The framework: why a long record of success can hide real danger

A summary of the paper; Mike’s own words are quoted.

The question #

Sharing lessons learned, with the goal of enhancing the effectiveness and safety of experimental flight test: that’s why the Society of Experimental Test Pilots exists. So Mike’s paper opens with a question: might we be missing something in this process?

And a thought experiment. Someone offers you ten million dollars, when you badly need it, to play a single game of Russian roulette: one bullet, six chambers, one pull of the trigger. Would you play? The question in full.

Where the ideas came from #

The ideas came out of a serious hang-gliding accident in 1995, on a landing Mike had made successfully some 1,500 times over fifteen years. After a lengthy analysis in which he could find no mistake in how he flew the approach, he was left with one conclusion: the error was the decision to attempt the landing at all. And if that decision was bad, so was every decision to land in the same or harder conditions over the fifteen years before, even though none of those landings had ended badly.

The full story: Case study · The 1995 top landing.

Two premises #

1. The primary determinant of safety is the quality of decision making.
2. When a negative outcome has potentially catastrophic results, the go decision must be made only when a very high degree of confidence in a positive outcome exists.

Success reinforces the go decision #

Closely adapted from the paper.

When this protocol is followed, it is a natural result that a very high percentage of outcomes will be positive, and therefore also likely that we will see a long string of consecutive positive outcomes. A feedback loop forms: each go decision is validated by the positive outcome as having been a good decision, and is reinforced, and becomes more likely to be made again under similar circumstances.

But it goes beyond that. Often the “lesson learned” from a string of successes, even if unconsciously, is that our criteria were unnecessarily conservative: not only were we right to go, but we had margin enough to go under somewhat less favorable conditions. This can lead to a gradual lowering of our standards.

Three factors #

Consecutive successes can mask go decisions that are fundamentally unsound, if the success came partly or entirely from good fortune. Three factors combine to keep us from seeing it.

I · Deceptive mathematics

Below a certain probability of success, repeated go decisions lead eventually to an unacceptable probability of failure. At 99 percent, eighteen go decisions carry a 17 percent chance that at least one fails: one chance in six, the same as the revolver.

II · Murphy’s Corollary

Nobody believes Murphy’s Law. But if it were true, so would be its corollary: anything that did not go wrong could not have gone wrong. We are surprisingly susceptible to that belief. We take a positive outcome as evidence that a positive outcome was inevitable.

III · The pressure to go

Deadlines, budgets, expectations and ego push us to go when we have doubts, and to validate the go decision afterwards because the pressure has been relieved. A no-go only increases the pressure next time. In either case, reinforcement can operate in the wrong direction.

Why the problem keeps coming back #

Apollo 1, Challenger and Columbia each followed a long run of successes with a known hazard: pure oxygen at pressure, O-ring erosion, foam strikes on the heat shield. In each, the investigation found a problem that should have been recognized and corrected, and was not. The case studies.

None of this is new; the accident reports said it first. What may be worth considering is why these problems seem so intractable. Why did we have Challenger after Apollo 1, and Columbia after Challenger? The Columbia board called for a “new culture” at NASA.

I will suggest that as decision makers we can become victims not merely of a culture, but of a deeper and more persistent human psychology.

The challenge comes from a powerful combination: the inherently deceptive mathematics of probability, the potentially self-deceptive psychology of judging our own decisions, and the inevitable pressure to go even when we have doubts.

The remedy #

Mike doesn’t claim an answer, but offers two strategies he has found helpful.

1. Be aware of the problem: continually aware of our tendency to succumb to Murphy’s Corollary and to the pressure to go.
2. Analyze successes, not only failures: set a quantitative safety metric and a target beforehand, then measure every outcome against it, to find the successes that were really good fortune.

That is what the opening question was pointing at. We share lessons learned from failures, to avoid repeating the mistakes of the past. What we may be missing are the mistakes of the future, the ones we haven’t seen yet. The remedy in full.

@misc{meier2026framework,
  author = {Meier, Mike},
  title  = {The Framework: Why a Long Record of Success Can Hide Real Danger},
  year   = {2026},
  note   = {Edition 1.0},
  url    = {https://meiersafe.com/framework/}
}