I like the red line, but I wonder if it's a good idea if it's "somewhat arbitrary." Is it more like "totally" arbitrary or closer to "I made a decision based on some averages and stats that are nebulous but close-ish?" Hospital capacity seemed like a good red line early on give the "flatten the curve" narrative, but that doesn't work state-wide. Big city hospitals can be way overloaded while the rest are quiet, leading the statewide capacity average looking fantastic. God, this data is complicated.
Yes! One of the reasons I like the "10 new cases per day per 100K" is because, having looked at this data SO MUCH over the last 4 months, I've seen states go slightly above it but then come back down without much worry. We end up seeing a lot of movement between 5 and 10 that doesn't seem to correspond to higher hospitalizations or deaths. Not that we should ignore states that are below 10 cases, but it does seem to be a fairly safe zone.
Definitely becomes less "arbitrary" if it seems to be a nice round number that presents a tipping point of sorts. Given how little we still know, though, it does make sense to pick a measuring stick and go with it. At least for an analysis like this one.
Perhaps I am misreading/misunderstanding, but you write:
"In every chart, I’ve added a red line. For cases, the line is 10 new cases per day per 100K residents. When cases rise above this line, it’s a sign that things *could* go very badly. For deaths, it is at 0.1 deaths per day per 100K residents (one death per million per day)."
It appears that the red line in the death charts is at 1 per 100K (per the y axis and title).
Enjoy the posts and your analysis. Am beginning to wonder whether most states are simply going to reach 400 to 800 deaths per million regardless of what they do, except if they are mostly rural or can otherwise be relatively isolated. The only issue is when it happens. Even if you get covid under control you can't eradicate it and it just needs an ember to start up again. We may be at the point where we can hopefully keep it from overwhelming the medical system and doctors have a better idea how to treat (though not cure) so that may play a role in lower mortality rates. Even in other countries like Spain and Israel we are seeing flare ups where things were seemingly "under control".
I like the red line, but I wonder if it's a good idea if it's "somewhat arbitrary." Is it more like "totally" arbitrary or closer to "I made a decision based on some averages and stats that are nebulous but close-ish?" Hospital capacity seemed like a good red line early on give the "flatten the curve" narrative, but that doesn't work state-wide. Big city hospitals can be way overloaded while the rest are quiet, leading the statewide capacity average looking fantastic. God, this data is complicated.
Yes! One of the reasons I like the "10 new cases per day per 100K" is because, having looked at this data SO MUCH over the last 4 months, I've seen states go slightly above it but then come back down without much worry. We end up seeing a lot of movement between 5 and 10 that doesn't seem to correspond to higher hospitalizations or deaths. Not that we should ignore states that are below 10 cases, but it does seem to be a fairly safe zone.
Definitely becomes less "arbitrary" if it seems to be a nice round number that presents a tipping point of sorts. Given how little we still know, though, it does make sense to pick a measuring stick and go with it. At least for an analysis like this one.
Perhaps I am misreading/misunderstanding, but you write:
"In every chart, I’ve added a red line. For cases, the line is 10 new cases per day per 100K residents. When cases rise above this line, it’s a sign that things *could* go very badly. For deaths, it is at 0.1 deaths per day per 100K residents (one death per million per day)."
It appears that the red line in the death charts is at 1 per 100K (per the y axis and title).
You are totally correct and I am deeply embarrassed! I've fixed it and issued a recall on all my emails.
No worries, thanks! Keep up the good work!
Enjoy the posts and your analysis. Am beginning to wonder whether most states are simply going to reach 400 to 800 deaths per million regardless of what they do, except if they are mostly rural or can otherwise be relatively isolated. The only issue is when it happens. Even if you get covid under control you can't eradicate it and it just needs an ember to start up again. We may be at the point where we can hopefully keep it from overwhelming the medical system and doctors have a better idea how to treat (though not cure) so that may play a role in lower mortality rates. Even in other countries like Spain and Israel we are seeing flare ups where things were seemingly "under control".