Post #1238: COVID-19 trend to 9/7/2021, pardon my artifact.

Posted on September 8, 2021


The Labor Day holiday left a big reporting “artifact” in the data.  The numbers nose-dive, purely due to the extra day of no or minimal case reporting.

But if I crudely remove that, it sure looks like the Delta wave has peaked, for the U.S. as a whole.

Best to wait a few days, and let any true impact of the holidays work through the data.  For example, right now, it’s obvious that pretty much the entire Midwest under-reported their cases a bit, based on the shapes of the lines.  So this might spring up a bit over the next few days.  That said, I’m not seeing any remaining artifacts that look big enough eliminate what appears to be a 9/1/2021 U.S. peak for this wave.

So, for what it’s worth, here’s how the numbers look.  The first chart is how they show up, straight out of my programming.  The extra day of zeros in this week, compared to last, puts a big divot in the end of the line.  The line ends up in the right spot, at the end.  It just hits a pothole on the way.

Data source for this and other graphs of new case counts:  Calculated from The New York Times. (2021). Coronavirus (Covid-19) Data in the United States. Retrieved 9/8/2021, from”  The NY Times U.S. tracking page may be found at

Here’s the same thing, but this time I filled in the divot.  The line ends up at the same place, I just cheated and overwrote the “divot” with the final position of the line.  Instead of that big divot, there’s a little two-day flat spot now.


In any case, that sure looks nicely peak-y now.  If I overlay the data (red) and my sketched-in line (blue bit at the end), it’s right on target.

My guess (and hope) is that by next week, it will be obvious to all that the Delta wave peaked.  But, again, better give it a bit of time to make sure this isn’t some fluke deriving from the Labor Day holiday.

Sometimes the picture is clear in one of my all-state spaghetti charts.  This is just the Delta wave, so far.  The flat spots at the right are an artifact of my “divot filling”, described above.

I’d say that’s not very clear.  What we have is most of the very-high-rate states peaking, and most of the rest of them not.

In logs, it at least looks flat, if not an outright peak.  For sure, looks like growth rates have slowed pretty much across-the-board, even if case counts continue to rise in some states.

Absent some sort of nationwide school-based outbreak, I’m guessing that 9/1/2021 will have been the peak for the Delta wave in the U.S.  We ought to see a bit of decline, and then we’ll probably transition to the winter wave in the middle of October.  I’m just trying to plan ahead.