Thread by @Humble_Analysis

It has now been 4 weeks since "cases" peaked in the northern Great Plains states; they have now fallen nearly 50% since mid-November despite mild restrictions and limited mask mandates:

Cases in the Rocky Mountain states appear to have peaked one week later (before Thanksgiving):

The southern Plains states seem to have hit a plateau before Thanksgiving and may have begun their inevitable descent:

The sunbelt, on the other hand, is still showing growth in cases, especially in the CA/NV/AZ region, with much more moderate growth (and perhaps a plateau) in the southeast:

The Northeast also looks as though we may be close to the top:

The Midwest region also looks as though it may have reach the peak for this season:

The Mid-Atlantic states are showing recent declines, outside of NC, but it is unclear if they are at the top, yet:

All-together, it should now be obvious that Covid cases follow a strictly regional-seasonal pattern and are indifferent to the variety of NPIs deployed as well as holiday travel (no Thanksgiving surge evident anywhere):

As a nation, cases seem to have peaked despite the assurances that Thanksgiving would lead to a spike in infections. Interestingly, the relationship between cases and deaths seems to have changed since this summer.

This summer, it appeared that there was a consistent 14 day lag between cases and deaths - but this has broken down in the fall:

@trvrb has suggested a 22 day lag, but that would imply that deaths increased BEFORE cases during the summer and this relationship has also broken down. The closest correlation currently between cases and deaths is a zero day lag.

A zero day lag between cases and deaths likely indicates that some portion of deaths are simply being labelled "Covid". This is confirmed by looking at all-cause mortality. In mid-November, the CDC reported 4K excess deaths in a week with almost 11K Covid deaths:

It appears that more than 50% of currently reported Covid deaths are simply mislabeled "normal" deaths.