between driving to the airport for an airline trip is five times worse for the car—not nearly as bad as it used to be.

If a GA trip is involved, the fatal risk is 240 times greater than the airline and nearly 25 times worse than cars if all traffic fatals are considered. If the automobile fatal accident rate were equivalent to GA's fatal rate, we would kill more than 800,000 people a year and your car insurance would cost $25,000. As a society, we would never accept that as a driving risk, but we're blasé about it in aviation. Why?

Several reasons, I think. The big one is that as participants, we just don't care. Cynical, I know, but still true, I believe. Five hundred people dead at the bottom of smoking airplane craters isn't the societal problem that 5000 killed by drunk drivers is. People just assume—including those of us in the industry—that airplanes are dangerous and, well, everyone knows that and accepts the risk. Even though we occasionally use the phony drive-to-the-airport logic King was talking about, we don't really believe it. That's the "big lie" he refers to and it's mostly told to people outside the industry.

The second aspect of the don't-care response is that as an industry, we're willing to give pilots only enough training to get them relatively competent, but not require them to do any serious recurrency because that costs money and time and will thin the ranks and shrink the industry. So we strike a balance between what we're willing to require for training, how many dead bodies we can tolerate and how much money we want to continue to trickle into the industry from people who haven't been driven off by having to do a $250 flight review. It's harsh calculus, but, as I said in another blog, perhaps it's the cost of doing business.

Also, most of us believe—correctly, I think—that some people will crash no matter how much you train them. We let these people buy and fly airplanes, too, because we want their money to keep the industry afloat and it goes against the basic American grain to deny people the right to do whatever they want, regardless of consequences.

I'm not necessarily making a moral judgment on any of this because I support SAFE's efforts. I am merely pointing out that the group—and the FAA—are pushing against strong inertia to nudge that 1.2/100,000 downward. That's part of the cultural change we've been discussing, but nothing at all will change unless we no longer accept the stasis. If 275 fatal crashes a year are viewed by industry participants—that's you and me—as "not that bad," that's exactly what we'll continue to have.

*Notes on methodology: 275 fatal accidents a year are assumed. There are about 198,000 GA airplanes registered in the U.S., less jets and helicopters. In 2008, there were about 31,000 fatal traffic accidents in the U.S. of which 60 percent or 18,600 were single-vehicle crashes. (This may include pedestrian figures, but for this rough study, it doesn't matter much.) The NHTSA estimates there are about 248 million passenger vehicles registered in the U.S., which average 12,000 to 15,000 miles a year. At an aggregate speed of 45 miles per hour—generous—that's 266 hours of driving a year.

Click here for a video on the Lindsay, Oklahoma accident.

Click here for a podcast with Rich Stowell and John King on the same accident.

Comments (179)

I don't think that comparing GA accidents to single-vehicle accidents is relevant, or fair to GA. Compare GA accidents/100K hours to auto accidents/100K hours. Yes, GA is about four to one worse. It's roughly on par with motorcycles, and none of this is news.

I agree that a lot of people downplay this. And I'll admit to being one that shrugs it off to some degree. I can't change the overall GA accident rate much. But I can and do make a concerted effort to not be one of the statistics. I'm in the FAA Wings program, I go to the seminars, I'm working on my instrument rating. I challenge myself to improve on each flight.

I recently joined a flying club, for the camaraderie as much as for the lower prices and better fleet. What can I do (as a low-time pilot) to encourage safer flying from my fellow club members?

Posted by: Brad Koehn | May 10, 2011 8:44 PM    Report this comment

With all that is available on electronic media there is no excuse for not immediately mandating a small amount of ground training annually, say three accredited one hour courses annually. It would go a long way to addressing key risk areas and would begin to improve the numbers. Paramedics, EMT's, firefighters, daycare workers etc all have continuing education requirements. It will likely need to mandated by insurers. Trending is what is most important. Worried about over- regulation? What happens the next time a GA aircraft is in midair with a airliner? The answer is more regulation than the industry can bear.

Posted by: BILL ELLISON | May 10, 2011 9:09 PM    Report this comment

I accept the risk. I was almost killed in an aircraft accident because of a poorly maintained aircraft. It wasnt my airplane.... Mine is in immaculate shape. I knew that i was training for my cfi in a dog. 30 minutes after the crash i was in another worn out plane with the sheriff and the local fbo owner pointing out the crash site.

My grandfather got his fighter plane shot up in Tunis in WW2. He parachuted to captivity from 1000 ft agl and landed on his back because the canopy hadnt stopped swinging back and forth. So i guess im lucky and I have the risk gene. I also fly as much as 10 hours a week. I fly low sometimes. I fly low Ifr. I fly around thunderstorms. I fly my wife and kids around. Sometimes i fly over large expanses of water.

I dont do these things because i have a big set of balls. In fact I fear that which I dont understand. I make calculated decisions and never ever give
up. No set of rules or standards will change my desire to discover. My experience and knowledge , in my opinion, keep me as safe as possible.

This type of decision making simply cannot be regulated in a free society. In fact, as soon as my freedom to fly is compromised I will be looking for another place
To live...

Every fatality that i have witnessed has been the result of a dumb decision at some point or the other. Does anyone propose to outlaw risk and stupidity? I Watched Ltc. Kevin Davis fly his f-18 into the ground at Beaufort, SC.

Posted by: Brad Vaught | May 10, 2011 10:22 PM    Report this comment

He was the best of the best , but he tried to pull too may G's in a 180 turn at 200'agl late in the show.... He knew the risk, and he died doing what he loved.

Flying GA is not for everyone. Its for that part of a percentage of us who have the passion and the right stuff. Thats unlikely to change.

Posted by: Brad Vaught | May 10, 2011 10:26 PM    Report this comment

In 35 years of driving I have had 1 traffic accident. (rear ended by a drugged out lady doing 55 while I was stopped for road work by a flagman) It was a startling experience, I was in the passenger seat my wife driving. She was looking in the mirror and saw it coming but could do nothing about it. She got out the words "oh,no" before the impact. It totalled my Jeep Cherokee, but Thankfully my wife and I were not badly injured. Can't say the same for the other driver, she was transported but recovered. My point is that in 15 years of flying I have had some hair raising times but no accidents or incidents. I worry about the other driver more than my driving, and I like to be PIC not so much a passenger for that reason. I have met a lot of Type "A" pilots, they will tell you how big "theirs" are and then want to go show you in a plane. Thanks in large part to those types we will continue to have accidents that make headlines in the periodicals. Without those accidents our flying/pilot magazines would only be 10 pages.

Posted by: Roger Mullins | May 11, 2011 6:32 AM    Report this comment

"Yes, GA is about four to one worse."

This is part of the problem. I've given you data here showing its 40 times worse. If it were but four times, we would have about 30 fatals a year, a level of accidents too trivial to even worry about. When you say it's not news, I would argue that you don't have a good grasp of the real risk.

How does this play out? For one, it elevates product liability insurance costs, which account for a large percentage of the cost of new airplanes and replacement costs. The smaller the industry gets, the more this number is over represented. It is a real factor.

SAFE believe the accident record is bad enough to actually keep new participants from coming into the industry. I don't really see that, but concede it's possible.

Posted by: Paul Bertorelli | May 11, 2011 6:40 AM    Report this comment

In pursuing your instrument rating, you are probably doing the single best thing to reduce your personal risk. If you follow that with reasonable recurrency, if you stall-proof yourself, and understand the factors related to VFR-into-IMC, you likely will vastly decrease your risk of dying in an airplane.

A decade ago, the Navy famously confronted this very problem, but the first step was understanding what the numbers were telling them and conceding the problem on a community basis. From there, they proceeded to the cultural change that drove down the mishap rate.

It's not clear GA can do the same, but SAFE's effort is to at least raise the issue.

Posted by: Paul Bertorelli | May 11, 2011 6:48 AM    Report this comment

the "industry" is continuously telling people flying is safe, from John King, to manufacturers, to flight instructors trying to get(and keep) students, to AOPA presidents during press interviews.

Those promoting and making money in the industry have kept telling the HOGWASH that flying is safe. Flying GA is inherently dangerous and that is why fatal accidents are part of the game.

Posted by: Mark Fraser | May 11, 2011 6:57 AM    Report this comment

More rules and more training won't help much. Flying will never be as safe as cars(and probably shouldn't be) it is a risk/reward problem. Yes, the risk is slightly higher, but the reward is much higher so we think is it worth it. Probably 90% of pilots are responsible and do more than the minimums, but you can never do enough training to really close the gap when 10% or so just do care(or are thrill seekers).

Get the FAA off our backs. Yes safety is important(deadly serious) but most of us are working on improving daily. The FAA can never make enough rules to keep us safe, but the can make enough rules to kill GA.

Posted by: Roy Zesch | May 11, 2011 7:16 AM    Report this comment

What would the GA accident rate be if you removed the egregious, illegal and stupid accidents? A quick review of recent accidents includes: a guy who flew his airplane which had just been grounded by a mechanic. The door was wired shut with a coat hanger. He crashed and killed his daughter. A doctor who crashed with 2X the amount of narcotics in his blood necessary for pain control. Killed his whole family. A guy buys an airplane and on the first flight flies it to the limit of its range to an area of IMC and runs out of fuel after the 3rd approach attempt. A night takeoff into low IMC without an instrument rating. All fatal. There are so many of these types of accidents that they have to skew the statistics. I see how better training could prevent many accidents - but I'm not sure training would have any effect on this sort of insane behavior. How do you prevent people with poor judgement from flying? There are too many of them and they kill not just themselves but their passengers. That is the tragedy.

Posted by: ROBERT THOMASON | May 11, 2011 7:35 AM    Report this comment

Like others, I look at the accident statistics and see where over 70% is "pilot error" consisting mostly of running out of fuel, VFR into IMC and low level buzzing.

Yes, I shrug.

If I fly my well maintained aircraft well within its design limits, file IFR rather than scud run, avoid TRWs with Stormscope and XM by a wide margin, attend FAAST seminars and always land with at least 1 hour of reserve fuel, what is the relevant accident rate for me?

Posted by: Ross Bennett | May 11, 2011 7:53 AM    Report this comment

Regardless of who says flying is safe, those of us who do it know it has certain risks, some of which can be mitigated and some that cannot. Most pilots are well aware of the inherent danger involved with flying an airplane.

GA pilots are free to fly when and if they want to, especially those who own their own aircraft. The greater percentage of GA accidents result from bad ADM, and usually a string of bad decisions leading down the path to the accident. Mechanical failure is a contributor as well, but to a far smaller degree.

The airline industry has a disciplined process that must be followed before the plane can launch. Those of us in GA must always be aware that rushing through the preflight can have disastrous results. Personally, I am anal about making sure the plane is airworthy before I risk my life, and sometimes those of my passengers, in it.

In the last few years, we’ve had three fatal accidents at my airport, resulting in four fatalities. Two were caused by component failure. Could they have been prevented by a more thorough preflight? Possibly. The other was a result of attempting the impossible turn when the engine failed on takeoff. No doubt, a bad decision that needlessly killed two people.

Discipline and risk management clearly reduce accident rates. We don't need more regulation. We are adults. We do, however, need to get rid of complacency and ramp up the personal discipline and decision making if we want to avoid being a statistic.

Posted by: MEREDITH HUTTO | May 11, 2011 8:16 AM    Report this comment

In the movie "Fifth Element" a computer voice tells the driver:"You have zero points remaining on your license" or something to that effect. If a "danger score" was assigned to current weather, condition of the airplane, proposed route, time of the day, pilot experience etc. then a swipe of the computer-readable pilot license at the FBO would tell the poor chap that for his chops the proposed flight is deadly and disallow the liftoff. Of course, in line with American values there should also be an override button labeled "Proceed with removal from gene pool anyway" that would allow the liftoff, automatically sign the waiver of insurance and notify the local search and rescue teams that a probable accident has been added to tha queue in status "Waiting to happen". Such Next-Next-Gen system would go a long way to both preventing accidents and providing more clear statistics, as well as lowering insurance rates for the rest of us. Seriously, the first step would be for AOPA and EAA to put the real statistics somewhere front-and-center on their websites, especially in parts designed for students. Now those two most authoritative GA websites do their best to hide the real danger quotient behind sleek PR Speak. The closest number I was able to get after searching far and wide on the Web was that GA is about 10 to 1 more dangerous than driving. Such info should be coming from AOPA and EAA directly, otherwise their campaigns to promote GA look like less-than-truthful advertising.

Posted by: Andrei Volkov | May 11, 2011 8:50 AM    Report this comment

Paul, the real question is how much "supervision" are we willing to accept. In the FAR 121 airlines a second opinion is required to ensure that the flight can be conducted safely and within the SOPs and regulations. The dispatcher and Captain must agree to fuel, weather, route, alternate and maintenance condition of the aircraft prior to release. The military, with few exceptions (may have changed since I flew my last flight long long ago) requires an Ops Officers concurrence/approval prior to launch and usually buy in from wingmen or crew. That goes against the grain for most GA pilots. It's their one area that is totally devoid of a second opinion. I suspect ultimately that it will be the insurance companies that will require a higher level of professionalism, but it'll be awfully hard to coral the cowboys. The pros ALWAYS have a back up plan, night single engine over mountainous areas or large bodies of water have none, yet you hear people continue to do it. I don't have an answer to that one.

Posted by: Burns Moore | May 11, 2011 9:04 AM    Report this comment

I thought that the whole future of GA was to try to REDUCE regulation (LSA's, Sport pilots, relaxing 3rd class medicals, etc)? That puts all sorts of people "in the air" into easily overloaded LSA's with a lot less training.

Paul, you can't have it both ways. You cannot promote the GA industry on one hand by praising LSA's and then in the next month suggest that GA needs higher requirements. It's the same air and the same risk.

I "don't care" about fatal accidents because I read the NTSB reports. They tell you EXACTLY what not to do.

Posted by: Mark Fraser | May 11, 2011 9:52 AM    Report this comment

It is a huge country with relatively few GA aeroplanes so it is unlikely that pilots have frequent experience of losing friends and helping grieving families. That is why the eyes glaze over when the subject is raised. Make it personal and the safety campaigns will possibly have more impact.

Posted by: Brian McCulloch | May 11, 2011 10:06 AM    Report this comment

I wish there were some way to compare the statistics of the risk averse to the risk ignorers. We have pilots at our field who are organized, disciplined, comply with the rules in letter and spirit, almost obsessive in their approach to flying safely. We also have the "whatever" types, who fly however the whim strikes and it's no surprise when they become a smoking hole. For the first group the odds are probably about the same as the auto statistics, for the latter they are probably truly scary.

I often get to work with the JAARS missionary pilots who fly in some of the toughest environments in the world, yet, because of their proffessional, disciplined approach to flying, have an incredible safety record considering what they do and the conditions in which they do it. They prove the rest of us can do better, we just need the will to do it.

Posted by: Richard Montague | May 11, 2011 10:14 AM    Report this comment

Brian, the problem is that no one "points and laughs" anymore when stupid stunts end in a self-inflicted death. Such things need to be made fun of, criticized, used as examples. You don't make it "personal" you call it "shameful". When PIC's shrug their responsibility it's a black eye on all of GA. It should not be tolerated nor empathized with. That's why "care" is counter-productive.

Posted by: Mark Fraser | May 11, 2011 10:16 AM    Report this comment

Not to mention, who is going to pay for all this extra training and more comprehensive flight reviews? With gas prices climbing and the cost of GA outside the reach of most people, it will just chase another segment of pilots out of mix with no guarantee that it will have any positive impact on student starts (especially where the cost is increased 40% to cover the more comprehensive training requirements). Like so much else, it comes down to a cost/benefit analysis and it seems highly unlikely to be worth it to marginally improve the fatality rate.

Posted by: Ken Appleby | May 11, 2011 11:12 AM    Report this comment

As an instructor I have always assume that my potential student was intelligent (capable of reading the news) and aware that flying is more risky than driving.

On demo flights, I always retard the power to idle during a normal descent to demonstrate that airplanes DO fly without power to put that understandable fear to rest. Furthermore, I do not say that flying is safer than driving (because it is not and they known it; I would loose my credibility right away). Instead I say that flying can be done safely and explain how to achieve increased safety.

I think that we err as an industry when we pretend. As a result, we loose both the people who call our bluff and the people who enjoy higher risk activities (since we say flying it is not that).

Risk management starts with accurate information. We are cheating people by feeding them inaccurate data. It is not about more regulations, it is about honest discussions and putting people in charge of their destiny by honestly telling them what it takes to minimize their risks.

Posted by: Flying Bug | May 11, 2011 11:39 AM    Report this comment

I guess I’ll take a different approach.
Flying is inherently dangerous, but unfortunately there are several things standing in the way of improving safety.
One is practice. To get better, you have fly. The rising cost of fuel has a major limiting factor on flying hours as the cost of gas directly relates to the amount of flying pilots do. With the price of gas going up, pilots are flying less, not more, and so they’re not staying proficient.
Two is situational awareness in the cockpit. We should be doing everything we can to leverage technology to help the pilot. Instead we (read FAA and lobbyists) put road blocks in place. There are several avionics suites under $5K that could increase situational awareness and the safety factor. I cannot use them unless TSO’s, even though I have a functioning panel of steam gauges, which are my primary instruments. TSO panels? 3-4 times as much.
I do believe that a majority of accidents are caused by cowboys, who take unnecessary risks. You can’t fix stupid. But we also have a bureaucracy, mired in 20th century rules and regs, that prevents us from becoming safer.
So unless you can change the price point of flying, you keep getting inexperience pilots, with less than optimal SA, who kill themselves and others.
This is the reputation GA will continue to enjoy unless we make it safer.

Posted by: John Wrenn | May 11, 2011 11:41 AM    Report this comment

My understanding of the problem lies in each individuals personal choice. I don’t think you can teach someone to not be a risk taker who is prone to the ’Evil Conevil’ syndrome. I consider myself a safe pilot but looking back to when I 1st started (when I was single) I took risks that I wouldn’t take today. I really don’t believe that more regulations that are created or the more restrictive or burdensome (mandatory) extra training, will solve this problem. I enjoy the freedom of flying and the more regulations and rules that are created by FAA the more I feel my freedom is being taken away. Are we truly free in this country anymore? The government seems to be trying to protect us from ourselves by taking away our agency bit by bit. I don’t think that is the answer to the problem at all.

Posted by: Jess Reynolds | May 11, 2011 11:43 AM    Report this comment

Car accidents are almost 100% due to driver error. Plane accidents are not. I don't know the numbers or the percentages of pilot error to fatalities (nor did I see any in the stats Paul provided) but I'm betting that it is less than 30 percent.
NTSB's database of G/A accidents with fatalities (or without fatalities) revealed that engine failures play a large part. Car drivers usually never have to worry about this problem and drift to the shoulder.
Auto engine failures in the last twenty years has been reduced almost to zero where airplane engines haven't made any real strides toward reliability in the same time period.
Using the average Paul put forward implies that every fatality is due to pilot error which is not true. Apples and oranges? Engine failures, structural failures, maintenance incompetence, unknown weather anomalies, bird strikes and a host of other conditions are beyond a pilot's control.
Additionally which sector of the GA community should we include. Take bush pilot operations in Alaska which shows how dangerous that occupation is requiring skills to avoid scenarios that your weekend pilots won’t encounter. But their fatalities are still in the G/A database. Accidents/percentages then must be smaller than presented in Paul's argument. Most of my pilot buddies regularly attend COPA "rust remover" seminars and have safety talks provided by either Transport Canada, search and rescue professionals or helpful aircraft maintenance people.
These do help.

Posted by: Don Ledger | May 11, 2011 11:43 AM    Report this comment

"I don't know the numbers or the percentages of pilot error to fatalities (nor did I see any in the stats Paul provided) but I'm betting that it is less than 30 percent."

You'd bet wrong. It's 80 percent for fatals, about 65 percent for all accidents. For fatals, mechanical failures account for less than 10 percent. For all accidents, it's about 15 percent.

Posted by: Paul Bertorelli | May 11, 2011 12:00 PM    Report this comment

As an IS-BAO auditor and Human Factors Educator, the focus in todays "Human Factors" arena doesn't just take into effect the pilot. It's anybody who comes into direct contact with a flight that is a factor in the outcome.

Want to know the real statistics? 95% of all aviation accidents have at least one direct human factors cause. I know what you are saying. What about the 17% of mechanical failures data I have seen? Well, that's an NTSB category, not a human factors category. We can knock that 17% number down if we note that in half of all mechanical failures, someone knew a problem exsisted before the flight ever left the ground or the MX issue was inproperly done. The leak on the floor, the weird smell on that last flight, the rough running engine etc. wasn't enough to deter the pilot from flying anyway. Most of what's left on the mechanial failure side can be attributed to the pilot failing to execute or follow proper emergency procedure AFTER the failure. Albeit under stress, it still goes to human failure.

What you have left is about 5% of undetectable, unavoidable accident outcomes. Know how much trainig resources we spend on "Human Factors?" About 5%. That's right, we spend about 5% of our training resources on a problem that kills us 95% of the time.

Posted by: Kim Barnes | May 11, 2011 12:27 PM    Report this comment

So we can knock off twenty percent from the original totals. How about when we compare the number of professional pilots to private (weekend) pilots. The pro pilot might have a dozen or more passengers on board (e.g. Beavers, C-206s - Beech 1900s that get included in the fatalities as well. BTW-that's something that can't be directly related to professional drivers of buses and transport trucks because those drivers normally survive as opposed to whatever autos they hit.

Posted by: Don Ledger | May 11, 2011 12:30 PM    Report this comment

Don Ledger has a point. It is difficult to determine the risk of flying by just counting dead people. The 700 people who die every year in airplanes aren't all pilots, just as the 31,000 who died last year in automoblies wern't all drivers.

That is why we must look at individual accidents and keep the body count out of it as so far as determining "Fatal" or "Non-Fatal" accidents.

Posted by: Kim Barnes | May 11, 2011 12:41 PM    Report this comment

Paul--I agree with your premise that GA accidents are understated, but also agree with the first poster that says that it is unfair to compare GA fatals to single-car auto accidents. Single car accidents are at limited speed. There is no question that most car accidents involve multiple vehicles (often colliding at opposing speed) while few general aviation accidents are mid-airs. Just because the TYPE of accidents are different doesn't make the danger different.

I wholeheartedly agree with you that the FAA information is poor--without knowledge of the number of hours flown, statistical analysis is a guess at best. The answer? Simply have every mechanic fill in a card with the number of hours flown for each aircraft at the annual inspection. Analysts have complained about this lack of information for at least the 48 years I've been flying. Gather the data, and we would have valid numbers in a couple of years.

Posted by: jim hanson | May 11, 2011 12:45 PM    Report this comment

Paul, it's fine to use auto accident rates as a baseline because it's something familiar, a risk level that the average person *should* be able to gauge (although I doubt that, actually). But there's still something that bothers me about the comparison.

For one thing, the argument is that a) GA is more dangerous than driving, b) driving fatality rates have gone down due to something or other, therefore c) more or better training/maintenance/supervision will improve GA fatality rates. Does that really hold up? We can gripe about training, maintenance, attitude, and lack of oversight in aviation but are those things better for the driving population? Good grief, no! The typical training level of drivers is abysmal. The average car on the road is not better maintained than the average light plane, and as for attitude and oversight, just about anyone can hop in a car at just about any time and go just about anywhere. So none of those factors can really be the magic bullet that will make GA safer. (But -- recent experience is one factor that should stack up dramatically in favor of drivers vs pilots. Could that be the controlling variable?)...

Posted by: Glenn Killinger | May 11, 2011 1:03 PM    Report this comment

The other thing that is fishy about the comparison is this: in general industry we carry out risk assessments that take into account not only the likelihood of a fault, but also the severity of the outcome and the availability of safe corrective actions once the fault has occurred. The likelihood of engine failure while driving could be roughly the same as while flying, but the severity of the outcome and the availability of safe options (pulling over and stopping) are very different.

But when we consider the 80 per cent pilot error rate in fatal accidents, studying the reliability of the mechanics or avionics is not the right approach either. We can measure the probability of a connecting rod failure. How do we measure the probability of a bonehead decision by a pilot?

If we could get to the root cause of that bonehead decision, we might be able to figure out how to fix it. Assuming that attitude and training are the root causes is really just a guess.

Posted by: Glenn Killinger | May 11, 2011 1:04 PM    Report this comment

On the other hand... and this may be exactly your core point, Paul ... I have seen in general industry that it IS possible to drive the incident rate down with an effort at cultural change and awareness, without necessarily having a rigorous analytical understanding of the underlying mechanisms.

BTW if the response to this blog is an indication, "don't care" is not the problem. Good article.

Posted by: Glenn Killinger | May 11, 2011 1:14 PM    Report this comment

Cars have improved tremendously in the last 20 years--front and side airbags, seriously engineered crush zones etc. Of course cars aren't as mass-sensitive as airplanes. But how many 70's Cessnas have airbag harnesses or ballistic parachutes? (And, mention ballistic parachutes and you will be crushed by all the negative replies from other pilots. It's a useless 35 lbs, you could die anyway like that guy in the smoking Cirrus, as a pilot you should never get yourself in that situation, as a pilot you should be able to fly yourself out of any situation, knowing the chute is there will just make you careless, and on and on.)

The real answer isn't training, it's technology. Make airplanes safer, just as we've made cars safer. Thing is, both the bureaucracy and the pilots themselves stand in the way.

Posted by: Patrick Underwood | May 11, 2011 2:01 PM    Report this comment

Talking about cars vs. airplane speeds and the accident rate misses the point, which is that the first place to begin understanding and recognizing risk is via statistical outcome and exposure. Comparing the two gives you an *understanding* of the relative risks, not a framework to directly address either one. That's a different problem.

Most pilots tend to assess risk the only way they know how, by intuition and emotion. I submit that this leads many of them to underestimate risk factors. For example, if pilots really understood that a third of fatals involve stall-related events, would they seek more understanding and mitigation of that risk?

No one knows. But we have a pretty good idea that pilots don't have the sense that for GA accidents, human factors and judgement are the overwhelming drivers. Above, Don said he thought less than 30 percent of fatals were pilot error, but that's shy by nearly a factor of three.

How can you assess risk if you don't know this? You fall back on your personal demons and emotion. That's not all bad, but some facts help, too,

Posted by: Paul Bertorelli | May 11, 2011 2:50 PM    Report this comment

I believe we as the flying community can immediately improve the fatal accident rate. How many stall/spin accidents would have been avoided if the accident airplanes were equipped with an AOA indicator? We've seen it proven (by Paul B. himself) that these are far more effective for avoiding an impending stall than an airspeed indicator. If 30% of fatals are due to stall/spin accidents, wouldn't the mass installation of AOA indicators show an immediate improvement in the fatal accident rate?

Posted by: JOHN EWALD | May 11, 2011 3:09 PM    Report this comment

Maybe you should concentrate on the comfortable level of risk. You compare GA flying to driving a car, which leads me to believe that GA is transportation in your mind. For myself, flying is an adventure/outdoor sport, so I compare it to other such sports, such as whitewater kayaking, mountaineering, offshore sailing, etc...

I don't expect flying to be as safe as driving and I don't want the burden of cost or beauracracy imposed on me to make it "safe" enough for the industry.

Posted by: Todd Smith | May 11, 2011 3:14 PM    Report this comment

GA flying will never be as safe as driving. The aviation environment is inherently riskier due to the fact that we are operating in three dimensions, thousands of feet off the ground, higher speeds, lighter structures, weather, etc. What I explain to non-pilots is that while statistically driving is safer than GA, that the stats don't tell the whole story.

Aviation accidents usually follow a chain of events that can usually be traced to a combination of poor decision-making, bad risk assessment, and lack of proficiency, all on the part of the PIC.

In cars we place our fates in the hand of every stranger aroud us. I can do everything right in a car and still die in a few milliseconds if the guy next to me decides to fumble with his cellphone at the wrong time. In short, pilots tend to dig their own graves, but that gives them a lot of control over when they dig, and they usually have at least a short window of opportunity to prevent an accident if they are alert enough. Drivers have relatively little control over their environment, and usually have almost no time to recognize and prevent the accident that kills them. I'll take my chances in the air over the freeway any day.

Posted by: Chris McLellan | May 11, 2011 3:21 PM    Report this comment

Just for the heck of it, I calculated the rate for all fatal accidents, not just single car. It's .05/100,000 hours, about 24 times lower than GA.

Posted by: Paul Bertorelli | May 11, 2011 3:38 PM    Report this comment

"How many stall/spin accidents would have been avoided if the accident airplanes were equipped with an AOA indicator?"

John Ewald,

Bullseye! AOA indicators in GA aircraft would eliminate the majority of (if not all) stall/spin accidents. Most of my flying time is in fighters with an AOA indicator -- they are invaluable. It has always boggled my mind that GA aircraft don't come from the factory with AOA gauges as standard equipment.

Posted by: Gary Dikkers | May 11, 2011 3:57 PM    Report this comment

Paul--thanks for taking the time to do the recalculation--we've gone from 45 times as dangerous to "only" 24 times as dangerous.

The airlines and corporate safety records are about the BEST we can hope for--good equipment, sharp crews, good training. While we can't equal those stats with the equipment most GA pilots fly, it is a more meaningful benchmark.

We've already had a breakdown of engine failures vs. total accidents. How about a breakdown of airline and corporate accident rates compared to cars? How about a breakdown of instrument operations accidents compared to VFR--comparing both to cars?

Having this information would not only set achievable goals, but would help tell us where to focus.

Posted by: jim hanson | May 11, 2011 3:59 PM    Report this comment

I am simply awestruck at the rationalization, at the misreading of the data, at the willfull acceptance of the terrible accident rate in GA. This culture appears to be insular, resistant to change, dodgy about the data and doomed to continue to kill far too many pilots and innocents. I wondered, during the SAFE symposium, how difficult this cultural change process will be. I am humbled. Sheesh.

Posted by: William Ross | May 11, 2011 4:10 PM    Report this comment

As one of the SAFE Symposium organizers, and a frequent contributor to Aviation Safety magazine, I would like to emphasize that our pilot training system is woefully deficient in risk management training. Risk management is a science that can be taught as a GA training application, allowing a pilot to identify, assess, and mitigate risk. The pilots on the right tail of the GA bell curve already practice it, the people on the left tail never will, but the vast majority in the middle don't know what they don't know because they were never taught it. Stay tuned to the Symposium web site in a few weeks (or less) for some bold ideas about how to work risk mangement (and other techniques) into our training with no new regulation.

I definitely believe that the "latent" GA market is turned off by our fatal accident rate but Paul is right to suggest recent data is lacking.

Robert Wright, SAFE Smposium Chair (and former FAA lead GA Executive)

Posted by: Robert Wright | May 11, 2011 5:35 PM    Report this comment

One of the reasons for a blase attitude about GA accident statistics, is that most of feel they don't pertain to "me". Within the constraints of $$$ and mission, we have great capability to control our own risk. For example, a properly maintained, fixed gear, single engine airplane is very safe. If one flys such an airplane only in the vicinity of an airport, during daylight, over open, flat terrain, with clean fuel, the fatality risk is extremely low. When we start adding other factors, such as x/c flights, IFR conditions, night IFR over mountainous terrain etc., the risk increases. The beauty is that as "private" pilots, we can almost completely choose our own level of risk that we find to be acceptable. While I think that education of risk factors is very important in developing a "safe" mindset, I would be very resistant of a "one size fits all" regulatory attmosphere. Risk matrixies and similar risk assesment procedures are great tools, but terrible masters.

Posted by: Dale Olsen | May 11, 2011 6:16 PM    Report this comment

Something like 95 percent of aircraft accidents have a human origin, which also relates to mechanical failure. Most of us are aware of the accident-waiting-to-happen type of pilot and can usually avoid flying with him, but that sort of thing does drive up the accident rate, along with the just plain dumb decisions.

I count myself among the silent majority, the sort of pilot who reads the accident reports on the basis that it's not clever to repeat somebody else's mistakes. I also wrote a book on aircraft accidents many years ago, for a number of reasons, among them the fact that I was approached by the publisher. I took on the project in the end because I've always throught that aviation incidents and accidents attract far too much hype and uninformed comment in the media, and this was a chance to help put the record straight.

I don't know if it had much impact on the general public, but the book was well received by my peers - and it made me a lot more careful as a pilot. We pilots amy appear callous to an outsider, but we have an inbuilt form of insulation against being too affected by an accident, even if it results in the loss of a friend. In my case at least, I read a report and think it wouldn't have happened to me because ... I don't fly aerobatics, or I don't fly that type of aircraft, or I'm too cautious, or some other excuse. Without that insulation we'd stop flying. Period.

Posted by: JOHN KING | May 11, 2011 6:40 PM    Report this comment

A disclaimer which would have appeared with my original comment but put it overlength: I have no connection whatsoever with the fellow named in the first sentence of Paul's article. I'm not even in the same country.

Posted by: JOHN KING | May 11, 2011 6:42 PM    Report this comment

I suggest that the major cause of aircraft accidents is; pilots start flying with zero experience. News? Not hardly. Those who survive enough mistakes (AKA; terrifying thrills) to be impressed with the fact that aircraft can scatter your hair, teeth and eye balls all over the county, become "experienced" and are safe pilots. IF he is to survive; the novice must also learn respect for weather, maintaining recent experience, aircraft maintenance and fuel over destination cushion. Therefore; I suggest that when you factor out those who do not have the experience to respect the fact that airplanes are terribly unforgiving of the arrogant and those who do NOT know their limits, then the fatality rate in GA is comparable to any means of transportation.

Posted by: George A Hutchinson | May 11, 2011 8:11 PM    Report this comment

Inexperience a major cause? Perhaps, but there's always the puzzling one at the other end of the scale. Think Steve Fosset.

I'd take comfort in the Darwin Factor eliminating the dross, except that the roadbound idiots tend to take out innocent people as well, something not quite so common in aviation accidents.

Posted by: JOHN KING | May 11, 2011 9:25 PM    Report this comment

i think getting to know my personal limits and the airplane limitations are somehow lowering my risk, reading the NTSB also helps on what not to do!. Human error is the main cause for an incident or accident, but as some comments above say You build your own accident, and ovbiously there will be more accidents caused by weekend pilots than ATP's, in the same order there are more accidents caused by teens than the UPS delivery guys.

Posted by: joe kawage | May 12, 2011 12:30 AM    Report this comment

There is a bunch of "what a dumb move; I'd never do that" response to every accident report by the rest of us, yet in the 40 years of experience, the stats haven't varied much: the stalls, flight into IMC, improper inspection, are about the same as in the 60's.
I chalk it up to a lack of respect for the air and for what is required for safe flight. Mr. Wilson, please bring on the risk assessment training. It is most needed. But what is even more needed, and will never totally be attained is the ability of someone to assess another individual's attitude towards risk management.
A cowboy attitude will still be out there, such as flying too low or under a bridge, but the guy taking off in a Tri-Pacer without a rudder (removed for recover), flying over loaded, without enough fuel, trying to get home in scud, and other stupid moves; these guys can be taught, screened and almost eliminated with better training and assessment. Yes, it will infringe on all pilots, and it will cost. But to improve the pilot pool to one that pays attention will be a better state of GA.

Posted by: Pete Jessen | May 12, 2011 1:07 AM    Report this comment

I don't know why we are comparing fatal rates per registered vehicle or per hour. We should be comparing rates per mile driven or flown. My plane flies three times faster than a car, so I fly 1/3 the number of hours to make the same cross-country trip. All the turboprops and light jets (that's still GA) fly 6x to 10x faster than cars, and get used more often. Maybe the accident rate per mile for GA would be even better.

Posted by: Marc Clemente | May 12, 2011 7:41 AM    Report this comment

I have two comments.

First comparing the fatal accident rate between GA and ground travel bogus. They are very different animals with very different consequences for stupid, careless, or reckless behavior.

In close to 35 years of flying (and 30 years as a CFI)I have known or had first hand knowledge of more than 20 fatal accidents. Only a very few were cuased by something other than the pilot. Nearly all of the fatals were caused by a total disregard for flight safety. Most were flight into terrain due to IMC conditions, no instrument rating, trying to slip in uder fog and rain VFR, low level aerobatics, buzzing, an various other Darwinian activities. All in spite of the fact that the pilots knew better and were aware of the potential results. These were unintentional suicides.

I have concluded that, for a large part, training as presently, given will not correct reckless or stupid behavior. Those that will ignore the traing of the teachers and the council of there peers will continue to kill themselves and, unfortunatly their passengers! How can we instill judgement and responsibilty in those who have neither? The common route of more costly recurrent traing and other recurring requirements can not adress this problem.

I wish I had an answer. All I have done is recommend that students that have these traits find some other way to injure themselves.

Posted by: James Hiatt | May 12, 2011 8:12 AM    Report this comment

Robert Wright hit the nail on the head. As one who started flight training in his 50s, private 2005, instrument 2009, I have fairly recent experience with the current flight training regimen and risk management is not at the top of the curriculum. The emphasis was on stalls, steep turns, landings, etc. I had a good primary instructor who showed me IMC in order to impress upon me a need to avoid it. That said, little was taught in the way of risk management, weather evaluation (other than avoid) or overall judgement. I have read as much on ADM as I can find published but there is precious little out there. Flight reviews have all emphasized airspace and FARs, not weather evaluation or flight planning. The FAA should change this.

The flight schools need to spend more time teaching judgement. Have a student plan a flight in which the correct answer is "don't do this flight." The other thing a new pilot needs after training is a mentor to fly with, both right and left seat. GA doesn't get the first officer experience one gets with the airlines but that is precisely where one learns the judgement necessary to keep safe. After training, we are basically turned loose with our "license to learn" hoping to stay alive long enough to achieve some mastery of our art.

Posted by: BYRON WORK | May 12, 2011 8:26 AM    Report this comment

More regulation is not the answer - it boils down to exercising good judgement. You can't legislate common sense. Nothing would have stopped that dimwit kid from landing his Warrior on the beach in Queens last month. Likewise, Cory Lidle's accident wouldn't have happened if he'd chosen not to fly on a marginal VFR day, in busy unfamiliar NYC airspace. These were pilots that made bad decisions, and didn't recognize the accident chain forming, which is understandable considering their experience level. But it's not just the newbies that get tripped up by bad judgement. A 20k hour pilot based at my field in CT was lost due to fuel exhaustion in a turbine Bonanza some years ago. This was a guy who worked in accident investigation. If somebody with those credentials and experience can make that kind of basic error, what could be expected of a new pilot? A similar FE accident happened to another pilot I met, who pushed a King Air to the ragged edge of its endurance. The ensuing crash (1/2 mile from the runway) killed a friend and his son. He told me he thought the plane had not been topped off properly.
In my ten years as a CFI, I've always tried to teach (and show) good judgement. Sadly, there is no shortage of examples of what not to do - just read thru the NTSB briefs from any given month.

Posted by: Will Alibrandi | May 12, 2011 9:45 AM    Report this comment

Ballistic recovery parachutes. Yes, yes, I can hear all the hot-shot pilots here making derisive comments about this. I cannot believe there are arguments made AGAINST the provision of recovery chutes on aircraft; probably made by the same category of people who believe they are safer in a car without seat belts - they will dumb down the pilot population or make them reckless etc.Complete nonsense. There can be no logical reason not to fit one to every GA aircraft and the only hurdle is paperwork and cost. They undoubtedly would save lives, measured by any statistic you could come with.

Posted by: PETER THOMAS | May 12, 2011 10:07 AM    Report this comment

If we were to double the number of hours we fly by increasing recurrent training, we'd not come near halving the accident rate per hour--maybe we'd knock it down 25%. Overall, the number of accidents would go up 50% due to the increased number of hours. Not to mention all the money we'd spend. And, we'd take out a flight instructor on some of those flights, not just the pilot. That makes no sense to me. If we're serious about saving lives, it has to be done in a fashion that doesn't put lives at risk. Simulators are the obvious solution here. And as for the bit about pilots' eyes glazing over when confronted with the statistics--I admit, I kind of skimmed down that section reading Paul's article. All those numbers, and I'd have to sit down and check the figures myself to even know he got it right. I'd say I'd rather just go flying, but you know that's not going to happen, I'm too busy . . .

Posted by: DAVID CHULJIAN | May 12, 2011 10:25 AM    Report this comment

"They undoubtedly would save lives, measured by any statistic you could come with."

Well, we can actually come up with data to shoot down this claim. We have good data on Cirrus crashes and while the Cirrus overall rate is similar to other airplanes, the fatal rate is about 1.6 to 1.9, compared to GA's overall of 1.2. Furthermore, the percentage of Cirrus accidents that are fatal is among the highest in GA. So for whatever reason, systemically, all of the Cirrus safety features, including BRS, haven't reduced its fatal accident rate. It could, in fact, be a little worse.

Would it be different if BRS were on every airplane? Who knows. My guess is not. One qualifier here: As I said, the GA rate may be higher than the 1.2 we use because of inaccurate hours-flown data. Maybe it's really 1.4 or 1.8. If that's true, than the Cirrus system produces only an average airplane, not a safer one. Further, when compared to like airplanes, the Cirrus may be about the same or a little better or a little worse. We don't have good hours for the other models.

One fallacy in BRS claims of "saves" is whether the occupants would have survived without using it. In many cases, I think the answer is yes.

Having said all this, I think BRS is a great idea. It can save lives and has. But let's not overstate its potential or demonstrated capabilities. It hasn't proven to be a magic pill.

Posted by: Paul Bertorelli | May 12, 2011 10:39 AM    Report this comment

To say that xx accidents were caused by yy always struck me as fairly bogus. Without any data recorders no one truly knows. Especially pilot error which really amounts to smoking hole, nothing mechanical blatently wrong at impact = pilot error.

Posted by: BYRON WARD | May 12, 2011 10:53 AM    Report this comment

I was a Vietnam era Cobra crew chief for one year. During that time, my outfit's flight casualties were 14 pilots killed, 1 wounded. Pretty gory stuff and I know some of the readers are saying that it's combat flying in a war zone(hostile fire, pressure to complete the mission, etc.). Here's the breakdown:

Hostile Fire - 2 KIA, 1 WIA

Loss of Control - 1, plus a marine officer friend he took up for the famous last words, "watch this", joy ride(Inverted flight(stictly forbidden) which broke the rotor head off of the mast. A/C fell from a thousand feet with no rotor blades.

VMC into IMC
Lost in low ceilings running out of fuel - 11.

Posted by: Edward Michalowski | May 12, 2011 10:54 AM    Report this comment

There are snakes that mimic poisonous snakes even though they themselves aren't poisonous. But so I heard they don't mimic fatally poisonous snakes. Why? Because predators that bite fatally poisonous snakes aren't around any more...the ones that learned are dead.

What does that have to do with GA? To me it seems like in cars there is a huge number of total accidents for every fatal. It's not so common to meet someone who has not once been in a car accident in a life time of driving. After a few years of driving that gives many people a lesson that it *can* happen to me. But in flying, your first accident has ~20-25% chance to be your last. That leads people down the garden path to believing that it's soley their superior skillz that have protected them from harm and therefore, it will never happen to me.

Posted by: BYRON WARD | May 12, 2011 10:59 AM    Report this comment

Paul--- One's chances of dying in a GA crash, IF involved in a crash, are much higher than dying in a car, IF involved in a crash---for various reasons already touched upon. It's not a surprise. To me, a more relevant question would be: What are the chances that if you fly, you will crash? And then I'd compare that stat with: What are the chances that if you drive, you will crash? It seems to me that you only compared fatality statistics, and that's not really news. Is flying less forgiving of stupidity? You bet. We already knew that too.

Posted by: Andrew Cleveland | May 12, 2011 11:35 AM    Report this comment

This last post spurred a thought. What have the advances in GA safety done to reduce fatalities? Do the lap airbags now common in Cessnas & Cirrus reduce fatalities? How about the honeycomb impact-absorbing seats of the Cirrus? How about the airframe parachute? Let's look at data from the *1970s* for autos if we want to compare the safety of autos to the majority of the GA fleet today. I bet it is much less severe! Today's cars are literally 100s of times safer, statistically speaking, than cars made 30 years ago. I think that needs to be given consideration here -- a 60-knot impact on the ground in a late-model SR-22 may be vastly more survivable than a 1970s-era ship with no shoulder belt and no headrest. I think this might make a strong case for modern aircraft (despite their increased cost). I wonder if the manufacturers see it that way as well?

Posted by: Christopher Rimer | May 12, 2011 12:05 PM    Report this comment

I've seen, heard and read proposals by none other than NASA (a government agency!) to establish a "small aircraft transportation system", which has as its intent to put more and more people into non-airline airplanes. The problem, given the risks inherent in aviation, is this: How long will it be before the "right" of everyone to have a flying machine, becomes the DUTY to do so? After all, look at what's happened with cars. Henry Ford's dream was to put everyone on wheels in their own car, but look at what the result was: We're now so dependent on private cars that for most of us, not driving isn't an option. Do we want the same thing to happen with aircraft?

Posted by: Alex Kovnat | May 12, 2011 12:16 PM    Report this comment

The number of comments tells you that people care - very much - about safety.
However, they are also defensive and extremely skeptical about costly and unproven proposals to improve safety - especially if they're self-serving proposals from flight instructors or regulators.
People care about safety, but they believe most "safety" proposals will be expensive and inconvenient, while doing little or nothing to improve safety. Deep down, they don't believe anyone knows HOW to improve safety, not really.
Regulators have had 100 years to reduce the accident rate, and by now we can assume they've done everything they know will help - and many things in addition that sounded like good ideas but actually don't help. Their track record of overreach, and burdening safety improvements with certification requirements, suggests that by now they're hurting rather more than they help.
More flight instruction? Same problem: flight instruction long ago reached the limits of what's known. You can only get so far with repeated preaching. The reality is that more-rated pilots have higher, not lower accident rates (the pattern reverses at ATP, which is driven by type of operation, not by training). Instructors don't know how to help, and more of the same won't achieve much.
The cries for a data-based approach come from the same basic demand: show me that you know HOW to improve safety, before you ask me to support your expensive proposal. Show me the data; show me the facts.
I hope to see some.

Posted by: Finbar Sheehy | May 12, 2011 1:03 PM    Report this comment

"Is flying less forgiving of stupidity? You bet. We already knew that too."

Of course you do. That's why you will never be in a fatal accident.

Here's the problem I see in risk assessment and the lack thereof. If you broad brush it by saying...I already know this...then to a degree, you suppress the curiosity that leads you to understanding whatever risk factors you're personally susceptible to but don't even realize. What's more likely to kill you but not the other guy?

Right now, I'm looking a Cirrus accident. Classic base to final stall. The recovered accident data show that at the point of departure, the airplane was at an indicated speed of 60.4 knots with a 4.98 degree pitch angle and a 31.73 degree roll angle.

What was the mechanism that led a wealthy, educated and well-respected middle-aged businessman to get an airplane that far into extremis without realizing it? If he had known that this particular point in the pattern was the highest risk portion of the flight, would his risk assessment antenna have been activated? Did he even have risk a assessment antenna?

Even in this thread, we have seen comments to the effect that only a third of GA accidents are caused by the pilot, when the reality is far worse. We have seen a comment that BRS is a great life saver but the data indicates it makes little difference in the overall accident rate.

Posted by: Paul Bertorelli | May 12, 2011 1:05 PM    Report this comment

What many in this industry seem to be saying is, hey, flying is dangerous. Get over it or don't do it. SAFE's view--again, correct--is we're doing ourselves a disservice if we don't at least attempt to give pilots the basis for risk recognition.

There are stupid people involved in flying who will kill themselves not matter what. But smart ones are doing it, too. The challenge is to get at them and find out why.

Posted by: Paul Bertorelli | May 12, 2011 1:05 PM    Report this comment

"Show me the data; show me the facts. I hope to see some."

What facts do you wish to see? We can point to the industry's CAST effort for Part 121 and show that the airline rate went from .034 to .005 over a 12-year-period. Industry/FAA partnership. Just luck, ya think?

First, you have to at least get past the knee-jerk reaction that training won't help, even though it has in the military, in the airlines and in commercial operations. It's just one of many solutions to effect culture change. The difference is the participants in those fields recognized they had a problem, had a desire to fix it and had some sort of industry--duty, economic--to work on it.

Posted by: Paul Bertorelli | May 12, 2011 1:19 PM    Report this comment

I wonder how many Cirrus fatals are the result of trying the save that expensive, beautiful airplane? Pulling the BRS totals the airframe.

Posted by: Patrick Underwood | May 12, 2011 1:20 PM    Report this comment

Paul wrote: “Further, when compared to like airplanes, the Cirrus may be about the same or a little better or a little worse. We don't have good hours for the other models.”

We do have good data on some other models. The shining “Star” of safety in the general aviation fleet is the Diamond Star DA40. Diamond Aircraft in general have only 0.16 fatals per 100,000 flight hours, vs. a single-engine average of 1.27. Not only do Diamonds have fewer accidents, those accidents are also less severe: 85% of all Diamond accidents don't result in any injuries - compared to only 53% of no-injury accidents for all General Aviation Aircraft.

The DA40 is the safest in the Diamond family. I believe its safety record comes from a combination of factors including forgiving flight characteristics, good situational awareness with the glass panel and the large wrap-around wind screen, a strong composite cocoon around the cockpit and aluminum fuel tanks within a composite wing guarded fore and aft by carbon spars. The last feature is why the DA40 almost never burns after impact. How many fatals are the result of fire after an otherwise suvivable crash? Another factor in Diamond's record could be something of a self perpetuating cycle; safety conscious pilots choose the DA40 for its safety record.

Posted by: DAN MONTGOMERY | May 12, 2011 1:37 PM    Report this comment

For what it's worth, traffic deaths throughout the developed world--with few exceptions--have trended downward since 1970. In the U.S, they declined about 40 percent. In other countries, more such as Switzerland (-80), Germany about the same, the U.K about 40 percent.

The GA overall rate has actually improved, too. In 1970, it was 18.10/100,000. Now it's about 6-ish, for a 60 percent decline. Some of that has to do with technology, but not as much as for cars with shoulder harnesses, crush zones, airbags, ABS and so on.

Posted by: Paul Bertorelli | May 12, 2011 1:42 PM    Report this comment

There are two institutions that we can learn from: Civil Air Patrol has a flight release system where you ask someone else for the keys to a plane and have to prove that you, the plane and the weather meet certain qualifications or they deny the keys. CAP also requires annual flight checks in make and model, and offers training, buying into the latest buzzword du-jure. Their safety kingdom is a bit self serving, but when they mandated 'safety training' we sent instuctors out to fly with the search pilots. A look at what their accident stats are might be useful because they carefully track hours per year and aircraft type. At one time they were also all hair on fire to dispose of all 550 legacy aircraft and buy new Cessnas with glass cockpits plus Mauls and Airvans, so some conclusions might be drawn from that.

Posted by: THOMAS M CONNOR | May 12, 2011 1:59 PM    Report this comment

The second source of data might be the Air Force Academy: They fly fixed wing piston aircraft, providing a fairly accurate data set to draw from.

Posted by: THOMAS M CONNOR | May 12, 2011 2:01 PM    Report this comment

"Classic base to final stall. The recovered accident data show that at the point of departure, the airplane was at an indicated speed of 60.4 knots with a 4.98 degree pitch angle and a 31.73 degree roll angle."


Lots of neat data there, but what was the angle of attack (AOA)? (I know, it must have been the angle at which the wing loses lift and quits flying.) I'd be willing to bet that if that Cirrus had an AOA indicator and the pilot had learned to fly using it, this "classic" base to final stall would not have happened.

Did the SAFE symposium say anything in support of more widespread use of AOA indicators for GA aircraft? That is one action that would have positive results.

Posted by: Gary Dikkers | May 12, 2011 2:05 PM    Report this comment

Well, the AOA was greater than 17 degrees. But then you knew that. There is a way to calculate AOA from the Cirrus stored data, because I got it from the Lindsay accident. The full report may have it, but I didn't get it.

While I'm a big believer in AOAIs, I'm a bigger believer in learning to fly the wing, which is hardly a rocket-science concept. Base-to-final, I want my eyes outside the cockpit, yet many pilots with glass are looking at the PPI or airspeed.

AOAIs didn't come up as a topic, nor is it listed in any of the recommendations, although it should be. But do note that teaching stalls relative to load factor and AOA is on the list and is probably the single most important and specific recommendation SAFE came up with.

Posted by: Paul Bertorelli | May 12, 2011 2:35 PM    Report this comment

I am an avid reader of NTSB reports. What do I see? Month after month of poor flying skills in combination with obliviousness and/or an arrogant disregard of the risks involved with the flight they are about make. The community's laissez-faire, only fertilizes these attitudes.

'Outs' is the pilot's distance to the grips of panic. Vigilantly monitoring one's situation in all phases of flight, you know how many 'Outs' you have left. How else would you know if it's time to deploy the BRS. Reports show that many hides that might have been saved by use of the BRS sys. aren't. They just aren't being deployed in a timely fashion or at all. Weather had nothing to do with Cory Lidle's accident. In a race between Lidle, his instuctor and a building, the building got to the scene of the crash first. The building won.

AOPA's et al criticism and hypocracy to the FAA's surprise changes in the written test, I find, most discouraging. I'm assuming that the questions were still related to flying, not about Quantuum Physics, Metallurgy, Organic Chemistry or Geo-politics. The folks just didn't know their stuff. The tests weeded them out. That's what tests are supposed to do. Give people a free pass to get behind the controls with only a rote knowledge of what's going on, I don't think so. Isn't this what this discussion is about? Situational awareness anyone?

Posted by: Edward Michalowski | May 12, 2011 3:15 PM    Report this comment

Regarding Paul's example of the Cirrus base to final stall spin accident. This is a perfect example of why the BRS system is a bogus, feel good, waste of 35 pounds of useful load. It won't save the life of the pilot in the most common of GA accidents. It won't prevent "controlled flight into terrain" in IFR. It will tend to cause inexperienced pilots, with more money than brains, to fly an airplane they aren't qualified to fly in conditions they don't have the training or experience for. Which is another way of saying no frame of reference for risk managament. Those who advocate for a AOA reference on the other hand are exactly right.

Posted by: Barton Robinett | May 12, 2011 3:36 PM    Report this comment

"There is a way to calculate AOA from the Cirrus stored data, because I got it from the Lindsay accident."

You can calculate an inertial based AoA with typical glass data (ahrs + gps). It's only the aerodynamic AoA in a stable airmass i.e. gust / shear will invalidate the calculation assumptions. I can think of some higher order cleverness that would at least detect shear / gust from a typical glass data record but I'm not sure that you could actually correct for it. I think Lindsey just assumed inertial.

Based to final eyes outside the cockpit, I wonder, would he have stalled if he was monitoring the airspeed???

I think one element of these presumably eyes out, energy losing accidents are trim problems i.e. if you're trimmed at ref and you are pushing (or pulling) you are either pulling g or off speed. If you're not trimmed on speed to start with and eyes out the idea that you can reasonably control the airpseed by "flying the wing" is a fallacy, imho.

The low altitude yank & bank special doubles down on this problem because if you're trimmed but pulling g again seat of the pants (really palm of the hand) sensing of airspeed _does_not_work. You can't untangle stick force changes due to speed from due to off trim.

I think these issues and most to all pilots weak understanding (because it's not taught) of them are a factor in at least some of these accidents. I don't know how you'd ever prove it though.

Posted by: BYRON WARD | May 12, 2011 3:47 PM    Report this comment

The late and great Gordon Baxter of FLYING magazine had a lament on the attitude of some individuals in GA. I can't remember it exactly but it went something like this.
The very characteristics which give some people the success and wherewithal to take part in aviation are the characteristics proving to be their undoing.

Posted by: Edward Michalowski | May 12, 2011 3:48 PM    Report this comment

Good start on the data front!
Sticking with the numbers, though, we do have to consider the cost-benefit tradeoff. When a military jet goes down, the bill for damage is in the tens of millions at least. When an airliner goes down, it's readily over $100M. When a weekend pilot goes down, the economic cost is in the order of $1M. At the margin, an extra hour of training has to lower the weekend pilot's accident rate by an awful lot, to match the benefit of an extra hour of training for an airline or military pilot.
But, returning to your point, which was that "training can make a difference," what cockpit procedure changes from Part 121, that are affordable to the weekend pilot (whether Sport, Private, Commercial, whatever - flies an average of 35 hours/year), operating single pilot, could be carried across?
In other words, it's not "more training" that's needed: it's "better/different" training that leads to different behaviors.
Again, I believe safety can be sold - if it can be demonstrated. Just look at all those Cirruses that sold because of the ballistic chute...

Posted by: Finbar Sheehy | May 12, 2011 3:54 PM    Report this comment

B Noel,
One of my gliding instructors taught me to make the base to final turn "eyes in" - and glider pilots are zealous advocates of flying "eyes out."
His point was that for the few seconds it takes to make the turn, the most important things are 1) airspeed and 2) coordinated flight. You can fly the base leg and final approach using the trim to set the speed, but in the turn that won't work.
It struck me as a sensible instruction.

Posted by: Finbar Sheehy | May 12, 2011 3:57 PM    Report this comment

I havent been able to convince people to tape yarn around the aircraft for heads up AOA/Yaw information.literally 10 cents. This is a fashion and status industry, not a science industry. Like the motorcycle helmet controversy, it will require enforcement, be in at insurance,rental, owner, employer ,manufacturer, or government level.
I have reviewed the least expensive approches at a website called project festoon.

Posted by: Francis Gentile | May 12, 2011 4:10 PM    Report this comment

"In other words, it's not "more training" that's needed: it's "better/different" training that leads to different behaviors."

That's true and why, I think, SAFE pitched this as "training reform." It's two problems, really. Reforming the initial training for new pilots while reducing the accident rate among established pilots. And there's where you run into the problem, because an hour of non-standard training every two years is not going to do it.

CAST (Commercial Aviation Safety Team) billed itself as a data-driven, disciplined approach to improving aviation safety. In the airline world, you can see how that can be layered on to briefings and recurrent training required of pilots.

Not so easy in GA. But perhaps not impossible. But the data part is important.

Posted by: Paul Bertorelli | May 12, 2011 4:24 PM    Report this comment

Paul, a couple of thoughts. One is that the typical 35-hour renter actually flies at least one hour/year with an instructor because of rental checkout regulations. The other is that an instructor recently told me that when he flies with a licensed pilot to do a rental checkout, he feels he can tell how the flight is going to go by the time the aircraft reaches the runup area. That suggests he's cueing off observable behaviors (or is wrong!). I don't know what to do with that, exactly, but I thought it was interesting. I wonder if other instructors agree? If so, I wonder if they can identify what the cues are?

Posted by: Finbar Sheehy | May 12, 2011 4:40 PM    Report this comment

Finbar - First I've heard of anyone teaching that but that is what I was trying to say, maybe not so succinctly.

If you are turning / pulling g you have to have airspeed in the scan.

Correction above

You can't untangle stick force changes due to speed from due to off trim.

Should read

You can't untangle stick force changes due to g from speed change from off trim speed.

Posted by: BYRON WARD | May 12, 2011 4:41 PM    Report this comment

Also the above is why I personally don't feel like an AoA gauge will do much of anything. The lindsey accident got to the stall from a nose high energy bleed, not by increasing-g at relatively constant speed until it stalled. IAW if he had just maintained speed, the increased load factor would have been a non-issue. So, if he had looked at the airspeed gauges he had unwinding he had would have gotten the situation back under control. Adding another needle moving, unwatched, I doubt would have helped. Speculative and of course, imho.

Posted by: BYRON WARD | May 12, 2011 4:50 PM    Report this comment

B Noel - There was another reason for the "eyes inside" instruction. At altitude, in a turn, the wingtip appears to move backward over the ground. In the turn to final, because of the low altitude it appears to move forward. If the turn is late, the eye can see that the turn will overshoot, and the pilot subconsciously does three things to compensate: 1) bank more, 2) pull more, 3) step on the inside rudder (because the wing "shouldn't" move forward like that in a turn). Usually it works out okay. Sometimes not.

Posted by: Finbar Sheehy | May 12, 2011 4:57 PM    Report this comment

Seatbelts and harnesses must also be bogus, because they can't save the pilot in a low-altitude stall/spin and can't stop a pilot from flying into terrain in IMC. As for a BRS making a pilot over-confident, I'd sure like to see the data on that. Obviously, there are plenty of things already causing pilots to be overconfident, chute or no chute.

Posted by: Patrick Underwood | May 12, 2011 6:47 PM    Report this comment

Many many years since I did any GA Pvt Pilot training, but I think if one's proficiency isn't at a level that they can't make the turn from the 180 to touchdown without every looking inside, they don't know their airplane well enough and don't have the requisite sensory skills yet. Yes more training. One should practice "turn to final stalls" until they can sense and recover in their sleep. Looking inside leads to midairs, and they hurt too.

I go out every three months on a late afternoon or evening and practice ILS, GPS & RNAV approaches when few aircraft are airborne. Always to a MAP, since they are the most demanding to do well. That's close to what the OpNav requirement for instrument proficiency was. If an Instrument pilot isn't getting close to that level of training, is he/she really proficient?

Paul, with respect to recurrent training, is it possible to break out the accident rate for pilots that attend events like BPPP recurrent. Every year I go I learn something new. Every bit the equal of airline or military recurrent.

Posted by: Burns Moore | May 12, 2011 7:05 PM    Report this comment

"Paul, with respect to recurrent training, is it possible to break out the accident rate for pilots that attend events like BPPP recurrent"

I don't know about BPPP, but COPA has good data on this. For its members, the accident ratio was 1 in 157. For non-members, 1 in 59. That's pretty stark and has to do with both the training itself and the underlying attitude that got those guys to join COPA or ABS or MAPA and so on.

Posted by: Paul Bertorelli | May 12, 2011 7:34 PM    Report this comment

And as for the airspeed, AAOI and head-in, head-out thing, I really think if a pilot is taught from the beginning, he can assess coarse energy state without need for any instruments. Look out the window, sense the bank and pitch, you know the power and you can hear the airspeed.

Is this sense, I think glass is sending us backwards. There is no reason you can get in a Cirrus SR22 and fly it around the pattern with reference to the instruments. This is one of the fundamental doctrinal issues that SAFE addressed.

Posted by: Paul Bertorelli | May 12, 2011 7:38 PM    Report this comment

"I think if one's proficiency isn't at a level that they can't make the turn from the 180 to touchdown without every looking inside, they don't know their airplane well enough and don't have the requisite sensory skills yet... Looking inside leads to midairs, and they hurt too."

This is an example of overconfidence (and misplaced priorities). The physics are clear: you can't sense your AOA by "knowing your airplane" - it's like trying to fly IFR without instruments; you don't have enough data. Most people can do both, to some extent - but not reliably.

After a few hundred base-to-final turns (most with glances at the ASI), intuition says "I can do this (and anyone who can't is a bad pilot)." Intuition is wrong. Intuition extrapolates a success rate between 99 out of 100 and perhaps 50 times that (at most). But, to avoid creating a measurable contribution to fatal accidents, you have to be closer to 99,999 times out of 100,000. Pilots just aren't that good. I imagine a lot of the skilled pilots who have died - like flight instructors who appear to have fatally stalled (stall-resistant) Cirruses - believed they were that good, and were very surprised when they were not.

On whether to go "eyes in" for the turn to final, you have to prioritize. Stalling and spinning from the traffic pattern is a common cause of fatal accidents, while midair collisions during the base-final turn are very rare (although most pilots are staring at the runway, not looking for traffic).

Posted by: Thomas Boyle | May 12, 2011 8:22 PM    Report this comment

Did we ignore the cliche You Can't Fix Stupid on purpose? Could that be because, as a variable you can't control for that an acknowledgment that it is a major fact effectively ropes off quite a few accidents from any attempt to have them be prevented? I think this is in major play. If we wrote a list mimicing the You Might Be a Redneck If...You Might Be a Dead Pilot If... That list would be pretty short and statistics show the majority of the accidents would tic on one or more items. What is the downside of being honest about this?

Posted by: Michael Mahoney | May 12, 2011 8:45 PM    Report this comment

Paul, in my opinion, you can provide some answers to lowering the accident rate. You certainly gave me the tools that prevented me and my passengers from becoming fatalities. Your training methods work. Just recurrent training isn't sufficient. People need to be taught how to handle high risk situations which can kill them. Instructors need to be taught how train people for handling the unexpected. For example, an instructor simulating an engine out at 2000 ft is nothing. The aircraft becomes a glider. On the other hand, simulating an engine out at 400 feet on takeoff is critical training. The response to this situation has to be instinctive and fast. If not, a fatality occurs. The only way to create instinctive response is to experience it and practice it. Sometimes things just happen. Being prepared is essential. In a automobile when "things happen" the result is usually not fatal. You are already on the ground.

Posted by: DANA NICKERSON | May 12, 2011 9:07 PM    Report this comment

Michael, may I respectfully disagree. If you can't fly a SEL piston aircraft by sound, feel, and visual inputs you don't have enough training in that regime. There is tremendous value in being intimately familiar with your aircraft. That means long periods of slow flight, with and without the stall warning blaring, angle of bank to learn and feel the "bottom falling out" and increase in angle of attack, subsequent stall, perceive and immediate recovery. If one cannot do that, they are not properly trained. I will agree that you fly a B747, 777 or 787 for multiple reasons, but in the SEL piston world you should. I don't expect to change your mind, it's already made up, but should be food for thought for others.

For what it's worth, we taught slow flight and approach to landing stalls in Advanced Naval Air Training Command and Airline transition training in all aircraft. Students could not progress until those skills are mastered and are essential.

Posted by: Burns Moore | May 12, 2011 9:34 PM    Report this comment

Whoops, should read ...I will agree that you CANNOT fly a B747, 777 or 787 . . .

Posted by: Burns Moore | May 12, 2011 9:36 PM    Report this comment

Burns, yes, I completely agree that a decently-trained pilot will be able to do it, almost every time. But the technique used in everyday approaches has to deliver a safe approach at a 0.99999 level of reliability - which is a lot more than just "almost every time." I have no doubt that those downed-Cirrus instructors could have demonstrated by-feel approaches time after time - in good conditions, where that is what they set out to demonstrate, where that was the focus of their attention. But that wasn't what they were focused on in the seconds before they died. Sensory inputs aren't reliable enough when distracted, or in turbulence, or when change is very gradual. In turbulent conditions, the bottom falls out all the time; that's not a reliable indicator. The stall warning is a good tool (if it's working), but in a rough-air turning stall the sight of the ground above the ceiling may happen about the same time as the stall warning. The sound of airspeed decay and control mush are useful when they happen quickly in a training stall, but recognition comes very slowly in an energy bleed. However, Paul will be pleased to know that this debate has established something: one of us is wrong about this, so one of us has been mistrained - and training could be improved!

Posted by: Thomas Boyle | May 12, 2011 10:10 PM    Report this comment

Thomas agree 100%.

Burns and/or Paul why not give it a try. Fly with a safety pilot that can block instruments from your view while he can see them. Try a 15, 30, 45, 60 deg bank descending turn by feel, sound, intuition, whatever, but no reference to instruments at all. Then see what your safety pilot says about how well you wired the airspeed. I think you may be surprised.

Posted by: BYRON WARD | May 12, 2011 10:50 PM    Report this comment

There is an estimate which says half of the fatalities are burned alive. Near 50 years ago the army eliminated most post crash fires with design guides containing inexpensive modifications other than the well known fuel cells. These inexpensive details are completely ignored in modern designs. Modified designs to eleminate airleron stall first appeared in 1930's design contest and despite improvment since are ignored in modern designs. The Wright brothers flew with a single piece of yarn on a stick, and today we have NO means to directly visualize the direction of airflow around the aircraft. Most of Langwiege's "Stick and Rudder" is a direct explicit plea for mandatory angle of attack indicators. Trial Lawyers appear to be just the lazy blowhards that today's designers and experts appear to be, because the lawyers could sue this industry in to the ground with a showing of negligence to use the written knowlage available. All of us can wear helmets, fire resistant gear and tape yarn all over our aircaft right now, we don't have to ask anyone's permission or spend much money. It would take social guts to to do that because people would rather die than risk embarrasement.

Posted by: Francis Gentile | May 12, 2011 11:02 PM    Report this comment

Francis, I know that glider pilots have experimented with using yarn for AOA and yaw. It works well for yaw, but as I recall it doesn't work so well for AOA because of interactions between canopy shape and sideslip angle - and I imagine yarn will work very poorly on anything with a tractor propeller. Few of us fly biplanes with interplane struts (and those would only be useful to someone looking to the side). What do you have in mind? (I agree that some of the design details I've seen in GA look pretty scary from a fire standpoint.)

Posted by: Finbar Sheehy | May 12, 2011 11:10 PM    Report this comment

Dear Thomas & B Noel, I've been doing this (flying partial panel, with and without airspeed) since sometime in June of 1970 in my 2nd or 3rd T34 ride. It appears to be a matter of where you set the bar for personal proficiency. We all have excuses for why we won't or can't do something. We train "Unreliable Airspeed" in all the big Boeings and there is good reason. Two of the only three B757 hull loses were from that cause, and should have been manageable, but weren't.

Personally, I wouldn't sign someone off from any level of training that didn't have complete mastery of slow flight, stall recognition and IMMEDIATE recovery.

I don't expect to change anyones mind, since at some point, we all become convinced of the soundness of our decisions. But honestly guys, what if one had a bird strike or an icing event cost you your pitot static system. Could you still safely fly the aircraft?

We should train so that no one single point of failure costs us our aircraft or life. I try to train to that standard, but then again, I don't expect to change anyones mind.

Posted by: Burns Moore | May 12, 2011 11:19 PM    Report this comment

Burns, again, we're talking about 2 different things. Yes, I'd expect that a well-trained pilot should be able to manage the airplane without the pitot, after a bird strike for example, with very good odds of success. But there's a reason we have airspeed indicators: the odds are good, but not nearly good enough to make a routine habit of operating that way. A 99% chance of success in an emergency is pretty good, but a 1% chance of dying on a routine basis is unacceptable. As for stall recognition, I fancied myself pretty sharp at that until the day I saw the world turn upside down in the time it took me to break the stall. It happens. Fortunately it happened to me at several thousand feet, and I was stalling on purpose. I agree, we should train so that no one single point of failure is fatal. Operating without looking at the ASI in a low-altitude turn, leaves no backup to the pilot's all-too-human perceptions - and no recovery when it fails.

Posted by: Thomas Boyle | May 12, 2011 11:40 PM    Report this comment

"Why We Don't Care About Fatal Accidents"

Because there are experimental aircraft. 'Nuf said about assuming the risks and/or why we don't need to "care deeply" that GA accidents do happen.

Posted by: Mark Fraser | May 13, 2011 7:07 AM    Report this comment

"Burns and/or Paul why not give it a try. Fly with a safety pilot that can block instruments from your view while he can see them. Try a 15, 30, 45, 60 deg bank descending turn by feel, sound, intuition, whatever, but no reference to instruments at all. Then see what your safety pilot says about how well you wired the airspeed. I think you may be surprised."

No I won't, because I have done this as an instructor, many times. I'm teaching in a Cub at the moment, where we fly the wing exclusively without any reference to the ASI, which is hopelessly inaccurate as a AoA surrogate anyway.

Even in larger aircraft, I have routinely covered the the ASI during pattern work. The flying pilot has a tendency to fly and land fast as a result, but he is in little danger of stalling. The point is not to imagine yourself capable of magically sensing airspeed/AOA through the seat of the pants. Obviously, a well trained pilot will use *everything* at his disposal to fly the right AoA. The point is to understand *by feel* when you are approaching an out-of-bounds condition which you can confirm via ASI if you like, or just fix by unloading the wing immediately. The actual numerical airspeed is irrelevant.

Burns and I are obviously old school here. But the new school--adherence to observation of glass-cockpit precise numbers--is killing just as many people, if not more.

Posted by: Paul Bertorelli | May 13, 2011 7:26 AM    Report this comment

I consider myself fortunate to have been trained by old-school military instructors. One of them was an old Navy head who trained me in one specific technique I still use and pass on. An engine failure on takeoff drill that's Pavlovian: Engine quits, lower the nose. Now. No reference to ASI, no other action...unload the wing immediately. If you don't do that, you won't be in game long enough to consider anything else.

Same principle applies in the downwind-to-base or base-to-final scenario. If you understand the broadest brush concepts of wing loading and bank angle, you can rely more on observed attitude and less on the abstract surrogate of airspeed. And your trained-in instinct will have you react accordingly when you realize you're positive pitch in a high bank situation at low altitude.

I'm not arguing to never use the ASI as the surrogate. Obviously, you must. What I am saying is that an over reliance on it hobbles a pilot's ability to instantly understand and react to adverse physics. Perhaps some of the Cirrus accidents drive this point home.

As Burns says, maybe hard to change minds on this one. But we're going to have to do that if we want the change the culture. That mind change relates more to broad based risk awareness than to this specific issue.

Posted by: Paul Bertorelli | May 13, 2011 7:41 AM    Report this comment

Paul, thank you. That may be one of best complements I've received, being called "old school". I also have been very fortunate to have been trained by old school, "mastery is the expected result" pilots both military and civilian. And I would love to take lessons from you in the Cub. I've flown Citrabrias and tail draggin Cessnas but never the cub.

Posted by: Burns Moore | May 13, 2011 9:01 AM    Report this comment

I'm not talking about flying a pattern with asi covered or some other known emergency. Bump the speeds, fly shallow turns and bobs your uncle.

Again say typical approach, say 500 fpm descent at ref, trim hands off parallel to a road.

Then cover the airspeed, and roll into a bank to reverse 180 and roll out over the road. Try it starting at a normal pattern turn then repeat with progressively steeper banks up to say 60 deg. That was the Lindsay accident bank.

I would take a friendly wager on whether you can really keep the airspeed wired by reference to whatever outside references, sound, feel, wing whatever.

The guys in Lindsay sure couldn't. All they had to do was glance inside at the airspeed gauge and they'd still be alive imho. Also agree it's hard to change minds.

Posted by: BYRON WARD | May 13, 2011 9:09 AM    Report this comment

The point is to understand *by feel* when you are approaching an out-of-bounds condition which you can confirm via ASI if you like,

you can rely more on observed attitude and less on the abstract surrogate of airspeed. And your trained-in instinct will have you react

Paul, those of us trained in the "old school" are preaching to the choir, you have to change the minds of pilots who think life or death is determined by what the ASI says. I was taught "attitude flying", which is what you and Burns are advocating too. It's not "instinct" that tells me the nose is too high in a turn, since every airframe has physical reference points on it that allow me to determine the attitude of the wing at any time when at low altitude. I find most pilots trained in the past 20 years or so have no clue about this concept or the basic understanding that aircraft attitude and configuration ALWAYS results in the same air speed. Learn the attitude and set up the aircraft configuration the same way EVERY TIME and the airspeed will always be close enough to be safe. Maybe not exactly the number you're shooting for but close enough. AOA indicators are an excellent tool, especially when trying to land on a carrier, at night, but they only work if you look at them. There's no comparision to the airspeed indicator which, as Paul has pointed out, is notoriously inaccurate in any condition except straight and level unaccelerated flight.

Posted by: Barton Robinett | May 13, 2011 9:39 AM    Report this comment

FLASH! FAA and SAFE have joined forces to totally eliminate GA accidents. By driving the dwindling number of GA pilots crazy with still more onerous over-regulation, flying hours are now zero. Congratulating themselves at their success, they have now turned their attention to aging aircraft by developing onerous regulations which will require rebuilding airplanes every year to zero time condition. To ensure compliance, all airplanes will be required to have Mode S transponders. A new database of airplanes 'in' annual will be developed. It is felt that by doing this, the few GA pilots that might escape thru their SAFE/AFS800 "dragnet" to try to fly will be caught and processed by the TSA/CBP/Air and Marine Operations Center. Helping, CEH has scared avgas manufacturers out of the market. The poor scared huddled masses who have been holding back taking flying lessons -- for fear of dying in an airplane -- are now applying for flight training in droves at their local FBO’s. Everyone in the US will now be a pilot. Textron, dismayed by the lack of need for airplanes, has closed Cessna down. Tree huggers are happy because flying is now totally GREEN. FAA Administrator Randy Babbitt congratulated all concerned just before he laid them off as no longer needed. WILL YOU GUYS GIVE ME A BREAK ! One razor blade cut at a time, it is YOU who are killing General Aviation in the United States. Go get a job, will ya. Stop counting beans. Being born can kill you.

Posted by: Larry Stencel | May 13, 2011 11:05 AM    Report this comment

I would take a friendly wager on whether you can really keep the airspeed wired by reference to whatever outside references, sound, feel, wing whatever."

I'll stipulate right up front I won't be able to keep the airspeed wired. *Airspeed doesn't matter.* It is a surrogate for energy state and my point is, I think I can do that exercise without reaching stall AoA. I don't care what the actual AS is. We do it constantly in the Cub, not to mention most LSAs.

I shall go try it this afternoon in the Cub and report back. Maybe I'll make a video.

And Burns, c'mon down and we'll fly the Cub. Really a Champ or Citabria by another name.

Barton, you said what I meant to say when describing "flying the wing." AOA is discerned via perceived pitch angle over the nose and wing on the horizon. I said instinct, but I should have said *observation.*

Posted by: Paul Bertorelli | May 13, 2011 11:15 AM    Report this comment

Paul if you can't keep the airspeed under control you don't know your energy state. i.e. have you put the nose in the right place. I don't doubt you can easily fly a turn like that if you don't care about airspeed - just don't pull at all and let the airspeed go up 30 knots.

That's not what people are doing in these accident scenarios. There's is not question after looking at the Lindsey traces that they had the nose too high and let the airspeed decay. Again the speed they started manuvering, if maintained, would have given them plenty of stall margin even at the peak g they pulled in the end in a heroic low altitude 60 deg bank.

Posted by: BYRON WARD | May 13, 2011 11:46 AM    Report this comment

Paul if you can't keep the airspeed under control you don't know your energy state. i.e. have you put the nose in the right place.

I think I see the problem... In visual conditions, I know where the nose is because I can see where I *put* it. You know, with pitch. I know how much power I have, I know I'm level because I can *see* that I'm level.

So I know my energy state just as well as if I were using an ASI, within the required margin of error which, for this case, is to avoid stall AOA, not fly 78.3 knots for an approach.

You seem to be saying get the airspeed right first, then you know what your pitch will be. This is classic control/performance vs primary secondary theory. Old argument. But that's not what we're talking about. You're bringing a scalpel to sledgehammer throwing contest.

I am saying pitch to the right sight picture and hang the airspeed. It doesn't matter because you are, by direct observation, avoiding stall AOA. And this is the only point.

For those not familiar with the the Lindsay accident, I have posted video and audio links at the end of the blog text.

Posted by: Paul Bertorelli | May 13, 2011 12:40 PM    Report this comment

Barton's comments seem to me to be approaching a resolution of this "eyes in/out" debate. He makes the point that you need a procedure that works - highly reliably (as Thomas says). Outside the applicable range of that procedure, you need a new procedure (or you just don't go there); if you don't have one, you're down to instinct and "feel," and the track record of those isn't bad - but it's not good enough, as Thomas and B Noel say.
To Paul's and Burns' point, a combination of power setting, pitch attitudes and bank angles may constrain the flight to a safe envelope. Tom Knauff, if I recall correctly, says you cannot stall a modern glider in steady flight, turning or not, if the nose is below the horizon.
The key to safety may be to identify sets of procedures (there may be more than one for a given situation) that work, identify their limits of applicability, and teach students not just to use the procedure but also to identify explicitly when NOT to rely on it - and what to do instead (e.g., go around rather than bank to 60 degrees).

Posted by: Finbar Sheehy | May 13, 2011 1:56 PM    Report this comment

"...go around rather than bank to 60 degrees."


There is nothing inherently unsafe about 60 degrees of bank. Stalling is a function of AOA, not bank angle.

I've flown airplanes where 60 degrees of bank was standard procedure in the final turn. We'd roll off the perch abeam the touchdown point going right to 60 degrees, letting the nose fall as we pulled the airplane around the turn to the runway. (That keeps the pattern nice and tight.) But as we did that, we kept a close eye on AOA, airspeed, and power setting, while watching the end of the runway and visualizing a curvilinear turn that took us to a point in space where we could roll wings level, start pulling back the power, and descend until starting the flare. Of the four things to watch, the three most important are AOA, power setting, and the end of the runway.

Posted by: Gary Dikkers | May 13, 2011 2:17 PM    Report this comment

"I think I see the problem...So I know my energy state just as well as if I were using an ASI, within the required margin of error which, for this case, is to avoid stall AOA"

I think I see the problem, too. To me if you're claim is you have a technique that you can *put* the nose where you want it and *know* your energy state, a reasonable margin of error would be minimum PTS standards for the pattern. Is that not fair? So if you're commercially rated +/- 5, private +/-10.

If you can fly a constant power, tightening turn up to say 60 deg bank with airspeed covered within +/-5 kts all day every day you are better than the best. That's a classic, one might say "old school" even, flight test manuver called a wind up turn. Which I've seen blown many times by test pilots with 99.9th percentile stick and rudder skills. Saying average ga joe can consistently pull off something test pilots struggle with is just beyond belief, for me.

"You seem to be saying get the airspeed right first, then you know what your pitch will be."

No I'm saying you cross check your airspeed to fine tune the pitch attitude.

"I am saying pitch to the right sight picture and hang the airspeed."

And I'm saying that's been tried, Lindsay is the recent example, they said hang the airspeed and hung themselves. If the tape was running on past accidents I'm pretty confident they would show it too

Posted by: BYRON WARD | May 13, 2011 2:51 PM    Report this comment

Think about it, from a standard 1.3 you have just about 55 deg bank in level flight before you stall. If you can control your airspeed at ref you're fine at any kind of reasonable bank for low altitude maneuvering. I think low altitude stall accidents are primarily airspeed control problems. Not over-g problems.

Posted by: BYRON WARD | May 13, 2011 2:52 PM    Report this comment

And to tie it back to the original blog post, how are we supposed to train to avoid something when we don't truly know the root cause?

Only two people truly know what they were looking at and thinking in that Lindsay crash, and they aren't here to comment. Majority of accidents there's no data at all.

Posted by: BYRON WARD | May 13, 2011 2:55 PM    Report this comment

"Think about it, from a standard 1.3 you have just about 55 deg bank in level flight before you stall."

Airplanes don't stall because of bank angle. I've flown at more than 90 degrees of bank and didn't stall. Of course the airplane was descending all the way through the turn. If you have altitude under you, the fastest way to pull an airplane around a corner is at 90 degrees of bank.

If you have lots of altitude under you and the airplane is capable, the fastest way to do a 180 degree turn is to roll to 135 degrees of bank -- and pull. That's called a "slice back," and is what Paul Tibbets did over Hiroshima to get away after releasing the first A-bomb.

Posted by: Gary Dikkers | May 13, 2011 3:21 PM    Report this comment

Gary, it's also the fastest way out of blind canyon, and you'd better not be staring at the ASI when you do it. Roll and pull fast enough and the airspeed never gets out of wack, do it timidly and you'll be looking straight down before you know it.
Again the problem here is a lot of pilots have been trained, yes TRAINED, to believe "air speed is life". "If I stay away from the bottom of the ASI, I won't die". Once this concept takes root it's very difficult, maybe impossible to rip out. On any given day watch the number of pilots in C-150's, 172's,Cheokee's and the like who can't get the airplane on the ground in less than 2000 feet. The reason is always the same, the approach speed is too high because they're afraid of the bottom end of the ASI and have added a "safety factor" to the only number they remember, stall speed. You can tell them over and over that a wing will stall in any attitude and at any airspeed and they'll not their heads yes, but they really don't believe it. Back to the original precept of this thread, can training change the accident rate? Yes.

Posted by: Barton Robinett | May 13, 2011 3:34 PM    Report this comment

I did say level turn. In the context of what we're talking about it's fair. You have to be _accelerating_ downward to pull less g. That's just not happening in a base to final turn.

Posted by: BYRON WARD | May 13, 2011 3:43 PM    Report this comment

"Back to the original precept of this thread, can training change the accident rate? Yes."

I believe this, too. At least in some form. But as we've demonstrated in this thread, people will cling to their ideas no matter what. So you've got your airspeed thinkers and your non-airspeed thinkers. Interesting cultural camps.

Posted by: Paul Bertorelli | May 13, 2011 3:48 PM    Report this comment

"There is nothing inherently unsafe about 60 degrees of bank. Stalling is a function of AOA, not bank angle."

I think it's important to throw load factor in there. As a military pilot--which I think you are--pulling that kind of bank in a pattern entry goes hand in hand with relaxing the pitch a little to unload the wings and preserve energy. This kind of thing can be taught to civil pilots, too. The theory is trivial to understand, but important.

Posted by: Paul Bertorelli | May 13, 2011 3:58 PM    Report this comment

Paul, I think what we can conclude here is more than just that there are cultural camps. I think we can conclude that there is no wide agreement on how to avoid spinning in from the base to final turn. Consistently. Every time.
So the idea that training can improve matters is only potentially true, until a training approach can demonstrate real success with issues like this.
Minds might change if we saw data. Initiate two schools - the Piper and Pitts schools. Train procedures and skills. Normalize for flight experience, and observe the accident rates...
(When I put it like that, I'm pretty confident where we'll find the dead bodies. Not so sure who will have had more fun.)
It's been interesting reading - Thanks Paul!

Posted by: Finbar Sheehy | May 13, 2011 4:03 PM    Report this comment

"I did say level turn. In the context of what we're talking about it's fair. You have to be _accelerating_ downward to pull less g. That's just not happening in a base to final turn."

Noel, in what world do you live where you're not in a descending turn from base to final? Of course the nose is down in that turn. Training the pilot to know that's normal procedure and that the amount of back pressure on the stick at that point is intended to SLOW THE AIRPLANE DOWN. But of course there is a limit to how much he can pull and airspeed has nothing to do with that limit.
What's the point of a demonstration of a level turn with a 60 degrees bank at constant AS except to demonstrate that it takes a lot of power, pulls a lot of g and if you pull too hard you snap and spin. Unless you're flying a Red Bull race that's not a "normal maneuver" for any private pilot, and certainly isn't applicable to the base to final turn. And, by the way you don't have to "be accelerating" because the nose is down. You can, and in fact should be slowing down on the base to final turn, with the nose down. Try it, it works. It's a matter of balancing air speed, attitude and power. It's training!

Posted by: Barton Robinett | May 13, 2011 4:42 PM    Report this comment

"can pull and airspeed has nothing to do with that limit."

In the sense that the wing will stall at any airspeed if you load enough g on it by adding back pressure in the turn. By training the pilot learns that as bank angle increases stall speed increases therefore to avoid the stall he must avoid pulling the wing to a higher AOA or increase AS, either way will work. By training he learns that if he overshoots the turn to final he MUST NOT INCREASE BANK ANGLE AND PULL to get the airplane back in line. Of course he can increase the bank angle AND allow the airspeed to increase by lowering the nose (unload the wing) or by adding power. Training.

Posted by: Barton Robinett | May 13, 2011 4:52 PM    Report this comment

Thanks Finbar for reading my post. For 2cents of yarn and tape it is certainly worth testing. On a low wing with no strut to get a way from propblast it would be the same as tuft visualization of stall developement which is routine for test pilots, except test pilots remove the good instruments and leave us who really need it, without(go figure). There are a few people who seem to get some use out of canopy yarns despite tractor blast, the Nemisis racer I believe, and a comment by the "See how it flies" author. This is possibly since at higher speeds the propblast is not that much faster than the airspeed. Yaw alone is a factor in turn stalls so it could be a help. Longer lengths could show small angles better. There are mechanical "telltales" that could be more aerodynamically dampened, as sailors have used since the beginning of recorded history. Exact chasing of an angle of attack can cause the phugoid ossilation if not dampened by airspeed, but we are talking about the "big hunks" of stall awareness here. During the CAFE/NASA GA contest I paid a 1000 dollar deposit but was not able to rent a C-150 and put a few stickers and yarn on it, because I could not find anyone who would rent or fly the thing, the non conformity being an anethma, similar to the glazed eyes described in the beginning of this article. It is all psychology and sociology, not science.

Posted by: Francis Gentile | May 13, 2011 5:13 PM    Report this comment

you can take the ball in oil turn indicator out of its frame and clear tape it to the windshield in front of ones face. A grid pattern can be marked on the windscreen to assist in perspective perception. There is so much that could be done. I have attempted a comprehensive review at project-festoon dot com.

Posted by: Francis Gentile | May 13, 2011 5:23 PM    Report this comment

Barton, a steady descent does not unload nor does a steady climb load, many many pilots are confused about that. Do you have to g-strain in a jet climbing out at a few thousand fpm? No - it's still 1 g flight. Same same for descent. You have to *accelerate* up/down to change the load factor.

I don't care if how you chose to fly a turn gaining, losing, or constant speed but you sure can't sell me a story about how you have the maneuver locked up if you can't achieve a *preplanned* outcome. Que sera sera airspeed control to me is not evidence of superior stick and rudder skills.

My basic point is, though

1) you have to keep the gauges in the cross check. Whether it's AoA or airspeed is irrelevant because if you can't hold your target airspeed you also can't hold a target AoA.

2) Finding proper pitch reference for a steep, low altitude descending turn is extraordinarily difficult. It's even harder if the bank isn't constant. Which is, imho, a big factor in why aggressive low altitude maneuvering, whether base to final or "impossible turn" ends tragically.

Posted by: BYRON WARD | May 13, 2011 7:41 PM    Report this comment

In the Lindsay accident the g was actually *reducing* prior to the stall. Writing it off as pulled a lot of g and stalled is missing some key lessons I think.

The recovered data shows that they just did not have the flight path under control at all in the final two turns. In fact in the final turn the g was slightly decreasing. *If* the airspeed stayed constant they would still be here. Instead, their issues visualizing and controlling the flight path resulted in actually trying to climb at idle power in a 1.5 g turn. That was destined to end badly, and it did.

I don't think that's the first maneuvering stall accident that ended that way, and I don't think it will be the last, unfortunately. Especially if instructors teach an "old school" don't ever ever look at the airspeed indicator philosophy. I really believe if you could freeze them 3 seconds before the stall and ask them what their airspeed was the answer would be 15-20 knots higher then what was actually on the gauge. If they had even glanced at it, they would have seen the discrepancy. And maybe, still be here.

Posted by: BYRON WARD | May 13, 2011 7:53 PM    Report this comment

I am sorry I do not understand this discussion about eyes in or eyes out. Why not go talk to your local ag pilot (crop duster) who on a good day does several hundred maximum rate turns each with a varying aircraft loads all within 150 feet of the ground. In one year that would be perhaps 20,000 such turns. In my 12 years experience I never once looked at my airspeed indicator as I seemed to be more interested in getting lined up correctly and not hitting anything. That is way way beyond any .99999 percent success rate.

Ed Renoux

Posted by: Edward Renoux | May 14, 2011 11:22 AM    Report this comment

I started ag flying (part 137) in the early 1970s and the fatal rate was about 10 times that of GA (part 91) based on hours flown (if I remember correctly). In the late 1980s when I quit the fatal rate was LESS than for the rest of GA. The accident rate was still the same but pilots were walking away because most of the old Stearmans and Supercubs had been replaced by purpose built aircraft with 30G roll cages, well designed 4 point harnesses and seats, and foam filled fuel tanks.

Ed Renoux

Posted by: Edward Renoux | May 14, 2011 11:31 AM    Report this comment

So....to learn how to crash less we talk to the guys that crash 10 times more often???

Ag planes operate with hazards that don't apply to average ga guy. How many approach to landing accidents are there where aircraft struck wires, met tower etc vs how many are there with inadvertent stall?

How many ga accidents in any flight phase that didn't involve illegal shenanigans end with aircraft struck wires, pole etc vs how many ended as inadvertent stall/spin?

For a ga pilot doing something remotely legal the odds of an accident by hitting something in the 1 second you glanced inside and back out are nonexistant compared to the odds of inadvertently stalling. Why is is verboten to take 1 second to glance at the airspeed when everyone accepts you will be inside for a moment to look at a checklist, take care of fuel selector, mixture, flaps, gear, cowl flaps, pumps lights, radios, etc????

If you want to reduce fatal accident rates you have to start with what actually causes fatal accidents.

Posted by: BYRON WARD | May 14, 2011 12:22 PM    Report this comment

Also, you don't have to look that hard to find inadvertent stall ag accidents; your local ag pilot is not exactly immune from it.

Posted by: BYRON WARD | May 14, 2011 12:47 PM    Report this comment

But Ed...how could you possibly judge the airplane's ability to turn without looking at an airspeed indicator? You must be some kind of...superhuman. (g)

Point is, you and every other ag operator flies by visual reference and feel. Obviously, some pilots are just lost without the ASI. I get it. But they could be taught not be by instructors who aren't similarly lost.

It's not an argument for eyes-in or eyes out, really, but the suggestion that one can develop skills beyond staring at instruments to judge airspeed. Unless they don't want to...

Posted by: Paul Bertorelli | May 14, 2011 12:47 PM    Report this comment

"If you want to reduce fatal accident rates you have to start with what actually causes fatal accidents"

Exactly. It seems there is some general consensus that base to final stalls are one area that pilot technique and proficiency can reduce accidents. These turns are no different than the turn around at the end of the field that an ag pilot does many times a day (presumably successfully). Why not talk to someone with many hundreds of times more experience in this phase of aircraft operation (other than the fact that they smell bad)?

>"Why is is verboten to take 1 second to glance at the airspeed when everyone accepts you will be inside for a moment to look at a checklist, take care of fuel selector, mixture, flaps, gear, cowl flaps, pumps lights, radios, etc???? " Those are very good things to do on downwind but the turn to final is probably not the optimal time to do them.

Ed Renoux

Posted by: Edward Renoux | May 14, 2011 1:25 PM    Report this comment

Paul's original question was whether training could reduce fatal accident rates. He followed that up by suggesting that if we adopted the approach taken by Part 121, we could reduce the fatal accident rate. I don't know, but this "eyes in/out" debate has clearly touched a nerve. I'm not sure why there's such resistance to the idea that watching the ASI during the turn to final, as a standard procedure. Perhaps it's a fear that "they're going to take all the joy and art out of flying by making us fly like robots," or a notion that "anyone is isn't good enough to do it by feel, shouldn't be flying at all." Perhaps it's "I do it by feel - and if I admit that's not reliable, then I'm taking risks I didn't realize I was taking - and I don't like that." The debate strikes me as a variant of the "art vs science" debate that goes on in cooking, for example, or medicine. To Paul's point, though, the Part 121 operators have answered this debate, definitively: if safety is the goal, train procedure (science), not feel (art).

Posted by: Finbar Sheehy | May 14, 2011 4:45 PM    Report this comment

Paul also noted that he can fly a Cub, or an LSA, entirely by feel. I used to fly hang gliders entirely by feel (I didn't have an ASI). But in these lightly-loaded aircraft it's common to fly the approach well above the stall speed at least until short final (LSA approaches are generally flown at 60-65kt, while their stall speeds are generally 40kt or less - and the same is true of sailplanes). Hang glider approaches, too, are typically flown at 1.4 - 1.5 times stall speed. If the airplane has half-decent pitch stability, trimming it for 1.4 or 1.5x stall speed is going to mean that you have a procedure that should keep you out of trouble as long as you don't keep pulling on the stick. That may feel like "art," but it's not really; it's a procedure. It creates a large margin for error, and thereby avoids any need for precision. Not to say that Paul, Edward and others can't fly more precisely by feel: but the procedure works pretty well even for those of us who can't.

Posted by: Finbar Sheehy | May 14, 2011 5:03 PM    Report this comment

deflecting a wingtip airleron down makes that part of the wing more likely to stall than the other wingtip. it is an inherent design flaw. The same effect causes adverse yaw which makes instrument flying much more difficult, an inherent design flaw for which solutions have always been available. The germans rotated thier instruments so all needles were upward for landing configuration, so inside outside, buffer airspeed and instructor knowlage are built in. The russian artificial horizon did and does not tumble at unusual attitudes at US built models did and do, which is ugh.. exactly when you need it most. Dragging aviation kicking and screaming is how progress will come, or a new start with new people on the outside.

Posted by: Francis Gentile | May 14, 2011 5:19 PM    Report this comment

Actually, stalls happen in all phases of flight related to departures, arrivals, approaches and so forth. The downwind to base or base to final turns are common stall points, but by no means the only.

Posted by: Paul Bertorelli | May 14, 2011 6:03 PM    Report this comment

It seems to me that one of the things that instruction could do better is to get pilots to make conscious choices about how they're going to manage airspeed and alpha during various phases of flight. There are several ways to do it; some are more appropriate to one type of aircraft than another, or to one type of circumstance or another. The 35-hour-a-year weekend pilot may want to consciously decide to fly "by feel" but limit "by feel" maneuvering flight (lazy 8s and maybe even the occasional roll) to not below 3,000ft, "by feel" normal flight to not below 1,500 ft, and have belts-and-braces by-the-numbers procedures below 1,500ft. The hotshot pilot who flies every day may fly "by feel" approaches in a Piper Cub at 1.3x stall, but by-the-numbers in the B777. The ag pilot may fly entirely by feel because loss of situational awareness is a bigger threat than anything else. But always with a conscious decision about which procedure to use, when, and why.

Posted by: Finbar Sheehy | May 14, 2011 6:11 PM    Report this comment

"One of them was an old Navy head who trained me in one specific technique I still use and pass on. An engine failure on takeoff drill that's Pavlovian: Engine quits, lower the nose. Now."

If I remember correctly, to create the Pavlovian response, Paul would first distract the pilot by saying something like "watch that airplane" while pointing out the window. When the pilot was sufficiently distracted he would pull the power back. If the pilot's immediate response wasn't to drop the nose, the pilot heard a rather loud retort of "PUT THE %^&% NOSE DOWN NOW!!!". This may sound funny but the training process wires the brain so when it happens for real, the response is immediate and lifesaving. When the engine failed on me at 500 feet, I heard the same "PUT THE %^&% NOSE DOWN NOW!!!" in my headset. Maybe instructors have bring back some of the old Navy training techniques. Maybe instructors are being too nice.

I will have to say that I had to resist the very strong urge to pull the yolk back as I watch the ground come up at me rather fast.

Posted by: DANA NICKERSON | May 14, 2011 9:13 PM    Report this comment

"Pavlovian response"


There is a lot to be said for that. When I was learning to fly in the USAF, I had an instructor pilot (IP) who demanded that every time I rolled out on final I reach over and make sure the landing gear lever was down, check the indicator lights, and then say, "Gear down, three green." If I didn't, he would start banging on the stick until I said it. It didn't take long to burn that into my brain.

Several months later I was landing at night at Phu Cat AB, Vietnam. At the point where I should have been putting the gear down, a Viet Cong 12.7 mm machine gun opened up and a string of tracers came floating up toward my airplane. I jinked a few times, and then continued my pattern. I rolled out on final and just as my IP had taught me, reached over to check the gear handle, and surprise, surprise, it was still up, The anti-aircraft fire had distracted me and I had forgotten to put the gear down.

I pushed the handle down, the gear came down and landed. As I rolled out on the runway, I silently thanked my IP for drilling such good habits into my brain.

Posted by: Gary Dikkers | May 14, 2011 9:35 PM    Report this comment

Gary, what airplane was than in Vietnam? O-2? OV-10?

I had a similar experience landing in a intense but dry snow storm. The airplane was so charged with St. Elmo's fire that a large blue streamer jumped from a BNC antenna connector on the pedestal, ran down down my leg and went out the rudder pedal. I was just inside the outer marker.

I was so distracted I didn't remember having put the gear down, but the habit of saying "final landing check, gear down," saved me.

Another habit that may help is that on takeoff, I always say to myself, this time, the engine really is going to quit and I treat the takeoff as though it's an engine failure drill. This can reduce those one or two (or more) seconds of denial that allow the airspeed to bleed off, increasing the chances of a stall.

Lot of ways to skin the cat with built-in habits that can become instinctual. Some of these relate to things like sight picture over the nose and wing.

Posted by: Paul Bertorelli | May 15, 2011 6:19 AM    Report this comment

Nothing like a 24 hour airplane ride from Singapore to crystalize one's thoughts on a subject. I think we're seeing a discourse on pilot basic foundation for flying. When a current new pilot begins flight training, he's taught to trust his instruments as a foundation for future instrument training and because the FAA requires basic introduction. When I enter the VFR traffic pattern I'm always around a 100 KIAS, plus or minus a couple. Why? Because that's where the airplane feels good. Controls feel tight and responsive, have enough energy to make the runway if the engine quits or emergency maneuver if 1gets in the way. Do I look or care if it's 95 or 105? No, basic performance capability is the same. I'm watching for traffic, drift, closure, weather and all the things that are really important. Turning off the 180 does one need to look at airspeed for precise adjustment? No, the airplane will tell you. What's import is drift, closure, traffic and how the airplane sounds and feels, and thats' all happening outside the aircraft. If it feels like you're settling, the runway or horizon is rising or things are getting quiet, it doesn't matter what your airspeed says, you need power.

Posted by: Burns Moore | May 15, 2011 7:50 AM    Report this comment

These "visceral" skills teach a pilot basic aircraft control. I think it would reduce a lot of RLOC accidents, stall/spin landing accidents, make the Missed Approach a much safer process and even make the PIC much more aware of CG issues by "telling" him that the a/c just doesn't feel right. As an instructor it's really easy to say " . . . well your airspeed is off by 3 knots" thereby giving the impression that they're really on top of the instructing process, but I just don't think it's all that important. Pattern control, traffic interval, not running into someone, aircraft response are all more demanding of attention. The topic of AOA gauges as the "perfect cure" has been made, and I used and depended on them to land on carrier, but we didn't really use or understand them 'till we went to boat, not as an initial training device, and they were also remote mounted up on the glare shield so they were part of line up scan. We're not flying to that level of precision and let's not make them into another crutch.

Posted by: Burns Moore | May 15, 2011 7:51 AM    Report this comment

I started flying in 2007. I was horrified to learn just how dangerous GA operations really are--you call it "the big lie," I call it "our dirty little secret." I have one very big advantage over most pilots--ninety percent of my flying is with my close friend/plane partner. I've got 400 hours, but probably 650 if we were to count co-pilot hours. Two pilots are a hell of a lot safer than one.

Pilots are risk takers by nature even if you like that label. Most of us are pretty smart, but not all of us have good judgement, hence the idiot who landed on the beach in Rockaway Beach near JFK recently.

Posted by: Thomas Reilly | May 15, 2011 8:54 AM    Report this comment

"It's not an argument for eyes-in or eyes out, really, but the suggestion that one can develop skills beyond staring at instruments to judge airspeed. Unless they don't want to..."

You shouldn't stare at airspeed, I never said that. I said you should cross check it.

Problem is to me these accidents happen, really all pilot-error accidents almost by definition, because the pilots perception precesses away from reality.

Your only definition of success in "flying the wing" seems to be haven't inadvertently stalled = success. That's nice except you're setting the max allowable split between reality and perception at a fatal accident.

Since since low altitude stalls are generally fatal first time out (except maybe ag operators) pretty much everyone alive can claim victory in the "flying the wing" technique.

To me to prevent fatal accidents we need to cage perception to reality before an accident happens, not scratch our heads over a smoking hole and say "well I guess he wasn't flying the wing right". So I say pilots should be taught always to use the tools they have to verify their perceptions *are* reality.

If you think your stick & rudder skills can put the nose right where it needs to be to hold airspeed, a quick glance down tells you whether you're right or wrong. If you think your seat of the pants can tell you whether you are coordinated or not, a quick glance down confirms. Etc Etc

Posted by: BYRON WARD | May 15, 2011 9:01 AM    Report this comment

"It's not an argument for eyes-in or eyes out, really, but the suggestion that one can develop skills beyond staring at instruments to judge airspeed. Unless they don't want to..."

Paul I know you can fly a cub around the pattern on a nice day by feel; btdt.
Teaching that to students that are then going to fly a cirrus, bo or equivalent at night, marginal VFR etc, or try an impossible turn, or tightening turn to base is just feeding them false confidence imho and they aren't going to find it out that their feel wasn't right until they are in the bottom of a smoking hole.

To me I would say give me some objective measures, aoa, airspeed, load factor, margin to stall, whatever and some tolerances. If you can can train a student to always stay "in tolerance" by feel in a typical high perf airplane in some classic accident scenario setups well then I'll come down there and get schooled. I'm not too interested in something that's floating around on self-reported feelings and perceptions - if it works it's measurable.

Posted by: BYRON WARD | May 15, 2011 9:45 AM    Report this comment

"if it works it's measurable."

You keep misconstruing the point we're trying to make here, Noel. You seem to think the point is we try to teach people to fly without reference to instruments when what we are really saying is visual perceptions and aircraft feel tell a properly trained pilot when he's getting into the red zone and either needs to act immediately, check instruments or other sources for additional data...or do something. It is one cue of several. It can be a Cub or a Cirrus or Bonanza. Perhaps not an A310.

Don't construe it as anything else. You will never have measure of which is more effective because nobody is doing that research. In any case, it is not either or but a totality of understanding all aspects of flying an airplane. In the Cirrus fleet, you do have a coarse measure of one thing: Glass panels, parachutes, harnesses, 26-G seats, stall resistant wings, and factory training have not, fleet wide, produced a lower accident rate, for reasons unknown. On the fatal side, they have produced a higher fatal rate and a higher percentage of accidents that are fatal.

I don't know what that proves, but I know what it doesn't prove.

Posted by: Paul Bertorelli | May 15, 2011 10:05 AM    Report this comment

One of the reasons for the terribly high drop-out rates of student pilots is that they come to understand just how dangerous flying can be. A lot of these drop-outs probably shouldn't be flying. Flying is a life altering experience, meaning that you must always be learning if you want to be safe. Flying is NOT a hobby, it is a profession. How can a guy who rents a Cherokee once every 6 weeks be safe?

Posted by: Thomas Reilly | May 15, 2011 10:58 AM    Report this comment

Paul, your definition of "properly trained" is ex-post. If there's a smoking hole, we conclude the pilot was not properly trained. If there isn't, then we don't know yet. If we're trying to prevent the creation of smoking holes, we need a metric that allows us to determine whether one pilot's technique is more likely to create a smoking hole than another pilot's, and then train pilots to use the safer technique. Further - and I think this is important - if multiple techniques produce good outcomes, we need to focus training on the techniques that are a) easiest to learn, b) least affected by the variability of pilot aptitudes and c) least prone to degradation through lack of practice.
Your own gear-check is a good example. You should, on most small aircraft, be able to "feel" whether the gear is down: it sounds different, and the drag changes the aircraft's pitch trim and speed response to pitch change. If you pay attention, you can tell with you eyes closed whether the gear is down, on many aircraft. But even someone like yourself - many hours, very current - can be distracted and consequently not notice. Your check procedure - at THIS time you ALWAYS not just mentally but CONFIRM gear down - caught the gap between perception (I "know" the gear is down) and reality. Same idea: if stalling at this moment is especially dangerous, don't rely on perception, which may be warped by distractions: CONFIRM the airspeed/alpha.

Posted by: Finbar Sheehy | May 15, 2011 11:09 AM    Report this comment

"One of the reasons for the terribly high drop-out rates of student pilots is that they come to understand just how dangerous flying can be."


I'll counter and say the dropout rate has more to do with the realization that learning to fly is both expensive and time-consuming. It takes lots of hours to gain a high level of proficiency, and even once you have that proficiency, it takes even more hours (and lots of dollars) to maintain a satisfactory level of currency.

The promotions for flying schools like to say, "Anyone can do it." But the truth is different. Anyone who can think in three-dimensions and has a relatively good hand-eye coordination can learn, but then there is the issue of the time and money that learning to fly consumes.

Posted by: Gary Dikkers | May 15, 2011 12:46 PM    Report this comment

Gary, We agree. I'm sure you'll also agree that we need to find these capable pilot candidates and get them in the air before we lose more "ghost town" airports. We can't reasonably expect various municipalities and other such entities to keep these facilities open when hours go by without a single takeoff or landing.

Posted by: Thomas Reilly | May 15, 2011 1:51 PM    Report this comment

A third world country is probably the place to design and develop a ten or more times safer GA system. Few prexisting ideas to overcome, minimal regulatory and liability risks, actual non-hobby need for inexpensive GA, and an illiterate and routinely intoxicated ab-intio pilot population whose overwhelmingly superiour safety record will shame the ego driven US centric GA community. Perhaps an island chain.
Imagine your plane is stolen with your daughter in it by a drunk non-pilot, how would you change your plane design to ensure a safe outcome? Blame and train has gotten us nowhere.

Posted by: Francis Gentile | May 15, 2011 3:39 PM    Report this comment

I agree with the concept of honesty - we shouldn't delude ourselves or the public. But my belief is that the relative risk of GA is downplayed in part because the general public sees it as even more dangerous than it is.

I did some statistics for a presentation several years ago, taking a somewhat different approach. Namely, presenting the risk vs. miles traveled. The relative risk of flying GA by this analysis was not quite as bleak. Data was from the 2007 Nall report, and 2006 traffic accident statistics from the US Dept. of Transportation pulled from the Insurance Information Institute (www.iii.org). Assumptions were: "Personal" GA has a 50% higher accident rate than all GA. Average cruise speed for small airplanes comprising "personal" GA = 140 statue mph.

Fatal accidents per 100 million miles traveled in 2006 were:
Automobile = 1.28
Small Airplane = 13.5
Motorcycle = 49.4

I don't see a hue and cry for more regulation and hours of recurrent training for motorcycle riders out there. And I am often surprised by friends and family members that think nothing of riding a motorcycle, but would not set foot in a small airplane.

Posted by: Will Rueger | May 15, 2011 6:13 PM    Report this comment

I guess the background assumption is that more people should fly. Presumably to reduce costs and availability versus an actual declining population. If all motorycycle riders suddenly started flying, the accident rate might drastically increase. Or one could attribute the reduction in accidents since WW2 to the idea that most of the people likely to kill themselves flying have already done so. And the people who are left are getting old and dying naturally. Young people somehow still know Patsy Kline died in a prop plane. I know more familys devastated by a GA deaths than automotive deaths, despite the smaller GA population. So you probably could improve safety by reducing the population, but I know of stall spin approach accidents by highly trained test pilots, cuz everybody lock the keys in the car once in thier life. Look up Stapp's career, who popularized "if it can go wrong it will" "all accidents are design flaws" and rode the rocket sled himself. We need more people like him.

Posted by: Francis Gentile | May 15, 2011 7:01 PM    Report this comment

I don't see a hue and cry for more regulation and hours of recurrent training for motorcycle riders out there."

Perhaps because you aren't looking for it. In Florida, for example, in 2008, the state passed a statute requiring a basic motorcycle safety course, plus the license test.

And of course, helmet law debates come and go. Insurers and hospital associations trot in their helmet safety stats and groups like ABATE deny the findings. And so it goes. In Florida, the statute allows no helmet if you have 10K in medical insurance. Under 21 riders, I believe are required to wear helmets and maybe even have "under 21" on the plate. I've seen a few of those. The states have a patchwork of laws and safety requirements.

Other than SAFE, there is no significant push of any kind to reduce the GA fatal rate, which takes us full circle: I don't think the community itself cares enough to make the effort.

Posted by: Paul Bertorelli | May 15, 2011 9:09 PM    Report this comment

I've had motorcycles all my life, but have never had a motorcycle license--it's not going to keep me safe. I continue to train as a pilot and take care on my motorcycle because I understand the risks. Continuing Ed is not going to solve anything, but conversations like this and programs like those offered through FAA, AOPA, etc. do, in fact, make a difference. I have been reading NTSB reports since I started flying because I want to understand why pilots fall out of the sky. Community pressure, our community, can influence our behavior.

Posted by: Thomas Reilly | May 15, 2011 9:29 PM    Report this comment

"...what we are really saying is visual perceptions and aircraft feel tell a properly trained pilot when he's getting into the red zone and either needs to act immediately"

Yes and I'm saying that argument in the end is another way to restate the tired standby that any pilot that crashes was clearly not properly trained. "old school" indeed. Nice for the ego of those of us still around, not so much for moving stats.

Anyway it seems like the closest we have to an objective measure is the red zone. Seems reasonable to define the red zone as where the stall warning should sound? I can set up scenarios that will falsify that by feel and visual, without any reference to instruments, you can detect the red zone.

But I guess at some point common sense has to kick in; pilots still inadvertently stall with a horn that announces entry into the red zone, in fact even with audio "stall stall stall", stick shaking and then push when you've ignored all that and treaded across the line anyway.

Maybe we should try to figure out why and how pilots perceptions got so out of alignment with reality in these accident scenarios. Until you figure that out any fixes are just guess and hope.

Posted by: BYRON WARD | May 15, 2011 9:46 PM    Report this comment

BTW the ntsb database has plenty of examples of ag inadvertent stall/mush/spin. I thought we established they had the red zone awareness mojo. Why do you think those accidents happened? Training again?

Posted by: BYRON WARD | May 15, 2011 9:54 PM    Report this comment

"Same idea: if stalling at this moment is especially dangerous, don't rely on perception, which may be warped by distractions: CONFIRM the airspeed/alpha."

Finbar, I agree completely.

I'm fairly sure I know how the perception warping goes down in at least some low altitude maneuvering accidents, like Lindsay, but I'd really like to be able to come up with a training exercise that - safely - demos it. I think that would be a very powerful teaching tool.

Posted by: BYRON WARD | May 15, 2011 10:36 PM    Report this comment

Lots of comments; too many to take the time to find who wrote it, but there was a comment that we don't install parachutes on all planes because we are overconfident, too cheap, etc. There is another reason. A parachute may give you a soft(er) landing, but removes your control over where it is. It could be in water, power lines or a busy interstate. I think I'd rather find my own landing site.

Posted by: John Worsley | May 15, 2011 11:15 PM    Report this comment

I'm doubt anything has been left unsaid, but if the stall warning horn is being used as last line to stay out of the "red zone" there hasn't been much communication, but perhaps I misread the intent. I think Paul initiated this thread to create some dialog and increase awareness and introspection on behalf of the GA light a/c community. It would have been encouraging to see inexperienced pilots ask more questions on technique and concept. Rather, some argue and justify a position that Paul, with a wealth of experience in this specific area, has shown is clearly not working. If the Cirrus accident mentioned earlier and Cory Lidle (and Thurman Munson) and his instructor really understood their aircraft's performance capabilities, they may still been with us.

Posted by: Burns Moore | May 15, 2011 11:20 PM    Report this comment


"According to the pilot, he was on his fourth aerial application of the day, with a load of 430 gallons of fungicide. He departed to the northeast and turned west. He said everything was normal until he got closer to the pecan orchard that he was planning to spray. As he got closer to the trees, the airplane began to "mush" and started shaking. He said he "hit" the dump lever, but not hard enough to activate it. The airplane then descended and impacted the trees. Examination of the wreckage did not reveal any preimpact mechanical malfunctions, nor did the pilot report any. The pilot stated that the engine was operating continuously, but he was not able to clear the pecan trees."

Ag pilot, 15,426 hours total time, 14,745 make/model, 355 last 90 days

Posted by: BYRON WARD | May 16, 2011 8:57 AM    Report this comment

You and Paul and that camp are trying to sell a story that you can teach a pp in 40 hours to aquire a feel good for life for all airplanes that a guy flying one airplane 15,000 hours doesn't have. I'm not buying it, sorry.

Posted by: BYRON WARD | May 16, 2011 8:58 AM    Report this comment

Here's another one:

"The pilot was applying dry chemical to a field. After refueling and reloading dry chemical, the pilot departed for the second flight of the day. A rancher, who owned the field the pilot was spraying, was in contact with the pilot via hand-held radio. The pilot and rancher discussed the terrain and application flight paths. The pilot elected to start the application at the top of a ridge and fly downhill. The pilot reported that the downhill pass felt okay, and he was going to make the second pass up slope. The up slope application pass was successful. The rancher then observed the airplane climb to about 500 feet above ground level over a saddle gap between two 6,300 foot ridges, and the airplane began a turn back to the field. Subsequently, the pilot reported over the radio that he was " in a bind and was going to crash." The rancher then observed the airplane "pitch over abruptly and enter a tight spin." The airplane made three turns prior to impact with terrain. The airplane was destroyed by post-impact fire. Examination of the airplane and engine revealed no anomalies."

Ag pilot, 27,000 hours total time, 5,000 make/model, 300 last 90 days

To me you have to come up with theories that include those accidents. Everyone dead clearly wasn't trained right is just rationalization & that's the thinking that got us here.

Posted by: BYRON WARD | May 16, 2011 9:00 AM    Report this comment

Dear Mr. Noel, I don't think anyone stated that any one parameter of aircraft operation would make one "bullet proof". As a graduate of the Naval Post Graduate School in Aviation Safety (and Accident Investigation), I know that there are an almost infinite number of variables that apply to any particular event. Ag flying is a much less forgiving environment than what I choose to do at my age and risk tolerance level. In fact I lost a flight school classmate who flew F4's in Viet Nam and yet died in an Ag accident. I will ad that I would not sign a prospective pilot for solo who could not master slow flight, but then to be honest I don't instruct anymore, and would only do so for a family member where time, cost and ego were not a factor.

Good Luck and Best Wishes,

Posted by: Burns Moore | May 16, 2011 9:58 AM    Report this comment

"Anyone can do it." But the truth is different. Anyone who can think in three-dimensions and has a relatively good hand-eye coordination can learn"

I was trained by an old school, exmilitary pilot with over 13,000 hours of instucting time. He told me at a very young age that not everyone can fly an airplane and we shouldn't sell airplanes while we pretend they can. That may be why only about 2% of the population has any interest in learning to fly. And may answer the question about why so many who start, quit. As an Architect I can assure you most people can't think in three dimensions. That's an important difference between being able to fly an airplane and drive a car and the fundamental reason, to go back to the very beginning of this thread, that comparisons to auto accidents are meaningless. The underlying skill sets required are mulitple degrees of difficulty apart.

As Mr Noel has amply demonstrated, there are two camps in aviation. Those who think the airspeed indicator will save your life, and those who understand that there is a real skill, developed only by training and experience, that allows a pilot, even in the dark, to know there are some things we just don't do. And yes I understand how the inner ear works and that we can't fly in the clouds without instruments. The point is, training a student to depend on instruments and NOT TRAINING him to learn to "feel" the airplane leaves him dangerously unprepared for the real world.

Posted by: Barton Robinett | May 16, 2011 11:00 AM    Report this comment

"In fact I lost a flight school classmate who flew F4's in Viet Nam and yet died in an Ag accident."

There is no room for error in ag flying. Ed Renoux said he made ~20,000 low-altitude turns a year. He could be 99.9999% perfect (meaning only one error out of 20,000), and that single screw-up could bite him. (Which must explain the urban legend that crop dusters can't get life insurance, or that it is impossibly high.)

Posted by: Gary Dikkers | May 16, 2011 11:39 AM    Report this comment

All this is reminding me of something my hang gliding instructor used to say - "You could fly in these conditions / make that approach / do that maneuver, if you were good..." He always said it with an emphasis on "good," drawing out the word a little. He never seemed interested in trying anything that required him to be "good."

Posted by: Finbar Sheehy | May 16, 2011 12:03 PM    Report this comment

I viewed several documentaries on Aviation this weekend. The Nova show about Louis Bleriot and then one on the Wright brothers were very insightful. Basically "safety" in GA is a NEW IDEA. Flying was created by people who accepted the risks.

Ironically, we probably would not have survived as a species if it was ingrained in us to do nothing unless it was 100% safe. Thank goodness for adventurers, inventors, and risky people who gave us GA in the first place...

Posted by: Mark Fraser | May 16, 2011 1:15 PM    Report this comment

Mark that's an interesting take on Risk vs. Safety in GA. There was a tragic Risk vs. Safety demonstration last month in OK involving a beautiful family of four (two young daughters) in a Baron. Very well covered on a Beech website, but right now I"m not in an intellectual or emotional frame of mind to debate abstract necessity of risk taking. I'm sick of reading about perfectly good airplanes flown into the ground (with innocent pax) by pilots unprepared for the challenge.

Sorry . . .

Posted by: Burns Moore | May 16, 2011 2:00 PM    Report this comment

"Those who think the airspeed indicator will save your life"

Indeed it will. There are really only two ways for a low altitude manuevering stall to go down, either the pilot knows his airspeed and pulls to much g for that airspeed, or the pilot knows his g (roughly) and lets the airspeed get out of control. Given that standard ref speeds, and typical climb speeds, give you more than 1.5 g protection, beyond a 45 deg bank level turn, I'm skeptical that the typical low altitude stall is a known-airspeed-but-over-g scenario.

I think far more GA accidents fall as relatively low g, but poor airspeed control. Exactly like Lindsay. Given that cross checking airspeed will highlight rfn that your feel is out of wack, and the only reason advanced not to cross check seems to be machismo*, and the consequence of wrong feel is fatal, I'll keep backstopping my feel with the airspeed indicator thankyouverymuch, and advocate the same.

*for GA joe. Ag is between a rock and a hard place

Posted by: BYRON WARD | May 16, 2011 2:27 PM    Report this comment

"..those who understand that there is a real skill, developed only by training and experience, that allows a pilot, even in the dark, to know there are some things we just don't do."

Seems like if that was true than stall/spin accidents would correlate better with skill (ratings as proxy) and experience then they actually do. IAW, why are commercial pilots statistically most represented in stall/spin accidents, followed by private, followed by student. It seems the reverse of what would happen if your theory was true. Even harder to fit into your theory is the statistic that 90% of the stall / spin accidents during training happened with an instructor on board, while only 10% happened solo. For example, a simulated engine out in a PC-12 that ended fatally with an ATP providing the instruction. Was he just missing this skill?

I guess to me I would like this idea a lot better if someone from that camp could propose a way to objectively determine if someone has that skill prior to them dying. Otherwise it's as Finbar says, ex-post & not a realistic solution to reduce fatals.

Posted by: BYRON WARD | May 16, 2011 2:27 PM    Report this comment

"Ironically, we probably would not have survived as a species if it was ingrained in us to do nothing unless it was 100% safe. Thank goodness for adventurers, inventors, and risky people who gave us GA in the first place..."

I think there is a huge difference between people who take risks to advance society and people who add blatently unnecessary risk to everyday life out of ignorance.

Posted by: BYRON WARD | May 16, 2011 2:34 PM    Report this comment

It was Wilbur Wright who famously said about aviation, "If you are looking for perfect safety you will do well to sit on a fence and watch the birds."

Posted by: Gary Dikkers | May 16, 2011 3:12 PM    Report this comment

Burns & B. Noel: You can't stop stupid in airplanes, cars, trucks, boats, or any activity . No law ever written can do that. That's WHY there will allays be stupid deaths and innocent collateral damage.

I was talking about cautious and calculating aviators (Like the Wrights) who accepted that even with the best information available. They accepted the risks that came from learning how to fly. My guess is that the dangers never changed at all; just a pretty face was put on it to keep students coming. What students have to learn is that the same dangers exist today as they did 100 years ago. They also have to accept that death is more likely if anything does go badly.

Posted by: Mark Fraser | May 16, 2011 9:26 PM    Report this comment

Is it not comparing apples and oranges comparing GA fatalities with car fatalities?

How fast does the average car drive in the US compared to the average plane? If one wants to look at fatalities it would be better to compare fatal car accidents for cars driving faster than 200 km/h to fatal GA accidents. I suspect then the statistics would show less of a difference between cars and GA (perhaps even the opposite trend as anyone driving more than 200 km/h in the US is breaking the law and clearly reckless).

Perhaps total number of accidents per 100'000 hours fatal or otherwise would be a better indicator of the safety of GA than looking solely at accidents causing fatalities. It is mechanically much harder to prevent death in an accident at 200 km/h compared to one at 50 km/h.

Posted by: Geoff Engelbrecht | May 17, 2011 5:46 AM    Report this comment

"As Mr Noel has amply demonstrated, there are two camps in aviation."

Not sure why this "two-camp" idea persists. I guess the point I was trying to make--and keep trying to make--is that survival is a totality of factors, a combination of knowing what will kill you by second sense and applying all that you've got to keep it from killing you. Or, the inverse, applying it in the name of skill and proficiency.

It is, however, absurd to think that any training system will always produce perfect results and when perfect results are marred by an accident, that's ipso facto proof that whatever you're arguing against obviously didn't work.

Training is, at best, a set of probabilities. You're equipping yourself with the ability to reduce the probability of a bad outcome while increasing the probability of a desirable one.

"Either or" leads to the kind of rigid thinking that works against improving the probability, in my view.

Posted by: Paul Bertorelli | May 17, 2011 6:21 AM    Report this comment

Paul the problem I have with the flying by sight and feel is it demonstrably doesn't always work.You really can't take sight pictures learned in florida, and transfer them to, say, tabernash.

The right place for the sky-ground line with upsloping terrain is different. Sight pictures learned at altitude are wrong low level. Force feedback through the controls changes significantly with cg. Double the response for the same pull at aft cg as forward isn't unusual for traveling airplanes

"Just knowing" is not humanly possible. The reality of how pilots fly is constantly trying something, seeing an error, and adapting. Rinse and repeat. Mostly that process goes on subconsciously. Sometimes it bubbles to the surface as a little pitch bobble or similar. We're wired to adapt and very very good at it.

And obviously most of the time it's sucessful.

But sometimes it isn't. And some of those times are fatal. So teaching feel and visual picture actually has the potential to set pilots up to fail in classic accident scenarios. An equivalent of "turn base over the barn", it gives skills that work so well for the familiar but crumble in the unfamiliar.

If SAFE or whoever wants to cut accident rates they're going to have to apply a little science about how airplanes fly and how pilots control them and translate into teachable terms. When is the barn there? When is the barn actually a mirage?

Posted by: BYRON WARD | May 17, 2011 4:36 PM    Report this comment

I would like to echo the angle of attack importance mentioned earlier. Even without a separate AOA meter, for a specific CofG and flap/engine setting, the stick/column indicates angle of attack. Go up, throttle back and find the column position that the plane stalls at with wings level. Do the same at 30, 45 and 60 degrees of bank. Speeds will be higher but the stick will be the same. Now find the position for minimum descent rate with wings level. If you ever have to turn with engine off, your (stable) airspeed should have the stick position for minimum sink in the turn. You will not stall and you will lose minimum height (for that bank angle). When convinced... pass it on.

Posted by: Ray St-Laurent | May 18, 2011 12:12 PM    Report this comment

Ah, Ray, an elevator angle of incidence indicator is a missing instrument, and icon based links to other instruments to put 'fly by table look-up' in the instrument panel, like IFR. It is on the list of cheap improvements at project-festoon dot com. Long ago I showed these ideas to John King, and he has become a proponant of 'scenario based training' and he is admitting danger as a spur to improvment. I probably can't take credit for John King. Efforts to convince instructors at the Edwards test pilot school did not fail technically, however thier private instruction was an escape from the statistcally successful military regimented training, so my regimen was not pleasing. They felt thier accident stats would be better than average, but I think thier sample set is too low. This may the problem with instructors.

Posted by: Francis Gentile | May 18, 2011 4:47 PM    Report this comment

Add your comments

Log In

You must be logged in to comment

Forgot password?


Enter your information below to begin your FREE registration