- C-19 incubation period is 2-14 days with highest contagion somewhere in the 5-8 day range after contact 1
- To date, it seems ~.86 > RØ < 1.29 (FL data) 2
- Memorial Day was May 25th meaning inflection point starts ~May 30 / 31.
- The rest of the numbers here are from worldometers
Using New York's outbreak as a guide - and I'm sure there is a better term elsewhere - I am dubbing something called the "Death Lag". I'll define this as the number of days (seemingly) between infection and death impacts in the 7-day moving averages. Let's take a sample set:
| Date | New case 7-day moving average | New deaths 7-day moving average |
|---|---|---|
| March 27th | 5,461 | 132 |
| March 30th | 6,731 (+23.26%) | 265 (+100.76%) |
| April 2nd | 7,872 (+16.95%) | 443 (+67.17%) |
| April 5th | 9,097 (+15.56%) | 633 (+42.89%) |
| April 8th | 9,830 (+8.05%) | 799 (+26.22%) |
| April 10th | 10,059 (Peak) (+2.33%) | 920 (+15.14%) |
| April 13th | 9,306 (-7.49%) | 956 (Peak) (+3.91%) |
Just eyeballing the metrics here but it looks like the trail seems to be somewhere between 5 and 8 days from peak (April 5th +15.56% new cases and April 10th +15.14% deaths are when the curves seem to align most). So let's conservatively put the lag at ~7 days against peak. Going back to our original numbers, we can consider "peak" infection is day 7 of the infection so "death period" is anywhere from 14-21 days after infection.
April was Florida's worst month by far until the latest outbreaks.
| Date | New cases reported | 7-day Moving average |
|---|---|---|
| April 2nd | 1,235 | 932 |
| April 3rd | 1,260 | 1,010 |
| April 4th | 1,277 | 1,072 |
| April 6th | 1,279 | 1,132 |
| April 17th | 1,413 | 969 |
Florida at-home orders began Friday, April 3rd and ended Monday, May 4th.
On April 3rd the 7-day moving average for reported new cases in Florida was 1,010. Two weeks later it was down to 969, three weeks later it was 826 and four weeks later it was down to 599 7-day moving average (sheltering works!).
Florida began re-opening in phases starting May 4th but wide-spread disobedience started about a week earlier. During the month of May you see spikes (>800 reports) on (with estimated infection dates):
| Date | New cases reported | Estimated infection point |
|---|---|---|
| Friday, May 1st | 1,038 | Saturday |
| Monday, May 4th | 819 | Tuesday |
| Thursday, May 7th | 826 | Friday |
| Saturday, May 9th | 802 | Sunday |
| Tuesday, May 12th | 941 | Wednesday |
| Thursday, May 14th | 808 | Friday |
| Friday, May 15th | 928 | Saturday |
| Monday, May 18th | 854 | Tuesday |
| Thursday, May 21st | 1,204 | Friday |
| Monday, May 25th | 879 | Tuesday |
| Friday, May 29th | 1,212 | Saturday |
Quick checks on the estimates:
- Saturday x3
- Tuesday x3
- Friday x3
- Sunday x1
- Wednesday x1
I don't really pretend to know wtf Floridians are doing Tuesday night (maybe this is a Monday / return-to-office piece?), but it seems clear that the weekends were the problem children (duh).
Based on just this information - you can make a reasonable guess that says (without social distancing) that initial bumps would occur May 30/31. However, one thing to note is that Memorial Day fell on a Monday so, based on the above, we can guess that there was a decent chunk of Floridians hitting the "please get infected" buttons starting on the 22nd with expected bumps on the 27th through the 30/31st.
On May 29th, Florida hit 1,212 newly reported cases which was only the second time in a month they went over 1,000 new daily reported cases. 7-Day moving average on the 29th was 721.
Based on our estimated numbers above and combined with the RØ of... let's say 1.1, we should be relatively able to predict what happens without social distancing (7-day moving).
| Date | New Cases | New Deaths | Estimated infection increase (+7) | Actual new cases (∆) | Estimated death increase (+14) | Actual new deaths (∆) |
|---|---|---|---|---|---|---|
| June 1 | 726 | 30 | June 8 @ 798.6 | 1,154 (+355.4) | June 15 @ 33 | 32 (-1) |
| June 8 | 1,154 (+58.95%) | 37 | June 15 @ 1269.4 | 1774 (+504.6) | June 22 @ 40.7 | 33 (-7.7) |
| June 15 | 1,774 (+53.73%) | 32 | June 22 @ 1951.4 | 3270 (+1318.6) | June 29 @ 35.2 | 39 (+3.8) |
| June 22 | 3,270 (+84.33%) | 33 | June 29 @ 3597 | 6589 (+2992) | July 6 @ 36.3 | 48 (+11.7) |
| June 29 | 6,589 (+101.50%) | 39 | July 6 @ 7247.9 | 8587 (+1339.1) | July 13 @ 42.9 | 72 (+29.1) |
| July 6 | 8,587 (+30.33%) | 48 | July 13 @ 9445.7 | 10,855 (+1409.3) | July 20 @ 52.8 | ??? |
| July 13 | 10,855 (+26.41%) | 72 | July 20 @ 11940.5 | ??? | July 27 @ ??? | ??? |
So it honestly looks like the RØ is much higher than 1.1 at the moment and the death lag might be closer to 20-27 days from infection. If we go with 1.25 and 21 day lag, the table looks like this:
| Date | New Cases | New Deaths | Estimated infection increase (+7) | Actual new cases (∆) | Estimated death increase (+14) | Actual new deaths (∆) |
|---|---|---|---|---|---|---|
| June 1 | 726 | 30 | June 8 @ 907.5 | 1,154 (+246.5) | June 22 @ 37.5 | 33 (-4.5) |
| June 8 | 1,154 (+58.95%) | 37 | June 15 @ 1442.5 | 1774 (+331.5) | June 29 @ 46.25 | 39 (-7.25) |
| June 15 | 1,774 (+53.73%) | 32 | June 22 @ 2217.5 | 3270 (+1052.5) | July 6 @ 40 | 48 (+8) |
| June 22 | 3,270 (+84.33%) | 33 | June 29 @ 4087.5 | 6589 (+2501.5) | July 13 @ 41.25 | 72 (+30.75) |
| June 29 | 6,589 (+101.50%) | 39 | July 6 @ 8236.25 | 8587 (+350.75) | July 20 @ 48.75 | ??? |
| July 6 | 8,587 (+30.33%) | 48 | July 13 @ 10733.75 | 10,855 (+121.25) | July 27 @ ??? | ??? |
| July 13 | 10,855 (+26.41%) | 72 | July 20 @ 13568.75 | ??? | August 3 @ ??? | ??? |
That's certainly a little bit closer across the board, still not perfect, but closer. The mortality rate's growth is a bit alarming as it seems like it's closer to 2.2 on the 21 day lag. Also the lag seems to be extra-laggy during the onset of the outbreak. The first two death multipliers should be closer to 1.1 but it balloons quickly with the next multiplier being 1.5 and then 2.2. If it follows this trend (fingers crossed it doesn't), I would expect the upcoming values for 7-day moving average of new cases and deaths in Florida with (Floor: 1.95 / Ceiling: 2.3):
| Date | New cases | New deaths (floor) | New deaths (ceiling) |
|---|---|---|---|
| July 20 | 13568.75 | 76.05 | 89.7 |
| July 27 | 16960.94 | 93.6 | 110.4 |
| August 3 | 21201.17 | 140.4 | 165.6 |
Let's see how accurate eyeball measurements end up being!