THERE'S SOMETHING PUZZLING ABOUT THIS PICTURE.

This Medium post by Jonathan Geach, M.D., offers optimism to the people who would like to go back to work.  "Moving the Goalposts — Four Reasons it is Safe to Open America."

I'm going to pass on weighing his arguments, as those are a little outside my wheelhouse.  There's an instructive figure borrowed from the Institute for Health Metrics and Evaluation contagion model that bothers me.

Here's something I wrote last week.

The best you can do, dear reader, is improve the precision of the mean fit, which involves reducing the margin of error. Too much of the reporting and the briefing focuses on those solid lines: actual and projected cases. Those clouds that suggest the upper and lower bounds on the projection matter. When you see the clouds becoming less tall, then perhaps the policy makers will start talking more seriously about reopening life as we used to know it.

That's what bothers me about the modelling. Note how the margin of error goes to zero as the forecasts go to zero, and there appears to be no margin of error estimate once past the expected end of new cases, somewhere around Independence Day.  Now, I'm way outside my area of expertise, and I remember just enough econometrics to be dangerous, but I seem to recall that one of the hazards of macroeconometric forecasting was that the forecast error got larger, the farther out of sample (which is to say, the greater the time between the mean of the actual data you were using to calibrate and the out-year forecasts you were making) your projection ran.

In the above figure, that mean occurs somewhere around the end of March, and the projections are being updated on the basis of the (more favorable than projected) returns coming in.  But the out-of-sample forecast error is also getting smaller, with the arrival date of the last case being the point estimate and there being no uncertainty about that last case coming after that day.  Call me puzzled.

SECOND SECTION: Rick Moran at Pajamas Media, also puzzled. "The researchers argue that policymakers may believe that policies well short of the draconian measures taken by China in Wuhan prevented the horrific loss of life when, in fact, the model giving those numbers in the first place was fatally flawed."