COVID-19 April 27, 2020

The number of new cases dropped significantly yesterday.  This is usually the case for the weekend, which has historically been the low point of the weekly oscillation cycle.  The question for the coming week is whether the numbers will begin to take off again or whether the numbers will relax back to the model trend.  If the numbers increase again, it will be interesting to see whether they do so exponentially, which is shown by the orange curve I've added back since yesterday.  Perhaps we are starting to see the rise of the foolish states.

The number of daily deaths plummeted yesterday to one of the lowest values in weeks.  This is following the high point roughly a week ago.  If the weekly oscillation is still operating, the numbers should increase over the next few days.  The question for deaths is whether they will rise back towards the model trend.

What is interesting is that while daily new infections are climbing daily deaths are dropping.  The rise in new infections could be consistent with the increase in testing.  The number of deaths is independent of testing.  If deaths don't rise within the next week or so, that means that any continued rise in the number of infections is mostly the result of increased testing and not an actual surge in the pandemic. 

Overall, the quality of the data has become increasingly poor. The amplitude of the weekly oscillation has grown, the reporting criteria have changed from state to state, and the way that earlier cases just now discovered are added to the pile contaminates the data.  If the data collection and documentation protocol were rated by scientific standards it would be a fail minus.  It really just goes to show how poor (absent) the coordination and standards are at the federal level.