Dr. Tom Frieden is a former Director of the Centers for Disease Control. Yesterday, he put out this tweet:
Looking at those 1-million increases in ever shorter time frames seems scary. But is it an accelerating curve?
If we graph those points from 1 million to 9 million with the number of days on the X-axis, we end up with a straight line, with the curve fit having an R^2 value of almost 99%. If we squint our eyes a bit, we can see a slightly increasing trend, especially with the final data point. But this is not yet certain enough to proclaim the pandemic is accelerating based on this data.
Consequently, the claim of an accelerating pandemic, based on this data set alone, is questionable as of this specific day. Other data, or newer data, may confirm an acceleration. We just don’t see it in this data set.
If we chart those data points, beginning at day number 1, we see the argument for an exponential curve – but the reality is the curve is exponential at the beginning of the epidemic. Perhaps due to mitigation steps or other factors the curve is largely linear after the initial phase.
I saw this occurring in my state some time ago. The cumulative deaths curve is linear, not exponential. And the new cases curve has been mostly linear.
The steepening of the new cases curve, at right, is due to a large food processing plant outbreak, and over 200 people at a church that held large gatherings.
We do not know what the future data points will look like and currently released data suggests an increasing number of positive tests. However, public health enthusiasts say this may be due to a LOT more testing – and now testing individuals with little or no symptoms. Up through May, you pretty much had to be already hospitalized with pneumonia, or have all standard symptoms and be in a nursing center, or have all standard symptoms and have known contact with a confirmed case (those were the rules!) Consequently, no one with mild or no symptoms was tested until this month. And cases have risen since about May 29th. Large civic protests began on May 28th – and officially, they are said not to be a factor.
More charts after the break …
Total cases in the United States – linear curve with slight up tick at far right. Data from Worldometers.info.
A log scale on the Y-axis is used to emphasize the “rate of change” in the data. The flatter this curve becomes, the lower the rate of change.
After early April, the cumulative number of deaths curve for the United States was linear and is now moving towards a leveling off as total deaths drop.
These charts may appear different than your perceptions from reading the news or social media. These charts are showing you actual data, however.
As always, I am not an expert and you should assume I am an idiot. These charts and comments are for Entertainment Purposes Only.