Some excerpts from my favorite book on motor learning (implicit > explicit):
"Modern coaches are subject to considerable pressure to nurture the skills of the learner to an expert level in the shortest possible period of time. In order to do this the coach will persuade the learner to adopt certain fundamental methods and techniques that may be in vogue at the time or that somehow have become embedded in coachlore as “best practice.” The coach may even persuade the learner to adopt certain techniques simply because they worked for the coach. Whatever the reason, the information is usually communicated in a manner that leaves the player consciously aware of how to execute the technique.
Often the default for the coach is to use explicit verbal instructions simply because the coach knows of few other ways in which to bring about the recommended changes in technique. It becomes very difficult for the performer to carry out the skill without at some point exerting conscious control over the movements and it is this reinvestment that can lead to skill breakdown, particularly if the performer is highly motivated to perform well, which is not surprising given the huge cost and the intense and enduring commitment that is required to become an expert (not to mention the rewards that come to those who are victorious).
As all performers know – and the Zen Buddhist teacher Daisetz Suzuki articulated – best execution of skills occurs when there is no interference from consciousness:
Thinking is useful in many ways, but there are some occasions when thinking interferes with the work, and you have to leave it behind … It is for this reason that the sword moves where it ought to move and makes the contest end victoriously. -D. T. Suzuki, Zen and Japanese Culture , 1959
Considerable evidence now shows that reinvestment can be avoided if motor skills are acquired implicitly without recourse to hypothesis testing or instructions and thus the subsequent accumulation of consciously accessible explicit knowledge about the movements. After all, if athletes do not have access to explicit knowledge of how they move, how then can they use such knowledge to consciously control their skills?
Despite the significant challenges to the coach in overcoming these difficulties the advantages associated with learning skills implicitly are significant. Recent work has suggested that skills that are learned implicitly are performed better in front of an audience, create less interference when the performer has a complex decision to make, and even appear to be resistant to physiological fatigue. Beginners who learned implicitly to perform a rugby pass maintained their passing accuracy when significantly fatigued at an anaerobic or an aerobic level, whereas those who learned explicitly did not. Furthermore, one year later, passing performance remained robust under physiological fatigue, despite the complete absence of passing practice during the year. Such advantages have been explained using an evolutionary framework which suggests that skills were learned unconsciously (or implicitly) far earlier in our evolution than they were learned consciously (or explicitly). Long before we could verbalize we needed to be able to run, throw, and hit with considerable competence. If not, we did not survive. Consequently, unconscious implicit processes evolved with a degree of resistance to psychological stress, distraction, and even physiological fatigue.
Goal-directed movement abilities – such as walking, reaching, or grasping – develop before the language abilities of children develop so it seems probable that implicit learning is by default the primary means by which children acquire motor skills.
Analogies can be used to present the key coaching points of a skill to be learned as a simple metaphor that can be reproduced by the learner without reference to, or the need for manipulation of, large amounts of explicit knowledge. This allows the coach to instruct the performer implicitly without resorting to the use of verbal instructions. Not only do analogy learners seem to quickly identify and mimic the fundamental form of the skill that they are trying to produce, but they do so without acquiring explicit knowledge of how they are producing the skill. There is empirical evidence available to show that in tennis (and table tennis) a right-angled triangle analogy can be employed to teach beginners a topspin forehand shot implicitly. The learner is instructed simply to strike the ball by bringing the racquet (or bat) squarely up the hypotenuse portion of the right-angled triangle. It is vital, though, that the concept of a right-angled triangle and its hypotenuse is familiar to the learner. Strangely, tennis coaches often tell beginners to “brush up the back of the ball” in order to impart topspin. How meaningful is such a concept? With respect to the right-angled triangle analogy, most learners for whom Pythagoras's theorems are meaningful will automatically take a Western-style grasp of their racquet, accompanied by a stance that naturally allows the hips to rotate and the racquet to travel from low to high with the kind of force necessary to generate topspin and so on. The learner is unlikely to be consciously aware that he or she is using any of these important causal rules that underlie a topspin forehand.
Moreover, as studies have shown, the learner is likely to display performance advantages that are characteristic of an implicitly learned skill, such as stable performance under pressure or even when physically fatigued.
A large literature demonstrates that analogy making is a powerful cognitive mechanism by which children gain understanding of their world so it is likely that the performance advantages of analogy learning in sport also hold for children.
The “chase a chook” analogy highlights one of the important advantages of analogy learning; it appears to allow many pieces of information about a skill (i.e., rules or instructions) to be presented to the learner in one manageable “chunk.” This contrasts with traditional coaching methods which involve the explicit presentation of many individual bits of information about how to move. Considerable practice is required before the learner can integrate these pieces of information into a manageable chunk, as is clear when one considers the length of time that it takes for most tennis beginners to learn how to make a ball toss with one hand, scratch the back with the racquet in the other hand, bring the back foot through alongside the front foot, flex the knees, snap the wrist, and strike the ball. And that is only a first serve. “Chunking” has been examined in depth with respect to its critical role in human memory facilitating the organization of very large amounts of information. An advantage of chunking in analogy learning may be that considerably more information than normal can be presented to the learner in a short period of time. Presumably, analogy learning should therefore provide a faster route to expertise.”
"The consequence of this aggressive search for effective skills is that, over time, the performer accumulates a deep pool of explicit rules and knowledge. With practice, the skills become expert and automatic but they are inescapably linked to an explicit, highly verbal mode of control. At inopportune moments (e.g., match point down in the fifth set) or when too much time is available in which to construct the necessary movements (e.g., a gently lofted catch to the outfeld), verbal modes of control can sometimes cause the normally fluent skills of the expert to regress to the awkward, error-prone movements of the beginner. In essence, conditions such as over-eagerness to perform well or too much time to think can result in reinvestment : the tendency to consciously attend to knowledge that underpins the skill in order to control the quality of performance. Most athletes have, at some point in their careers, faced this problem.
There is now evidence that shows that the larger the pool of explicit knowledge that a performer has accumulated about how to perform his or her skills, the greater the chances that reinvestment will occur, especially under pressure.
Similarities exist in the sporting world, where athletes can occasionally become so conscious of flaws in their technique that they begin to display pseudo-clinical skill disorders, such as the “yips” in golf or “dartitis” in darts. These athletes are likely to have very high propensities for reinvestment."
"Would-be superstars, weekend warriors, or seasoned sport celebrities aggregate explicit knowledge about how they perform their skills by testing hypotheses and taking instruction regarding the most appropriate way to move. Conscious access to such information is often disruptive to performance, especially when performers are highly motivated to succeed. There is evidence that practical advantages are associated with forms of implicit motor learning that avoid the accumulation of explicit knowledge about the mechanics that underlie the skill. The coach has substantial influence over the physical and social learning environment of the performer, child, or adult. The willingness to use an implicit motor learning approach raises practical obstacles that a coach must navigate and, although he or she cannot always constrain the environment to provide a comprehensive implicit skill-learning experience, much can be done to bring forward the unconscious and push back the conscious during performance. Implicit motor learning theory asserts that, in particular, early skill-learning experiences should be maximally constrained to be implicit , the learning environment should discourage active testing of hypotheses about performance outcomes, and any form of instruction allowing explicit access to rule structures underlying the skills should be strictly rationed."