I’m trying to spend more time doing chores and outdoor work instead of blogging, so I might post less often.
But I do note here that the model has been updated substantially, with the following notable changes:
- Better hospitalization predictions (they were grossly overestimating needs before last weekend
- Death model depends on average of last 3 days of actual toll, rather than the last day—this prevents the predictions from varying wildly from one day to the next, particularly for states with small mortality to date
It appears that overall we are coming to the apex in the hotspot areas, and some of the places not yet hit hard may end up staying under control.
We have a ways to go before we start loosening up. But I think in the last part of April we will have a much clearer idea of what the exit strategy is. The testing capability will be better, and HCQ may prove to be a viable option for treatment.
I’ll finish here with some North Dakota predictions.


They have 20 cumulative deaths on 4/5 and 29 on 4/6. But the state page shows the current total at 4. WTF?

ND has no stay-at-home order.
Interesting…