Games with Models, 5/22/20 Data


The reduction of hospitalizations of people testing positive for COVID-19 was back on track in today’s data release from the RI Dept. of Health.  New admissions, while up from the previous couple days, were still in line with the past week while discharges jumped up beyond the recent average.

Meanwhile, the increase in new cases was among the smallest in months despite a jump in the number of total tests.

The problem point continues to be deaths.  Revisions to prior days again makes the one-day increase look bigger than it was, but it still came in at 13, which is pretty much in line with the daily average for the past two/three weeks.  The stubbornness of this number in light of decreases elsewhere in the data, along with the fact that the majority of deaths are still occurring outside of the hospital, raise questions about what the numbers actually reflect.  Are these people dying because of COVID-19, or are they people dying who happen to test positive for COVID-19?


(See here for my original methodology and here for a subsequent modification I made.)

Projections versus actuals (date of report).

  • Cases:
    • Projection for 5/22: 13,712
    • Actual for 5/22: 13,736
    • Projection for 5/23: 13,872
  • Hospitalizations:
    • Projection for 5/22: 245
    • Actual for 5/22: 242
    • Projection for 5/23: 235
  • Deaths:
    • Projection for 5/22: 562
    • Actual for 5/22: 579
    • Projection for 5/23: 585

  • Ruth H

    I believe the numbers are absolutely not accurate. I do not trust how they are getting cause of death…. at all. (

    • Lou

      You don’t trust the cause of death because they a hired a lobbyist to make sure RI gets their appropriate share of COVID-19 relief? I’m not seeing the connection. Is there more to the conspiracy theory that we are not privy to? I suspect you would also complain if they did not advocate for our share.

  • ShannonEntropy

    Ruth is absolutely correct. The other day** on p. A1 of the ProJo was a sad story about a grandson who loved his Nana and lots of happy memories and then BAM! she got Alzheimer’s Dz and in a few years became “blank” in his words: non-verbal and vegetative almost

    In March the virus hit her nursing home but she tested negative. By April her condition had deteriorated and she was on a morphine drip, obviously terminal. But they re-tested the place and this time she was positive

    She died a day later, and guess what her death certificate listed as primary cause of death ?? Yep — Covid-19

    The moral of the story — conveniently overlooked by the ProJo “journalist” —- is that you can’t believe ANY of the “Official Stats” being reported by the medical establishment or the State

    That’s why I think you guys parsing these numbers and trying to guess what’s coming next is just a big waste of time. And donut forget… yer talking to guy whose username is based on a branch of the fields of applied mathematics and electrical engineering called Information Theory that’s a measure of the uncertainty associated with random variables. Case closed

    ** Like many others these days, I’m beginning to lose track of what day it is, but weirdly enough I’ve been so busy that I’m weeks behind in my reading. So that article is prolly from early this month