Today's society is full of desires and expectations to constantly learn, track and improve. This often leads people to look at what the experts do, in the hope for something to emulate. Over the years this interest in the experts and becoming like them has led to bestselling books, with indicative titles such as The Talent Code: Greatness Isn’t Born. It’s Grown. Here’s How and Talent Is Overrated: What Really Separates World-Class Performers from Everybody Else.

In the field of psychology, it has always been a question for debate whether nature or nurture is what molds people and create great characters. Some of these popular science books clearly speak for the importance of nurture. But do these books live up to their promises and do we know what leads to expertise and thus why some become experts while others stay ordinary?

Many years ago, the Danish physicist Niels Bohr said:

An expert is a person who has found out by his own painful experience all the mistakes that one can make in a very narrow field.
-Niels Bohr, Danish physicist

These words lead us to a few questions:

• How long might it take for one to make all the mistakes in any given domain?

• What would be the best approach to making all these mistakes and learning from them?

• How does Bohr’s view fit in with the contemporary research of learning and expertise?

This article is an attempt at answering such questions by explaining at least some of the science on learning and expertise development.

The 10,000-Hour Rule

You might very well be familiar with the “10,000-Hour Rule”. This rule was made famous by Malcolm Gladwell in his bestselling book “Outliers”. Gladwell’s point was as compelling as it was simple: becoming an expert isn’t about being gifted, but primarily about putting the effort into practicing — for 10,000 hours.

Gladwell repeatedly argued by examples, that the 10,000-Hour Rule has shown itself again and again through history. The Beatles — Gladwell loosely calculated — played just around 10,000 hours before they moved beyond mediocrity and started becoming famous. And likewise, Bill Gates and many of the most successful entrepreneurs of Silicon Valley had early access to computers, securing them their 10,000 hours of practice at a young age.

Gladwell never stated directly that the rule will guarantee expertise, but it might be the impression left on most readers. Whether or not Gladwell has done his own 10,000 hours of writing-practice is unknown. His writing, however, has been hugely successful and popular. Such popularity and mass appeal are often the strengths of popular science. The weakness, on the other hand, is that the correctness and precision of the writing might suffer in the quest for catchy selling points. And so, the researchers specializing in expertise and learning have not been uniformly impressed with Gladwell’s arguments.

What Gladwell’s popular science got wrong

One critic of Gladwell’s concept of a 10,000-Hour Rule is Swedish research psychologist Anders Ericsson. Ericsson stands behind the research on which Gladwell based his 10,000-Hour rule. While it is to be expected that a lot of details get lost in translation between scientific papers and popular science books, Ericsson’s critique has two main points.

The number 10,000 comes from Ericsson’s study of expertise development in violinists playing in Berlin in the ’90s. But it is neither a magical number nor a rule. Ericsson’s group of researchers divided the violinists into groups by skill level. The researchers defined one group, as those showing “promise for international performance as violinists”. The number 10,000 was simply the average number of practice hours accumulated for this one group of violinists at the age of 20. Importantly some of these had played thousands of hours less and some thousands more than this average. Even more importantly these violinists were not experts at the age of 20 — they just had the potential to become experts. Safe to say, turning this number into a general rule for becoming an expert is quite a stretch.

Secondly, Ericsson and his colleagues measured the accumulated experience of the violinists by ‘deliberate practice’. This, in turn, got lost in translation and ended up being just ‘practice’. According to Ericsson’s research, this represents a critical difference. While practice could be casual play with friends, the concept of deliberate practice is less inclusive and is literally more deliberate and more focused. But we will get back to that.

What Gladwell got right

Gladwell did at the very least get one thing right: developing expertise in an established field requires extreme amounts of practice. Although it comes at a price, the simplicity of the rule makes it easy to understand and thus easy to be inspired by.

Since Gladwell attracted both positive and negative attention, he surely has helped spread the word of the research. Gladwell probably spread his concept of the 10,000-Hour Rule to far more people than what teams of researchers would have ever achieved. While being less exact and scientific he spread a message of both the importance of practice and its potential for development. This might, in turn, have set up a market for Ericsson’s own 2015 book Peak.

Here we will leave Gladwell behind for now and turn to the world of academic research.

The characteristics of your domain

Which domain you seek expertise in will in large part determine your odds of success. It could even determine whether success is possible. It might seem intuitive — if not directly obvious — that gaining experience over the years will be the path towards expertise. However, the evidence shows that what seems obvious isn’t always so. This is exemplified by the domain of clinical psychology.

Consider the difference between an inexperienced, newly graduated psychologist and a well-established psychologist with 20 years of practical experience in the field. The latter would statistically be the more expensive option and most likely the one preferred by most potential clients. Obviously so, you might interject. Surely those 20 years of experience will make a difference — right? Not according to large scale meta-analyses which show no link between psychologists’ level of experience and level of skill. In other words, judging from the best available evidence, it seems the skill building stops at the end of education — at least the skills which manifest themselves as a felt improvement by the clients. This doesn’t necessarily mean that expertise is not achievable in psychology. What it does tell us is that if you want to be an expert in this domain the odds are stacked against you, and statistically speaking you cannot expect to gradually become an expert as time passes.

How can this be the case? The research of James Shanteau holds some strong hints. For decades he has done research looking at how, when and why expertise is developed. Experts, as defined by Shanteau, are simply those “who have been recognized within their profession as having the necessary skills and abilities to perform at the highest level”.

Shanteau points out that the role of the task characteristics is at the same time crucial and often overlooked. So, while it matters what kind of person you are and what your mindset is, no combination of these two will be a guarantee for success by themselves. Shanteau specifies that domains dealing with human behavior are often hard to predict and therefore a difficult setting for an expert. On the other hand, experts do well in domains that are relatively static — for instance mathematics, physics and accountancy. Already we have a hint that the lack of expertise development in areas such as clinical psychology might not solely be due to a lack of effort.

Professor Robin M. Hogarth is another front figure in the research on learning and expertise, specializing in intuitive judgment without advanced aids. He makes a distinction between “Kind Learning Environments” and “Wicked Learning Environments”. In the former there is a clear connection between the cues in the environment and the resulting effects, enabling an agent to learn to recognize patterns and predict the outcome of a process. In the latter, this link is less clear, if at all existent. This is perhaps best described by an illustrative and rather famous example.

Hogarth gives the example of a physician working at a hospital in New York in the beginning of the 20th century. The physician acquired a reputation for being an expert in detecting typhoid fever at very early stages. Since he could base the diagnosis directly on the appearance of the tongue, no decision aids were needed. In his assessment, he would palpate the tongue of patients with his bare hands and often reached the conclusion, that the patient was indeed infected with typhoid fever. Only later was it discovered, that the physician’s impressive diagnostic accuracy was the result of his own hands being an extremely potent carrier of the disease. For all the physician could see, he was doing very well — and that is the nature of a wicked environment, in the worst case.

Surely a lot of work and challenges will lie between detecting illness in the pre-penicillin era and a clean-cut Kind Learning Environment such as accounting. Here the game of poker can be considered a middle zone — this location on the spectrum explains the longstanding debate over whether poker is gambling or a skill game with experts. In the research jargon, the game of poker is characterized by a combination of an inaccurate feedback loop and high validity. The former simply means that the feedback you get — how much you win and lose during the game — will be affected by luck in the short term and therefore is an inaccurate measure of skill. The latter means that the decisions you make in the game has validity and decides how the game turns out for you — in the long run at least.

So, the nature of your domain and what kind of expertise you aim for will set some boundaries for the likelihood that you will become an expert. Moving on from this point leads to the topic of how to plan your practice for it to be most effective.

Practice — deliberately

Here we will return to Anders Ericsson and the concept of deliberate practice. As mentioned in the beginning deliberate practice sets itself apart from just practice, by being defined by specific principles.

Goal setting

Here goal setting refers to relevant and well-defined goals for your practice. This is not to be confused with end goals, such as running a marathon or learning to play an instrument at a high level. The goals for your practice must be points along your path that are relevant to where you currently are. If you want to improve your tennis game, you shouldn’t be playing the same match against your partner each week and expecting to suddenly beat him. Instead, your goal could be to practice your backhand return of fast serves, as you notice that this is where you lose points.

One American basketball coach observed that his players missed too many free throws in their games. Realizing that these young players failed under the pressure, the coach set up a drill to work on this. During training, players would suddenly have to run around the court or do series of pushups if they missed their free throws. While this is anecdotal and the results hard to verify, the rationale is sound. This is deliberate practice with a well-defined goal: shooting your best at free throws while under pressure, to optimize the similarity between practice and games.

Leaving the comfort zone

When we leave our comfort zone we start to struggle, and we might hit a wall. Getting past such a wall will often require us to rethink our strategies and methods. With these changes comes an opportunity for growth in skill. Ericsson says in his book Peak: “This is a fundamental truth about any sort of practice: If you never push yourself beyond your comfort zone, you will never improve.”

Here we can think of how this relates to empirical evidence. What happens to us when we never leave the comfort zone? Scientific studies of not just psychologists but also physicians and other professions find that those with decades of practical experience are not more skilled than those with far less experience. One suggested explanation for this is that practitioners fall into a routine where they rarely experience something that’s new and thus are rarely challenged. The largest study of psychologist’s own experience of development over their career has shown that the therapists grow more comfortable as they become more experienced. This could support the explanation before as the therapists simply stop experiencing the feeling of being challenged and therefore might no see a reason to try new strategies.

Focused effort — an unpleasant state

Deliberate practice is further defined by a focused effort in practice sessions. This connects logically to the above points — a focused effort is simply necessary because deliberate practice takes place mostly outside your comfort zone. The effort that goes into applying such a focus makes the practice demanding and thus deliberate practice is almost by definition not very pleasant. Again, we can think of the example from Tennis. While your backhanded serve-returns should dramatically improve with your deliberate practice, it is almost guaranteed that playing matches will be more fun. It is the focused effort that enables you to succeed outside your (previous) comfort zone and find new strategies, and therefore a focused effort is crucial to your development.

Get feedback!

The research on learning and expertise has led to different models and varying conclusions on how best to learn and improve. However, if there is one factor on which the research seems to converge towards a consensus, it is the importance of feedback. When Shanteau lists the professions where experts do well and where they don’t, a clear pattern shows itself. The professions with great experts are professions where feedback is readily available. Similarly, this goes into Hogarth’ terminology of a Kind Environment, with a focus on the feedback not being misleading. In Ericsson’s research, feedback is part of the deliberate practice approach and can be optimized if a trainer is available to guide the process.

Feedback is often interpreted in a limited sense of the term, i.e. something you get at various time from a superior in your occupation or as part of the last team building exercise. But feedback can be understood more fundamentally as almost anything that comes as the result of something you do. Feedback can be asking friends or colleagues on an ongoing basis for what they think you could do better. If you’re playing golf it can be keeping track of your golf putting from various distances. If you’re into tennis you can take notes to whether your fast-flat serve or your slice serve makes you the most points.

Here it’s interesting to take another look at clinical psychology. We established earlier that psychologists feel less challenged as they gain expertise. Another possible explanation for the stagnation in development could be a lack of available feedback in the profession. Earlier studies amongst psychotherapy clients have shown that clients are often unwilling to bring any negative feedback to their therapists. This disrupts the feedback mechanism available to the therapist and makes it difficult to identify ineffective or harmful elements in therapy.

In recent years the practice of Feedback informed treatment (FIT) has increasingly been implemented in psychotherapy. The principle behind this is a short measurement scale that the client can fill out after each therapy session, securing that feedback is obtained. The psychologists of these clients can, therefore, pay special attention to clients who are not progressing and can ask some of these directly as to what could be changed in therapy. Interestingly some of the studies suggest that the therapists with access to the FIT system improve over time — in contrast to the general tendency where therapists do not improve. Thus, these results indicate that continuous development as a therapist is hard rather than impossible. While research is an ever-ongoing discipline and the FIT system isn’t finally verified, this at least speaks to the importance of feedback.

Feedback is simply the mechanism which lets you know when you did something good and when you did something bad. If we return to the words of physicist Niels Bohr in the beginning, we can see these mistakes — when they can be identified as such — not as something negative, but as small steps towards expertise.

Limits of deliberate practice

Where Gladwell might have painted an idyllic image of his 10,000-Hour Rule, something similar could be said about the impression left on leaders of Ericsson’s book Peak. To be sure Ericsson is far less sensationalistic than Gladwell, a bit less anecdotic and backed by research to a much higher degree. He never says directly that deliberate practice comes with a guarantee for success, but the many anecdotes included are all examples of great successes brought about by deliberate practice. Characteristically, a great emphasis is put on the observation that rarely — if ever — has anyone become an expert without great amounts of practice. It’s important to understand, that this is an observation of the limits of talent as a factor, not of its lack of relevance and potential in the first place. Readers missing this distinction could be left with the impression that deliberate practice is all there is between them and becoming an expert in whatever domain they desire.

There are limitations to the potential of deliberate practice and some of these are more tangible than others. A self-evident example is the sport of basketball, where physical stature is virtually crucial to success. Since height is exclusively genetically determined, no amount of practice will make up for this. But there are many other and less obvious domains in which deliberate practice is only one part of the bigger picture.

Sports researchers have argued that talent can predispose players to gain more from each practice and therefore get motivated to practice more. Further players with good genetics might be particularly injury free and therefore practice even more. Both effects will result in more training, which when observed through the lens of deliberate practice-glasses will be judged as the reason for their improvement. As far as this is what happens the effect of superior talent could be confused with the effect of superior discipline in continuous practice efforts.

But these are all theoretical and hypothetical limitations. Luckily as deliberate practice has gained much focus over the years many other groups of researchers have taken it up as their subject. So, let’s look at what empirical tests of the effectiveness of deliberate practice show.

What the data says about Deliberate practice

The biggest study to date looking into deliberate practice is a meta-analysis published in 2014, by three US-based researchers. The researchers pooled the findings from 88 scientific studies in their analysis and looked at the effects in a range of domains including music, games, sports, professions, and education. Their results might surprise readers of Peak and other sources.

The research on deliberate practice is based on the rather abstract statistical term ‘explained variance’. For readers not familiar with this, a quick ad hoc view will do. Whenever a sample of study participants are tested these will naturally vary in their characteristics — in this case, their abilities in the domains studied. A simplified way of looking at explained variance is as a measure of how much of an individual’s level of ability can be predicted from his or her amount of deliberate practice.

Across all domains tested the researchers found an overall average effect of 12%. So, while on average you can explain and predict 12% of a person’s skill level from how much he or she has practiced deliberately, 88% is left to be explained by genetics, environmental factors, and randomness.

This overall effect size has made some headlines — at least in the scientific community — but disguises a significant variability between the domains covered. On one side of the spectrum, deliberate practice is found to explain 26% of the variance in the category of games. For music, the number is 21% while it is 18% for sports. But here a very significant difference shows up as the number for education is found to be 4% while for professions it is 1%.

These are results which profoundly raises questions about the validity of the messages in books such as Peak and Talent is overrated. If the pattern for violinists is as clear as what Ericsson and his colleagues originally found, this simply doesn’t seem to generalize to other contexts. This brings us back to the importance of the domain. Could playing the violin be a very special activity in the way that practice is unusually important? The meta-analysis comes close to this topic when looking at music, but even here a value of only 26% explained variance is found.

The authors make certain reservations and do not claim that 12% is a conclusion set in stone on the effect of deliberate practice. Though they point out that even if very conservative statistical assumptions are applied, the overall effect would at most be found to be 19%. Even in that case, four-fifths of the individual difference in skill level is explained by factors other than deliberate practice.

Factors mentioned by the authors themselves include starting age and general intelligence. Earlier research on chess players has found that if two players have had equal amounts of deliberate practice, the one who started playing earlier will most often be the better player. This suggests that the age at which you start practicing matters — and not just because starting earlier gives you more total practice time. General intelligence is widely found to be a predictor of success in domains such as music, academia, and professional occupations. Finally, the authors recommend that future research looks into the role of individual differences — since the empirical evidence across domains does not support the claims made by Ericsson and colleagues.

Conclusion

For conclusion, we will return to the questions from the introduction.

How long might it take for one to make all the mistakes in any given domain?
Gladwell presented the 10,000-Hour Rule and spread this far by telling his readers that this was the clear pattern amongst experts known to all, such as Bill Gates and the Beatles. When we look at the reasoning behind this rule, it is clear that this is weak, especially since the number itself is arbitrary.

If we define expertise — roughly as Shanteau does — as being the best in a field, it follows logically that something needs to set one apart from the rest. This something could be one’s amount of accumulated practice. It’s tempting to think that expertise comes from working harder than the rest and therefore that a rule of some x number of hours can be established. But this logic hinges on the assumption that practice is all that matters. Since the research shows that this hypothesis doesn’t hold up, we simply cannot prescribe some x number of hours across individuals. It seems all we can say about the duration is that expertise development is a long process.

What would be the best approach to making all these mistakes and learning from them?
As we have seen both from the literature and empirical evidence, there is a strong risk that when we stop making mistakes, we also stop learning. Thus, if we want to improve, we should not hide from the situations where we risk making mistakes but seek out these situations instead. Part of this is to break our routines because we grow accustomed to our usual challenges and stop making mistakes.

Making mistakes is central to the question, but only insofar as we can recognize and admit them. This means we should make an effort to spot our own mistakes and improve from them. Further, it means we should think about whether our domain allows us to spot mistakes and what kind of errors we might be making, without receiving proper feedback to inform us.

How does the view of Bohr fit in with the contemporary research of learning and expertise?
The wording of Bohr’s definition is interesting. Had he phrased it as ‘a person having made all the mistakes’, it could be said to be incomplete, simply because you still need to understand this and adapt your approach. But Bohr puts emphasis on the expert finding out ‘by his own painful experience’. This difference, between just making mistakes and finding out which mistakes are made, is what the research shows is crucial for learning. He isn’t explicit about changing one’s behavior when finding mistakes, but here we must give him the benefit of the doubt.

But Bohr’s isn’t a definition in the first place. He was far ahead of just defining expertise as he implemented a description of the road to expertise in the very definition. And Bohr was far ahead of the research world too, as his definition still seems to holds up.

Thanks for reading along. Being no expert writer, I welcome your feedback in the comments.

Note: this post has been posted to medium.com also - link in comments