Thursday, October 20, 2016

Ericsson notes

Anders Ericsson et al
Role of Deliberate Practice (aka 10,000 hours aka talent doesn't exist)
http://projects.ict.usc.edu/itw/gel/EricssonDeliberatePracticePR93.pdf

raw notes:

Ericsson claims that the level of expert performance/competency is directly related to the time committed to deliberate practice (10000 hours theory)
The biggest issue is that you can dismiss a case that does not correlate as “not sufficiently deliberate,” but this is not very worrying considering the strength of the positive accounts


Talent:
The science of talent is so complex, muddled, and non-conclusive that the concept itself is barely useful— rather, it is decidedly non-deterministic. The relationship between high performance and genetics is laughably feeble compared to its relationships with upbringing, motivation, and labor. In a mental sport (as opposed to a physical sport in which height and build, obvious genetic factors), natural ability so much less important as to be sufficiently useless, particularly when this ability cannot be specifically identified.

Throughout the history of the olympic games records have improved steadily and dramatically in every single event, even those with static rules and conditions.

The fastest rate of typing in 1904: 82 wpm
in 1923: 147 wpm
1959: 176 wpm

Perfect pitch is actually instructable in all young children and even adults (although adults will require a substantial amount of training).

A high IQ (already reflecting both genetic and environmental factors) has next to no real correlation with success in competitive chess.

At the time it was written, the best violinists in the country considered Tchaikovsky’s violin concerto unplayable. Today it is standard repertoire. Musical ability is so increased that in blind tests, world-class pianists of 75 years ago are lackluster at best today.

This all shows that the standard for excellence is steadily increasing as information accumulates. The human genome remains the same. At the time, the best violinist in Russia was obviously considered “greatly talented” but was objectively worse than most professional violinists today in terms of ability and performance. Today’s violinists are no more talented but instead reap the benefit of optimized methodology.

Neither does talent manifest as “learning more faster.” Expert status is uniformly achieved in an average of 10 years from initiation. Prodigies that begin earlier actually achieve mastery slower due to development and maturity. Even outliers (such as Fischer) don’t undercut 9 years. In terms of efficiency in a similar deliberate routine, the exceptional are not so exceptional.

Finally, it is possible that talent is a gateway limitation as opposed to a performance cap limitation but this is unlikely.
While aptitude tests do predict short term success, any advantages quickly disappear over time as experience accumulates. Aptitude has very little influence on performance over time. In the following model, predisposition to optimizing practice is more valuable or identifiable with talent.

Thus talent is, as far as empirical or pragmatic consideration for method go, a sufficiently useless concept outside of spurring activity and motivating someone to achieve.

-

more time in practice does not = increased performance
in fact, in many cases it can decrease it due to neglect and loss of attentiveness (performance and experience has a negative correlation in doctors).

Duration, structure, and intensity of training is the worth.

optimal practice requires:
motivation (incl focus)
information
optimization
feedback

stagnation despite motivation indicates a change in method is necessary
a teacher is there to monitor and shift the task to accommodate
the average personal-tutored student performed at the 98th percentile of students taught conventionally in a classroom setting
by focusing effort and solving for specific problems

transition from student to top professional is marked by transitioning from a mastery of the mechanics into an insight into the structures and innovative use of to arrive at new interpretations
this, I think, goes back to top-down to bottom up learning
this bit of the article is far less specific and I don’t think it is very well constructed
creativity is a wrench in all of this
but creativity is not a high priority among performance criteria

in early stages, exceeding 2 hrs a day shows greatly reduced benefit
practice without full attention defeats the purpose of practice

like training muscles, any training activity must actively push the boundaries of our ability in order to increase it.
strain, then recovery time. Both are essential.
Progress is limited by the limits of either. (stagnation/burnout)
novelists tend to work for only 4 hours in the morning, intensely, then break completely from their (long-term) creative work. This allows for recovery. Music students that go too hard too fast tend to give up their instrument at the first sign of fatigue.
The key, then, is slow adaptation. Burnout motivational or otherwise is a symptom of too-rapid a change.
Just as deliberate practice itself, in order to allow for our full attention and learning, starts with low-level tasks then slowly builds in scope, difficulty, and complexity, the level of commitment in terms of a schedule must also start small then gradually grow as our capacity increases naturally through adjustment. This can be optimized to avoid the consequences of overreaching and burning out while avoiding those of underreaching and stagnating.

mental fatigue is real https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3562167/
but much more subjective than measurable. The brain makes you perceive it very early 'cause it's potentially sooo damaging.
(in terms of performance, mistakes increase over time even with focus. A tougher mental load will also dull reaction time, though it will not increase rate of mistakes.)

novelists tend to work for only 4 hours in the morning, intensely, then break completely from their (log-term) creative work. This allows for recovery. This is matched by expert pianists and is accepted as a general upper limit for sustainable concentrated individual practice.

The key, then, is slow adaptation. Burnout motivational or otherwise is a symptom of too-rapid a change.

Just as deliberate practice itself, in order to allow for our full attention and learning, starts with low-level tasks then slowly builds in scope, difficulty, and complexity, the level of commitment in terms of a schedule must also start small then gradually grow as our capacity increases naturally through adjustment. This can be optimized to avoid the consequences of overreaching and burning out while avoiding those of underreaching and stagnating.

again, high levels of motivation are correlated with high frequency of practice and performance

getzels and csikszentmihalyi (1976)
Artists are drawn to painting most often because it allows for isolation. But when faced with the need to network to support themselves, a large number retreat not just from painting professionally but from painting altogether at the very moment that they commit to an alternative day job that prevents them from a quality routine. This indicates that the activity itself is not inherently motivating but rather producing quality work.

Practice is decidedly not motivational in itself. Thus motivation must precede devoted practice. It may come easily from social setting or goals short or long-term, most often through performance in competition etc.


Violinists
in young violinists in a top school, number of competitions entered correlates with skill as does repertoire in memory and length of practice sessions
practice alone is perceived as the most valuable activity. Averages 3.5 hrs a day (sessions averaging 1.3 hrs) among top performers, not varying greatly weekday vs weekend. Alone before lunch was preferred.
Average of 60 hrs (~7 per night) of nighttime sleep which also correlated with more effort compared with other students. Additionally, they nap more (3hr a week and usually between 2-4 after practicing).
They spend 3.5 hr a day on leisure, worse performers spend 4.3 and the average person spends 5.2. Direct negative correlation with time spent with music. Interestingly, the top performers were far more accurate in estimating their leisure time because they were better at time-management.

Experts should be able to accurately report/estimate time spent practicing because deliberate practice requires close scrutiny of time spent.

Practice needed to achieve a high level of mastery is massive. Practice needed to maintain or regain a level once achieved is significantly less.


Scientists
80 hrs a week in preparation
focused and extended work developing and refining generated theoretical solutions to specified general problems
knowledge-transform as opposed to knowledge tell
more eminent scientists publish significantly more. Darwin, Pavlov, Skinner, etc all started every day with a few hours of writing.
generally speaking, motor activities are enhanced in the afternoon, intellectual in the mornings.

Ericsson uses the word eminent for going past national notoriety and into genius by way of making new contributions. He claims that this is achieved by first achieving expert then directing concentrated effort at a specific goal (which is honestly not much of a claim lol)



2 comments:

  1. Are you into "classical" music?

    ReplyDelete
    Replies
    1. I listen to glenn gould, nancarrow, and a little debussy regularly.

      Delete