Using the forecast I’ve shared in other blogs www.dailykos.com/... and now on Medium medium.com/… I’ve looked 5th the progress in reducing the march of COVID, and it’s quite faint.
I originally looked at a Peak Fatality Day (that’s when we potentially would see the number of daily fatalities start sustainably declining) back earlier in March, and it was end-April, now it’s sometime 1st week of May. That’s a good thing, but when you hear “bending the curve”, that peak needs to shift to August, September, not early May.
That means around 4th of May, we would see a peak fatality of 650,000 people in the US. That’s my pessimistic view, and very hard to get my head wrapped around. Even if it were 10x less, it’s still mind-boggling when 31 March we saw a daily fatality rate of near 1,000. It’s hard to forget that it was around 1 March it was 1, and 15 March it was 11. It’s growing by a factor of 10 around every 12 days.
Today’s grim picture:
You can probably see that the orange line keeps lifting higher above the actual line, from the past. This is because I keep re-projecting.
Today, April 1st, we are around 97% likely to exceed all US Fatalities in the Iraq War.
If you are watching other forecasting, like the Institute for Health Metrics and Evaluation, since their original Press Release which lit up news desks around the world like crazy, they have revised their forecast up by around 18% from 80,000 to 93,000. I (and you) can get their results easily, by simply assuming that we will only have 2.735 Million cases in the US. I’m not sure why that’s their assumption. We may have ramped up testing kits now and distributed them sufficiently to collect the backlog of actual cases, and we should see around 1 Million cases around 5th April, and about 10 Million by around the 15th of April. I hope I’m wrong, but I still hear about restaurants in the US having seated dinners. All I know is we can’t have more than 327,200,000 cases, roughly the entire US population, but only .7% of the US population as cases seems, well, a little low.
I continue to be mystified by very low forecasts presented, but then I may just be having deja-vu: the HIV/AIDS Crisis is a model for it. Here’s an article from March 1990, from the Journal of the American Medical Association, stating that the HIV epidemic would have around “200 000 cases”, and burn out in the mid-90’s. Off by a factor of 10 or so, and of course the HIV/AIDS Crisis is long from over.
I found a new source of data ourworldindata.org/… here, which has beautiful charts, and an easy-to-download spreadsheet of actual fatality info (makes projections easier), but unfortunately I can’t seen to find state-level daily fatality numbers for the United States. If I could, I’d make a visual and numeric projection state-by-state, of how long each state is towards a peak / decline. I have the dread that most states are in their first weeks of a 6-week cycle and a sharp peak, while CA and WA will have a very flat curve and in a 6-month cycle. If anyone knows of such an aggregated daily fatality data source by state, feel free to identify it for me.
Stay Calm. Critical decisions need a clear head.
Stay Sane. Incessant COVID noise from deranged people will derange you.
Stay Home. There’s no safety in numbers, for the time being.
Disclaimer: I’m not a statistician, nor am I an epidemiologist or physician. I do have substantial training in science and mathematics. My assumptions are based on publicly available data, and interpreted without domain-specific expertise to temper my judgements.