From The Editor-In-Chief
In a memorable scene from the The King and I, a frustrated Yul Brynner confesses to Buddha, "There are times I almost think I am not sure of what I absolutely know." Although he was referring to the conundrum of advising his son, the statement perfectly captures the sentiment in Nassim Nicholas Taleb's The Black Swan: The Impact of the Highly Improbable. Taleb would have congratulated the king on taking the first step to avoid what Taleb calls epistemic arrogance, which causes people to overestimate what they know and underestimate uncertainty.
Ironically, we saw that principle at work while putting together this issue. We quickly discovered that, although we
thought we understood the concept of "rare event," in reality we had no working definition, and it was not trivial to arrive at one. We were struggling with this task, when three events occurred in rapid succession: the Icelandic volcano Eyjafjallajökull erupted, the Deepwater Horizon oil well began spilling into the Gulf of Mexico, and the Dow Jones Industrial Average dropped 1,000 points, earning the nickname Flash Crash.
These events—combined with a little more history and Taleb's insights—helped us arrive at our final definition of rare event: an event that occurs more frequently than conventional models would have us believe and that when it does occur, its consequences are catastrophic. We were also struck by the importance of knowledge in understanding rare events. Massive blackouts, financial meltdowns, cancer and genetic abnormalities, and terrorist attacks—all hold lessons that can help us better prepare. Paradoxically, knowledge is also the key to understanding what we still don't know. As guest editors Thomas Fowler and Martin Fischer point out, when Vesuvius erupted, no one even had a word for "volcano," but now scientists know enough to understand the why and where. They also know more about what they
don't know.
Lessons alone are not enough to make a difference, however. Attitudes about such events must also change. As this issue demonstrates, disciplines face different challenges in modeling rare events, but overall, rare events have elicited one of two reactions: We never thought it would happen, or our models showed that it could happen, but it was so unlikely that it made no sense to include it in any budget. The second reaction is particularly troubling. A British Petroleum failure mode analysis of the original design identified the failure that did occur. The budget for mitigation resources reflected that possibility, but executives believed the low likelihood of the occurrence allowed them to not spend the money on preparatory mitigation.
Decision makers in both the public and private sectors, as well as analysts, engineers, and scientists must understand that it is no longer acceptable to consider rare events as external to their design, analysis, and operating plans. In this issue, we talk about how the perception of such events changes as knowledge about them increases and offer a mechanism for understanding this evolution. An event starts as rare and completely unexpected—Taleb's black swan-and ends as tamed through prediction, mitigation, or even prevention to the point that it is no longer a black swan. Our evolutionary framework is not meant to classify an event as rare or not; it is simply a reminder that varied knowledge requires tailored analysis and design. It also does not promise that the event will not happen or that anyone can predict when it will. The three recent events might well have happened anyway, but response plans, response resources, and better computer rules could have been in place to more productively mitigate their effects. So can we model the unforeseen? The short answer is yes, and more important, we can build and manage more resilient complex systems by recognizing the importance of rare events and setting aside the resources to mitigate their effects.
H. Gilbert Miller, Ph.D.
Corporate Vice President and Chief Technology Officer
Noblis
hgmiller@noblis.org