Fun with COVID-19 reporting in MA (and elsewhere)

The state uses the currently known count for the last Saturday (a number 4 days in the making) and the Sunday 2 weeks prior (a number 17 days in the making) to determine their 2-week increase in case count per city/town. This is downplaying the count as explained in my post https://funwithmacovid-19reporting.blogspot.com/2020/10/why-weekly-community-numbers-reported.html   I'm going to start doing the more true calculation going forward for my weekly Greater Greater Somerville reporting.  Just thought I'd give you preview of how this adjusts the numbers: Area - State reported - More realistic number Arlington -    5.6 -   7.5 Boston  -     23.1 -  24.5 Cambridge -  8.9 -   9.9 Chelsea -     57.8  - 60.5 Fitchburg -  48.8  - 50.2 Holliston -   16.2  - 17.1 Lawrence -  82.6  - 88.4 (a top range of 40+ for my color coding really may not be sufficient) Lowell -       42.1...

Ok, well... last week's cases keep climbing and we now see... There is no weekday that saw fewer than 1848 cases...for comparison, There was no day from May 21 through October 18 that saw more than 1,000 cases There was no day from May 5 through the start of last week that saw 1848 cases Wednesday and Thursday currently sit at 2119 and 2291 cases respectively.  To get close to these number, you have to go back to April 29 which has 2187.  To get close to 2291, that's going back to April 24 with 2277 or April 23 with 2406. The average case count is now at 1620.3.  I realized today that I've been counting back toward April and have to ask, why am I not counting forward from when we closed down originally... or counting the number of weeks/days that reached the current average? We closed on March 17....  we had 31.8 cases the preceding week March 17 itself saw 249 cases reaching 1620.3? Sunday-to-Saturday averages only have 4 weeks greater than 1620.3... those start Apr...

The other day, I wrote a post about how Baker had done real harm to MA in relation to his handling of COVID-19. As if on cue, Baker's team rolled out a new color-coding for their neighborhood spread map. The new color coding says that a city/town's color coding will no longer be based solely on the rate of infection in that location.  Instead, a location with 50k or fewer people can have up 10 cases and be grey and a location with over 50k can have up to 15 cases and be grey, meaning that to be green, a location with under 10k people would need more than 5 times as high an infection rate as a location with 50k people.  That seems strange that a smaller population would need to see a much higher infection rate to be considered the same risk... and that measure continues throughout the new color coding.  For yellow, a population of <10k needs 16-25 total cases, which, if you have 9,999 people (to make for the smallest avg cases with 25 cases), you would have an average ...

THERE'S A NEW REPORT LAYOUT! (there will be a separate post about current counts/prediction follow up) It's always interesting to see a new layout because it gives insight into what the state's priorities and story-telling is.  Today's layout shows that they continue to be either sloppy with their design OR they want to try to confuse people by putting deaths and turn-around time for test results under the same heading line, similar to how all testing/cases are under one and how all information on hospitalizations are under a second line. We now have an "average age of deaths"... but we don't have which average it is, whether it's mean, median, mode, or whether it's overall or for the newly reported or the last 7 days or last 14 days or any kind of detail around it.... same with the new "average age of cases that were hospitalized". They're no longer showing the percent of individuals tested testing positive and only showing the perce...