[dvt]

I don't think that Epstein is being very charitable with Gladwell's 10,000 figure. Gladwell:

a) Doesn't make the 10,000-hour figure a rule as both Epstein and Repanich would have us believe.

b) Fully concedes that there are myriads of other factors that have an effect on how one reaches "expert status".

I'm also not sure how Epstein can claim that the chess master study is somehow disanalogous with the athlete one. An expert athlete might have some physical advantage over your "average Joe": speed, lean muscle, endurance, height, etc., etc. Similarly, one can say that an expert chess player may have some neurophysiological advantage over the "average Joe": better-formed synaptic pathways, higher attention span, etc., etc. I don't think there's any difference between an expert chess master or an NFL quarterback.

Both might have some genetic advantage; both have trained extensively. The idea behind Gladwell's figure is not that you can practice X for 10,000 hours and then you will instantly be an expert at it, but merely that after 10,000 hours (of deliberate practice plus a number of contingencies) you can expect to be somewhere in the realm of expertness.

There are plenty of other studies that favor this hypothesis; in particular, some very interesting double-blind identical twin studies[1].

Outside of extreme cases (e.g. I am 4'11'' and want to play in the NBA; I have an IQ of 90 and want to be an astronaut), I think Gladwell is right on the money: practice is more important than talent.

[1] http://www.indiana.edu/~jkkteach/P335/shanks_expertise.html

[jacques_chester]

The problem is that the chess study shows that some people never became chess masters, no matter how many hours they practiced.

There was a hard upper limit to their performance that no amount of training could surmount. Insofar as "expert" is defined by relative ranking, some people will never become experts. Why? Because they are outranked by the "naturals".

The general problem with the Gladwell/Ericsson hypothesis is that it sorta-kinda suggests that practice is both necessary and sufficient to produce expert-level performance.

Everyone agrees on necessity. It is the question of sufficiency that is demonstrably false. People are different and are better suited to different endeavours; no amount or sort of training or practice can change the boundaries of your phenotypic potential.

You can move a long way from an average baseline in the direction of expertise -- humans are very malleable. But there are hard limits. Elite performers in any field are elite because they were better suited and then did the practice.

[jliechti1]

I think you are missing part of the results of Ericsson's research. He makes a very specific point to distinguish practice from deliberate practice. 10000 hours of practice is very different from 10000 hours of deliberate practice.

My interpretation of Ericsson's work is that natural talent can give certain individuals a head start in the beginning, but once you get into the domain of expertise, this head start at the beginning is very tiny and the effects of practice completely overpower it in comparison. However, I think in areas like sports - genetics do play more of a role. If you are less than 5 feet tall, you probably will not be an NBA point guard, even after 10000 hours of deliberate practice.

One of his papers for further reading:

From The Role of Deliberate Practice in the Acquisition of Expert Performance

"Consider three general types of activities, namely, work, play,and deliberate practice. Work includes public performance, competitions, services rendered for pay, and other activities directly motivated by external rewards. Play includes activities that have no explicit goal and that are inherently enjoyable. Deliberate practice includes activities that have been specially designed to improve the current level of performance."

http://graphics8.nytimes.com/images/blogs/freakonomics/pdf/D...

[jacques_chester]

I'm familiar with the distinction that Ericsson made.

The chess study is based on people who deliberately practiced.

It still didn't matter. Some people are simply better at chess.

Seriously proposing that all humans everywhere have perfectly identical potential in all possible tasks is nonsensical. That is the strongest form of Ericsson's hypothesis. If you weaken the hypothesis -- eg. "some people take less time" -- then why only publish the exact 10,000 average and fail to show variances in the original study?

And if there is variance in time-to-expertise ... what exactly does the "training is required, but the time needed varies based on talent" finding tell us over ... training is required, but the time varies based on talent?

Because that's already kinda known.

[dvt]

And some people will never play in the NBA no matter how many hours they free-throw. Again, I'm not sure how the situations are different.

In some (marginal) cases, you lack a necessary feature of being an expert in X. I don't see how this breaks Gladwell's position.

[jacques_chester]

Because performance is not usually a step function. It is a continuum. If it's a continuum, variance in training time to reach expertise is not surprising.