Since I left you hanging last week, here’s a mid-week update.
Yesterday’s number was 104. It’s down a bit, but not in line with the model, that says it should be much lower. The model continues to lose power, so I’m going to have to figure something else out.
I built a new model based on deaths, using the same polynomial function as I did for infections. Quite the chart – it says we should be at zero. This epidemic is basically done from a death perspective. That’s great! Let’s hope this model holds – it has a higher r-squared and thus hopefully more predictive than the infections model.
One tactic Nate Silver likes to do is look at week-over-week changes and month-over-month. Here are those – a great-looking trend!:
|This Week||Last Week||This Week Last Month|