Was it really pilot error – or was it something else?

Not too long after I started working at FLYING, we started running a monthly column called Pilot Error. As best I recall, we had to clear every column with the legal department. They thought the title of the column implied that we were passing judgment and they wanted our assurance that an error on the part of the pilot caused the accident.

The way to fix that was to punt. The title of the column became Aftermath which is about as far from judgmental as you can get.

I thought a lot about that then, and I still think about it. The oft-quoted statistic is that about 85-percent of the accidents in private aviation are caused by pilot error. I always had the nagging suspicion that what that really means is that in 15-percent of the accidents they can find cause with something other than the pilot so that just naturally means that the rest get blamed on the pilot instead of on some failure or fault in the training and regulatory system. Pilots don’t err on purpose, though, they err because they don’t know better.

Cessna crash
Pilot error – or is it?

I always thought we should put more effort into learning why a pilot made a fatal mistake. That is often difficult to do unless you know the pilot. So I am going to take a different approach to this essay on pilot error; I am going to base it mostly on pilots that I knew who crashed. There have been over fifty of those but I’ll only look at some of them. I’ll also look at one pilot who didn’t crash. That would be me.

John mentioned patience in the cockpit in a recent post. At some point, when addressing accidents, I opined that patience is the only virtue and impatience is the only sin. That is overly simplistic when it comes to air crashes but it is something to think about.

I also once wrote an article in FLYING about relating the seven deadly sins to accidents. If you can’t name the seven, some might suggest you get a life and others might issue an attaboy. Myself, I was just glad when the issue with that article was no longer current. In the print magazine business, we were convinced that when you placed the latest issue on top of the previous one all was either forgotten or forgiven.

When looking at the big fatal accident picture, two things stand out. Weather and loss-of-control accidents are the biggies with both a factor in some accidents. The reason those two loom so large is that such a high percentage of them are fatal. I think that your chances of survival are actually better in a midair collision.

A close friend and his wife were lost in a Cessna 340 that plunged from FL 210 to the ground in Colorado (elevation about 5,000 feet) in two minutes or less. It was night, the pilot had reported that he was in and out of clouds, the altitude wandered, the pilot reported a gyro failure and that was all she wrote.

The NTSB called this a loss of control for undetermined reasons with spatial disorientation on the part of the pilot. Weather and light conditions (dark) were also mentioned.

I think his friends and associates put a lot more effort into trying to determine why this well-trained, experienced pilot became spatially disoriented and lost control of the airplane.

I have written about it and thought about it a lot. There was a financial settlement that was supposed to be confidential but I believe the maker of the airplane, the autopilot, the vacuum pumps and the deice system participated though the NTSB found no evidence that anything failed. Trouble was, there was little evidence. The airplane virtually disappeared into the ground when it arrived at probably 300 knots in a near vertical attitude. The settlement made it possible for the survivors to feel that the pilot did not make an error.

This pilot used his airplane both in his medical practice and for pleasure. Like so many of us, he had discovered that a private airplane could virtually manufacture time. Spend the day at a watering hole and then fly well over 1,000 miles home that night and be on deck for work the next morning. That has proven to be fatal many times, as it did for him.

The question is whether or not a pilot is in error when he does this? Is impatience a factor? I suppose you could say that a patient pilot would arrange things so that he would not be tempted to play or work all day and fly all night.

There is more to it than that. I had made a night flight with this pilot, in a Mooney, not too long before his accident. We were flying above clouds and at one point he asked me if I was comfortable doing this in a single-engine airplane. I told him I was and he said something to the effect that he felt better in his twin.

Cessna 340
“I’m safe because I’m in a twin (or have a chute)” is simply wrong.

I have always felt that a pilot who will do something in a twin (or with a parachute) that he wouldn’t do in a regular single is making a basic error. If nothing else, he is planting the seed of complacency, which is not one of the seven deadly sins but is somethng that has cause a lot of fatal errors. I’m safe because I’m in a twin (or have a chute) is simply wrong. Private airplanes are only as safe as the pilot.

In another case, an acquaintance was flying a P210 (one serial number off mine) and lost control in similar circumstances to those just discussed. The main difference was that it was daytime.

In both cases the ceiling was quite high, 10,000 feet or more above the surface, but the airplane was operating in or in and out of clouds while flying in the low flight levels.

Enough of the P210 was recovered for the investigators to determine that both vacuum pumps had failed. The pilot apparently realized this, but might have thought his HSI was electrically operated, as was the one in the simulator where he had trained. Trouble was, the HSI in his airplane was vacuum so it had gone south with the other gyros.

Because both these pilots had sought out training above and beyond any requirements perhaps the accidents were related as much to training error as to pilot error. Complacency raises its head again, too. I bought the best training so I must be a safe pilot. I can recall at least one fatal accident that occurred on the first flight after a formal recurrent training session.

Those two were high-speed losses of control. Real quick I can think of three pilots that I knew who departed from controlled flight in V-tail Bonanzas while they were maneuvering to land. Two were experienced pilots but neither of these was the type pilot who thought a lot about the fine art of flying.

I could feel a relationship to these accidents because one of my duties as a flight instructor at Central Flying Service in Little Rock was to check pilots out in Bonanzas, all of which were V-tails at the time. We actually did a good rental business with the airplanes.

Claud Holbert, proprietor of Central, had checked me out in a Bonanza. A key item in the flight was a demonstration of a departure from controlled flight in the approach configuration. Gear down, flaps down, left turn, slowly increase the angle of attack and watch what happened.

Without much aerodynamic warning, the airplane started rolling left and pitching nose down as it stalled. You didn’t have to stick with it for long to realize that if it happened in the pattern, your name was going to be in the paper.

I did that on all check outs and while you might think it foolish to do with no control wheel on my side that was actually no factor. The throw over wheel was in front of the other pilot but was center-mounted so with my left hand I could briskly reduce the angle of attack.

It was both this experience and the fact that I grew up as a pilot doing spins that I think most stall/spin accidents are training system error and not pilot error. Looking back at the Bonanza checkouts, who can argue that a pilot should not see at least the beginning of a spin so he will know what happens when you lose control at low altitude? You don’t have to see a windshield full of twirling earth many times to understand that if you cause this to happen below, say, 1,000 feet you are autodead.

Confusion has always been a big factor in serious accidents, especially of the IFR variety. A confused pilot does not perform well and can at times become so bumfuzzled that he cannot perform at all.

Confusion comes simply because the pilot is trying to do something he can’t do or doesn’t understand. Again this is not strictly pilot error, it can be training system error. Over the years some have gone to great lengths to get the FAA to dilute flight training requirements, especially for the instrument rating, and this has been a fatal mistake for many pilots. Skimping on training for something as complex as instrument/IFR flying is purely foolish and dangerous.

Asiana crash 777
Pilot error times three? Not entirely.

When the Asiana 777 hit short of the runway at SFO and crashed on a clear day, there were three pilots on the flight deck. Did the fact that three of them couldn’t safely reach the runway without glideslope information (and a coupled approach) represent pilot error times three? Not entirely.

That crash was more representative of a training system failure. The crew was put in a position of trying to do something that that they had apparently not been trained to do, or at least not adequately trained. Sure, student pilots successfully fly visual approaches every day here but in Korea there is little or no private flying so pilots from there are exposed only to airline (or military) flying and that’s all they know how to do. If the ceiling had been 200 feet, they would have made a coupled approach to a runway with a functioning ILS and the accident would not have happened.

What does that have to do with pilot error accidents? Simple, a lot of accidents attributed to pilot error happen when a pilot has not been taught the proper reaction to a situation or problem.

I think a good lesson was learned with the Cirrus and the parachute. The fatal accident rate was terrible for a number of years with fatal crashes outnumbering chute pulls by a lot. Cirrus then went all-out on an educational campaign to encourage pilots to pop that chute if in doubt about what to do next. The result is that the fatal accident record improved dramatically as the number of chute pulls started outnumbering fatal accidents.

I have been writing about all this since 1958 and my experience with it goes back to when I soloed, in 1951. Pilots have changed a lot over that period of time and there were marked improvements in the safety record until ten or 15 years ago when it reached a plateau. (A recent improvement may or may not be real because the drop in the number of fatal accidents might well be the result of a corresponding drop in flight activity.)

When I started out, pilots were a different breed than they are today. Many, or even most, had either flown in World War Two or fought in that war and then learned to fly on the GI Bill. In either case, many had a fatalistic attitude toward flying as well as a lot of other things.

Three pilots I knew crashed V-tail Bonanzas for undetermined reasons with the accident charged off to pilot error. Two of them were suffering from terminal cancer and the third had no medical because he was diabetic. Maybe they were fatalistic about their flying (and life) and ran with the old thought about cussing the torpedoes and full speed ahead.

All three of those pilots were experienced but only one had much formal training. He had been a military pilot in World War Two. If the frequent mention of V-tails suggests that everyone I knew over the years flew a Bonanza, well, a lot of them did.

At the time, the only ground schools were at larger FBOs and the few colleges and university aviation programs. With the private pilot written (knowledge) test a 25-question true-false challenge, who needed ground school? The written tests for advanced certification were a lot more complex. I took a lot of them and the available prep materials were a lot more basic than the subject matter.

Flying and staying alive at that time was based more on common sense and experience than on education. We didn’t know a lot about weight and balance nor were we privy to performance charts. Look at the runway length, consider the load, temperature and wind and head out if it looked okay.

Piper takeoff chart
These were once routinely ignored (or unavailable). There’s no excuse for that now.

The fatalistic attitudes did lead to a lot of hazardous behavior. It was quite common to drink and drive and while most pilots didn’t extend that to flying, more than a few did. Scud running was very common and if you were good at it you had a chance of surviving.

In the good old days, then, almost all of the pilot error accidents were just that. In most, though, the basic error had more to do with the decision to start the flight than anything else. Now, the available training and information should ensure that no pilot is out there doing something he doesn’t know how to do or shouldn’t do. Pilots are still the basic humans they were 60 years ago, though, and the training and regulatory business still has notable soft spots.

Consider a couple of things the next time someone starts talking about most of the fatal accidents being caused by pilot error.

We teach people to fly and then send them off flying in airplanes that will spin. Yet in most cases we never show them a spin so they don’t really have a frame of reference when folks start preaching about low-speed loss of control with evangelical zeal. When one of these pilots loses control at low altitude and hits the ground in a spin, or the beginning of a spin, is it really pilot error?

No too long ago, and in response to Congressional pressure, the FAA increased the flying hour requirement for airline copilots six-fold. Prior to that, the FAA had reduced the experience requirements for an instrument rating. The airline crews and the private pilot fly in the same IFR system and often the same weather, and Mother Nature honestly can’t tell the difference between a single-pilot Cirrus and a two-crew 737. The complexity of an IFR flight is the same regardless of who is doing it yet if that inexperienced private pilot flying alone messes up, it’ll be called pilot error. Is it?

25 Comments

  • I have also often thought that pilot error has been used when design error should be listed as the real problem. If a pilot takes off and the engine quits due to fuel starvation and he stalls and spins in, it is “pilot error”. Never mind that the airplane has a 5 position fuel selector down on the floor or under the seat or some such other out of the way place. Newer designed have addressed this pretty well, although there really is no excuse for having a fuel selector at all.
    I also happen to believe that any machine which requires a checklist to be operated safely by the occasional user is inherently unsafe; no matter how much training you do.

  • When I was younger, I came across the following quote:

    “When the situation is obscure, attack” – General Heinz.

    Whether you agree or disagree with the General’s answer, it addresses the fundamental question: “what do I do when I don’t know what to do?”

    In medicine (I am an ER physician) the answer is the “ABCs”: keep the airway open, keep the patient breathing, keep the blood moving. Now you have bought yourself time to think.

    In flying (I am a CFI) I teach “Wings level, nose on the horizon, full power.” I tell my students that if the aircraft can fly, it will fly in this configuration. Now you have bought yourself time to think.

    Where would Air France Flight 447 or TransAsia Airways Flight 235 be today if this had been their immediate action plan?

    This action plan is succinct and easy to apply – just what students, low time pilots and low time-in-type pilots need in a crisis. I wish we would push this concept harder. It is THE thing to do when you don’t know what to do. It’s not perfect, but it represents the highest chance of success when you can’t figure out what is wrong or how to fix it.

    I think this would save lives.

    Typically, when instructors address this subject with their students they say: “First, fly the plane.” This is excellent advice on priorities, but it does not address HOW to fly the plane.

    Try asking your student (or a pilot friend) what their first action would be in an emergency. If they answer “fly the plane,” by all means congratulate them on having the correct priorities, but then ask them for an ACTION. If they ask for more information about the emergency, tell them that there is no further information.

    How many have a plan? The ones that don’t are at risk. We need to help them.

  • Richard, your thoughtful essay highlights the challenge for general aviation of complexity: complexities of aircraft characteristics and design, of aerodynamics, of weather, of the air traffic control system, and– above all– the complexities of human beings, their emotions, their motivations, their health, and their limitations. Some articles I’ve read over the years went deeply into all apparent factors in an airplane crash. These analyses made clear that while a given crash might have only one cause, usually there were more. Training interacts with aircraft characteristics, the latter interact with weather, the reaction to which is influenced by ATC actions, and all of this is affected by the pilot’s recent difficulties in marriage or business. Until all aircraft are autonomous (“robotic”) and simply will not fly beyond their capabilities, we will have tragic airplane crashes. Have we reached the lowest achievable crash rate for autonomous human pilots? Perhaps; I can only hope not.

  • Excellent article. Thoroughly enjoyed it. Will definitely think about it every time I climb into the cockpit. Thank you.

  • Richard,

    After +9800 hrs PIC in jets & props, and 22 years in my T-33 doing airshow acro demos, I still like seeing your wisdom.

    thanks,

    BIG DOG

  • Great article. My flight instructor required me to do spin recovery. Knowing that it was not required I did it because I wanted to have the experience. Afterwards, I never approached a landing or takeoff without realizing how quickly things can go bad.

  • Excellent article. I feel that the impact of night flying on the overall accident situation in GA is worth analyzing. I believe that there is room for real improvement in our situation by improving training and currency standards.

  • AS always Richard’s insights are golden. But I would humbly suggest Peter Row’s advice on “how to fly the airplane” as most valuable. Thank you both.

  • Great article sir. I’m going for my CFI next month and have often wondered how to approach safety with my students. I’ve been flying since 2001 and have seen my share of issues with pilot training (my own and others). In the most dangerous case, it was a pilot friend of mine who froze on the controls during an instrument approach in IMC at night. It got so bad that I had to take over just before he entered a power on stall WHILE STILL IN IMC. How do we 1. Keep flight training “fun” while 2. Letting the pilot see what can go wrong without scaring the crap out of them? When the crap hits the fan, the fun is over. It’s time to act with precision in a definite way.

    I was fortunate, I had a healthy respect for the airplane long before taking my first flight. Maybe it was my experience riding -and crashing- dirt bikes I saw what happened when you didn’t respect the machine and your environment while not killing myself during the process.

    I had to work through my fear of flying, I did that by facing it head on. I first spun with Rich Stowell in a super decathlon a year or so after getting my private and ended up loving it. I credit that one flight lesson with saving my life on at least two occasions.

    Thank you for your years of wisdom!

  • Thanks Richard however I think you, and almost everyone else, miss an important reality.

    At some point prior to being involved in a fatal crash the pilot comes to believe that they are going to die. This may be 15 minutes or a few seconds but whenever that happens at that time the pilot stops responding in a useful manner.

    Obviously it is not possible to test for this in a simulator or anywhere else but if you TRULY believe you are going to die you cannot recite a check-list or anything else.

    Therefore a pilot CANNOT respond correctly just when it is necessary for them to do so.

    • Pilot induced accidents.
      The under trained Certificate holder and most likely current and legal to fly, is usually not Proficient. Are they allowed to fly ? Yes. Who is overseeing this type of pilot ? Only himself.
      Not having recent experience and being unfamiliar with your AC systems is a trap, no matter the total time flown in the past.
      A simple cure to stay alert and ahead of the AC is continuing education and training.
      A well trained and proficient pilot has no time to panic.
      You are describing someone who is not a PIC ( pilot in command).

  • I have always felt that preplanning for emergencies is crucial. “What will I do if there is smoke in the cabin, and in what sequence?” “What will I do if I get an assymetrical flap deployment?” “What do I do if the engine quits after I move the fuel selector valve?”The list goes on and on, but doing this, visualizing it, and then memorizing and yes, practicing each response, so they become instantaneous, makes us far more prepared when something unplanned for happens. Last thought….why, when airlines do not even allow any non-flight specific conversation below 10,000 feet, does private avaiation now think it is ok to bop around while listening to music while flying? I just have the strong sense that rocking out and proper piloting are mutually exclusive.

  • Hello Mr. Collins !!!
    If we were blessed to have you as FAA Administrator, most of the discussion about LOC and Pilot Error would have been a thing of the past.
    FAA Administrators are trying to improve flight safety, but unfortunately they hit and miss as time goes by without encouraging results.
    Last year there was a proposal to change the traffic pattern so pilots avoid LOC by not executing a 90 degree turn, just make a constant shallow bank turn and you are not going to stall/spin close to the ground. We had a CFI forum at the local FSDO when the discussion came up, my reaction was ” Navy guy flying a desk at the FAA came up with this life saving idea”. We had a good laugh.
    I know there are many excellent CFIs who teach the right thing including emergencies and stall/ spin avoidance/ recovery. Their students have the potential of becoming safe pilots.
    Cheers !!!

  • In real IMC, zero zero take off, in a Cessna 340, very stable in pitch, very unstable in roll, canted tip tanks.
    The DG showed about a 5 degree left wing down, the Turn and Bank showed a slight right turn. Things we’re going bad, and it is extremely difficult to figure out which instrument is correct. Since my power and pitch were proper, and I was showing increase in altitude, and up on the rate of climb, I was feeling OK with pitch, but concerned about roll and getting a complete upset close to the ground, in real IMC. Bad feeling.
    It was made ever worse than losing an instrument, but because a slow failure is much harder to detect, and much harder to figure out.
    The AI was showing a climb, and wings level?
    Which to believe. Before GPS.
    It’s much easier in a training environment to figure it out, and you know the instructor will intervene if you’re getting upside down, but in IMC with no instructor, it is extremely difficult to figure out which is good and which is bad, with only very slight indications both ways.
    So I tried to keep them both showing about 2-3 degrees opposite, IE, I tried to split the different.
    A quick glance at the compass was of course wasted time, and as long as I was climbing, and the opposite errors seemed small, I just kept splitting the difference, but not an easy thing to do.
    I can certainly understand Air France 557 pilots panicking and doing the wrong thing, but remember they had been taught that the computers wouldn’t let things to completely wrong, and would not let the pilots override the conditions.
    The startle factor when it really happens is difficult.
    The tendency to have your focus narrow and peripheral things blur is almost overwhelming.
    The day I did the zero zero, which I have trained extensively for from my first instructor and first zero zero TO made me barely able to keep things together for.
    There was ice on the runway, and the day was very cold, and I couldn’t even do a very good run-up because locking the brakes and increasing power only made the tires slide.
    So I hurried through the taxi and run-up.
    Back in the air, with things going badly, I just kept splitting the difference, for a long three minutes of so when the gauges slowly began to come together, especially the AI.
    I fly regularly from this airport on this trip, this day starting south, but waiting to a clearance back to the North.
    For that reason apparently the tower folks thought I had done a 180 degree turn as standard, even though they hadn’t cleared it, so didn’t say anything to me for a while, and then finally blessed the turn by saying you’re cleared for heading 340 degrees, which it turns out exactly what I had done. From 170 degrees to 340 degrees inadvertently, and luckily for me.
    I finally topped out of the clouds at about 10,000′ and saw I was heading 340 degrees, which really makes you want to walk down that old long dirty road, kicking at turds, eyes down, and muttering to yourself.
    My thought was that probably the AI was sluggish because of the cold, and by the time I reached 10,000′ everything had warmed up including the AI.
    About three trips later, the AI failed completely. Fortunately this time I was visual.
    This error, if it had been a crash, would have been simply pilot error, loss of control in IMC.
    But actually, a cold and failing AI.
    So your point is extremely pertinent, and I feel more sorrow for the crew of 557.
    This unbelievable fatal accident was blamed primarily on the pilots, but was primarily a fault of Airbus in not putting pitots on an airliner that all knew would be flying in -50 degree conditions often. Nor should such a lack of ability for the pitots to be able to create enough heat for standard cold conditions, and even more, is absolutely the fault of Air France, Airbus, and the designers of the inadequate lack of heat to the pitots.
    Even worse, Airbus and Air France knew of this inadequate pitot system and was slowing retrofitting the planes with new and better pitots which could handle the expected and even worse cold temperatures.
    Additionally, the pilots believed that even holding completely back on the controllers would still not have resulted in the computers allowing the plane to stall.
    That was a “pilot error” that was just bad design, and bad execution of the design, inadequate heat on the pitots, and no sense of urgency on Airbus and Air France to ground the planes until equipped with proper pitots.

    • Why 0/0 T/O ?
      Why rush and make an incomplete run up on ice ?
      Compass reading, waste of time ?
      Splitting the difference did not do the trick, but compass would have done that just fine.
      ATC knew what your airplane was doing and failed to query you/correct your flight path.
      The latter was going to alert you about looking at the turn and slip indicator.
      All of the above constitutes pilot error due to poor ADM.
      Air France Flight 447, was clearly pilot(s) error big time.
      They failed basic airmanship and CRM, by not prioritizing. ” I have the flight controls- You have the flight controls- I have the flight controls”.
      Among other errors the power setting was disproportional to the desired attitude and both pilots made large inputs, one full back pressure the other forward without each others knowledge.
      Sadly many lives were lost.
      You are blessed to have survived !!!
      Fly Safe !!!!

  • ” A superior pilot uses his superior judgment to avoid situations which require the use of his superior skill.”

    Frank Borman

Leave a Reply

Your email address will not be published. Required fields are marked *