(56j) Black Swans and Meteors - "Worst Case Scenarios" and Why We Need to Ignore Them

The term “worst-case scenario,” typically is taken to mean “the most unpleasant or serious thing that could happen in a situation.” There is nothing about the term or its common definitions that gives any guidance on how to determine the worst case. Indeed, a PHA or a QRA team is asked to speculate, with no basis, what the worst case could be. No matter how unpleasant or serious someone imagines an outcome to be, someone else can imagine something worse. At what point do we reach the “worst case”? The phrase itself has no actionable meaning.

Worst case scenarios are not “Black Swans.” As popularized by Taleb^[1], the term “Black Swan” describes inconceivable events, the probabilities of which are impossible to calculate, that have profound, typically catastrophic effects, and that are typically only rationalized by hindsight. It has become popular these days to refer to any rare external event that results in catastrophic consequences as a Black Swan. It is not their rarity, however, that distinguishes Black Swans, but their inconceivability. For some, the phrase leads to abandoning the principle of risk management in favor of consequence management and the insistence on planning for the worst conceivable case. This is an expensive approach and is not risk based.

The point of PHAs and QRAs is to identify hazards and assess their associated risks. Risk is the product of consequence and likelihood. Worst case scenarios do not address risk. Instead, they focus on consequence, ignoring likelihood. Adding the qualifier, “credible,” seems to address likelihood, but usually is of no real help. It simply shifts the lack of rigor from “worst” to “credible”. For it to mean anything, the term “credible” must be defined. Without an agreed upon definition, the term “credible” is as meaningless as “worst” and still leaves a PHA or QRA team burdened with making up that meaning as they go along.

For some, credible means conceivable. Meteors, although rare, are conceivable. They occur with predictable likelihood and have quantifiable consequences, but most would not fault a PHA or a QRA team for dismissing a meteor strike as not credible.

This paper develops a framework for giving meaning to the term “credible” and considers a number of “meteors”, both natural and man-made, that are occasionally posited during PHAs or QRAs: earthquakes, tornadoes, hurricanes, airplane crashes, pipeline ruptures, train derailments, and meteors. With probabilities quantified and the meaning of “credible” defined, “worst case” and “worst credible case” can be used in analysis with rigor.