One thing that watching our response to COVID brought home to me is that dealing with randomness is really hard, in lots of different ways. Some ways in which that played out:

Group randomness versus individual randomness

There are lots of actions that you can take that make a difference in terms of your chance of getting sick or of dying from COVID at an individual level; and there are also lots of actions that we can decide to take as a society that affect its spread at a group level. Since COVID is a transmissible disease, almost all of the individual measures affect the group, too, but the effects are very different. And it feels to me like, at least in the United States, people seem to gravitate towards one mode of analysis or the other, with conservatives preferring to focus on the individual level and liberals preferring to focus on the group level.

Hidden randomness

A lot of COVID-related randomness is in principle measureable but, in practice, not actually measured. Sometimes, we have good measures: e.g. vaccine trials give fairly reliable numbers for drug effectiveness. But even those numbers aren’t dependable in a fixed way over time: different COVID strains change effectiveness numbers, and vaccine effectiveness degrades over time even without a change in strain. And how do we take future strains into account?

But a lot of random effects are much harder to measure. There are a lot of people trying to figure out the risks of harms that are directly caused by COVID, but what about the potential harms to children from having to spend time at home instead of having to play together, the harms to grandparents who couldn’t touch their children and grandchildren for a year or two, the harms to parents who were expected to both work and look after their kids and their kids’ schooling, the harms to people that come from having their worlds constrict? Having those problems be hard to measure doesn’t mean that they aren’t real and important.

Rare events are hard to think about

I think humans are not very good at distinguishing between types of low probability rare events. You should, in general, react very differently to something that has a one out of a hundred chance of happening and something that has a one in a million chance of happening, but those both feel like they belong in a “very rare” bucket.

Also, our media environment, especially our social media environment, makes it very easy for anecdotes to spread, spreading the perception of bad events; so not only is it hard to react appropriately to a probability, we’re getting bad probability signals in the first place!

Not believing in uncertainty

I think that, at a gut level, many people don’t believe in uncertainty, at least when it comes to themselves. (They’re the main character of the story, after all, so everything involving them is plotted out and will turn out well, or at least turn out badly in a dramatically appropriate way.) Sometimes this involves overestimating certainty in something: this random COVID cure will definitely work. And sometimes this involves discounting things that are probable but not certain: vaccines don’t prevent COVID 100% of the time, so therefore they’re useless. And, of course, frequently those two go together: the patient who refuses to take vaccines, then gets very sick and goes to the hospital, and then demands treatment from a drug that they are sure will work, because that’s the way the story in their head is written.


In general, I like trusting experts; but COVID has shown the limitations of that, with the CDC, FDA, and for that matter my local doctor’s office behaving badly. They emphasized the importance of washing hands and underemphasized (and still continue to underemphasize!) the importance of masks and ventilation, and their discussion of distancing was sorely lacking in nuance. And, even though we now have an amazing ability to develop vaccines, the approval process means that COVID strains run through the population months before vaccines for them are improved: the disease’s OODA loop is faster than the OODA loop for the approval process.

Of course, some people got the masking and ventilation question right early on: e.g. I was pretty convinced by Zeynep Tufekci that masking and ventilation are important, and she was in fact correct in that. But I absolutely don’t want to take from that the idea that, if you read a range of people’s opinions on COVID-related subjects, think about them, and come to the conclusions that make sense to you, then you’ll end up in a good place: that also describes a bunch of people who ended up being convinced that a deworming drug was a great cure, or for that matter that vaccines are harmful instead of helpful. (And, for those of you who, like me, are convinced that Ivermectin is a bad COVID treatment, what is that belief based on? I was surprised when I read a discussion of studies of its effectiveness.)


I find it particularly hard to reason about Long COVID. It clearly exists, and it can be very serious. But also I’ve run across a few reports of studies that claim that the rate of symptoms of Long COVID isn’t as high as I would have expected compared to the base rate. And a big part of that is because the base rate is higher than I would have thought. Which makes me wonder: for what other diseases X is Long X also a thing?

And I don’t have a feeling at all for how the prevelance of Long COVID is affected by vaccination. Vaccination doesn’t completely prevent Long COVID, but that’s not a surprise, it doesn’t completely prevent anything. My default assumption is that vaccination makes Long COVID less serious, but I don’t have any concrete data to back that feeling up.

At least the medical establishment is acknowledging that Long COVID is real. And, if it turns out that there are lots of other similar syndromes out there that the medical establishment has been underplaying, hopefully people suffering from those will get better treatment now.


I’m spending a lot of time talking about uncertainty in the context of COVID, but of course nothing about our life is certain. So we do what feels right to us in a given context, and whatever happens happens.

The problem with COVID is that it’s new, so for a long time I didn’t know what felt right! At first, it was at least clear that COVID was unusually serious, so holing up and going to significant lengths to try to avoid getting affected seemed like the right way to react. But, after vaccines became available and effective, it wasn’t so clear that treating COVID as exceptionally serious continued to be the right thing to do.

What I really wanted was an analogy with something familiar: if I can accurately say “COVID is like X” then I can behave around COVID like I would behave around X, and I’d feel comfortable enough with that choice. The obvious choice for X is the flu; but is COVID for vaccinated people about like the flu? And, to be clear, I don’t see this as meaning that I don’t have to worry about COVID: one year I came down with the flu, had it turn into pneumonia, and was unable to work for about a month and recovering for a while after that. This is much worse than what happened to me when I came down with COVID, and I strongly suspect that, if this had happened a hundred years ago, I would have died as a result.

It feels like the answer is: vaccinated COVID and the flu probably are comparable? In terms of chances of dying or serious illness, I think they’re in the same ballpark, which is serious enough to not treat them lightly but also for me to not hole up trying to avoid them. The main thing that makes me unsure there is Long COVID: how much do I, as a vaccinated person, have to worry about it? And, for that matter, is Long Flu a thing as well that we just hadn’t recognized? Beats me.

Of course, not everybody is in my situation. For the elderly, COVID is quite a bit more dangerous than it is for me; but that’s true for the flu, too, so probably the flu is still an okay analogy? And presumably immunocompromised people have to worry more about the flu than I do as well, so maybe the analogy holds there too? I’m not sure about any of this, though.

Also, this analogy, assuming that we accept it, cuts both ways. If COVID is like the flu, and if COVID has gotten us to change our behavior, maybe it should also have us be more careful around the flu? I think this is probably true for me: in retrospect, I probably went out in public too much in the past when I had the sniffles, and I also probably didn’t do as much as I should have to protect myself from others when I wasn’t sick. So quite possibly, a decade from now, I’ll still be wearing a mask when I’m taking the train during the winter; I’ll certainly be wearing a mask most of the time when I feel like I’m probably sick but feel like I need to get some shopping done.

What to worry about

There’s one more way in which my behavior relating to COVID has been changing recently: what I worry about or get angry about. When COVID first came on the scene, I spent a decent amount of time developing and having opinions about what we, as a society, should do to try to reduce the effects of COVID. And I’m not going to say that that was a bad idea back then: COVID was new, important, and dangerous, and it wasn’t at all clear how we would react to it or how that would play out, so it’s natural to spend time thinking about that.

But it’s two and a half years later now. I could spend time continuing to have a strong opinion about, say, whether we should still require masks for people riding transportation; but is it helpful for me to develop and maintain an opinion about that?

The conclusion that I’m coming to is: no, it’s not helpful, it’s actually probably harmful. Because, concretely, what would I be trying to accomplish by feeding those opinions? Coming back to what I said at the top of this post: there are choices that matter at an individual level, and choices that matter at a group level. And I’m not in a position to affect the group level choices; so would I be trying to get out of spending time thinking about them?

If it’s just intellectual curiosity, then I’m all for that, but there are a lot of things to be curious about and I could probably find a better focus for my curiosity. In particular, if developing a strong opinion about correct group choices is just going to lead me to be angry when we make the wrong choice, then that’s probably harmful for me. Also, at least in the US, right now beliefs about the correct approach towards COVID feel way too strongly tied to political group membership; I don’t think that’s healthy, I’d rather sidestep that by not participating.

So I’m still thinking about how I should respond to COVID as an individual (when to mask, when to get vaccinated), because that is something that I have control over. But I’m trying not to spend much time worrying about whether the country as a whole is making the right choices in how we open up, because I don’t have control over that.


This chain of reasoning generalizes; and, to some extent I think those generalizations are pointing in a useful direction. It feels to me like lots of us spend a lot of time worrying about things that we can’t control, and at times fixating on catastrophic situations; that worrying is not healthy, we should do less of it!

But also, this distinction between individual choices and group choices is an artificial one. Groups are made up of individuals; so sometimes general public opinion matters, and sometimes individuals are in a position where they make decisions that affect groups.

Which leads us to the flip side: being at least somewhat informed about matters that affect society broadly is a good idea, I think? I don’t have much patience for the idea that a single vote is unlikely to matter so you shouldn’t pay attention to politics.

So if you want to be one of those people who makes individual decisions that has an outsized effect, then more power to you. But if you’re doing that, then please take that desire seriously. Of course, part of taking something seriously is digging into the pros and cons of various approaches to the issue in question. But also a big part of it is to dig into what actions would actually be effective towards getting us closer to your preferred position.

Just convincing yourself that position P is the best one isn’t going to do squat towards your desired end; and telling people who disagree with you that they’re wrong and bad people is also not effective. Exactly what is going to be effective is a very difficult question to figure out; but if you really want to make a difference, then that’s what you’re signing up for.


Anyways, that’s where I’m at: I’m chosing to behave as if COVID is kind of like the flu, I don’t know how accurate that is, but I also don’t think it’s helpful for me to spend time worrying about whether I’ve gotten it right, and I certainly don’t think it’s helpful for me to spend time worrying about whether the country as a whole has gotten that right.

Post Revisions:

This post has not been revised since publication.