A few notes:
The spread of this disease is profoundly non-linear. Looking at cause and effect, and extrapolating the future from any given window of time is likely to produce very large (order of magnitude) errors. And has.
My hoped for “resistance” to infection and disease at about 35% is not supported by the facts. There is no factual reason to hope for this. The folks who predicted a fall wave based on people moving indoors have been more correct than anyone else.
Most people who extrapolated from the “case” explosion staring in May-Jun as testing ramped up (not shown on this graph) to produce curves like we are now seeing, and are claiming modelling success, are lying. “Cases” went up dramatically and stayed there for about 10-12 weeks before the disease as measured by deaths started climbing. In many places, cases and positive rates were falling as the last wave was increasing. This is still true. Nobody knows if the current wave will continue to grow for 12 more weeks, or will flatten and start to fall in the usual pattern, or some new pattern, starting today. Sadly, we see few signs of peaking in the data yet.
The policy problems we have created by ignoring the false positive rate of the PCR-RT at a 40+ cycle count remain. A hypothetical administration, wanting to show better results, would lower the PCR-RT cycle guidance to, say 35.
There was a 3 month period when the US seemed to be doing dramatically worse than almost all (or all) reasonable comparisons. I tried to explain this with the admittedly poor data I had in my models. What I did was remove the US (an outlier) , and then tried to do a Monte Carlo analysis of what I believed to be potential “causes” of disease spread and actual results to determine the sensitive variables. For the most part I failed to find much stastical correlation between policy and outcome. Variables that did seem to have an effect were limiting travel (globally and locally), limiting group sizes and population density. While I couldn’t measure it, there seems to be strong scientific reason to believe the 3’ (6’) practice is effective.
I failed to effectively model large quarantines with enough comparables effectively. I simply don’t know what they do.
One thing I did notice, and pointed out at the time, was that US failure to suppress the first wave was a statistical mistake – we were treating the summation over time of a many smaller graphs, each resembling places with smaller geography, as a single unit. In fact what we were seeing was the virus moving geographically, and hitting larger and larger populations, at the same time as we got better at treating it, and maybe better at preventing it. This explains, for me, the duration of the USA “bad time”, and the failure to suppress the first wave, but not the magnitude difference.
Of course, there were lots of explanations in the news: Southern and Mid-Western Americans are too stupid to follow rules, America all up is too stupid to follow rules, socialized medicine, the lack of a one-size fits all strategy, the lack of power at the federal government, unlike the more progressive China to weld shut the doors to people’s apartments, and the superior morality of people protects them from disease. Trump.
All nonsense, as we know can observe.
I’ll point out that global masking started as the same time as the geometrical explosion of wave 3. Policy and Karen wise, we still think whatever it is we’re doing with masks works. And there’s no real proof that it doesn’t. We can’t compare with wave 1 because the virus had not spread geographically yet.
There’s also no proof it doesn’t make things worse, which is what the data shows if you assume (statistical dangerous) correlation.
One constant though all of this: Fauci hasn’t a clue.