I see some really troubling trends in the data over the last week, and it's not what you expect.
We need to be avoiding indoor restaurants, bars, churches and gyms nationwide right now. Most states will not enforce limits top down until it's too late.
This is a log scale graph, which is useful for looking at how the rate of increase is changing.
The mostly straight blue line is the US total, showing our most recent doubling has taken 18 days. That's bad, and we're all aware of it.
The soft lines in the background are the states in the western US. By and large, they're now rising at a pace similar to the national average, and vary from fairly low overall rates on the west coast to seriously high rates UT, MT and WY.
No big surprises here, but remember that CA WA OR are rising at a rate very similar to the rest of the state. WA's governor recently announced measures to tame the rise before the absolute level gets too high.
A similar situation in the south. Most states are seeing doubling on a 2-3 week basis. While the overall rates are lower than most of the country, they're heading for the same disaster, just a few weeks behind.
But look at what's happening in the northeast.
Yes, their overall rates are below the national average, but look at that rate of growth. New York has seen its cases double in just the last 10 days, and most of the northeast is seeing similar explosive growth. 10-day doubling means an 8x increase per month. I don't think people in the northeast are really internalizing what those rates mean by year end.
These have all been log charts to show that rate of increase. I do that, because the usual linear scale charts make it look like NY has nothing to worry about...
Here's the same data on a linear scale, with North Dakota added. It doesn't look like it, but New York is on course to reach North Dakota's current level in just under 30 days.
Linear scale graphs give a false sense of security until the overall level reaches the "hockey stick" curve.
Remember that all the external factors are favoring an INCREASE in the rate of spread: colder, drier weather, holiday travel and gatherings, etc. Large scale vaccine availability still looks like a summer thing.
Finally, remember that our reporting on the rising death rate lags a full month behind infections. (22 days for mean death, plus more than a week on average of delays in death reporting.) What we did yesterday has already determined what our mid-December death count will look like.