Fragility, Not Probability
Taleb argues that a low probability of harm can still justify protection when the consequences would be severe. The decision depends on exposure to loss and the cost of precaution as well as the chance of the event.
THE IDEA
Fragility, Not Probability
We check people for weapons before they board the plane. Do we believe that they are terrorists: True or False? False, as they are not likely to be terrorists (a tiny probability). But we check them nevertheless because we are fragile to terrorism. There is an asymmetry. We are interested in the payoff, and the consequence, or payoff, of the True (that they turn out to be terrorists) is too large and the costs of checking are too low. Do you think the nuclear reactor is likely to explode in the next year? False. Yet you want to behave as if it were True and spend millions on additional safety, because we are fragile to nuclear events. A third example: Do you think that this random medicine will harm you? False. Do you ingest these pills? No, no, no.
If you sat with a pencil and jotted down all the decisions you’ve taken in the past week, or, if you could, over your lifetime, you would realize that almost all of them have had asymmetric payoff, with one side carrying a larger consequence than the other. You decide principally based on fragility, not probability. Or to rephrase, You decide principally based on fragility, not so much on True/False.
Let us discuss the idea of the insufficiency of True/False in decision making in the real world, particularly when probabilities are involved. True or False are interpretations corresponding to high or low probabilities. Scientists use the term “confidence level” for a statistical procedure. A 95 percent confidence interval comes from a procedure that, under its assumptions, fails to cover the true parameter in 5 percent of repeated samples. It does not assign a probability of error to a particular conclusion. A confidence level does not measure the size of an effect or the cost of an error, which matters especially when losses could be extreme. If I tell you that some result is true with 95 percent confidence level, you would be quite satisfied. But a claim that a plane is safe with 95 percent confidence is not the same as stating its crash probability. Even a 99 percent survival probability would leave a 1 percent chance of a crash, which would be alarming (Taleb reports a commercial-aviation crash probability of less than one in several hundred thousand at the time of writing and attributes improvements to learning from errors). A probability alone cannot determine the decision; the consequences and costs of acting also matter.
You have taken probably a billion decisions in your life. How many times have you computed probabilities? Of course, you may do so in casinos, but not elsewhere.
When to use it
Risk assessment and safety regulations (e.g., nuclear power, aviation, pharmaceuticals)
Personal health choices (e.g., avoiding unproven treatments, taking preventive measures)
Business strategy (e.g., investing in redundancy, hedging against tail risks)
Public policy (e.g., preparing for low-probability but catastrophic events like pandemics or natural disasters)
FROM THEORY TO EVERYDAY LIFE
In practice
Airport Security Screening
Airport screening illustrates Taleb’s argument: even when an attack is unlikely, its possible consequences provide a reason to consider precautions alongside their costs.
Nuclear Reactor Safety
Vast sums are spent on additional safety measures for nuclear reactors even though the chance of an explosion in the next year is very low. The asymmetry of consequences (disaster vs. cost) justifies the precaution.
Unknown Medicine
Taleb uses refusing an unidentified pill to illustrate avoiding an uncertain downside. Without knowing what it contains, a person lacks a sound basis for treating the risk as small.
A MOMENT TO REFLECT
Put it to work.
Recall a recent important decision you made. Did you focus more on the probability of outcomes or on the potential consequences? How would considering fragility change your approach?
Think of a situation where you avoided a low-probability, high-impact risk. Was the cost of avoidance justified? What if the consequence were less severe?
In your work or personal life, identify a decision that was made based on a confidence level (e.g., 95%). How would the 'fragility not probability' lens alter that decision?
Often confused with
CONTINUE THE CONVERSATION
Source texts
- Antifragile (Nassim Nicholas Taleb)