April 1:

“Three Wave Model V”

So what if the third wave is really bigger? What if twice as many people are going to die in it? What would that look like?

Here ya go. Doesn’t look like the data, does it?

No photo description available.

Now that we have the 3WM, we can test it by using it to calculate some values, and then compare those to what we observe. Here we go: The Deaths Projection.

For reference, the Deaths Projection Graph shows you the three component waves of the 3WM (peaks along the bottom). It also shows you the NY data superimposed on its wave (W2).

When you combine all three waves, and add up their effects, you get the big red “S” shaped curve, finishing at about 8900 total deaths.

No photo description available.

March 31:

This graph will become increasingly important in the coming days. It assumes that we can think of the USA in two components: New York and the US49. It also is the perfect place to look to get a quick appraisal of the entire country.

Here is how to look at it. As long as the data points continue to fall along the traces, the NY & US49 assumption is holding, and everything is going as we expect. If the data diverge from the traces, then something new is happening.

For example, note how the red diamonds fall right along their trace. The red diamonds are deaths for the *whole country* minus New York. So, in one glance, you know everything is tracking in the rest of the country. That is reassuring, and it matters. Because if it ever diverges, we need to go looking for what is amiss.

Now consider the blue diamonds. Notice how they have jumped off their trace, though appear to be tracking well after. This indicates that something is amiss with New York. There are more deaths there than we expected. It is a sudden change, almost like a book-keeping error, or something. Time will tell.

Note how the sudden NY increase is also reflected in the total deaths curve (top curve). If those round dots do not return to the underlying trace, then it means our overall deaths estimate will be off by that amount.

See how handy this curve is? You can evaluate how the whole country is doing in a single place. And unlike the Big USA Cases model (which is subject to the testing whims of various regions) this graph is rock solid, because it tracks deaths, and there is no ambiguity about that.

Death is the ultimate statistic. One out of one people will die.

No photo description available.

March 30:

No photo description available.

March 29:

This is a very important graph, because it confirms our earlier hypothesis that we can model the US as the combination of New York + US49. (At least for now, as a very good approximation.) The new data fall right on the projected traces… made two days ago.

March 27:

No photo description available.

March 26:

Leaving the last post up for accountability.

Frank Giordano noticed a discrepancy in my data, so I went back and reconciled. My totals were correct, but the daily numbers were off. Usually they settle by the next day (when I typically reconcile), but apparently some folks made some older reconciles as well. It’s an issue with the world reporting numbers with GMT midnight as the cutoff, but the US reporting at various times. Somebody out there has a headache.

Here are the two reconciled graphs.. one smoothed, one not.

My intial gut is that New York needs to be a sharper peak. But I will come back to this later. Right now, people are sending me state updates in a flurry…

No photo description available.

No photo description available.

March 25:

Important mathematics to understand. Changing the “amplitude” of a peak (height) does not change the horizontal location of a peak.
So, the New York numbers are large, and are certainly affecting the vertical lineup in this graph, but not the horizontal. Our timeline is still intact.

I didn’t change the model at all. Only added today’s data.

No photo description available.

March 24:

Just added the data. I have repeatedly said that the deaths graph is the best way to track the progress of the epidemic.

Stickin’ to my guns on that.

Prior to this, we’ve acknowledged that the discrepancy between my initial projections and the data is the situation in New York. So yes, I am working on something (and you are gonna like it). But this graph remains valuable, because it SHOWS us the effect of New York. Why would I want to obscure the graph that illustrates this so perfectly?

I am not changing it today.

No photo description available.