This post is a day late although not all that short, as I warned it might be. This is because I have spent the last week visiting the best place on Earth, my true home, which is New York City. I will return again soon, and soon after that I will once again be able to live there. A great moment. I hope to write about the trip, but I need to get this post out quickly, so that won’t make it in this week.

This week had three (other) big discrete things happen. The CDC issued unexpectedly sane guidelines for vaccinated people, and thanks to the day of delay, we also have Biden announcing a date that everyone will be eligible for the vaccine. Finally, on March 7 the Covid Tracking Project stopped collecting data, requiring another phase transition in much of the data and leaving me without a source for detailed positive test rate data that I’m happy about.

We also had a bunch of opening up around the country, despite what is obviously about to happen.

Let’s run the numbers.

The Numbers

Predictions

Last week: 4.2% positive test rate and an average of 1,827 deaths after subtracting the California bump, using Covid Tracking Project’s final week of data.

Last week’s prediction: No prediction due to some combination of ‘somehow I forgot to do this’ and the expected lack of data collection making it difficult to fairly evaluate the prediction. We’ll start again fresh now.

(My (highly unreliable guess you should not trust to match what I would have said) would likely have been to see small declines, to something like 3.9% positive rate and 1,650 deaths, which absolutely does not count for anything as a prediction but does give a sense of vaguely where expectations were a week ago.)

This is the testing trends chart from Johns Hopkins, a plausible new data source.

I hate when I want charts and all I get are graphs, and also note that phase jump up and down, and also when you highlight a day it doesn’t tell you what date that is so unless you can work backwards from the final day (and this doesn’t make it that easy to know what the final day is, if it might or might not have updated yet) everything is super fuzzy.

The phase jump here appears in their data in some states and not in others. It’s clearly not ‘real.’ For now, I’m going to presume that the new 4.0% rate at the end is now calibrated however the old numbers were calibrated at the Covid Tracking ...