In 2012 (imagine some wavy flashback squiggles here) I made a mailing list called Science Friday and each week for a year I emailed a question from an “Are You Scientifically Literate?” quiz and we talked about them and it was great.
I thought it could be fun to do the same thing with PutANumOnIt’s “100 Ways To Live Better”.
So I polled the daily beemail readers (the scale was “dear lord those sound idiotic” to “omg i eat that stuff up like cookie monster”) and they were into it so it became a thing.
And that thing is now also here. It started March 2, 2020, right before the pandemic lockdown, and went for exactly one year. Enjoy!
Any life advice that isn’t given to you personally is not designed to be followed to the letter. Try to resonate with the philosophy that generates it instead. Remember that directional advice (e.g., “be more…”) may need to be reversed before consumption.
That Scott Alexander post is a classic and a great attenuator / sanity checker on internet advice generally (even though it might also apply to itself!).
Verdict: Great meta advice A++ would advise again.
Peanut Gallery:
Collect feedback from everybody. Play games with close friends where you have to give each other constructive criticism and ways to improve. Collect anonymous feedback from internet strangers on Admonymous.
Verdict: Sure, I’ll start: admonymous.co/dreeves.
Write that comment. You know the saying about letting people suspect you’re dumb rather than opening your mouth and removing all doubt? F**k that. We know you’re dumb. You get less dumb by saying things and getting feedback.
Verdict: I feel like there’s a stark dichotomy. Either you should be way less shy and chime in way more, or you’re the opposite and should constantly remind yourself to pause and reflect. So I guess this is good advice, subject to madhack #1 (consider reversing it). I do see a lot of brilliant things that y’all say in reply to these emails that it can be hard to convince you to repeat in the forum.
In particular, it would be really great to get more thoughts from you in the coronavirus thread. But, ok, I get it, you’re shy. So we can stick with the protocol where you reply to this email and if it’s particularly insightful I can twist your arm to repeat it publicly (or get your permission to quote you at least).
Peanut gallery:
Learn some improv, at least to get the basic gist of it. Take a class or read Impro (book by Keith Johnstone). Improv mindset is a great way to approach many social situations including most interactions on the internet. A good comment/reply often starts with “yes, and”.
Verdict: Yes to the “yes, and” mentality. And I’ve been hearing about that book for years in the context of figuring out how social status works (even if you have no interest in improv itself) so it’s probably pretty great.
That’s the opposite of good improv. You think that the categories in this [list] are arbitrary? A piece of advice doesn’t apply to your special situation? You’re probably right, but writing this in a comment will just make readers annoyed and make you frustrated when nobody responds.
Verdict: I originally disagreed in my reaction in the daily beemail because I wanted to encourage responses but pointed out that if you’re posting publicly that may be another story. Maybe a better version of this one is “don’t derail discussions”. Derailing is for Beeminder graphs, which is totally ok, as I’ve shamelessly argued on the blog.
Peanut Gallery:
There are more great podcasts than you’ll ever have the time to listen to. [Jacob links to 11 he likes.] If it sucks after 10 minutes, skip half an hour ahead. Still boring? Delete and move on. Obviously, do the same for books.
Verdict: I’m on board. I suspect I do too much slogging through things out of irrational completionism or something. This is often related to the sunk cost fallacy which I have an ancient article about that I think is still valuable, especially for people who are all about not falling for the fallacy. I just realized a cuter title for it would’ve been “The Sunk Cost Fallacy Fallacy” though I see I was not the first to think of that. Anyway, my old article has a cosmo-style quiz and everything.
Peanut Gallery:
Free will. The anthropic principle. Solipsism. The simulation hypothesis. Moral realism. They’re fun to argue about through the night but don’t judge anyone too much based on the positions they take and don’t treat any of them too seriously as guides to actually living your life. It should all add up to normalcy in the end.
(Knowing that I know about Betteridge’s Law of Headlines will hopefully attenuate the wrath of my friends who are philosophers here.)
Verdict: Either banal or heretical. On the banality side, yes, treat everyone as operating in good faith and appreciate that when smart people have an opinion that feels infuriatingly wrong, there’s probably a fundamental piece you’re missing. And it’s probably an unknown-unknown, so your rightness feels unassailable, but isn’t. (Except when it is, of course.) Same for politics and religion. And on the heresy side, well, the people who take these things seriously as guides to actually living their lives are the most impressive people. (Except when they infuriatingly conclude that it’s wrong to have babies or something.)
But to defend this madhack, I think there’s something profound in “it should all add up to normality in the end”. The rationality folks call it Egan’s Law:
Apples didn’t stop falling when General Relativity supplanted Newtonian mechanics. The purpose of a theory is to add up to observed reality, rather than something else. Quantum mechanics is not some strange alternative universe, told as a story to delight us with its counterintuitiveness; quantum mechanics is where we have always lived. Science sets out to answer the question “What adds up to normality?” and the answer turns out to be “Quantum mechanics adds up to normality.”
A weaker extension of this principle applies to ethical and metaethical debates, which generally ought to end up explaining why you shouldn’t eat babies, rather than why you should.
Find a medium of expression and express yourself publicly every day for three months. If you’re good with words, write 100 tweets. An artist — post 100 sketches on Instagram. Music/dance person — 100 TikToks.
Verdict: Yes, and beemind it! I beemind sending these beemails, for example (which also means beeminding my way through this list). And tweeting at twitter.com/bmndr. And the blog. And beeminder.com/changelog counts, I think.
If you can’t stand social media, go with an email list. In fact, here’s what I recommend: email everyone in your address book individually (important that it be individually, for non-obvious reasons I won’t go into here) saying you’re starting a newsletter / art exhibit / whatever kind of thing and to reply “sure” if they’d like to receive it. Get enough “sure”s — even if just a handful — that it will feel like expressing yourself publicly. Then just do that consistently. (And if you tell your audience that your consistency is thanks to Beeminder, you’ll be our new best friend!)
Peanut Gallery:
crickets
Either this was a dumb idea or you’re all chicken.
chicken noises
UPDATE: My chicken noises may have worked on Clive Freeman:
Tell a bad joke or a pun as soon as you think of it, even if it’s just to your exasperated spouse or coworker. It takes 20 bad jokes to think of a single good one, and you only start making good jokes once you remove the unconscious filter stifling your generative brain.
Verdict: I’m phoning this one in. That “Babble and Prune” article sounds potentially interesting but… that’s all I’ve got. I’m refusing to come down as either pro or anti on bad-joke telling. Also I’m a dad so my opinion on this question is meaningless anyway!
This did lead to an unrelated meta madhack though. Normally I’d have thought harder and actually read that “Babble and Prune” article. But when I was due to send out the madhack, I had an important Beeminder bug to fix and my partner, Bee, suggested I strategically procrastinate on this. So I took her advice and I guess it worked like a charm. Beeminder made me publish this despite having no interesting commentary on this madhack. Definitely a better outcome for Beeminder than me spending half the day down a Wikipedia rabbit hole learning about language acquisition in babies or whatever that LessWrong article would’ve led to!
PS: Whaddya know, this turned into one of the best, simplest madhacks of all! I’ve now updated the title from “Tell Bad Jokes” to “Strategically Procrastinate”. It’s a classic commitment device to cap the amount of time you spend on something: don’t start till that much time before the deadline!
If you can’t give it up completely, try to constrain the bandwidth of how much you hear about politics. Don’t start your day with the front page of the Times. Unfollow anyone whose posts are more than 20% about politics or the outrage du jour. And don’t jump into online arguments, it’s vice masquerading as virtue.
Verdict: Oh my yes. Better to read books about history than news articles about the present. I think Aaron Swartz was the first to convince me of this.
And I have more advice! When you can’t bear the temptation to find out the details of a news item everyone’s talking about, restrict your reading about it to Wikipedia. It does a great job of filtering out the speculation and sticking to facts. And, amazingly, when news is happening, the relevant Wikipedia page gets updated by the minute as the story unfolds.
That was actually one of my gratitude entries when I tried a gratitude journal: “You can’t use the world’s most comprehensive encyclopedia as your Slow News source because it’s pretty much real-time.”
Peanut Gallery:
Lawrence Evalyn: “I cannot think of a single life change that has been as good for me as totally rejecting the role of ‘news’ in my life. I definitely get by via herd immunity, though: all my friends know I don’t follow the news, so they tell me about things that might genuinely interest or impact me… like the existence of COVID-19.”
Clive Freeman: “Definitely agree with this! I try and sip the daily news through a tiny straw — the Wikipedia current events portal. It’s brilliantly non-partisan, and entirely global, reducing UK/US news to a couple of two-liners, alongside a dozen similar extracts ranging across the rest of the world. I’ve no idea how items are selected, I just know that if it’s really important it’ll show up, and meantime it reminds me that the rest of the world is much larger and more interesting than my particular corner.”
PS: One more related link, David Hainemeier Hansson, of Ruby on Rails fame, reflecting on his decision to stop following the news ten years ago.
Binge a show/video game for a couple of weeks, then take a break from TV for a couple of weeks. Trying to limit yourself to an hour a day is less fun and more addictive.
Verdict: No no no. First of all, the verb “binge” is a pretty good clue that this isn’t healthy! But, ok, maybe you don’t mean it in an unhealthy way. Live a little once in a while, etc. Fine. But clearly there’s an addictive element if the concept of binging is even on the table. So having some insurance against that is only prudent.
Can you tell where I’m going with this? If you want to occasionally binge a TV show, that’s a perfectly reasonable choice… if you make it dispassionately. Pick the average number of hours per week you’re ok devoting to TV / video games / whatever and beemind it. If you agree with Jacob that binging for a couple weeks and then abstaining for a couple weeks is more fun then great, but abstain first to build up safety buffer. And let Beeminder keep you honest.
I claim it’s even more fun that way because you’ve actually earned the binge. Enjoy it guilt-free!
Peanut Gallery:
(Moved to the Beeminder forum.)
Should you watch that movie / play that game / read that book? Use the ratio:
([# who rated it 5/5] + [# who rated it 1/5]) / [# who rated it 3/5].
This doesn’t apply to everything, but it applies to many things, including media. There are too many options out there to waste time on mediocrity, and everything great will be divisive.
Verdict: Um, sure, endorsed! Beeminder is plenty divisive and it’s super great, right?
Peanut Gallery: Holy cow, Clive Freeman was inspired by this madhack to write code to test this! His conclusion: directionally interesting, but probably not actually practical without reading reviews, etc. And a conclusion that’s very unsuprising in retrospect: the books (Clive tested this with books) that score highest by Jacob’s divisiveness metric are those with a political or religious axe to grind and that split people along ideological lines.
Unless one of them is your friend or boss, you should spend 100x less time thinking and talking about billionaires than you currently do.
Verdict: Shrug! I guess it’s plausible I agree with this on the grounds that the question itself seems worthy of roughly zero thought. Are people too obsessed with billionaires? Wait, don’t answer that, turns out I don’t care. The things I’m obsessed with are utterly ridiculous to most humans so if you want to obsess about billionaires or other celebrities, as I hear people do, who am I to judge? Maybe that’s my counter-advice on this one: Don’t be ashamed of your obsessions!
Peanut Gallery:
Facebook is for event invites only, not for scrolling. The people you met offline are not going to be the people posting the best stuff online, so the timeline content is worse than what you’d get on Twitter/Reddit/blogs. And the algorithm is designed to fuck with your brain.
Verdict: Yes. Except I’m not totally sure how else to see adorable pictures of my friends’ kids, which is my one reason to care about Facebook. (Apparently some people hate seeing baby pictures on Facebook so I should clarify that I’m definitely not being sarcastic. Also who are these misanthropes who don’t want to see their friends’ baby pictures??)
But, yes, scrolling through Facebook’s timeline, seeing whatever subset of your friends’ posts Facebook decides to show you (especially if you friend anyone-and-everyone you’ve vaguely heard of), feels kind of like being manipulated.
Peanut Gallery:
Half a dozen people saying some variant of “yes, kill it with fire”. (Obviously a biased sample.) Recommendations include News Feed Eradicator and Socialfixer. (Thanks to Paul Fenwick for first recommending the latter.)
Don’t keep watching a bad TV show just because your friends are talking about it; it’s a terrible time trade-off. You can read a recap or even better — bring up richer topics of conversations. And don’t pay money for bad movies just because “everyone is watching them”. Doing so is defecting against your friends since they’ll now have to watch it to not feel left out.
Verdict: Seems obvious enough. Sunk cost fallacy! And I like the game theoretic take on breaking out of the bad equilibrium where everyone watches a dumb thing mainly because everyone else is watching it. See also madhack #6.
Peanut Gallery:
Habits are reinforced by your habitual environment. That’s a big part of why retreats work: they take you away from your usual surroundings and people. If you want to start meditating, doing pushups, intermittent fasting, etc, try starting on a vacation where the new circumstances make it easier to integrate new habits.
Verdict: Thumbs up. Engineering your environment is a solid, classic lifehack. Beeminder (here I go) is an especially powerful and flexible way to do that. Bringing long-term consequences near and all that jazz. You know what I’m talking about.
Peanut Gallery:
Are you really going to give up on expressing yourself, learning from mistakes, attracting like-minded people, building a reputation, and changing the world because someone may someday try to cancel you? They can smell the fear, you know.
Verdict: I’m afraid of giving one. Wait, no, expressing that fear is totally a political stance! I can’t win here. Ok, but seriously, I don’t actually feel particularly afraid and I guess I agree with this madhack. (Although there seems to be a lot between the lines — fear-smelling? — that’s way too nuanced too opine on in anything short of an essay.)
I do have a million opinions but immodestly think I’m exceptionally good at knowing when not to have an opinion and embracing uncertainty. It’s really nice to have trusted friends with whom you can try out tentative opinions in a way that would feel scary on the open internet. I don’t know if that’s because of cancel culture or whatever or if that’s just always true.
So as for opinions, since Jacob is making me feel like a chicken if I end my commentary there…
On the most topical topic I can think of at this moment, I really like these proposals for police reform: “A Comprehensive Reboot Of Law Enforcement”.
Weirdly, I can’t think of any controversial opinions I have about COVID-19 right now [June 2020]. A vaccine would be nice.
My strongest political opinion is that we need government action on climate change, starting with carbon pricing. Also I’m pro-nuclear.
Is that enough opinion-expressing to satisfy this madhack? Probably not, but I’m now realizing that I have another madhack that I think overrides this one: “Keep your identity small”.
Peanut Gallery:
You just read 1000 words. Close your eyes and count to 10 to break the dopamine loop and make sure that reading a listicle is really the thing you want to be doing most right now. If not, this post will still be here when you get back.
Verdict: True but mostly useless. Not just useless in the sense that I’m magnanimously spreading these out for you over a [year] but even if you are binging these (see madhack #11) the admonition to take a deep breath and ask yourself if this is what you want to be doing misses the point.
It’s like asking how procrastination can be a problem; if you know what you should be doing then you should just do it. (Plenty of lucky people are in fact baffled by how anyone else suffers from procrastination. Beeminder seems especially absurd to those people!) Point being, it doesn’t occur to you to step back and reassess like that. You need something external to prompt you to do that. Inserting this meta-madhack may happen to accomplish it at this moment of your binge, but you need something more general.
Like Beeminder.
Beeminder is constantly snapping me out of addictive or procrastinating behavior. It’s not the only way or necessarily the best way (this may depend on your personality) but it’s an extremely reliable way.
Humans are made to walk. Set up your life to encourage walking by acquiring soft-soled shoes, good audiobooks, and/or a dog. If you’re not enjoying walking and not getting your 10,000 steps you can get there with good design choices.
Verdict: First just beemind that shizzle, then try the above ideas to make it as enjoyable as possible. I often hear skeptics suggest that Beeminder is so harsh or negative or stressful and they’d rather take more positive approaches — like the ideas above to encourage more walking. My answer to that is: why not both?? Beeminder can be your insurance or safety net, to make sure you don’t fall below a minimal acceptable level and then you can also do all those other things to make Beeminder irrelevant. It’s still nice to collect the data and see the graph. If you’re correct that you don’t need that Negative Nancy, Beeminder, to get your steps (or whatever it is) then wonderful. You just have a nice graph and no stress or derailments. If you’re wrong, then — by definition of “minimal acceptable level” — you’re much better off than in the Beeminderless counterfactual.
It’s like Pascal’s Wager maybe?
Crap, that just went full on proselytizing straight out of the gate, didn’t it? That’s not actually the intention of Madhack Mondays. I’m just really into Beeminder, it turns out.
Peanut Gallery:
Wrestle while naked and covered in coconut oil. [link redacted]
Verdict: I’m going with “if you want to”. I guess I haven’t tried it so I’m just fundamentally ignorant!
(I sure was embarrassed to have to include this one in the daily beemail which I concluded like so: “hides face, hits send Wait, this is even more embarrassing if anyone’s just now tuning in so I should link to doc.bmndr.com/madhack for how I backed myself into this corner of having to talk about coconut oil wresting in a Serious Professional Publication For Our Business.”)
Peanut Gallery:
Buy a $20 bar of soap on Amazon just to see how it feels. If it doesn’t do much for you, go back to $4 bars. Liquid soap has a low ceiling, so don’t bother. See Shopping For Happiness.
Verdict: This specific thing sounds dumb to me but Jacob’s post on shopping for happiness is great if you like nerding out about your own utility function, as I emphatically do.
My favorite quote:
Converting money into happiness seems perfectly straightforward: maximize the ratio of money-equivalent resources spent, divided by the integral over time of hedons experienced relative to a neutral hour. There, we’re done!
(The whole thing is pretty funny. If you’re me.)
There are also lots of other reasonable-sounding madhacks sprinkled in there, like not owning a fancy car but splurging on renting one while on vacation. Or buying seasonal treats (egg nog, candy corn), not making them a regular thing, and not binging your favorite show. In general, break up your indulging as much as you can to avoid acclimating. Eat popcorn with your non-dominant hand to force yourself to pay more attention to the enjoyment you’re deriving.
I mean, I’m generally/vaguely skeptical but this is worth at least experimenting with. So that makes it a pretty ok madhack. And I guess the real madhack here is to read “Shopping for Happiness” and think about counter-intuitive ways to change how you spend your money.
Oh, and one tangential opinion of my own about soap: Don’t buy antibacterial soap. It aids and abets bacteria (via evolutionary pressure) and doesn’t offer any kind of protection over plain old soap anyway. I’m out of my depth on this but if anyone clueful wants to concur or set me straight, I’m all ears.
Ooh, and bonus pandemic madhack, via Astral Codex Ten:
To make sure you’re washing your hands long enough, sing this song to the tune of “Rudolph the Red-Nosed Reindeer” while washing:
Novel coronavirus
Has a lipid outer shell.
Washing your hands correctly
Sends it back to virus hell.
Gallons of steaming water
Won’t slay the beast if used alone.
You need a good surfactant —
The role for which our soaps are known!
(Typed from memory, having sung that to myself countless times now. bows extravagantly)
Phew. Who knew there’d be so much to say about soap? My first reaction to this madhack was “verdict: eyeroll. next.” I mean, I still have zero interest in trying a $20 bar of soap. Money aside, it’s just too implausible to me that I’d derive any utility at all from better soap to bother with this one. But that’s just me!
Peanut Gallery:
Verdict: Dudes/dudettes, shower when you want to. I shower in the morning because my hair is greasier than a teenager’s and it’s embarrassing to be seen in public before I’ve showered. I’ve heard the theory that that’s a vicious cycle, that my hair is so greasy precisely because I strip it of all its natural oils every dang day and so it’s trying to compensate, thus inducing me to shower more, etc. Plausible! But also showers are great and I’m so filthy rich that I can take them every day just because.
I guess shower timing seems silly to me as a madhack because it’s self-reinforcing. If it feels nice to shower before bed then you’ll just do that. It’s not like it’s something you wouldn’t think of. Is next week’s going to be “you know what tastes good? cake”?
Actually, I have a related parenting madhack: never make your kid wear a sweater or jacket because you think they’ll be cold. That’s a battle of wills that entirely goes away by letting Mother Nature impose Her own consequences. Let your child die of pneumonia and maybe your next one will be more evolutionarily fit! Just kidding, being cold doesn’t cause any kind of illness, why do people still think that? It makes you seek out a sweater, or jump around to burn more calories. We do sometimes make our kids step outside before deciding whether they want to wear a jacket, but it’s always 100% their choice. This works even for toddlers. If they don’t feel cold you can just believe them. Unless you made it a pointless will-battle and they refuse to admit they’re cold but even then, they’ll cave soon enough. Bide your time, beleaguered parent!
(Update: Also you can just have a jacket along in case they change their mind; you don’t have to be all “you refused a jacket and NOW YOU SHALL SUFFER THE CONSEQUENCES”.)
Peanut Gallery:
Doctors are fallible humans, they have biases and make mistakes. It’s your job to be educated about your diagnoses and the drugs you are prescribed. If you’re confused, ask for details or a second opinion.
Verdict: Jacob speaks the truth. Dicts the ver, if you will. (I see by your expressions that you emphatically will not.)
Adam Wolf and I were talking about this recently, about the extent to which a couple hours of internet research can put you ahead of actual doctors. I think you’d be surprised. I mean, a couple hours is an exaggeration but by ignoring everything that’s not relevant to your own specific case you can get beyond the understanding of most doctors in like a matter of days absorbed in the medical literature, I think. That’s what I found when trying to figure out what to do about my Crohn’s disease when I was in grad school. But then I went to the Mayo Clinic and those people really knew their shit. So to speak.
Peanut Gallery: A chorus of “OMG people trying to follow this are going to conclude that wifi signals are giving them coronavirus”, etc. Fine. Consult your doctor before reading the previous paragraph.
In ⚽ / 🎾 / 🏓 / 🏐, keep your eye on the center of the ball through the hit. The goal/court/table doesn’t move, only the ball does.
Verdict: As one could probably predict from everything else one might know about me, the closest thing to a ball sport I’m into is ultimate frisbee. To me this tip sounds equivalent to “in bike racing, remember to pedal as fast as you can”. Or maybe there’s some kind of clever “the enemy’s gate is down” thing going on there? (That was a nerdy sci-fi reference, in case any normal people wandered over when we started talking about ball sports.) I mentioned this one to Adam Wolf who’s even nerdier than me and he suggested this madhack may be equivalent to “don’t keep looking at your damn cards, you know what they are”.
So, yeah, I think you’re on your own with this madhack. Good luck!
Peanut Gallery: Split between “has merit” and “if it were this easy, these sports wouldn’t be so hard”.
Even if you end up throwing a couple apples out once in a while, it’s hugely valuable to have a tasty fruit closer at hand than junk food.
Verdict: Yes. Which reminds me of a classic Beeminder blog post by Alys. Also beeminding junk food as a do-less goal is great. I still wear an abacus bracelet to help me keep track.
Peanut Gallery: Garrison Taylor suggests mixed frozen berries since they keep indefinitely.
In case you missed it, humanity has fully optimized apples. Snapdragon, Zestar, and Cosmic Crisp if you can find them, Honeycrisp or SweeTango as backup, Fuji in a pinch. All other cultivars are a distraction.
Verdict: Poppycock! Why, back in my day, apples were so Delicious that, no, I’m kidding, Honeycrisp apples are the best I’ve ever tried and Fuji are pretty good, so most likely Jacob knows his way around apples and those other ones — which I haven’t tried — are even better. Basically, I don’t know about them apples but this advice is highly plausible.
Peanut Gallery: Clive Freeman points out that there are over 2,200 varieties in the UK and we’re still searching over a vast space to improve them, so, optimized-schmoptimized. Others chimed in with their favorites, including: #1848 from the University of Minnesota’s Apple House (Adam Wolf), Envy (Braden Shepherdson), and Gingergold (Mirabai Knight).
PS: We just did some taste testing and think maybe Splendor is the best?
PPS: AppleRankings.com says SweeTango is best.
Programming note: I peeked ahead and some of the upcoming entries in PutANumOnIt’s list are not family-friendly enough for a Serious Professional Publication For Our Business, which apparently I’m passing this off as. (The original #27 in PutANumOnIt’s list isn’t family unfriendly but is a bit naff, as one of the Brits in the audience put it.) I’ve decided to just sub in my own madhacks when that happens. This one is particularly meta and philosophical.
Even if a norm is dumb (like the definition of “family friendly”) it doesn’t make it ok to violate it. I had this epiphany when philosophizing about norms with Cantor (our now-11-year-old) a couple years ago. We discussed examples like norms about truthfulness and how some norms dictate some dishonesty. (Side epiphany: most norms are invisible to us!) Like you can’t say “you’re uninvited from my house because I’m feeling anti-social”. Or how you have to pretend things are no skin off your nose even when that’s not accurate. More honest, direct norms may be superior but only if everyone coordinates on them. If you switch to them unilaterally you are genuinely being a jerk and actually harming people and relationships.
I’ve thought about that surprisingly often since then. Another random example: My old article on gift-giving is a fine argument that gift-giving norms are dumb but not a good argument for not giving gifts (except among people you establish different norms with).
This is probably related to Scott Alexander’s “Be Nice, At Least Until You Can Coordinate Meanness” but I’m not sure how yet.
PS: Another great Scott Alexander article on this topic: “Give Up Seventy Percent Of The Way Through The Hyperstitious Slur Cascade”.
Peanut Gallery: There was some skepticism about my example of not being able to uninvite people to your house because you’re feeling anti-social. Fred Antell points out that if people expect the dishonest norm and get honesty, they read way too much into the honesty. Like not only must you be feeling anti-social but you must outright hate them if you were willing to say that directly. But, he says, this is mitigated if you can establish a reputation for extreme honesty!
Luke Barone-Adesi: “The invisibility of norms is the main reason people advocate travel to broaden your mind. Seeing from other people’s fishbowls makes bits of your own become less invisible to you.”
The #1 measure of an exercise program should be “is this fun enough to keep me coming back to the gym?” I don’t care how “efficient” HIIT is, it’s for masochists.
Verdict: Seems obvious enough, though there’s nothing wrong with masochism when it comes to exercise. I’m an extremely masochistic exerciser myself, though even for me I’m rarely motivated to do HIIT on my own. I need to be chasing things. (I chase bikes when on skates and sometimes cars when on a bike.) There’s pretty much no limit to the amount of pain I’ll willingly cause myself if there’s a real-world objective involved.
So I guess that’s my madhack: Make exercise social and you get things like HIIT for free by trying to crush / avoid getting crushed by fellow athletes. This is not a very helpful madhack though. I’m basically saying “compete in sports!” and if that’s not your idea of fun then getting some high intensity interval training (that’s what HIIT means, btw) is not going to convince you. But the general madhack is good: make fitness fun.
Related bonus madhack: Combine a sport (running, skating, biking) with transportation and leave juuuust enough time to get to your destination. Automatic HIIT between all the stoplights.
I should also mention the classic from Prof Katy Milkman: Take your favorite podcast or a page-turner novel and arrange so that you can only consume it while at the gym. Temptation bundling, she dubbed it.
Programming note: Madhack #29 from PutANumOnIt (use blackout curtains to get more sleep) is fine but boring so I’m swapping in my own again.
Intermittent fasting is the bees’ knees.
Bee and I have started a thing we call Midweek Megafast. We consume nothing but water + electrolytes between dinner on Wednesday and dinner on Friday. A 47-hour fast! This will be the 3rd week in a row we’re doing it. My verdict: I feel like I have more energy, more concentration, more time, and get a nice dip in my weight enabling me to pig out afterwards. It’s great! Even the hunger pangs go away on the second day. Not that I mind hunger pangs that much anyway. Also I hear stressors like fasting and being cold (and exercise, obviously) help you live longer? (That’s just a thing I heard on the internet. Not entirely implausible but who knows.)
I realize some people get hangry and I’m lucky that I have zero emotional side effects. Also, as I mentioned in the previous madhack, I’m kind of a masochist. For many people, no list of nebulous benefits are going to compare with the sheer unpleasantness-bordering-on-pain of skipping like 5 meals in a row. (I get it. That’s kind of how I feel about meditating.)
But that’s the point of a madhack: It’s weird but might be worth a try!
UPDATE: This has evolved into Waste-away Wednesdays (no eating between Tuesday night and Thursday morning, or the milder version: no eating between Tuesday dinner and Wednesday dinner).
FURTHER UPDATE: This further evolved into
Alliterative Alimentation which has been going strong for most of 2022 over a year and counting.
Verdict: This is my own madhack and thus comprises unambiguous objective truth. Just kidding, I have low confidence in any of this being a good idea. I personally like it and it kinda makes some vague sense to me in terms of the science, which I know little about. Not medical advice, obviously.
Peanut Gallery: Thanks to Caelan Mitchell for pointing out that the theory that stressors may improve longevity is related to Hormesis. Other folks pointed out various fasting protocols that the Midweek Megafast is similar to. I mostly prefer the quasi-random one induced by Beeminder: If I’m above my weight graph’s bright red line in the morning then I can’t eat until I’m back on!
When I sent out the previous madhack in the daily beemail, everyone who thinks intermittent fasting is idiotic and/or insane was very polite and didn’t reply. (I then — in what’s now this very parenthetical — emphasized that I was interested in counterarguments so don’t be shy! That yielded plenty of anti-intermittent-fasting sentiment.)
Relatedly, I included a parenthetical in the Midweek Megafast madhack that was vaguely negative on meditation. I thought of expanding on that for a madhack [1] and then I remembered a great admonition from one of our kids’ teachers: Don’t yuck my yum. So that’s madhack #30. Let people enjoy things they enjoy. It’s pretty much never a good idea to make arguments against enjoying something. Unless there are negative externalities, obviously.
If you must do it, like if someone presses you to opine, like in the footnote below, keep your negativity as specific to yourself as possible.
Verdict: Right? Of course right.
Peanut Gallery: So many new opinions on intermittent fasting, my goodness. But those are all off-topic for this madhack.
[1] In fact, one of you tried to pin me down on the meditation question, which, uh oh, now you’re all curious. Fine, since we’re already in a footnote: When talking about fasting, I said that for many people, no list of nebulous benefits are going to compare to the sheer unpleasantness-bordering-on-pain of skipping like 5 meals in a row. I added that I get it, it’s kind of how I feel about meditating. What do I mean? That all the nebulous mindfulness and anti-stress benefits don’t feel, from my perspective, like they can make up for the sheer boredom and opportunity cost. It just feels wasteful (for me) to spend a chunk of time doing (what feels to me like) nothing.
There’s a Chinese proverb: “The best time to plant a tree is 20 years ago; the second best time is now.” I guess if you want to get pedantic, 19 years ago is the second best time? Depending how you discretize? Math it up however you need to, the point stands. It’s a nice mantra to neutralize the what-the-hell effect, which as far as I can tell is the actual technical term in psychology for letting a small deviation from rationality spiral into an egregious deviation.
(Then David Ernst pointed out that what I’ve called the “what-the-hell effect” sounds like the sunk cost fallacy — doubling down on a bad decision instead of cutting your losses.)
So that’s the madhack: When you start to succumb to the what-the-hell-effect or the sunk cost fallacy or are just kicking yourself for not having adopted some no-brainer of a madhack years ago, say out loud, “second best time to plant a Chinese tree!” and give yourself a mental clean slate.
UPDATE: This is now a major motion blog post!
Content warning: weight loss
If you’re not obese, have you considered that losing 20 pounds will not actually solve all your problems? If you can’t lose weight easily, keep your weight stable and work on the insecurities that make you scared to take your shirt off.
Verdict: I don’t know how to respond to that and mostly don’t want to touch it with a 10-foot pole (do you need to work on insecurities? I’m far too insecure in my understanding of psychology to say) so let me focus on the part I emphatically agree with: “keep your weight stable”. I have a whole beemindery pitch about that that I shall now try out on you. Deep breath, wavy flashblack lines, aaaand…
For most of my life I ate whatever I wanted and was thin. Then one day I got old and fat. Well, half true — and I would’ve unambiguously gotten fat (truly, there’s no ambiguity there) but of course I have a Beeminder goal that prevents that. Instead I’ve had to stop eating whatever I want. I achieve that by intermittent fasting, so that I actually can still eat whatever I want, just not all the time. That might not be the healthiest way to modify one’s diet (and this madhack is not recommending it! that was madhack #29) to avoid gaining weight but the point is, you should have the Beeminder goal so that if/when this metabolic tragedy happens to you, you don’t wait until you notice you’re overweight to make whatever dietary / fitness / lifestyle changes make sense. You’re forced to gradually make them day by day because you’ve hard-committed to keep your weight below Beeminder’s bright red line.
Point being, becoming overweight and then losing weight is total insanity. Whatever shıt you have to do to lose weight, it’s drastically easier and healthier if you do that shıt to NOT GAIN IT IN THE FIRST PLACE. It’s too late, you say? You’re already overweight? Second-best time to plant a Chinese tree! Hard-commit right now to not gaining any further weight. How or whether to lose weight is a secondary question.
~~
I tried that pitch out on 2 or 3 of you and got some quibbles about who it doesn’t apply to and I kind of want to double down on it. Or at least there’s a generalization that’s applicable to everyone. You can start with “what’s an upper bound on your healthy weight?” For many people that’s unambiguously something lower than their current weight. For me personally, I don’t know that it’s less than my current weight but I’m fully certain that I don’t want to gain, so my current weight is an upper bound. For others the answer might be your weight plus five kilograms or something.
But all of us can take the answer to that question, apply the “max” function with our current weight, and set a Beeminder goal with a flat redline at that number. If your weight is akratically low, make another Beeminder goal to get yourself up closer to that upper bound!
And, sure, in some cases it’s silly to bother with the upper bound if you’re certain your weight will never approach it. Or at least that it won’t do so without you noticing, so you can cross that bridge if you come to it.
~~
Ok, that was a lot of words (it’s a dicey subject, ok?) to say something pretty simple: Establish an upper bound for your weight and hard-commit to not crossing it as you age. And if you’re losing weight, do the same thing to ratchet yourself, always locking in your losses.
I know my opinions on weight loss infuriate some of you but “don’t yo-yo” is something we can all agree on and that’s ultimately all that this madhack (as opposed to my other more infuriating ones) boils down to!
Once in a while, try eating only a short list of simple foods for several days. For example, carrots+almonds+yogurt+water. You’ll eat less without being hungry, and afterward you’ll savor flavorful foods a lot more.
Verdict: I believe this is based on Stephen Guyenet’s “The Hungry Brain”, which you can read about on Slate Star Codex / Astral Codex Ten.
(I guess I’m doing the extreme version of this where my “short list” is the empty list. I’m up to 71 hours on my midweek megafasts now! Still liking it a lot.)
Peanut Gallery: David Gessner says this reminds him of the Stoic practice of voluntary discomfort (which he has beeminded!) and offers this Seneca quote:
Set aside a certain number of days, during which you shall be content with the scantiest and cheapest fare, with coarse and rough dress, saying to yourself the while: “Is this the condition that I feared?” It is precisely in times of immunity from care that the soul should toughen itself beforehand for occasions of greater stress, and it is while Fortune is kind that it should fortify itself against her violence.
Others mentioned the Shangri-La diet, from Seth Roberts.
Having declared a moratorium on butt-related madhacks (even perfectly benign, family-friendly ones like this would-be madhack #34 would’ve been), I’m now free to make up any madhack #34 I like. Ah me, do I have a lot of outrageous opinions I could inflict on you!
Instead, here’s a simple madhack from one of my favorite youtubers, CGP Grey: Weekend Wednesday. His video is pretty pithy and persuasive but let me make my own contribution to the internet by turning it into my own typed words that you can read quietly and probably much faster.
Claim: Five workdays in a row is too much. You’re drained by the end of the week and your productivity is plummeting. Then you need Saturday just to recover and so it only feels like one day of actual relaxation on Sunday before the cycle starts over.
Solution: Swap Wednesday and Saturday. Now you just have two days of work, a break, then three days of work, followed by another day off.
Advantages: (1) More frequent days off, (2) a non-work day (Wednesday) that’s off-sync with the rest of the world so you can do recreational things with less crowds, (3) a workday (Saturday) with few interruptions, and (4) weekend days you can enjoy more because you’re less drained.
Verdict: I don’t actually know! This “being drained by the end of the week” concept isn’t one I relate to, so the whole thing feels inapplicable. Also I have kids in school so I’m somewhat stuck with the standard workweek if I want to stay in sync with them. So I haven’t been tempted to implement this for myself. But that’s just me! I figure at least one of you will want to try Weekend Wednesdays.
(Thanks to both narthur and Mary who pointed me to this idea.)
Peanut Gallery: Philip Hellyer tried taking Wednesdays off once, in addition to normal weekends. He said that in practice it meant every workday was either a Monday (ramping up) or a Friday (ramping down), which was ridiculously unproductive for him, at the time. But he adds that that was before Beeminder!
Learn how caffeine and alcohol affect you. I know people whose quality of sleep improved dramatically once they stopped having coffee with friends after lunch; it turned out they are metabolizing coffee very slowly and it affected them 10 hours later.
Verdict: Sounds sensible to me! I beemind caffeine and don’t consume it anywhere near bedtime. I don’t beemind alcohol because, weirdly, I’ve never consumed enough of it to experience any effect at all. The concept of altering my mental state weirds me out too much, I guess. I’m pretty square, I know. Especially when I say “square” like that like I’m from the 60s or something. In general nootropics seem worth experimenting with but I’ve been too timid to go beyond caffeine so far.
Slightly related update on my Midweek Megafasts: 48-hour fasts now seem trivial but last week I got headachey by the end of the 71-hour version, so, first of all, I did break the fast slightly early, like a reasonable person, and, second, I’ve decided this week to try Midweek Minifasts instead: Only dinners on weekdays, ie, a 23-hour fast every day for 5 days.
I might go back to just fasting when Beeminder tells me to — i.e., based on my scale weight — but figure I’ll keep experimenting with more routinized systems a bit longer first. It does seem to free up attention to precommit to simple rules like Midweek Megafasts or Midweek Minifasts.
Peanut Gallery:
Philip chimes in that he’s adopted Bee’s genius method of coffee-making: Buy a hand-grinder and gradually reduce the proportion of caffeinated beans in your mix. He says he got down to zero for a while but settled on combining a 1kg bag of decaf with 250g of caffeinated for a consistent 20% — a nice happy medium.
As originally sent on 2020 Nov 1, when we switched from Daylight Savings Time to Standard Time in the US.
I’m doing an early Madhack Monday today because I have something timely. Timely, get it? (Well, for Europeans it’s a week out of date.) Ok, here it is:
As any computer programmer can attest, off-by-one errors can be dang confusing. At least I’m personally bad enough at mental arithmetic that it’s easy to confuse myself about whether I’ll get tired or wake up an hour earlier or later after the clocks change for daylight savings time.
I think the right concept handle for this is body time. The clocks fell back last night which means that, relative to the clocks, body time is an hour ahead. If you usually go to sleep at 11pm then at 10pm clock time, it will be 11pm body time. So that’s when you’ll get sleepy. Same thing in the morning. You wake up at your normal body time and clock time is an hour earlier than that.
Body time is +1 hour after falling back in the fall and -1 hour after springing head in the spring.
Easy peasy. And especially helpful if you have children if you want to be like “ok, time for bed, it’s 9pm body time” at 8pm tonight!
And now PART II, in which I alienate all of you by expressing a highly unpopular (in my nerd circles) opinion…
Daylight Savings Time (yes I say “savings”; sue me; but that isn’t the unpopular part) is a brilliant (mad)hack solving an otherwise utterly intractable coordination problem. Standard business hours are 9-5 and even if you’re self-employed you probably are, for example, on an ultimate frisbee team or somesuch that can’t start at 5pm because of all the nine-to-fivers on the team. There’s a whole web of interdependent schedules and there is just no way to induce a critical mass of people to shift their activities earlier in the day so as not to waste so much daylight by sleeping in hours past dawn.
UNLESS you resort to the outrageous hack of just literally changing the clocks.
Sure, it makes life a living hell for computer programmers, and the original rationale of saving energy on lighting surely doesn’t apply. (Also I guess it literally kills people, but so do a lot of things that are obviously still worth it. Like ice cream, probably. Cars I don’t know. If they’re worth it, I mean.)
Point being, more daylight in the evening is a big deal. And the idea that people could just choose on their own to wake up earlier when the days start getting longer is all wrong. I mean, yes, you can personally do that, but it does you no good unless everyone else (like the rest of your ultimate frisbee team) does it too. And your frisbee team can’t do it unless all the businesses do it and the businesses can’t do it unless the trains and buses do it and every other sport and club and social group and… Like I said: massive, intractable coordination problem.
I think it’s kind of awesome that we were able to solve the problem at all.
“Ok, fine, permanent daylight savings time then!” says everyone I know. I don’t think I would mind that but it does mean starting the day in pitch darkness in the middle of winter. I think ideally I’d love to see a system of referring to times that was relative to sunrise. This is a can of worms though. Or a can full of cans of worms, one of which is how much nerds despise the concept of timezones.
Peanut Gallery: Oh my goodness so many opinions! There ended up being a fun thread about this in the Beeminder forum so I shall redirect you there for the peanutting.
PS: Briefly in 2022 it seemed like the US was going to try switching to permanent DST. Apparently we tried that in the 70s and people hated it.
I’m swapping in my own again because Putanumonit’s #37 is
(Hi, funny story here. In the original version of this madhack that I emailed, I forgot to finish that sentence. Embarrassing! It seemed culture-wars-adjacent (“is masculinity good” blah blah blah) so I didn’t want to touch it with a 10-foot pole and then I didn’t even want to say I didn’t want to touch it with a 10-foot pole because saying so is kind of touching it with a 10-foot pole. While agonizing about that little paradox, I got distracted, and … yeah. So I guess that was sort of a way to not touch it with a 10-foot pole. Just not even finishing the sentence about why I wasn’t using Putanumonit’s madhack! Moving right along…)
I read CS Lewis’s “The Inner Ring” recently and found it wise, possibly valuable. I’m still reflecting on whether I could make a madhack out of it somehow.
But while reflecting on that, I thought of a meta-madhack: When you encounter an insight that seems valuable, share it with people in your own words. Why?
Ironically, when I first started listing the above points I almost stopped because it seemed too obvious. Everyone knows you haven’t really understood something till you can teach it, etc, right? Remember the classic joke about what mathematicians mean by “obvious”? If you gradually or circuitously conclude something’s obvious, it isn’t. Blog that shit.
Anyway, I’m still not sure if I’m impressed enough by “The Inner Ring” to bother re-articulating it, though I don’t doubt that it’s advice some people need (and probably some people should reverse — remember madhack #1!).
Peanut Gallery: Philip points out a classic xkcd relevant to point 6. And a fan of CS Lewis’s essay summarizes it like so: “When you feel yourself tempted to belong to some group or reach the inner circle of some organization, consider whether it’s actually just the sense of belonging that you crave, and consider carefully the ethics of that situation.” They also pointed out that in legal writing it’s considered better to paraphrase a court’s opinion rather than quote it.
We also discussed when it could possibly make sense to reverse CS Lewis’s advice. I came up with this, though it’s pretty speculative and should not be construed as me disagreeing with CS Lewis (God forbid):
Some people suffer pathological scrupulosity and have imposter syndrome or maybe they’ve internalized a high-school rebelliousness kind of attitude where they self-identify as a lone-wolf type who doesn’t need anyone else and sees all attempts to fit in as despicable brown-nosing. There presumably exist people for whom “no, it’s ok, this group is valuable and putting in effort to be a part of it is valuable” is good advice. Or just someone who’s too unambitious but not at all in danger of going ethically astray. Maybe they should be encouraged to seek status more. Maybe that would result in achieving more and being a better person due to the kinds of circles they’d be seeking status in. Madhack #1 FTW!
PS: I love this graph I found (via Jacob Falkovich again) on Twitter:
Not a single hungry child in Africa was helped by you finishing a meal you didn’t enjoy.
Verdict: True fact. I mean, I could steelman the argument for not wasting food and for clearing your plate with appeals to promulgating a culture where we maintain a sense of gratitude for the food we have. Probably that could be persuasive.
But pragmatically, this madhack is hard to argue with. Here’s how we put it in our blog post, “The Dirty Plate Club”:
If you have food left on your plate but are full, it’s more wasteful (waistful, har har) to put it in your face than to put it in the trash. Eating food your body doesn’t need helps no one, least of all you.
It feels more wasteful to throw it away, because, well, we were raised that way. And maybe also because of cognitive dissonance. The real wastefulness was serving yourself too much food in the first place and by cleaning your plate you don’t have to admit that you did that.
[… the crazy part about the dirty plate club with historical and cross-cultural commentary…]
In theory you can achieve the same effect more efficiently by serving yourself slightly less than you feel like eating. (Actually, go ahead and do that too!) But that’s hard to predict and doesn’t break the habit of eating on autopilot when you happen to have too much food in front of you.
I also realized we have a much earlier blog post making a similar point, back from before we’d publicly launched Beeminder and just blogged any random little thought that occurred to us, apparently: “No Preemptive Eating”.
Madhack Monday is growing up so fast! We have finished with the Body section of PutANumOnIt’s “100 Ways To Live Better” and are now embarking on the Stuff section. The first stuffy one seems obvious and boring though (get extra chargers and cables so you have one handy in every room) so let me do my own again, in honor of the first Beeminder Community Discord Video Night, which we had yesterday, thanks to narthur.
Ready? Approval Voting is the answer to voting. I mean, yes, if we were voting for the best voting mechanism, I’d give my vote to everything except Plurality Voting (the kind most of us are familiar with from political elections — one person, one vote). But if we eliminated that abomination and just rounded up the usual suspects — Instant Runoff, Range Voting, Borda Count, etc — I’d put my full support behind Approval Voting.
(I’m also assuming that Quadratic Voting or anything else that uses any kind of currency is off the table. Those are my favorite of all, but I appreciate that those are repugnant to normal humans.)
If you twist yourself into knots you can convince yourself that Range Voting or Instant Runoff or whatever are somewhat better than Approval Voting but (a) I’m not sure it’s true and (b) if any are better, they’re not enough better to overcome the sheer simplicity of “vote for as many options as you like”. It’s the exact same mechanism — the option with the most votes wins — as plurality voting, just removing the constraint that you can only vote for a single option. Most polls (like the ones built into the Beeminder forum software) already support it. It’s also trivial to do with a show of hands — or emoji-reactions in Slack or Discord — for each option. Even easier than plurality voting because you don’t have to check that each person is voting only once. Everyone votes for everything they like, most votes wins, end of story.
Proponents of other voting mechanisms will malign Approval Voting as not letting voters express rich enough preferences or point out ways in which it still allows for strategic voting. Every possible voting mechanism has similar problems! (Literally every possible voting mechanism — there are famous impossibility theorems about this.) The strategery in Approval Voting is at least simple and transparent: pick your threshold for approval. Maybe you strategically decide to vote for only your favorite even though a few other options are quite decent. Fine. Or maybe you decide to vote for everything except one option you despise, including some also-kinda-bad options. It’s better than the more complicated strategizing you’ll have to do if you can assign everything a score or give a full ranking.
Also it turns out Approval Voting is in fact the special case of Range Voting (aka Score Voting) where the range of scores you can give to each option is just 0 or 1. I guess only nerds will find this fact interesting. And one of the big advantages of Approval Voting is that it’s 100% non-nerd friendly.
So there you go. Save democracy, vote for Approval Voting.
Peanut Gallery:
Highly beemindable workout idea for the dead of winter if you have stairs in your house: Put 100 marbles or legos or whatever in a bowl at the bottom of your stairs. Then start a timer and see how fast you can carry them all upstairs one at a time. My record so far is 17m31s, edging out our 11-year-old by 2 seconds.
Rules:
I worked out a technique where you stay super crouched as you bring your second foot up to meet your first foot at the top, before pivoting and dropping 2 stairs down. I.e., minimize how far you raise your center of mass. So much saved energy! Not as much saved energy as ignoring this madhack altogether, but, I mean, are you going to leave this gauntlet just sitting there?
Of course I don’t know how standardized house heights are so if you beat me I’ll assume your house is shorter.
And for those who are not freakishly competitive you could just beemind a number of times you go up and down the stairs per day or per week — tracked via marbles/legos — without timing it and without doing it 100 times in a row or until you throw up. I assume that’s a thing people could do!
Peanut Gallery:
Old: buy 20 of the same pair of black socks so you don’t have to worry about matching. Bold: buy 20 colorful pairs and don’t worry about matching.
Verdict: Two big toes up.
Peanut Gallery:
Ask people to stop giving you non-consumable gifts. A physical thing that’s not exactly what you need costs more in storage and opportunity cost than it’s worth.
Verdict: Look at the impeccable timing of this madhack [it was right before Christmas]. I do like the idea of establishing a norm that gifts should be consumable. I think that mitigates the deadweight loss, as economists say, of gift giving. I have a lot of opinions about this from an econ-nerd perspective. I even once wrote an article about it.
But, yes, prefer consumable gifts! Also a trip or seeing a play counts as consumable. Anything that doesn’t involve permanently owning a physical object.
Peanut gallery:
Buy some cryptocurrency, maybe 2-3% of your net worth. Barbell investing makes sense. As a bonus, checking Coinbase every day provides the same excitement as checking social media but takes a lot less time.
Verdict: I’d be more inclined to short cryptocurrencies at this point [December 2020, haha] but I don’t think you’re here for investing advice so let me pick a fight with Scott Alexander instead!
Scott wrote “A LessWrong Crypto Autopsy” a few years ago in which he lamented the rationality community’s failure to capitalize much on their (plenty impressive) prescience about Bitcoin.
Let’s say Scott’s right that Rational Consensus said to invest $10 in 2011. The fallacy that bugs me is imagining that one could’ve thereby turned $10 into $100k just because at some point later the price went up by a factor of 10,000. Scott cites Gwern Branwen as particularly prescient here, and he was, but even he originally said that a $10k price for Bitcoin would make sense if Bitcoin became the world currency. It makes zero sense now. So when the price hit, probably, $1k, or maybe even $100, the rational thing would’ve been to sell everything.
Following the Rational Consensus meant buying in for a small amount when the price was $1, and mostly cashing out when the price went up by, say, 100x. Maybe you’d repeat that reasoning and not sell quite everything. Then the price would go up another 100x and you’d make a few more thousand dollars.
But in no way is it a rationality failure to have not made a million dollars because you (like me!) thought Bitcoin was cool in 2010.
In other words, it feels highly instrumentally rational to have invested a small amount in 2011 and cashed out at a million dollars profit in 2017 (or 2020, not so much in between) but that’s massive hindsight bias. Unless the consensus is still to invest now. But it’s not, at least not in any significant way. It might even be worth shorting it now, if there were a good way to do that.
If so, then, if you had (quite rationally) invested $10 in 2011, the time to have cashed out is also in the past. Probably long before making profits worth writing home about.
Also a quibble with Scott’s post: It feels misleading to say Gwern predicted a $10k bitcoin price. He specifically predicted a 99.95% chance of it not hitting even $5k. I think Scott acknowledges this later in his post.
PS: A logical corollary of this sounds particularly weird but, as usual, I’m ok with the bullet-biting: If you made millions of dollars investing in Bitcoin, that was irrational! Getting rich from a price bubble means you merely got lucky. Piling on to a price bubble in hopes that it lasts a bit longer is not a good investment strategy.
I haven’t exactly refuted the claim in the original madhack because I don’t know that Bitcoin is a bubble. I just think it is.
See also 3blue1brown on how Bitcoin actually works, which is brilliant.
Reversion to the mean seems obvious enough. If two supergeniuses have a baby, you expect that baby to be smarter than average but less smart than its parents. Its parents are anomalies after all. Here’s a less obvious application that’s worth keeping in mind when someone tells you a compelling-sounding anecdote about alternative medicine. Someone with a randomly fluctuating chronic condition (as chronic conditions tend to be) will get more and more desperate as they fall deeper into a random trough and so it will tend to be at the trough’s nadir that they try something like healing crystals or whatever. So then, by regression to the mean, they’re utterly duped by post hoc ergo propter hoc. In their unassailable subjective experience, the healing crystals were a slam dunk, unambiguously precipitating a drastic improvement!
If you’re meeting a friend for lunch who makes less than half your income, you should pick a place in your price range and pay for both of you. And if a friend who makes double offers to do the same, accept it graciously.
Verdict: What are we, communists? Just kidding, this is fine. I even have a whole mini essay proposing something very vaguely similar (half similar? not really similar) about friends and money but I’ve decided it shouldn’t be public. How’s that for a tease?
Peanut Gallery:
Try a much harder mattress. Try a much softer mattress. They all have 100-day free trials now, there’s no excuse for spending thousands of hours on a less-than-perfect mattress.
Verdict: This is my skeptical face. Counter-madhack: Don’t believe the mattress hype. If you’re not waking up in the middle of the night uncomfortable then there is exactly nothing wrong with your mattress. Although… what ever happened to waterbeds?
Peanut Gallery:
PutANumOnIt’s Madhack #47 is something about tea. Yawn.
Here’s something of my own, inspired heavily by the rationality community:
How To Disagree:
See also the Scott Alexander classic, “Varieties of Argumentative Experience”.
PS: And the Rogerian Argument Model (HT Theo Spears).
PutANumOnIt’s Madhack #48 is only two sentences but I couldn’t get through them without my eyes glazing over. Something about car maintenance. Instead I’ll tell you about something else I first learned or at least first appreciated from reading another PutANumOnIt post, in which the phenomenon was called, idiosyncratically, “triangle distributions”, for reasons which are about to become clear…
A few months ago we were chatting in the beehive about support. I made the point that the perceptions we have from our support interactions can be misleading. The users we talk to in support may not be representative of the whole userbase. I even coined the term “invisible bread and butter” for the conjecture that most of our revenue is from people we don’t see in support at all.
(This turned out to be wrong. Only a third of revenue comes from invisible users. It goes up to just over half if we count invisible plus nearly invisible users.)
Anyway, I ended up making a scatter plot comparing number of support interactions with revenue. It turns out there’s not much correlation between revenue and more support interactions.
Now we’re getting to the actual madhack: to beware of bogus correlations that can appear if your data is filtered in some way. Here’s how it can happen. Imagine if we were all ruthlessly mercenary in support (we aren’t!) and just fired any user if their ratio of revenue to support cost were too low. That would be like drawing a diagonal line on that scatterplot from the bottom left (low revenue, low support cost) to the upper right (high revenue, high support cost) and disappearing everyone below it. If we were just quietly doing that — we aren’t! — or implicitly doing it by just being meaner to such users and driving them away — I hope we aren’t! — then we’d end up seeing a scatterplot with an utterly spurious correlation: more squeaking = more revenue!
This is like the classic example where, among everyone you’ve dated, the more attractive people seem to be bigger jerks. But really it’s just that you eliminated everyone whose niceness + attractiveness didn’t meet some threshold. If they were mean and ugly then you never even considered them. If they were nice but ugly or mean but gorgeous, they had a shot. Draw that as a scatterplot and it looks like prettier people tend to be meaner and nicer people tend to be uglier. Even though in the general population — without your implicit filtering — niceness and attractiveness are entirely uncorrelated!
Ok, stats lecture over! Next time we can talk about the amazing Simpson’s Paradox, which I have an amazing way to elucidate involving Simpsons characters…
By popular demand, this week’s madhack is that you should understand and beware of Simpson’s Paradox! Also I have a really good way to remember it and make it intuitive, thanks to Michael Wellman (my PhD advisor).
First, here’s a classic example of how Simpson’s Paradox can rear its confusing head in medicine:
You have an Ailment and your doctor says there are two possible treatments, Pill and Surgery. Your doctor further informs you that those treated with Pill have a much higher success rate than those treated with Surgery. “Well that’s a no-brainer,” you say. “Let’s go with Pill.”
“Wait,” says the doctor, “I forgot to mention there are two types of Ailment: Mild and Severe. Fortunately you have Mild.” You ask if Pill is still the right choice. “Let’s check,” says the doctor. It turns out that among those with Severe Ailment, Surgery has a higher success rate. And among those with Mild Ailment… Surgery also has a higher success rate. What’s going on? If Surgery is more successful in both subgroups then how can Pill be more successful for the group as a whole?
Here’s another example, with some very hypothetical college admissions rates:
A university notices that it’s accepting men at a higher rate than women. There are two departments in the university: Carpentry and Psychology. The dean goes to each of them to see which one’s being sexist in its admissions. The answer is neither! The Carpentry department accepts women at a higher rate than men and so does Psychology. The numbers all check out: each department favors women in their admissions but the combination of both departments has a greater fraction of the male applicants being accepted than the fraction of female applicants accepted.
Finally, here’s the exact same phenomenon with Bart and Lisa Simpson (this is not actually why it’s called Simpson’s Paradox):
If you look at all the exams where neither Bart nor Lisa cheat, Lisa scores higher than Bart. (Lisa is the smart one, in case you don’t know the TV show.) If you look at all the exams where they both do cheat… Lisa also scores higher than Bart. If you look at all the exams overall, Bart scores higher on average than Lisa.
Suddenly it’s obvious, right?
Lisa never or hardly ever cheats and Bart almost never doesn’t cheat. Lisa averages like 97% without cheating and gets a 100% if she does cheat. Bart gets like 12% the one time he doesn’t cheat and gets 99% by cheating otherwise. So Lisa’s overall average is something like 97.1% (bumped up slightly from the time she cheated and got a 100). Bart’s overall average is 98.4% (dragged down a bit from 99% by the couple times he didn’t cheat).
So Lisa runs circles around Bart but Bart beats her, slightly, overall, by cheating. Paradox resolved!
It’s the same dynamic with the medicine example. The surgery is more effective but only severe cases with worse prognoses opt for it. If you opted for just the pill then you were a mild case. So of course the biased sample of those opting for the pill have more success. (Calling the treatments “Pill” and “Surgery” was a big hint. It’s particularly baffling if presented as just treatments A and B.)
With the college admissions example, it’s a similar story. Carpentry both attracts mostly men and is also very easy to get accepted into. Psychology is the exact opposite: it attracts women and is highly selective. Maybe the carpentry department accepts every woman in the rare event that one applies, plus most but not all men. And Psychology rejects almost everyone, women and men alike, but especially men — those they reject 100%. So each department has a blatant pro-female bias, and yet, because of the self-selection where men are opting for the non-competitive major, the overall acceptance rate for men across both departments is much higher.
Simpson’s Paradox is a case of Correlation Does Not Imply Causation. The non-invasive treatment is correlated with but does not cause better outcomes, for example. Recall that Berkson’s Paradox (last week’s madhack) says that correlation may not even imply correlation! Namely, spurious correlations can appear due to filtering of your data. Statistics is tricky business.
UPDATE: This is now a major motion blog post!
Beware Buying Before Beeminding! Are you about to buy something that requires ongoing time or energy? Some things don’t. A meal, for example, or a trip, just happens and then is done. But a musical instrument or a bike or a book requires investment beyond the actual purchasing. Some things, like a video game, require rationing lest they be even worse than a waste of money. So ask yourself, might this purchase be a waste of money (or worse) without a do-more (or do-less) goal?
Peanut Gallery:
Order weird clothes off the internet. It doesn’t make economic sense for anyone to open a shop of “J-pop streetwear” or “African athleisure” in your town, but someone from South China will send them to you for cheap. It’s easier to stand out by being weird than by spending more on the same style that everyone around you wears.
Verdict:
Utterly blank stare. I’m vaguely embarrassed that I can’t even fathom dedicating the modicum of mental energy it would take to do this. Even more embarrassing is that the reason for that is because I’ve essentially never bought my own clothing. I just wear what my cofounder (also my spouse) gives me. Wait, now that I say that, why don’t I just get her verdict on this madhack…
Bee says if your goal is just to stand out then you can, say, wear your clothing inside-out or buy crazy things from thrift stores for even cheaper. Probably there’s a lot more to “fashion” than standing out but, well, Bee’s saying she’s with me on not caring about that. She says just wear what’s comfortable the end.
Peanut Gallery:
Do blind tastings of wines, then just keep buying the $10 bottle you like best. Novelty is good, but let’s be honest: you can’t really tell different Malbecs apart that much.
Verdict: Yes, not that I know what I’m talking about. But this reminds me of the famous study about how you can fool wine connoisseurs by dying white wine red.
I know that sounds implausible and I wouldn’t be surprised if it doesn’t replicate but there’s something to this. Expectations profoundly shape our experiences. An utterly mind-bending demonstration you can try on yourself is the McGurk Effect.
UPDATE: Scott Alexander with the rest of the story on whether wine is fake.
Not a hypocrite, a Hippocrat. As in Hippocrates — primum non nocere. First do no harm. It’s surprisingly general!
(Thanks to commenters on Astral Codex Ten — plus the usual peanut gallery in the daily beemails — for helping me collect these!)
Another way to express the common theme here, and possibly a better concept handle for it, is Ratcheting. You want to make forward progress on something without risking any backward progress. Like in example 9, you want to improve your water fetching with a fancier bucket but you should be ready to fall back to your current bucket in case the new one is worse.
PS: See also Ratchets in software development.