"28. Some new variant not currently known is greater than 25% of cases: 60%
Hold. So far we’ve had Alpha, Delta and Omicron within two years, so 60% seems reasonable."
"13. WHO will designate another Variant Of Concern by year's end (75%): HOLD
Sell to 70%. Feels high to me, I dunno."
Here is my case for these being more likely.
Mutations are a function of number of replications, and the graphs going vertical everywhere with Omicron mean there is a lot more COVID replicating out there* than there ever was with earlier strains. Pretty soon there is going to be a huge selection advantage to escape immunity gained by Omicron, and since apparently Omicron infection gives Delta immunity it won't just be Delta coming back.
Is there anything wrong with how I'm thinking about this?
I think we'll still be doing enough sequencing somewhere to catch something that shows up in 25% of cases, even as most people stop testing, and infections get much more mild.
Buy 28 to 70%.
I have a gut feeling WHO will want to keep talking about COVID to counter people probably being very over it and going back to life as usual, but less confident.
Remember that I force myself into bins of 50/60/70/80/90/95/99 (or the inverse of these) in order to make grading easier, so when you (for example) sell from 5% to 3% you're not necessarily disagreeing with me. I don't know if there's a better way to handle this; maybe next year I'll break this custom even though it would make it harder to compare to past cycles.
That's fair. When I do +/- 5 it's an indication that I disagree a small amount with the anchor so I think that's real disagreement (e.g. mostly if you go 70% but mean 75%, and I say buy to 75%, if you'd said 75% I'd probably have gone to 80%).
In the 95%/99% buckets in particular I think that more detail is needed. It seems to me like thinking 3% vs. 5% is a big (40%!) difference, and 1% vs. 2% is even bigger. So either we shouldn't pick questions that are in the 1%-5% range, or we should be willing to be more specific.
With so few cases of you explaining your logic, and only calibrations for evaluation, it's hard to tell if we disagree in a particular case. I don't even do the calibration check, and no commenter has ever cared enough to bother either.
People seemed to think it would eventually converge on its most transmissible form. Sure, transmissibility is related to immune escape, and Omicron had a lot of immune escape. But that being said, just based on the physics and biology of viral reproduction, it seems like there has to be an overall transmissibility threshold. Otherwise, why doesn’t measles not have an R0 approaching 1000? or 10000? It’s had a lot of time to evolve.
So will we see variants with immune escape? Very likely. Will also they find an additional way to increase viral load significantly relevant to Omicron and thereby also replace Omicron and the most contagious strain? It just seems to me like Omicron has to be very near the peak of viral replication and load.
I agree with almost all your choices (among those that I have some opinion about). Except for vitamin D. It is currently not recognised as effective for anything except bone health and the positive studies are very much like ivermectin studies. Also Scott has much deeper understanding of medical regulatory environment, so I trust him more on this.
"28. Some new variant not currently known is greater than 25% of cases: 60%
Hold. So far we’ve had Alpha, Delta and Omicron within two years, so 60% seems reasonable."
"13. WHO will designate another Variant Of Concern by year's end (75%): HOLD
Sell to 70%. Feels high to me, I dunno."
Here is my case for these being more likely.
Mutations are a function of number of replications, and the graphs going vertical everywhere with Omicron mean there is a lot more COVID replicating out there* than there ever was with earlier strains. Pretty soon there is going to be a huge selection advantage to escape immunity gained by Omicron, and since apparently Omicron infection gives Delta immunity it won't just be Delta coming back.
Is there anything wrong with how I'm thinking about this?
I think we'll still be doing enough sequencing somewhere to catch something that shows up in 25% of cases, even as most people stop testing, and infections get much more mild.
Buy 28 to 70%.
I have a gut feeling WHO will want to keep talking about COVID to counter people probably being very over it and going back to life as usual, but less confident.
Still I'd buy back up to VOX's 75%.
*Including in my body currently. Yay!
Remember that I force myself into bins of 50/60/70/80/90/95/99 (or the inverse of these) in order to make grading easier, so when you (for example) sell from 5% to 3% you're not necessarily disagreeing with me. I don't know if there's a better way to handle this; maybe next year I'll break this custom even though it would make it harder to compare to past cycles.
I aim for halves, thirds, quarters, eights and sixteenths. Imperial mindset dies hard.
That's fair. When I do +/- 5 it's an indication that I disagree a small amount with the anchor so I think that's real disagreement (e.g. mostly if you go 70% but mean 75%, and I say buy to 75%, if you'd said 75% I'd probably have gone to 80%).
In the 95%/99% buckets in particular I think that more detail is needed. It seems to me like thinking 3% vs. 5% is a big (40%!) difference, and 1% vs. 2% is even bigger. So either we shouldn't pick questions that are in the 1%-5% range, or we should be willing to be more specific.
With so few cases of you explaining your logic, and only calibrations for evaluation, it's hard to tell if we disagree in a particular case. I don't even do the calibration check, and no commenter has ever cared enough to bother either.
> Hold. So far we’ve had Alpha, Delta and Omicron within two years, so 60% seems reasonable.
My gut says lower here, maybe 20%. I asked about six months ago if there was a theoretical asymptote for Covid R0. https://www.lesswrong.com/posts/mXBBHAEXj2JyPC6Dt/is-there-a-theoretical-upper-limit-on-the-r0-of-covid
People seemed to think it would eventually converge on its most transmissible form. Sure, transmissibility is related to immune escape, and Omicron had a lot of immune escape. But that being said, just based on the physics and biology of viral reproduction, it seems like there has to be an overall transmissibility threshold. Otherwise, why doesn’t measles not have an R0 approaching 1000? or 10000? It’s had a lot of time to evolve.
I think we have approached the transmissibility threshold with Omicron, where the viral load is higher than Delta https://health.ucdavis.edu/news/headlines/omicron-10-things-to-know-about-the-very-contagious-covid-19-variant/2022/01#:~:text=Omicron%20is%20highly%20transmissible%20compared,with%20the%20Delta%20variant.
So will we see variants with immune escape? Very likely. Will also they find an additional way to increase viral load significantly relevant to Omicron and thereby also replace Omicron and the most contagious strain? It just seems to me like Omicron has to be very near the peak of viral replication and load.
When two of the three comments are moving the same question in opposite directions, that's a sign of good predicting.
I agree with almost all your choices (among those that I have some opinion about). Except for vitamin D. It is currently not recognised as effective for anything except bone health and the positive studies are very much like ivermectin studies. Also Scott has much deeper understanding of medical regulatory environment, so I trust him more on this.