I am taking stock of my first year on Substack, and my first year with the resources necessary to focus on attempting to be a public intellectual of sorts.
Really liked your takeaways, particularly #6 and #9. Curious what you see as the main limitations to individual post metrics and where you do see value there. Also why does no one click on links? Seems kind of weird to me because posts that center around sharing links seem quite popular with readers, yet even on those click rates are low.
I don't know the technical details of how link clicking is tracked but I wonder how reliable it is (does it work with ad/tracker blockers that are increasingly common? Privacy focused browsers?).
I'm either a very abnormal reader or else the tracking is poor, because I'd guess that I click half our more of the links on nearly every post.
I just assume that not everyone is as voracious for new information. Normal people don’t click link after link. I find it reassuring to see a link but rarely check that it backs up a claim, unless I am really into a subject.
Still I would think Zvi's readers would be pretty high resolution thinkers. When I put "On Bounded Distrust" into a word counter it was about 34 pages single-spaced!
This is my feeling as well. I don’t click links unless I’m particularly interested and/or skeptical, but it’s reassuring that it’s possible to check the source of information.
Tyler linked to a post thinking about why no one clicks on links, and also a bunch of people answered with their takes. I'll write a bit more about it in the January roundup, probably. It's kind of like, the link needs to exist so people know it exists, but given we know it's there no need to click.
I kind of think a bunch of TV wants to work that way and can't? As in 'now we are going to show the awkward conversation because we have to, but it's not fun and doesn't contain unexpected info so it would kind of be better if you knew to skip it.' Ya know?
The limitations on post metrics are basically that clicks tell you a lot more about whether certain people linked to it and timing queues and not much about whether people actually read or approved of it, and likes/comments are about what some extreme outliers do, and neither tells you if your post actually did the thing you wanted it to do on any level. They work with extreme outliers (e.g. if you have 5x or 10x your normal views, or something) but then all they really are telling you is 'went viral.'
Yeah, that was actually me who wrote that post! Felt like I learned a ton about people and links from reading the responses to it.
Re: post metric limitations, I totally agree. Will look forward to your future thoughts on both of these topics should you touch on them in the roundup.
Three monitors, each 30", two vertical, new computer with high end stuff custom built by a friend on a generous budget, mainly. Also some other stuff but that's the biggest impact.
My post stats suggest that a little under half of my readers click at least one link. Perhaps you write more thoroughly, so that readers feel little need to supplement their understanding of your writing.
Ben Thompson has said that he likes to use a very current of-the-moment hook on his posts, from which vantage point he drives into a broader and more timeless theme or idea. This is a great trick which leverages ephemeral interest into more timeless ideas. So sometimes he will sit on an idea and have notes and outline for it, until there's a topical news story which illustrates it, and then he'll write the actual post.
my comment is in response to your saying "I wish I had structured them around and talked more about things that would endure, and less about things of the moment." So that's one way to thread that needle.
Surely if it registers a view it would register a click. Indeed, possibly more so. On Substack I sometimes see a click without a view - suggesting that uBlock disables the js but can’t disable the redirect for each link.
I like reading your posts. Regarding #4 Twitter, I don't tweet much, but it has dramatically declined for me since Musk took over. Tons of "For You" spam tweets, interesting people have stopped tweeting or are tweeting a lot less, and every time Musk does something everybody tweets their reaction to it and it fills up half my feed. I don't care about him, his companies, or his actions so that's all useless to me.
I'm kind of hanging on by a thread there, if you and just a few other people I follow stopped tweeting I'd probably just stop using it entirely.
Obvious question: Have you tried using Tweetdeck, as I keep saying? If you keep seeing 'for you' and other things you don't want, yeah that's terrible, but there's a way to make it stop.
No. I should probably try it. Actually the "For You" thing is recent, it just started doing that in the past couple weeks. Never saw that in the pre-Musk era. I guess its part of his pursuit of "engagement".
re: how many people go back to read things months or years later?
Some of us would like to, if we had a better way to find 'now what was that substack article about car seats again? I don't remember who wrote it, but I would like to re-read it now given that I just read <something else>?'
It may matter a lot when it comes to attracting new subscribers. I cannot be the only person who reads a bunch of articles to see what it is I would be getting if I subscribed.
The statistics say that view rates after a while are VERY low, with the possible exception of greatest hits like the car seat one.
I now while writing this grabbed a snapshot spreadsheet of everything, so I'll be able to do an over-time comparison after a while. Could be interesting.
1) Link click-through rate. I'll be frank, there's nearly zero possibility of me ever clicking on a link to TVTropes or YouTube; if you have to explain a joke, it's not funny anymore. (I find you humorous, but typically not from these add-ins.) Twitter links are similarly highly painful - I __do not__ enjoy that medium for acquiring information, and place high value on you parsing through that swamp so I don't have to. When your in-post summaries are good enough to begin with (especially embedding tweets as images), I see no reason to "check the primary source", as it were. Other classes of links tend to be outside my interest (MR) or things I already knew about (am cross-subscribed to Scott Alexander and Matt Yglesias). Part of why I like DWATV is that there's often enough meat on the post-bones that I don't have to click around to a buncha other links. Feature Not Bug.
2) Gaming. It's a shame this hurts the blog in an unambiguously bad cost-benefit trade; you __are__ still Kind Of A Big Deal as a famous gamer, and I particularly appreciate hearing your thoughts on the industry, whatever the manifestation. Shunting such posts to a wholly separate other Substack that I'd need to keep up with...is a Trivial Inconvenience perhaps too far. Understandable, if sad. I'd hope to at least have such posts mentioned in regular roundups, in such case.
3) Evergreen keyword posts/what's missing. One thing I've found frustrating while trying to share your work is that...you're The Covid Analysis Guy, but unless someone continuously reads the covid posts, it's hard to point them in a one-stop shop direction? There's your Omicron Model post, and the Long Covid Post Is Long post, but in many ways those are __too__ general...the sort of anxious people I'd want to reassuringly link to you, don't necessarily have the ability to derive from first covid principles the answers to more specific questions they want. Some of the highly-locally-temporally-relevant points have, indeed, been buried in random covid posts that are hard to track down for later reference. I think some sort of index post or wiki would serve well, to capture this value for posterity reading. It's some of your most important work, and at least what lured me to the blog to begin with. (Think the navigation post for "Immoral Mazes" sequence, for example.)
3x) Sub-point, I'd also agree that 2022 and arguably 2021 didn't see any Truly Classic Hits, along the caliber of old reference posts like Something Was Wrong, The Thing And The Symbolic Representation Of The Thing, Slack, Out To Get You...obviously those are exactly the kind of "short posts that take much longer to write" that you're cognizant of missing. (Twitter For Dummies comes closest, and I found it a great steelman, but continue to be skeptical; the Bounded Distrust game is very important, but increasingly expensive to play continuously.)
4) Subscription revenue. You don't have any paywalled posts or subscriber benefits; incentives are not aligned towards subscription. This should not be surprising? The entire reason I do is because you don't have an obvious Donate/Patreon button on WordPress, and I continue to feel like paying for the value you've given/continue to give to my regular reading. When you pay more for something, you tend to get more of it...and unlike some other bloggers who seem pathologically driven to write regardless of incentives, you seem much more potentially-responsive to direct monetary compensation for throughput. A significant chunk of Important Thing Happened I only hear about through your writing, and that's worth incentivizing...somehow. Substack subscription revenue is a messy lossy kludge towards that goal. (I honestly preferred reading on WordPress too, Substack's UI is atrocious.)
5) Reach. I think this is always going to be inherently limited, if you want to maintain the same quality standards. It's not __just__ about post length, although that's a nontrivial factor. Your writing style is...very...particular? Dense, complex, interwoven, assuming a huge deal of audience foreknowledge/shared assumptions. It's Tryhard Reading which is just gonna make a lotta people bounce hard, no matter how good or concise the content. If Public Intellectual is something you aspire towards, I think it's important to ask "which public?" and "how intellectual?", since those are frequently in tension. I do also think one comparative advantage you have is being a less-discovered gem. There's a __sane__ number of comments on posts, a minimum of spam or random "what is this" people wandering in, you don't face the same sort of conformity pressures that e.g. Scott does now that he's Made It Big. Better is better, but big isn't always better, along every axis!
Anyway, just my reactions to your introspection. Which I do look positively on, and I hope you're feeling more encouraged than discouraged on the margin. You do good work, in many ways unique work, and I think it'd be a particularly costly loss-to-others if your output decreased.
I would be extremely interested in a general analysis of Elon, especially his strategy / decision making. You've made some comments about his behavior in the past that got my attention, and this seems like a perfect combination of something I think is important to the world and that I'd really value your perspective on. Hopefully others agree, and thanks for all the hard work you've put into this blog already!
I’ve started reading you blog about a year ago, so might as well do a retro :)
First of all, thanks, your writing does add to my life! I get two kinds of stuff out of your posts:
* _slow news_ type of content: I am not a huge fan of news news, and I don’t Twitter, so your takes on what’s going on form a significant part of my “what is the current thing?” intake.
* bits of world modeling. To expand on this:
I have a draft post of my own, where I essentially ask to recommend me a book on how the world of humans work. Specifically, why humans are so good at solving coordination problems. I know a fair amount of economics, so things which are explainable by the market are somewhat clear to me. A lot of things are not markets (nation states, nationalism, in group / out group), I am still trying to find a master model to explain them. I got a lot of bits and pieces from various rationalist blogs (X is not about Y, blue-red tribalism, signaling, etc), but the unified model evades me. I want “microeconomics of politics” and “macroeconomics of politics” pair of books, and in your posts you always tantalizingly hint at some future post which would focus on long-term modeling :-)
Don’t see too much of that, to be honest, but I am picking more individual pieces of the puzzle (as an example, from the vaccine CA post, the idea that you can optimize a) total social surplus b) surplus of those who are worse of c) equality).
> links
You have a lot of links! I think they fall into three categories:
a) cultural references
b) sources for direct quotes (eg, you both embed and link tweets)
c) links to extra stuff which is not directly covered by your text.
I want to ignore a) as we seem to have little intersection culture-wise, I mostly ignore b) as I am willing to free ride on others fact-checking your quotes, and c) I am interested in. It often is hard to figure out whether a link is a, b or c without clicking on it. In my own writing, I now default to writing full urls (as opposed to linking random words), as the url helps the reader to understand at a glance what’s on the other side and whether it is worth a click.
Finally, while I really appreciate the content, I often fine that your writing style sometimes gets in a way of understanding. This is all supper fuzzy in my mind, so I am not sure I can give usefully actionable feedback, but the overall feeling is “oh dear, there’s new post by Zvi, gotta find a calm mental space to read that, as that’s work”. I think part of that is your writing being information-dense, which is great and much appreciated. But part I think is due to over-reliance on context, which I often miss. It’s sometimes like reading a math paper, where at some point you are like “ok, what the hell is M in this formula, did we define that 666 pages ago?”
As a specific example, in this very post, “paper is true” section made zero sense to me on the first read. It took me reading the whole post (and skimming through the post the post links to) that I got what is this about at all. What’s more, I did read Scott’s post which introduced the rock, so the context *was* in my head, it just wasn’t evoked by references.
For the gaming posts: Substack lets writers define and assign different post categories, and readers can then deselect categories for which they don't want to receive notifications. There might be a way to make a category opt-in instead of opt-out, so it doesn't surface by default but is there for those who want it. But if you want to build a dedicated gaming audience, a separate Substack or other site might be the way to go.
+1, I came down here to the comments to mention this Substack feature--I remember the Dispatch substack taking major advantage of Substack's subscription-within-subscription silo ability before they outgrew the platform and moved to their own website.
I think DWATV would massively benefit from "COVID," "Balsa," and "Gaming" categories that readers can choose to receive emails for or not.
Comments from someone who really likes your stuff: if it's on a topic I'm really into, it's fantastic. I read a lot of your long form Covid posts. They were best in class. For topics I'm less interested in, your posts are too long to read. Maybe this reflects poorly on my attention span, all the same though. Overall great stuff.
Your summaries are often great, and always at least good. I occasionally sub when I really enjoyed a particular article.
Consuming news about the world often wears down my patience and mood. I am happier with the less news I consume. I'm down to an average of ~20 minutes a day of intentional news consumption. Being on a website like Twitter that is literally a bunch of news sounds like pure torture to me. I will happily find other people who read Twitter and then summarize it for me. I think I am not alone in feeling this, and your substack is very useful to people like me. But also it is very hard to find you. I'm not sure if some form of advertising would pass cost benefit for you, but it might. Right now it seems like you are trying to sell gold nuggets right outside a gold mine, or selling fish in the harbor. You are close to where news is produced, but people who want to consume news summaries aren't hanging around where news is produced.
______
There was a topic a while back that I asked you about. Some trade group that liked the Jones act called the marginal revolution bloggers traitors for opposing the Jones act. You had said you were planning to write about it, but I didn't see it. Not sure if I missed it, or if you changed your mind. I did think it was important because it revealed that something you might see as low hanging fruit is actually fiercely defended and isn't low hanging fruit at all. Which might make Balsa far less effective if you've made the same mistake elsewhere.
________
You asked why people aren't more upset by student loans, I think most people I know get told the sob stories, and never hear the "milking the system for all it's worth" stories. Your post on the topic was helpful in that regard. I am personally less upset, because COVID restrictions had me more upset and more personally impacted. The government never has to put unfunded mandates on their books, so I think those often wind up costing way more. For example, things like the ADA have probably cost trillions, and have outright destroyed value when the costs of compliance are greater than any possible profits and things are simply shut down. What's another trillion in student loan debts when its all gonna get eaten by inflation in a few years? Hard to care about all of the fiscal irresponsibility at once.
A trillion here, a trillion there, pretty soon you're talking real money.
The Jones Act is a big deal, and actually Balsa's #1 planned target. The reason I didn't write about the incident is that I decided to wait and 'do it right' after I had done all the proper research, the way I did with the Dredge Act at minumum.
I like the idea of more focused posts, they seem like your best and most valuable work to me. Hot takes can be fun but are never going to be reliable. If you’re writing tweet replies they could just go on Twitter. I might make an exception for subjects where you have real expertise.
That reminds me, you wrote a lot about Ukraine this year and I never understood why. I don’t remember a single thing from those posts.
I’m not a fan of your dismissive style, though I can live with it. I think the colonoscopy episode adequately demonstrates that you’re not immune to overconfidence. More importantly, it never helps with persuasion so I’m uncomfortable sharing your posts with people who don’t already agree. I expect it’s born of understandable frustration, but still.
I don't find many of your gaming posts _directly_ interesting, but I greatly appreciate your modeling skills (and explaining them in writing) and the gaming posts are interesting for me modeling you (and your modeling). I don't personally mind them being on the same 'newsletter'/blog. It's easy enough (for me) to ignore them or save them to be read later. I can understand why lots of other people might feel differently, especially if their email habits/workflow are like most/many others.
I do often find myself a little confused by your vocabulary. I often can't tell what's 'poetic' versus a (technical) ' improper noun' versus your own terminology. (I think I asked you before about some concise source for understanding just the finance subset of these words/phrases and ended up with a friend lending my their 'series 7' study guides as my best option.)
As for links, the 'meme' links have seemed annoyingly superfluous – an inline or embedded version (or a static image instead of a video or GIF) would work better, but they're visually identical to all the other links so there's a bit of a 'lemons market' effect. The Twitter links are also mostly superfluous as you're mostly screen-shotting the relevant tweets anyways. Other links seem more like citations than even a weak recommendation to be read. It's definitely not been obvious to me that you were recommending what you're linking (except, obviously, when you recommend them explicitly). It also doesn't help that there's often a lot of links in what are often pretty (if not very) long posts.
Really liked your takeaways, particularly #6 and #9. Curious what you see as the main limitations to individual post metrics and where you do see value there. Also why does no one click on links? Seems kind of weird to me because posts that center around sharing links seem quite popular with readers, yet even on those click rates are low.
I don't know the technical details of how link clicking is tracked but I wonder how reliable it is (does it work with ad/tracker blockers that are increasingly common? Privacy focused browsers?).
I'm either a very abnormal reader or else the tracking is poor, because I'd guess that I click half our more of the links on nearly every post.
I just assume that not everyone is as voracious for new information. Normal people don’t click link after link. I find it reassuring to see a link but rarely check that it backs up a claim, unless I am really into a subject.
I think there is definitely something to this. People vary in the level of resolution they demand for claims.
https://www.econlib.org/library/columns/y2022/klingemotions.html
Still I would think Zvi's readers would be pretty high resolution thinkers. When I put "On Bounded Distrust" into a word counter it was about 34 pages single-spaced!
This is my feeling as well. I don’t click links unless I’m particularly interested and/or skeptical, but it’s reassuring that it’s possible to check the source of information.
Tyler linked to a post thinking about why no one clicks on links, and also a bunch of people answered with their takes. I'll write a bit more about it in the January roundup, probably. It's kind of like, the link needs to exist so people know it exists, but given we know it's there no need to click.
I kind of think a bunch of TV wants to work that way and can't? As in 'now we are going to show the awkward conversation because we have to, but it's not fun and doesn't contain unexpected info so it would kind of be better if you knew to skip it.' Ya know?
The limitations on post metrics are basically that clicks tell you a lot more about whether certain people linked to it and timing queues and not much about whether people actually read or approved of it, and likes/comments are about what some extreme outliers do, and neither tells you if your post actually did the thing you wanted it to do on any level. They work with extreme outliers (e.g. if you have 5x or 10x your normal views, or something) but then all they really are telling you is 'went viral.'
Yeah, that was actually me who wrote that post! Felt like I learned a ton about people and links from reading the responses to it.
Re: post metric limitations, I totally agree. Will look forward to your future thoughts on both of these topics should you touch on them in the roundup.
I think you should be really proud of all of this Zvi!
care to elaborate more on the "hardware upgrade" thing?
Three monitors, each 30", two vertical, new computer with high end stuff custom built by a friend on a generous budget, mainly. Also some other stuff but that's the biggest impact.
My post stats suggest that a little under half of my readers click at least one link. Perhaps you write more thoroughly, so that readers feel little need to supplement their understanding of your writing.
also i think doing gaming on a separate vertical and just putting links to it in your main feed (or like a paragraph excerpt) makes alot of sense
Ben Thompson has said that he likes to use a very current of-the-moment hook on his posts, from which vantage point he drives into a broader and more timeless theme or idea. This is a great trick which leverages ephemeral interest into more timeless ideas. So sometimes he will sit on an idea and have notes and outline for it, until there's a topical news story which illustrates it, and then he'll write the actual post.
my comment is in response to your saying "I wish I had structured them around and talked more about things that would endure, and less about things of the moment." So that's one way to thread that needle.
I click on links plenty, but also have uBlock dialled to the maxx, so maybe it's that.
Surely if it registers a view it would register a click. Indeed, possibly more so. On Substack I sometimes see a click without a view - suggesting that uBlock disables the js but can’t disable the redirect for each link.
I like reading your posts. Regarding #4 Twitter, I don't tweet much, but it has dramatically declined for me since Musk took over. Tons of "For You" spam tweets, interesting people have stopped tweeting or are tweeting a lot less, and every time Musk does something everybody tweets their reaction to it and it fills up half my feed. I don't care about him, his companies, or his actions so that's all useless to me.
I'm kind of hanging on by a thread there, if you and just a few other people I follow stopped tweeting I'd probably just stop using it entirely.
Obvious question: Have you tried using Tweetdeck, as I keep saying? If you keep seeing 'for you' and other things you don't want, yeah that's terrible, but there's a way to make it stop.
No. I should probably try it. Actually the "For You" thing is recent, it just started doing that in the past couple weeks. Never saw that in the pre-Musk era. I guess its part of his pursuit of "engagement".
"For You" is a rebrand of what used to be called "Home."
re: how many people go back to read things months or years later?
Some of us would like to, if we had a better way to find 'now what was that substack article about car seats again? I don't remember who wrote it, but I would like to re-read it now given that I just read <something else>?'
It may matter a lot when it comes to attracting new subscribers. I cannot be the only person who reads a bunch of articles to see what it is I would be getting if I subscribed.
The statistics say that view rates after a while are VERY low, with the possible exception of greatest hits like the car seat one.
I now while writing this grabbed a snapshot spreadsheet of everything, so I'll be able to do an over-time comparison after a while. Could be interesting.
Thoughts from a happy paid subscriber:
1) Link click-through rate. I'll be frank, there's nearly zero possibility of me ever clicking on a link to TVTropes or YouTube; if you have to explain a joke, it's not funny anymore. (I find you humorous, but typically not from these add-ins.) Twitter links are similarly highly painful - I __do not__ enjoy that medium for acquiring information, and place high value on you parsing through that swamp so I don't have to. When your in-post summaries are good enough to begin with (especially embedding tweets as images), I see no reason to "check the primary source", as it were. Other classes of links tend to be outside my interest (MR) or things I already knew about (am cross-subscribed to Scott Alexander and Matt Yglesias). Part of why I like DWATV is that there's often enough meat on the post-bones that I don't have to click around to a buncha other links. Feature Not Bug.
2) Gaming. It's a shame this hurts the blog in an unambiguously bad cost-benefit trade; you __are__ still Kind Of A Big Deal as a famous gamer, and I particularly appreciate hearing your thoughts on the industry, whatever the manifestation. Shunting such posts to a wholly separate other Substack that I'd need to keep up with...is a Trivial Inconvenience perhaps too far. Understandable, if sad. I'd hope to at least have such posts mentioned in regular roundups, in such case.
3) Evergreen keyword posts/what's missing. One thing I've found frustrating while trying to share your work is that...you're The Covid Analysis Guy, but unless someone continuously reads the covid posts, it's hard to point them in a one-stop shop direction? There's your Omicron Model post, and the Long Covid Post Is Long post, but in many ways those are __too__ general...the sort of anxious people I'd want to reassuringly link to you, don't necessarily have the ability to derive from first covid principles the answers to more specific questions they want. Some of the highly-locally-temporally-relevant points have, indeed, been buried in random covid posts that are hard to track down for later reference. I think some sort of index post or wiki would serve well, to capture this value for posterity reading. It's some of your most important work, and at least what lured me to the blog to begin with. (Think the navigation post for "Immoral Mazes" sequence, for example.)
3x) Sub-point, I'd also agree that 2022 and arguably 2021 didn't see any Truly Classic Hits, along the caliber of old reference posts like Something Was Wrong, The Thing And The Symbolic Representation Of The Thing, Slack, Out To Get You...obviously those are exactly the kind of "short posts that take much longer to write" that you're cognizant of missing. (Twitter For Dummies comes closest, and I found it a great steelman, but continue to be skeptical; the Bounded Distrust game is very important, but increasingly expensive to play continuously.)
4) Subscription revenue. You don't have any paywalled posts or subscriber benefits; incentives are not aligned towards subscription. This should not be surprising? The entire reason I do is because you don't have an obvious Donate/Patreon button on WordPress, and I continue to feel like paying for the value you've given/continue to give to my regular reading. When you pay more for something, you tend to get more of it...and unlike some other bloggers who seem pathologically driven to write regardless of incentives, you seem much more potentially-responsive to direct monetary compensation for throughput. A significant chunk of Important Thing Happened I only hear about through your writing, and that's worth incentivizing...somehow. Substack subscription revenue is a messy lossy kludge towards that goal. (I honestly preferred reading on WordPress too, Substack's UI is atrocious.)
5) Reach. I think this is always going to be inherently limited, if you want to maintain the same quality standards. It's not __just__ about post length, although that's a nontrivial factor. Your writing style is...very...particular? Dense, complex, interwoven, assuming a huge deal of audience foreknowledge/shared assumptions. It's Tryhard Reading which is just gonna make a lotta people bounce hard, no matter how good or concise the content. If Public Intellectual is something you aspire towards, I think it's important to ask "which public?" and "how intellectual?", since those are frequently in tension. I do also think one comparative advantage you have is being a less-discovered gem. There's a __sane__ number of comments on posts, a minimum of spam or random "what is this" people wandering in, you don't face the same sort of conformity pressures that e.g. Scott does now that he's Made It Big. Better is better, but big isn't always better, along every axis!
Anyway, just my reactions to your introspection. Which I do look positively on, and I hope you're feeling more encouraged than discouraged on the margin. You do good work, in many ways unique work, and I think it'd be a particularly costly loss-to-others if your output decreased.
I would be extremely interested in a general analysis of Elon, especially his strategy / decision making. You've made some comments about his behavior in the past that got my attention, and this seems like a perfect combination of something I think is important to the world and that I'd really value your perspective on. Hopefully others agree, and thanks for all the hard work you've put into this blog already!
I’ve started reading you blog about a year ago, so might as well do a retro :)
First of all, thanks, your writing does add to my life! I get two kinds of stuff out of your posts:
* _slow news_ type of content: I am not a huge fan of news news, and I don’t Twitter, so your takes on what’s going on form a significant part of my “what is the current thing?” intake.
* bits of world modeling. To expand on this:
I have a draft post of my own, where I essentially ask to recommend me a book on how the world of humans work. Specifically, why humans are so good at solving coordination problems. I know a fair amount of economics, so things which are explainable by the market are somewhat clear to me. A lot of things are not markets (nation states, nationalism, in group / out group), I am still trying to find a master model to explain them. I got a lot of bits and pieces from various rationalist blogs (X is not about Y, blue-red tribalism, signaling, etc), but the unified model evades me. I want “microeconomics of politics” and “macroeconomics of politics” pair of books, and in your posts you always tantalizingly hint at some future post which would focus on long-term modeling :-)
Don’t see too much of that, to be honest, but I am picking more individual pieces of the puzzle (as an example, from the vaccine CA post, the idea that you can optimize a) total social surplus b) surplus of those who are worse of c) equality).
> links
You have a lot of links! I think they fall into three categories:
a) cultural references
b) sources for direct quotes (eg, you both embed and link tweets)
c) links to extra stuff which is not directly covered by your text.
I want to ignore a) as we seem to have little intersection culture-wise, I mostly ignore b) as I am willing to free ride on others fact-checking your quotes, and c) I am interested in. It often is hard to figure out whether a link is a, b or c without clicking on it. In my own writing, I now default to writing full urls (as opposed to linking random words), as the url helps the reader to understand at a glance what’s on the other side and whether it is worth a click.
Finally, while I really appreciate the content, I often fine that your writing style sometimes gets in a way of understanding. This is all supper fuzzy in my mind, so I am not sure I can give usefully actionable feedback, but the overall feeling is “oh dear, there’s new post by Zvi, gotta find a calm mental space to read that, as that’s work”. I think part of that is your writing being information-dense, which is great and much appreciated. But part I think is due to over-reliance on context, which I often miss. It’s sometimes like reading a math paper, where at some point you are like “ok, what the hell is M in this formula, did we define that 666 pages ago?”
As a specific example, in this very post, “paper is true” section made zero sense to me on the first read. It took me reading the whole post (and skimming through the post the post links to) that I got what is this about at all. What’s more, I did read Scott’s post which introduced the rock, so the context *was* in my head, it just wasn’t evoked by references.
Apropos of nothing, thanks for your work on rust analyzer!
For the gaming posts: Substack lets writers define and assign different post categories, and readers can then deselect categories for which they don't want to receive notifications. There might be a way to make a category opt-in instead of opt-out, so it doesn't surface by default but is there for those who want it. But if you want to build a dedicated gaming audience, a separate Substack or other site might be the way to go.
+1, I came down here to the comments to mention this Substack feature--I remember the Dispatch substack taking major advantage of Substack's subscription-within-subscription silo ability before they outgrew the platform and moved to their own website.
I think DWATV would massively benefit from "COVID," "Balsa," and "Gaming" categories that readers can choose to receive emails for or not.
Comments from someone who really likes your stuff: if it's on a topic I'm really into, it's fantastic. I read a lot of your long form Covid posts. They were best in class. For topics I'm less interested in, your posts are too long to read. Maybe this reflects poorly on my attention span, all the same though. Overall great stuff.
Your summaries are often great, and always at least good. I occasionally sub when I really enjoyed a particular article.
Consuming news about the world often wears down my patience and mood. I am happier with the less news I consume. I'm down to an average of ~20 minutes a day of intentional news consumption. Being on a website like Twitter that is literally a bunch of news sounds like pure torture to me. I will happily find other people who read Twitter and then summarize it for me. I think I am not alone in feeling this, and your substack is very useful to people like me. But also it is very hard to find you. I'm not sure if some form of advertising would pass cost benefit for you, but it might. Right now it seems like you are trying to sell gold nuggets right outside a gold mine, or selling fish in the harbor. You are close to where news is produced, but people who want to consume news summaries aren't hanging around where news is produced.
______
There was a topic a while back that I asked you about. Some trade group that liked the Jones act called the marginal revolution bloggers traitors for opposing the Jones act. You had said you were planning to write about it, but I didn't see it. Not sure if I missed it, or if you changed your mind. I did think it was important because it revealed that something you might see as low hanging fruit is actually fiercely defended and isn't low hanging fruit at all. Which might make Balsa far less effective if you've made the same mistake elsewhere.
________
You asked why people aren't more upset by student loans, I think most people I know get told the sob stories, and never hear the "milking the system for all it's worth" stories. Your post on the topic was helpful in that regard. I am personally less upset, because COVID restrictions had me more upset and more personally impacted. The government never has to put unfunded mandates on their books, so I think those often wind up costing way more. For example, things like the ADA have probably cost trillions, and have outright destroyed value when the costs of compliance are greater than any possible profits and things are simply shut down. What's another trillion in student loan debts when its all gonna get eaten by inflation in a few years? Hard to care about all of the fiscal irresponsibility at once.
A trillion here, a trillion there, pretty soon you're talking real money.
The Jones Act is a big deal, and actually Balsa's #1 planned target. The reason I didn't write about the incident is that I decided to wait and 'do it right' after I had done all the proper research, the way I did with the Dredge Act at minumum.
I like the idea of more focused posts, they seem like your best and most valuable work to me. Hot takes can be fun but are never going to be reliable. If you’re writing tweet replies they could just go on Twitter. I might make an exception for subjects where you have real expertise.
That reminds me, you wrote a lot about Ukraine this year and I never understood why. I don’t remember a single thing from those posts.
I’m not a fan of your dismissive style, though I can live with it. I think the colonoscopy episode adequately demonstrates that you’re not immune to overconfidence. More importantly, it never helps with persuasion so I’m uncomfortable sharing your posts with people who don’t already agree. I expect it’s born of understandable frustration, but still.
I don't find many of your gaming posts _directly_ interesting, but I greatly appreciate your modeling skills (and explaining them in writing) and the gaming posts are interesting for me modeling you (and your modeling). I don't personally mind them being on the same 'newsletter'/blog. It's easy enough (for me) to ignore them or save them to be read later. I can understand why lots of other people might feel differently, especially if their email habits/workflow are like most/many others.
I do often find myself a little confused by your vocabulary. I often can't tell what's 'poetic' versus a (technical) ' improper noun' versus your own terminology. (I think I asked you before about some concise source for understanding just the finance subset of these words/phrases and ended up with a friend lending my their 'series 7' study guides as my best option.)
As for links, the 'meme' links have seemed annoyingly superfluous – an inline or embedded version (or a static image instead of a video or GIF) would work better, but they're visually identical to all the other links so there's a bit of a 'lemons market' effect. The Twitter links are also mostly superfluous as you're mostly screen-shotting the relevant tweets anyways. Other links seem more like citations than even a weak recommendation to be read. It's definitely not been obvious to me that you were recommending what you're linking (except, obviously, when you recommend them explicitly). It also doesn't help that there's often a lot of links in what are often pretty (if not very) long posts.