Posted by Later today I’ll show you the death tracker graphs for the US and the world.
For now I’ll share with you that I have taken the data from the IHME study and put it into an Excel spreadsheet.
Their models have predicted a minimum average, and maximum range for required hospital space and equipment over time, as well as daily mortality. This is based on actual data up to a few days ago, with predictions based on effective dates for various social distance measures (both actual and predicted).
The variation between minimum and maximum is huge, because so much is unknown about spread rate and mortality rate. The predictions are very, very sensitive to these inputs.
However, the mortality numbers are going to be very large when this is all said and done. They are going to be in the five to six figures—even with today’s mitigation efforts being maintained.
It appears that the virus either (a) spreads very, very easily with a very small fatality rate, or (b) it spreads less easily with a larger fatality rate. The long strategies for dealing with either situation will be different, and hopefully we will learn which situation applies in next month.
But for the short term, we need to continue with the heavy mitigation we are doing now, for there is still far too much virus out there to deal with effectively. If we don’t, we will have a lot of loss of life—including that due to non-Covid patients not getting proper medical care in an overwhelmed system.
In my Excel spreadsheet, I record the actual cumulative death from day to day for each state and DC. This gets mapped against the min/max/average prediction curves in the IHME model.
Each state then gets a score depending upon where they are relative to the curves. Negative numbers mean they are under the average, with numbers less than -1 being under the minimum curve. Positive numbers means they are over the average, with numbers greater than 1 being over the maximum curve.
I then take a weighted average of these scores (with more recent scores being weighted higher) and then grow the numbers until they arrive at a final value. The final values are arrived at several weeks after mitigation has attenuated the number of new infections and the subsequent mortality trickles to nothing.
The number in parenthesis above is the running prediction. It is over 127,000 right now.
The trouble we have right now is we have 17 states where the death so far (3/29) exceeds the maximum levels in the study on the same day. Included in these are New York (27,298 predicted) and Florida (11,150). The predictions are inflated for these at this time.
The people who have died so far have generally contracted the virus in the early part of March, before most people in the country realized we had a big problem on our hands.
Another 9 states are above average, but less than the maximum predicted. Georgia (4,199) is in this group.
Another 16 are below average, but above the minimum predicted curve. Louisiana (2,723) is in this group, but very close to the average—as is Illinois (2,699) Michigan (4,645) is slightly below average. New Jersey (5,453) is moderately below average.
And 9 are below the minimum curve in the model. Ohio (1,619) and Texas (3,330) belong to this group
I think a good day of realization for a lot of people was Wednesday, March 11—the day the NBA cancelled their season, the NCAA restricted March Madness attendance to family only, and President Trump gave his Oval Office address. Things really started to change after that day.
In a few days, we should start seeing the deaths from people who contracted the virus around 3/11. If some degree of mitigation was happening, the daily mortality rate should drop by then. The model attempts to predict when this drop-off might occur. The final numbers will depend on the timing of that point.
Stay tuned! And stay safe!
Weight Betting Odds 