Monday, May 4, 2020

How Bad Will the Pandemic get in the USA?

Apparently you can publish papers based on statistical models of how bad you think a pandemic is going to be. That's just plain stupid. So I'm going to do a little of my own math and predictions for the Coronavirus Pandemic of 2020.

100,000 USA deaths by May 19th. 

I arrived at that on May 1st by estimating about 1900 Covid-19 deaths a day until we reach 100,000, and that's 19 more days from the 1st. I estimated 1900 deaths a day based on recent deaths of 2200-2400, and a general down trend in cases and deaths over the last two weeks, but a very minor one. I think deaths could be under 1000 per day by May 19th, but as long as we're seeing 25,000+ new cases per day, we're going to keep seeing a lot of deaths.

Second point, the University of Washington IHME predictions are terrible! Why? Because they've been screwing up Colorado Covid-19 deaths since March. We haven't had any zero death days since mid-March, but there are four days with zeros. This is simple data entry to feed the models. How are they getting this wrong? In short, their models are so broken they have basically no validity.
https://covid19.healthdata.org/united-states-of-america/colorado
Example 2 for IHME: We're going to have over 120,000 Covid-19 deaths in the United States, but the upper bound of the model is 115,000 or so, with a median prediction that we are going to pass this week.

Final question, how many deaths will we see in the United States and world wide before this is over? I'm not sure, I think it will be on the scale of World War 2, but I'll leave that for another post.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.