Retro: using an old NYT page posted here awhile back...

[PeterB]

This one:

https://www.nytimes.com/interactive/2020...erica.html

... playing with the parameters a bit, I got that we will have another peak June 13th, with 3.8 million dead in the U.S. ::o ... could this be based on data that is now inaccurate? Perhaps. But still...

[pdq]

What parameters did you use? With only 14 days of social distancing (which is way less than what was done), they were guessing 1.3 million deaths by October (at a 1% death rate). With a 30 day intervention period (everything else the same), that drops to 700,000 deaths.

Here in MN, we went into stay-at-home with widespread closures Mar 28. We’ve loosened somewhat (something I don’t necessarily agree with) with a date of closer-to-normal of maybe June 1. At 63 days of intervention (again, with everything else the same), the NYT thingy predicts about 65,000 US deaths before October. Nationally, we’ve blown well past that already (and locally here, the daily numbers are, at best, plateauing.)

So if anything, I think the NYT was a little overly optimistic. Of course we don’t know what would have happened if we did nothing - maybe it wouldn’t have been that bad. But I think they made a pretty good guess, considering all the unknowns back then.

[PeterB]

Here are the parameters I used:

2 weeks of social distancing, starting 2/14, 1% death rate, 10% rate of hospitalization, moderate intervention level, no effect of warm weather (sorry, but I don't think this will matter), R naught of 2.2. None of these parameters are (to me) really that far off, except maybe the weather.

If I change the start date of intervention to later, like 4/15, I get the same date for a peak, with 1.9 million deaths.

Edit: I think it depends on how you define length of interventions. If I lengthen the interventions to 40 days, starting 3/15, I get much the same numbers, just a different date for the peak...