Correction: I somehow hadn’t read the section on “What should I do?” in sufficient detail. I now think that section is really really wrong, and terrible advice for most people, and was already at the time, and I very sorry for anyone who followed the advice in this post. Please, don’t follow it. Most of you are probably being too cautious.
The implied statement
“You accept you’re probably going to catch it (and in all likelihood at least soft prefer catching it now versus later). [If you don’t lock down]”
Is very wrong and even the most basic of precautions will very likely bring you below 50% infection risk. Do not assume you will get COVID unless you lock down. It is very easy for you to avoid the biggest sources of risk without locking down. Please continue seeing your closest friends (and also please continue avoiding your oldest relatives). Don’t unnecessarily socially isolate. Track your microcovids and see where that leads you, I bet it will cause most of you to discover you have been far too risk-averse.
Again, I am really sorry for promoting this without reading it carefully. I now think this was a very substantial mistake.
Here are the comments I sent to Robby and some others about my thoughts on the post:
The section I disagree with most, and that strikes me as quite wrong, is the following:
I broadly agree that this is the decision point and it’s basically all or nothing. Everyone needs to choose essentially one of three orientations.
The Protected—You protect yourself now and during the crisis, whatever it takes.
The Willing—You don’t protect yourself much now, but at some point shift to being in the first group if you haven’t caught it.
The Able—You accept you’re probably going to catch it and at least soft prefer catching it now versus later.
And also that everyone who can, should build up supplies again and start getting ready for an extended lockdown, possibly with supply line issues.
With the specific section on: “[If you don’t lock down] You accept you’re probably going to catch it and at least soft prefer catching it now versus later.”
I really don’t understand what model generates the “you are either locking down or you are going to catch it” prediction. Even just from a very simple model, you are not going to go above 75% of people infected, which means that as long as you are in the top 25% of carefulness, you have at least a better than 50% chance of not being infected.
And being in the top 25% of carefulness is trivial! You are probably going to reach that by wearing a KN95 instead of a cloth mask, while still going to work with random office workers and grocery shopping and hanging out with your friends who are similarly careful.
In practice, I think given the carefulness of our broader social network, we could probably throw 100 person parties every week, and still have it so that it’s less than 50% likely that any given person is going to catch with, given really modest contact tracing attempts (this is importantly different from throwing 100 person parties with people selected randomly from the population, that would be much more risky).
Overall, it seems that for the vast vast majority of people we know, just keeping up basic precautions, while still seeing your friends (mostly outdoors), probably working in-person if you are in a small organization, will easily keep COVID risk to an acceptable level even through the worst case scenarios of 85% infected.
Another perspective on this is that microcovids go up directly proportional to prevalence, and we were already at 2%-5% infected in early January, with probably at least 20% historically infected. It is not really possible for your microcovid risk to go up by more than a factor of 5, no matter how bad it gets, because you can’t have 10-25% of the population being infectious. In any geometric progression you reach heard immunity long before that (i.e. if you have an R of anything less than 3, you need to first have a period where you get 7% infected, and before that a period where it gets 1.5% infected, which already brings you very close to herd-immunity territory).
This means that no matter how bad it gets, the cost of any activity can’t really go up by more than 5x (in terms of microcovids, there are some additional considerations in terms of hospital overload, though that’s less relevant for a younger population). And I really doubt that the difference between “lockdown” and “no-lockdown” is actually within a factor of 5 for anyone. Every single time I have done microcovid calculations for people, they’ve found that they were drastically underspending, unless they were doing something very risky and dumb.
Overall, it strikes me that the binary advice of “lock down or expect to get it” is wrong for basically everyone I can think of, and errs far far too much on the side of risk-aversion.
Some quick clarifications on some statistical points, now that I am rereading this:
Even just from a very simple model, you are not going to go above 75% of people infected, which means that as long as you are in the top 25% of carefulness, you have at least a better than 50% chance of not being infected.
This was assuming something like a modest correlation of 0.5 with “carefulness” and “becoming infected”. In practice I expect it to be higher, and the above an underestimate (i.e. being in the top 25% of carefulness will give you a much better than even chance at not becoming infected, even if we reach herd immunity).
It is not really possible for your microcovid risk to go up by more than a factor of 5, no matter how bad it gets, because you can’t have 10-25% of the population being infectious. In any geometric progression you reach heard immunity long before that [...] (i.e. if you have an R of anything less than 3, you need to first have a period where you get 7% infected, and before that a period where it gets 1.5% infected)
I do think that R=3 and higher makes it possible to get a bit higher than 10-25% of the population being infectious, but not very much. But also, I would be really surprised if we reach R=3 and no model of the new strain I know of would predict that, so it’s meant as a strong upper bound. In practice I think the new strain, even if vaccinations don’t really pick up, will at most double or triple the microcovid costs.
Note: I wrote the OP, and I’ve asked Oliver to credit me in the post. I originally kept my name off it in case I needed to add controversial or crazy-sounding advice to the post at some point. But I no longer see scenarios where that happens, and leaving the post authorless prevents people from using the post to update about my epistemics/rigor/COVID knowledge/etc.
Oliver’s arguments make sense to me, and make me less concerned about the B.1.1.7 strain. I’m still a bit confused about why others disagree/disagreed with this?
Correction: I somehow hadn’t read the section on “What should I do?” in sufficient detail. I now think that section is really really wrong, and terrible advice for most people, and was already at the time, and I very sorry for anyone who followed the advice in this post. Please, don’t follow it. Most of you are probably being too cautious.
The implied statement
Is very wrong and even the most basic of precautions will very likely bring you below 50% infection risk. Do not assume you will get COVID unless you lock down. It is very easy for you to avoid the biggest sources of risk without locking down. Please continue seeing your closest friends (and also please continue avoiding your oldest relatives). Don’t unnecessarily socially isolate. Track your microcovids and see where that leads you, I bet it will cause most of you to discover you have been far too risk-averse.
Again, I am really sorry for promoting this without reading it carefully. I now think this was a very substantial mistake.
Here are the comments I sent to Robby and some others about my thoughts on the post:
The section I disagree with most, and that strikes me as quite wrong, is the following:
With the specific section on: “[If you don’t lock down] You accept you’re probably going to catch it and at least soft prefer catching it now versus later.”
I really don’t understand what model generates the “you are either locking down or you are going to catch it” prediction. Even just from a very simple model, you are not going to go above 75% of people infected, which means that as long as you are in the top 25% of carefulness, you have at least a better than 50% chance of not being infected.
And being in the top 25% of carefulness is trivial! You are probably going to reach that by wearing a KN95 instead of a cloth mask, while still going to work with random office workers and grocery shopping and hanging out with your friends who are similarly careful.
In practice, I think given the carefulness of our broader social network, we could probably throw 100 person parties every week, and still have it so that it’s less than 50% likely that any given person is going to catch with, given really modest contact tracing attempts (this is importantly different from throwing 100 person parties with people selected randomly from the population, that would be much more risky).
Overall, it seems that for the vast vast majority of people we know, just keeping up basic precautions, while still seeing your friends (mostly outdoors), probably working in-person if you are in a small organization, will easily keep COVID risk to an acceptable level even through the worst case scenarios of 85% infected.
Another perspective on this is that microcovids go up directly proportional to prevalence, and we were already at 2%-5% infected in early January, with probably at least 20% historically infected. It is not really possible for your microcovid risk to go up by more than a factor of 5, no matter how bad it gets, because you can’t have 10-25% of the population being infectious. In any geometric progression you reach heard immunity long before that (i.e. if you have an R of anything less than 3, you need to first have a period where you get 7% infected, and before that a period where it gets 1.5% infected, which already brings you very close to herd-immunity territory).
This means that no matter how bad it gets, the cost of any activity can’t really go up by more than 5x (in terms of microcovids, there are some additional considerations in terms of hospital overload, though that’s less relevant for a younger population). And I really doubt that the difference between “lockdown” and “no-lockdown” is actually within a factor of 5 for anyone. Every single time I have done microcovid calculations for people, they’ve found that they were drastically underspending, unless they were doing something very risky and dumb.
Overall, it strikes me that the binary advice of “lock down or expect to get it” is wrong for basically everyone I can think of, and errs far far too much on the side of risk-aversion.
Some quick clarifications on some statistical points, now that I am rereading this:
This was assuming something like a modest correlation of 0.5 with “carefulness” and “becoming infected”. In practice I expect it to be higher, and the above an underestimate (i.e. being in the top 25% of carefulness will give you a much better than even chance at not becoming infected, even if we reach herd immunity).
I do think that R=3 and higher makes it possible to get a bit higher than 10-25% of the population being infectious, but not very much. But also, I would be really surprised if we reach R=3 and no model of the new strain I know of would predict that, so it’s meant as a strong upper bound. In practice I think the new strain, even if vaccinations don’t really pick up, will at most double or triple the microcovid costs.
Note: I wrote the OP, and I’ve asked Oliver to credit me in the post. I originally kept my name off it in case I needed to add controversial or crazy-sounding advice to the post at some point. But I no longer see scenarios where that happens, and leaving the post authorless prevents people from using the post to update about my epistemics/rigor/COVID knowledge/etc.
Oliver’s arguments make sense to me, and make me less concerned about the B.1.1.7 strain. I’m still a bit confused about why others disagree/disagreed with this?