From the abstract: “Overall, the size of the effect was substantial, but considerably smaller than the original study's effect size.”
Regarding small sample sizes: “
The current sample was obtained from over 4 years of recruiting efforts (summer of 2014–autumn of 2018). While our sample size is 33% larger than Ericsson et al.'s, it is still small. Conducting expertise research, by definition, means studying a small subset of the population, thus there are relatively few participants from which to sample, making large-sample replications nearly impossible. However, if replications of expertise studies never enter the scientific record because of small samples, then we probably will never provide additional evidence to support or refute the original studies. This leaves only the original study (with an even smaller sample size) as the only indicator in the scientific record.”
“To be clear, evidence from small samples—in original studies or replications—should be interpreted with caution. However, when the original study (with a small sample) has already entered the scientific record, replications with similar sample sizes (since that is all that is feasible) should also be allowed to enter the scientific record rather than be suppressed from publication. This is probably the best way to accumulate knowledge in this area.
”
Oooo! And they link to actual data they collected in the study! https://osf.io/4595q/
> Practice still mattered: It accounted for 26 percent of the difference between good violinists and the less accomplished students. But the original study claimed that practice accounted for 48 percent of the difference.
This is very different from "10k-hour rule confirmed".
The study concludes that practice is positive, but it is a small part of being the best, practicing a lot won't make much of a difference to achieving elite status, although it is probably a requirement for it. The study also points out that the top players actually had practiced less than the average of good players.
And this is for violin playing (plus another meta-analysis on sports). While the 10k-hour rule is most often associated to Bill Gates, and technical, academic or business success (at least in HN circles). Very different domains where physical dexterity matters little and learning processes have nothing to do with internalizing complex movement.
The numbers discussed in the study are roughly around the 10k-hour ballpark (practice before the age of 18). But again, there's nothing indicating that this generalizes to anything other than violin playing. I think it is fairly obvious that different disciplines take different time to master, although I haven't looked into the studies about it.
Aren’t many of these populations self selecting anyway?
It would be something else if the took a random sample of people off the street, got them to use ~10,000 hours to master something and then got positive results.
"replications with similar sample sizes (since that is all that is feasible) should also be allowed to enter the scientific record rather than be suppressed from publication."
Fake it 'til you make it.
That should be the slogan of the modern scientific process...
Ofc it's still true, as long as you don't take it too literally ...
Essentially it means it's gonna take a whole lot of practice and repetition for your brain to build the necessary neural links and reinforce them by forming a proper amount of socalled Myelin [0] so that they fire faster, more effectively and more efficiently than ever before.
And it doesn't matter what kind of skill you are talking about - be it physical or mental, eventually it boils down to the neurons in your brain.
If anybody happens to be into the current Age of Empires II scene, this [1] is what 10000 hours of practice gets you there for example (the guy has been the World #1 for a good while now).
> this [1] is what 10000 hours of practice gets you there for example (the guy has been the World #1 for a good while now).
This is false since it misses that tiny part called "ability", so there are plenty of people that can invest as much time and never get as good, but then I guess that's also taking it too literally. Though the solution to the latter part is not to create fake rules with such literal numbers to make an otherwise obvious point
This is very true which is why athletes with certain body types will simply never win in certain roles. With that said, I'm pretty confident that a swimmer with the wrong build could, with lots of practice, get within a small margin of a gold medal olympic swimmer that has the right build to win. Without that practice they would be laps behind.
The studies discussed in the article state that practice accounts for just 26% (violin playing) and 18% (athletic performance) of what counts for being elite.
So no, the whole point is that a swimmer with the wrong build wouldn't get anywhere close to gold just with lots of practice.
The article seems to be conflating "you need to practice 10,000 hours to be an expert" (which is debunked, per the study) with "you need to be an expert to enjoy something."
That's a little odd to me.
You shouldn't feel compelled to become an expert at something just because you enjoy doing it.
As expertise increases, so does the requirement for deliberate practice. On the two ends of the spectrum, there are those who commit to it, and those who don't. Ultimately, individuals who will reach expertise are those who lean towards a high(er) level of commitment.
One can easily observe this in any organized activity. There will be a huge variance in expertise (= skills) among people who have a long experience.
So many people conflate necessary and sufficient conditions. If you look at the world's best and notice that they all practiced for at least 10k hours, that doesn't mean that practicing for 10k hours will make you one of the best.
People didn't just invent that idea. The likes of Bill Gates and Malcolm Gladwell have pushed it as the separator between the haves and have-nots for years.
It's a PR friendly way to sell "We're not billionaires because of luck, early access to computers, or a complete lack of empathy! We're billionaires because you don't work hard enough."
"Which is why I find these debunkings of the 10,000-hour rule to be a complete relief."
Actually, I find the debunking to be the opposite. It basically means, if you aren't already blessed - no amount of practice or training will get you very far anyways.
My observation is society will bend over backward. Do literally anything. Rather than actively embrace what is known about IQ as being true.
Any book offering something along the lines of “just work harder and you will get there!” stands a good chance about being widely adopted and celebrated.
Because it offers a pseudoscience and publicly acceptable theory that is in opposition or replaces IQ.
High intelligence is incredibly heritable and enables people to learn complex skills much faster than others.
If you add in ADHD and Aspergers, I would guess I can out learn anyone but only at topics that interest me.
You are not going to beat someone whose brain literally cannot turn off and who ruminates obsessively about complex topics all day. No matter how much you “practice.”
Not sure how you think “IQ” is particularly meaningful to this, though. At best it measures one small contributor to all the aspects of a person. “Working hard” is also required. So is “natural inclination or ability”. Psychological and physical traits also play a role. Practice is also important. IQ alone is no more likely to make someone successful as any other piece of the puzzle.
But then there are types of brains which are naturally attuned to wanting to solve certain problems. If the owners of those brains get lucky they get a brain that wants to solve problems that they can get paid very well for. It doesn't feel like hard work then. It's like what your brain was naturally wired or specialised to do.
I agree, but you have to realize that very often people call it some other names, like "talent" even though it actually means the exact same thing, it is much more politically correct.
And it is indeed what it is. IQ is very heritable and also predicts performance for a lot of stuff, I would say almost any activity; even for the one that do not necessarily interest the high IQ in person.
But society would rather ignore that because it fears if we agree to this truth something like eugenics will gain power. Society doesn't want to think about the fact that reproduction is not exactly a right but more of a privilege on a finite planet with finite ressource.
In fact, most of our current problems, even concerning the climate, can be traced back to having too many people consuming too many ressources. And people want to reproduce, they don't want to be told that they can't because they don't pass the threshold.
So, we have this giant travesty, that culminates on top of massive hypocrisy that runs everywhere.
And yes, for most people even if they practice hard for over 10K hours in any activity, they are not going anywhere without the "talent" (in other words, IQ high enough for the relevant activity plus some other variables).
You seem to believe that we don't know that, demonstrating precisely what we are talking about.
Nobody ever said that this is a guaranteed thing, and that's completely beside the point. Most things in life are not 100% certain, but we work with "highly probable" or "lots more chance" all the time.
Using high quality ingredients doesn't guarantee a good cake, but if you don't start there, your chances of having a good cake are much lower.
You keep repeating that as if you know something special that others don't, but how it looks to me is that you are the one lacking the depth to understand the original argument.
One thing I don't understand is the argument you're trying to make here. A condition being heritable doesn't make it 51% likely to be genetically determined. For instance, people's environments are heritable, so the "heritability" of "IQ" doesn't say anything about how society should deal with the distribution of ability. It doesn't say we should allocate resources to improving ability and it doesn't not say that. It literally begs the question. Given that: what do you take the parent comment to mean?
So far as I can tell, we've never spoken before, so I'm not sure I follow the "repeating" thing, either.
The repeating part is about you saying basically the same thing to the previous commenter, I don't get how you do not intuitively understand that?
It's not exactly clear to me what you are asking but you seem to be making the traditional nature vs nurture argument, and you seem to think the environment is important enough to impact the IQ of an individual and more largely its abilities.
It's not clear what you mean by people's "environments are heritable" but from my point of view it was barely true in another era but nowadays with increased mobility it is just not true. Most people are very likely to have kids in a very different environment than they grew up in. And that is the case for 4 generations now, so I don't think there is any heritability there appart for a few privileged rich, I guess (then again, they have largely been replaced in the last century or two).
I have to say, I don't understand where you are going because we have a number of twin studies that clearly show that the environmental impact is not very large, at least for a single generation.
There are also plenty of various fields, related to animals or plants that directly work with genetics, to change stuff regardless of the environment. GMO would not be viable if genetics weren't more important than environnement...
I do think environment plays a role and is important for the very long term. But on the human scale, it takes multiple generations to have a significant change in the average so compared to the outsized influence of the base genetic material.
With everything we know about genetics, while there is no guarantee, children are largely a mix of their parents' traits, including intelligence. The higher the parents' intelligence, the higher the chance of this intelligence showing up more in the children (because intelligence is not localized at a single place in the ADN, but is spread out all other).
I really don't understand people contesting that kind of thing, since it never bothers anyone when historically tall people, make, you guess it, taller people than the average. How surprising!
When it comes to the allocation of ressource, it is pretty simple: the humans who get more ressource thrive better, are allowed much more influence on the environment directly and indirectly wich can have large impact.
It is not a hard argument to make that to benefit society the most, you want those ressource to go to the smartest humans available. If you do not, you dilute your benefits and may actually prevent those individuals from taking over because of poor ressource distribution. In the long term your society ends up dumber and worse off.
This is basically where we are and exactly why society is getting restless in some places, it is not a tenable situation in the long term, because even though it may not look like at first glance, everyone is guaranteed to be worse off over time.
I don't know what previous commenter you're referring to.
I worry that you're confusing heritability with genetic determination, failing to account for the ways in which environments are inherited along with genetics, tacitly claiming that IQ is fixed at birth by genetics, ignoring epigenetics, and, of course, working under the assumption that IQ testing is statistically meaningful to the point where the kinds of comparisons you're trying to make are valid.
Your argument seems to condense down to "tall people have tall kids, so smart people have smart kids, so any demographic IQ pattern we see must primarily be fixed by genetics". That argument --- I'm paraphrasing it and apologize to the extent it doesn't capture what you're saying --- is weak in a bunch of places that you're not acknowledging, let alone rebutting. But more importantly, it's not a scientific argument; it's just a subjective, intuitive claim. Your intuitions are different from mine.
I don't know what previous commenter you're referring to.
You wrote this, to the parent I replied to.
tptacek 14 days ago | parent | prev | next [–]
Every time someone here mentions the heritability of IQ, I
wonder what they think that term means, because it clearly
does not mean what most people intuitively think it does.
The "issue" you are talking about has already been addressed ; it is not something that has an effect large enough to be relevant, at least for a small number of generations. We could talk about the very long-term effects, and the necessity of territory/environment control but that's an even more politically loaded subject and it doesn't change what I say.
You seemed even more confused about probability than you think I am about genetics, even if you lack that intuition, it does not make its effect go away.
Then again, since you seemd to have "forgotten" a previous comment of yours, I am inclined to think that you are either arguing in bad faith or don't really understand what you are saying.
I would be happy to argue those things with a trained geneticist but since my argument is probabilistic it wouldn't even matter that much. Research seems to say the same thing I do, abeit with a lot more politicaly correct wrapping.
What effect size are you talking about? The heritability effect size? What is the "effect size" of a simple ratio of genetic variance to trait variance? What confounding effect do you think I'm suggesting to you?
Every time someone here mentions the heritability of IQ, I wonder what they think that term means, because it clearly does not mean what most people intuitively think it does.
I think the 10k rules can be a decent guideline, not something that's written in stone.
Some very talented individuals can get there much faster, untalented much longer - just like with everything else, there's some distribution involved, and the mean might be somewhere around 10k. You'll still have the tails on both ends...
I'm a musician, and I've seen people get "good" really fast. They've learned 80% of the instrument in as little as 2-3 years, and you couldn't really distinguish them from someone that's been playing for 10 years - prodigies like that pop up from time to time.
But I've neve met a prodigy that just half-assed and got there in that time. Every single one that I've met, have been practicing obsessively during those formative years. At my peak, I could play guitar for 8 hours a day - and every good musician I've ever talked with, have the same stories of playing/practicing long hours for some time.
I like to think that if you're talented, you can turbocharge that time. You don't need to spend as much time as others on certain things, and can progress much faster. A good friend of mine has perfect pitch, and learned every note on the instrument in just a couple of days. On the other hand, I've met people that have played for years, that struggle to find almost any note - just to use one example.
With that said, it is the last 10%-20% that takes time to master. That's where all the nuances come in, and what differentiates a master from a good musician.
I've been practicing guitar pretty hard for 4 years. Certainly not 8 hours a day since I have a job and wife, but at least 1 or 2 every day. I feel like maybe I'm not practicing the right things.
When you were playing guitar 8 hours a day, do you remember what you were doing for those 8 hours? Learning new songs? Practicing scales? Transcribing songs? I know all those things are important, but I'm still trying to figure out what I should really be spending my practice time doing.
I think my goal is just to be able to play with friends for fun, no gigs or anything like that, and sound decent, rhythm or lead. And maybe write a few songs.
At my height of playing, I was between ages 14 - 20, so basically from Jr. HS to college. I had all the time in the world, and was really obsessed with the instrument.
I played in multiple bands, so a ton of my spare time would go toward that. Writing and playing songs, basically living in the rehearsal space. When it came to practice it was mostly learning/playing scales, modes, chords, rhythm, etc. - and of course trying to play more and more difficult (cover) songs, as well as trying to write stuff. I spent a huge amount of time just writing stuff in guitar pro (software), and later composing stuff via MIDI in DAWs (Cubase at the time).
My main tips would be to use a metronome at all times, learn as much music theory as you can, learn to sight read, and just have fun by challenging yourself with music. Finding someone to play with is, IMO, priceless.
The author does not know what it is meant by "deliberate practice" nowadays.
Everytime they use it, they should have written "rote practice" or, simply, "practice".
As do most people who quote the 10,000 hours thing. Gladwell greatly simplified K. Anders Ericsson's work [0]. Most people just read Gladwell's catchy headline and miss the point. Doing something over and over != deliberate practice. In Ericsson's words:
>Expert performance can, however, be traced to active engagement in deliberate practice (DP), where training (often designed and arranged by their teachers and coaches) is focused on improving particular tasks. DP also involves the provision of immediate feedback, time for problem-solving and evaluation, and opportunities for repeated performance to refine behavior.
I always find it difficult to square away the critical comments on HN or even studies like this with my own experience.
Focused practice on a specific thing I want to get better at or learn has always resulted in rapid improvement in that specific thing versus just trying to wing it or just haphazardly and irregularly repeating the thing as a part of a larger activity.
What might take days or weeks to learn can take hours if I consciously think about what I'm lacking, what I want to learn and what methods to use to bridge that gap, then continuously reevaluating what I do and how effective it is.
I'm not sure that'll ever turn me into a superstar programmer, but I'm fairly convinced it's the most effective and efficient way to become the best programmer that I can be within my own limitations.
I can always strive to be the best version of myself. How that version fares in comparison with others is of little relevance to me.
The most naturally gifted pianist who ever lived may have never seen a piano. One doesn't need any study to confirm practice, rightly done [1] improves performance, only to attempt to quantify it.[2] The better adage is one I first heard from a Little League coach: "Practice like you play."
[1] Practicing incorrectly can lead to bad muscle memory that can hinder performance. E.g. learning to touch type after years of 4 finger hunting and pecking v learning how to touch type the first time you start using a keyboard.
[2] Sitting in a waiting room years ago I read an article in a science digest of some sort extolling how physicists explained why the air smells differently after rain. They used time-laspe photography, zoom, and optical spectography (???) Or something to capture spores being released into the air. I was baffled. Once you know of the actual existence of molecules, spores, and the like why would you need to prove to your eyes what your nose already tells you. If you think it's a bad example I am not knowledgeable enough to say you're wrong, but I do belive humans in general have a disposition to more readily accept evidence received by the eye than we are any other sense except touch.
I've never heard it said that after 10,000 hours anyone can be great at anything. I've always heard it said that it takes 10,000 hours to master a field - whether that be a subject, aport, art, or what-have-you. In this context "master" just means you have all the basics, you're competent and people would recognize your competency. That's a looooooong way from being great.
What is even the thing you are supposed to reach after 10k hours? I would guess that it is essentially impossible to reach world-class performance in anything by practice alone, there will always be gifted people with a head-start over the average person. But you can certainly become quite good at something with enough practice. But also probably not at everything, there are almost certainly things that you are just not made for, where no amount of practice will make you really good.
And all things considered, 10k hours is not that much, it is about five years in a full time job. A developer with 10k hours will not even have 5 years of job experience if you take into account time programming during school and university years if you start programming as a child or teenager. With that you have a good chance of not being a bad developer but less than 5 years of professional experience will almost certainly not make you a really good developer unless you started out as one of the rare gifted ones.
Malcolm Gladwell is mostly to blame for the misinterpretation of Ericsson’s original publication. The effect size of deliberate practice can be refined through tighter study methodology; that’s fine, it’s how science works. But the 10k hour heuristic was just wrong from the beginning; and Gladwell and Ericsson had a very public disagreement about this. Gladwell is a great storyteller but he’s no scientist.
Part of my professional activities involve coaching young chamber music groups and this idea of “10k hours of practice is all you need” is rampant in the music education community. There is just something qualitatively different about the great soloists that most can’t acquire even with 20k hours of deliberate practice. And that’s ok. I tell these teens that there’s no limit to “better.” And if your goal today, the next day and the day after that is “better” you are already on a track to attain more than the vast majority of humans.
> Which isn’t nothing. But it also means that a great many other factors — like genetics, personality, life history, etc. — makes up the majority of the difference. “Almost across the board, practice should improve one’s performance,” Macnamara told me in 2016.
> Practice matters, yes. *But at the same time, it’s unlikely to bridge the gap between natural superstars and your average player.*
Exactly.
Some people are just born to be better at a thing than others. Practice and coaching definitely makes you _better_, but things like the 10,000 hour rule are just cruel ways to make people that don't have "it" feel even more down about themselves.
This is especially the case with physical appearance. It's a real shame that the ripped six-pack (for men) and super slender beach body (for women) are the physiques to aspire to. For many (most?) people, getting there requires brutal levels of restriction and determination that some people that are more genetically predisposed to attain and maintain that physique than others.
That physique isn't even indicative of peak performance! In fact, there are many folks that look "perfect" but have horrible health markers underneath.
It's even worse when those people become influencers and try to tell others that will almost never look like them how to do so. Just cruel stuff.
So, yeah, I'm glad we're finally putting this "advice" to bed.
What about the quality of practice? I'm using quality very broadly here to include things like: the content of the practice, good spaced repetition, focus, etc. Perhaps 10k hours of mediocre practice doesn't get you far, but 5k hours of high quality practice does?
No? Being a mediocre hiker is perfectly adequate for a joyful day out with friends. Being a mediocre cook is sufficient for adding joy to your meals. Being a mediocre singer is enough to be a joyful and contributing member of a church choir.
Thinking expertise is required to enjoy things definitely can blind you to the joy that can be found in mediocrity.
It's ok as long as it's not an excuse to not put in the work. It's very often used as an excuse to not even try. Nobody wants to be mediocre, just like nobody wants to be fat, but plenty of people will tell you they're perfectly happy in both of these scenarios.
It's like most things in life, people don't care if you're mediocre or an expert, just don't brag about either
> It's ok as long as it's not an excuse to not put in the work.
Hmm? I'm perfectly happy with my hiking performance that is usually about 30% slower than the estimated time for the marked trails i do. I see no reason to improve it because I don't make a living out of hiking.
Not all parts of your life have to be ultracompetitive.
We probably don't have the same definition of mediocrity then, the way I look at it is that if you go for a hike you're already in the top 25% which puts you way way way out of mediocrity. Most people haven't done anything more straining than walking up a couple of flights of stairs in their last 5 years
Again, nobody in the entire world prays for a mediocre life, a mediocre health, a mediocre salary, a mediocre partner. If you could put in the same amount of effort and get better results you'd accept on the spot. There is joy in things, no matter how good you are at them, but being mediocre at something certainly doesn't bring more than being average or good at it.
> Again, nobody in the entire world prays for a mediocre life, a mediocre health, a mediocre salary, a mediocre partner.
There are plenty of people for whom mediocrity would be a big step up. If you have terrible health, having mediocre health sounds delightful. If your partner is abusive, having a mediocre one looks like a grace.
Your sample selection for "the entire world" seems to be leaving a lot of people out.
Nobody's saying you ought to hope for mediocrity. But plenty of people do.
In fact, for most things that aren't important to me, mediocrity is the level I hope to attain. Sure, I could probably achieve higher, but if it's something I don't value, putting in that additional effort would be at the cost of something I value more. If my achievement drops below mediocre, it's usually bad enough to cause problems, so mediocrity is my "not very good, but not bad enough to be a problem" level goal for things that don't matter to me.
> We probably don't have the same definition of mediocrity then
For a hiker, I'm worse than average :) Since the estimates in our mountain guides are 'for the average person'. Of course I'm better than someone who just walks to/from the car.
As anecdata I'm pretty sure I did a 3 hour estimate hike up a mountain in 1.5 to 2 hours when I was 19, enthusiastic and in top shape. So they really are averages.
It's ok as long as it's not an excuse to not put in the work
After a point, it isn't about 'putting in the work'. If you are a mediocre cook, that's good enough. No need to put in any more work than that. You'll enjoy your food well enough and be able to feed yourself and your family healthy meals.
There is a lot to be said about something being good enough and not put any more effort in. This does include things like jobs: It doesn't matter much if you excel at mopping a store nor put the work in to be better: All anyone needs is good enough and effort is better used elsewhere.
And I'll brag if I want to. I generally don't, but I'm not going to say I suck at (hobby) when I don't believe it either. Society does not benefit by understating abilities.
Good enough isn't mediocre though, mediocre is lower than that. Mediocre is, by definition, insufficient, unsatisfactory, &c. Mediocre is what you use when something is about average but you expected better, you cannot simultaneously expect better and be content.
Mediocre is "just good enough not to be so bad it causes problems." And for the many, many things in my life that are not important to me, "bad, but not so bad it causes problems" is, in fact, "good enough."
I do. At least in some areas of my life. Maybe the issue is in the loaded word “mediocre.” It has a very judgemental connotation, whereas “good enough” is an assessment of satisfaction with one’s level of attainment. Professionally, I’ll never be satisfied with a given level of knowledge and ability; but in other realms, mediocre (good enough) is perfectly fine. In the kitchen, I’m mediocre. And I don’t try. I can’t imagine a life where every pursuit had to be pursued with equal striving toward expert status.
Wanting to be mediocre and enjoying being mediocre are different. League of Legend players want to be amazing. I don't think it makes them enjoy the game.
I feel like wanting to enjoying yourself and wanting to win competitions are two completely different drives that are even antagonistic.
I think this can actually be a very good idea - this is a thing often discussed in rock climbing, where for most people, no matter how hard and long you train, there will simply always be someone (most likely a 13yr old child) that is just incomprehensibly better than you. So I am glad that I find profound joy in my climbing and trying the hardest I can, while still knowing that globally I will probably always be mediocre at best.
Much better link to the full text of the actual replication study, without the vox-splaining: https://royalsocietypublishing.org/doi/10.1098/rsos.190327
From the abstract: “Overall, the size of the effect was substantial, but considerably smaller than the original study's effect size.”
Regarding small sample sizes: “ The current sample was obtained from over 4 years of recruiting efforts (summer of 2014–autumn of 2018). While our sample size is 33% larger than Ericsson et al.'s, it is still small. Conducting expertise research, by definition, means studying a small subset of the population, thus there are relatively few participants from which to sample, making large-sample replications nearly impossible. However, if replications of expertise studies never enter the scientific record because of small samples, then we probably will never provide additional evidence to support or refute the original studies. This leaves only the original study (with an even smaller sample size) as the only indicator in the scientific record.”
“To be clear, evidence from small samples—in original studies or replications—should be interpreted with caution. However, when the original study (with a small sample) has already entered the scientific record, replications with similar sample sizes (since that is all that is feasible) should also be allowed to enter the scientific record rather than be suppressed from publication. This is probably the best way to accumulate knowledge in this area. ”
Oooo! And they link to actual data they collected in the study! https://osf.io/4595q/
Fun little chart from midway through the paper: https://royalsocietypublishing.org/cms/asset/bc80eddb-371e-4...