Games with Models, April 21 Data


The emerging question when it comes to the daily data releases for COVID-19 in Rhode Island continues to be:  which metric should we be watching?  Today, the number of cases continued to outstrip my prediction as did, tragically, the number deaths.  However, the number of hospitalizations, which is the focus of my model, went down one from yesterday and was, therefore, substantially lower than my prediction.

The reason the model I developed shows what looks like an upward correction and the predicted decline spreading out has to do with the number of cases.  The rate of growth is slowing, but not as quickly as it had been, which means there are more “active” cases than expected, and that’s the foundation of my hospitalization predictions.  (By the way, here’s the original post of this series, with my methodology.)



The three key metrics now look as follows:

  • Cases:
    • Projection for 4/21: 5,311
    • Actual for 4/21:  5,500
    • Projection for 4/22: 5,731
  • Hospitalizations:
    • Projection for 4/21: 296
    • Actual for 4/21: 271
    • Projection for 4/22: 302
  • Deaths:
    • Projection for 4/21: 162
    • Actual for 4/21: 171
    • Projection for 4/22: 178

The level of power Governor Raimondo is claiming for herself is huge, and it is noticeable in subtle ways.  For example, in an article on GoLocal about how she’s had her hair done even as Rhode Islanders are prevented from doing the same, the governor tells “hairdressers, barbers, gig economy workers” to file for unemployment “until I can find a way to get you open safely.”

Notice that it’s all up to her.  She is going to find a way.  She is going to tell you when you can work again.  If she is going to do that, she should give Rhode Islanders detailed information and data about how she is making decisions so we can judge for ourselves whether what she is concluding is reasonable or ridiculous.