Happy Thanksgiving! 

Even in a year like this one, we have much to be thankful for. Most of us, even if we get sick, will fully recover. Treatments continue to improve. Given expectations, this week’s numbers were surprisingly good, to the extent that I have theories later to explain why. The market is at all time highs. The center held. We are at peace. Vaccines are on their way soon and look to be super effective. Life goes on. Life finds a way. A year from now, we likely will be able to give thanks for no longer living in quite such interesting times. 

As bad as many things have been, they could have been far worse.

Short term, even though positive test rates on average declined for now, things remain rather grim. The holiday is about to hit us, followed a month later by Christmas. Positive percentages are down, which is great, but there are reasons to suspect that is somewhat of a mirage this week. Raw positive test counts are still up, and deaths are doing exactly what one would expect from a lagging indicator, giving us the most since early May.

Still, it is good to give thanks. Gratitude is hugely underutilized. I will conclude this post by giving thanks. 

First, let’s run the numbers.

The Numbers


Last week’s prediction was a 13.4% positive rate on 9.5 million tests, versus 12.3% positive rate on 8.9 million tests last week with 1,209 deaths per day. Instead, we got pleasantly surprised, and got an 11.6% positive rate on 10.4 million tests, but 1,644 deaths per day. I did not predict deaths.

Median prediction for this next week is weird, but I’ll say with wider than usual error bars 12.8% positive rate on 11 million tests, and a rolling average of 2,100 deaths. 


Sep 24-Sep 309349902619360
Oct 1-Oct 779711032308400
Oct 8-Oct 1478212172366436
Oct 15-Oct 2180415912370523
Oct 22-Oct 2889517012208612
Oct 29-Nov 495619772309613
Nov 5-Nov 11108927122535870
Nov 12-Nov 181255293428181127
Nov 19-Nov 251761416933961714

This is slightly worse than I expected, especially in the West and Northeast. If you look at the 14-day and 21-day lagged CFR, week over week, we do see small increases, but that is to be expected as well, as positive test percentages are rising and thus we are missing a larger percentage of cases, and also as some hospitals are under increased strain. The best we can reasonably hope for is to avoid further increases in those ratios, and for case counts to stabilize. Even if that happens, we are still in for multiple weeks of large increases in death counts across the board. 

If we avoid breaking 2,000 deaths per day for the next week, that would be sufficiently surprising that I would expect it to be due to a lack of reporting the week of Thanksgiving rather than there not being that many deaths. My guess is we will be caught up on reporting by Wednesday, and thus be substantially above the 2,000 threshold. But while case counts have large error bars, in the short term death counts mostly don’t. Things will get worse before they get better. 

Positive Tests

Sep 24-Sep 30554969293210630027214
Oct 1-Oct 7567429724311017034042
Oct 8-Oct 146828412574411799538918
Oct 15-Oct 217557114985113323843325
Oct 22-Oct 289498318188115812357420
Oct 29-Nov 411268425291716709870166
Nov 5-Nov 11157495387071206380108581
Nov 12-Nov 18211222452265255637150724
Nov 12-Nov 18269230433294294230170595

Test counts are up a lot, so the raw positive counts make things look somewhat worse this week than they were.

Is it plausible that the Midwest would have peaked this past week if we didn’t have Thanksgiving coming up? I think it is definitely plausible. A lot of people have already been infected recently, in a way that both builds substantial immunity and gives others cause to adjust behavior. The hospitals are overflowing in many areas and deaths are piling up, which causes the important adjustments, both private and from government action. As I’ve noted many times, the government restrictions are mostly not genuine attempts to solve the pandemic, but they do have some impact. In the case of the Midwest, this includes some new mask mandates, which as discussed this week is the most effective of the options that actually get implemented around these parts.

Positive Test Percentages

9/17 to 9/232.20%5.96%7.13%4.11%
9/24 to 9/302.60%6.17%6.18%4.27%
10/1 to 10/72.61%6.05%6.74%4.23%
10/8 to 10/142.57%8.14%7.09%4.75%
10/15 to 10/222.95%8.70%7.85%5.36%
10/22 to 10/283.68%9.87%8.58%6.46%
10/29 to 11/44.28%12.79%8.86%7.04%
11/5 to 11/115.56%17.51%9.89%8.31%
11/12 to 11/186.99%18.90%11.64%10.66%
11/19 to 11/257.00%16.62%10.41%11.75%

The improvement in the Midwest is even clearer here, and we also see the South headed in the right direction. The more I look at these numbers, the more I think this past week must have been heavily influenced by pre-Thanksgiving dynamics. This is too big a turn around to mostly be a permanent reversal.

Test Counts

DateUSA testsPositive %NY testsPositive %Cumulative Positives
Sep 17-Sep 235,737,9195.2%610,8020.9%2.09%
Sep 24-Sep 305,833,7575.1%618,3781.1%2.18%
Oct 1-Oct 76,009,8455.2%763,9351.3%2.28%
Oct 8-Oct 146,322,8655.7%850,2231.1%2.39%
Oct 15-Oct 216,439,7816.5%865,8901.2%2.52%
Oct 22-Oct 286,933,1567.5%890,1851.4%2.67%
Oct 29-Nov 47,245,6008.6%973,7771.6%2.86%
Nov 5-Nov 118,285,49510.6%1,059,5592.4%3.13%
Nov 12-Nov 188,924,33812.3%1,155,6702.9%3.47%
Nov 19-Nov 2510,422,94111.6%1,373,7512.9%3.83%

Big expansion in testing capacity. That likely was at least partly caused by a wave of people wanting tests in advance of Thanksgiving.

Machine Learning Project

The project does not seem to have been updated. Its numbers still only go out to November 10. I’ll keep checking, but will skip this section entirely in future weeks if there is no update available.


I put America on the graph to show the contrast. Europe has imposed sufficient restrictions slash adjusted behavior sufficiently, and infections are now on the decline in most places. An important exception is Germany. While Germany is still doing better than other places, it seems they only imposed at most enough rules to stabilize the situation. Their positive test percentages strongly suggest that they have not even done that. If trends continue, soon they will have a bigger problem than other European countries.

Some Speculations about Thanksgiving Dynamics

I have a few theories about the dynamics surrounding Thanksgiving, and why we are seeing the numbers we see this week. None of this was something I’d thought through before this week, which should make you a little suspicious this is fitting theories to a graph, but I do think the theories make sense.

One theory I have is that the social dynamics of family gatherings are causing some holiday gatherings to go forward despite many or all people involved realizing it should be cancelled. 

The Onion lays out the case. If the elderly cancel, they worry they are being selfish and rude, or that they are implying that the young are not staying safe. If the young cancel, they worry that they are implying they wanted to cancel all along and never cared all that much, or that they don’t care enough to stay safe. Worrying about the safety of others is highly suspicious on multiple fronts. 

Note that no one has to actually think any of those things. They only have to worry that others might think those things. 

Even that isn’t necessary. One might reasonably be worried that others might see them as doing something that could cause others to think that they did not care, and thus judge that they did not care about whether others thought they cared, which would itself be quite blameworthy in many cases. And so on.

There is also the worry that if one cancels, that one is imposing a decision on others, especially if one is the host. Cancellation imposes costs, potentially high costs, in addition to bad feelings. 

Plus if you break your tradition, especially voluntarily, that weakens the tradition for the future.

This suggests that outright banning gatherings can be a win for all concerned. Choices are bad. Forcing the choice most people want can be a big win. Sorry, mom, I can’t go. It’s illegal. A lot better than ‘they strongly suggested.’ What even is that?

Another theory is that the holiday is driving positive tests down temporarily, and that this means what we saw this week was plausibly a false peak even ignoring the spike we likely get from Thanksgiving itself. 

A lot of people are about to go see Grandma. When you are about to see Grandma, several things happen. 

Hopefully you are going to spend the prior week being relatively careful. It seems like the least you could do. 

Then there is a chance you get tested for Covid-19 even without any symptoms. Many states require this before interstate travel. Even if it isn’t required, you still have a good reason to do it.

We don’t need this as an explanation for long lines and long waits, but it seems likely to have been a contributing factor. With a lot of people both being careful and seeking out Covid tests without symptoms, that’s going to drag down the positive test rates. If tests available is also a limiting factor, which seems likely given long waits and linear increases in tests over time, this will drive down positive test counts as well. 

Then after Thanksgiving we should expect the opposite. The gatherings will presumably create a lot of new infections, as will travel to and from the gatherings. On top of that, with no pending exposure to anyone vulnerable, and having been reminded of the joys of face to face contact, a lot of people will no longer have as much reason to play safe. 

Then we do it all over again for Christmas.

This suggests both that this week’s numbers were artificially low, and then next week’s will not only fully reverse this but be unusually high. We are going to enter a period where it is exceedingly difficult to know how things are developing. 

Can We All Agree Andrew Cuomo Is The Worst?

Last week we highlighted several stories of prominent politicians telling everyone to cancel their Thanksgiving plans and otherwise stay safe, while simultaneously flouting the rules themselves. 

Those examples had two flaws. They did not involve Andrew Cuomo, and they were insufficiently on the nose. 

Good news is that now we can fix both of these problems at once. It seems the following events took place on Monday, November 23:

First, he tells the story of how his daughter in Chicago is not coming to his Thanksgiving dinner. While telling this story, he cries on live television. 

Several hours later, Cuomo reveals that his 89-year-old mother and his two other daughters are traveling for the holiday, and will be joining his Thanksgiving, which is fine because they got Covid-19 tests.

Several hours of outrage after that, Cuomo claims he is cancelling on his family. 

I am, once again, not here to tell private citizens that they are being irresponsible, or making the wrong choice, if they value the holidays with their families enough to take the risks of becoming infected. I worry that many are unaware of how bad things have gotten, or how risky it is to have a traditional Thanksgiving dinner, or that social dynamics make such gatherings hard to avoid even when people would prefer to cancel them. Eyes should be open. But if in the end people decide to live their lives, then I respect that decision.

Cuomo’s situation is dramatically different than that. Cuomo is setting an example for 19 million New Yorkers and 330 million Americans. When others see that the person lecturing them on their irresponsible behavior, telling them to skip the holiday with their family and crying about not seeing one of his daughters, is instead flagrantly doing exactly what he is telling everyone else to avoid, that’s not a decision that primarily impacts him and his family. That’s a decision that sends a clear message to the whole country and the world that the rules aren’t real. That they don’t apply to people who count, the same way they didn’t apply to the protests or the election victory celebrations. 

Here is a post summing up many similar examples, most of which this column hasn’t covered because they are from other countries.

Here’s one that’s just in: The Mayor of Denver. The tone deafness here is astounding.

We need not wonder why the people decline to follow such rules, or such leaders.

A lot of people posted this week that we should be alarmed by full airports, because half as many people as usual travelled in advance of Thanksgiving. One could instead view this glass as half full. Only half as many people travelled in advance of Thanksgiving. 

Similarly, the partisan divide is real, but not as big as one might have expected (source is NYTimes via Twitter so no link):

Given everything, it could have been much worse. 

After all, mom has a plan.

And if you look at the details, the obsession with surfaces is unfortunate, but that’s totally not mom’s fault. If we ignore that last line, otherwise the plan mostly seems at least as good as the CDC’s. Temperature checks, good ventilation, putting the vulnerable in a different room. Long tables to increase distance. Focus on avoiding face to face close contacts. If this was par for the course I wouldn’t be that worried. All things considered, not bad! 

Supreme Court Throws Out Religious Restrictions in New York

Amy Barett has joined the other non-Roberts conservatives to form a new majority on the Supreme Court, and now she has made her impact felt for the first time. In a 5-4 ruling that is a reversal of similar previous 5-4 rulings, the court held that Andrew Cuomo (again, the worst) imposed Covid-19 restrictions that likely violated the First Amendment’s restrictions against freedom of religion, because rules against houses of worship were harsher than rules against other businesses. Thus, they provide injunctive relief that prevents Cuomo from closing down houses of worship or imposing restrictions on the number of people per location. 

Restrictions on percent of capacity seem to still be allowed, provided they are not based on rapidly moving colored zones, and the plaintiffs said they were complying with such rules already, although some other places are most definitely not doing so. 

I am not a lawyer, so I may have the implications wrong here, but it’s clear that Cuomo’s hands are being tied. 

I am of two minds about this.

On the one hand, severely restricting houses of worship is physically necessary to halt the spread of the pandemic, in a way that halting many other things is not. People gather, by default quite tightly packed because that enhances the experience in some ways, and in most services they spend much time talking, chanting and singing. That’s asking for an outbreak.

The counterargument that ‘well, our locations have operated at reduced capacity for a while without an outbreak while infection rates in New York were low’ is not a compelling statistical response (and also leaves out the important words ‘that we know about and that we have been forced to admit’), and the court citing this shows they do not understand how risk works. Which is unsurprising. If an action has a low probability of a damaging event each time, and so far it has been fine, that does not make it harmless. Would the court be fine with someone guessing your bank password so long as they have not yet succeeded? 

As in a previous unrelated case I saw last month, Gorsuch has written his own concurrent opinion whose central arguments seem to show key factual misunderstanding. In the previous case, he was wrong about legal procedures in a state and used it as a precedent. In this case, he has no idea why a wine or bicycle shop might not have the same risks as a house of worship. 

Attending religious services is important – some would say essential – to many people, but as commonly practiced it is also unusually dangerous. 

On the other hand, freedom of religion is important, the precedent of letting churches and synagogues be shut down leads to places that should scare us, Cuomo is the actual worst, De Blasio is right behind him, and they were really asking for this to happen.

Time and again in press conferences, both Cuomo and De Blasio singled out houses of worship to go after, and especially singled out Jews. I don’t know how much history you know, but let’s say that when the government starts singling out Jews as shameful people who are bringing down ruin on us all and whose religious practices must be stopped in its official pronouncements, and tells synagogues they have to close or else, that is generally not a great sign. It also does not help one’s legal argument.

The restrictions really were rather over the top compared to other locations, limiting thousand-person locations to ten people at times when New York was a relatively safe zone. And to what end? It is not like we were going for suppression, nor could we have done so given travel between states.

More importantly, well, there’s this:

In a red zone, while a synagogue or church may not admit more than 10 persons, businesses categorized as “essential” may admit as many people as they wish. And the list of “essential” businesses includes things such as acupuncture facilities, camp grounds, garages, as well as many whose services are not limited to those that can be regarded as essential, such as all plants manufacturing chemicals and microelectronics and all transportation facilities. See New York State, Empire State Development, Guidance for Determining Whether a Business Enterprise is Subject to a Workforce Reduction Under Recent Executive Orders, https://esd.ny.gov/guidance-executive-order-2026. The disparate treatment is even more striking in an orange zone. While attendance at houses of worship is limited to 25 persons, even non-essential businesses may decide for themselves how many persons to admit.

If you tell me you are not imposing a restriction on acupuncture or on camp grounds not because they are safe, but because they are “essential” then you make a mockery of the term essential. You are taking the power to shut down most of life, and applying that power based on political expediency and what your tribe wants to signal it approves. If you establish you can do that, what protects freedom of religion? What protects freedom of speech?

Those freedoms are at the heart of a free society, and they are under increasingly severe threat.  There are damn good reasons that such restrictions must pass “strict scrutiny,” and Cuomo’s actions here do not pass strict scrutiny. The court did not frame this as a reversal of its previous rulings allowing restrictions in other places to stay in place. It merely said that New York’s particular rules were over the top, and didn’t stand up to strict scrutiny. That seems at least plausible to me.

One Night Stand

The governor of Pennsylvania announces that for one night and one night only, on Wednesday the 25th, bars and restaurants cannot serve liquor after 5pm. 

This makes very little sense. Why would you spend this much of our attention and ability to absorb rules on something that only lasts for one evening? That transparently is going to have very little impact on anything? Josh Jordan figures it out. Being nonsense is the point. Ruining people’s night if they are travelling for Thanksgiving, preventing people from enjoying their holiday, won’t help much, but it will be noticed as someone who wants to act as a Very Serious Person doing a thing that punishes people doing bad things, and therefore is an excellent Sacrifice to the Gods. Because there is no story in which this is a good idea, there can be no confusion as to the goals here. One wouldn’t want anyone to be fooled into thinking you were enacting policies in the hopes that those policies might work. If you did that, you might support policies that could possibly work over policies that definitely won’t work in other areas. That doesn’t sound like a very loyal or trustworthy ally at all.

I do want to acknowledge that this doesn’t quite make zero sense. The night before Thanksgiving is one of the biggest drinking nights of the year. Making travelling for Thanksgiving worse potentially could lead to less travel (although by the time this was announced, it was far too late to change plans). Infections on Thanksgiving weekend are likely to overwhelm testing capacity and potentially hospitals, and there’s a good chance they are the peak of the pandemic, so limiting infections at exactly that time is high leverage. All of that makes the play so ingenious. There’s a story you can tell about how you’re actually modeling the world, while the magnitudes involved make it obvious that you don’t care about such things and your impact on the world will be negligible. You can have your turkey and be one too.

Do We Have Tools That Might Work?

Yes. We have several tools that would definitely work. Alas, we have banned those tools, in part because they work. But what about the tools we have not banned? Do any of those work?

Study looks at the effectiveness of various preventative government measures. They find that mask mandates are the best tool available in terms of cost-benefit, which makes sense, and that it is also useful to limit gatherings to ten people and to close restaurants and gyms. They also find data that confirms my suspicion that it is actively harmful to impose bad restrictions, such as closing low-risk retail like bookstores, presumably due to the substitution effects driving other risky behavior. They also notice that restricting gatherings to 100 people backfired, and speculate (I think correctly) that this is because it sounds like permission to gather 99 people. Half measures are bad. 

The big surprise, which also surprised researchers, was that it was effective to close parks and beaches. On its face, this doesn’t make a lot of sense. One theory they offer is that this is driven by beach parties and park gatherings, and even outdoors a few giant events dominate. I like their speculation that the pattern might be driven by closures of national parks, as it suggests another model. If you close the parks and beaches, you drive people away entirely. Those people were going to be moving around and doing things, and taking other risks. When you close a bookstore, this theory goes, you screw with the locals. Close the beach, and you screw with people everywhere, who now stay away, or even drive locals to leave because they were there for the beaches. And the people you drive away are the ones you want to drive away.

I’d also worry a lot about decisions on what to close being both very non-random in their timing (they’re caused by anticipation of future cases) and correlated to lots of other things as well. If you impose strong restrictions you’re likely doing many other similar things and your area likely has many other different characteristics, beyond the current infection levels. 

It is also possible that closing beaches early on was effective in sending a message about how bad things were, and in setting an example to get others to take less risk elsewhere. Every news source seemed to converge on ‘show picture of a beach’ when they needed to show risky behavior. If you don’t then shut that down, you’re sending the message you aren’t serious. Shut it down, and people get it, because you’re ruining everyone’s day. 

Sacrifices to the Gods do have that benefit. They’re not all bad, because they teach everyone who the Gods are these days. Choose wisely, and good things can happen. At a minimum, you can keep worse Gods at bay. When the beaches were actually closed in practice, it was at a time when getting the “no really there’s a pandemic” message out there was important.

Thus, all this stuff is very hard to disentangle. I still think that closing parks or beaches going forward would be a bonkers move.

All I Want For Christmas Are Covid-19 Vaccine and a PS5, But They Mispriced Them and Now They’re All Sold Out

There is also the small matter that the Covid-19 vaccine continues to not have been approved for emergency use. They applied for emergency use on Friday, November 20. The F.D.A. will hold a meeting to discuss approval on December 8-10. Why not sooner, exactly? We asked America’s finest news source.

I think the explanation in the link is essentially correct. The FDA has rules and procedures and schedules, and damned if they are going to change them for the pandemic. Now that there is an EUA application in, they will begin looking at the data submitted and discuss at the next meeting. 

Contrast that with what we would have done if we actually wanted to stop the pandemic and save lives as rapidly as possible subject to safety and efficacy concerns.

Pharmaceutical Company Runs Experiments

We now have three working vaccines! Ah, ah. I love it! 

Well, probably.

The Oxford/Astrazeneca vaccine was tested with two full doses, and also tested with a half dose followed by a full dose. Alas, these were done in different locations, that otherwise followed different procedures, so we can’t be confident which dosing strategy was superior, but at least we tried multiple different strategies. The initial half-dose so far has been 90% effective, whereas the double full dose was “only” 62% effective. Which would still be well above the threshold we were looking for a month ago, but at this point would be rather disappointing. Their procedure for detecting infections was more sensitive than those of Pfizer and Moderna, which might help explain the difference in measured effectiveness. AstraZeneca claims that there were no serious illnesses in the vaccine groups. 

Getting to run two different dosing regimens here was potentially a big deal. The half-dose strategy superficially looks more effective, although we don’t have enough data to know for sure even if they weren’t coming from different locations. Even if it’s not more effective, so long as it’s equally effective, using 25% smaller combined doses lets us inoculate more people faster and cheaper. 

How did they manage to do such a thing when our entire regulatory structure and all of ‘medical ethics’ and pretty much everyone and everything else conspire, both implicitly and explicitly, to ban experimentation? By accident! They messed up and initially underdosed by half, noticed people weren’t getting as many side effects as expected, rechecked and then corrected for the second shot. Then in other places, they avoided their mistake.

Once the mistake had already happened, it would have been even more irresponsible to not continue than continue, so they got to finish. Now the new half-dose is likely the only ethical thing to do, given this result, so I am guessing that’s what will be done going forward. I guess this is what it takes to run experiments these days. It’s a real shame the whole thing is massively corrupted by other factors.

I tire of beating the dead horse that is ‘if we ran orders of magnitude more and bigger experiments we still wouldn’t be running enough experiments’ but it is good to see what it takes to try something even slightly nonstandard.

I do very much enjoy all the theories developed in hindsight about why the half dose strategy might be better, whereas as far as I could tell no one was suggesting such a strategy at all until the results came in that it worked. 

AstraZeneca has some huge logistical advantages over the other two vaccines. They can manufacture a ton of doses, as in billions of doses in 2021. The vaccine can be stored in a normal refrigerator.  They are going to only charge cost, which I am completely against for reasons that regular readers would find obvious but which definitely has its practical advantages. 

If you think that ‘the dose was off by a factor of two by accident in one of the trials and now they’re claiming that doing that was super effective’ might be somewhat of a red flag and you’re suspicious, you are not alone. This article raises various serious concerns about what AstraZeneca has been up to, and the data they’ve chosen to share. 

I am not sure how serious to take these objections. The source does seem to be operating the procedure “find things to raise alarm about” and throws a bunch of stuff at the wall, including things that aren’t worrisome at all. But assuming the article is factually accurate, I think the answer is still that it doesn’t look good. They released subgroup data without releasing other subgroup data. They combined different studies with different protocols to get significance, again while holding back other potential data. Getting the dose that wrong is quite the unforced error to make. There are some other issues too. 

It certainly isn’t clean the way the other tests were. Pfizer ran one test, with one endpoint, all announced in advance, released all the data, and got a result. Moderna did the same. We should feel pretty great about those results. AstraZeneca instead did a weird mix of things some of which were released and some of which were not, in different places, in ways that seem potentially vulnerable to various forms of p-hacking and manipulation. 

Since then, I’ve seen several sources I trust confirm in various ways that this doesn’t look good. Most telling is that the stock price is down on the announcement, rather than going up. 

I still think the vaccine probably works. If offered the vaccine I would happily accept it. Events still have not exactly inspired confidence. 

Russian vaccine also reports very good results. The data here seems rather sparse, the study seems like it was even more too small than the others. I don’t know to what extent we shouldn’t trust the Russians here, but I’m inclined to mostly believe that the data is legit and the vaccine probably works. One concern is that Putin is declining to take the drug before it is approved. When he takes it, or keeps finding excuses not to, that will be strong evidence either way. Hopefully his desire for the vaccine now speeds up their approval process! 

One other note is that I don’t have a good sense of the extent to which the different vaccines are competing for the same scarce resources. Having three vaccines is better than two, but how much better depends on whether producing one makes it harder to also produce the others. Either way, every time we get an additional vaccine, it does not seem like anyone’s timeline estimates change at all. That means that either we are dealing with such rapid scaling that a one time multiplier doesn’t change things much, or the other vaccines were already factored in because we (correctly) assumed they would probably work, or the timelines don’t are not actually physical estimates. Take your pick.

Vaccine Trials Available Now

The Pfizer vaccine is expected to be distributed on December 12. At that time, there will be enough vaccine doses for between two and three million people, and allocation will for now be determined by the states. It is expected that health care workers will get priority, after which it will be some mix of essential workers and the most vulnerable.

The rest of us still do have an option, if we want it. We can enter a vaccine trial. 

From the comments on last week’s post, we have links to some of the Covid-19 vaccine trials that may still be taking sign-ups. I have not investigated further, but figured I would signal boost this in case anyone was interested. A 50% chance at being vaccinated is much less than half as valuable as a 100% chance, but it beats the hell out of a 0% chance, and you get to help accelerate the end of the pandemic.

If you do join such a trial, you have full permission to boast about it in the comments and gain status thereby, ideally while also reporting back interesting information in the process.

Here are the links:

1. https://www.ensemblestudy.com/ – this one is for the two-dose Johnson&Jonhson trial.

2. https://www.covidvaccinestudy.com/ – the Pfizer vaccine, I think they’re still enrolling?

3. https://velocityclinicalresearch.com/ – not sure which vaccine they’re doing

4. https://www.c19vaccinestudy.com/ – Oxford/Astrazeneca

5. https://www.coronaviruspreventionnetwork.org/clinical-study-volunteer/ – generic website that creates a registry of potential volunteers.

6. https://www.fredhutch.org/en/research/divisions/vaccine-infectious-disease-division/research/immunology-and-vaccine-development/seattle-vaccine-trials-unit/seattle-covid-studies.html – sign up for a local trial in the Seattle area, Oxford/AstraZeneca

In Other News

Toilet paper shortages once again being reported. People seem to be following the logic of something like ‘the pandemic is going badly so there may be a run on toilet paper again because people are not smart.’ I had more to say and spun off a post that was up briefly on Wednesday, but objections were raised that I hadn’t sufficiently considered so I quickly took that post down and no longer endorse its conclusion that wet wipes are a superior good... 

Media coverage of Covid-19 in the United States might be slightly biased towards reporting bad news. 

Regeneron: Covid treatment used by Trump authorized by FDA. Good drug if you can get it at the right time. Better late than never.

Big wedding this past week in Brooklyn. That’s one way to handle a pandemic. I doubt this even conformed to fire safety codes. Cuomo called it disrespectful. When asked why he didn’t stop it, DeBlasio said “It’s a big city.” Either way, at some point shouldn’t one assume that everyone in that community already had it or is immune some other way, or at least is going to get it no matter what you do, and move on? 

Once again we have the reminder that it has not been ‘proven’ that the Covid-19 vaccines prevent infectiousness, in this case by the boss at Moderna, and thus we have not ‘proven’ that they will stop the spread of the virus. This is technically correct, which as we all know is the best kind of correct, but the vaccines turning out not to greatly reduce infectiousness would be, shall we say, a day of very low probability. It’s physically possible, but it’s not all that plausible that it would actually be the case. If you want to give me action on that at reasonable odds, let me know.

British people, and presumably people everywhere, dramatically overestimate their risk of death from Covid-19. That doesn’t mean correcting those estimates would make people’s decisions better rather than worse, especially since they likely underestimate the risk of long Covid and don’t think about anything like the true cost if they end up infecting others. It also seems unlikely to me that when people tell you a number that this means what it would mean if I told you a number. They’re not doing math.

Visitors to Britain can now pay money in order to get useful medical help from a private source, in the form of a Covid-19 test that can shorten their quarantine. Interesting concept. Raises the question of why this wasn’t an option before, but good news is good news. Meanwhile in New York City, you can get a Covid-19 test so you can go to a restaurant. Which makes me kind of want to go to that restaurant.

Guardian raises a key point that isn’t getting enough attention, which is that if mRNA vaccines work, potentially this opens the door to lots of other new vaccines for other things. If we were smart, we’d do another of these Warp Speed things now, because vaccines are so amazing that it’s worth doing that when there is no pandemic. Now imagine reacting reasonably to a pandemic.

In an interesting experiment, Vermont will ask students next week whether they would prefer remote learning. Looking forward to the results. If you think this is something different, our models of schools and children greatly differ.

I recently wrote a piece defending polling and models based on polling, but one does need to remember that when you ask if someone wears masks or does other desirable things, you might get a bunch of people going “um, yeah, sure, of course I do that.” 

Rapid antigen tests not only work, there are places they have been actually used and did a lot of good work. Good to see some places being less insane than others.

Yet another observational study of Vitamin D and Covid-19. Of the asymptomatic patients about a third were Vitamin D deficient. Of the severely ill, ninety-six percent were Vitamin D deficient. Fatality rate difference was 21% vs. 3%. I do realize it’s all observational, but come on. Seriously, supplement Vitamin D. I continue to take 5000 IU/day.

Yet another statistical analysis pointing out that lots of inaccurate testing would rapidly end the pandemic, for little cost and purely on a voluntary basis, but we’ve banned such testing. 

Yet another call to run challenge trials so we can actually learn things. And yet another reminder that doing so would have meant the pandemic would already be over, since the Moderna vaccine was designed in two days. In a pinch my understanding is they could have done it in one.

Another write-up of how Covid-19 was a reasonably standard virus and could have been dealt with in standard ways

Correction to previous information about Covid-19 and mental health. The numbers remain bad both for Covid-19 patients and for other patients, and the overall mental health situation still seems highly disastrous in general, but the numbers gave the false impression things were several times worse than the true numbers would indicate. Some commentators also caught this. My apologies.

“Today was a very, very odd day. I testified before [the Senate]. They held a hearing on hydroxychloroquine.

El Paso is now transporting hospital patients up to ten hours away.

The problem lies not in our candles.

Lawsuit: Tyson managers bet money on how many workers would contract COVID-19. The managers are being accused of having knowledge, keeping records and attempting to model the world. Highly scandalous. It would have been much better for them if they had still imposed horrifically risky and abusive conditions on their employees, but had done so via only implicit collusion with plausible deniability, like one does in a proper maze. Bad form indeed. 

Joe Rogan did an episode focused on Covid-19. It’s mostly pretty good on both ends, and provides both good info and good suggestions of some things a functional civilization might do. Joe puts a lot of focus on maintaining health, and asks quite pointedly why we aren’t helping people strengthen their health and in particular their immune systems as part of our response to Covid-19. Overall, not that much here that regular readers of this column wouldn’t already know, but a reasonable place to point regular folks who aren’t up for something like these posts, to get them up on the basics and hopefully taking the biggest wins, especially supplementing Vitamin D. 

LessWrong post of the week that isn’t about dealing with Covid-19 but also kind of is: Pain is not the unit of Effort. We also got Embedded Interactive Predictions on LessWrong which means I can now experiment with them there. This week I didn’t have time.

This isn’t Covid-19 but there was this claim by some Israelis to have reversed the human aging process. Using oxygen. In particular, they are claiming they can lengthen telomeres and the accumulation of resulting senescent cells. Huge if true! As I understand it, that would be two down and five to go for the SENS project, and would probably imply a few extra years of life expectancy and be worth a lot in life quality along the way. Again, if this is real. I assume it probably isn’t, so no one should get excited at this time. I assume I have some readers who can explain why this is nothing to get excited about, but seems worth asking for them to do that.

Thank You

To all the health care workers. Thank you.

To all those maintaining the supply lines. Thank you.

To those working to manufacture the vaccines as quickly as possible. Thank you.

To Pfizer, Moderna and AstraZeneca/Oxford. Thank you.  

To all those working on all the other vaccine candidates, even if they didn’t work out. Including SputnikV in Russia, and the one in China. And anyone working on all the other treatments, or running any of the experiments or studies, again whether or not your particular effort paid off. Thank you.

To all those working on or expanding capacity for or administering or fighting for the right to do Covid-19 testing, especially rapid testing. Thank you. 

To all the essential workers. Thank you.

To you, the reader, for being here. Thank you.

To all the commenters, yes all of them. Thank you.

To all those who have thanked me for doing these columns. You keep me going. Thank you.

To everyone who helped push masks, Vitamin D, airborne transmission, doing things outdoors and other key information when official sources were saying otherwise. Thank you.

To Robin Hanson in particular, who tried to actually figure things out and propose the best solutions available, and made serious attempts to make that happen. Also, without our debate and the reaction to it, I would never have felt free to start writing these columns. Thank you. 

I’d also single out Tyler Cowen for his work at Marginal Revolution, which has provided lots of useful thoughts and information. And for his rapid grants. Thank you.

To the Covid tracking project, and all other similar projects. Thank you.

To the financial security and opportunity to keep me and my family safe, and give me the necessary time to write this column each week. And to the freedom of speech to say what I think each week. Thank you.

To China, South Korea, Japan, New Zealand, Australia and all the other places that beat the pandemic. You showed us it can be done. Thank you.

To everyone working to reform or contain the damage from the FDA, CDC, WHO or any other member of the Delenda Est club, or otherwise find ways around regulatory barriers. Thank you.

To the owner of the New England Patriots, who flew a jet in to get protective equipment to health care workers. And everyone else who did what was necessary in the face of banditry, piracy and regulatory obstruction to get people what they need. Thank you. 

(Don’t worry, I still hate the New England Patriots.)

To every politician and public figure and corporate leader, or anyone else, who called upon people to take precautions and also accepted skin in the game and practiced what they preached. Thank you.

To those politicians who did the best they could to help people, based on their model of what would physically help, even if I disagree with that model. Thank you.

To everyone doing what they need to do to keep themselves, their families and friends and their communities safe. Thank you.

To everyone doing what they need to do to keep themselves, their families and friends and their communities sane and thriving through all of this. Thank you.

To my in-laws, who helped us get out of New York City in March and have been invaluable keeping things going out here in Warwick. And to our kid’s nanny, without whom disaster would have rapidly ensued. And of course to my amazing wife Laura, for far too many reasons to list here. Thank you.

To my father Solomon, who did his part in all this to help make things better however he could, even if I can’t talk about it on the internet. And for teaching us a legit immunology class over zoom, and helping me understand the science whenever I needed it. And for keeping sane through everything that has happened in that tiny apartment. Thank you.

To my temporary home here in Warwick, New York, thank you. You have exceeded almost all of my expectations. This place is highly underrated. 

To my cofounder Kathleen Breitman, and my coworkers Alan Comer and Brian David-Marshall, as we continue to fight to make the game Emergents a reality. Thank you.

I’d also like to thank India, clarity, disillusionment, consequence and silence. Especially clarity. But screw frailty. 

Let’s not forget the internet. In particular, Amazon, Instacart, Google and Netflix. Couldn’t have done it without you. Thank you.

And also basically everyone anywhere who modeled the world and is now doing a thing to try and make the physical world better, regardless of whether I think it’s misguided or nonsense. To all the schmucks who think for themselves. To all the live players. You rock. Thank you.

Finally, to everyone I’m forgetting. Thank you too. Comments to thank those I missed are encouraged.

Next week’s numbers will be misleading, and will need to be interpreted carefully and skeptically. The holiday will slow down testing and reporting in strange ways. Infections from the holiday will lag by about five days. The backlog of testing demand will partially resolve itself. Even more than usual, we won’t know as much as we would like. 


New Comment
14 comments, sorted by Click to highlight new comments since: Today at 3:02 AM

You missed one! Thanks to Zvi for keeping us all informed every single week for months :)

For real though! I mean, the concept of a “rationalist newspaper” has been floated around for a while now. But now I have a much better sense of what it might look like. This is a great way to get news; I feel like I’m learning about the world in the process too.

I second the motion!

This isn’t Covid-19 but there was this claim by some Israelis to have reversed the human aging process. Using oxygen. In particular, they are claiming they can lengthen telomeres and the accumulation of resulting senescent cells. [...] I assume I have some readers who can explain why this is nothing to get excited about, but seems worth asking for them to do that.

Ok, there's a lot going on here.

First, general epistemic comments. Paper is here. There are some major red flags: only ~20 patients in analysis, no control group, tested a bunch of different cell types and endpoints. In this case, I think the lack of a control group isn't too alarming - we have a pretty decent prior idea of what "normal" looks like in old people, and in some ways using the initial conditions of this particular group as the "control" is better anyways, especially with such a small sample size. The garden of forking paths is a bigger concern. The effect sizes and p-values are strong enough that I still think there's plausibly a real effect here, but definitely take it with a sizable helping of salt.

The main measurements I'd pay attention to are the senescent cell counts post-treatment (the "post-HBOT", taken "1-2 weeks" after the treatment concluded). Hyperbaric oxygen will definitely have short-term effects, but it's mainly the longer-term effects which are interesting here, so post-HBOT is the thing to look at. Telomere length measurements in general are... kinda tricky to interpret. In normal operation, they're effectively a downstream measurement of DNA damage rates (which are what "really" seem to matter for aging), but some interventions can lengthen telomeres without significantly reducing the damage rates. Senescent cell counts, on the other hand, seem to be more directly relevant, based on my current best understanding. Put that together, and we can ignore most of the forking paths and just focus on post-HBOT senescent cell changes.

And in this case, the post-HBOT senescent cell changes are exactly where the most dramatic results are. They looked at senescent cell counts in two cell types. One had ~37% drop in senescent cell count, the other had ~11% drop. Those are definitely not "the problem is solved" kind of numbers, especially when the 37% drop is only in one cell type, but it's substantial.

The bigger question is how long the effect lasts. The study only checked in 1-2 weeks after treatment, which is a bit less than the typical half-life of senescent cells. What we really want to know is whether the effect persists after 6 months or a year. Given the mechanisms involved (i.e. hyperbaric oxygen, defense activation), I would expect a priori that it probably wears off after 1-2 months, and that in the long run the hyperbaric oxygen exposure accelerates aging overall.

I did an analysis of how convincing the Oxford-AstraZeneca claim of 90% effectiveness is.

Unfortunately I inferred the numbers of infections in each group incorrectly according to this - the infections were split 3:27 between the half-full group and the full-full group, not 2:28 as I'd calculated. (Note that the naive interpretation of the numbers doesn't come to 90% or 62% effectiveness so I assume they're doing some corrections or something else which alters the result slightly.)

That means the 8:1 Bayes factor I originally calculated (in favour of half-full being more effective vs the two different regimens being equally effective) comes down to 2.9:1. In my book that isn't enough evidence to overcome the prior against the half-full dose regimen being more effective.

The above assumes that everything else about the groups is equal.

Having read the report linked in the OP I think the actual update should be noticeably lower, particularly as the half-full treatment group were younger than the full-full treatment group (or at least only the latter included anyone >55 years old).

(I mean which interpretation will the evidence favor, not on whether they go ahead with the half-full as the standard dose)

There's also a Metaculus question about this:

Thank you to members of the LessWrong community for their efforts early on in the pandemic.

I am so glad these posts exist. They have been a rare breath of sanity this year.

That said, I disagree with your point specifically about camp grounds. Obviously taking a camping trip is nonessential (and that this fact dominates the optics, relative to closing houses of worship), but about 1 million Americans live full time or near full time in RVs (I plan to join that number within a year or so, so this is very much on my radar), and the camp ground closures early on in the pandemic caused many of them to be evicted, suddenly and without warning, at a time when there was no where else for them to move to. After at most a week or so, that made many of those people have trouble getting sufficient access to electricity, fresh water, and proper sanitation. Today, with restrictions on interstate travel, it could be much harder for those people to simply go elsewhere (say, to stay with relatives, or to visit states that still have camp grounds open).

I'm sure there is some way to have written the restrictions to get around this (though I'm not sure if that would be enough to keep camp ground operators afloat financially), but given the general lack of reason and nuance in what gets shut down, I don't really think anyone was going to put in the effort to do so. 

Ah. When I think camp grounds I think... camp grounds for camping trips. Which isn't dangerous at all, but also isn't essential.

So camp grounds are de facto semi-permanent housing, or collectively serve that function because people rotate between them? 

If that's true, and there's no alternative, then it would make sense to let people already there remain, and if that requires keeping them open, so be it. I don't think it changes the larger argument. 

Of course, the correct answer is actually more like "we don't have to think about this as essential because it's safe so there's no reason to shut it down" and calling it essential is being perverse at least once and potentially twice because of that. 

Of course, the correct answer is actually more like "we don't have to think about this as essential because it's safe so there's no reason to shut it down" and calling it essential is being perverse at least once and potentially twice because of that. 

No argument there!

I hope you don't mind my saying "You're very welcome."

(Healthcare, public safety)

And thank you very much for providing these regular figures and interpretations.

From an old Garrison Keillor Thanksgiving special:

Thanks to all those who when thanked

Don't sit there as if insentient

Or mutter don't mention it

And don't act like your heartfelt mercy

Were some kind of heresy

Which makes you want to grab 'em by the ears and yank 'em

Thanks for people

Who will let you thank 'em!

again, if this is real. I assume it probably isn’t, so no one should get excited at this time. I assume I have some readers who can explain why this is nothing to get excited about, but seems worth asking for them to do that.


The Fight Aging take on this (one of my favorite longevity blogs) is that if hyperbaric oxygen chambers actually effected the relevant biometrics, we would expect to see more miracles from people going through them.  Their take is that the biometrics of immune cell senescence and telomere length just aren't that great at tracking aging.

Thank you.

New to LessWrong?