Post #1112: COVID-19 trends to 4/18/2021

Is this the peak of the U.S. fourth wave?

Below is the U.S. and regional graph, in logs.  There really hasn’t been a trend to speak of for more than two weeks now.  But now, most of those regional graphs are starting to bend downward.

This past few days, all the places that were hotspots seemed to be getting a break.  The thing that started to look like a peak, for Michigan, a few days back, looks like even more of a peak now.  Minnesota seems to have gotten to a plateau.  All the other places where the U.K. variant is known to be highly prevalent are doing, well, pretty much nothing.  Like so:

Even Colorado, which was the outlier among Mountain states, and the Mountain state with the highest incidence of the U.K. variant, got little break:

The situation isn’t uniformly rosy.  Maine, which had largely been bypassed by COVID-19 in the earlier waves, now has a strong upward trend in new cases per day.  Both Oregon and Washington still have upward trends (but fairly low absolute rates of new cases per day).  Otherwise, you really have to look hard to find any dark clouds on the horizon.

Separately, vaccinations continue apace.  As I noted some time ago, these daily vaccination numbers have some issues.  In particular, they have a strong regular weekly cycle, so if you catch Sunday between your snapshots, it’ll look like the pace of vaccination has slowed.  (Which this next pair of snapshots does.)

I keep hearing news coverage about vaccination rates slowing down, and vaccine “hesitancy” and blah blah blah.  Sure as heck isn’t showing up in the national data.  I wonder of that’s a case of the news media cherry-picking a few places that make a good story.  But overlooking the actual U.S. trend.

Over the past three days, the elderly added another 0.33 percentage points per day, and the U.S. averaged about 3.7M vaccine shots per day.

Today’s data glitch:  I guess it had to happen sometime.  Month after month, states would find old cases and add then dump those into their case counts, generating a “speed bump” in their data.  On 4/17/2021, Missouri found over 7,000 old duplicate cases, and removed them from their case counts, per this documentation, creating, I guess, a “speed dip” in their data?  The problem is that  you can’t plot negative numbers on a log scale.  So I’m tossing those 7500 or so cases back into the Missouri data, and getting on with it.

Post #1110: William and Mary, no COVID-19 uptick following spring break days

William and Mary had a COVID-19 outbreak following St. Patrick’s day.  There was some concern that we might see another such outbreak following two back-to-back spring break days last week (April 6th and 7th).  In my last post (Post #1099), I said I’d check in again to see whether there was an uptick following spring break.  Continue reading Post #1110: William and Mary, no COVID-19 uptick following spring break days

Post #1107: COVID-19 trend to 4-13-2021

Not much to say, really.  No material change from prior trends.  US trend is slightly up.  Michigan is almost up to 80 new cases / 100,000 / day, so they’re right where they were at the peak of the U.S. third wave.  The five states with the highest proportion of the U.K. variant are moving in five different directions.  And the elderly continue to get vaccinated, defying all prior survey-based estimates.

U.S. trend remains slightly up.

Michigan has almost reached 80 new cases/ 100,000 per day.  So they’re back into U.S. third wave territory.  But Minnesota isn’t anywhere near its U.S. third wave peak.  And I’ve made the argument that this is cause-and-effect.  Part of Michigan’s problem is that it has a low proportion of the population immune to COVID-19 due to generally low rates of infection in the earlier waves.

The dates are unreadable, but this one spans 10/1/2020 to the present, to show the U.S. third wave.

But it’s not as if all the states with high incidence of the U.K. variant are following Michigan’s path.  Michigan remains an outlier.


The elderly continue to get vaccinated.  Last three days, that rate was a bit under 0.4 percentage points per day.  Almost sure to break 80% in the next day or two.  Far more than was suggested by survey data.



Post #1106: COVID-19 trend to 4/11/2021


Opinions vary on what’s going to happen next in the U.S. fourth wave of COVID-19.  At one end of the spectrum, there’s the CDC director, talking up “impending doom” from the U.S. fourth wave.  And you’ll see sporadic newspaper stories where some expert will opine that a massive U.S. fourth wave of COVID-19 is approaching.  Or might be approaching.  Or at least remains enough of a likelihood that you should keep up your COVID-19 hygiene and get yourself vaccinated.

My post today is a reaction to a news article proclaiming that we’re all going to go the way of Michigan.  With, as far as I can tell, zero hard analysis to back that up.  Certainly, nothing like a statistical analysis of the state-level data (Post #1101), or a detailed outline of the factors that make Michigan particularly vulnerable (Post #1105).

Near as I can tell, that’s straight-up fear-based journalism.  The article more-or-less blames the people of Michigan for being a bunch of COVID-19 hygiene slackers.  When, in fact, that’s just not objectively true.   Or, at least, they’re no worse than 35 or so other states. Continue reading Post #1106: COVID-19 trend to 4/11/2021

Post #1105: COVID-19 trend to 4/9/2021, Michigan as a perfect storm

Source for this and all other graphs of new cases:  Calculated from The New York Times. (2021). Coronavirus (Covid-19) Data in the United States. Retrieved 4/10/2021.  The NY Times U.S. tracking page can be found at

I think the graph above appropriately characterizes the situation.  There’s an outbreak in Michigan.  And everything else is a jumble of modest up or down trends.  The resulting U.S. average result is still “no trend to speak of”, as shown below. Continue reading Post #1105: COVID-19 trend to 4/9/2021, Michigan as a perfect storm

Post #1104: Sidewalks, one last try at clarifying the issue.

I’m just going to make a few simple points here that may have been missed in my last post.  I made the mistake of thinking that everybody knew this.  That’s my error, and I’ll try to correct it now.

These point are that:

  1. The Town of Vienna strenuously objected to the narrow interpretation of the Robinson bequest.  They didn’t agree with it.
  2. A lawsuit in this case is the professional way to have differences in interpretation of a legal document resolved through the legal system by a disinterested, trained legal expert (the judge).
  3. My opinion of what the intent of the document was does not matter.
  4. The legal impediment to using the money is what matters.  Unless you change that, the Town of Vienna’s current plans are as good as it gets.

Continue reading Post #1104: Sidewalks, one last try at clarifying the issue.

Post #1102: Yet another sign that the pandemic is nearing an end.

Source:  Washington Post.

I’m sure nobody recalls this, but as far back as July 2020, it was clear that there was no valid reason to engage in “deep cleaning” of public surfaces as a way to limit spread of COVID-19 (Post #766).  All of that was just “hygiene theater”, as described in this article, in the Atlantic.

Continue reading Post #1102: Yet another sign that the pandemic is nearing an end.

Post #1101: New statistical analysis: When the facts change I change my mind. What do you do?

This is a quick redo of the analysis of the analysis of post #1092.  That was a simple regression using factors such as vaccination rates and prevalence of the new COVID-19 variants to try to explain recent trend in daily new COVID-19 cases.  The data points were U.S. states.  If you want the CAVEATS and the background, look back to that post.

The brief summary is that once you have enough states to look at, the picture snaps into focus.  It’s no longer muddy.  It’s exactly as it has been described by our public health establishment.  With (now) 40 states’ worth of data, it certainly does appear that we’re in a race between the new, more-infectious COVID-19 variants, and overall immunity of the population via prior infection and immunization. Continue reading Post #1101: New statistical analysis: When the facts change I change my mind. What do you do?

Post #1100: COVID update, more and interesting data on variants of concern and Michigan

National, regional, and state are about the same as they were yesterday.  Really, there’s nothing new to say there. There is no national trends to speak of.

Every day, new case counts go up in about half of states.  And they go down in the other half.

Regarding vaccination rates, I now know that the reported data have a regular weekly cycle, and that the “snapshot” data that the CDC post every day are substantially incomplete (see Post #1094).  In .  Also, looks like a lot of vaccination facilities were closed for Easter, which will put a little artifact into the data timeseries.

I think the only thing I can safely conclude, given all that, is that we still haven’t reached the upper limit of vaccination of the elderly.  The fraction with at least one shot continues to crawl upward.

I find that kind of surprising, given that most states are now beyond the phase where the elderly were given priority.  I can’t quite figure out whether this is just the lag between scheduling the shot and getting it, or whether there really are that many elderly individuals who waited beyond their priority period before they decided to get vaccinated.

There is significant interesting new information regarding COVID-19 variants.  The CDC has updated its state-level incidence data, greatly expanding the number of states that it reports incidence data for. Prior to this, just 17 states were listed.  This now shows 44 states.

Note that these rates reflect the state of the world as of about five weeks ago, on average.  And that some of these variants have been growing, as a proportion of the total, at a rapid clip.  If a snapshot of today’s rates were available, you’d see much higher numbers.

The point is that the highest incidence of the U.K. variant B.1.1.7, as of that time, was Michigan And Minnesota was right up there as well.  This finally resolves an apparent discrepancy between the CDC’s count of variant cases, and that table of incidence rates.  Michigan hadn’t been listed before, presumably due to small sample size.  Now that they have enough samples sequenced, Michigan is at the top of the list.

Despite that, it’s still hard to make the case that the variant is driving the outbreak.  Note Florida and Texas are also high on the list, and not much is happening there.  It’s almost as if the rule is U.K. variant plus cold climate yields outbreak.  And I’ll note that for the U.K. itself, with a combination of that variant, lockdown, vaccination, and warmer weather — the U.K. is seeing gently falling new case rates.

In any case, given the greatly expanded count of states for which this information is now available, I’m going to redo my simple statistical analysis (Post #1092) and see if anything changes.