K. Anders Ericsson is THE guy when it comes to expert performance and training. Everyone else refers to his work in their pop-sci books and articles (Freakonomics, Outliers, The Talent Code, and more), and he’s the ultimate origin of the Gladwellian “10,000 hours” meme (which trope Ericsson explicitly disavows).1
Sadly, I haven’t been able to find out who is carrying on his work on expertise and deliberate practice, given that he died in 2020 - he doesn’t have any regular co-authors, I couldn’t find any info about his Phd’s or postdocs or research at the FSU site, and GPT came up with nothing. I fear his intellectual torch may not have been passed, which would be a great loss to the world.
Nevertheless - let’s begin.
“This is a book about the gift that Wolfgang Amadeus Mozart, Sakakibara’s schoolchildren, and Ray Allen all shared—the ability to create, through the right sort of training and practice, abilities that they would not otherwise possess by taking advantage of the incredible adaptability of the human brain and body. ”
“More than two decades ago, after studying expert performers from a wide range of fields, my colleagues and I came to realize that no matter what the field, the most effective approaches to improving performance all follow a single set of general principles. We named this universal approach “deliberate practice.”
Ericsson wants to distinguish between 3 types of practice:
“Putting time in,” or just faffing around - which is probably most “practice hours” empirically. Think a kid halfheartedly playing scales, or hot cross buns, or an athlete in the pain cave running to crank out a 5 miler just for the sake of getting their miles in, or somebody doing something like “driving” or “doctoring,” and assuming that “hours put in” does something positive to skill levels (it doesn’t)2
Purposeful practice - has well-defined, specific goals, and some plan to get to those goals. It is focused, and has feedback mechanisms so you know whether you are wrong or right, performing better or not.
Deliberate practice - is purposeful practice, guided by a coach, mentor, or other expert who knows the road to getting better, and can give you feedback and training advice to get there by fixing specific flaws or following a specific regimen.
And what is the primary outcome of purposeful or deliberate practice?
Better mental representations
“Much of deliberate practice involves developing ever more efficient mental representations that you can use in whatever activity you are practicing.”
“in essence these representations are preexisting patterns of information—facts, images, rules, relationships, and so on—that are held in long-term memory and that can be used to respond quickly and effectively in certain types of situations. The thing all mental representations have in common is that they make it possible to process large amounts of information quickly, despite the limitations of short-term memory.”
“In any area, not just musical performance, the relationship between skill and mental representations is a virtuous circle: the more skilled you become, the better your mental representations are, and the better your mental representations are, the more effectively you can practice to hone your skill.”
Expertise consisting of better mental representations is true even for physical skills, like athletics, but is obviously true for things like chess and music and mathematics as well.
My favorite fun story about the power of mental representations is Janet Starke running a study on volleyball players in the 70’s - she put together thousands of photographs of women’s volleyball games, and made slides where the volleyball was in the frame, or had just left the frame. This meant the orientation and body language of the players between the two were nearly identical, because people haven’t changed much in the time it takes the ball to leave the frame. Then she had competitive volleyball players look at the slides for fractions of a second, to see if they could determine whether the ball was in the frame or not - the brief flash of time that the picture was shown was too quick for people to consciously look for and decide if the ball was there.
The difference between top players and novices was huge - top players could reliably determine whether the ball was in the frame or not, and the better the player, the more quickly they could determine this. One member of the Canadian national volleyball team was able to do this in 16ms, which is an *astounding* speed.
Conscious human reaction time is generally estimated at around 200ms - the time it takes somebody to see a light, then hit a button their finger was already on indicating they had seen the light. Even the blink reflex operates at about 100-150ms. 16ms is so fast it would look like a flash of light so brief you couldn’t even be sure it had happened, and this player could not only determine whether the ball was in the frame in that time, she could tell you things like “that was the Sherbrooke team after they got their new uniforms, so the picture must have been taken at such and such a time.”
THAT is the power of good mental representations - the ability to create an interior schema so strong that you can understand and react subconsciously far faster than what’s normally thought of as the possible human reaction time.
Indeed, Ericsson calls out mental representations as one of the great blind spots in training optimization today - we know it seperates the elite from the sub-elite, but we have no real understanding of them, good ways to elicit them, or ways to craft training programs to explicitly hone them.
This ties into a tension running throughout the book between the “blank slate” view of human talent and capability, where anyone can be a winner, and the idea of innate and differentiated levels of talent between people.3
Ericsson wants blank slate to be true - indeed his ultimate vision and aspiration when writing this book was “deliberate practice” becoming more widely adopted in every domain, for the skills of the top 5% in “expertise” to become honed and available to the median practitioner - “Imagine a world in which 50 percent of the people in these professions learn to perform at the level that only the top 5 percent manage today.”
That is indeed a wonderful vision, and I hope it someday becomes true, but I personally think there ARE important differences in innate talent, and that achieving this world will require a lot of gengineering or technological enhancement, and not just more deliberate practice.
But deliberate practice IS powerful, and it CAN be used to improve anyone from where they are now (but perhaps not all the way to elite levels - maybe we can hope for a world where 50% of people perform at today’s top quintile, which would still be an amazing improvement).
So what more can we learn about it?
I’ve gone over some of the interroceptive cues you should look for to see if you’re in deliberate practice mode in my review here, and Ericsson confirms all of those. You should not be comfortable,4 you should be striving and failing, and you will find blocks and plateaus.5
How to overcome those blocks and plateaus?
“Getting out of your comfort zone means trying to do something that you couldn’t do before. Sometimes you may find it relatively easy to accomplish that new thing, and then you keep pushing on. But sometimes you run into something that stops you cold and it seems like you’ll never be able to do it. Finding ways around these barriers is one of the hidden keys to purposeful practice.”
“The best way to get past any barrier is to come at it from a different direction, which is one reason it is useful to work with a teacher or coach. Someone who is already familiar with the sorts of obstacles you’re likely to encounter can suggest ways to overcome them.”
“In all of my years of research, I have found it is surprisingly rare to get clear evidence in any field that a person has reached some immutable limit on performance. Instead, I’ve found that people more often just give up and stop trying to improve.”
Seeing as he’s directly pointing to “coaching” as the differentiator between purposeful and deliberate practice, maybe it’s time to go over the coaching insights from Peak and Talent Code now.
Master coaching insights
Ericsson on finding a good coach: don’t ask people who are elite for advice, ask them for their coaches. To evaluate a coach, talk to former students, ideally finding a former student who was at your current skill level when they came to that coach.
Most elites started with completely average coaches as children - these coaches didn’t have high levels of domain skill, but were great at getting engagement from kids and keeping them motivated, and this is generally a good strategy to follow - your kids can move on to higher skilled coaches when appropriate and desired
Master coaches aren’t dominant, alphas, like heads of state6 - they’re more like farmers or gardeners, careful, deliberate cultivators of talent - “They possess vast, deep frameworks of knowledge, which they apply to the steady, incremental work of growing skill circuits, which they ultimately don't control.”
Of more than 2k discrete “acts of teaching” observed when watching master coaches at work, only 7% were compliments, 6.6% were expressions of displeasure, and 75% were “pure information.” What to do, how to do it, when to intensify an activity - one of the most valuable forms of teaching was a three-part instruction where they modeled the right way to do something, showed the incorrect way, and then remodeled the right way.
Details matter - “The coach would spend two hours each morning with his assistants planning that day's practice, then write out the minute-by-minute schedule on three-by-five cards. He kept cards from year to year, so he could compare and adjust. No detail was too small to be considered.”
Master coaches have vast task-specific and domain knowledge, sharp perception, and seek and establish deep and mutually trusting personal relationships with their mentees
As soon as somebody gets a handle on a new move or technique, a master coach layers in an added difficulty - they keep momentum and progression going
“In the most literal sense, master coaches are the human delivery system for the signals that fuel and direct the growth of a given skill circuit, telling it with great clarity to fire here and not here. Coaching is a long, intimate conversation, a series of signals and responses that move toward a shared goal. A coach's true skill consists not in some universally applicable wisdom that he can communicate to all, but rather in the supple ability to locate the sweet spot on the edge of each individual student's ability, and to send the right signals to help the student reach toward the right goal, over and over.”
How to keep motivation high
“How do you keep going? That is perhaps the biggest question that anyone engaged in purposeful or deliberate practice will eventually face.”
Indeed, this is often THE distinguishing characteristic between talent tiers, because motivation leads to more hours being put in, and after initial quality filters, hours-put-in is usually the determining factor for where you will rank in a given talent hierarchy.
As the original “10k hours study” on 10 violinists that Gladwell latched onto and memed to the stratosphere showed us: “Specifically, the music-education students had practiced an average of 3,420 hours on the violin by the time they were eighteen, the better violin students had practiced an average of 5,301 hours, and the best violin students had practiced an average of 7,410 hours.”
So how DO you put in those hours? Broadly, you do some combination of weakening reasons to quit, and strengthening reasons to continue.
“Maintaining the motivation that enables such a regimen has two parts: reasons to keep going and reasons to stop. When you quit something that you had initially wanted to do, it’s because the reasons to stop eventually came to outweigh the reasons to continue. Thus, to maintain your motivation you can either strengthen the reasons to keep going or weaken the reasons to quit. Successful motivation efforts generally include both.”
Weakening reasons to quit:
Set aside dedicated time every day
Do it as soon as you get up
Plan and keep track of all your time in the week
Look for things that interfere with practice and mitigate them (phones!)
Staying healthy via general physical maintenance, sleep, diet, and activity
Limit practice to an hour or so per session
Strengthening reasons to continue:
Map it back to your original intrinsic motivation - you wanted to play this sport / instrument for a reason, one initially grounded in wonder or admiration or intrinsic motivation - tap back into that
Incorporate it into your sense of identity, and then training hours are an investment in your future self, and this aspect of yourself
Believe that you can succeed and perform at the level you’re aiming for
And finally, of course, harking back to the environmental cues:
“One of the best ways to create and sustain social motivation is to surround yourself with people who will encourage and support and challenge you in your endeavors.”
He points out that music students often hang out and befriend and date other music students, and serious athletes likewise, and that this is a good idea for a number of reasons.
You can also refer back to some of the environmental cues that you can cultivate for yourself that I covered in the Talent Code review.
What can you do for your kids?
The psychologist Benjamin Bloom ran a study that examined the childhoods of experts in various fields.
The first thing you want to do is make sure your kids don’t quit.
This is accomplished by getting a teacher or coach who is great at keeping kids engaged and motivated rather than optimizing on the coach having elite skill levels.
Another trick - Bloom found for performers who ran into obstacles when young, like injuries or some other reason they couldn’t practice for a while, when they started again, their skills had noticeably declined, and discouraged, they wanted to quit. Their parents said they could quit, but first they had to work back up to how good they were before, and by the time they got there, intrinsic motivation kicked in and they wanted to keep going.
Having older siblings is a good idea too:
“Mario Lemieux’s experience points up another salient feature of prodigies’ early experience—how many of them had older siblings to be inspired by, to learn from, to compete with, and to model themselves after. Judit Polgár had Susan and Sofia. Wolfgang Mozart had Maria Anna, who was four and a half years older and who was already playing the harpsichord when Wolfgang first became interested in music. Tennis great Serena Williams followed in the footsteps of her sister Venus Williams, who was herself one of the best tennis players of the current era. Mikaela Shiffrin, who became the youngest slalom champion in history during the Olympic Games in 2014, had an older brother, Taylor, who was a competitive skier. And so on.”
This actually lets me work in another factoid I love - you know how younger sibs have to run faster to keep up with their older sibs? Motivation is high, and entirely unconscious. It turns out elite sprinters are overwhelmingly younger sibs!
From Epstein’s The Sports Gene:
“Here are the birth-order ranks of the world-record progression in the 100-meter dash, with the most recently set world record first, the previous world record second, and so on.”
Usain Bolt (second of three children)
Asafa Powell (sixth of six)
Justin Gatlin (fourth of four)
Maurice Greene (fourth of four)
Donovan Bailey (third of three)
Leroy Burrell (fourth of five)
Carl Lewis (third of four)
Burrell (fourth of five)
Lewis (third of four)
Calvin Smith (sixth of eight)
“In all, history's fastest runners were born, on average, fourth in families of 4.6 children. We find a similar result with the top-ten all-time NFL running backs in rushing yardage, who score an average birth rank of 3.2 out of families of 4.4 kids.”
That’s fun, right? One of the more entertaining arguments for having a big family, IMO. “Honey, we NEED six kids, because that maximizes our chance of having an elite sprinter!”
But back to kids:
“Generally when they’re in their early or mid teens, the future experts make a major commitment to becoming the best that they can be.”
This is the switchover from extrinsic motivation, or doing it because they want to please you, to intrinsic motivation, or doing it because it’s becoming part of their identity.
As they become praised for their burgeoning skills, as they take part in the cameraderie and approval of team endeavors, many children begin to see themselves as soccer players, or swimmers, or musicians, or whatever. This is also generally the time when you seek out a higher level coach, and your kids begin to take responsibility for their own practice and advancement.
But that brings us to an important point:
Don’t coddle your kids
One very clear and recurring lesson we get from this book is that people need to be uncomfortable, people need to strive and reach and fail, to learn and grow in skill.
That includes kids. It is *especially* important for kids to strive and push and not be comfortable, if they’re going to have any skills or be worth anything at all as adults.
And yet, every single trend seems to point the other way in our current culture.
The median kid in the USA spends more time sitting and more time in air conditioning than ever before. They’re driven everywhere. Physical activity is at an all time low, and many schools have made PE optional. Most school’s “education” is a farcical joke aimed at the slowest kid in the class, and they’re STILL failing to teach things like basic literacy to big chunks of the populace. It’s a bad show all around.
The median kid spends their life sitting for 10 hours a day commuting or in school, and sitting and staring at screens another 7+ hours a day at night at home, and sitting is the new smoking. There’s no competition in any domain, because the adults in their lives have decreed that “everyone is a winner.” When do these kids ever reach? When do they ever strive? When do they ever fail?
They largely don’t. This is a recipe for dumb, fat, weak kids who have no skills, and we’re currently at a nadir after a long road aiming right where we are today.
“A larger-scale example of this phenomenon occurs in Japanese schools. According to a 1995 study, a sample of Japanese eighth graders spent 44 percent of their class time inventing, thinking, and actively struggling with underlying concepts. The study's sample of American students, on the other hand, spent less than 1 percent of their time in that state.”
This anecdote about Japanese schools points to an important truth - YOU need to provide a framework for struggle, reaching, failing, and trying again to your kids, because our schools will not do it. It is *entirely* up to you, to inculcate the right learning mindset, for your kids to undertake deliberate practice and learn excellence in any domain whatsoever.
So take these lessons from Peak and apply them to your kids’ lives.
Also, if I may insert a bit of self-insert advocacy as an aside - you want a reasonable “screen time” compromise? Get them a treadmill desk, and advocate that they can only have screen time as long as they’re on it, moving. Did you know hunter gatherers are 5x more active than sedentary moderns, and don’t suffer from anxiety, depression, diabetes, eating disorders, or a bunch of other things? Moving matters a lot, for both mental and physical health - exercise matters. You want to hear the full story, read my review of Dan Lieberman’s Exercised.
One last point - Ericsson points out a few domains where it is better to start young.
Childhood critical windows
Music - perfect pitch acquisition window closes at 6 generally, train between 4-5yo
Tennis - the full serve range of motion is usually seen in people who began between 8-10yo
Ballet - leg and hip turnout should be learned and practiced between 6-10yo, hip calcification beginning at 8-12yo can prevent being able to do full turnout later
Baseball pitching - the full shoulder and arm range of motion is usually seen in people who began between 9-13yo
Language acquisition - fluency better in people exposed between 0-7yo to second or third languages
What if you don’t have a coach, but want to improve in more commonplace domains in life?
“The hallmark of purposeful or deliberate practice is that you try to do something you cannot do—that takes you out of your comfort zone—and that you practice it over and over again, focusing on exactly how you are doing it, where you are falling short, and how you can get better. Real life—our jobs, our schooling, our hobbies—seldom gives us the opportunity for this sort of focused repetition, so in order to improve, we must manufacture our own opportunities.”
“Much of what a good teacher or coach will do is to develop such exercises for you, designed specifically to help you improve the particular skill you are focused on at the moment. But without a teacher, you must come up with your own exercises.”
“To effectively practice a skill without a teacher, it helps to keep in mind three Fs: Focus. Feedback. Fix it. Break the skill down into components that you can do repeatedly and analyze effectively, determine your weaknesses, and figure out ways to address them.”
He even gives some specific feedback in terms of being better at business or management:
“For anyone in the business or professional world looking for an effective approach to improvement, my basic advice is to look for one that follows the principles of deliberate practice: Does it push people to get outside their comfort zones and attempt to do things that are not easy for them? Does it offer immediate feedback on the performance and on what can be done to improve it? Have those who developed the approach identified the best performers in that particular area and determined what sets them apart from everyone else? Is the practice designed to develop the particular skills that experts in the field possess? A yes answer to all those questions may not guarantee that an approach will be effective, but it will certainly make that much more likely.”
What is the future of deliberate practice?
Ericsson has pointed to real gaps - understanding mental representations, crafting more deliberate coaching plans, and the like, but closing the gaps is hampered by execution challenges.
He’s also pointed to a real opportunity - most training, even training at the “elite sport” level, hasn’t fully incorporated the insights from deliberate practice. He talks about medical continuing education requirements, and this is one area he was trying to make a difference, but honestly MOST learning and training opportunities in life are structured badly, are knowledge rather than skill based, and don’t take advantage of what we know would drive the most effective changes in skill and expertise.
Obviously we are reading things like this to extract insights we hope to use for ourselves and our loved ones - but there’s a major “external world” opportunity too, in many domains.
I personally (of course) think the gaps are going to be filled by AI.
What are the problems with eliciting mental representations via asking?
So many problems:
You have to ask significantly after they’re actually doing things, because they’re not able to answer in the moment while they’re playing a sport or musical piece
“Doers” aren’t necessarily the most narratively introspective or articulate people to try to put inchoate, System 1 reactions and judgments into words
As we’ve seen with the volleyball study, most mental representations are purely subconscious, because they’d be far too slow as conscious schemas, so you’re only ever going to get a “conscious mind” confabulation that probably has little if any relation to the real unconscious representation
How would you even judge in what ways somebody’s mental representation was better using only words? You can see talent differentiation in actual performance, but dialing in on “what one person sees and responds to that the other doesn’t” sounds extremely difficult to extract from words alone, and it’s likely many things are informing a gestalt judgment, not just one or two things
Why is AI the answer here?
It can listen to a multitude of people’s (mostly fake) blather and verbal explanations about what’s going on, and extract commonalities tiered by skill level
It can empirically test the limits of your personal mental representations by flashing relevant pictures or audio clips at varying speeds and having you make judgments, and it can do this in a massively parallel, itereated way to actually map out the boundaries of what you do and don’t see and react to, and it will have reference standards for every level of performer it has info about
Using the map of your boundaries it’s created, it can create a custom pedagogy and training program that can fill those holes in for some people, allowing them to progress from sub-elite to elite
It can integrate and analyze the mapped boundaries with the verbal explanations to see if there actually is any correlations between them, to craft better coaching instructions, as well as training to elicit better verbal explanations from the performers
Honestly, this idea is so exciting I’d love to build it, because you can get a good enough start on this today for it to probably drive value, and it’s pretty simple to build.
But I’m not sure the money and impact is there to be worth it - the US doesn’t really care at the Olympic level, they only pay $38k per gold medal. There’s money in pro sports, but that’s definitely a “do you have the personal contacts” sort of thing that’s probably difficult to break into. And really, you probably want to start at the collegiate level because most of the value lies in promoting a broader talent pool to “elite” status, so that’s even harder to get paid for, because you want results you can point to but won’t have them for a few years. Maybe Singapore or Hong Kong would be interested, which both pay ~$700k per gold, but even then, how much could you really make from them? A couple million? Over the couple of years it probably takes to upgrade “sub-elite” collegiate schemas to elite level? Then how do you prove that they wouldn’t have become elite by themselves? You need unambiguous talent tiers and to RCT a statistically significant number of talent-matched athletes to measure your lift. Quite hard to assemble, long to payoff, a very low payoff, quite a bit of investment and runway burned before you see any results. Overall a pass, even though it’s pretty exciting and interesting in terms of the idea itself.
What would you get from reading the book yourself?
So much, I left out a lot in the interests of keeping this review shorter and more actionable. Overall it’s a fun and accessible book with decent depth written by an expert at the top of his field, and I think it’s well worth reading.
Much more depth on specific training domains like music, athletics, etc
The story of how the Top Gun naval air combat program 5x’d our dogfighting effectiveness
The story of how Paganini continued playing a masterful concert piece on the violin as first one, then two, then the third string broke in succession, leaving him to finish on only a single string, to great plaudits and astonishment
More flavor on Dan Maclaughlin and the “Dan Plan” to put in 10k hours and become a pro PGA golfer7
Fun anecdotes about Ben Franklin coming up with various purposeful practice techniques to improve himself
Various arguments about willpower not mattering, everyone being capable of bettering themselves massively, and so on, which I found naively endearing
Various debunkings of “innate talent” examples that varied widely in quality (Mozart, Donald Thomas, Mario Lemieux)
A case study of “deliberate practice” tuned pedagogy as a hugely impactful educational intervention, that definitely seems worth replication if you do any classroom teaching
The original study Gladwell latched onto was data on only 10 violin players, relying entirely on years-later retrospective self-estimates of practice hours, and the violinists did “some of the restrospective estimates several times, and there was no perfect agreement” according to Ericsson himself. So more garbage quality than usual, even for the social sciences!
Not just that, but “10k hours” is totally fake. Practice hours before age 18 in the “elite” violin players was ~7400. By 20 it had gotten up to 10k, but even at 20, they weren’t anywhere near the top of their field, and were still learning - Ericsson points out that most elite musicians don’t peak until 30, or around 25k hours.
“But no. Research has shown that, generally speaking, once a person reaches that level of “acceptable” performance and automaticity, the additional years of “practice” don’t lead to improvement. If anything, the doctor or the teacher or the driver who’s been at it for twenty years is likely to be a bit worse than the one who’s been doing it for only five, and the reason is that these automated abilities gradually deteriorate in the absence of deliberate efforts to improve.”
“The final problem with the ten-thousand-hour rule is that, although Gladwell himself didn’t say this, many people have interpreted it as a promise that almost anyone can become an expert in a given field by putting in ten thousand hours of practice. But nothing in my study implied this. To show a result like this, I would have needed to put a collection of randomly chosen people through ten thousand hours of deliberate practice on the violin and then see how they turned out. All that our study had shown was that among the students who had become good enough to be admitted to the Berlin music academy, the best students had put in, on average, significantly “more hours of solitary practice than the better students, and the better and best students had put in more solitary practice than the music-education students.
The question of whether anyone can become an expert performer in a given field by taking part in enough designed practice is still open, and I will offer some thoughts on this issue in the next chapter. But there was nothing in the original study to suggest that it was so.”
“All of the singers, both amateur and professional, felt more relaxed and energized after the lesson than before, but only the amateurs reported feeling elated afterward. The singing lesson had made the amateurs, but not the professionals, happy. The reason for the difference lay in how the two groups had approached the lesson. For the amateurs it was a time to express themselves, to sing away their cares, and to feel the pure joy of singing. For the professionals, the lesson was a time to concentrate on such things as vocal technique and breath control in an effort to improve their singing. There was focus but no joy.”
“This, then, is what you should try when other techniques for getting past a plateau have failed. First, figure out exactly what is holding you back. What mistakes are you making, and when? Push yourself well outside of your comfort zone and see what breaks down first. Then design a practice technique aimed at improving that particular weakness. Once you’ve figured out what the problem is, you may be able to fix it yourself, or you may need to go to an experienced coach or teacher for suggestions.”
From Coyle’s The Talent Code:
“Instead, the teachers and coaches I met were quiet, even reserved. They were mostly older; many had been teaching thirty or forty years. They possessed the same sort of gaze: steady, deep, unblinking. They listened far more than they talked. They seemed allergic to giving pep talks or inspiring speeches; they spent most of their time offering small, targeted, highly specific adjustments. They had an extraordinary sensitivity to the person they were teaching, customizing each message to each student's personality. After meeting a dozen of these people, I started to suspect that they were all secretly related. ”
It didn’t work out. Dan eventually got to around 6k hours in 2015, and his peak handicap was 2.6. That’s a good handicap, but PGA Pros usually average around +4-6 handicap, with the plus meaning the other way past 0, so even at 2.6, Dan was 6-9 strokes off, which is a pretty significant gulf.
Given how strict he was about allocating only “deliberate practice” hours, he probably had 10k hours in the Gladwellian or layman sense. Broadly, he got to “top 5% of golf players,” when he needed to be at “top 0.1%” to average what a PGA card carrier averages.
Dan gave up in 2015 citing an injury and back pain.