Games with Models, 6/8/20 Data


Well, if we’re going to see a protest-driven bump in COVID-19 numbers in RI, we haven’t gotten there, yet.  The number of new cases has been under 100 for three days, hitting 51 yesterday.  If we consider a 14-day average illness, the number of active cases is now under 1,500 for the first time since early April.

In fact, all the numbers continue their declines, including deaths.  A headline slipped through my Twitter stream this morning proclaiming 27 newly reported deaths since Friday, but that way of tracking the deaths is becoming absurd.  Revisions will likely bump these numbers up a little, but as of today, Rhode Island hasn’t seen double-digit deaths for more than a week.

The best news, if you’re following my daily hospitalization charts, is that the numbers are now coming in well below my last few projections.  As a note on how to read the chart: The latest projection is the red line, with earlier projections getting lighter as you go back (for five days).  Basically, the formula had been projecting a short plateau that hasn’t happened.



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

Projections versus actuals (date of report).

  • Cases:
    • Projection for 6/8: None made
    • Actual for 6/8: 15,642
    • Projection for 6/9: 15,705
  • Hospitalizations:
    • Projection for 6/8: None made
    • Actual for 6/8: 146
    • Projection for 6/9: 146
  • Deaths:
    • Projections for 6/8: None made
    • Actual for 6/8: 799
    • Projection for 6/9: 802

  • Mario

    Everything is still on track for this whole thing to end in early July, probably with close to, but not exceeding 900 deaths. We may get our first 0 death day (recorded, not reported) within a week. I wanted to say a few days ago that we may never see another 3-digit case day (luckily I didn’t because the next day was 112), and that might be true now, too.

    We didn’t even get the small hospitalization increase I was expecting, although that one is harder to understand. I have another one showing up now, too, largely because admissions dropped so fast some time ago that a discharge dip shows up later. So I have it going back up to 180 on the 17th, but that probably won’t happen, and wouldn’t be a concern if it did.

    The protests were a big risk, and I expect some places will get punished for it, but I didn’t think that was too likely here for a few reasons, and so far, so good.

    So maybe 61 new cases, 6 deaths, 139 hospitalizations.