As we saw in the last installment, The Confidence Man works by making every assertion -- no matter how ridiculous, heinous, or false -- with extreme confidence. By following this practice, with never a waver or inkling of self-doubt, roughly 40% of people will follow along in adoration. If you as a forecaster could pull off this confidence trick, you'd be on the Mount Rushmore of Forecasting!
But you can't ... and for good reason.
Forecasters operate in a loop of continuous feedback. Their forecasts are published and stored in the organization's system of record. And when the future rolls around (each week, each month, or whatever the forecasting cycle), their forecast of what is going to happen can be compared to what actually happens. By further comparing to what a no-cost naive model would have forecast, their performance can be properly evaluated. (Hopefully, that performance is not chronically worse than the naive forecast.)
Unlike other professions such as politician, tv celebrity, or sideshow huckster, it is hard for a forecaster to be a BS artist. Expressing unwarranted confidence in a forecast will get you into trouble -- as you'll be quickly exposed and lose credibility. So how do you express, in an appropriately confident manner, the uncertainty and risk inherent in a forecast? And is there even a role for forecasting in periods of extreme uncertainty?
What the Best Say About Forecasting Under Uncertainty
Forecasts are uncertain enough during "normal" times. But what about when major disruptive events, like a global pandemic, make a mockery of historical patterns and leave you little basis for projecting the future?
Times like we've seen so far in 2020 raise legitimate concerns about the role of forecasting. Is there any use in trying to forecast in this kind of environment? Or do we need entirely new approaches to survive the public health crisis and economic turmoil?
To find out, why not turn to two of the very best in the field: Nassim Taleb and John Ioannidis.
In a compelling "battle of the blog posts" organized Pierre Pinson and Spyros Makridakis for the International Institute of Forecasters, we get a preview of the positions from Taleb and Ioannidis. Quoting from the announcement of the debate:
Nassim N. Taleb [On single point forecasts for fat tailed variables] believes that all efforts and resources should be directed to halt its spread and reduce the number of infected and deaths without any concern about forecasting its future course as the uncertainty of doing so cannot be measured and the risks involved are highly asymmetric. John P. Ioannidis [Forecasting for COVID-19 has failed], on the other hand, claims that more reliable information is needed to make multiple billion-dollar decisions and that forecasting has failed us by being too pessimistic about the future growth of the pandemic and by exaggerating its negative effects. Reflecting upon it, they both see a failure in forecasting here, with John P. Ioannidis seeing those forecasts as necessary but unreliable to help policy makers, while Nassim N. Taleb sees the failure in the very idea one may use such (single-valued) forecasts as input to decision-making in view of the properties of the underlying processes.
Taleb and Ioannidis have followed up their initial blogs with full opinion papers to appear in a special section of the International Journal of Forecasting devoted to "Epidemics and forecasting with focus on COVID-19." Preprint versions of the papers are now available:
- Ioannidis, Cripps, & Tanner, "Forecasting for COVID-19 has failed"
- Taleb, Bar-Yam, and Cirillo, "On Single Point Forecasts for Fat-Tailed Variables"
Learn More About the Debate by Attending the Virtual International Symposium on Forecasting (October 26-28)
If you are not already a member of the International Institute of Forecasters, now is the time to join.
Membership ($145 for regular, $25 for student) provides full access to the Virtual ISF 2020, as well as one-year subscriptions to both IIF journals: the aforementioned International Journal of Forecasting and Foresight: The International Journal of Applied Forecasting.
What's more, membership gives you free registration for the Virtual 2020 Foresight Practitioner Conference on December 2.