Hi everyone. This will be my last post for a while. At this point, you would have to be a moron not to realize the virus is spreading out of control, so I don't think I need to continue to update. Here is my final take on what is happening:
/1

TLDR: Bad news. Cases are increasing, and mortality is increasing everywhere with an early case increase. I don’t see how things improve until there is a major policy change. But we now have better treatments, and drugs, so hopefully we won’t see the same level of mortality.
/2

We are seeing a rapid Coronavirus growth that cannot be explained by increased testing. The first state to see a dramatic rise in cases was Arizona at the beginning of June. See the green line in Figure 1.
/3

Based on data from March, I expected mortality to start spiking about 10 days later. But mortality remained flat while the state government declared they were only allowing the virus to spread among young people. The growth in cases was blamed on an increase in testing.
/4

As our President says, “the more you test, the more you find”. But that can't explain the data. Compare the increase in cases (green line) to the increase in % positive (pink line). Those can only consistently increase if the virus is spreading through the population.
/5

Meanwhile in Florida, cases also started to increase about a week after Arizona.
/6

On June 16th, Mike Pence wrote a highly misleading editorial in the WSJ claiming that the apparent case increase was solely due to testing. But from this graph, you can plainly see that both cases and percent positive were increasing by June 16th.
/7

Soon thereafter, it became evident that no amount of testing chicanery could explain the rapidly growing case count. The virus was spreading. Spreading very quickly.
/8

Florida Gov. Desantis was quick to point out that the average COVID patient age was dropping dramatically. The new line was “only young, low risk people are catching it, which is actually a good thing, because it means that they will become immune”.
/9

Was evidence sufficient that the elderly and high risk were protected? Plotting average age of positive coronavirus tests, we see that patients are getting younger. I'll use the term patients and cases going forward, but realize that many of the “patients” have no symptoms.
/10

Does this mean that only young people are getting the virus, not old people? How realistic is it to only infect the low-risk population? Plotting that data by age shows that cases are increasing in all age groups.
/11

If we separate the cases by age, we see that cases are increasing in all ages, but fastest in 20 year olds. However an increase in 20 year old patients doesn’t help dramatic increase in patients in their 60s. In fact, all age groups are testing positive more than in March.
/12

So we know that cases are increasing for all ages, why do we keep hearing about increasing cases and dropping mortality? Remember, Arizona was the first state to have a significant increase in cases. Did we see an increase in mortality in Arizona? Here is the same graph.
/13

You can see mortality reached record highs about 3.5 weeks after cases. Why is there now a 3-4 week lag, compared to a 10 day lab in March? In March, we didn’t have enough tests, people were tested at hospital admission. Now we are testing people, sometimes pre-symptoms.
/14

This data ruins the narrative that only young people are getting the disease, so officials needed another excuse. They started saying the increase in mortality isn’t due to new cases, but is instead due to a backlog of old cases from previous months finally being reported.
/15

This part is a little confusing, so let me try to explain. There are two important dates: the date that new fatalities are reported (reported by news). But those fatalities occurred earlier than reported.
/16

So while you might hear about 100 new deaths in Arizona today, those deaths occurred sometime in the past few weeks and were finally reported today. Here is a diagram to help explain.
/17

These are both useful graphs. The upper graph is very useful for noticing if cases have recently increased, but is prone to biases. For instance, see that almost no data is reported on Sunday and Monday, because the clerks are not working the days prior.
/18

The lower graph tells you when all known cases occurred. It is clear that by June 10th, cases had started increasing. But this plot suffers from an artifact where it always looks like cases are dropping off recently, because those will continue to be reported in future days.
/19

We can animate this plot to show how new cases are added over time, and you can see that new reports almost always occurred within the past 2 weeks, usually within the past several days.
/20

This week, Florida reported 308 fatalities in a three day period. This is a huge increase, and Desantis attributed it to reporting a “backlog” of cases from April and May. What does the data say?
/21

You can see from the graph that almost every fatality occurred in July. This is, IMO, the most grim, scary and important graph. It means that we should expect to see an acceleration in fatalities in Florida for at least a month after cases peak, and cases haven’t peaked yet.
/22

So now we know that mortality is actually increasing in the bellwether states Florida and Arizona. How do the rest of the states look?
/23

It does not look good. You can see that cases are rising in the majority of states, and mortality is rising in all states that saw the earliest increase in cases. These states recorded a record high average mortality on Saturday: TX FL CA SC AZ NV TN UT MT.
/24

We can pool all of the data from the states, and we get this graph:

There isn’t really any good news here. Cases started rising nationally 4 weeks ago, and we're seeing mortality now. This means that mortality will continue to increase for another month.
/25