On a logarithmic scale the reported cases plots are almost horizontal, meaning that the growth is approaching linear, not exponential as has been the case until recently. That suggests the crest is near. Death rates, while still increasing, are showng definite downward trends, also suggesting a nearing crest in coronavirus cases.
Here's an interesting (anecdotal) report, one of several I've seen recently suggestive that the unreported/asymptomatic cases are much higher that the 80% number I use in creating these charts. If, for example, the unreported/asymptomatic cases were 90% of the total rather than the 80% I use, the death rates would drop by a factor of two. Slopes wouldn't be affected, which is why I worry about more about consistency than bias.
The projected deaths chart shows how badly the tail is wagging the dog. Consistency matters, folks. The only things you can take from this data are two: one, cumulative deaths have been consistently less that projected since April 5, and two, when (if) the trend line stabilizes, it will probably end up somewhere between the projected (blue) and best case (green) dashed lines.
99.25% of Americans remain coronavirus-free tonight.
