(1200 words, 4 minutes to read)
This series is an attempt to pay homage to some particularly amazing blog posts I’ve read, which had a profound impact on me. Here are the first two and the third one.
In no particular order:
4. Scott Alexander’s “Searching for One-Sided Tradeoffs” on Slate Star Codex
Scott centers this post around an example of an admissions officer at a medium-tier university. If you think about it, they’re in a tough situation: the best students won’t even apply to this medium-tier university, and the worst students will never pass the simplest filter. So it seems likely that student quality will be narrowly distributed around some mediocre level. Does it even matter very much who they choose?
It does, because “quality” isn’t one thing; it’s a combination of numerous factors, and so if you sacrifice one you can excel at another. If all the students in the admissions list have roughly the same SAT score of 1200, you can build an amazing math department by admitting students who got low verbal scores. And you might find that students below average in other qualities, like social skills or sports ability or conscientiousness or whatever, will have SAT scores higher than 1200; this is, again, because “quality” consists in some combination of all of these things. You can make this trade if that’s what you want.
Scott points out three ways the admissions officer can succeed in his job: 1) insider trading, which is basically just being really smart and hard-working and mining the right tail of the distribution better than the competition; 2) bias-compensation, which is where you aren’t bothered by some quality that others see as a downside; and 3) comparative advantage, where you have some quality that makes you more attractive along some specific axis, so you can attract those who value that quality particularly highly. (This is sort of the flip side of bias-compensation: if you are high in trait X and only a select subgroup A prefers X, then it is your comparative advantage, and people in A can win by finding you through bias-compensation.)
To succeed, our admissions officer can: (1) work really hard to mine the right tail; or (2) benefit from some built-in bias of other schools (e.g. if other admissions departments are racist then ours can win by finding highly-qualified minorities); or (3) build their department as (for example) “the best algebraic topology department in the country” by sacrificing other qualities in favor of this one narrow math subfield, because over time this will attract the very best algebraic topologists and make the department more than mediocre.
\\
The admissions officer is a good example, but the second half of the post is about how to apply this to real life, which is of course why I care about it so much. In real life, “insider trading” = “being really smart and hard-working” is something we’re all trying to do all the time, but in spite of our unflappable human confidence, 50% of us are worse than the median person at that. But the other two strategies are worth thinking about seriously too. Here are two examples that Scott doesn’t cover, but that come to my mind readily: relationships and jobs.
In (romantic) relationships, people talk about how someone they like might be “out of their league”, and often what they’re talking about is physical attractiveness. But “quality” in relationships is more than physical, obviously varies wildly from person to person, and is about as complicated and multi-faceted as something can be. You would do well to employ Scott’s admissions strategies here. In searching for a partner, one way to bias-compensate is to have non-majority preferences, e.g. if there is a bias against people with red hair, you with your redhead fetish have an advantage. Or, if you are the redhead, you have a comparative advantage with some relatively high-“quality” mates who prefer red hair. Now replace “red hair” in the example with “a passion for death metal” or “a bookshelf full of manga” or whatever particular trait makes you stand out. This could be your comparative advantage! Let it show! Jacob at Putanumonit notes research suggesting that “…controlling for average rating of attractiveness, the variance in ratings correlated strongly with more people wanting to date you.” It’s better to get 5/5 stars or 1/5 stars than to get 4/5 stars. Some people giving you a 1/5 means that you aren’t some people’s taste on some dimension, but this very fact may be what gives you an advantage with some other subgroup. Sometimes it’s a quirk of your partner, that others see as a bug, which to you is a feature that is part of why you love them so much. Or, maybe you personally sacrifice quality X because you care much more about quality Y, and you would do well to find a mate that has the same preferences.
Jobs are not mono-faceted either. People often talk about it as though the salary is the top priority, and maybe for some it is, but if you’re willing to accept a pay cut a lot of other advantages become available to you. Maybe your job pays poorly but it’s interesting and you love the daily work; or maybe it’s not interesting but you love the people working there; or maybe you can’t stand your coworkers but it allows you to travel; and so on. For me, I know that there are jobs I could be doing right now that would pay me 50-100% more than what I’m doing now; academia is not known for high stipends. But that trades off against the fact that my work is interesting and I get to travel often (that is, I used to in the BC era (“before COVID”)). I also have a work-life flexibility that almost nobody else has: I can go to work at 8AM or 11AM or 2PM and nobody bats an eye, provided I am doing interesting research on a reasonable timescale. My current job required me to relocate to Japan, which many of my colleagues would see as a bug, but to me is a feature. So rather than taking a pay cut, I effectively got a pay raise for having the unusual preference of moving to a strange, far-away country.
The effect of Scott’s blog post on me was to see my decision-making on important choices in a new context, which has allowed me to think them through more clearly. To boil it down, an actionable strategy Scott advocates for here is to not look for things that have no downsides; rather, find things with “downsides” that you can handle better than others, or if you’re lucky, that you positively prefer. Sacrifice some qualities for others that you care about more. It’s how you do better without, in some broad general sense, having to be better.