The limits of probabilistic modelling
The term “black swan” was used in 16th century discussions of impossibility (all swans known to Europeans were white). Explorers arriving in Australia discovered a species of swan that is black. The term is now used to refer to events that occur though they had been thought to be impossible. In risk analysis, these are also called “unexampled events” or “outliers”.
They are more likely when your risk models include “fat tailed” probability distributions, where extreme events (in the tails of the distribution) have higher probabilities than a normal probability distribution. Fat-tailed distributions are infrequent in most measures of physical populations, but can be caused by mechanisms such as preferential attachment in social networks and “winner takes all” mechanisms. They can also be caused by correlation between different risk categories, which lead to systemic risks.
Methods for coping with black swans include stress tests, horizon scanning exercises, improving your “safety imagination”, logics of precaution, and increasing system resilience.
This submodule is a part of the risk management module.
Black swans, or the limits of probabilistic modelling
We recommend the following sources of further information on this topic:
Book The Black Swan: The Impact of the Highly Improbable by N. Taleb (ISBN: 978-0812973815)
Book Managing the Unexpected: resilient performance in an age of uncertainty, Karl E. Weick & Kathleen M. Sutcliffe, 2007 (ISBN: 978-0787996499)
Short course on resilience engineering by David Woods, a set of videos of a class taught at Ohio State University