Why practice doesn't make perfect

Summary: In any field, 10 years seems to be the minimum time needed to attain world-class level performance. However, most employees and even quite a lot of hobbyists have more than ten years or 10,000 hours experience in a certain field -why then are so many of them just mediocre at it? Is it lack of innate talent, or perhaps something else?

A while ago I was discussing with some acquaintances the rule that ten year of practice is necessary to achieve world-class performance. One of them said: then we must be world-class, for we have worked in our field for over ten years!

My acquaintances may have been world-class in their field (I wouldn't know the ranking order in their profession). Generally observing people around me, however, I must unfortunately conclude that 10 years of practicing anything does not necessarily guarantee top performance. Not all people who have played tennis for ten years can get into the semi-finals of Wimbledon, and not all professors win Nobel prizes (and are even not nominated for them!), even if they have been active in their fields for over twenty years. Apparently, while ten years of experience in something seems a necessary condition, it apparently is not sufficient to make one's performance excellent.

To understand why one's performance does not necessarily improve with practice, one should first realize that great performance is the exception rather than the rule in human learning. After all, most of us do most things decently, but rather mediocrily. Einstein might have been a great theoretical physicist, but he was not a very good teacher, husband, cook, actor, public speaker, juggler, driver, and so on. He wasn't even very good at experimental physics. The general pattern in human learning would be that we improve our performance in something until we reach a sufficient level, and then we stop improving.

At first sight, that may seem silly. Why stop improving when we could continue getting better?

Actually, the reason to stop improving is probably entirely practical. Our brain just 'automates' a process that we do well enough. This means that the different steps of the process are fired off' in the right order without us needing to think about it. This has some very nice advantages. First of all, it doesn't cost too much brain power, we can perform it efficiently under stress and time constraints, even when fatigued, and we can free the rest of our brain for planning and learning other things. It is incredibly handy that we can spend the time we are tying our shoes or brushing our teeth pondering how to ask our boss for a raise, instead of trying to create improved ways of shoe-tying (or just having to continually concentrate on how to move the toothbrush) all the time. Having habits and automated reflexes for things that are relatively unimportant or easy saves valuable brain power for things that are important and complex; this is similar to a good process engineer who would not pay much attention to the color of the paint in the factory hall, in order to be more aware whether the meters are indicating that an explosion is imminent.

The process of learning, in most cases, is like a river starting to form. During the first few attempts the water will try all sides to find a good way; but after a while the stream finds a decent way, and as that path is transporting more and more water, it hollows the ground, which creates an even more attractive trajectory for water, which draws even more water, and so on. The final path may not be the best path available- most likely, it is just the first accidentally encountered path that was good enough. However, it is good enough. And for most purposes, that is sufficient.

As an analogy, consider a young German from Berlin who wants to visit Amsterdam. Since there are no road pointers in Berlin indicating in which direction Amsterdam lies, he decides to execute the old heuristic and drive to Rome first, since, as the saying goes, all roads lead to Rome. In Rome there may be a sign or a friendly Italian pointing the way to Amsterdam, so the German learns to reach Amsterdam via Rome.



Since Amsterdam is such a great place, the German travels there many times during the following years. He learns the route (via Rome!) by heart, and buys a faster car, so he gets in Amsterdam quite a bit faster than his first tentative journey. But of course, he is still taking a huge detour and will continue to do so until he decides to find a faster way, or an acquaintance accidentally points him towards a road map.

While this story may seem somewhat absurd, similar trial-and-error and consolidation of the first more-or-less effective pathway seems to be the way in which the brain learns. A good-enough path is stored, and as it is traversed many times, it becomes faster and automatic; but it may not be the best path (or method) possible. The only way to continue improving is to regularly ask oneself (or others/books/teachers) whether it would be possible to increase performance by doing things differently.

So... are my two acquaintances mentioned before really at the top of their field? As I have tried to make clear, not necessarily. Ten years of experience may just be one week of experience repeated 520 times just deepening the inefficient initial riverbed; or it may be a systematic effort that gradually elevates the practitioner to world-class levels. As a rule of thumb, however, work in general does not tend to build much skill; for a part because most jobs consist of doing the same things over and over, but also for other reasons, which can also learn us much about talent and excellence (or lack thereof). But that may be good to discuss next time.