More virus analysis
To me, it's 100% obvious how the virus behaves as it does: once you reach a critical percentage of the population that is infected (and keep in mind that the numbers posted provide a lower bound to the actual infection rates), you move from the slow burn stage to the exponential growth stage.
I'll just number them:
- Pre-infection - The virus is not in the community. In the outlying counties there just wasn't anyone there that was infected.
- Slow burn - things are basically kept in check with preventative measures and various types of lockdowns.
- Exponential - All you need is one super-spreader event to release the virus into the community. At this point preventative measures won't stop the spread of the virus, but can slow the spread of the virus. More on this later
- Peak and recovery - The virus has chewed through the available hosts and the infection rate starts to drop down again. This is the controversial one that I get constant downvotes on.
#1 is easy. No virus.
#2 also makes sense I think. As long as you are under a certain critical number of people and you can actually track them you can stay one step ahead of the virus. This is also a painful step from an economic standpoint because you have to have pretty slow interactions between people to slow the spread of the virus.
#3 happens when you have a super-spreader event and you get the exponential growth. Once you have more than a certain number of people out in the community, it will spread. At this stage I don't believe that any amount of lockdown or preventative measures short of enforced stay-at-home orders will stop the spread of this. While I'm saying that the safety measures will not prevent the spread of the virus, they can slow the spread of the virus.
Let's take a deeper dive into this point. The virus was basically in my stage 3 here in Nashville when we started tracking it. Same with Memphis. The difference is how each of these cities handled it. At its peak around 0.9% of the residents were being tracked as active cases (again, this simply provides a lower bound). Memphis, on the other hand, peaked at around 0.6% of its residents in the same state.
- Memphis: https://i.imgur.com/TZd2RaI.png
- Nashville: https://i.imgur.com/kOdOYfo.png
Let's ponder why they are different for a second. Nashville and Memphis both accelerated when the safer-at-home order was lifted. Nashville started going downtown to bars. Nashville accelerated faster than Memphis because we were uniquely more asshole-ish. We made the national news for Broadway being open.
#3 is also the point where "flatten the curve" become critical. Memphis did a far better job of spreading out the virus than we here in Nashville did. The numbers show exactly that.
#4 is where I get all the haters. First off, let's take a look at Macon County: https://i.imgur.com/JL8TX22.png. If you take a look at the infection rate, as represented by the red line, it traces what appears to be a sigmoid function. These functions are used in modeling many natural behaviors. The communicability of viruses are one of these phenomena that can be modeled by this.
Now if you look back at the Memphis and Nashville graphs above, you can see they are on the very same curve, just playing out in different time-scales. Macon really messed up and got everyone sick all at once. They did not flatten the curve -- their curve was a freaking spike.
What I'm saying here is that some areas of the state (and I would imagine the nation, but I don't have data to back me up) are well past the worst of this. Many of the smaller counties are still well on the upswing.
Regardless of the cause of this (and I've linked to scientific studies in the past to corroborate this), there seems to be an inflection point that gets reached around 1.8%-2.4% where the number of active cases starts to level off and start to dip. This is consistent over many counties in this state. Seriously, take a look.
(why the variation? I have several theories: political affiliation tends to make some people more or less likely to get tested [i.e. more conservative denier areas may get tested less often, ergo the number would be lower], how quickly tests were broadly available in the area, etc.)
If you scroll through the graphs and exclude the counties that have had prison infections (which throw off the numbers substantially because the testing protocols do not behave the same as self-selected testing out in the wild), there's only a single county, Putnam County, that seems to be going up out of character from the remaining counties.
Let me take a step back to re-emphasize that I'm not saying we are done with this. We are not done with this. We all need to keep being careful. I specifically am calling out to the counties where they are well below the threshold I'm seeing, places like Greene (https://i.imgur.com/kpt6rZA.png) to get in line and start to flatten your curve.
Opening schools in areas that are not past the peak is folly. All you're doing is quickening the pace the virus is spreading. You are not flattening the curve. The virus doesn't care about political affiliation.
What I am saying is that I feel that we are seeing the light at the end of the tunnel. Stay careful, but know there is hope.
Full graphs here: https://imgur.com/a/AWV2ntz