Forecasting Misery, or Miserable Forecasting?

Time keeps on slippin,' slippin,' slippin'
Into the future
--Steve Miller Band

Back to the drawing board indeed. After just a couple of days, the 'benchmark' UW-IHME forecasters have revised their COVID-19 death estimates (along with hospital loads) down again. The new projection predicts about 60,000 deaths inside of a range of 31,000 to 126,000. That's a 25% decline off of the past weekend's forecast, and orders of magnitude lower than its initial doomsday projections.

Not surprisingly, the forecasters do not provide ready access to their past predictions, or how indicate just how erroneous they have been.

Mainstream media, on the other hand, has begun publishing puff pieces about how difficult forecasting can be rather than on how miserably wrong these projections have been--not to mention the mammoth costs associated with implementing draconian countermeasures in response to predictions that have been way off.

It is safe to say that forecasters sporting this type of track record in the private sector would not be forecasters for long.

Meanwhile, with actual COVID-attributed US deaths below 13,000 and case counts and hospitalizations slowing, it seems likely that the IHME forecasters will be taking their projections down further yet in order to maintain some semblance of relevance.