10000 hours

[hyperpape]

I happened across a blog post on the idea that the key to expertise is a lot of practice--10000 hours is the figure typically quoted. Turns out that while it is common for experts in diverse fields to have put in 10000 hours of practice, there's a huge huge individual variance--a finding which seems to have escaped Malcolm Gladwell. Moreover, athletes do not seem to follow the same pattern. Try to contain your surprise. http://www.sportsscientists.com/2011/08 ... crets.html

[RobertJasiek]

An amount of time alone says little. It matters how you use it. E.g., a dan's some 20000h for rules study equal much less than 1 rank. The European Champion has said that he does not know precisely what made him strong except studying a lot of everything he got in his hands. It is similar to my experience when I improved a lot, except that a few particularly good books helped me much more (factor up to 500:1) than the vast rest.

[quantumf]

Nice article. That 10,000 thing was becoming a bit cliched. Seems the 10,000 may be one of those causation!=correlation things.

[hyperpape]

An amount of time alone says little. It matters how you use it. E.g., a dan's some 20000h for rules study equal much less than 1 rank. The European Champion has said that he does not know precisely what made him strong except studying a lot of everything he got in his hands. It is similar to my experience when I improved a lot, except that a few particularly good books helped me much more (factor up to 500:1) than the vast rest.

In minimal defense of the 10000 hours idea, its proponents usually talk about "effortful practice" as a necessary condition of expertise.

[Joaz Banbeck]

More than just 'effortful practice', but as Kageyame might phrase it, 'correct practice'.

There is a trite phrase that 'practice makes perfect'. This is not necessarily true. Rather 'practice makes permanent'. One can reinforce bad habits as easily as good ones. One must practice the proper things to get good results.

Ten thousand hours of correct practice may make one an expert.

[Tami]

And who's counting? Was Jimi Hendrix born with a practice-o-meter attached to him or did Lee Sedol's parents and trainers stand by with a stopwatch?

I think really the important thing is how you use your time: effortful study, where you're stretching yourself and actually taking on new ideas and attempting to do things you cannot yet do, is going to be much better than repeating stuff you already know or, gasp, goofing off.

Another important thing is that it takes time to show improvement, because of the way relational memory works. You can try until you're blue in the face, but sometimes the only thing that will see progress is time. What seems hard now often feels easier a few months later. You can study tsumego for ten hours a day if you like, but you won't notice the difference for many days. It's just the way learning works.

[EdLee]

Rather 'practice makes permanent'. One can reinforce bad habits as easily as good ones.

Very true. 10,000 hours of bad habits will make them quite permanent -- very polished bad habits.

[Mef]

If I recall, the original thesis was that 10,000 hours was more or less a necessary (but not sufficient!) condition of becoming an expert, which was loosely derived from the fact that it meant someone spent a significant number of years doing something as a full time occupation. The problem is that many try to flip the statement around and think that 10,000 hours is the only thing required to be an expert. Regarding athletes, I would imagine that they put in 10,000 hours toward athletic activity quite early on, even if it's not dedicated to their final sport of choice. The time spent training though should readily transfer, as so many traits are favorable across most athletic activities (strength, speed, agility, etc).

Edit: I went ahead and read the link now, it seems to be more or less what the author is saying, though I disagree with much of what he says about sports....I think that only tracking the hours one has put in since entering competition is a naive way of measure "athletic practice"

[Ortho]

If I recall, the original thesis was that 10,000 hours was more or less a necessary (but not sufficient!) condition of becoming an expert, which was loosely derived from the fact that it meant someone spent a significant number of years doing something as a full time occupation. The problem is that many try to flip the statement around and think that 10,000 hours is the only thing required to be an expert. Regarding athletes, I would imagine that they put in 10,000 hours toward athletic activity quite early on, even if it's not dedicated to their final sport of choice. The time spent training though should readily transfer, as so many traits are favorable across most athletic activities (strength, speed, agility, etc).

Edit: I went ahead and read the link now, it seems to be more or less what the author is saying, though I disagree with much of what he says about sports....I think that only tracking the hours one has put in since entering competition is a naive way of measure "athletic practice"

Agree totally that this is the correct interpretation. I think that an interesting aside that I read somewhere else as well, however, is that people have often achieved some large percentage of the competence that they will acquire in something after a much smaller period of time--1000 hours at most. I can't lay my hand on where I read this right now, but it is extremely unusual for someone who goes on to become an expert to not already be outperforming expectations after 1,000 hours and it is very unusual for someone to be quite bad at something for 1,000 hours and then go on to be exceptional. Of course, there is some obvious survivor bias involved in that (people usually keep doing things they are good at and quit things they aren't), but I don't think that people usually need to practice something for that long in order to know whether they can eventually become an expert in it.

[hyperpape]

The time spent training though should readily transfer, as so many traits are favorable across most athletic activities (strength, speed, agility, etc).

This is plausible, but also potentially trivializing.

If I study logic or combinatorics for two years, you might expect that it would be of some assistance to computer programming, and I won't quite need my full 10000 hours. But now what about calculus, engineering, machinery--you can make a case that each of these things should contribute a bit, if not as much as the first two subjects. And you can imagine other subjects that have an effect too, though they'll be even smaller.

Now, either those effects are all quite small, as if studying 100 hours of logic saves you nothing more than the hours you'd have had to study logic while learning to program, or you will stop seeing experts have 10000 identifiable hours of practice. Everyone will have a different background, so some will achieve expert status quickly because of transfer, while others will take forever.

But then the whole point of single-minded focus on one area of study starts to look wrongheaded.

[Solomon]

Only 2,483 hours left until I become an expert!!

[Mef]

If I study logic or combinatorics for two years, you might expect that it would be of some assistance to computer programming, and I won't quite need my full 10000 hours. But now what about calculus, engineering, machinery--you can make a case that each of these things should contribute a bit, if not as much as the first two subjects. And you can imagine other subjects that have an effect too, though they'll be even smaller.

Actually, engineering is a great example, and I would consider mathematics and problem solving to be the equivalent of strength and speed. Many starting engineering professions will be looking for a mechanical engineer, but are willing to hire, electrical, chemical, etc, because at the end of the day the bulk of what you need to do is mathematical modelling and problem solving (which any engineer has spent a lot of time investing effort into). Transferring skills from one engineering field to another is not terribly difficult as long as they can handle the equations (often times the equations even look the same, the variables just stand for different things). Likewise, looking at financial modelling/investment firms and the like -- they will often search for candidates from any field that would that would require a strong mathematics background.

If you have a very strong mathematical background and good intuition/problem solving abilities, it takes a significantly reduced effort to go from being in one technical field to another.

[Kirby]

How many hours of forum posting is necessary to become an expert at it?

[Kirby]

More than just 'effortful practice', but as Kageyame might phrase it, 'correct practice'.

There is a trite phrase that 'practice makes perfect'. This is not necessarily true. Rather 'practice makes permanent'. One can reinforce bad habits as easily as good ones. One must practice the proper things to get good results.

Ten thousand hours of correct practice may make one an expert.

I think you make a good point, but this type of thinking has deterred me from playing games in the past. I would worry, if I play more games that aren't of high quality, I'll reinforce bad habits, and get stuck with them.

For better or for worse, I think I'm starting to get over this. I have been trying to play every day. Interestingly, despite lack of study, I have been improving (I think) faster than when I studied a lot without playing. It could very well be that I'm just reinforcing bad habits. I don't really know.

But my motto these days is, "just do it". I want to get good at playing go. So I "just do it". I play go. If that doesn't work out, I don't care. At least I am playing go, and I like that. When I was freaking out about learning bad habits, I didn't even play go. And what's the good in that? I'd rather be bad at go and be able to play it often than to be good at go and never play it.

[Ortho]

I think that as long as you are looking for mistakes, "practice makes permanent" isn't a big deal. The real problem comes up when you are training yourself to make mistakes, doing something wrong and doing lots of repetitions of it, without discovering that you're doing it wrong.

This is a big debate in language learning. On one hand, some people say that you should just speak as much as you can and try to do it with people who will point out your mistakes so you can improve. There is another camp that says that you should learn to say things perfectly and never say anything that you can't say perfectly. IMO, the first group get so much more exposure that they end up running circles around the latter group, just like people who play a lot of casual games of go get so much more experience, that they see and identify so many more mistakes that it must be better.