Strong +1 for writing the post-mortem—there are incentives not to write that kind of thing, but I appreciated it. I also get value out of your weekly posts, so thanks for writing those.
That being said, I do have a few minor quibbles with the post-mortem. I think I interpreted your December (and future month) claims more strongly than it sounds like you interpreted them.
I haven’t re-read all your posts carefully to check for places you might have said this, but I don’t remember seeing you say in the 1-2 months after your “We’re F***ed” post that you thought it was significantly less likely that we’re f***ed. E.g. I commented on a January post https://www.lesswrong.com/posts/Rhy5g75NdRKdw9ibB/?commentId=2TjkS4XLA5PaAjoxG#2TjkS4XLA5PaAjoxG saying that the UK evidence seemed pretty strong that we were possibly not f***ed.
I was also somewhat surprised that in the post-mortem, you said you thought the 70% prediction evaluated to false.
From that post:
The new strain has rapidly taken over the region, and all signs point to it being about 65% more infectious than the old one, albeit with large uncertainty and error bars around that.
I give it a 70% chance that these reports are largely correct.
There is no plausible way that a Western country can sustain restrictions that can overcome that via anything other than widespread immunity.
I haven’t looked into the details of this much at all since February, so I don’t have strong takes on whether the UK strain is closer to 40% or 55% more infectious (and “more infectious” can be defined in different ways). Comments on this metaculus question https://pandemic.metaculus.com/questions/6089/50-transmissible-variant-to-infect-10m/ suggested that people, at the time at least, believed the number was likely close to 55% (note: the overall probability on the linked question also includes the probability of other strains infecting >10M people by June 2). There’s also a decent amount of room for different interpretations in “largely correct,” and a 10% difference in infectiousness matters a lot, but at the time at least, I would have predicted that 55% was close enough to 65% to hit your “largely correct” bar.
One reason I pushed back on some of your posts in Dec—Jan is that I thought the post was directionally wrong for the LessWrong community (I think LessWrong readers as a whole were much too cautious as of Dec 2020). I think people in the community deferred to the reasoning of “We’re F***ed” some, and that it would’ve been better if you had more quickly updated / told people that you had updated.
Overall though, I appreciate you writing these posts, so I hope my comment doesn’t come off as too critical, and thank you again for doing that.
I agree that 55% would be an ambiguous evaluation, but 40% is substantially different than 65% and I’ve been using 40% in my recent models. Central mistake is the same in both cases.
I don’t remember what my update timeline was, I’d have to go back and look. Makes sense I would be too slow.
Strong +1 for writing the post-mortem—there are incentives not to write that kind of thing, but I appreciated it. I also get value out of your weekly posts, so thanks for writing those.
That being said, I do have a few minor quibbles with the post-mortem. I think I interpreted your December (and future month) claims more strongly than it sounds like you interpreted them.
I haven’t re-read all your posts carefully to check for places you might have said this, but I don’t remember seeing you say in the 1-2 months after your “We’re F***ed” post that you thought it was significantly less likely that we’re f***ed. E.g. I commented on a January post https://www.lesswrong.com/posts/Rhy5g75NdRKdw9ibB/?commentId=2TjkS4XLA5PaAjoxG#2TjkS4XLA5PaAjoxG saying that the UK evidence seemed pretty strong that we were possibly not f***ed.
I was also somewhat surprised that in the post-mortem, you said you thought the 70% prediction evaluated to false.
From that post:
I haven’t looked into the details of this much at all since February, so I don’t have strong takes on whether the UK strain is closer to 40% or 55% more infectious (and “more infectious” can be defined in different ways). Comments on this metaculus question https://pandemic.metaculus.com/questions/6089/50-transmissible-variant-to-infect-10m/ suggested that people, at the time at least, believed the number was likely close to 55% (note: the overall probability on the linked question also includes the probability of other strains infecting >10M people by June 2). There’s also a decent amount of room for different interpretations in “largely correct,” and a 10% difference in infectiousness matters a lot, but at the time at least, I would have predicted that 55% was close enough to 65% to hit your “largely correct” bar.
One reason I pushed back on some of your posts in Dec—Jan is that I thought the post was directionally wrong for the LessWrong community (I think LessWrong readers as a whole were much too cautious as of Dec 2020). I think people in the community deferred to the reasoning of “We’re F***ed” some, and that it would’ve been better if you had more quickly updated / told people that you had updated.
Overall though, I appreciate you writing these posts, so I hope my comment doesn’t come off as too critical, and thank you again for doing that.
I agree that 55% would be an ambiguous evaluation, but 40% is substantially different than 65% and I’ve been using 40% in my recent models. Central mistake is the same in both cases.
I don’t remember what my update timeline was, I’d have to go back and look. Makes sense I would be too slow.