(Dark humor alert: Sarcasm is not a healthy form of humor. Sometimes though, a healthy dose of ridicule accompanied by pointing and laughing is necessary to emphasize the ridiculousness of something. So, check out the following research…)
Breathtaking reports on social media claim there is scientific proof that the Sturgis motorcycle rally was a massive super spreader event.
Incontrovertible proof I tell you.
Study is described at Wall Street Journal on 11/21/20: CDC Study Links Sturgis Motorcycle Rally to Covid-19 Spread in Minnesota. Hosts of other news sites are doing their own re-write of the story.
Guesses are somewhere around 460,000 people attended the huge motorcycle rally held every year in Sturgis, South Dakota. This year it ran from August 7-16.
The lack of mask usage, the closeness of the event, and massive number of people was dangerous I tell you.
Coverage at the time and headlines now say this was obviously a super spreader event that would obviously cause 7 or 8 gazillion infections and just over half a bazillion deaths.
CDC research shows there, uh, weren’t, uh, quite that many infections.
Source of this out-of-breath coverage?
Analysis from CDC shows in Minnesota during August to September there were 52 people who got sick while at the motorcycle rally. Those people in turn infected another 35 people.
The study interviewed 80% of all the people who were diagnosed with Covid in Minnesota during August and September.
Of the people who tested positive from the rally and those who were infected by them, CDC sequenced the genome on 38 people, confirming they got the bug in Sturgis.
Primary infections – 52.
Secondary and tertiary infections – 35.
Yes, you read that correctly. A massively whopping 52 people were infected from the rally.
An astoundingly underwhelming 35 additional people were infected by those with primary infection.
In terms of the critically important spread rate, that implies the R0 factor was 0.67. In other words of every three people who were infected at the rally, two additional people were subsequently infected, either secondarily or tertiary.
An R0 factor above one means the virus is expanding rapidly, while below one means the spread will slow. The CDC study suggests the strain is not all that virulent.
Actually, since the number of subsequent infections includes both secondary and tertiary infections the R0 is substantially below 0.67.
Hold on, I’m not done laughing.
Got to wondering – how many infections there were in the state?
What portion of infections in all of Minnesota were caused by this cruel, horrible, irresponsible super spreader event?
To find out I poked around for about a minute and a half, finding this page at the New York Times which summarizes lots of data by state. In particular this specific linked page shows the data for Minnesota, including daily reported of new cases from March through current. Data for each day gives number of new infections and the seven-day average.
I listed the daily infections from August 20th through 31st, then added those up. For September I use the seven-day average on 9/7, 9/14, 9/21, and 9/28, extended that out, then added the infection on 9/29 and 9/30. Adding all that up gives the following info:
- 33,049 – infections in the state from August 20 through September 30.
- 26,439 – 80% of the total infections, which is the number of people the CDC interviewed.
Yeah, you read that right. The CDC interviewed approximately 26,000 people and was able to determine that 53 of them were infected at the rally. Yeah, 53 of 26,000.
They were able to identify a whopping 38 more people who in turn were infected by those 53 over the next month or so. Yup, 38 of 26,000.
That means that for every one person who picked up the bug at the rally combined with those who were subsequently infected by them, there were 304 other people who got sick during the last half of August and September.
Here’s the proportions again:
- Infected from rally? – 1
- Infected from some other source? – 304
That means 0.33% of the people who were sick during that timeframe in the state of Minnesota contracted the bug from the rally (87 / 26,439 = 0.3291%, rounded to 0.33%). A piddly, inconsequential one third of one percent. Not even 1 out of 100.
Keep in mind this is a time that public health officials are unable to identify the source of infection in anywhere from 50% to 75% of cases.
How do we compare infections from the Sturgis rally in relation to the unknown number of sources of infection? That means comparing 0.33% from one particular known source to 50%-75% unknown. The Sturgis rally is not even a rounding error in the infections in Minnesota. The infection rate would have had to been 10 times higher to merely be a rounding error in relation to the unknown sources.
The source-unknowns are between 150 and 230 times more than the known from Sturgis (50% / 0.33% = 152, rounded to 150; 75% / 0.33% – 228, rounded to 230). Let’s round that to 200 times.
For every 1 person in Minnesota infected from Sturgis, there are likely around 200 people in the state for whom the public health authorities have no clue where the person picked up the bug.
Here is another way to look at the tiny number of cases tied back to the rally – the average daily infections from August 20 through August 31 in the state were 842 people. The 87 people infected in total from the rally are approximately 10% of one day’s worth of new infections in September. Roll all the people infected from the rally into one day’s reporting and there were still nine times more people infected from other sources on that day.
The Sturgis rally, as documented by the CDC report, was inconsequential, trivial, minuscule in relation to the number of infections in Minnesota.
Okay, now I’m done laughing at the report and the news coverage.
Previous post discussing laughable research on the Sturgis rally: Stats on infection coronavirus rates not showing what you would expect. Time to open the economy.
The silly research estimated there were 266,796 new infections caused by the rally. Yeah, quarter of a million. Please stop laughing.
Same post discussed believable research which looked at the relationship between infection rates and the date each state imposed restrictions. The results showed in all but one of the states the infections had either peaked or were already declining by the time any restrictions would have had any impact on the infection rates. In other words, allowing for a few days for restrictions to have an impact, reduced number of people getting sick after incubation period, then dropping number of positive tests, the lockdowns had no impact on reducing infections.