Part of me is worried that this wave is subsiding too fast and that we might get hit one more time before we're through this disaster. As always, great analysis!
Really helpful comparison as always. I actually think the bump in cases after Christmas has some real increase in infections behind it, in addition to reporting delays. I had looked into this carefully just for Wisconsin, using cases and tests dated to the test date instead of the reporting date, and it looks like cases and positivity rate both went up modestly in the weeks after the holidays, before resuming the downward trend.
It looks like adding the 100M Target line to the vaccine graphs shifted the colors of the States by one for each region for the vaccine graph compared to the case and death graphs. Might want to put the 100M Target line last in the order, rather than first, to avoid this.
I’ve been tracking California county-by-county since March 2020, and I’ve seen a few things. First, counties are getting very sloppy with reporting as the months have gone on, tending to be perfectly happy to report straight zeros for a week and then 30+ deaths in a day. They also do this with cases — Tuesday is “catch-up day” for many counties, except it’s not always Tuesday and it’s not as if it is 0-0-0-all three days. Thus seven day rolling averages are much more useful. Second, the average age of death is moving up, not down, even though the average case age is moving down. Third, there are now seven counties with above 10% infections. LA is one of them. And when LA hit 10%, their new cases quickly began to fall. Fourth, I expect the new “floor” to remain around the level of spring peaks because more counties seem to now have solid community spread, so even with the more populous counties of the southland (LA, Orange, San Diego, Riverside, and San Bernardino) seeing declines in new cases, the floor is now raised. Fifth, the counties love to change their data with little warning — I have a gap of 2374 cases versus state reporting and where those are is isn’t easy to to find, but it’s been that number for four months. When it changes, it is usually because a county restated numbers from the summer or early fall. Sacramento county moved 3000 of their cases off their books and onto other counties, which was a PITA to clean up (I can’t just re-download the data in most cases) Sixth, on a county-by-county basis, the spikes look weather driven, but the trend upward preceded the weather in some counties, so shrug? Last, the California prison data I mentioned before is holding — effectively no new cases in the institutions with 68%+ cases (one or two, but those are likely just new prisoners or transfers), and the facilities with over 50% cases don’t seem to be experiencing many new cases either. The prisons also do weird things to the county numbers, as Lassen has 16% cases (highest in the state) but that is because the county population is 37k, and the two prisons hold 8000+ prisoners more than half of whom were cases.
Part of me is worried that this wave is subsiding too fast and that we might get hit one more time before we're through this disaster. As always, great analysis!
Really helpful comparison as always. I actually think the bump in cases after Christmas has some real increase in infections behind it, in addition to reporting delays. I had looked into this carefully just for Wisconsin, using cases and tests dated to the test date instead of the reporting date, and it looks like cases and positivity rate both went up modestly in the weeks after the holidays, before resuming the downward trend.
https://covid-wisconsin.com/2021/01/24/christmas2/
It looks like adding the 100M Target line to the vaccine graphs shifted the colors of the States by one for each region for the vaccine graph compared to the case and death graphs. Might want to put the 100M Target line last in the order, rather than first, to avoid this.
I’ve been tracking California county-by-county since March 2020, and I’ve seen a few things. First, counties are getting very sloppy with reporting as the months have gone on, tending to be perfectly happy to report straight zeros for a week and then 30+ deaths in a day. They also do this with cases — Tuesday is “catch-up day” for many counties, except it’s not always Tuesday and it’s not as if it is 0-0-0-all three days. Thus seven day rolling averages are much more useful. Second, the average age of death is moving up, not down, even though the average case age is moving down. Third, there are now seven counties with above 10% infections. LA is one of them. And when LA hit 10%, their new cases quickly began to fall. Fourth, I expect the new “floor” to remain around the level of spring peaks because more counties seem to now have solid community spread, so even with the more populous counties of the southland (LA, Orange, San Diego, Riverside, and San Bernardino) seeing declines in new cases, the floor is now raised. Fifth, the counties love to change their data with little warning — I have a gap of 2374 cases versus state reporting and where those are is isn’t easy to to find, but it’s been that number for four months. When it changes, it is usually because a county restated numbers from the summer or early fall. Sacramento county moved 3000 of their cases off their books and onto other counties, which was a PITA to clean up (I can’t just re-download the data in most cases) Sixth, on a county-by-county basis, the spikes look weather driven, but the trend upward preceded the weather in some counties, so shrug? Last, the California prison data I mentioned before is holding — effectively no new cases in the institutions with 68%+ cases (one or two, but those are likely just new prisoners or transfers), and the facilities with over 50% cases don’t seem to be experiencing many new cases either. The prisons also do weird things to the county numbers, as Lassen has 16% cases (highest in the state) but that is because the county population is 37k, and the two prisons hold 8000+ prisoners more than half of whom were cases.
I speculate that low average age is the biggest reason that Utah's deaths have stayed so low despite modestly high numbers of per capita infections.