Some observations to start: 1. Disciplines that are held to require 'skill' are complex. 2. Humans are not good at actively recalling many elements at the same time, much fewer than the total that constitutes a complex discipline, in fact that is possibly the definition. 3. Being good at a 'skilled' task then requires managing many elements. 3 seems like a contradiction, but this is where the critical component of skill comes from--the 'sense' heuristic. Human brains operate by association and fuzzy evaluation, so this is the means by which we are able to make judgement calls. When looking at a painting or piece of music or position in chess or whatever, it's not possible to weigh on every single component that influences the decisions made. This fuzzy decision making is what we call 'intuition', 'game sense', 'artistic eye', and so on. Mechanics account for the rest of the distinction between the novice and more accomplished. The dilineation between the two is hazy--how much of drawing is accomplished by mechanical action of moving the arm and wrist around, and how much due to higher level functions? Obviously, on closer inspection it unravels so let's say there's a general distinction because we said so. This is the basic structure, however. The heuristic is the decision-making process, and sometimes there's a mechanical component of execution. Anyhow, the much less straightforward element is how to improve that heuristic. In part it can only be done by looking at what other people have done at least for artistic practices because art is a means of communication, which of course is intrinsically social. And plently of the time in strategy actions are based off what the other players are likely to do, so observing someone better is definitely a major aspect of improving. But being able to evaluate what lessons to come away with is maybe a cyclical sort of situation; in order to tell what's important, you need to know what is in the first place... Since anyone can learn isolated facts, and clearly there's more at play than the individuals' capacity for recollection, the difference has to be whether that information is being meaningfully used, 'slotted' into a larger model. As for how one would go about more effectively building up that model/sense...