There are three ways to crash. The first involves factors well outside our control, such as a rope break at 150 ft with nowhere to land straight ahead. We tend to disproportionately worry about these kinds of events and yet they account for a very small proportion of accidents or fatalities. The second involves deficiencies in basic airmanship. This includes situations like confusing the gear handle for the spoiler handle and fruitlessly cycling the handle up and down while screaming down the whole length of the runway. Lack of experience or recency tend to be the significant factors. However, complacency or an unfamiliarity with a new aircraft systems could still bite experienced pilots.
The third kind of accident involves critical errors in judgment or decision-making. This is where many of us get into serious trouble. Accidents related to “when to stop soaring and start landing” fall into this category. Also relevant is when to abandon a task or determine that the conditions are too challenging.
Matt Wright’s and Dale Kramer’s accidents come to mind. Both cases were similar in that other experienced pilots flew in the same conditions and chose to abandon their respective flights. Both pilots were driven, pursued difficult weather and then drove themselves in situations that were unrecoverable.
Why is it so hard to avoid this trap?
The underlying issue is that glider flying is unforgiving of overstepping safety margins. And yet the margins are not clearly demarcated. The glider flies until it doesn’t and departs into a spin. We can clear a tree line over the last field before the airport, or over the top of a ridge, until we can’t. The outcomes are either we are unscathed or end up plastering ourselves into the ground.
Sure, we sometimes scare ourselves. We may realize that we “cut it too close” on a flight. But tragically, even this experience will not always help us.
Worse yet, as we expand our margins, we can slowly erode our way into the territory where we are at serious risk. Ideally we scare ourselves before we go too far, but not everyone is so lucky.
And more importantly, as our margins become thinner, the more we enter into the gray area where we might lose control of the situation. There are limits to our control; even the best pilot can only keep his airspeed within certain bounds, suppose +- 1 knot. If his aircraft is exactly at the edge of stall, he would have a 50 percent chance of failing. And extending this logic to other situations, just how low can we reliably clear a treeline? Or how low can we thermal near a field? And if we get away with it once, twice, ten times, twenty times, does it mean we can do it 100 times or 1000 times?
The problem is that we only need to fail once for it to be game over.
The psychology of these kinds of situations is insightful. I found the paper, Decisions from Experience and the Effect of Rare Events in Risky Choice (Hertwig et al. 2004) to be especially relevant to this discussion.
To summarize their findings, they point out that decisions from “description” or “experience” are processed very differently. A doctor who has many years of providing vaccinations is likely to have never observed a patient have an adverse effect. A patient reads online that 1/9000 people experience complications. Both doctor and patient are processing the risks of adverse effects, but they are doing it in different ways.
Both kinds of judgment have been studied experimentally. An example of a decision made from description is playing a simple lottery:
Suppose you can take $2 for sure.
You can flip a coin. If you land on heads, you will win $5. Land on tails, you win $0.
Which do you choose?
In this case, the description of the lottery provides clear probabilities and the payoffs.
In contrast, decision-making from experience would have you learn the probabilities and payoffs over time. For instance, I could give you three urns that each contains a total of 100 balls; red or white. The urns have 30, 50, and 70 red balls respectively.
You are given time to sample from each urn. You can take out a ball, mentally note its color and then put it back into the urn. You are allowed to repeat this process as many times as you like.
Finally, you are told to select an urn and pull out a single ball. If it’s red, you will win $5. If it’s white, you will win nothing.
If you spent a sufficient time sampling from the urns earlier, you will readily pick the third urn.
Where decision-making from experience fails.
Now consider an urn that has 99 red balls and 1 white ball, but the white ball is radioactive. You might sample from that urn 20, 30, or even 50 times and never find that white ball. And maybe those red balls give you a pretty big payout each time. It would be awfully tempting to keep pulling the red balls from that urn.
That is until you stumble into the white ball and you die.
The fickle nature of safety risk in gliding is such that failure is rare, yet catastrophic. Much like sampling from an urn with a radioactive white ball.
And studies show that in such examples, people systematically under-weigh the risks of failure (Hertwig et al. 2004). When people rarely experience bad things, it’s especially easy for them to think it can’t happen to them.
But what about those times we got scared after a “close call”? Or when our friends crash? Usually this gives us pause and gets us thinking about our margins and choices.
The trouble is that when people base their decisions on experience, they over-represent recent events in their memory.
“Rare events have less impact than they deserve not only because decision makers have not encountered them, or have encountered them less frequently than expected, but also because they have not encountered them recently” (Hertwig et al. 2004).
Unless you were recently scared, or someone recently had something bad happen to them, your experience is not necessarily going to help you avoid making a bad decision.
This is especially evident in glider clubs. Someone crashes, maybe even gets hurt and there’s that cathartic moment when everyone becomes safety conscious. Some time passes and everything return to “business as usual”. Then two years later, there’s another accident, and the process repeats itself.
We can’t help ourselves.
What can we do about it?
One of my club members has a mantra, what can possibly go wrong? It’s usually easy to figure out well in advance that we might get ourselves into serious trouble. Suppose we’re watching our final glide bleed away and become marginal. Well if this continues, we may come up short, or we might just squeak it in. What are our options? What is our plan? How low are we comfortable pressing on?
There’s even more risk soaring in mountains or on ridges. On the Blairstown ridge, there have been countless incidents and accidents. Often time when I pass an accident site, I think about the poor fellow and what happened to him. This keeps reminding me of what could possibly go wrong.
When flying near places where I had scared myself, I visualize what happened. Passing through Snyders, I shudder when I remember how I fell off the ridge to the field below. Or how I crossed over Rt.81 uncomfortably low making the transition over to Bear Mountain. These incidents are seared into my memory and I routinely remind myself about them before entering comparable situations (maybe I should take that thermal an extra 500ft, thank you very much!).
This certainly makes the case that experience is helpful. But the point is that we have to actively summon past experiences for them to be useful.
Short safety talks in competitions or in clubs can also be helpful. Done well, they can remind people of what can go wrong and get them thinking about it in their routine flying.
Training for judgment can be difficult. We try to provide simple tools for success. For example, when we start cross country soaring we are usually given rote heuristics for when to land in a field (3k- choose area, 2k- narrow down, 1k- commit to land). There’s a lot to be said that having strict decision rules is good for keeping us within acceptable margins.
However, this does not work in situations that vary considerably. For instance, ridge soaring does not lend itself well to making decisions by rote. Attempting a transition at a certain altitude can work well with one wind direction, but not if the wind shifts 30 degrees. Experience lets you predict the conditions and make the necessary adjustments. This is what makes it fun, but with room for judgment there is even more room for error.
And if things start going to hell in a hand basket, where and when to leave for a field can be tricky. There are places where you have to be high enough to clear the ridge to make it to fields downwind. Other times, there are fields on the upwind side, but you might have to glide out a considerable distance. Where and when you can make these fields will change depending on the wind speed, direction, expected sink and more. There’s a lot of judgment that goes into maintaining these margins.
It is impossible to prepare for or train for all of these situations beforehand. The only way to deal with these circumstances is to build experience incrementally. Even better is to be guided by experienced coaches or mentors who can alert you as to what can go wrong.
Another way to explore the limits is to fly in the simulator. Condor lets you fail without the consequences. You can see what happens when you fly too slowly near a ridge, turn downwind into the mountain, wait too late to commit to landing and all sort of other common and fatal accidents. This is useful toward learning the boundaries and the limits of what you can get away with.
Decision-making research shows that glider pilots will nearly certainly under-weigh the likelihood of crashing. And unless we continually remind ourselves of what can possibly go wrong, we are susceptible to making really bad decisions. Experience is worth nothing unless we use it. We must learn from our mistakes and the mistakes of others. We must train incrementally and stringently maintain our margins.
And above all, remember that we are our own worst enemies.
2 Replies to “Experience Can Kill You”
We’re flying modern, heavy, fast, high performance gliders which essentially have 1930’s technology. In those days an outlanding was a non event due to the low stall speed and low glider mass. Would anyone fly motorless sailplanes if there was a light, *reliable* means of auxiliary propulsion with good performance that could be used?
You’ll know when that day arrives because a tow release will not be an available option on the aircraft.
Unfortunately the regulatory environment (self launching sailplane engines must be certified) precludes the experimentation necessary to improve the technology to that point. Anyone who has operated a self launcher will understand that the “certification” of the powerplant is a sick joke.