Hacker News new | past | comments | ask | show | jobs | submit login
Mental processes of chess masters revealed how people become experts (2016) [pdf] (pdx.edu)
265 points by lainon on Aug 25, 2017 | hide | past | web | favorite | 61 comments

Related - "Deliberate Practice and Performance in Music, Games, Sports, Education, and Professions: A Meta-Analysis" - http://scottbarrykaufman.com/wp-content/uploads/2014/07/Macn...

Gives some context for the kinds of fields where deliberate practice improves performance.

Your mention of deliberate practice prompted me reinvestigate the guy who was doing the 10,000 hours of deliberate practice to become a golf pro. Unfortunately, from what I can tell, he did get to 10,000 hours, but it doesn't seem like it panned out. Anyone have further sources of information?



He did not get to the 10,000 hours, he had to quit after 6,000 due to back injury.

Source: https://www.theatlantic.com/health/archive/2017/08/the-dan-p...

Survivor bias seems to be a thing in Pro Sports

Not everybody can get to 10k hours without getting hurt and/or managing to work through it

6000 seem like just totally close to 10K for all practical purposes...

What do you mean? It's barely more than half of the goal.

There was a recent HN discussion on it, I believe - I think he got to around 6,000 hours, and was doing really well at golf actually, but then injuries started happening.

Maybe a slightly different topic.. but from what a lot in the running field think, its all about how fast your body recovers.

There are 10,000 amazing marathoners in the world. If you gave them all the same training, they would all be absurdly good, as they got better and better. But 9,990 or so of them would get injured and wash out.

So it seems at least in some sports, your limiter is not how genetically gifted you are the SPORT, but how genetically gifted you are at recovering from training so that you can train more.

So perhaps this guy has the genetics to be a pro golfer, but his body can't take the strain and recover between sessions.. so he would never quite make it.

Well, when you consider that even Tiger Woods had to take an extended break from golf and completely redesign his swing to prevent permanent shoulder and back problems, it doesn't seem to be a 'genetic' thing per se, but rather a human mechanics limitation.

Some sporting activities are just outside the scope of design of our bodies, it would seem. Much like trying to use a Japanese chef knife to prise open jar lids. It will work up to a point, but not being the core design of the tool, you are guaranteed of breaking the knife at some point, or at least severely damaging it.

It would appear that the human frame is not suited to short, explosive, twisting action of the arms and shoulders while the lower body remains planted.

I'm not convinced that the recovery is genetic (well it is, but that's not the point.) People give themselves repetitive motion injuries in music all the time. It causes tons of otherwise amazing musicians to wash out.

It's become such a problem, that it's basically become the biggest issue the pedagogy community as a whole is actively trying to address. This is interesting, because there has been a lot of progress made in injury-preventive technique.

Even in running, a quick Google search revealed that injury preventive techniques exist.

So I guess, I'm not convinced that injury isn't technique related and that not being injured doesn't have more to do with luck (in terms of natural technique) than genetics.

>So I guess, I'm not convinced that injury isn't technique related and that not being injured doesn't have more to do with luck (in terms of natural technique) than genetics.

Certainly luck is a factor, but genetics is an important factor as well. Many studies have linked genetics to a predisposition for inflammation, wound-healing, and other injury-related phenomena.



> It's become such a problem, that it's basically become the biggest issue the pedagogy community as a whole is actively trying to address. This is interesting, because there has been a lot of progress made in injury-preventive technique.

Could you please give me some references for this? Guitar has loads of famous people who encourage people to do things that will cause injuries. Having some actual research to fight against this would be really useful.

An egregiously bad example: Paul Gilbert (who is like 6'5" and has enormously long arms and fingers) actually SELLS an especially long guitar strap that is practically guaranteed to give you RSI. I wish I were joking: https://www.youtube.com/results?search_query=paul+gilbert+gu...

Side note: this made me hugely disappointed in Paul Gilbert. Prior to this, I thought of him as an amazing guitar player with a very wry/dry sense of humor who was simply making fun of the guitar stereotype crowd. Actually encouraging something which is damaging in order to profit from it crosses the line whether he is serious or joking--too many people will listen and believe him.

Jerome Lowenthal, the pianist and pedagogue has focal distonia, and speaks about it in his teaching. Opera singers get vocal nodes from incorrect use or overuse of the vocal chords... musicians often turn to Pilates, Feldenkreis, or other techniques- but mostly it is about finding the posture that is the most natural and the motion that is most efficient. Cellists have special orchestral chairs, and violinists use better and better chin and shoulder rests these days- all in service to this.

ansolutely. an expert knows the most efficient motions from refining inefficient ones or from a coach who can spot issues. inefficient motions create repetitive use injuries.

I wouldn't discount the not insignificant amount of luck involved either. Anecdotally, I know of one baseball player who got called up to the majors only to have his career ended by taking a bad hop to the face just before his rookie season.

If his body cannot take the strain and recover between sessions then wouldn't that suggest that he does not have the genetics to be a pro golfer, but merely a good one?

Kind of related is that one of the biggest advantages of using performance enhancing drugs/steroids is the greatly reduced recovery time. Even if someone tests clean before an event because they genuinely are clean rather than being on something undetectable, they still benefit from all of the extra training they managed to get in in-between events when they were juicing.

> genetically gifted

No need to narrow this down to genetics. Just say "how fast you recover", thanks.

I mean, unless you give specific insights into how detrimental slack after training and superb wound healing are directly genetically correlated, you might as well not mention it.

Edit: Also, you seem to presuppose a benefit from wounding. I do doubt that.

Well compared to where he started or the average golf player but still very far away from a anything resembling a pro. Which in general is the scientific consensus. Which is you can get a lot better at things by practicing but being that best is always a matter of talent.

Since you mentioned both deliberate practice and 10,000 hours: are you familiar with Anders Ericsson's criticism of the 10,000 hours idea? Do you think it's still a concept worth talking about?

I liked Gladwell's response, that 10,000 hours was never intended as the goal it turned into. Rather, it was intended as an order of magnitude. More than that however, it was intended to call attention to the support structure being expert requires. I.E. nobody can get 10,000 hours in on their own.

This. I'm listening to Outliers right now (where he builds up that concept). I can understand how people think Gladwell is a load of crap... if you focus on his points out of context. It was never "get 10,000 hrs and you'll be an expert", but instead "once you get to a certain level of talent/innate ability, it's deliberate practice that makes the difference."

like the constant factors usually omitted in complexity analysis

Academia has put its nose up far in the air, it's become a career-crime to write anything for the public, so it falls to journalists to explain scientific results. Gladwell did a good job, then got manure tossed at him from scientists for daring to speak their results; and laypeople who flatly misunderstood what he said, or never read the articles/books and imagined what he said.

But the root problem is inequality - science has begun to see itself as an immaculate priesthood that mustn't mix with the plebs. Scientists should have been presenting their results to the public, but snobbishness now prevents this.

I'm having a bit of trouble interpreting these numbers.

If I look at deliberate practice in musical proficiency amongst the general populace, it'll look like deliberate practice is responsible for almost all of the proficiency. However, if I look at the top 100 musicians in the world, they would have all practiced thousands of hours and I expect practice to have diminishing returns, so it should be responsible for very little of the variation in their ability.

I would expect also that in more multifactorial activities which are often the basis of professions that it is both A) difficult to ascertain what expertise is (often reducing to an empty social credential) B) difficult to evaluate what good deliberate practice is.

The meta-analysis talks about such factors as the predictability of the environment correlating with the effect size of deliberate practice. I think that's just the tip of the iceberg.

> However, if I look at the top 100 musicians in the world, they would have all practiced thousands of hours and I expect practice to have diminishing returns, so it should be responsible for very little of the variation in their ability.

I think you're talking about two different things here. There are lots of musicians and artists and writers who have mastered their craft but have no amount of creativity. The opposite is also true. There are quite a few writers and musicians and artists who are rather poor from a technical perspective, but they manage to use what they have to create enduring pieces of art. Peanuts would be a great example for art. Snow Crash for writing.

Technical skill is one thing, but the ability to create an emotionally engaging piece of art is something else. Deliberate practice can only help with the former. As for the latter, John Cleese has an interesting perspective seen here: https://www.youtube.com/watch?v=Pb5oIIPO62g

As basically a lifetime musician (and second generation) I'd like to give some counter-points to your hypothesis regarding the "diminishing returns" notion and deliberate musical practice.

I'll try to keep this light hearted - so as a guitarist, one learns the rules, and practices them, then grows and grows and grows. Eventually, one learns how to break a rule, but over time, bend it into something that resembles an existing rule or otherwise still interests an audience, and create rules whereby none existed before. Sometimes minor mistakes or oddities become wholesale innovations in the right hands (ex: Jimi Hendrix, Tom Morello) but are very much a byproduct of deliberate, routine, intense attention and practice.

My experience is that music performance has an incredible power to both be-shaped-by-and-reciprocally-shape the human brain. The inherent elasticity and plasticity cross cultures. Generations. One of the most consistent observational effects though is that practice makes better in most cases.

There's one or more quotes about this, one of them supposedly by Picasso: "Learn the rules like a pro, so you can break them like an artist."

I think you're right. What separates the average from the good is practice. What separates the good from the great is something different.

My guess is that 'low-latent inhibition' plus a great memory plus long practice make for mastery.


(And those who force themselves to practice are going to get injuries like repetitive strain which are really a kind of 'inner rebellion' against this forcing.)

Never heard of LLI as lying on a spectrum that includes OCD, ADHD etc so thank you for that link.

The top 100 didn't all practice the same way, have the same experiences

If you want to improve this chess, I recommend this playlist by International Master John Bartholomew:


I recommend https://lichess.org over chess.com for a far superior UI.

Lichess is excellent. I run my in person games through it's analysis. They have a great feature where you practice the correct move where you had made a mistake or "blunder".

And it's open source: https://github.com/ornicar/lila

Open source, awesome!

I'm so impressed with their UI. You can watch games in real time and there's a dashboard showing you your Ping + Server Processing Time per Move (usually 2ms).

I'd have to recommend http://www.net-chess.com, but only because I wrote it.

But seriously on improving your chess. I think "how" you improve your chess play greatly depends on "why" you want to improve your play. Very few people actually plan to become professionals, so it does not make sense to try to copy their learning strategy, and in fact, a professional player probably isn't even good coach for the "casual but still serious" player.

I have been playing in the "serious but casual" arena for about 20 years now, the highest USCF rating I've achieved was about 1750. I used to be very active in the local chess community being cheif TD of at least one USCF rated tournament a week, sometimes 2 or three, and would play in about as many. I had seen many people start well below my rating and excel past me, and many stay well behind. I'd say the main difference between the 1800+ and the lower players is the amount of time they spend studying openings, games resulting from specific openings, and specific endgames. I've tried to study such things, but honestly every time I've tried I ended up reminding myself that I'm mostly playing for fun, and there is not point in torturing myself for something I just do for fun. So, if you actually enjoy sitting down and reading random books on theory, that's probably the way for you to go. But just because you don't want to do that, know there's a lot of fun to be had just playing people your own skill level, and you will (slowly) improve.

On a side note, there's some players who learn all the ins and outs of one archaic/unknown opening and use it all the time. These people usually rise to a certain level and platou, they might break 1600, or even 1800, but never 2000 without a significant change in strategy. The reason those openings are unknown is because they're usually weak. And a good player will take a slight advantage in the opening and leverage it for a win. At lower levels, a slight advantage in the opening usually isn't game determining.

Thanks for your advice and congrats on making a chess site!

There's also this paper which gives a general overview on which aspects of the game you should focus improving on once you mastered the basics of chess.

"Training in chess: a scientific approach"


Fyi... this a 2006 publication. (Not sure why HN submission says (2016).)

Being 11 years old isn't an issue because the chess memory observation is interesting data. The age is significant because virtually every pop psychology & brain book in the last 10 years mentions this chess study as one of the anecdotes. (Similar to how all the pop psych books mention the Invisible Gorilla, Marshmallow Experiment, Stanford Prison Experiment, etc)

Hey want to provide some quick clarification on your point.

First, you are dramatically, 110% right that pop psychology books plunder really valuable research such as this for ultimately banal insights that they manage to hide in a 300 page book that could be 10 pages.

K. Anders Ericsson, the absolute giant in the field of expertise research, is quite dismissive of Malcolm Gladwell and upset at his mis-interpretations of his work. Go look at his book, "Peak", which goes into depth on this (and funny enough looks like every other stupid pop psychology book, although it is not).

Finally, if you want to explore the best research we have from the present through the last 40 years, highly recommend you check out "The Cambridge Handbook of Expertise". It's 40 chapters written by nothing but PhDs.

They compare what expertise means across History, Firefighting, Surgery, Ballet, even Truck Driving and Software Design.

Got any links for when K. Anders Ericsson slams Gladwell? Was it in a blog post or a podcast or something?

Just looked at the physical book. I work in DevOps but just in general want to say physical books are great and highly underrated for how quickly you can access info / other things.


Lol you'll see in the comments I even wrote "hates Gladwell!".

In fairness to Ericsson, he is very civil and it's a very respectable "takedown".

One of the books that has had the biggest influence in my life is "The Art of Learning" by Josh Waitzkin, the child chess prodigy of "Searching for Bobby Fischer" fame.


What are some key learnings you got from that book?

I heard the audio book version, which was read by the author, and found it to be an excellent piece of literature. Highly recommend it to anybody.

Big takeaways for myself are the differences in what it takes to be "good", "great", and "among the best"

Also, once you become a master at something, then your opponents are more or less equally matched to you, and the final winner comes down to psychology

And, don't stick to one thing for too long if you don't want to. Put it down when you're finished, and begin mastering something else that's more fun / interesting.

My experience was the complete opposite. I genuinely do not understand why this book is recommended. It's been years since I've read it but whenever i see it recommended it irritates me. Especially as people who do never list anything of actual significance that's to do with learning, other than clichés.

You're not the first person who has told me that.

It's been ~5 years since I read it, so I can't remember too many specifics. I was ~6 months into starting to learn to program and I think I would have probably given up if I hadn't read Waitzkin's book. He talks about how certain people tend to give on something if they struggle in order to protect their intelligence, to keep from feeling stupid or like a failure. It may be common sense but it's something that really clicked with me as I realized that it was something that I had done my whole life up until that point.

I kept going and eventually made a career transition into engineering. But something else happened that was actually much more important, and that was that I developed a much deeper understanding of and appreciation for the whole learning process. I was so tuned in to the whole thing that I would have these insights where I would I could pinpoint almost all of the things he talks about in the book in my personal experience. Years later, I’m able to make parallels in my present experience to all the things I experienced then, which in many cases provides some encouragement because I know I’m on the right path.

So what kind of deliberate practice is required to become a highly skilled developer? Not just to know a language and library. Code kata used be a thing years ago.

IMO code katas are a waste of time. Practically speaking, prolific development means getting large chunks of the system done quickly, and the best way to do that is to reuse what others have done. Don't code your CSS from scratch, use Bootstrap or similar. Don't create an auth system, use the one that comes with your framework.

Counterintuitively, don't automate all the things until that automation is sure to pay off. Sometimes manually testing something is just faster, if you only need to test it a few times before the code never changes again (when was the last time you rewrote your password reset flow?)

To get prolific, don't do katas. Katas are the same problem solved in different environments. Do different problems in the environment in which you are productive. That is what will make you better and faster.

"Katas are the same problem solved in different environments."

Are you sure about that? I thought it was more solving the problem, then solving it again more efficiently / differently in the same environment. Like using recursion the second time around, not changing languages.

"Don't code your CSS from scratch, use Bootstrap or similar. Don't create an auth system, use the one that comes with your framework."

Definitely true for productivity, but a lot to be gained in experience by knowing how to roll your own.

Regularly solve problems! Try making minimal implementations of things you're interested in. I've implemented toy versions of lots of things just to learn how they work.

I always start off thinking it'll be easy, and every single time I'm amazed by the number of details and edge cases involved. It's usually where you'll get exposed to more interesting algorithms, since most applications are able to leverage libraries which usually hides the magic.

The meta-analysis linked above [1] claims that deliberate practice explains only 1% of variance in performance for professions (programmers are in this category in the article)

[1] https://news.ycombinator.com/item?id=15101548

It helps to read other people's code bases.

In the introduction I would begin telling the story of Najdorf playing blind chess: http://www.blindfoldchess.net/blog/2011/12/after_64_years_ne...

Apparently he did that one also as a means to inform family in Poland that he was still alive (http://www.poloniachess.pl/najdorf2017/index.php?page=najdor...).

I post-war times, doing something highly newsworthy was pretty much like today making it to HN front page.

Magnus has some youtube videos playing with his back to a series of boards, then after he wins he writes down every move that was played on every board.

So...(start internal monologue)

- The 10x engineer is so because of 10x practice (numbers not precise nor accurate). Companies - need to keep challenging people to get experts. Bootcampers - keep practicing.

- As with the soccer study in the article, boys ended up as the "bigger kids" in computer science and due to motivational factors, there's more boys in the tech industry today [1]

- Refutes a large portion of [2]; did the google engineer read these studies before?

[1] http://www.npr.org/sections/money/2014/10/21/357629765/when-...

[2] http://gizmodo.com/exclusive-heres-the-full-10-page-anti-div...

"Talent is Overrated" is a great pop exploration of this and similar findings. If you find this interesting, I highly recommend that book as a jumping off point for exploring similar findings.

Another very good book is "Peak: Secrets from the New Science of Expertise" by Anders Ericsson. I'm in the middle of reading it now. As I suspected (since he mentions studies on chess masters), he's part of this article, and in fact he's the author of 2 of the 5 "More to Explore" sources listed at the end of the article.

Just posted above on K. Anders Ericcson and the book Peak is great.

Main point is the 10,000 hour rule comes from Ericcson's study that the best violinists in the world practice about 7,000 hours by the time they are 18. They average between 3-4 hours a day. This last point is really interesting because the ceiling for what qualified as good practice consistently was no more than 4 hours. So if you do 4 hours of top notch practice / studying, it would seem like no matter your field you are done for the day.

Continuing on, the 10,000 hour rule was stupidly expanded by 3,000 hours just to make it easier to remember, and additionally that number is COMPLETELY ARBITRARY anyway.

The only reason violinists practice that much is because everyone else is competing that hard. Old, well defined fields like Violin or say Chess have pretty clearly defined ways on how you get better, meaning the "secrets" are somewhat known. Therefore the only way to get better is via practice, which of course everyone does, which drives up these insane numbers.

In other fields which are much newer and far less defined, expertise is up for grabs and there are no well-defined broadly agreed on ways to get better. If you do 800 hours of the "right" practice you can be far ahead of anyone else, since no one really knows what's right.

That being said, Ericcson's research still shows lots of correlation before time and skill, so alas no easy shortcuts.

EDIT: between time and skill

If I have an intellectual hero, it was Ericsson's post-doc advisor Herb Simon. In "Peak", Ericsson talks about work they did together on the mental representations used by expert chess players. I recently reread some of a lecture Simon gave in 1982 ("Reason In Human Affairs") and it is pretty much golden. Simon is the person responsible for the idea of bounded rationality, but in that talk he also captured ideas about information overload/"quality of information"/"media ecology" that were talked about later by Neil Postman and Nassim Taleb and are especially relevant today although they haven't quite sunk into our culture yet:


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