Nothing both unexpected and important happened this week, other than that we might be on the verge of at least partially solving obesity.
There’s plenty else to talk about, but everything was on trend. That doesn’t mean my predictions were perfect or the speed of trends didn’t give us a few surprises, nor does it mean I could have written this post a week ago. It does mean that long term expectations haven’t much changed.
Vaccine efforts continue to be far too slow to improve, but to still improve slowly, including both distribution and approval. All current numbers continue to steadily decline, as they should until the new strains threaten to reverse that, which should still be a few weeks away.
While we can hope for unexpected good news, the more realistic and modest hope is that the lack of major news continues but things continue to steadily improve, and vaccine access improves dramatically in March after the approval of Johnson & Johnson combined with accelerating production across the board.
The Covid Tracking Project will be ending soon, and I still don’t have an great alternative source I love, so if you do have one (either for downloading data, for grabbing good data manually slash copy/pasting into a spreadsheet, or both) please suggest it in the comments.
In the meantime, let’s run the numbers.
Last week: 7.7% positive rate on 12.7 million tests, and an average of 3,041 deaths per day.
Prediction for this week: 6.9% positive rate and an average of 2,750 deaths. Current trends should continue, with minimal impact yet from new strains, changed policies or increased vaccinations.
Results: 6.4% positive rate and an average of 2,968 deaths.
The 6.4% number isn’t shocking and was definitely in my range of plausible outcomes, but it’s pretty great. It’s even better than it looks, because our test counts are in rapid decline as well, from 12.2 million to 11.3 million, which presumably reflects lower demand. Things are improving rapidly, if not rapidly enough to stall the rise of the English strain, but things could plausibly be 25% better week-over-week, and that’s going to buy us time to improve further. This is reason to be hopeful.
The 2,968 deaths are highly disappointing, as that’s only a few percent less than last week. The 14-day-delayed-from-diagnosis-to-deaths rolling 7-day CFR went up from 1.29% to 1.57%, but on the other hand the 21-day-delayed-from-diagnosis-to-deaths rolling 7-day CFR went down from 1.34% to a new low of 1.25% (technically the low was 1.23% yesterday).
When looking at regional totals, I noticed that Indiana added 1,507 deaths this week that happened previously, as the result of an audit.
When you take those away, that drops the average death rate down to 2,752 deaths per day. Bullseye.
Prediction: 5.7% positive rate and an average of 2,450 deaths. Things should still continue to improve.
Dec 10-Dec 16
Dec 17-Dec 23
Dec 24-Dec 30
Dec 31-Jan 6
Jan 7-Jan 13
Jan 14-Jan 20
Jan 21-Jan 27
Jan 28-Feb 3
Feb 4-Feb 10
That’s a very large uptick in the Midwest, so the question is what is going on there. There certainly was never a corresponding jump in positive tests. Looking at the details, the explanation is that more thanall of it comes from February 4 in Indiana, with 1,518 deaths. The Covid Tracking Project explains that these additional deaths come from a historical audit, so they’re not from last week at all. For this week’s update I’ll leave them in place, but will be editing the extra 1,507 past deaths out of the weekly charts going forward. At that point, things look normal.
Dec 31-Jan 6
Jan 7-Jan 13
Jan 14-Jan 20
Jan 21-Jan 27
Jan 28-Feb 3
Feb 4-Feb 10
Excellent progress. Not enough to overcome the new strain, but still very good, and even better accelerating improvement.
Dec 17-Dec 23
Dec 24-Dec 30
Dec 31-Jan 6
Jan 7-Jan 13
Jan 14-Jan 20
Jan 21-Jan 27
Jan 28-Feb 3
Feb 4-Feb 10
Last week there were signs the growth in vaccination rate might be stalling out. Fortunately that has not happened, and supply continues to increase at a good clip.
The current pace is not fast enough by any means, but if we can continue to grow at 20% week over week with the Pfizer and Moderna vaccines alone, it won’t be long before we get where we need to be.
It doesn’t feel like things are accelerating, because eligibility was expanded sufficiently that there are no appointments anywhere and no one has any supply, but that will change. As with many growing phenomena, more of the action takes place later in time than it would seem to people who have been paying attention for a while.
Globally things are mostly similar to last week, with Chile putting on a strong showing.
It is clear that the United Kingdom has contained the English strain through its countermeasures. The countermeasures in place elsewhere do not appear to be sufficient to do that, but the existence proof of the option is definitive. Things are generally improving everywhere, with Spain improving especially rapidly. As always, there are other countries with things going on but I post the same list each week to remain consistent and avoid cherry picking, with the USA on here as a comparison point.
Max says this is obviously very bad news, but whether something is good or bad news depends on your prior expectations.
If you previously had your head in the sand and were trying to pretend that the English Strain wasn’t more infectious, this is indeed very bad news.
If, however, you previously had accepted that the English strain was more infectious, and the question was how much more infectious, then news of the answer could be good or bad. In this case, it’s good.
This is an estimated 37% increase in infectiousness of the new strain versus the overall mix of strains. My initial version got this calculation wrong and forgot that there's already a lot of the new strain in Denmark, which pushes this into the 40s. Even that small difference is big, although other official estimates are still over 50%. Compared to 50%, even 45% is effectively much better. The difference is enough to give us somewhat of a puncher’s chance of things not being so bad, both buying us time and reducing how bad it is when the time comes.
This is a common pattern. For a while in March of 2020, the stock market was reliably moving in the wrong direction on news. They’d see that things were shutting down and update that things were worse than they thought, whereas from my perspective things shutting down meant things were better than I thought, because I already knew that they should shut down and now they were at least both aware of the situation and thus actually shutting down. Which was good news.
This is an attempted Bayesian analysis of various vaccines and their likely efficacy against various severities of disease and against various strains. I say attempted because it seems like these take relatively generic prior distributions, and don’t adjust them much based on our understanding from other data points, including data from other vaccines. The result is that these distributions are likely underconfident, and in some cases biased, but it’s still a worthwhile exercise.
Last week I went over how we know the Johnson & Johnson vaccine is safe and effective, and there are millions of doses waiting to be distributed, and there’s no good reason we can’t start that process yesterday.
I do realize that there is a difference between, as Scott Alexander discusses, the FDA’s need to be legible and reliable, and follow proper procedures, versus my ability to apply Bayesian reasoning.
Mostly, it’s a call and response. You say ‘why are we letting people die for no reason?’ and they say ‘Thalidomide!’ and ‘people won’t trust it.’
So basically, one time someone had a drug that wasn’t safe. We didn’t approve that drug because our existing review process made it look unsafe, so in response to that we created a more involved and more onerous process, as opposed to noticing that the previous process actually worked in this case exactly as designed. Then we use this as a fully general excuse to freak everyone out about everything that hasn’t gone through this process, and then use that freak out (that, to the extent it exists which it mostly doesn’t, is directly the result of such warnings) as our reason to force everything through the process. Neat trick.
Oh, and did I mention that the ‘safety data’ that requires three weeks to review is, and I quote it in its entirety, ‘nothing serious happened to anyone at all, and no one was struck by lightning.’ Either J&J has created a safe vaccine, or J&J is committing a fraud that will be caught and get everyone involved arrested within three weeks, or they’re committing a fraud so effectively that the review won’t catch the fraud and won’t help. Those are the only possibilities. If the data isn’t fraudulent then the drug is safe, period.
It’s one thing to have Asymmetric Justice. I get that the FDA has to act as if damage from permission is orders of magnitude more important than damage from lack of permission, despite how awful that is for people’s lives, but how many orders of magnitude are we talking, and once you pick a number can you then show your work?
The question is actually kind of weird and terrible – if they move the meeting or issue the EUA before the meeting it resolves to ambiguous and is cancelled, so the question it’s asking is ‘conditional on having the meeting without approval do they then issue the EUA within 7 days’ and that’s not the thing we actually want to know. We want to know if the EUA will eventually be issued one way or another, whereas this question is actually mostly measuring whether they drag their feet so long it takes more than a week while they hash out the exact minimum age and procedural guidelines, or something.
Also, I broke down and joined Metaculus for this question because 90% seemed crazy low, and this is not how odds work this is not how odds work at all:
But then I checked another question, and hey, you know what gets one even more internet points than good odds?
On the actual J&J vaccine, I don’t know what more there is to say. As with Moderna and Pfizer, they’ve already done the actual approval process and confirmed that it’s going to get approved before they applied, and now we’re delaying in order to make it clear we are Very Serious People who Follow Proper Procedure and are not In Bed With Industry and Putting People At Risk or Destroying Trust in Vaccines by going ‘too fast.’ Or something like that.
We have now done this three times. It’s one thing to have the first vaccine application point out that there’s weeks of lost time. It’s another thing to not have fixed the problem months later.
Waiting on Novavax
From the USA article on Johnson & Johnson, we learn exactly what is holding up Novavax, and the problem is a straightforward case of Not Clinically Trialed Here:
Waiting on AstraZeneca
If Novavax needs an American trial, I guess we have our answer on AstraZeneca.
Not counting implicit arguments from full-on anti-vax people, I have yet to see one single person argue that we should undo the approval of AstraZeneca in the UK or EU.
Nor did I see a single call for stopping the use of the AstraZeneca vaccine in South Africa. That’s presumably because this would be a really stupid move. There was a small and therefore underpowered study of the AZ vaccine against the South African strain, which is easy to misinterpret since it couldn’t conclude anything definitively due to lacking sufficient power. The AZ vaccine still clearly provides some protection against infection by the South African strain, and likely greater protection against severe disease, and it’s safe, so what’s the downside to continuing to give it out?
Certainly if the choice is between ‘some reduction in deaths, hospitalizations and severe disease’ and ‘no thanks, we prefer the full quantity of deaths, hospitalizations and severe disease’ this choice should be clear enough.
You would, of course, be wrong about that. South Africa did stop the use of the AZ vaccine (WaPo), and again, I have not heard any experts point out how suicidal that decision is, although shifting future vaccine purchases to other vaccines did make sense.
This follows the pattern that ‘expert opinion’ on what to do is primarily defending following elite decision making. I’ve been underestimating this effect, but the pattern is increasingly clear. If the properly respectful elites with the power (e.g. basically anyone powerful except the Orange Man) decide not to do X yet, the ‘experts’ will talk about how we must follow proper procedures before doing X or else risk catastrophe, or how X is damaging or unethical or something. Once the properly respectful elites decide to do X, whether or not they followed the procedures the ‘experts’ were defending earlier, the ‘experts’ might briefly decry the skipped procedure, but they will immediately pivot to defending X. There’s presumably some value of X sufficiently awful that the ‘experts’ wouldn’t do this, but we’ve yet to prove that by example.
So we have the same vaccine sitting in storage unused in Baltimore, while failure to secure and deliver enough of that same vaccine for current use is a continent-wide scandal in Europe.
And in case it wasn’t sufficiently clear last time, the reason AstraZeneca didn’t apply is that they’ve been told not to and that they’d be turned down. That’s how the process works, you’re in constant communication and only apply when you have what is necessary.
First we had a shortage of eligible arms into which to put shots, resulting in lots of needless delay. We were so obsessed with prioritization and what was fair that we threw doses away while letting others sit on shelves.
Then we got to the current phase of the vaccine rollout, where there is an abundance of arms and not enough shots to put in those arms. Now we are thankfully using all our doses, but with a train wreck of a prioritization scheme that places lots of people at minimal risk (e.g. anyone with a BMI of 30 in many places) in competition against seniors for shots, while destroying the legitimacy of the prioritization system. It’s worse than nothing, but far better than too much, and the focus rightfully has shifted to finding more doses faster, increasing production, cutting vaccine doses where we can and approving new vaccines quickly.
Possible is not the strongest word, but the other part of this statement is claiming relatively high confidence. The key question is whether the 250M doses target is real and how much of that is one-shot J&J doses. If we hit that mark, then our chances to get everyone in the ‘eager to get it’ group done by April look reasonably good.
Peter Hotez, in the linked thread, has been hearing the other half of the news reports, and also echoes the same range I have for how long we have before the new strains take over:
We can and should do our best to move the needle but the most important practical question in so many of our lives is which of these two reports is right. Will we have sufficient doses for ‘the eager’ in April, or will it be waiting until at least June?
There continues to be very high confidence that the necessary doses will be distributed by the end of May. As usual, they are overconfident and C is nothing like 90%, but the level of confidence still speaks volumes – Scott Gottlieb is making an explicit prediction of outcome B, and Peter Hotez is soft expecting option D.
Doing quick math, if Scott’s other half of the equation is right and there’s only ‘deep’ demand for 50-75 million vaccinations, then compare that to our current pace of 1.5mm+ doses per day, or 750k people vaccinated per day with 44 million shots already administered. Getting to 150 million shots from here would only be two months (mid-April) even if we didn’t accelerate from here. So it seems like the end of infinite demand, and the shift to a potential demand shortfall, is likely rather near.
I could not be more contemptuous and sick of the ‘nothing changes when you get vaccinated’ position. Or even this nonsensical ‘we don’t know’ position. I took a graduate level course in epistemology where it was called into question what it means to know that some proposition P is true, and whether we can actually know anything at all, and that’s all quite interesting as philosophy, but for practical purposes yes we damn well know this stuff.
My guess continues to be that this hesitation will fade with time. Wait for the people who actively want the vaccine to get it, and then everyone who can’t do their job without a vaccine to get the vaccine, and then we’ll have many months of tens of millions of people as clear evidence that the vaccines work, and a lot more people will follow. Thus, I’d be inclined to mostly wait on the public service campaigns until then, when people will be able to act if convinced and we’ll have an easier time convincing them.
I also notice that there is a large part of me that thinks, once it’s easily and widely available, you know what? Straight up, just f*** ‘em if they don’t want the vaccine.
Suppose it is June 15, and there are open same-day appointments outside of work hours at half the pharmacies in America. The good stuff, too. Anyone who wants a free shot of Moderna or Pfizer can get a free shot, no questions asked.
Remember, we seem to have vaccines that are almost 100% effective against death and severe disease, and everyone’s had the chance to get vaccinated, and can change their mind at any time.
Why not simply go back to normal? If the other half of America doesn’t want the vaccine, all right, fine. Some of them can keep hiding in their houses for another year if it suits them. Others can walk around vulnerable and shake everyone’s hand. This large part of me really, really doesn’t care. Still do the PSAs and all that, sure, but it’s their lives on the line. Either we’ll still contain the pandemic anyway, or the ones who are out there unvaccinated taking big risks will infect each other, and that will burn itself out, and then either way we can all move on with our lives.
I don’t endorse that position on reflection or anything, but it seems important to have the courage to note that it is there.
In particular, the whole ‘we’ve told people that the vaccine won’t stop them from getting Covid or from passing Covid to others or let them resume their normal lives, so why the hell should we expect them to take it?’ counter-argument is getting stronger every time Zeynep retweets another example, and her list is getting long.
Vaccine Allocation By Politics And Power
New York joins the club letting those who technically have a ‘comorbidity’ of a BMI of 30 get vaccine priority while seniors continue to be unable to book appointments. How the wheel turns. Looks like the list at least excludes smokers.
The National Center for Health Statistics estimates that, for 2015–2016 in the U.S., 39.8% of adults aged 20 and over were obese (including 7.6% with severe obesity). Luckily in my first reading I'd mixed up which statistics are referring to BMI of 25 vs. 30 (they use the same words in different places to refer to both, but still totally my fault), and it's only that 39.8% that are eligible in this fashion via the BMI threshold, if everyone is scrupulously honest and doesn't game the system in any way.
Right away, the main mailing list I am on had one person note that he was four pounds short of qualifying, and a plan was quickly formed (I believe and hope in jest but these days who can know) to eat some carbs and put on a little water weight to get over the top. There seems to be a lot of somewhat serious talk about putting on a few pounds to hit the cutoff point.
Of course, a better strategy is to just lie, it’s not like anyone is going to weigh you. This isn’t a boxing match.
I wonder how much long term damage this is going to do to people’s health by demotivating healthy lifestyle choices, the same way I worried earlier about the impact on smoking rates. I don’t think these impacts will be worse than the health effects of the pandemic, but if you think that kind of damage is impossible here, you have not done the math.
You can’t undo eligibility that’s been granted. I mean, sure, you can, but it’s pretty terrible. What’s done is done. We’d be better off without eligibility rules at all, and relying on people to use their judgement and their willingness to invest time in the process as allocation methods (despite the ability to do this being anti-correlated with actual need to get the vaccine, both in terms of exposure and vulnerability), since we are unwilling to use price.
The WHO is suddenly enamored with the ‘frozen food’ hypothesis for the origin of Covid-19, which would claim that the virus didn’t even originate in China, and I will leave interpreting what is happening there as a (very easy) exercise for the reader.
If you’ve already had Covid-19, the second dose is likely not helpful to you. The first dose will have been enough. The second dose also knocks a lot of people on their ass for a day, so if it isn’t helpful one would actively want to avoid it for one’s own sake, plus it gives that dose to someone else who still needs it.
The best way to improve productivity and efficiency is to do the thing, in this case Pfizer using their experience to double production. Which they could have done last year. It occurs to me that if we had offered to pay $100 per dose for mRNA vaccines all of last year and then tossed them into the trash after testing their composition rather than even bother storing them we’d be done with this pandemic by now.
The UK vaccine rollout is considered a success, and by the standards of other results, it is indeed a success. This interview explains how they did it, which was essentially ‘make deals with companies and pay them money in exchange for doses of vaccines.’ They still did much less than a sufficient amount of that across the board, and it doesn’t seem like anyone involved grokked anything like the full speed premium or the need to overspend, but they at least threw some effort in the right directions.
How do we know politics has returned to normal? “News stories” like this one on Fox News, complaining that Biden was travelling by plane despite CDC warnings not to travel. That’s how you keep life, liberty and property relatively safe while the legislature is in session.
Cases are declining across the world, here’s an analysis of why that comes down to behavioral changes combined with ~30% of people being immune due to prior infection and that allowing us to turn the corner. He points out that temperatures of various places don’t much correlate with current trends. Of course, as the new strains take over, this progress will by default be lost and reversed, and current levels overall remain quite bad.
There used to be a robust consensus that the right to free speech, both in law and in practice, from threats both public and private, was one of our core values. I would argue it was the core value. There is now an increasing consensus against the right to free speech. This has progressed from the rapidly expanding umbrella of ‘hate’ speech (regardless of its truth value), to speech in support of the political outgroup or against ingroup causes that the Powers That Be have decided to proclaim as false or dangerous.
Increasingly it also applies to health claims that are in conflict with Officially Recognized Sources of truth. Like the WHO, FDA and CDC.
Professional retaliation for speech in some cases has extended to merely having joined (likely to reserve a preferred username) social networks that have arisen as alternatives to this censorship, even the person in question ever having posted anything at all on such networks. Why would you have access to a way to say something, if you didn’t have something to say?
I write these posts on WordPress and they are reposted on LessWrong, which has strong norms around front page posts, but a very strong commitment to not censor personal blog posts, and which provides a robust backup in case someone tries to get my content taken down. Ideally I’d fully self-host as well but I haven’t put in the work. We live in times when even outlining my model of the physical world of a pandemic makes it feel necessary to see the entirety of my technology stack and where pressure might be brought to bear. A few years ago such considerations wouldn’t have occured to me at all.
It has been my belief for a while that if I was posting my full content on Twitter, Facebook or YouTube, that it is likely my content would have been taken down and censored.
It is not a helpful policy to take down information for not lining up with the official government position. That is not how one inspires confidence and trust, or identifies error and discovers truth. There’s also the small matter that the official positions frequently reverse themselves.
If a regular person says vaccines don’t provide immunity, they’ll be censored. If a newspaper writes that we don’t know that vaccines provide true immunity and therefore everyone needs to keep social distancing and mask wearing forever even when everyone involved is fully vaccinated, that’s the mark of a Very Serious Person, and also the recommendation of the CDC.
Shall we follow Zeynep’s example, and violate this policy a bit more by making more true statements about the physical world? Let’s take a closer look at the policy and see what else we can find.
There were a lot of opportunities to violate this policy with true statements, so I’ll limit myself to two (a previous version had a third which I had misread).
The group of young children can potentially be infectious to others, but are effectively immune and cannot die from COVID-19.
The specific treatment of getting vaccinated or having previously been infected then recovering results in immunity.
These aren’t trivial gotchas. Both of these cases seem somewhat important.
The original source of Covid-19 infection was not human-to-human transmission. It is generally believed to have been caused by consumption of an infected bat from a Chinese wet market in Wuhan, but there are alternate theories as well, including the WHO’s would-hilarious-if-it-wasn’t-real new ‘frozen food’ hypothesis, and also those who claim it escaped from a lab (which is another claim explicitly prohibited by Facebook). I’ve chosen to consider such questions of origin out of scope for these posts, but we know for sure it wasn’t human-to-human transmission.
On the flip side, love this exchange:
And also I wonder if they’ll be enforcing this one against a lot of journalists and “experts”:
I’m glad to hear that it’s now official that the vaccines are effective in preventing Covid-19, and I’m confused why Very Serious People keep telling everyone otherwise while also wondering how to get more people to want to take the vaccines. Life is so weird.
It is well-known that I am Against Facebook but these concerns have nothing to do with my existing core objections to Facebook, other than taking ‘Facebook decides what I see and don’t see, is manipulating me for its own profit and I can’t count on seeing anything or having anything I say be seen’ to an entirely new level.
Since I am only on Facebook purely to exchange contact information for communication elsewhere, plus the occasional messenger conversation that can’t be avoided, I don’t have a read on the effective level of enforcement of these rules, or whether such enforcement is being done in a reasonable fashion. I do know that I continuously see reports of suspended or even terminated Facebook and Twitter accounts with no explanation or one that makes little sense, and no practical ability to appeal the decisions.
I do think that being kicked off Facebook will often be a blessing in disguise, but that doesn’t make it right. If you value your Facebook account, please reconsider that position, but meanwhile guard and bite your keyboard’s tongue.
I am thankful that I am in a position where I need not worry about retaliation, for the same reasons I am in a position to devote time to writing this. Most others are not so fortunate.
To anyone who says ‘this isn’t the government doing it so it doesn’t violate free speech’ or ‘there are plenty of other places to go all you have to do to have a platform is Create a Full Alternative Stack (including your alternative to AWS, if not yet an ISP), think about the practical implications of that. Does anything even need to be said?
Zeynep reports that the University of Berkeley has banned outdoor exercise. How do they think that is going to go? Do you think students are going to obey your restrictions in ways that make them take less risk? Do you think they’re going to cooperate with your contact tracing?
Israeli hospital claims it might have a cure for Covid-19, dubbed EXO-CD24. They’re saying it helped 29 out of 30 patients in serious condition. I have no idea how promising things are at this point, but figured I’d pass it along. They are coming out of Phase I and applying to begin Phase II.
Biden has not yet appointed an FDA commissioner. The key qualification should be whoever promises to get the vaccines approved quickly, the tests approved quickly, and also things like filling in the vials of vaccine approved quickly. Once again, I suggest Solomon Mowshowitz, or failing that, Scott Gotlieb is always available.
I mean, none of this is wrong or anything, on the margin this is good advice where available. Now can we have a talk about what it means when there’s ‘no evidence’ of something, and why one might want to act as if that thing was probably true anyway?
As someone who is involved in both Metaculus and the Good judgement project, I think it's worth noting that Zvi's criticism of Metaculus - that points are given just for participating, so that making a community average guess gets you points - applies to Good Judgement Inc's predictions by superforecasters in almost exactly the same way - the superforecasters are paid for a combination of participation and their performance, so that guessing the forecast median earns them money. (GJI does have a payment system for superforecasters which is more complex than this, and I probably am not allowed to talk about - but the central point remains true.)
Metaculus points are not money, so positive points on a question doesn't mean you're a top predictor. However, they aren't meaningless either. It's about winning MORE points than the competition to win on the leaderboards. The incentive system is good for that (though there are some minor issues with variance-increasing strategies or questions with asymmetrical resolution timelines).
The thing that I was more surprised by, looking at the scoring system, is that Metaculus is set up as a platform for maintaining a forecast rather than as a place where you make a forecast at a particular time. (If I'm understanding the scoring correctly.)
Metaculus scores your current forecast at each moment, from the moment you first enter a forecast on the question until the moment the question closes. Where "your current forecast" at each moment is the most recent number that you entered, and the only thing that happens when you enter an updated prediction is that for the rest of the moments (until you update it again) "your current forecast" will be a different number. Every moment gets equal weight regardless of whether you last entered a number just now or three weeks ago (except that the very last moment when the question closes gets extra weight).
So it's not like a literal betting market where you're buying at the current market price at the moment that you make your forecast. If you don't keep updating your forecast, then you-at-that-moment is going up against the future consensus forecast.
So the scoring system rewards the activity of entering more questions, and also the activity of updating your forecasts on each of those questions again and again to keep them up-to-date.
The problem is that metaculus points reward some non-obvious combination of making good predictions and being active on the platform. I only care about the first of those, so the current points system doesn't help me much.
I can't look at a user's points score and figure out how much I should trust their predictions. Or possibly I could, but only by diving into the small print of how scoring works.
I say that as somebody who uses metaculus and believes it has potential. The points system is definitely a weak point
There's no single metric or score that is going to capture everything. Metaculus points as the central platform metric were devised to —as danohu says — reward both participation and accuracy. Both are quite important. It's easy to get a terrific Brier score by cherry-picking questions. (Pick 100 questions that you think have 1% or 99% probability. You'll get a few wrong but your mean Brier score will be ~(few)*0.01. Log score is less susceptible to this). You can also get a fair number of points for just predicting the community prediction — but you won't get that many because as a question's point value increases (which it does with the number of predictions), more and more of the score is relative rather than absolute.
If you want to know how good a predictor is, points are actually pretty useful IMO, because someone who is near the top of the leaderboard is both accurate and highly experienced. Nonetheless more ways of comparing people to each other would be useful. You can look at someone's track record in detail, but we're also planning to roll out a more ways to compare people with each other. None of these will be perfect; there's simply no single number that will tell you everything you might want — why would there be?
Someone who is near the top of the leaderboard is both accurate and highly experienced
I think this unfortunately isn't true right now, and just copying the community prediction would place very highly (I'm guessing if made as soon as the community prediction appeared and updated every day, easily top 3 (edit: top 10)). See my comment [LW · GW] below for more details.
You can look at someone's track record in detail, but we're also planning to roll out a more ways to compare people with each other.
I'm very glad to hear this. I really enjoy Metaculus but my main gripe with it has always been (as others have pointed out) a lack of way to distinguish between quality and quantity. I'm looking forward to a more comprehensive selection of metrics to help with this!
I actually think it's worth tracking: ConsensusBot should be a user, it should always update continuously to the public consensus prediction in its absence, and it shouldn't be counted as a prediction, so we can see what it looks like and how it scores.
And there should be a contest to see if anyone can use a rule that looks only at predictions, and does better than ConsensusBot (e.g. by deciding whose predictions to care about more vs. less, or accounting for systematic bias, etc).
You can also get a fair number of points for just predicting the community prediction — but you won't get that many because as a question's point value increases (which it does with the number of predictions), more and more of the score is relative rather than absolute.
I think this is actually backwards (the value goes up as the question's point value increases), because the relative score is the component responsible for the "positive regardless of resolution" payoffs. Explanation and worked example here: https://blog.rossry.net/metaculus/
You don't care, but if the goal is to motivate better communal predictions, giving people the incentive to do more predicting seems to make far more sense than having it normed to sum to zero, which would mean that in expectation you only gain points when you outperform the community.
This seems to me to be very non-obvious. Do we want more low-quality low-effort predictions, or less high-quality high-effort predictions? Do we want people to go for the exact correct probability as they see it, or give a shove in the direction they feel strongly about? Do we want people to go around making the actual community prediciton to bank free points? Who will free points motivate versus demotivate? What about the question of who to trust, and whether others would update their models based on the predictions of those who are doing well? Etc.
If I have time a post on the subject would be interesting. Curious if there are writings detailing how it works and the reasoning behind it, or if you'd like to talk about it in a video call or LW meetup, or both.
The scoring system incentivizes predicting your true credence, (gory details here).
I think Metaculus rewarding participation is one of the reasons it has participation. Metaculus can discriminate good predictors from bad predictors because it has their track record (I agree this is not the same as discriminating good/bad predictions). This info is incorporated in the Metaculus prediction, which is hidden by default, but you can unlock with on-site fake currency.
There was a lesswrong post about this a while back that I can't find right now, and I wrote a twitter thread on a related topic. I'm not involved with the reasoning behind the structure for GJP or Metaculus, so for both it's an outside perspective. However, I was recently told there is a significant amount of ongoing internal metaculus discussion about the scoring rule, which, I think, isn't nearly as bad as it seemed. (But even if there is a better solution, changing the rule now would have really weird impacts on motivation of current users, which is critical to the overall forecast accuracy, and I'm not sure it's worthwhile for them.)
Given all of that, I'd be happy to chat, or even do a meetup on incentives for metrics and issues generally, but I'm not sure I have time to put together my thoughts more clearly in the next month. But I'd think Ozzie Gooen has even more to usefully say on the topic. (Thinking about it, I'd be really interested in being on or watching a panel discussion of the topic - which would probably make an interesting event.)
Not really. Overall usefulness is really about something like covariance with the overall prediction - are you contributing different ideas and models. That would be very hard to measure, while making the points incentive compatible is not nearly as hard to do.
And how well an individual predictor will do, based on historical evidence, is found in comparing their brier to the metaculus prediction on the same set of questions. This is information which users can see on their own page. But it's not a useful figure unless you're asking about relative performance, which as an outsider interpreting predictions, you shouldn't care about - because you want the aggregated prediction.
So, to "win" I need to participate in every possible market, regardless of my own knowledge, as long as I can make a prediction with positive value regardless of outcome (or at least a hugely favorable spread for going with the consensus?
That sounds like a flaw.
Yes, but it doesn't take much time to just predict the community median when you don't have a clue about a question and don't want to take the time for getting into it. However, as another commenter points out, this means that Metaculus is rewarding a combination of time put in + prediction skills, rather than just prediction skills.
What are you hoping to "win"? This isn't a market - you don't need your relative performance to be better than someone else's to have done well. And giving people points for guessing the community prediction is valuable, since it provides evidence that they don't have marginal information that causes them to believe something different. If people only predict when they are convinced they know significantly more than others, there would be far fewer predictions.
The wording here makes me worry we're Goodharting on quantity of predictions. And the best way to predict the community prediction is to (of course) wait for others to predict first, then match them...
If the user is interested in getting into the top ranks, this strategy won't be anything like enough. And if not, but they want to maximize their score, the scoring system is still incentive compatible - they are better off reporting their true estimate on any given question. And for the worst (but still self-aware) predictors, this should be the metaculus prediction anyways - so they can still come away with a positive number of points, but not many. Anything much worse than that, yes, people could have negative overall scores - which, if they've predicted on a decent number of questions, is pretty strong evidence that they really suck at forecasting.
Looking at my track record, for questions resolved in the last 3 months, evaluated at all times, here's how my log score looks compared to the community:
Binary questions (N=19): me: -.072 vs. community: -.045
Continuous questions (N=20): me: 2.35 vs. community: 2.33
So if anything, I've done a bit worse than the community overall, and am in 5th by virtue of predicting on all questions. It's likely that the predictors significantly in front of me are that far ahead in part due to having predicted on (a) questions that have resolved recently but closed before I was active and (b) a longer portion of the lifespan for questions that were open before I became active.
I discovered that the question set changes when I evaluate at "resolve time" and filter for the past 3 months, not sure why exactly. Numbers at resolve time:
Binary questions (N=102): me: .598 vs. community: .566
Continuous questions (N=92): me: 2.95 vs. community: 2.86
I think this weakens my case substantially, though I still think a bot that just predicts the community as soon as it becomes visible and updates every day would currently be at least top 10.
Anything much worse than that, yes, people could have negative overall scores - which, if they've predicted on a decent number of questions, is pretty strong evidence that they really suck at forecasting
I agree that this should have some effect of being less welcoming to newcomers, but I'm curious to what extent. I have seen plenty of people with worse brier scores than the median continuing to predict on GJO rather than being demoralized and quitting (disclaimer: survivorship bias).
I think that viewing it as a competition to place highly on the leaderboards is misleading, and perhaps even damaging.
I'd think the better framing for metaculus points is that they are like money - you are being paid to predict, on net, and getting more money is better. The fact that the leaderboard has someone with a billion points, because they have been participating for years, is kind-of irrelevant, and misleading.
In fact, I'd like to see metaculus points actually be convertible to money at some point in some form - and yes, this would require a net cost (in dollars) to post a new question, and have the pot of money divided proportionate to the total points gained on the question - with negative points coming out of a users' balance. (And this would do a far better job aligning incentives on questions than the current leaderboard system, since for a leaderboard system, proper scoring rules for points are not actually incentive compatible.)
The fact that the leaderboard has someone with a billion points, because they have been participating for years, is kind-of irrelevant, and misleading.
There are many leaderboards, including ones that only consider questions that opened recently. Or tournaments with a distinct start and end date.
(And this would do a far better job aligning incentives on questions than the current leaderboard system, since for a leaderboard system, proper scoring rules for points are not actually incentive compatible.)
This is true, but you can create leaderboards that minimize the incentive to use variance-increasing strategies (or variance-decreasing ones if you're in the lead). (Basically just include a lot of questions so that variance-increasing strategies will most likely backfire, and then have gradually increasing payouts for better rankings.)
I agree that what you describe sounds ideal, and maybe it makes sense for Metaculists to think of the points in that way. For making it a reality, I worry that it would cost a lot. (And you'd need a solution against the problem that everyone who wants a few extra dollars could create an account to predict the community median on every question to get some fraction of the total prize pool for just that.)
If points could be converted to money enough to motivate real predictions, I would expect a flood of people who do nothing but information cascade to bank points, and it's not obvious what to do about that. As it is, it felt (to me) like there was a tension between 'score points' and 'make good predictions or at least don't make noise predictions' and that felt like a dealbreaker.
I agree that actually offering money would require incentives to avoid, essentially, sybil attacks. But making sure people don't make "noise predictions" isn't a useful goal - those noise predictions don't really affect the overall metaculus prediction much, since it weights past accuracy.
This is an estimated 37% increase in infectiousness. Compared to 50%, that’s much much better. The difference is enough to give us a puncher’s chance of things not being so bad, both buying us time and reducing how bad it is when the time comes.
Unfortunately, this is an incorrect conclusion from the data referenced in the tweet. It seems the 37% number was obtained by dividing 1.07 by 0.78, which rounds to 1.37. However, while 1.07 is the R of the B.1.1.7 variant, the 0.78 is not the R of the other variants, but the overall R (it says so right in the tweet!), which includes B.1.1.7. As B.1.1.7 is a sizable portion of total cases, it already skews overall R upwards quite a bit, and this means that the 37% number is an underestimation.
The latest conclusion from the SSI that I am aware of is, as is also mentioned in the article linked in the post, that B.1.1.7 is 55% more infectious (using the Danish generation time estimate of 4.7 days).
It does look like you are correct. My math from that still had it in the low-mid 40s rather than 55%, but that depends on details. If it's 55%, as I've noted before, that makes it too fast for us to stop in time unless things change fast.
One of the weird things about the "Ministry of Truth" problem is how, somehow, liberals have a perception that other liberals will think that "Facebook should ban 'disinformation'/'pseudoscience'" is the responsible position. How does that happen? I don't think we were ever explicitly told by somebody. Nobody made a cogent, persuasive argument in favor of that position.
If I had to try and trace it back, I think it went a little something like this:
People talk a lot about how much stupid stressful crap is on Facebook.
Mark Zuckerberg/Facebook gets a crappy public image, not for producing the crap, but for profiting off the crap.
People blame MZ/Facebook for a lot of social ills: pseudoscience, Trumpism, depression/anxiety/teen suicide, privacy concerns, hate speech, etc. They're in everybody's bad books. Yet because there are so many competing notions for what FB ought to do to correct the problem, they don't have any obvious fix.
Facebook doesn't want to be in the bad books. They strategize about how to project the image of a responsible, upstanding company. For a while, their position is "we'd be irresponsible to police 'disinformation,'" which is based on the fear that people would in fact call them irresponsible if they did. Asymmetric justice applies: Facebook is safer doing nothing (or less things) than doing something (or more things).
COVID-19 changed the game, giving a much more unified concept for what "responsible social media behavior" would look like. Somebody else, the freakin' CDC, makes the policy for what constitutes improper speech. Facebook just enforces it. It's a time of crisis, when public health concerns take priority over privacy/free speech/democratic concerns. Facebook can look like a noble guardian of public health. The few remaining "free speech" advocates and uncredentialed contrarian science enthusiasts can not only shove it, but get framed (by a gigantic corporation!) as the new villains in the story.
It'll be important for people who are worried about being cast inappropriately as a villain to look for people who might have been deemed as such manage to alter their perception. How does Zeynep Tufekci manage to get in and stay in the good books? How do we raise controversial but important issues with our families and friends when they're so much less fluent in the issues than we are that we risk looking unbelievably overconfident just for talking about what we know?
I think these things are possible, and I think it'll help if we focus more of our attention on tractable solutions to them than on overemphasizing the current climate. After all, politics isn't the mind killer. Fear is the mind killer. How do we let some of this fearful stuff pass over us and through us, and yet remain?
IMO 2020 wasn't a turning point, and Facebook is not special. The events that happend lately have been a predictable development in a steadily escalating trend toward censorship. I'll note that these censorship policies are widespread across every social media platform, and infact extend well beyond social media and apply to the entire infrastructure stack. Everything from DDoS protection services, to cloud service providers, to payment processors have all been getting more bold over the course of several years about pulling plugs on people saying the wrong things or providing platforms for others to say the wrong things. Here's how I think it went down:
1.From 2010-2020 Social media and other SV companies gained a tremendous amount of power by gaining control over social media networks. 2. By virtue of all being near each other, they formed a political monoculture/ingroup. 3. They found themselves capable of deplatforming anyone they disagreed with. 4. They started banning people, starting with the most deplorable and the outgroup and working their way up from there. This seems to have become especially noticeable sometime around 2015. 5. First the deplorables complained about this by making appeals to free speech, which made free speech low status. 5. Then the outgroup complained about this and made appeals to free speech, which made supporting free speech an outgroup identifier. 7. Everyone falls in line because otherwise they might get unpersoned if they're mistaken as a member of the outgroup or the deplorables by defending free speech. 8. The overton window of acceptable speech continues to shrink as opinions on the ever changing fringe continue to get silenced in a process that's not too different from the evaporative cooling of group beliefs.
I expect that the trend towards more censorship will continue unabated, especially on public social media platforms.
I personally have a very tough time fitting your interpretation into my model of the world. To me the popularity and actions of Facebook et al. are mostly disconnected from our ability to communicate with family and close friends.
In my opinion the timeline seems to be a little more as follows:
People are on Facebook and Twitter and other social media platforms both to stay in touch with friends and to complain about the outgroup.
COVID-19 hit, significantly reducing quality of life everywhere. People realign their political discussions and notions of outgroup along COVID-lines - are you a believer in lockdowns and masks and science or the opposite? This temporarily supersedes other political discussions, not because people have wonderfully unique and insightful opinions on COVID countermeasures but because this is the biggest event happening and as such is necessarily political.
After approximately one year of lockdowns and countermeasures people have sunk significant parts of their public profile into their thoughts regarding COVID. A large portion of the public, as well as officials, will support silencing opposition if only to retain a coherent public image (after all, if communication on COVID is not more important than free speech, what have you been doing all these months?).
Facebook rises to the occasion and offers to selflessly censor people according to criteria set by the WHO.
I'd like to couple this with a prediction that Facebook will not start censoring older messaged by the WHO and other Respected Officials. I see Facebook's cooperation more as a power grab with plausible deniability than a desire for certain messages (officially endorsed) over others (crackpot/other). It only exists through the support of the very serious people, so it is counterproductive to start challenging them on their own history.
Lastly I think that if you genuinely want to have a heart-to-heart with your friends and family it is silly to restrict yourself to communicating via Facebook. Call them, start a blog, meet somewhere outside for a walk if you want. This has the twin benefit of you not having to worry about issues being 'controversial' as defined by Facebook, and them not having to publicly change their thoughts over your message. Also it is much less embarrassing if it turns out you were unbelievably overconfident all along.
President Biden says that the U.S. will have enough coronavirus vaccine to inoculate 300 million Americans by this summer. Biden says Moderna and Pfizer will deliver the doses by the end of July, more than a month earlier than initially anticipated.
I also notice that there is a large part of me that thinks, once it’s easily and widely available, you know what? Straight up, just f*** ‘em if they don’t want the vaccine.
This is how I was planning to act at that point, and basically as soon as I'm able to get an official vaccine. Once it's readily available I'll feel no guilt about continued cases (assuming no major vaccine escape, that would be a different story). Even once I've gotten the official vaccine, I'll want to propagate the norm that vaccinated people should live their lives as if they were, you know, vaccinated, so I intend to act that way, unless there's a reason I'm not considering.
The loss of life and health of innocent people who got suckered into a political issue without considering the ramifications?
I mean, the group of people who holds out on getting a vaccine as long as possible will definitely be harder to convince than the average citizen. But with these numbers (death rate, long term health conditions, effectiveness of vaccines) around are you seriously suggesting trying to help them is not cost-effective? From the post I think you're talking about tens of millions of people in the USA alone, if not 100M+.
The loss of life and health of innocent people who got suckered into a political issue without considering the ramifications?
By now, everyone has had a year to consider the ramifications of their decisions. People are free to make their own choices about the vaccine and their response to covid in general. If they make their choices based on their political affiliation or in-group signaling, so be it.
But with these numbers (death rate, long term health conditions, effectiveness of vaccines) around are you seriously suggesting trying to help them is not cost-effective?
I am seriously suggesting it is not cost-effective for me to try to influence others to get the vaccine. Most of the people I know have either already decided to get the vaccine at their first opportunity, or decided they will never get it. In November/December, as the vaccines were starting to get approved, I had some discussions with my few friends who I thought might be on the fence, but they weren't moved much by my arguments. I don't actually think I know anyone that I could convince at this point.
What has scaling up involved, practically speaking?
We committed ourselves to Covid-19 in March. We took a huge risk, because nobody knew then that any vaccine was going to work. Of the $800m (£579m) we needed, we put in $270m and the rest we raised from the Gates Foundation and various countries. We dedicated about 1,000 employees to the programme and deferred all product launches planned for 2020 for two to three years, so that we could requisition the facilities allocated to them. Then it was a question of equipping those facilities and getting them validated for use, which we did in record time. By August we were manufacturing and stockpiling a vaccine that we predicted, correctly, would be approved around December.
If the vaccines need to be adjusted to protect against future emerging variants, how much of a challenge will that be?
It would be simple now that the processes are up and running. We grow the virus in living cells, so we would simply change the master clone – the virus with which we infect those cells – and that then propagates through them. It would take us two to three months to start producing the new vaccine at capacity.
Reiterating a Zvi point:
What I find really disappointing, what has added a few months to vaccine delivery – not just ours – is the lack of global regulatory harmonisation. Over the last seven months, while I’ve been busy making vaccines, what have the US, UK and European regulators been doing? How hard would it have been to get together with the World Health Organization and agree that if a vaccine is approved in the half-dozen or so major manufacturing countries, it is approved to send anywhere on the planet?
Instead we have a patchwork of approvals and I have 70m doses that I can’t ship because they have been purchased but not approved. They have a shelf life of six months; these expire in April.
I wonder how much long term damage this is going to do to people’s health by demotivating healthy lifestyle choices,
Afaik building muscles is a good way to gain weight.
Now it would be fun to make vaccines that contain the mark of the beast, and 5g masks.
We face increasing levels of Komogorov Complexity, and are told to say that thunder proceeds lightning. More and more true facts are only stated privately in whispers, or implicitly and esoterically within other statements.
Huh? Ctrl-f for "complex" only shows up in the comments. The title of the linked SSC post is "Kolmogorov Complicity And The Parable Of Lightning". It's not really related to Kolmogorov Complexity, except for a) being the same Kolmogorov, and b) being a pun on it.
Ah, you were talking about this article. Me and Daniel were saying that "Kolmogorov Complexity" never shows up in the linked ssc article (thinking that Zvi accidentally wrote "Kolmogorov Complexity" when he meant "Kolmogorov Complicity").
The UK vaccine rollout is considered a success, and by the standards of other results, it is indeed a success. This interview explains how they did it, which was essentially ‘make deals with companies and pay them money in exchange for doses of vaccines.’
A piece of this story you may find interesting (as an example of a government minister making a decision based on object level physical considerations): multiple reports say Matt Hancock, the UK's health Secretary, made the decision to insist on over-ordering vaccines because he saw the movie Contagion and was shocked into viscerally realising how important a speedy rollout was.
We remove false claims about how and where COVID-19 can be transmitted and who can be infected. This includes: [..] Claims that COVID-19 can be transmitted from anything other than human-to-human transmission
Interestingly, the Danish version of the Facebook policy page does not include this clause (only forbidding claims about transmission by mosquitoes).
If, however, you previously had accepted that the English strain was more infectious, and the question was how much more infectious, then news of the answer could be good or bad. In this case, it’s good.
This is an estimated 37% increase in infectiousness.
The question of how much more infectious B.1.1.7 is is pretty useless without also referencing a generation time estimate. Different agencies/countries use different values for that, so the numbers for the relative R number R_B.1.1.7 / R_old they give are not directly comparable. I expanded on this in a comment a while ago [LW(p) · GW(p)].
In the meantime, the Danish SSI also published a report in which they also stress that the numbers of how much more infectious B.1.1.7 is can't be compared across countries due to in particular different generation times being used. This report is from January 21., and in it they estimate relative R to be 1.36 as of January 14.
A newer report from February 3. mentions that SSI now estimates a relative R for B.1.1.7 of 1.55.
The SSI uses a generation time of 4.7 days, the English PHE uses a generation time of 6.57 days.
The quoted conclusion of 37% increase in infectiousness from the original post is unfortunately a mistake, see my comment here [LW(p) · GW(p)].
The question of how much more infectious B.1.1.7 is is pretty useless without also referencing a generation time estimate.
Generally true, but in using contact tracing data the English analysis is answering the "how much more infectious" question directly rather than relying on inferring from relative growth rates and estimated generation times.
The 37% error does revise my estimate a bit for how confident I should be that it is <50% (although even correcting that it is probably still under 50% according to Zvi) but I still expect it to end up that side of the equation. If I was answering that survey now I'd be at 20% or so.
I've covered this extensively in previous weeks. If they are claiming that they could not have spent more money to accelerate production, then (1) that means they spent too much money almost by definition and (2) I do not believe them.
I have one correction on the obesity/overweight numbers, unless I misunderstood the claim being made. In most contexts, including the NCHS numbers cited above, the cutoff for overweight is a BMI of 25, not 30. The cutoff for the vaccine is a BMI of 30, so only ~40% of people qualify, not ~70%.
I swear I checked this multiple times before and saw people explicitly say it was 30 that was the second threshold and that 35 was severe obesity, but now I look and it seems you're right. So still outrageous, but not quite as completely patently absurd.
I'd be interested in seeing a write up on whether people who've had COVID need to be vaccinated. I have a friend who was sick with COVID symptoms for 3 weeks and tested positive for SARS-CoV-2 shortly after the onset of symptoms. He is now being told by medical professionals that he needs to be vaccinated just the same as everyone else. I tried to look up the data on this. Sources like CDC, Cleavland Clinic, and Mayo Clinic all state that people need to be vaccinated even if they have had COVID. However, their messaging seems to be contradictory. There are many appeals to "we don't know". The reasoning doesn't appear to be any more complex than "vaccine good" and "immunity from infection 'not known'". There is no discussion of things I would expect like, the difference between testing positive with no symptoms, had symptoms but never tested, or tested positive and never had symptoms. While I can imagine reasons why immunity induced from the vaccine and from infection would be different, my prior is that most of the effects are going to be the same. There is repeated reference to not knowing how long immunity developed from infection lasts, but by definition, we have had less time to see how long immunity from the vaccine lasts. So our evidence about the vaccine would be weaker. I could say a lot more, but I'll leave it at that.
To avoid any confusion: My actual model is that if you've had COVID19 then the vaccine would act as a booster. So I'd say people who've had it should get vaccinated eventually but should be among the lowest priority. That should be modulated by the probability that you had COVID and the fact that asymptotic COVID may be less likely to develop immunity. On the other hand, having had asymptomatic COVID is probably evidence that you will be asymptotic if you get it again. That is not the message that is being given to the public.
I believe the quote in the Janelle Nanos tweet (after "Meanwhile, in Boston, priorities are straight:") was taken out of context here. The full article shows how Dr. Ivers was trying to point out the inefficiency of the state's rigid system and offer improvements:
For weeks, Dr. Louise Ivers has been advocating for Massachusetts to speed-up the pace of its COVID-19 vaccinations. But it’s not just the slowness of the rollout that is causing the Boston doctor consternation when it comes to the state’s vaccine push.
The executive director of Massachusetts General Hospital Global Health and interim head of MGH’s Division of Infectious Diseases told Boston.com that while she’s been disappointed by the state’s vaccine efforts, she isn’t completely surprised by the sluggish and fragmented rollout based on the response to the virus over the last year.
Ivers told Boston.com she believes that if the pace of the vaccine were ramped up with more flexibility to start new phases as others plateau, that some of the issues around equity that the state has seen would “settle a little more carefully.”
“It’s quite complicated — you spend a lot of operational resources and planning and logistics to make sure that you only vaccinate 75-year-olds,” Ivers said. “There’s a lot of time and energy spent on making sure that a 74-year-old doesn’t accidentally get vaccinated.”
Instead, Ivers said the state should be moving more quickly to expand vaccine access to those 65 years old and up, as well as groups with comorbidities.
Yes, I think we are all in agreement on the topic. On my first reading, seeing the isolated quote between the other two examples of poor vaccine responses made me think this was another example of a poor response, and the quote itself can be interpreted that way if read alone (i.e. We think only vaccinating 75-year-olds is the correct policy, and it's hard but necessary work to enforce it).
I’m not saying that major governments keep doing this without any good reason, because I’ve already documented that. There’s the trivial case that everyone thinks it’s not ‘safe’ to give the vaccine to children yet. Many also have legitimate concerns about people who are old and in sufficiently poor health, and governments keep imposing various age rules on eligibility citing safety concerns.
Minor point: I think you might've misread this one? The Facebook policy would allow a claim like "it's not safe to give the vaccine to children" because it does identify a group based on personal health, age, or disabilities.
I do have limits to how far I'd go. The impression I have in my head is that the two ways to jump in line are 1) malfunctions where they have to give them away or throw them away 2) areas where the demand is so low that they're having to choose fairly lax rules or throw them away. My hope is that there's a way to get a vaccine that wouldn't have gone to anyone truly in need without having to do anything particularly illegal
I'm not familiar with Metaculus scoring in detail, but it is POSSIBLE that they are trying to simulate a subsidized rather than an unsubsidized prediction market in order to encourage betting on a wide range of issues rather than just picking off the handful of positions that look most egregiously mispriced.
Update: Metaculus published a document explaining their scoring rule, which seems to say something similar. The key quote is probably this:
Our positive sum Scoring Rule — in which for a given question the total points awarded across users is generally positive — encourages broader participation on more questions by not penalizing forecasters who participate in more questions