The Politicalmath take
This post contains mostly meta commentary on my own account. Skippable.
I wanted to eventually write about this tweet, the most controversial take I’ve posted:
Politicalmath unexpectedly praised me as a straight-shooter, so it makes sense to address it now in the spirit of reconciliation. This post discusses what I’ve learned since tweeting the politicalmath take, how I’ve changed the focus of my account, and why I wouldn’t post this take today.
Why I posted politicalmath’s bad takes
Most of the takes I post are suggested by users who DM or tag me in tweets. That’s what alerted me to the pair of politicalmath tweets. Slate reporter Mark Joseph Stern flagged an existing tweet that featured the same pair of takes I ran (with h/t).
As always, I searched out the underlying tweets to make sure they weren’t being misrepresented. The March 18 “putting down my marker” tweet was a little-seen prediction politicalmath made after a dialog with one of his friends, who asked him directly how deadly poli then believed COVID-19 would be.
I wouldn’t have posted that take on its own. Lots of people threw out idle estimates in mid-March, and why should I feature a guess 11-tweets into a thread on the timeline of an account with 1000 followers? Like, Clay Travis exists. I’d much rather post someone actively using a large audience to spread misinformation.
The April 26 “put those people on a list” tweet made me happy to run the take.
Twitter pro tip: don’t ever post anything like that. It’s a sure-fire way to go viral in an unflattering way. I think human beings are hardwired to spread gossip about at least a couple things:
People you don’t like getting hoisted with their own petard
Examples of apparent hypocrisy
Announcing that you think we should “make a list” implies a built-in audience of people who dislike you, and so of course they’re going to look for the thing that makes it look like you belong on your vindictive list, and they’ll probably find it. Even Caesar’s wife has a couple of eyebrow-raising IG posts.
Anyhow, in verifying the “put those people on list” tweet, I determined it came from a thread where politicalmath attacked Noah Smith for something Smith had said a month earlier—that while COVID-19 was then hitting New York City, it would be hitting your city next. Politicalmath opined Smith was “shit at this” and emblematic of “shitty people” who “don't care about the truth.” In my opinion, this rant hasn’t aged well, nor has the earlier dialog between politicalmath and Smith, where poli contended that New York City was sort of sui generous but Smith believed it was more of a harbinger of what could and would happen elsewhere without strong mitigation. North and South Dakota arguably tested this disagreement and have so far achieved overall per capita deaths 85% as large as New York in spite of the advantage of having six months more medical knowledge. Smith was arguably wrong about the timing of the spread—none of us knew how seasonal it would be—but I don’t think his comments have proved to be “dangerously wrong” as poli contended.
So I still think it was a bad take.
But I wouldn’t run it again given the chance.
What happened after posting the take
I mute tweets after posting them so that my mentions are at least sort of useful.
I didn’t realize there was any controversy about the thread for several hours, and then noticed that politicalmath was very active in the replies to my tweet. I’m not sure that helped him; it’s very hard to frame the “put those people on a list” take in a way that doesn’t sounds like “nooooo! I meant him, not me!”—the sort of hypocritical-sounding thing that just makes the tweet even more viral.
wint @dril"im not owned! im not owned!!", i continue to insist as i slowly shrink and transform into a corn cob
May 6 changed everything.
Politicalmath had sent my account a puzzling assortment of messages asking whether I had seen the context of his tweets, why I didn’t feature takes of Noah Smith or people who had predicted that 2.2 million would die (answer: they didn’t), with poli’s messages becoming increasingly depressed before he signed off by wishing me safety. I tried to wish him safety in return, but I’m not sure that message went through; he blocked my account.
Tons of people were tagging me, and the most alarming messages were about what politicalmath had tweeted late the previous night: that his manager had been contacted over his Twitter activity and that he intended to delete his account because he couldn’t risk his livelihood and family comes first.
Honestly, I didn’t believe it at first. I’m a skeptical person by nature, and I’ve grown up on message boards and Wikipedia where “delete my account” threats are part of the ordinary drama of the platform. That said, I did see at least some of my followers who seemed to think it would be a fine thing for someone to call up an employer for swearing at Noah Smith on Twitter dot com, so I posted a self-deprecating warning to my followers that no one should do something like that:
Phil @philllosoraptorI'm pissed that this happened. Polimath is a good dude, smart, kind and a friend to lots of us. Because that trash Bad COVID-19 Takes account screenshotted him to dunk on him, this happened. You people who did this are the scum of the Earth and I hope karma bites you. https://t.co/JSV0Cn8Con
I was disappointed when one of my followers still replied that they still had no sympathy, so I added this explanation:
I dislike internet mobs. In small social groups, shame is a pro-social mechanism for encouraging people to avoid harmful behavior. In small groups, the mechanism has proportional limits. A finite number of people in a small group can shame another, and they can’t get too carried away with it because everyone is a repeat player. Folks remember over-reactions. But shame becomes toxic on social media when thousands of people can turn to attack a stranger, unrestrained by any realistic consequences. (I’m sure tons of people have made this observation, but I first remember reading it by Daniel Solove.)
TLDR (with twist ending): I feel bad for bean dad.
The problem is almost certainly worse with BadCOVID19Takes than other accounts because it’s a matter of life and death. Some shame is good, but not when folks’ zeal knows no bounds.
Even though people were just angry at me posting politicalmath’s tweets, it felt weird to see random people curse me by my Christian name and see others suggest that maybe someone should call my work. I realized that whatever blowback I experienced was a pale imitation of what politicalmath experienced.
By the afternoon of May 6, I was messaging a mutual friend who told me that I should delete the tweet and apologize. I didn’t want to do it because it was actually a bad take and thought it would encourage other accounts to send their followers after me.
Politicalmath’s rebuttal and why I didn’t post it
Politicalmath posted a rebuttal or explanation on the afternoon of May 6, which is strangely not indexed by Twitter’s search function. Parts of it are well-taken and it’s worth a read:
As explained above, I agree the March 18 tweet wasn’t a particularly bad take.
When I first read his explanation, I found it a little goofy that a Twitter user with ~40,000 followers would act as if a Twitter dialog were a private conversation, but I’ve come to realize that politicalmath uses twitter quite differently from me. He’s an extremely extroverted and social person, and the way he interacts is more like I associate with Discord or the topical message board communities that used to exist before Reddit superseded them.
I thought it would be fair to append a link to politicalmath’s response to my tweet, and that this might make everyone feel better about the incident, but my pride stood in the way. I do not agree that the tweets were taken “entirely out of their context.”
Most people use the expression “out of context” when the meaning has been distorted by the lack of context. For example, when someone quotes words to condemn another, but obscures the context that the words were actually satire or a quote from another source. I take this sort of misrepresentation very seriously, and so does poli. But the context doesn’t change the meaning of the words here. Politicalmath really was guessing that only 10-20,000 Americans would die of COVID-19 and he really was opining that someone should make a list of people with incorrect predictions.
By “out of context,” politicalmath meant that I didn’t explain that these tweets were in the middle of longer threads, and that by not linking the threads I was unfair to him.
Maybe so! But I didn’t want to link to politicalmath’s rebuttal without expressly disagreeing that the tweets were “taken out of context,” and I figured that including such explanation would draw more unwelcome attention to both of us and might even be taken as an escalation, so I moved on.
I followed polticalmath closely after then. At first, for angry, defensive, petty reasons. For example, I fumed when one of his friends commented as if I had tried to get politicalmath fired, and was upset that poli didn’t set the record straight that I hadn’t done that; it was different folks with a different grudge.
But I eventually realized politicalmath was genuinely hurt by the whole affair. It wasn’t just online drama, and so it’s not surprising that he wouldn’t have anything nice to say about the person who (from his perspective) randomly brought him so much unhappiness. I empathized with how he said it drained the joy from Twitter. Running BadCOVID19Takes became a chore. People I know and respect condemned my account. I stopped posting as much, but I didn’t spend time on other things—I just doomscrolled more. I dreaded DM’s even more than I normally do by virtual of being an asocial procrastinator. And I know that much more bile was sent his way.
I also started to identify with him personally. We’re both non-Trumpy moderate conservatives (I found it amusing he initially though I targeted him for his politics). We’re almost exactly the same age—I doubt we’re more than one grade apart based on his Oregon Trail generation references. He also like craft beer, but he’s not the snooty whale-hunter sort, more like the enthusiastic “try this regional brewery kölsch with grilled brats!” sort (note: I might just be projecting now). Finally, most weirdly and affectingly, I learned his father died within days of when mine did, a bit over two years ago. I tend to be a quiet, stoic person and would have never posted the full spread of reflections that he did, but I’ve thought about it a lot.
More relevantly, his COVID-19 takes are generally not bad. While politicalmath is defensive of Florida governor DeSantis to a fault (and I mean that, poli), he also resists unfounded narratives that have cropped up on the right. For example, when many conservatives (even non-hacks) were suggesting that Pfizer and/or the FDA conspired against Trump by delaying the vaccine announcement, poli examined the timeline and concluded otherwise. Pfizer and FDA were working in good faith as quickly as possible to modify the approval plan—it turns out the researchers did not foresee that the accelerating pandemic would cause the trial to essentially blow past interim milestones. If more conservatives followed commentators like him instead of grifters and corona-charlatans, I sincerely believe fewer people would die.
How BadCOVID19Takes has changed since May 5
Almost immediately, I started becoming more rigorous about redacting usernames from tweets. I only leave the name for verified accounts, users with many followers, and people engaged in bad coronapunditry or antivax nonsense essentially full time. This is unpopular with my followers; almost every time people ask why I covered up the name of someone who “deserves” to be shamed but I think this is the right approach. Some users look up the original tweets anyway, and that’s usually more than enough public shaming than needs to be inflicted.
I’ve also tried to specialize more in “debunking”-type posts. An account called @BadBadTakesTake (which I initially thought was a politicalmath sock puppet) influenced a lot of my thinking of how a “bad takes” account should be run, partially by criticizing my account. These tweets made a big impression on me:
My account was originally inspired by @BadLegalTakes, which has a fine format in simply posting screenshots of bad legal takes. Lawyers, litigators at least, love more than almost anything explaining why people are wrong, so BadLegalTakes has a built-in army of volunteer explainers. (Note: I’m a litigator.) I also liked how BadLegalTakes seems sphinx-like and inscrutable in its silence.
But this didn’t always work with BadCOVID19Takes. For months I would often overload tweets with both the bad take and documents showing why they’re wrong. This is sometimes an amusing presentation (particularly for side-by-side bad takes), but it’s bad for engagement, it’s unsearchable, and often results in posts that are incomprehensible without taking a lot of effort.
The @BadEconTakes approach works better for my account. I tend to provide context for any tweet that isn’t extremely obvious. I am encouraged that some people say these “context” posts make the account worth following, because it takes much longer to research and present exactly what errors bad takes make.
Why I wouldn’t post the May 5 tweet today
But none of this explains what I said in the first section—that politicalmath’s “put these people on a list” tweet was indeed a bad take, but I wouldn’t feature it today.
Essentially, I’m increasingly reluctant to drag someone in a way that produces more heat than light.
First, it’s kind of cheap to run such a take. Politicalmath’s “list” tweet was mostly just assholish, and a bad idea for reasons unrelated to COVID-19, which poli himself likes to explain when he isn’t irrationally angry with someone:
A lot of users and accounts exist to drag people for general assholery, and my account doesn’t need to be one of them. It stretches and dilutes the focus of the account.
Second, fighting the virus is my ultimate goal, and I believe tribalism undermines it. Unlike the corona-skeptics, I believe incremental changes in behavior can marginally improve outcomes and that we really are all in this together. A lot of celebrities and politicians have forfeit their moral authority through viral hypocrisy, so I’m not optimistic governments in America can do much better than they’re doing now—unless people insist on it.
We can do better. We can choose to forego risky behavior if we believe the goal is worthwhile. We can be persuaded it’s worthwhile.
Tribalism interrupts persuasion. It may be satisfying to bash Republicans, or whoever, as anti-science Neanderthals, but it isn’t actually helpful. Many conservative people live near you, whether you know them or not, and to the extent they’re more inclined to engage in risky behavior right now (probably!), they’re the most valuable hearts and minds to win.
We can’t remake the population to our liking or rewrite decades of history and politics. But we can meet people where they are. We can convince them based on our shared values that sacrifice now will ensure more precious lives make it to the other side of this epidemic. I believe folks can be persuaded.
I’ll take imperfect allies wherever I can find them, so I now give wider berth for intemperate, dumb remarks by usually reasonable commentators that bear only tangentially on the pandemic itself. Sure, it might be fun to post that Nate Silver take where he sounds like an absolute megalomaniac, but we’d be lucky to have a public discourse dominated by people who try in good faith interpret evidence, even if it sometimes exceeds their grasp.
Bottom line: dragging often-reasonable people does not win hearts and minds.
Cynics might wonder if I’m now soft on politicalmath for having grown to kind of like him. Cynics might be right! I’ve seen a lot of motivated reasoning in the last year, and I always doubt that I have a good grasp on anyone’s motivations, including my own.
I’ve made mistakes, and this might be one, but I think it important to build bridges to potential allies because we’re going to need all the cooperation we can get in the months ahead.