Hacker News new | past | comments | ask | show | jobs | submit login
The Problems with Deliberate Practice (2020) (commoncog.com)
150 points by ZephyrBlu 86 days ago | hide | past | favorite | 67 comments

I'm not really sure what I'm supposed to do with this article - it seems written as a hit piece on Deliberate Practice, but doesn't live up to that by more or less pointing at various works and saying, "That seems crazy!"

Anders Ericsson passed away in 2020, but fortunately in 2019 he wrote a comprehensive article describing his journey through the discovery of deliberate practice and how it evolved over the ensuing 25 years. I would suggest reading this in addition to the submitted article, if you're interested in better understanding the "Problems with Deliberate Practice":


Edit: The author has this to say: "Is deliberate practice good enough to work? My opinion — one that’s informed by painful experience attempting to implement the results of Ericsson’s research — is that, yes, deliberate practice does work, but it’s really difficult to turn the principles into a practice program if you are in a field where no ‘highly-developed, broadly accepted training methods’ exist."

Edit2: I was too harsh, edited to tone down the language!

I think you'll be surprised to learn that most applied expertise research in the years since DP was first discovered has focused on training methods that do not depend on mature pedagogical development, which DP does. This is the primary point that I make, and realising this was what led me down the alternative path in the year since I wrote this post.

In sum: DP simply has too high a bar to clear. Want to do DP? Well, has there been a few decades of pedagogical development in your field so that you can be coached like with chess or math or tennis? NO? Ok, you're tough out of luck. No DP for you!

So the real question is: if you are working in marketing or management or leadership or computer programming, all of which are domains that matter to our careers, these domains do not have good pedagogical development — so how do you get good?

The answer to this, at least from one branch of expertise research, is to extract mental models of expertise from the heads of domain experts, and then turn those into training programs. This is a far more tractable training approach. See: https://commoncog.com/blog/accelerated-expertise/ for a summary (note: Paul Feltovich is one of the founders of the field of expertise research, and a contemporary of Ericsson; he's one of the authors of the book).

For a full survey of the approach, see the Oxford Handbook of Expertise https://www.oxfordhandbooks.com/view/10.1093/oxfordhb/978019... (as opposed to the Cambridge Handbook, which is the Ericsson school.)

As a self-taught web developer switching careers in his 30's I was interested to learn more about deliberate practice. I remember my extreme disappointment when I found out that it's only applicable to well-developed fields where there are established pedagogical practices. It's not that the common-given advice "just build stuff" isn't true, I just wanted to take the extra mile and choose projects in order to maximize my personal growth.

Finding someone who went through the extra mile to research for ways to apply DP in the relatively unexplored territory is such a breath of fresh air so, thank you for this!

> In sum: DP simply has too high a bar to clear. Want to do DP? Well, has there been a few decades of pedagogical development in your field so that you can be coached like with chess or math or tennis? NO? Ok, you're tough out of luck. No DP for you!

I actually read "The Cambridge Handbook of Expertise and Expert Performance" and if i recall correctly it's not that simple, , Expert practitioners can develop themselves better methods (it is argued the reason world records keep getting broken is not because our genes become better but because of better training methods ) , the main thing about deliberate practice is detecting points you can improve (e.g. failing to answer a coding question in a job interview), developing a method to improve the sub skill (trying at home with less stress and more time, going over a piece of knowledge you forgot, e.g. how BFS works), going over "socially accepted good enough" to "the best you can be" (e.g. you might fail some interviews and think maybe the interviewers don't like you, or you could double check your answer and find better solutions which might help if you will interview to more demanding companies such as facebook or google).

There is another interesting book called "the psychology of problem solving" (see http://www.al-edu.com/wp-content/uploads/2014/05/Davidson-St...) where Ericsson argues iirc that deliberate practice is a form of problem solving.

You mean these? https://i.imgur.com/2tWApHA.jpg

I've not done more than cherry-pick from them, admittedly, but I've always had a pipe dream of writing articles about concepts within, as a way of teaching myself.

I don't think practicality is a necessary component of truth. It may be true that, should you stumble upon a way to deliberately practice something, you will grow your mastery of that thing regardless of talent, and it may also be true that finding a way to deliberately practice something is nearly impossible without decades of effort.

What I believe you're talking about is called "tacit knowledge" and the work around how people can effectively transfer that knowledge from one person to another. One classic anecdote is of the pilots who were able to find success by "pretending" to be their instructors -- doing exactly as they thought their instructor would do, without really knowing why, and obtaining more positive evaluation results than by trying to do everything "as themselves".


I've actually been putting some of the CTA methods mentioned in the Oxford Handbook to practice! See: https://commoncog.com/blog/john-cutlers-product-expertise/ for one accounting; I'm currently trying it out with an international Judo coach right now, on technique analysis. (We can't meet due to COVID, since I cannot fly to him, so what I want is to extract his ability to perform video analysis, so that I can perform it on myself and on other competitors).

If I were to summarise why the CTA/Oxford Handbook approach is more tractable, it is this: you don't 'stumble' into a way to deliberately practice something — you do 'pedagogical development'. But pedagogical development is hard! One of the things that I've found most perplexing about Ericssons's work is that, in Peak, he provides an accounting of a skill domain (memorisation competitions) that experienced pedagogical development during his time. Thanks to that pedagogical development, DP techniques could be used in the domain. But he spent no time talking about the pedagogical development!

So I've concluded that it's simply too hard to do good pedagogical development. Better to adopt CTA methods, which already work, and use those to design training programs for myself.

And while it's too early to tell (check back in a few years) I think it's more fruitful.

Okay, I'll admit I was too harsh initially, you and I are certainly both interested in this topic, and I suppose I'll keep reading as you write more about your exploration here. :)

Good luck, if I ever find more free time I hope to do a lot of the same work you're doing here, I find this topic extremely fascinating!

I just read that review of Accelerated Expertise. I didn't know which link to click in all of this comment thread so I just picked that one at random, and -- wow -- that was one of the most interesting things I've read in a while. Thanks for posting!

Thanks for the additional explanation! Could you give a short definition of what you mean by "pedagogical development"?

Pedagogy means 'methods of teaching'. Pedagogical development means that there is a some body of knowledge around how to teach the skills of the domain. For instance, in tennis, there is some understanding of how to teach serves, and a coach may be able to break down your technique into subskills and assign you specific exercises that are known (have been developed) to work on building those subskills, before building it back up to the full serve.

This body of knowledge is built up over time, usually through trial and error. It is then passed from coach to player and coach to coach.

The dirty little secret of DP is that it cannot be done in domains where no good pedagogical development exists. This is a definitional thing. You may read more about it in Peak, Ericsson's popular science book on the topic (summary: https://commoncog.com/blog/peak-book-summary/)

I want to repeat this, so it is very clear: you cannot do DP in a domain where there is no good pedagogical development.

So: rock music, jazz, some aspects of computer programming, software architecture, marketing, leadership, management — all of these are domains with under-developed pedagogical development. Don't get me wrong — there are experts in all of these domains. But they got there through trial and error and reflection, not DP (see: cognitive transformation theory for an explanation of how that occurs here: https://commoncog.com/blog/the-hard-thing-about-learning-fro...)

If you want to do DP, you would have to come up with training techniques for all of these skills, and test them over a long period of time. In other words, you would have to do pedagogical development yourself before you can do DP.

One other implication: anyone who says "oh, just go do DP" is likely someone who has not a) looked at the research closely, or b) not actually tried to put it to practice.

Applying DP is a lot harder than you might think.

Anyway, once you understand this, you may now pursue two different lines of inquiry:

- How do good coaches do pedagogical development? I recommend John Danaher to give you a taste: https://youtu.be/ktuw6Ow4sd0?t=2396 (on developing new techniques) and https://www.youtube.com/watch?v=ktuw6Ow4sd0&t=6211s (on his creation of the leg-lock system)

- How do you develop training without pedagogical development. And for that I recommend my tacit knowledge series (https://commoncog.com/blog/the-tacit-knowledge-series/) and my summary of Accelerated Expertise (https://commoncog.com/blog/accelerated-expertise/)

Thank you for that first link, a really interesting read and timely for me. I will probably buy the book

Incidentally, Ericsson's 2019 paper cites a Hacker News comment. We've come full circle.

'10,000 h' succeeds as a meme because it tells people what they want to believe, that with enough practice, anyone can covet the skills of genius. It’s not so much that people want to become world-class musicians or top physicists, but rather that they have the potential to become those things if they want to, by practicing enough

Hacker news, (2017). Hacker News Post by "Paulpauper". Available at: https://news.ycombinator.com/item?id=13855422 (accesseed March 4, 2019).

Thanks for sharing this it's a very helpful summary on deliberate practice.

“ First, it seems improbable that in the narrow area of expert performance we find that practice dominates, whereas in everything else, we find that genetic factors have an equally large (if not greater) effect on outcomes. (This latter bit — that genes matter in determining IQ, mental health, personality traits, aptitude, and others — is so well replicated so as to not be in serious dispute any longer.”

Now that’s a pretty huge claim, and I don’t think that linking Guardian articles backs up that claim as scientific consensus.

Neurons + culture (aka, practice) are essentially the biological solution to DNA being too slow. It’s not “improbable” that practice dominates here, in fact, I’d say that our genetics specifically evolved for practice to be more important there.

I've had similar thoughts on this.

In sports, most people accept that genetics plays a huge role (if you want to play in the NBA, having tall genes certainly helps a lot). Why wouldn't they also matter for intellectual pursuits? If the effect of genes is something we can easily see with our own eyes - arm length, bone width, etc. - we have no difficulty accepting it, but if we can't see it - certain type of brain structure, certain hormone profile that favours some activites over others, etc. - we want to believe that genes don't play a role.

Besides, the role of the genes can also be a second order effect. The book Sports Gene by David Eppstein tells a story of sled racing dogs that were bred for motivation to run (apparently it can be bred). Naturally dogs that like running also want to train more and become better at it. Could it be that the motivation to do things like mathematics or programming are also heavily influenced by genetics?

Well, the consequence of what you're writing is that if you're somewhere in life where you don't want to be you're pretty much out of luck 'cause genes (a little exaggerated, probably more true for future dreams).

One word: play.

I didn’t go to Unix school or do practice reps of building networks. I stayed up late nights building weird and sometimes nefarious projects.

Similarly, I have mediocre results when my schedule only allows BJJ classes but markedly better results when I can just go “play” Jiu jitsu with others.

Just go get reps doing something you love, in an experimental and playful manner.

The standard BJJ classes (warm-up, technique of the day, sparring) are pedagogical black holes (BB here), anything else decent enough is better.

Can you provide some suggestions on improving (or just pointing out what makes these pedagogical black holes)?

I am a MA practitioner myself (Aikido, 32 years and counting) and while I am always interested in the pedagogical aspect even if I rarely held a lesson myself.

(And yes, I know that BJJ is very different from Aikido, but at the end of the day, I see lots of other arts - including Karate, Judo... Aikido etc. - being taught as warm-up/a handful of techniques from the curriculum/sparring...).

A lot of martial arts training is focused on practicing techniques in isolation and developing "perfect" form (punches/kicks in the air, doing the same throw repeatedly against a cooperative partner). The problem is that any technique you want to apply against an opponent in a competition (or self-defense situation) also has a significant tactical/decision-making component, which is not really trained by that approach. You have to learn when to do a technique, how to set it up, and how to adapt the movement to your specific opponent. You can learn the tactical skills in sparring but it's difficult because there's so much going on.

So when you try to jump from isolated technique practice to sparring, there's a disconnect because you haven't built the tactical skills. Even worse, the movement you practiced in isolation may be different from what works in sparring, so all that isolated practice can actually be counterproductive. You can see this pretty clearly in combat sports like kendo and fencing which have a strong component of tradition -- the movements of high-level competitors tend to be very different from the "correct" form taught in class.

I'm not sure how much this would apply to aikido, which as I understand it is not particularly focused on competitive or self-defense applications.

If you're interested in pedagogy for martial arts (or sports in general), I highly recommend "Development of Technique and Tactical Skill" by Luis Preto and Spyridon Katisgiannis.

"Principles-Based Instruction for Self Defense" by Rory Miller is also pretty good, if a bit rambling.

Thank you for the pointers. I have already put the Preto's book in my Amazon basket.

Yes, I am aware that Aikido is radically different from most other martial arts (some dojos do practice a sort of "freeflow randori" but it's usually just having many ukes throwing attacks at you, and there is not much resistance to your techniques... and in any case it never really dominate the lesson in terms of time).

Regarding your answer: personally I think this is a sort of paradox which cannot really be solved when teaching "how to fight": if I go to a "translator school" I will dedicate lots of time in actually translate texts. There will always be a practically infinite quantity of German texts to pick up, and I am expected to always tackle each with full energy and apply any trick I know. At the same time, there will never be a German text that will sue me from its hospital bed, or send the police to arrest my teachers and close down the school on Manslaughter charges.

Even if we look at fighting arts/sports that were designed to freely apply full force (e.g. Judo) what you learn there is targeted to face an opponent which is wearing Judo Federation approved garb, is alone and is in your same weight class and probably level of expertise. And there is at least a referee around, fight happens on a sort of "consecrated ground" etc. etc.

It's been a few years since I read it, but Josh Waitzkin's Art of Learning discusses how he went from complete novice to winning the world championship in Tai Chi Push Hands within two years.

It's a pretty good memoir for anyone interested in chess, martial arts, or just learning in general.

It is an interesting book and I know people who know him since they trained in the same gym, but along with the solid substance (meat) that he proposes, my impression is that he throws a lot of smoke at the problem (see for example the few interviews he gave to Tim Ferris). He is a smart guy and the book is one I recommend, but he overcomplicates what should be pretty straightforward.

I met many high-level, sometimes world-level practitioners of various sports and at the risk of coming across as glib, I have yet to find one with higher-than-average intelligence (there are, I just have not met them). That's why when some NFL or NBA players are presented as semi-savant, it makes me chuckle and want to say to the interviewer or commentator, "that's what you want to believe, not what it is. They are simply much better, but very rarely in an intellectual way".

Yeah I get what you mean, but it's also pretty rare for people to achieve what he did in martial arts in such a small time frame.

And the specifics he outlines about how he rapidly improved to the point where he could compete at an international level will surely benefit anyone looking to improve at their art.

This is for the common gym (and follows what was developed as class structure in the early days in Brazil), not for the more enlightened gyms that exist but they are not (or not likely) for the common person. Say, for 100 gym, 90 are set up the way I describe above.

The technique section is like going to a seminar whose topic you didn't choose and whose content you won't likely see again for a few months or years. Like, today we do partial differential equation, and we will do them again in 6 months from now (apart from some of you who like them and decide to practice with no support).

There is very little effort to present the common processes that make a technique work or not. For example, in ML you have broad categorization of approaches, such as supervised/unsupervised/reinforcement learning. Within supervised learning, you then have classification/regression. Then within those buckets, you have linear, non-linear. Within non-linear, you have bagged or boosted or neither. They share commonalities at different levels and they help the practitioner learn the techniques much faster and then apply those technique in a creative way.

Likewise, in mount and side control you have similar need for hip control, shoulders on the mat, diaphragm under pressure for the more pressure-oriented jiu-jitseiro. But more often than not, the side control technique of the day is presented as just the technique of the day with very little connection to other techniques or common processes that make the technique work. A waste of potential deeper learning.

More often than not, the instructors/coaches/professor love their voices and presentations way too much -- 5 minutes of explanation regarding some details of little relevance while students are getting cold and unfocused, then 5 minutes of drills that take 3 minutes to start because students have forgotten what the coach has shown (and you can imagine how long it will remain in the mind of the average practitioner after the class is over).

Then, we have sparring, which in JJ is for some reason way too high stakes. In judo, you see top people getting thrown or give points all the time in randori/sparring with no ego bruised (well, more in judo than US wrestling, while in Russia for example you see more learning-oriented sparring, in the US it is grind, grind, grind), but in JJ, especially after people get their blue belt, they absolutely do not want to get chocked, arm-barred or swept by equal or lower belts. This creates defensive games that make thee knuckles white, but give little back in terms of improving the game of the practitioner.

Positional sparring is pedagogically much better than free-sparring, since the same position or situation is presented or needs to be faced multiple times and that creates lower stakes, less defensive sparring and more "learning". That's the drilling that matters (and see Dave Alred's repair sessions https://leadersinsport.com/performance/dave-alred/) and what for example Roger Gracie, the best JJ sport practitioner of all time) does.

Many other points to discuss, but the main point is that coaches are not trained in coaching, they have no idea how to present ideas or concepts (clearly, there are exceptions and they are very often the most successful gyms, although there is selection bias at work), they are "just" good at JJ and they went into coaching.

Some might say: but some people are getting very good at JJ. I agree, but "sink or swim" is a selection approach, not a teaching one.

I've actually tried to frame my Computer Science instruction style based on my martial art experiences (BB Aikido, Purple BJJ). How I view it is that CS and MA share many properties in different environments. Both require a need for understanding technical skills (for loop or an arm bar), areas where to apply these skills (simulations or triangle->arm bar transition), and finally have an element of "problem-solving" (real world coding or sparring).

Most of the issues I see with Deliberate Practice criticism are the 10k hours rule and limitations in training "problem solving". 10,000 hours is obviously an arbitrary number, but signals the importance of time on task. With regard to training problem solving, I don't think we really have a solid "here is how you train it", but instead that we focus on training patterns so we can identify opportune times to apply them. THAT is where I can see DP being useful, since the goal moves from problem solving to lower level Bloom levels (Remembering/Applying those Patterns).

how is sparring any different from just "playing" jiu-jitsu?

the wikipedia page on BJJ ( https://en.wikipedia.org/wiki/Brazilian_jiu-jitsu ) claims that live drilling ( https://en.wikipedia.org/wiki/Aliveness ) plays a major role in training and development, which suggests that it should be part of BJJ classes, but the way it is described, suggests that it is more similar to "playing", which, apparently is not part of BJJ classes. so how is sparring done that it doesn't have the desired effect?

i also wonder why other sports (in particular other martial arts) should be any different.

ps: i take it that BB means Black Belt?

Not a black belt myself, but agree with GP. Sparring frequently tends to be "people trying to murder each other on the mats in pursuit of their egos". There's a level of dynamism in good sparring that can immediately get shut down if one or both practioners are overly motivated by "winning". You need to take a lot of risks to learn in practice even if it means getting clowned by someone who you could beat if you had to, but you do see people sticking to conservative games in order to dominate. Which often doesn't teach them (as they aren't learning new aspects) but it also doesn't teach their opponent (as it's very hard to learn new stuff when trapped in someone's well-rehearsed "A Game").

The advantage of positional sparring (i.e. go from one position and stop when someone gets out of it to some outcome - not necessarily a tap) is that everyone gets to sharpen a single position. The ego motivation isn't as strong there as it is with full sparring.

Yes, that's my approach when I am teaching. Positional sparring is way superior to free sparring for learning. Lower stakes, people are more relaxed, the same position is presented multiple times and that's where learning happens.

And I explain it to students beforehand. I say to students: there are interesting games going on in positional sparring and one is that who is trying to escape (say, from mount) is in a better mental, but not physical, position. You are already mounted, now if you "lose", that's ok, but if you "win" (that is escape the position or finish), it is a big win! And for the who's in mount, finishing is ok (already in mount), but letting the mounted escape is a big "loss". Winning when all, including yourself, know that anything else than winning is a bad loss, is being in the "ugly zone", like Dave Alred says, high pressure, low margin for error, when things are not pretty. But that's where one learns (it sounds like it is high stakes, but it not, because the activity/drill is repeated multiple times).

I do positional for takedowns, side control, mount, from flat half guard, favorite guard (and I say, this is your favorite guard, you should be hard to pass!).

I like this approach. I always wanted more positional sparring and less free sparring. I've noticed that high pressure "expected to win" situation (which, frankly, I love in mount/side control as I've got a better game, relative to expectation of a submission, while my back control/submissions game underperforms).

I've noticed the defender from 'bad positions' does better in specific training than they do in sparring in the same position, probably because they can put everything into getting the "W" - while the attacker in sparring would probably be cooking them or switching to a different position.

There are still weird ego-driven things going on here. My least favorite: being on bottom open guard and having people disengage and run around to win - it's like, bro, if we weren't positional sparring I can just stand up and we can see who has better takedowns. Personally I think this one should begin with at least one grip to avoid time-wasting.

Do you do the "5 minutes on top / 5 minutes on bottom" approach? Or alternating? Or "3-5 people down with a line of people coming in". I think each of them has good/bad elements (without "3 wins and you're out", the latter tends to reduce to the top 3 students annihilating the newer people and not facing each other :-) ).

Yes, I have found positional sparring to lead to the quickest improvement from blue to black (novices have different needs, typically)

"I've noticed the defender from 'bad positions' does better in specific training than they do in sparring in the same position" - It varies, but my rationale is that when somebody is passing your guard and then mounting you, you are typically dog tired and mentally you go from bad position to worse. In positional sparring, the guy in a bad spot starts from that position, so mentally is in a good place, and not as tired or positionally compromised.

"My least favorite: being on bottom open guard and having people disengage and run around to win" - Agreed, like typically new guys not engaging and wasting a couple of minutes by running around. In that case, I simply say, "come here my friend, let's practice", and they start getting in. And the same when I am the coach, I just encourage a more low-stake environment (which does not mean low effort or low quality) and I must say that I never got any pushback. "Let's practice, that's how we improve".

"Do you do the "5 minutes on top / 5 minutes on bottom" approach?" - I typically do two bottom and two top for each position (one top and bottom with one person and one top and one bottom with another), with rounds of 2 to 3 minutes, encouraging students to push the pace (2 minutes go by real quick and I don't like wasting opportunities to improve). My coach likes the line of people getting in, winner stays in, and 3 wins and you are out.

I think that both have a place, but I have prefer the two minutes round because it creates good competition (you pass my guard, we restart, now I am trying to avoid that same pass, while with the line there is not this immediate opportunity of a comeback).

Thanks, good thoughts there. I'm off BJJ for a while (lockdowns in Australia), and won't be getting to decide how classes go for a long time if ever, but it's good to hear people thinking about it rather than blindly replicating "what I saw in Brazil once".

I've only been training a little over 3 years (definitely not a BB), but maybe the BB meant there are many different types of sparring. Many (most?) schools warm up, show a move, (maybe drill it), then let everyone spar away. The problem is when sparring the person with more skill will severely limit the positions the less skilled person sees. If the less skilled person happens to be very strong or athletic they can in turn limit the positions the more skilled person sees just by holding on, which turns the sparring into very limited training value.

For me, what has helped really learn is drilling. I know lots of BJJ people hate it, but I like winning and drilling works. Even those who hate drilling will admit it.

Positional sparring. Start in some position and work from there. I purposely let less skilled people start in advantageous positions so they get to work from a dominant position (where most submissions come from), and I get work escaping from bad spots.

Finally, flow rolling, which at higher training levels doesn't look that different from sparring. But, it's sparring where both sides give and get so that each sees many positions. At my level we also move much slower. Think about how hard it would be for someone to learn Chess if all they played was bullet chess? You need that bit of time so that your mind can start to build patterns, and then speed up.

I'm fortunate that I lucked into a school that teaches a bit differently.


I should add that free for all sparring still happens, it's just not the focus of every class. Open mat days and end of class is often when people go more free for all.

I don't think this is the correct model nor should we try to generalize it. Deliberate practice works by systematically forcing us to continually engage in the part of the activity that we suck the most at. That is the place where we can most readily improve and addressing it is consequently the best usage of our time (disregarding some side effects and constraints). This sort of engagement requires a brutal degree of discipline and integrity to accurately assess performance. Most people are unwilling or incapable which is why teachers are valuable.

I did martial arts for years, but never quite deliberately. Its easiest for me to talk about music where I put in 3-5 hours a day for years - its so fucking hard to hear what you are playing instead of what you intend to be playing or what other performers have played. Music lives in your head more than the real world. After a few years, undirected practice was stealing my time, holding me back.

> Just go get reps doing something you love, in an experimental and playful manner.

This may have addressed some problems you faced (maybe the practical application of otherwise the Platonic ideal technique), but that is a specific problem. I'd wager the more common problem is one of mindlessly playing with no regard for performance because that's the easy trap. At the end of the day, training BJJ to improve is an entirely different endeavor than going to class 2-3 times a week (though both are perfectly reasonable activities of course).

I agree that 3 hours on hannons (pattern excercices) on the piano a day might hold you back. But that's not playful. I built my own metronome. Disasemble the piano! Really. I admit that I will never be a good pianist, but have made good music, with others, on keys. It is really difficult to be mindfull and conscious of your production. I try swapping modalities. Loop record some chords on top of euclidean rythms... Swap to the drums. Obviously the dilettante approach can also be a dead end. But I've been learning music, more and more this way for decades. I'll never be a virtuoso, but have recordings where I think, wow, was that really me? Does help to play with others who help 'carry' you. I have a similar approach to software.

PS. I did have to save for the basement studio.

>I didn’t go to Unix school or do practice reps of building networks. I stayed up late nights building weird and sometimes nefarious projects.

This is how i learned Linux/Unix, AWS, bash scripting and all sorts of things.

And then .. one day I found myself ... older. Work would kick my ass and I had spent my 20s eating like shit because i spent too much time sitting at a computer.

So time once used for computer play became gym time, jogging time, swimming, hiking, yoga, lifting time.

Then a little older, and starting to buckle down the finances because we got married and now wanted a house so we stopped going out so much and started staying in more and watching movies, playing board games. The money i once used to build new computer or have spare machines or pay for EC2 instances and ELBs and S3 buckets, that all dried up.

Then i have a house to take care of - which takes time.

Then i have a kid to take care of.

All of a sudden I'm 40, with a house, a kid, a 8-10 hour a day job, i get up at 6am and go jogging.

The stress of life has boiled into an anxiety problem - so i can't over caffeinate anymore (which used to keep my ADHD at bay, to an extent). I have to exercise daily to keep in shape and htat anxiety at bay. I'm getting older and stressed out and maybe i need to eat healthier and spend more time in the kitchen making sure I'm getting whole, nutritional, well balanced meals and not a lot of quick, processed shit. This makes my child's health better too.

And now i have no time for "play". And when i have time for play - one more computer screen is the last thing in the world i want to look at.

and now i understand how STEM basically exploits young men who are willing to spend 8 hours a day working, plus a couple hours of overtime daily or weekly, ontop of all the free training they're constantly giving themselves via "play"..then looking at people 20-30 years their senior and expecting them to keep up.

What you love - also just changes. Work is the best way to ruin any passion. The toxic co-workers, the preachy HR departments, the exciting projects ruined by bureaucracy, the small companies that expect too much of you, the large companies that lock you into narrowly defined roles and treat you like a number, the endless meetings (zoom or IRL), the jam packed calendars to the point you don't know how you're going to have the time to ramp your brain up to take on something complex.

You start to realize.. hey, i like gardening. I like making furniture. I l ike writing poetry. I like playing video games with my kid or teaching him to swim or ride a bike. I like taking care of my yard. I like just... putting a pair of headphones on and listening to a record, carving time out to read on a regular basis, maybe catch and old movie with the wife after t he kid's asleep. I like some fine dining, or a couple glasses of red wine 1-2x a month.

All the things that bring you the joy that computer-play once did, do absolute fuck all to further your career.

Then what? what do you do then?

I see people talking past each other here and I suspect it has to do with the level of achievement one is picturing.

Many smart (and talented) people work really hard at chess with good methods for many years and don’t become grandmasters. Talent must play a role. On the other hand, probably anyone can become much, much better than the average person on the street.

So people saying talent matters, and people saying deliberate practice alone can ensure success, can both be correct for slightly different measures of success.

As the article says, the very top levels of any prestigious competitive field have people with both talent and good training methods. Talent can separate people at the very top. That doesn’t mean that becoming very good is out of reach for less talented people.

I'd like to share a view on this that I found for myself, I am curious if anyone finds it interesting. For years I thought of talent as an effort multiplier. "Intelligence is speed" etc.

I recently remapped it in my head slightly and migrated to a definition that talent is a degree of competence and productivity which you can achieve in a given field "effortlessly".

Let's say you are a top 5% software engineer off high school, because of your specific circumstances of nature and nurture. You will still need hard deliberate work, introspection and patience in overcoming plateaus if you want to climb the remaining 5% and become better or the best.

So, some people have a head start, some have a very big one, but true improvement in a field is only possible through deep work, and if you can organise yourself to do it, that's a meta-skill for 21st century.

That reminds me a lot of the story of the manager explaining the best engineers superiority in terms of dividends on a higher percentage of time dedicated to work..

I don't really buy into this theory since in my observation a lot of people get sidetracked and get their dividends on skills that cease to have meaning or otherwise miss the important pieces. I.e. they become convinced of a particular paradigm and use it as the wrong hammer on many nails. In authors they become obsessed with what amounts to retelling the same story to the bored audience. Also, people seem to ignore the net zero nature of being the best or even a pro. If you choose some paths every other amateur ends up as a pro by the standards of when you started, with others you spend years training alone only to suddenly be very relevant due to some shift.

I think there's obviously a confluence of nature and nurture which is going to lead to the final outcome in terms of skill attainment.

I think the extent to which genetics varies probably varies a lot depending on the domain. For instance, something like long-distance running depends a lot on physics and biology: how efficient you can be at moving your body through space depends a lot on how tall you are, how long your limbs are, and things like the width of your pelvis. You will still need a lot of practice to reach your full potential, but what level you reach, and how fast you reach it will be heavily influenced by genetic factors.

However, that being said, practice is still absolutely magic. In the spirit of the 80/20 rule, it's amazing how much a little bit of deliberate practice can improve an untrained skill. Especially when developing a skill from absolute zero, a few weeks of deliberate practice can often take me from the realm of absolute failure, to the point where I can actually enjoy practicing it.

I've also sometimes reflected on the fact that skill attainment through practice seems more difficult as one gets older. I think this is almost certainly true to some extent, as neuroplasticity is real, and for instance a 40 year old will never have as easy a time learning a language as a 5 year old. However I think it's also a bit overstated.

When you are a kid, every thing you are learning is relatively new, so when practicing a new skill you have a more uniform point of comparison in terms of how fast you learn things. As an adult, you have certain skills you might have been practicing daily for 20 years and have completely mastered. So when you sit down and manage to play Für Elise for the first time at a piano it seems absolutely rudimentary in comparison, even if you are making real progress, which can be demoralizing.

There's also the factor that adults tend to be more subconscious about failing in public. We've spent years learning how to avoid situations where we look bad, or can't do what we set out to do, unlike kids who aren't expected to be able to do much. It's an approach which is really self-defeating if your goal is to actually learn something new.

Age factor is absolutely real in well studied skills such as chess.

There are no instances of people starting to play chess after say age 25 and becoming a grandmaster say after 20 years at 45. Plenty of people who start at 10 or earlier and become grandmasters before 25.

Lack of deliberate practice is surely not the only explanation.

Starting at age 15 is already considered very late for chess. (Anecdata: I started at 10 and peaked at FM coming close to IM norm once)

Internet is awash with late bloomer examples in various fields (such as writing). However when you look closer you see that the persons in question had serious related practice at an earlier age..

I am yet to find an example of someone taking up a completely new skill after 40 and achieving notable greatness. Like champion of your city at tennis or master in chess, etc.

So my formula is skill = STARTING_AGE X coef1 + GENETICS X coef2+ DP X coef3

Only questions is what values to give coef1, coef2 and coef3

Oh I totally agree. My only point would be, I think older people tend to over-estimate coef1 and under-estimate coef3. In other words, since aging has an effect on speed of skill acquisition, people over 30 might be more likely to blame that for the fact that their struggling and give up on something they could pick up quite decently with a bit more practice.

Also, when you talk about "related practice at an earlier age", I also think this is nothing to scoff at. As I have gotten older I've often found that a very real route to solving novel problems is to figure out how to apply something tangentially related that I already know.

There's also a correlation does not imply causation in this. People get worse at acquiring new skills compared to a younger person but there's lots of things that change with age outside of physical fitness. For instance, a kid delegates all worries outside of studying and games to the parents. Experts at fields might also underperform in some school subjects in favor of focusing in those they enjoy the most, so they can optimize their time even further.

And then we come to the society component. Anyone seeing a 12 year old shipping an app in the app store will omit the fact that the kid might be receiving training from an expert (maybe even a family member) since the age of 4. They won't judge it as "a person who learned programming for 4 years and shipped an app" but "a 12 year old that shipped an app". You might think the difference is semantic but if I were to told you that I was learning programming for 4 years and shipped an app you'd be like "what do you want now, a cookie? You're 30".

People who believe that over a certain age you're not gonna do anything good might consider suggesting software development as an activity to a 40 year old as a hobby, not as a profession. I cannot imagine them telling the same thing to a child. Don't get me wrong. I do believe that physical change exists and that my 15-year old self would run circles around me. But everyone gives advice on the grounds of seeking masters, not professionals. Translate that belief to the hiring process and it's easy to see where ageism starts. Someone might say that hiring a 40 year old who wants a junior developer position is "settling for less" in spite of the fact that your product can be shipped just as efficiently.

> the effect sizes from deliberate practice far dominate when compared against innate talent, genetics, or other factors

We won't settle this in the comments section of HN so I'm not even going to try.

I'm just curious how this is percieved by ordinary pepople in your home region.

I'm from Romania and generally people here see innate talent as a higher limit to performance, and practice is the way to reach it.

I don't know if it's general consensus in the states, but this is my mental model as well.

Interesting, personally I view talent as being more about the effectiveness of practice than a specific higher limit. Though if the effectiveness of practice is low then that obviously places some practical limits on the maximum amount of 'mastery'.

If you want alternatives to DP that work much better, it might be worth it to take a look at:

1. The Tacit Knowledge Series (which argues that methods from Naturalistic Decision Making are more useful when you are in a skill domain with no good pedagogical development) https://commoncog.com/blog/the-tacit-knowledge-series/

2. Accelerated Expertise — which covers methods from the military to accelerate expertise acquisition (and, yes, some of it uses DP principles, but it isn't really DP) https://commoncog.com/blog/accelerated-expertise/

Not to bring politics into a HN discussion, but based on their recent outcome in Afghanistan, I'm not sure if the US military is an organization I would currently turn to for insight.

Military itself should seldom be judged by the decisions of the government. The entire mission in Afghanistan has long stopped being military oriented, and military was present to preserve some semblance of US control.

One could argue that there is sufficient proof now that US application of military in attempting to "democratize" nations has so far failed, but that should not come as a surprise since military is, by definition, not "democratic". Democracies deploy military exclusively to preserve peace (by discouraging invasion, for instance). This is still not a slight against US military, but against US foreign policy instead. Whether that makes it a net win in other (economic, geopolitical) measures for USA is another matter altogether.

As such, I do not think you should discount US military as a top-notch organization with a lot of know-how in building up expertise.

Most of these techniques are co-developed in industry applications, and are seeping out from military research into business. In tech, John Allspaw has done a huge amount to take these training methods and skill-extraction techniques and adopted/applied it to resilience engineering. (https://twitter.com/allspaw and https://www.adaptivecapacitylabs.com/)

Not paying attention to research with military provenance seems rather silly. What matters more are the questions: does this work? Where has it been tried? And can I apply it to my company or career?

Do you think I get an incomplete picture by reading Accelerated Expertise (the actual book, not just the summary) before the Tacit Knowledge series? I like reading books more than blog posts.

I would recommend the series first — AE assumes familiarity with the recognition primed decision making model (or RPD) which I cover in the series.

If you want to read books to prepare yourself, read:

- Sources of Power by Gary Klein

- Power of Intuition by Gary Klein

Before AE.

And then be prepared to download LOTS of papers to read concurrently while reading the book. AE is not written for the layperson.

I have at least a passing familiarity with recognition primed decision making, although I haven't specifically read Klein beyond his adversarial collaboration with Kahneman. Maybe I could benefit from a perspective that is purely Klein's.

I do enjoy a good not-for-laypeople book. Thanks for the information!

“Second, the deliberate practice research makes an awfully attractive claim: that anyone, given enough determination, can become good at any skill. They merely have to do lots and lots of difficult deliberate practice.“

I am not sure if people on deliberate practice agree on precisely that claim, but perhaps they would agree on “Anyone can become better at any skill, given that one does deliberate practice”.

Whether deliberate practice can make you good relative to other people is a separate question that is beyond the scope of deliberate practice. There, one factors in people with different innate and qualities and practice habits all competing with each other. The ones who are at the very top will likely have both innate qualities and good training regiments.

The thing is that self-limiting thoughts and doubt don't help you. Are you good enough? Do you have what it takes? Is this minor failure a sign that you should give up? Will you ever get there? How do you know if you could even get there?

I would rephrase it: If you want to be good you need a lot of practice. The way to mastery is paved with failures. You need to invest a lot of time and energy before you can even begin to evaluate. Don't waste time on too early self-evaluation. Don't waste energy with useless doubt. Your body and brain are very adaptable much beyond what you normally experience.

In the end it's a balance of doing the things right and doing the right things.

I think that a second part of the issue comes from achieving mastery in creative fields, where there are no clear evaluative goal posts for what success looks like.

Talent certainly differentiates the top levels competitive environments when everything else is held equal. However, in these environments, I can see where deliberate practice can move someone who is less talented into the top nth percent of performers. They may not be among the best, buy they're close enough that someone who is untrained or less disciplined doesn't stand a chance against them. Given that the distribution of material rewards generally doesn't follow a winner takes all model, like what you see with sports, deliberate practice can should eventually create significant competitive and material advantages.

What happens, though, in creative fields such as writing, where engaging in deliberate practice doesn't necessarily improve your craft or your end product? A part of me questions whether the deliberate practice that many prospective writers, such as MFA students, do on a daily basis really helps them write a best seller - or even a quality literary work.

The same could also apply to tech startups. Would it be a good idea for early-stage entrepreneurs to do daily coding challenges? Or, over a long period of time, would they be better off trying to consistently create something that they then test on the market?

> Can you become a brilliant theoretical physicist with an average IQ (95-105)?

> [...]

> I have an IQ below 105.

> I also graduated with a degree in Mathematics and received Highest Honours.

> [...]

> And the answer is ‘No'!


> What I’m willing to bet, however, is that you’ve heard of deliberate practice in the context of Malcolm Gladwell’s ‘10,000 hour rule’ — the mistaken notion that 10,000 hours of practice would turn anyone, at any age, for any skill, into a master practitioner.

Ericsson spent a lot of time and page space trying to distance himself from this "rule", but it's clear from his writing that he introduced it.

See his 2007 HBR article (which lists it as a necessary condition):

> By now it will be clear that it takes time to become an expert. Our research shows that even the most gifted performers need a minimum of ten years (or 10,000 hours) of intense training before they win international competitions.

It's a confusing enough situation that Macnamara et al review the many places he discusses 10k hours (or 10 years) in their book chapter The Deliberate Practice View.

Note he says "a minimum of ten years (or 10,000 hours)" That is it might be 20K, 30K or more hours of practice. So 10k hours is more like the starting line, not the finish line that Gladwell proclaims.

> That anyone, given enough determination, can become good at any skill.

That "given enough determination" part is glossed over a bit too easily. Plus, even if it's just an approximation, people radically underestimate how much 10k hours of practice is.

I've learned a wide range of skills throughout my life so far, to various levels of proficiency. One or two things I've been able to achieve what I feel most others would considers an expert level, and have learned a very large number of things ranging from beginner to intermediate.

Anecdotally the confirmation I've found about 10,000 hours is that my area of expertise is something that for years I thought I was naturally bad at. But I got tremendously interested in it, and began studying and practice literally every day for years, and to this day continue to practice and study this area.

As far as beginner/intermediate skills, two typically talent based things I've been able to teach myself are life drawing and juggling. I have awful hand eye coordination, and still learned enough juggling to impress people at parties when the occasion comes up. My hand writing is illegible, I also have awful fine motor skills and I've still been able to create surprisingly realistic sketches.

Of course I'll never be an expert at these things because I'm just not interested enough in them to get anywhere near the 10,000 hour mark. Four ball juggling will likely remain my threshold indefinitely.

The number of things I've gained proficiency in that I had a prior belief and initial experience that they were impossible for me to do is just to high for me to throw out the wisdom of deliberate practice.

Deliberate practice might be hard to define, but at the very least it means "practice that is not easy". In order to do something every day with a reasonably high frustration/failure rate means that you have to have an incredible drive to continue. Most people I've known simply can't endure finishing a day of work to go sit down and struggle with something for 1/2 hour a day. If you want 10k hours you have to do this without cessation, and no end in sight. In my experience no amount of will power can force this to happen either. You can force yourself past the beginner and into the intermediate stage, but expertise can only come from an insatiable drive.

It's also important to realize that expert and world famous are not the same thing. There are many true experts in guitar who are unknown, 10k hours doesn't get you to being John Frusciante.

I agree that 10k hours is often severely underestimated. Maybe if we all collectively phrased it as "2 hours a day for 14 years" the pedagogy surrounding the above might take a different tone.

How do you measure this? You can't take a group of 100 people and randomly assign half to 10,000 hours of deliberate practices.

I always felt like this conversation had different connotations at different levels. Not everyone can be Simone Biles, but we can all probably learn to do a backflip.

Or more to the point, I think the average person thinks they could never be good enough to impress their friends, and this is probably wrong. With much less than 10,000 hours of deliberate practice you can probably impress people. Write music that satisfies your own soul. You won't be Mozart, but how many of us in the modern world want to be Mozart?

The article seems to mix up "expert" with "world class" a bit. I do think that becoming one of the best in the world in any popular field probably requires suitable genes. But, to me, "expert" is more like someone who's usually the best person in the room at something. Quite different goals.

10,000 hours is enough for someone having extraordinary talent to be able to realize a small portion of that and have it come out completely astonishing.

It's also enough for a mediocre performer to reach a plateau early then end up becoming somewhat of an authority without anything astonishing, and with no talent to spare.

It takes a lot of perseverance either way which can truly blur the lines but these are two opposite ends of the spectrum.

I always had an inkling that the "talent is overrated" argument (that a certain amount deliberate practice is sufficient to achieve mastery) is something of a tautology: only those talented individuals, who are capable of such a rigorous training routine, will achieve mastery.

On the other hand, critics of natural talent explanations often point out that too is tautological: someone is talented if they're successful and vice versa. Clearly there aren't proteins for violin bow control, so something happens in between.

Talent doesn't have to be "natural", just uncontrolled factors. Something that is "in the person" at the time of training. This could have come from past experiences, genetic, even things like diet, who knows. The point is, it's still not understood and not controllable.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact