Lesson 18 · How are movement skills acquired, developed and improved?
The practice that looks worse and works better
This lesson contains the most counter-intuitive finding in the whole module, and it is worth stating before anything else: the practice arrangement that produces the best performance during the session is usually not the one that produces the most learning. A coach watching a session cannot tell the difference, because what they are watching is performance and what they want is learning, and for a while those two run in opposite directions. The same reversal shows up in feedback, where the most helpful-seeming correction can be the one that prevents a learner from ever developing their own.
Two groups learn three tennis strokes over the same six sessions, with the same total number of repetitions.
Group A does 30 forehands, then 30 backhands, then 30 volleys. Tidy, focused, one stroke at a time. Group B does the same 90 repetitions in a shuffled order, never knowing which stroke is coming next.
By the end of session six, group A is clearly performing better, and it is not close. Two weeks later both groups are tested. Group B wins comfortably.
Write down what you think happened, and be specific about what group B was doing on every repetition that group A was not. Then: which group's sessions would you rather be observed coaching, and what does your answer suggest about why this mistake is so common?
- Distinguish massed and distributed practice, and choose between them from the skill and the learner.
- Distinguish whole and part practice, and link the choice to the classification of Lesson 17.
- Distinguish blocked and random practice, and explain why the one that performs worse in the session produces more learning.
- Explain the difference between performance and learning, and why a coach can only observe one of them.
- Describe the performance elements — decision-making, and strategic and tactical development — and how each is trained.
- Describe the six types of feedback and match each to a stage of learning.
- Explain how too much augmented feedback can prevent a learner developing their own error detection.
- Massed practice — practice with little or no rest between repetitions or sessions.
- Distributed practice — practice with substantial rest between repetitions or sessions.
- Whole practice — practising the skill in its entirety.
- Part practice — practising components of the skill separately before recombining them.
- Blocked practice — repeating one skill many times before moving to the next.
- Random practice — practising several skills in an unpredictable order.
- Performance — how well the skill is executed right now; temporary and observable.
- Learning — a relatively permanent change in capability; only detectable later, in a retention or transfer test.
- Task-intrinsic feedback — information arising from performing the skill itself.
- Augmented feedback — information added by an outside source such as a coach or a video.
- Knowledge of results — feedback about the outcome.
- Knowledge of performance — feedback about the movement that produced the outcome.
Two of the three practice decisions are relatively straightforward, and both are settled by the skill and the learner rather than by preference.
| Massed | Distributed | |
|---|---|---|
| What it is | Little or no rest between repetitions or sessions | Substantial rest between repetitions or sessions |
| Suits | Skills that are short, discrete and not physically demanding; motivated learners at the associative or autonomous stage; limited time before a competition | Skills that are physically demanding, dangerous, or require high concentration; learners at the cognitive stage; younger learners |
| Risks | Fatigue, which degrades fine control first and therefore embeds poor technique; boredom; higher injury risk | Needs more total time, and long gaps allow some forgetting between sessions |
The default for a cognitive learner is distributed, for a reason that comes straight from Lesson 16: the movement is consuming all of their attention, so they fatigue mentally as well as physically, and a tired beginner practises errors. And practising an error is not neutral — it is practice. Each repetition of a faulty pattern strengthens that pattern, which is why the phrase "practice makes perfect" is wrong and the accurate version is that practice makes permanent.
Whole and part is settled largely by Lesson 17's second classification. Practise whole when the skill is continuous, when it is short and simple, or when its parts are so interdependent that separating them destroys the timing — a golf swing broken into parts becomes three movements that no longer join. Practise in parts when the skill is serial with genuinely separable elements, when it is complex enough to exceed a beginner's attention, or when one element is clearly the fault. Part practice is not complete until the joins have been practised, and the standard progression is part, then progressive-part — combining elements two at a time — then whole.
Quick check: Distributed rather than massed practice is most appropriate when:
This is Card 01, and it is the most important idea in the lesson. Blocked practice repeats one skill many times before moving to the next: group A. Random practice mixes several skills in an unpredictable order: group B.
Blocked practice produces better performance during the session. The learner settles into a rhythm, each repetition is a small adjustment of the last, and the improvement is visible. Random practice looks messy: every repetition is worse, the learner never settles, and a coach watching would conclude it is going badly.
And yet the random group learns more, because of what they are doing between repetitions. In blocked practice, the learner is repeating a movement they have already retrieved and can simply adjust. In random practice, the intervening repetitions have displaced it, so on every attempt the learner must retrieve the movement plan from scratch and reconstruct it. That reconstruction is the learning. It is effortful, it makes each individual attempt worse, and it is precisely what a match will demand — because a match never gives you the same shot thirty times in a row.
Two qualifications, so this is not applied as a rule that all practice should be random. Blocked practice is right early: a cognitive learner has no movement plan to retrieve, so randomising gives them nothing to reconstruct and simply prevents them forming a pattern at all. And the sensible progression is blocked to acquire, random to retain, moving the learner along as soon as the basic pattern exists. For an associative or autonomous learner, blocked practice is usually the weaker choice and is chosen because it is comfortable to run and looks productive.
True or false: Because blocked practice produces better performance during a session, it also produces better long-term learning than random practice.
The syllabus names performance elements separately from the movement itself, and the reason is the one Lesson 17 established: in an open skill, executing the movement is only half the task. The other half is deciding which movement, when, and why — and it has to be trained deliberately rather than hoped for.
| Element | What it is | How it is developed |
|---|---|---|
| Decision-making | Selecting the right response from the available options, under time pressure and with incomplete information | Practice that requires a genuine choice from unpredictable information — small-sided games, conditioned games, drills with two or more legitimate responses. It cannot be developed by a drill with one correct answer |
| Strategic development | The overall plan for a contest, decided in advance — how a team intends to play, how an athlete intends to approach an opponent | Analysis of the opposition, video review, planning sessions, and rehearsing the plan in training so it is familiar under pressure |
| Tactical development | The adjustments made during a contest, in response to what is actually happening | Conditioned games that create the situation, questioning rather than instructing, and giving players the authority to change something themselves |
Strategy and tactics are regularly confused, and the distinction is simple enough to be worth getting right: strategy is decided before, tactics are decided during. A team's strategy might be to press high; the tactical decision is whether to keep pressing when they go two goals down with twenty minutes left.
The connection to the rest of the lesson is direct. Decision-making cannot be developed by blocked practice of a single correct response, because there is no decision in it. Random and varied practice is not only better for retaining the movement — it is the only arrangement that trains the decision at all, which is why the two halves of this lesson are really one idea.
Match each practice arrangement to what it is best for.
- Blocked practice
- Random practice
- Part practice
- Distributed practice
- Acquiring a new pattern in the cognitive stage, when there is nothing yet to retrieve
- Retention and transfer, and the only arrangement that trains decision-making
- A serial skill with separable elements — provided the joins are practised too
- A physically demanding or dangerous skill, or a beginner who fatigues mentally
The syllabus names six types, and they are easier to hold — and far easier to use — once you see that they are three pairs answering three different questions, so any single piece of feedback has one answer from each pair.
| The question | One answer | The other |
|---|---|---|
| Where did it come from? | Task-intrinsic — from performing the skill itself: what you saw, felt and heard. Available on every repetition, free, and often the only kind available during play in competition | Augmented — added from outside: a coach, a video, a timing gate, a teammate |
| When did it arrive? | Concurrent — during the movement. Useful for slow or continuous skills where it can still be acted on, and impossible for a movement lasting under a second | Delayed — after the movement. Allows the learner to form their own judgement first, which is what develops error detection |
| What was it about? | Knowledge of results — the outcome. "It went wide." Often already obvious from the task itself | Knowledge of performance — the movement that produced it. "Your shoulders opened early." Usually the more useful, because it says what to change |
Knowledge of results versus knowledge of performance is the pair that most improves real coaching. A player who has just shot wide already possesses the knowledge of results; telling them it went wide adds nothing. What they lack, and what only an observer can supply, is the knowledge of performance that explains it. Feedback that only repeats what the learner already knows is not feedback.
The repair is to fade the feedback deliberately: reduce its frequency, delay it, and ask the learner what they thought before telling them. That question is the single most valuable coaching habit in this lesson, because it forces them to consult their own task-intrinsic feedback before hearing a better answer.
Two truths and a lie. Which statement is false?
The syllabus asks for practice methods and feedback types for the different stages of learning, so this table is the dot point's own structure. Read it downward for a stage, and across for how each decision travels.
| Cognitive | Associative | Autonomous | |
|---|---|---|---|
| Massed or distributed | Distributed — they fatigue mentally, and a tired beginner practises errors | Either, according to the skill's physical demand | Massed is workable, since the movement costs little attention |
| Whole or part | Part, for a complex skill, to stay within their attention capacity | Progressive-part into whole, with the joins practised | Whole, in context |
| Blocked or random | Blocked — there is no movement plan yet to retrieve | Shift to random, as soon as the basic pattern exists | Random and varied, at full game complexity |
| Source of feedback | Mostly augmented — their task-intrinsic feedback is not yet interpretable to them | Augmented, but reducing; prompt them to consult their own first | Mostly task-intrinsic; augmented sparingly and on request |
| Timing | Immediate, so the attempt is still available to them | Increasingly delayed, to make room for self-assessment | Delayed, often after a session rather than after an attempt |
| Content | Knowledge of performance, on one element, in simple terms | Knowledge of performance, more detailed and precise | Knowledge of performance only where it is genuinely not self-detectable, plus tactical content |
One direction runs through every row: from external to internal. The coach starts as the learner's source of structure, error detection and motivation, and the entire purpose of good coaching is to hand each of those over. A coach who is still the primary source of error detection for an experienced athlete has not finished the job — and, worse, has made it harder to finish, because the athlete's own detection has had no reason to develop.
Fill the gaps. Type one word in each box.
Practising several skills in an unpredictable order is called practice. Feedback added by an outside source such as a coach or a video is called feedback. Feedback about the movement that produced the outcome, rather than about the outcome itself, is called knowledge of .
Text description
- Two parts, one above the other.
- Above, performance plotted across six practice sessions and then a retention test two weeks later, for a blocked group and a random group: the blocked line rises higher and more smoothly during the sessions, the random line stays lower and more erratic, and at the retention test the two cross over so that the random group scores clearly higher, with a note that the coach can only observe the session portion of this graph.
- Below, the six types of feedback arranged as three question pairs: where it came from, task-intrinsic or augmented; when it arrived, concurrent or delayed; and what it was about, knowledge of results or knowledge of performance, with an arrow above the three rows showing the direction coaching should travel from external and immediate toward internal and delayed as the learner develops.
Activities · put the method to work
This can be done with paper balls and three targets at different distances, and it takes twenty minutes plus a follow-up. It is also a Lesson 15 investigation, so design it as one.
Write the research question with all four elements from Lesson 15: measure, population, conditions and comparison. The comparison is built in here, which is unusual and convenient.
Split the group. Half throw 10 at target 1, then 10 at target 2, then 10 at target 3. Half throw 30 in a randomised order called out to them. Same total repetitions, same targets, same time.
Score the practice, and predict. Record each group's accuracy during the practice, and write your prediction for the retention test before you run it.
Retention test at least 24 hours later, both groups in randomised order. This delay is not optional: without it you have measured performance and not learning, which is the entire point of the exercise.
Then write up the limitations honestly, because with a class-sized sample there are several: the groups were not matched beforehand, the numbers are small, and one enthusiastic thrower can move a group mean. Say what your result can and cannot support — and note that if your result comes out the other way, that is a finding about your design rather than a refutation of the effect.
Odd one out. Three of these are augmented feedback. Which one is task-intrinsic?
For each: classify it on all three axes from Card 07, name the stage of learner it would suit, say what is wrong with it for the learner described, and rewrite it.
To an experienced player who has just shot wide, immediately: "That missed."
To a beginner learning a tennis serve, called out during every single attempt: "Toss higher — no, knees — watch the ball — follow through."
To an associative-stage swimmer, after every length, for eight weeks: a detailed technical correction from the coach on the wall.
To a squad after a match, three days later: "You didn't want it enough."
Scenario 3 is the one with the subtlest fault: each individual correction is accurate, useful and well-intentioned, and the pattern across eight weeks is still damaging. Explain exactly what the swimmer has been prevented from developing, and what the coach should do instead without withholding help.
Explain how practice methods and types of feedback should differ for a learner in the cognitive stage and a learner in the autonomous stage.
Explain requires how and why. The organising idea that turns this from two lists into one argument is that every decision runs from external to internal: the coach begins as the learner's source of structure and error detection, and the job is to hand both over.
Plan: name the direction; practice for the cognitive learner with reasons; practice for the autonomous learner with reasons; feedback for each; then the dependency risk that explains why the fading matters.
A cognitive learner should practise in a distributed arrangement, because the movement consumes all of their attention and a mentally fatigued beginner practises errors, which are then strengthened rather than corrected. Complex skills should be broken into parts to stay within their attention capacity, and practice should be blocked, because there is as yet no movement plan for random practice to make them retrieve. An autonomous learner can tolerate massed practice, since the movement costs little attention, should practise whole and in context, and should practise randomly at full game complexity.
A cognitive learner needs mostly augmented feedback, because their own task-intrinsic feedback is not yet interpretable to them; it should be immediate, so the attempt is still available, and should be knowledge of performance on one element in simple terms, since knowledge of results is usually already obvious. An autonomous performer relies mainly on task-intrinsic feedback, which they interpret more sensitively than an observer can; augmented feedback should be sparing and delayed, and should address only what is genuinely not self-detectable, together with tactical content.
Practice and feedback should differ between these two learners in a single consistent direction: from external to internal. In the cognitive stage the coach supplies the structure and the error detection that the learner cannot yet supply themselves, and the whole purpose of the process is to hand both over by the autonomous stage. For a cognitive learner, practice should be distributed rather than massed, because the movement consumes all of their available attention and they fatigue mentally as well as physically; a fatigued beginner practises errors, and since every repetition strengthens the pattern performed, practice makes permanent rather than perfect. A complex skill should be broken into parts, so that each part falls within their attention capacity, with the joins between the parts practised specifically once the elements are secure. Practice should be blocked rather than random, because random practice produces its benefit by forcing the learner to retrieve and reconstruct the movement plan on each attempt, and a cognitive learner has no plan yet to retrieve; randomising too early simply prevents a pattern forming at all. Their feedback should be predominantly augmented, since their own task-intrinsic feedback is not yet interpretable to them, delivered immediately while the attempt is still available to them, and should be knowledge of performance directed at a single element in simple terms, because knowledge of results is usually already apparent from the outcome itself. For an autonomous learner every one of these reverses. Massed practice is workable because the movement costs little attention; practice should be whole and in context, and random and varied at full game complexity, because reconstructing the movement plan on each attempt is both what produces retention and what a competition actually demands, since no match presents the same situation thirty times consecutively. Feedback should be mainly task-intrinsic, because the performer detects errors within the movement more sensitively than an observer can judge them from outside, with augmented feedback given sparingly, delayed rather than immediate, and confined to what is genuinely not self-detectable along with tactical content. The reason the reduction matters, rather than merely being possible, is that frequent and immediate augmented feedback creates dependency: the learner stops attending to their own task-intrinsic information because a more accurate source is always available, so their error detection never develops, and in competition, where augmented feedback is unavailable, they have nothing to draw on. The practical habit that prevents this is to ask the learner what they thought before offering an opinion, which forces them to consult their own feedback first.
A coach's blocked sessions produce visibly better performance than a colleague's random sessions, yet the colleague's players perform better in matches. Justify a recommendation to the first coach about how to arrange practice and feedback.
Justify means support a recommendation with evidence and reasoning. So the answer must actually make a recommendation, not merely explain a phenomenon, and it must show why the first coach's own evidence — which is real — does not support his practice.
Plan: concede the observation is genuine; separate performance from learning; explain the retrieval mechanism; give the recommendation in specific terms; qualify it for beginners; add the feedback half; close on what evidence he should be collecting instead.
Performance is how well a skill is executed at a given moment and is temporary and directly observable. Learning is a relatively permanent change in capability, and it can only be detected by a retention test after a delay or a transfer test in a new context. A coach standing at a session can only observe performance, and blocked and random practice are the clearest case in which the two move in opposite directions.
In blocked practice the learner adjusts a movement plan they have already retrieved, so each repetition is easy and performance rises. In random practice the intervening attempts displace it, so the plan must be retrieved and reconstructed on every repetition; that reconstruction is effortful, depresses immediate performance, and is what produces retention — and it matches competition, which never presents the same situation repeatedly. The recommendation is therefore to retain blocked practice only while a skill is being acquired by a cognitive learner, and to move to random and varied practice as soon as a basic movement pattern exists, accepting a visible fall in session quality as the expected cost.
The first coach's observation is genuine and should be conceded before anything else: blocked practice does produce visibly better performance within a session. The error is not in what he has seen but in what he has concluded from it, because performance and learning are different things and only one of them is observable at a session. Performance is how well a skill is executed at a given moment; it is temporary and directly visible. Learning is a relatively permanent change in capability, and it can only be detected by a retention test conducted after a delay, or by a transfer test in a new context. Blocked and random practice are the clearest case in which the two diverge, and his colleague's match results are precisely the transfer test that his own sessions cannot provide. The mechanism explains why. In blocked practice the learner repeats a movement plan they have already retrieved and need only adjust, so each attempt is a small correction of the last, the repetitions are easy, and performance rises quickly and smoothly. In random practice the intervening attempts at other skills displace that plan, so on every repetition the learner must retrieve and reconstruct it from scratch. That reconstruction is effortful and makes each individual attempt worse, and it is also the process that produces durable learning — and it corresponds to what competition actually demands, since no match presents the same skill thirty times consecutively. The recommendation is therefore as follows. Blocked practice should be retained only while a skill is genuinely being acquired by a learner in the cognitive stage, because such a learner has no movement plan for random practice to make them retrieve, and randomising at that point prevents a pattern forming at all. As soon as a basic pattern exists, the arrangement should move to random and varied practice, and the coach should expect and accept a visible fall in the quality of his sessions, since that fall is the cost of the mechanism rather than evidence that the change is failing. Without that expectation stated in advance, he will abandon the change within a fortnight for exactly the reason he adopted blocked practice in the first place. The same reasoning applies to his feedback. Frequent, immediate, detailed augmented feedback likewise improves performance within a session and impairs learning, because the learner stops attending to their own task-intrinsic feedback when a more accurate external source is continuously available, and so never develops the error detection that competition requires — where augmented feedback is unavailable by definition. He should therefore fade his feedback deliberately: reduce its frequency, delay it, and above all ask the player what they thought before offering his own assessment, which forces them to consult their own information first. Finally, he should change the evidence he collects. If he judges his coaching by what he sees at a session he will continue to select the arrangement that flatters it, so he should introduce a simple retention measure — the same test repeated a week later, without warning — and judge the two arrangements on that. That single change makes the rest of the recommendation self-verifying rather than a matter of taking someone's word for it.
Multiple choice · 5 questions
MC 1. Practising several different skills in an unpredictable order is called:
- A. blocked practice
- B. random practice
- C. massed practice
- D. part practice
MC 2. Random practice produces better retention than blocked practice principally because:
- A. it allows more total repetitions in the same time
- B. it is more enjoyable, so learners practise for longer
- C. the learner must retrieve and reconstruct the movement plan on each attempt
- D. it reduces the physical fatigue caused by repeating one movement
MC 3. The information a gymnast gains from the feel of the landing through their own body is:
- A. augmented feedback
- B. knowledge of results only
- C. concurrent augmented feedback
- D. task-intrinsic feedback
MC 4. A coach who gives detailed, immediate feedback after every single attempt for months risks:
- A. the learner becoming dependent on it and failing to develop their own error detection
- B. the learner acquiring the skill faster than is safe
- C. converting a closed skill into an open one
- D. reducing the learner's task-intrinsic feedback to zero permanently
MC 5. Which best distinguishes strategy from tactics?
- A. Strategy applies to teams and tactics to individuals
- B. Strategy is decided before the contest, tactics are adjustments made during it
- C. Strategy concerns defence and tactics concern attack
- D. Strategy is taught by the coach and tactics are invented by the players
Short answer · exam style
SA 1. Remember Band 3 Distinguish between massed and distributed practice, and between whole and part practice. (4 marks)
SA 2. Understand Band 5 Explain the six types of feedback, and how the feedback given to a learner should change as they progress through the stages of learning. (5 marks)
SA 3. Evaluate Band 6 A coach reports that blocked practice works better because his players perform better in those sessions. Evaluate this claim. (6 marks)
📖 Comprehensive answers (click to reveal)
Multiple choice: 1 → B. 2 → C. 3 → D. 4 → A. 5 → B.
Activity 1 guidance: The 24 hour delay in step 4 is the entire experiment and should not be negotiable; a retention test run at the end of the same session measures performance, which is exactly the error the activity exists to expose. Expect the blocked group to score clearly higher during practice — that part is reliable — and expect the retention result to be noisier than the textbook version, because a class of twenty produces groups of ten and one outlier moves a mean substantially. That noise is worth using rather than apologising for: it is the Lesson 15 point about reporting the spread, and a group where eight of ten random-practice students improved and two collapsed is a more interesting finding than a mean. If the result comes out the other way, the useful response is to examine the design rather than to conclude the effect is not real: the commonest causes are groups that were not comparable to begin with, a randomisation that was not genuinely unpredictable to the thrower, and a retention test in which the blocked group was tested in blocked order, which hands them back the advantage the test was supposed to remove. Students should also notice that this activity is itself random practice for the random group and blocked practice for the other, so the groups have had unequal exposure to the testing format — which is a real confounder and is worth naming.
Activity 2 guidance: Feedback 1 is augmented, immediate and knowledge of results, and it is useless to this learner because an experienced player already possesses the knowledge of results — they saw it miss. Feedback that only repeats what the learner already knows is not feedback; the rewrite supplies knowledge of performance, or better, asks what they thought first. Feedback 2 is augmented, concurrent and mixed, and its fault is volume: four cues exceed a cognitive learner's attention capacity, so most will be lost and the learner is additionally told that everything they did was wrong. The rewrite selects one cue, gives it before the attempt rather than during it, and holds the rest for later sessions. Feedback 3 is the subtle one and the point of the activity: each correction is accurate, useful and well-meant, and the pattern is the problem. Over eight weeks the swimmer has been given no reason to attend to what the stroke felt like, because a more accurate account is guaranteed at the wall, so their error detection — the capacity that defines the associative-to-autonomous transition — has not developed, and in a race, where no coach is available, they have nothing. The alternative is not to withhold help but to fade it: ask what they noticed before responding, give feedback after every third length rather than every length, and progressively delay it. Feedback 4 is augmented, delayed and neither knowledge of results nor knowledge of performance — it is a judgement of character, it identifies nothing that could be changed, and three days later it cannot be attached to any specific moment. The rewrite names an observable behaviour in a specific passage of play and what should have happened instead.
SA 1 (4 marks): Massed — practice with little or no rest between repetitions or sessions, suiting short discrete skills and motivated learners at later stages [1]. Distributed — practice with substantial rest, suiting physically demanding or dangerous skills and cognitive-stage learners who fatigue mentally [1]. Whole — practising the skill in its entirety, appropriate for continuous skills and for skills whose parts are too interdependent to separate [1]. Part — practising components separately before recombining, appropriate for serial skills with separable elements and for skills that exceed a beginner's attention capacity, provided the joins are then practised [1]. A definition without any indication of when it is used earns half the pair at best.
SA 2 (5 marks): Task-intrinsic and augmented correctly distinguished by source, with the observation that task-intrinsic is available on every repetition, including in competition, where augmented feedback is often limited to breaks, timeouts or between attempts [1]. Concurrent and delayed correctly distinguished by timing, with a reason why delayed feedback supports learning — it leaves room for self-assessment [1]. Knowledge of results and knowledge of performance correctly distinguished by content, with the point that knowledge of results is often already available to the learner and knowledge of performance is usually the more useful [1]. The progression across stages described on all three axes — augmented to task-intrinsic, immediate to delayed, frequent to sparing [1]. The reason the change is necessary rather than merely possible: continuous augmented feedback creates dependency and prevents error detection developing, leaving the athlete without resources in competition [1]. An answer that lists six definitions with no progression caps at 3.
SA 3 (6 marks): Concedes what is true — blocked practice genuinely does produce better performance within a session, so the observation is not in dispute [1]. Defines performance as temporary, immediate and observable [1]. Defines learning as a relatively permanent change in capability, detectable only by a retention test after a delay or a transfer test in a new context, so a coach at a session can observe only performance [1]. Supplies the mechanism — blocked practice allows the learner to adjust a plan already retrieved, while random practice forces retrieval and reconstruction on each attempt, which depresses immediate performance and produces retention, and which also matches competition, where a situation is never repeated identically [1]. Identifies when blocked practice is genuinely correct — for a cognitive-stage learner with no movement plan yet to retrieve, since randomising too early prevents a pattern forming at all [1]. States what evidence would settle it — a retention or transfer test rather than in-session observation, since judging coaching by what is visible at a session will always select the arrangement that flatters the session [1]. An answer that simply asserts the coach is wrong, without conceding the observation or supplying the mechanism, caps at 3.
Performance is not learning, and a coach can only watch one of them. Performance is how well it goes right now — temporary, visible. Learning is a lasting change in capability, and it shows up only in a retention test after a delay or a transfer test in a new context. Blocked and random practice is the case where the two run in opposite directions: blocked looks better all session and loses two weeks later, because random practice forces the learner to retrieve and reconstruct the movement plan on every attempt, and that reconstruction is the learning. A session that looks good is not evidence that it worked.
Two of the three practice decisions have clear rules. Distributed for beginners and for demanding or dangerous skills, because a fatigued learner practises errors — and practice makes permanent, not perfect. Part for serial skills with separable elements, whole for continuous skills and for anything whose parts are too interdependent to separate — and part practice is not finished until the joins are practised. Blocked to acquire, random to retain, moving on as soon as a basic pattern exists.
The performance elements have to be trained on purpose. Decision-making cannot be developed by a drill with one correct answer, so random and varied practice is not just better for retention — it is the only arrangement that trains the decision at all. Strategy is decided before the contest; tactics are the adjustments made during it.
Six feedback types, three questions. Where from — task-intrinsic or augmented. When — concurrent or delayed. About what — knowledge of results or knowledge of performance. Every piece of feedback has one answer from each. And the same reversal applies: constant, immediate, detailed augmented feedback improves the session and creates dependency, so the learner never builds their own error detection — and in competition, task-intrinsic feedback is the only kind there is. Fade it deliberately, and ask what they thought before you tell them.
One direction runs through all of it: external to internal. The coach begins as the learner's structure, error detection and motivation, and the job is to hand all three over. That sentence answers most questions on this dot point.
Lesson 19 closes this inquiry question by making you do the research: how movement skills are acquired, developed and improved in a sport of your own choice — what the research says, how it is applied in practice, and what further questions it opens.
Mark lesson as complete
Tick when you've finished the practice and review.