I still can’t believe Kevin Hasset and that civic model. I know he’s out there trying to say that he didn’t intend it to be a forecast. But 1) it appears some in the White House took it that way and 2) if you want to ‘smooth data’ why not just use a moving average and why plot any points of the cubic fit (not model) beyond where we have measured current data? I’ve always felt Kevin just worked too hard to please. I remember his press conference from mid-18 where he rolled out all these different plots of why the economy was amazing and it was only because of Trump and on almost every one there were tricks played with the way the data was shown chosen or shown in order to drive home a point. Basically data warfare. At the time I was like “I get that you aren’t making this up, but really, Kevin?” And the second I heard about “Kevin’s cubic model has deaths going to zero in 11 days” the exact same thought occurred to me.
In other news, despite a lot of people giving IHME a lot of grief (and they do deserve some regarding their initial hospitalization projections), their death modeling in that First March release has held up quite well. We’ll see if their recent revisions factoring in opening back up hold up as well or whether factors like warmer weather turn the tide.
