Ditch the 10,000 hour rule!

April 21, 2014 11:21 AM

Obsessive practice isn't the key to success. Spaced, interleaved, varied practice is.
posted by shivohum (44 comments total)

This article is pretty interesting in light of my current obsession with Titanfall, in two ways.

First, each time I play I have to choose between playing a specific gametype, or playing random. I've found myself drifting more frequently towards random gametypes, because the variety keeps it fresh and I get bored slower.

Second, in order to increase your generation (e.g. prestige level) the player has to hit level 50, *and* cap 500 kills with a specific gun. I am Gen2 right now and I tell you what, I am sick of playing with this gosh dang shotgun. It's been forever shooting dudes with it and I'm not very close to being done with it. I don't think I've gotten very good with it, either.
posted by rebent at 11:33 AM on April 21, 2014 [1 favorite]

The headline and framing Gladwell stuff is weird, isn't it? The 10,000 hour thing is talking about a sheer long-term commitment to practice. It's not really discussing how that practice happens from day to day. So in the surgery example, one group took all the lessons at once in a day, while the other group took them over four weeks. That's talking about something orthogonal to Gladwell's assertion that you need to spend a lot of time on something to master it. Both groups spent the same amount of time on the lesson, it was just handled in different intervals. I suspect the author didn't even have Gladwell in mind when he wrote this, and a Salon editor added it in to drive page clicks.
posted by naju at 11:41 AM on April 21, 2014 [34 favorites]

When I'm trying to learn something new in music, I play it a bunch until I feel like I'm not improving. Then I take a 5 minute break on another instrument and go back to the first one. Gain a lot quicker that way.
posted by Ironmouth at 11:44 AM on April 21, 2014 [4 favorites]

naju: I think the assertion that 10 hours in one day will actually learn less than 2 hours a day for 5 days.
posted by idiopath at 11:50 AM on April 21, 2014

I'm just quietly outliving all the other bass players.
posted by Devils Rancher at 11:51 AM on April 21, 2014 [46 favorites]

The tidbit about kids who practiced 2' and 4' shots being better at 3' shots than the kids who practiced 3' shots is pretty interesting.
posted by straight at 11:52 AM on April 21, 2014 [4 favorites]

Ironmouth: I know exactly what you mean by "until I feel like I'm not improving." When I hit that point, I quit practicing that piece, and know that I will play it much better once a sleep period has passed.

The effect of a sleep cycle on what you gained the previous day is sometimes amazing.
posted by seyirci at 11:55 AM on April 21, 2014 [15 favorites]

naju: I think the assertion that 10 hours in one day will actually learn less than 2 hours a day for 5 days.

Right, but the length of time spent is a constant. It doesn't address whether you can spend 5 hours total on the material vs. 10 hours total on the material. It's not claiming that 5,000 hours of spaced, interleaved practice is as effective as 10,000 hours of massed practice. You're still doing the 10,000 hours regardless as far as we know.
posted by naju at 11:56 AM on April 21, 2014

Just think, if we sporadically discuss this topic in a varied manner for another 9000 hours, we'll finally master the key to success and then everyone can just put it into practice and be masters of everything.
posted by forgetful snow at 11:57 AM on April 21, 2014 [40 favorites]

All Salon's headlines are just trying to get clicks anymore, and I'm getting increasingly peevish when I end up visiting them even by accident--and this is coming from someone who used to be a paying subscriber.
posted by Sequence at 11:59 AM on April 21, 2014 [9 favorites]

This seems obvious to me, but then, I've never been impressed by the "ten thousand hours" thing and I've never PRACTICE PRACTICE PRACTICEd my way into any skill in my life.

A whole book about how traditional ideas of "how you learn things," including PRACTICE PRACTICE PRACTICE, are full of crap, is Ellen Langer's wonderful _The Power of Mindful Learning_. Breaking up and varying how you practice something will surely cause you to approach it more mindfully.

It's a wonderful book.
posted by edheil at 12:00 PM on April 21, 2014 [6 favorites]

I practice until I get sloppy then I get away from it for a while. If you practice without focus then you run the risk of teaching yourself how to do it badly.
posted by any major dude at 12:01 PM on April 21, 2014 [2 favorites]

False dichotomy. The 10,000 hours rule is not a claim that doing just anything will get you expertise in a given field. Of course there are ways to practice that will yield worse results than intelligent programming. Trust me, both sets of surgical residents in this study are going to have put in their 10,000 hours before they are set loose on the general public.
posted by teh_boy at 12:03 PM on April 21, 2014 [1 favorite]

Planetside 2 does something vaguely similar, Rebent, though you get a series of medals for clearing out to 1162 kills with a particular weapon. Though the rewards for the medals themselves pale in comparison to what 1162 kills nets you. And that pales in comparison to seeing 1162 enemies fall like wheat before you over a period of days/weeks as you paint the map purple for your glorious Sovereignty...

There's also a nifty series of graphs that let you track your hourly, daily, weekly, monthly, quarterly, or yearly stats for things like score/hour, kills/death, etc. and its fairly awesome to go back and look at the steadily increasing graphs over the course of a year and a half.

I've found myself hopping around too much though, and this article really rings true with me. There's something along the lines of 7 to 9 weapons per class that are unique coupled with 9 weapons that are shared across classes (which are incredibly niche things to boot). I unlocked almost all of the things ages ago and the challenge each day now if figuring out which set is actually going to work and which set is going to make me want to claw my eyeballs out in frustration as I repeatedly fail to get into the right mindset for that range/engagement time/style of play. Definitely good enough yet where I can just pick any of them up and adjust for the night to its style of play.
posted by Slackermagee at 12:03 PM on April 21, 2014 [1 favorite]

I do the same thing as Ironmouth, except that I'll mix learning a tune with learning other tunes and scales / intervals and various exercises (often on different instruments). I'm gratified to find out that this fragmented (excuse me, interleaved and varied) approach isn't merely lack of discipline but a more useful holistic approach - even if I do still need to continue working toward that 10,000-hour goal.
posted by Greg_Ace at 12:04 PM on April 21, 2014

Saturday on MeFi: "Not 21 days"

How Long does it actually take to form a habit? Answer: Not 21 days. ...Maltz's work influenced nearly every major "self-help" professional from Zig Ziglar to Brian Tracy to Tony Robbins. And as more people recited Maltz's story -- like a very long game of "Telephone" -- people began to forget that he said "a minimum of about 21 days" and shortened it to: "It takes 21 days to form a new habit." A study debunks a popular self-help myth.

Your favorite self-help time span sucks.
posted by Celsius1414 at 12:04 PM on April 21, 2014 [11 favorites]

forgetful snow: "Just think, if we sporadically discuss this topic in a varied manner for another 9000 hours, we'll finally master the key to success and then everyone can just put it into practice and be masters of everything."

10,000 hours to rule them all.

In all seriousness, I thought it was pretty well-established that the brain requires periods of rest to encode experiences into long-term memory? This is why students are advised against "cram" sessions, because it simply won't stick. That said, I don't think the general message here is particularly novel. However, I suppose the devil's in the details and it's interesting to tease out optimums for spaced learning, which is ultimately the purpose of this research.
posted by tybeet at 12:04 PM on April 21, 2014 [2 favorites]

Team Fortress 2's class system is a good example of leaving it when you don't feel you are advancing. Just pick a different class, blamo, new set of skills to learn.
posted by Twain Device at 12:06 PM on April 21, 2014

naju: I think you misread me. With pauses / rest periods, fewer total hours are required to achieve the same result.
posted by idiopath at 12:10 PM on April 21, 2014

Hypothesis: 10,000 hours practicing X will make you an expert at X

Experiment:
Group A: spend 12 weeks practicing X
Group B: spend 12 weeks practicing W and Y

Result: Group B has learned X better than group A

therefore, to reach the same level of expertise as those who study X alone will take less than 10,000 hours - perhaps 30% less time.

New hypothesis: 6,000 hours of practicing W and Y will make you an expert at X

Thus: 10,000 hour rule refuted
posted by rebent at 12:12 PM on April 21, 2014 [3 favorites]

Its important to keep in mind that the 10000 hour thing was about people who have MASTERED a complex skill, not just being reasonably competent. Thus, unless you are world-class at whatever example you'd like to use, anecdata probably doesn't refute that particular set of assertions. Also, it does seem to make intuitive sense for me at least that periods of rest and reflection will allow faster overall learning, as long as I am immersed in the effort.
posted by sfts2 at 12:15 PM on April 21, 2014 [2 favorites]

The Dan Plan is about a guy in the Pacific Northwest who is putting in his 10,000 Gladwell hours on golf, starting as a complete rookie. He just passed 5,000 hours and has some interesting reflections on his progress. I've read the first bit of his blog…he spent months just putting, putting putting and putting. Then he progressed the same way with other clubs…I think he was a year or two into it before he hit the longest clubs at all. He has a single digit handicap now but the road from there to 'professional golfer' is probably as long as what he's already accomplished.
posted by Kwine at 12:28 PM on April 21, 2014 [2 favorites]

therefore, to reach the same level of expertise as those who study X alone will take less than 10,000 hours - perhaps 30% less time.

New hypothesis: 6,000 hours of practicing W and Y will make you an expert at X

Thus: 10,000 hour rule refuted

I may be misunderstanding but it seems like a fallacy to jump to any of this. Is the author making any such statement that less overall time is required to be good at something? I think that requires modification of the studies that are being cited and some arguments from the author in that direction.

The author is coming to 2 different conclusions in 2 different kinds of studies:
1. 12 weeks of X compared to 1 week of X will lead to stronger retention of material
2. 12 weeks of W and Y compared to 12 weeks of X will lead to better performance

He emphasizes that the stronger retention is a result of spending the same amount of time on material, but spaced out. He also emphasizes that the better performance is a result of the same amount of overall time, but on varying materials. In neither case is he claiming that less amount of time is necessary; quite the contrary. And then also, we're talking about mastery, not adequacy. It may be the case that spending 11.5 weeks on W and Y leads to less mastery than 12 weeks on W and Y, in which case you're not getting to where Gladwell wants you to be.

And then, Gladwell didn't address any of that "the nature of the practice" stuff with his 10,000 hours statement, so it's all orthogonal.
posted by naju at 12:41 PM on April 21, 2014

Thus: 10,000 hour rule refuted

The 10,000 hour "rule" (i.e., the finding that practitioners rack up apx 10K hours of deliberate practice before achieving expertise) refers to complex skills, like playing a musical instrument or driving a car. Part of deliberate practice means that it reflects the rich complexity of the skill. Playing note A really well, and playing note B really well, etc. won't help someone play a complex piece of music. It also means getting feedback on success or not, identifying problems and trying to fix them, trying new challenges. It is not just simple repetition.
posted by neutralmojo at 12:44 PM on April 21, 2014 [4 favorites]

This really seems like it should be a TED talk.
Why isn't this a TED talk?
posted by rocket88 at 12:58 PM on April 21, 2014 [3 favorites]

I'd like to take this opportunity to plug an app I have no stake in whatsoever (except that I use it all the time and love it like my own child), Anki. The "skill" in this case is memorizing flashcards, but spaced repetition is the philosophy and the results are unbelievable.
posted by telegraph at 1:13 PM on April 21, 2014 [7 favorites]

That's talking about something orthogonal to Gladwell's assertion that you need to spend a lot of time on something to master it.

Now I know the word orthogonal. Thanks, naju!
posted by jwhite1979 at 1:36 PM on April 21, 2014

Nobody ever masters anything. You just might get to be relaxed at doing something that people need or like, and if you do you'll then worry about something else in your life.
posted by colie at 1:55 PM on April 21, 2014 [1 favorite]

Doesn't matter how you practice if you lack talent. (I tried it.)
posted by Segundus at 2:09 PM on April 21, 2014 [2 favorites]

I think we can all agree that if someone types Print "Hello World!"; again and again for 10,000 hours consecutively that they likely won't learn something other than answers to questions like what does carpal tunnel syndrome feel like and what is the most muscle memory intensive way to write that. The contingency in Gladwell's initial statement of 10K hours is that it is both deliberate practice and actual practice. Doing Y and Z activities to improve your ability to do X is not at all surprising. Learning two skills is going to help with the neuroplasticity of skill X. If I got really good at lacrosse and yodeling I can guarantee my coding would improve because I would have learned something about how to learn in the process and that is directly transferable to any skill - ergo a liberal arts education is not completely useless - it tells you "this person can figure out how to learn things at a set level of proficiency."

Want to code even better than someone who knows lacrosse and yodeling? Practice coding as well. Want to code better than someone who only knows code? Get a liberal arts education and apply your knowledge of form and structure to code - you'll likely be able to pick up nuances of differing languages.
posted by Nanukthedog at 2:17 PM on April 21, 2014 [4 favorites]

Doesn't matter if you lack talent, if you practice.

I've taught over 600 architecture and design students, some with a natural talent from the outset, others with no such thing. Those who have succeeded are those who practiced their basic skills, at least 10.000 hours, regardless of initial talent. Actually, some of those who came in with great skills have failed dramatically, because they felt no need to practice the first 1-2 years, and then it was too late to catch up with the less talented but more hard-working.

And like many above are saying: it's not like you practice one single thing (bean-bag-throwing) for 10.000 hours. You practice a number of different, interrelated skills. And at a certain point, it stops feeling like practice - it feels like the thing you really enjoy doing for fun and for the social aspects and for the great feeling when something does what you want it to do.
posted by mumimor at 2:46 PM on April 21, 2014 [8 favorites]

Team Fortress 2's class system is a good example of leaving it when you don't feel you are advancing. Just pick a different class, blamo, new set of skills to learn.

W+M1 LOLZ

(Do they still say that? I usually loved playing Pyro. I frankly thought the occasional mad cackling suicide run was almost as fun as airblasting ubered heavies off the tower in orange_x_2.)
posted by Foosnark at 2:47 PM on April 21, 2014

Also, The Mundanity of Excellence: An Ethnographic Report on Stratification and Olympic Swimmers, which notes that there is a lot of stratification amongst competitive swimmers, and it tends to be discrete: People don't continuously advance between levels of competitive swimming, they do so in big leaps. It's not that they're doing something over and over and over again, they're figuring out how to do something differently.
posted by straw at 3:07 PM on April 21, 2014 [2 favorites]

Since I've never read his books, I don't know what Malcom Gladwell said specifically about 10,000 hours. I'm a little surprised that the premise of this article would come as a surprise to anyone who has worked to improve any skill.

It's always been obvious to me that cramming is a suboptimal strategy for learning material, and that the best way to learn is a steady, periodic and varied approach to the material you're trying to learn. I mean, that notion of how to study *ought* to be something every high school graduate takes with them after they leave school; it's pretty fundamental. I think I understood the concept it (even if I didn't always abide by it) after a few hard classes in junior high.

It's also been apparent to be for a very long time that the same principle applies to physical training, illustrated by the opening anecdote about the bean bag toss. In my experience for example, training to be a good runner doesn't just involve running at the speed or distance for the race you're training at. It requires fast days on the track, long slow days, days where you're mixing up the pace or the terrain, days where you're not running at all but doing some other physical activity such as swimming or cycling or rowing or lifting weights, and (especially important) some days where you simply rest.

I've heard the 10,000 hours meme many times, but figured it was just short hand for "practice a lot and practice well." That doing the wrong thing for 10,000 hours should be avoided seems to go without saying.
posted by scelerat at 3:28 PM on April 21, 2014 [1 favorite]

I haven't read Gladwell's books, but I did read an excerpt of Outliers that referenced this concept a long time ago, and it's clear from all of the rebuttals that almost nobody has read Ericsson's original paper (PDF), because it already responds to most of the critiques.

As mentioned upthread, Ericsson is referring to complex tasks, such as playing chess, sexing baby chickens, and musical instruments. The time is important as a rule of thumb, but much more important is the concept of deliberate practice. Just strumming on your guitar for 10 hours a day won't make you much better, but focusing on notes you hit a little bit flat or sharp, and improving those aspects of play will. Ericsson also says that many people can't focus on this type of deliberate practice for much more than 2 hours a day.

Actually, Josh Fouer makes a much better explanation of the 10,000 hour rule in his book "Moonwalking with Einstein" than anything Gladwell will ever write.
posted by KGMoney at 4:12 PM on April 21, 2014 [3 favorites]

Also, The Mundanity of Excellence: An Ethnographic Report on Stratification and Olympic Swimmers, which notes that there is a lot of stratification amongst competitive swimmers, and it tends to be discrete: People don't continuously advance between levels of competitive swimming, they do so in big leaps. It's not that they're doing something over and over and over again, they're figuring out how to do something differently.


I've noticed this in my experiences learning to play guitar over the past several years. I've never made linear progress, it's always a plateau of weeks of struggle struggle struggle leading up to "holy shit skill breakthrough" then a new plateau of struggle struggle struggle, then three weeks or a month or two later, "holy shit" breakthrough! again. And the brain really seems to need some space of "not doing that particular thing" for at least a few days in between - like there are background processes that need catching up...
posted by stenseng at 4:27 PM on April 21, 2014 [3 favorites]

That doing the wrong thing for 10,000 hours should be avoided seems to go without saying.

My research confirms your theory
posted by thelonius at 4:29 PM on April 21, 2014 [1 favorite]

pottery anecdote.
posted by GuyZero at 5:14 PM on April 21, 2014 [4 favorites]

Blog post about spaced repetition I stumbled across previously, including literature review. (and +1 to learning instruments several at a time it just does go quicker)
posted by yoHighness at 6:50 PM on April 21, 2014 [2 favorites]

(I say blog post above but it's to that Salon article what a 3 course meal is to throwing a single pea)
posted by yoHighness at 6:54 PM on April 21, 2014

This matches up with my own observations in mastering video games.
posted by JHarris at 8:26 PM on April 21, 2014 [1 favorite]

Massed versus distributed learning is one of my areas of research.

There has been about 130 years of research (Going back to Ebbinghaus in the late 1880s)which has consistently demonstrated that distributing practice over a longer period of time is more effective than cramming study into an intensive training session.

I don't think this will really surprise anyone here. What is interesting is how robust this body of research has been over time, and over a wide range of subjects and tasks, including motor skills. In fact, it has been referred to as the most robust and consistent area of research in experimental psychology (Dempster, 1988).

More recent research has started looking into the relationship between the amount of time between study sessions (called the intersession interval, or ISI) and the time between training and testing (called the retention interval, or RI).

What this research has found is that there is a strong relationship between the ISI and the RI and, to maximize learning, the ISI should be about 10-30% of the RI.

The real world practicality of all of this is that "cramming" is very effective when you are tested immediately after the cram session, however it is far less effective when the test comes after some delay. Think back to tests you crammed for at college--you probably passed the test, but couldn't have recalled the information if you had been asked a couple of weeks later.

Distributing practice, however, might be less effective when the test comes immediate after training, however the learning that does take place is far more durable when tested after some delay.

Another interesting aspect is that we still don't know why it is that all of this is happening.

If you're interested in reading more, Melody Wiseheart's lab has done a lot of interesting work in this area (and has links to articles which are free to download).
posted by FunGus at 9:05 PM on April 21, 2014 [13 favorites]

The way I'd describe what I've noticed is, the more times you repeat some task you're hoping to train in succession, the less impact it has. This is all IMO of course, far be it from me to step on researchers' toes, but the impressions made by each repetition becomes shallower and shallower. It never goes to zero, but it does become quite little. I've found that as boredom with a task increases so does the decreased effect of learning. It could be described like those parts of the brain that retain the training get tired out, or that they're just unable to retain too much of the practice in a short period, or maybe that the learning has to be digested, internalized, before it can continue beyond a certain level.

To continue those analogies, once you let some time pass, the parts of the brain that record the training rest and once again become able to accept change/the brain's ability to record impressions becomes restored/the digested effects of practice are committed to longer-term memory and short-term memory again becomes available to accept the results of practice.

Well, that's just how I consider it.
posted by JHarris at 11:38 PM on April 21, 2014

I've read the Gladwell book and I don't hate it, but he did miss a lot of nuance in his arguments. Yeah, the Beatles did 10000 hours of live performance in Hamburg, but it wasn't about learning how to put on a show, the point was that there was a lot of experimentation involved at that time. The bar owner actually didn't care about what they played all night as long, as it was kid's "youth" music. So John and Paul started playing the fast songs slow, and the slow songs fast. It eventually turned into a long involved song-construction seminar. The real unique trick about the Beatles was that John and Paul were both interested in the craft of songwriting since they were quite young.
posted by ovvl at 6:28 PM on April 22, 2014