Coronavirus: 410,000 plus deaths by January 1, 2021?

Institute for Health Metrics and Evaluation (IHME) at the University of Washington recently released an estimate of COVID-19 deaths in the United States. Their estimate is 410,451 deaths by January 1, 2021. Seriously? I previously went after Nobel Prize winner Michael Levitt for what I thought was a ridiculously low estimate of 175,000 deaths. I personally think 410,451 deaths is laughably high considering that as of Sept 6th we had 193,250 deaths. 

The Hill (Sep 4) wrote:

The model created by the Institute for Health Metrics and Evaluation (IHME) at the University of Washington posted an update Friday predicting an additional 224,000 Americans will die by the beginning of next year.

The model says that as many as 122,000 of those deaths could be avoided with safety measures, including near-universal mask use, but it warns that easing restrictions could cause the death toll to be more than 620,000.

So their range is really between 288,451 and 620,000.

US News & World Report (Sep 4) writes: 

The institute also projects a dramatic increase in the daily death count, which could rise to as high as 3,000 per day by December. . . . The institute's previous death toll projection was 317,697 by Dec. 1.

Currently, we're averaging around 855 deaths a day. Since COVID-19 cases have been declining since July 24th, you'd think that the average number of deaths would continue to decline for at least a period of time. Let's say that we have an average of 855 deaths until mid-October. To get to 410,451 deaths, the daily deaths would have to increase to 1,137 for the rest of the year. That would be a 33% increase in daily deaths from the current level of 855. Since US News & World Report mentions that they're predicting deaths will go as high as 3,000 daily deaths by December, you have to assume they believe that daily deaths will trend down from the average 855 before spiking back up later in the year.

If we're averaging say 50,000 cases a day which is resulting in 855 deaths a day (these are just high level estimates), is the Institute for Health Metrics and Evaluation saying we're going to average 175,000 daily cases by say mid-November -- assuming a couple week lag between increase in cases and resulting deaths -- to get to 3,000 daily deaths in December?

I suspect they're making a couple assumptions here:

1. That this virus will start spreading like the flu when winter hits.

2. That we won't be able to protect people in nursing homes.

3. That very few people will continue will social distancing.

Sorry, I don't buy it. I'll definitely reference this blog post in the future to assess my criticism of their forecast.