BJMB
Brazilian Journal of Motor Behavior
Special issue:
The role of practice in motor learning
Anderson,
Steel
2022
VOL.16
N.2
179 of 193
It's not the type of practice that matters, it's the attitude: The impact of playful practice on
motor skill learning
DAVID I. ANDERSON
1
| KYLIE A. STEEL
2
1
Marian Wright Edelman Institute, San Francisco State University, San Francisco, California, USA.
2
School of Science and Health, MARCS Institute, Western Sydney University, Sydney, New South Wales, Australia.
Correspondence to: David I. Anderson. Director, Marian Wright Edelman Institute
San Francisco State University,
1600 Holloway Ave.
San Francisco, CA 94132
email: [email protected]sfsu.edu
https://doi.org/10.20338/bjmb.v16i2.278
ABBREVIATIONS
KR Knowledge of Results
PUBLICATION DATA
Received 13 12 2021
Accepted 28 02 2021
Published 01 06 2026
ABSTRACT
The traditional approach to practice has focused on the physical structure of practice, manipulating parameters
like duration, spacing, task variations, practice order, and whether tasks are practiced in parts or as a whole and
physically or mentally. The emergence of the deliberate practice framework shifted the focus to the learner’s
attitude or mindset toward practice. It argues that the most effective practice involves a consciously effortful,
workman-like approach to remedying weaknesses in performance. In the current paper, we build on the notion
of deliberate play that arose in response to the deliberate practice framework. Rather than view deliberate play
as a forerunner to deliberate practice, we argue that a playful approach to practice can benefit the learning
process at any stage of learning or skill development. We draw on contemporary research in motor learning and
development, in developmental and evolutionary psychology, and in education to highlight the benefits of a
playful approach to practice on motor skill learning. We end with practical suggestions for encouraging a playful
approach to practice and learning.
KEYWORDS: Deliberate play | Deliberate practice | Goalless learning | Mindfulness | Play
INTRODUCTION
What is the most effective way to organize practice to facilitate motor skill
learning? This question has motivated and puzzled researchers and practitioners at least
since the origins of the modern field of motor behavior. Theories imported from
experimental and developmental psychology, in combination with the practical needs of
learners and society, have influenced approaches to answer this question and the volume
of research devoted to it. For example, the publication of Hull’s (1943) theoretical treatise
Principles of Behavior
1
spurred considerable interest in the topic of distribution of practice
according to Adams (1987), during what Adams referred to as the Middle Period of
research on motor learning
2
. Similarly, the publication of Schmidt’s (1975) schema theory
stimulated substantial interest in the virtues of constant versus variable practice, ultimately
leading to a major proliferation of research on the scheduling of variable practice relative to
the contextual interference continuum
3
. On the other hand, the exigencies of World War II,
which were significantly different from previous conflicts, spurred research on whole versus
part-whole practice given the military’s need to train personnel quickly to acquire a range
of complex skills
2
.
These examples practice distribution, constant versus variable practice, the
BJMB
Brazilian Journal of Motor Behavior
Anderson,
Steel
2022
VOL.16
N.2
180 of 193
Special issue:
The role of practice in motor learning
scheduling of variable practice, and part-whole versus whole practice relate to the
physical structure of practice. Further, these practice variations can be controlled easily by
the experimenter or practitioner by manipulating parameters like time, spacing, number of
practice variations, the interleaving of practice trials, and the components of tasks. The
ease with which these parameters can be varied likely accounts for the large volume of
research these practice variations have attracted, particularly during the latter half of the
last century. Notably, that volume of research has decreased considerably since the
Millennial (see Magill & Anderson, 2021)
4
.
Coincident with the decline in interest in how the physical structure of practice
influences learning was a surge of research inspired by Anders Ericsson’s expert
performance approach and its introduction of the concept of deliberate practice
5
. The
questions raised by the expert performance approach were particularly interesting to
scholars and practitioners interested in the identification and development of talent in sport
6
. The deliberate practice idea shifted focus away from the optimal structural parameters of
practice toward the attitude or mindset the learner had to practice and the nature of the
activities the learner engaged in during each practice session. Although the definition of
deliberate practice has shifted since the term was introduced
7
, deliberate practice has
some defining characteristics. For example, deliberate practice is designed specifically to
improve the current level of performance and is often directed to overcoming weaknesses.
Practice is adapted as skill improves and it requires focused attention, immediate
feedback, repetition, and high levels of effort. Deliberate practice is neither inherently
enjoyable nor intrinsically motivating and is often frustrating. In essence, it can be
characterized as a workman-like approach to practice
5
. The dour characterization of
practice offered by the expert performance approach caused many sport scientists to
question whether deliberate practice was the primary way in which learners developed
expertise in sport.
Following interviews with athletes, parents, and coaches, which indicated that
many athlete’s first exposure to sport was fun and playful, Jean Côté introduced the term
deliberate play
8
as an alternative or complement to deliberate practice. Deliberate play
involves activities that are intrinsically motivating, provide immediate gratification, and are
designed to maximize enjoyment. However, unlike deliberate practice, which is typically
structured and monitored by a teacher, deliberate play is informally regulated by modified
rules that are established, enforced and further modified by the participants themselves.
Deliberate play became a central feature of Côté and colleagues’ Developmental Model of
Sport Participation, which argues, among other things, that early exposure to deliberate
play prior to later engagement in deliberate practice has a beneficial effect on talent
development in sport
9
. We take this idea a step further in the current paper.
We argue that exposure to playful practice, or more specifically a playful attitude
toward practice, has intrinsic value regardless of the learner’s stage of learning or skill
development. From a broad perspective, the various approaches to practice can be
organized along a continuum from pure play to deliberate practice. However, there is no
obvious reason why more playful approaches to practice need to precede more deliberate
approaches to practice, even if this seems to be the case when skill development is
viewed ontogenetically and the challenge in the practice context needs to be adapted as
skill increases. In the following sections, we briefly outline the key features of play and then
discuss how playful practice might enhance motor learning. We end by describing how
BJMB
Brazilian Journal of Motor Behavior
Anderson,
Steel
2022
VOL.16
N.2
181 of 193
Special issue:
The role of practice in motor learning
practitioners might encourage playful practice. Before continuing, we must confess that the
ideas expressed in this paper are still embryonic; they require further development.
However, we hope this paper will serve as a starting point for debate and discussion on
the potential role of playful practice in motor skill learning and development.
CHARACTERISTICS OF PLAY
Play is difficult to define, though scholars often view play as the opposite of work
10
. Extensive research on play in anthropology, psychology and education also indicates
that it is an important contributor to learning and socialization throughout life
10
, potentially
via a “playful” learning mechanism
11,12
and evolutionary processes
13
. Play, which is
prevalent in many species, follows an inverted-U development, peaking in the juvenile
period when the pressure to perform up to particular standards is low and waning
thereafter
13
. All forms of play tend to follow exploration, ontogenetically. For example,
infants spend more time exploring than playing and toddler play only begins to displace
exploration when the environment is familiar, though infants, toddlers and children explore
new objects before playing with them
13
. The structure of play is highly variable and
appears to lack immediate purpose, yet play is intrinsically motivating and involves
experimentation with objects, the environment, one’s own body and motor patterns, and/or
with other organisms. The animal’s preoccupation during play appears to be driven more
by means rather than ends (goals); further, in play, the combinations of motor patterns is
greater than in almost any other form of behavior
12
.
Scholars consider play adaptive because it increases behavioral variability, which
in turn leads to the discovery of novel or innovative action patterns or combinations of
behaviors that can be propagated to other individuals via observational learning
1113
. Play
appears to stir the imagination and encourage a sense of adventurousness. One can
observe novel behaviors first appearing in play in later goal-directed activity. According to
Bruner
11
, play provides an excellent opportunity to try combinations of behavior that an
animal would never, under functional pressure, attempt. This pressure free context
maximizes variability because it minimizes the consequences of one’s actions, and
therefore risk, and encourages an extension of behavioral limits. Wide variation in
behavioral patterns then provides a base upon which selection can operate. Bruner
14
argues that for behavior to be highly flexible play must precede it. The benefits of play are
generally realized over the long term, thus if the interest is in short term gains in
performance under narrowly constrained contexts, the value of play is less obvious
10
.
HOW PLAYFUL PRACTICE MIGHT ENHANCE LEARNING
Given the aforementioned characteristics of play, we now speculate on the
potential ways in which playful practice might enhance learning. We build our case by
interpreting findings from the motor learning and educational psychology literatures relative
to the characteristics of play and, by extension, playful practice.
Enhanced Exploration
As noted earlier, exploration and play are closely connected during ontogenetic
development. Exploration features centrally in contemporary theorizing about motor
BJMB
Brazilian Journal of Motor Behavior
Anderson,
Steel
2022
VOL.16
N.2
182 of 193
Special issue:
The role of practice in motor learning
learning and development inspired by the ecological approach to perception and action
and the dynamical systems perspective. Gibson made a distinction between movements
that are performatory (executive) or exploratory (investigative)
15
. Performatory movements
accomplish some behavior in the usual meaning of the term whereas exploratory
movements serve to reveal and pick up information. The distinction between these two
types of movements led Gibson
16
to claim “We must perceive in order to move, but we
must also move in order to perceive” (p.223); a claim which highlights the value of
exploratory movements prior to and during goal-directed behavior in detecting and utilizing
important sources of information for the control and development of action. Bernstein’s
description of learning and practice as a search for optimal solutions to motor problems
further reified exploration as a major contributor to behavioral change
17,18
.
Exploratory behavior subsequently became recognized as a critical agent in motor
skill development, particularly during infancy
19,20
, and motor skill learning
2123
. Perhaps
the most systematically developed description of the role of exploration in behavioral
change is Newell and colleagues’ characterization of learning and practice as a search for
task solutions through the perceptual-motor workspace
22,23
. Drawing on the work of
Newell, Gibson, Bernstein, and others, Hacques and colleagues have argued that skill
acquisition should focus on how performers can develop exploratory behavior rather than
learn a specific movement
21
. They argue practice should lead performers to develop
exploratory activity that reveals more reliable information and that it should be undertaken
in safe environments where learners can explore even when they are close to or beyond
their maximal action boundaries.
Given what we know about play, it seems ideally suited to developing the types of
exploratory activity and competence advocated by skill acquisition researchers. Even
though exploration typically precedes play, it does not disappear once play emerges.
Rather, exploration and experimentation remain highly visible components of play
11
. The
safety afforded by play would seem to encourage the type of limit-testing exploration
advocated by Hacques and colleagues
21
. Further, because play is directed toward objects,
the environment, the body, and others, it would seem ideally suited to uncovering critical
perceptual invariants that characterize the important relations among the task,
environment, and learner necessary for learning.
Mindfulness
Play and mindfulness share several characteristics. Mindfulness is a state that can
be deployed toward the performance or learning of any activity, yet is rarely explored in the
motor learning context. Like play, mindfulness is a difficult concept to define, however, like
play, mindfulness also has characteristics that facilitate its recognition. Langer
24
refers to it
as “a flexible state of mind in which we are actively engaged in the present, noticing new
things and sensitive to context” (p. 220). Being in the present moment, being aware, and
attending to experiences in a nonjudgmental, nonreactive, and accepting way are the most
common characteristics scholars assign to mindfulness
25
. Central to this paper’s thesis,
Langer
26
notes that whereas work is often accomplished mindlessly, play is almost always
mindful. In other words, playful practice might be the most reliable way to encourage the
characteristics of mindfulness that are presumed to enhance performance and learning.
The benefits of mindfulness have been studied extensively in educational settings
27
and more recently in elite sport performance
25
. Although scholars have conducted far
less research on the effects of mindfulness on motor skill learning, some notable studies
BJMB
Brazilian Journal of Motor Behavior
Anderson,
Steel
2022
VOL.16
N.2
183 of 193
Special issue:
The role of practice in motor learning
suggest a mindful attitude can enhance learning. For example, Kee and Liu
28
showed that
learners with a stronger disposition toward mindfulness tended to perform a rollerball task
more skillfully. Moreover, mindful individuals used more adaptive learning strategies and
reported higher enjoyment during practice. In addition, Zhang et al.,
29
showed that
learners who were assigned to mindfulness training while they practiced throwing darts for
eight weeks showed significantly greater improvement in performance than learners
assigned to an attention control group. The performance of the mindfulness group was
also significantly higher than that of the control group at post intervention and on a follow
up test. Further, the mindfulness group, but not the control group, reported significant
improvements in mindfulness, experiential acceptance, and flow at post intervention and
follow up.
Langer
24,27
describes several studies in which learners were encouraged to adopt
a mindful approach to practice. To foster mindfulness, learners were instructed to be as
creative and playful as possible and to vary their approach to practicing the tasks as much
as possible. In one skill acquisition study, participants were taught a new sport called
“Smack-it-ball.” The sport is like squash except that players wear a small racket on each
hand like a baseball glove. Half of the participants were instructed in how to use the racket
using absolute language and the other half received conditional language. Both groups
were shown and told how to hold their hands, however the conditional language group
were also told that the demonstrated method was only one possible way to hold their
hands. After considerable practice, the researchers surreptitiously changed the weight of
the ball. The participants who had learned the game “mindfully” adapted to the changed
ball much better than those who had received more traditional instructions. Interestingly,
the effect was much greater for females, who appeared to be trapped by their original
learning, than for the males.
In summary, a small but growing number of studies suggests a more mindful
approach to practice can benefit motor skill learning. Importantly, play, or a playful attitude,
appears to encourage aspects of mindfulness that researchers believe are key contributors
to an effective learning environment.
De-emphasis on explicit instructions
Learners are rarely exposed to explicit instructions during play. Langer’s smack-it-
ball study highlights the negative effects explicit instructions can have on motor learning. A
number of other studies show that explicit instructions not only degrade the ability to adapt
to novel task variations, or changes in context, but also mar the rate of improvement during
practice and the level of performance in retention
3032
. In one of the clearest
demonstrations of these negative effects, Wulf and Weigelt
32
showed that learners given
explicit instructions about when to apply force to the platform of a ski simulator performed
much more poorly during practice compared to participants who received no instructions
about how to move. The differences between groups were particularly pronounced on a
test in which the participants were stressed by being told an expert ski instructor would
evaluate them. Green and Flowers
30
used a very different manipulation to highlight the
negative consequences of explicit instructions. They instructed some of their participants
to look for specific deviations in the trajectory of a falling object as they learned a
computer-based catching task. The participants could use the specific deviations to predict
the landing location of the object. Rather than enhance performance and learning, the
BJMB
Brazilian Journal of Motor Behavior
Anderson,
Steel
2022
VOL.16
N.2
184 of 193
Special issue:
The role of practice in motor learning
explicit instructions had the opposite effect, with instructed participants performing
substantially worse than non-instructed participants.
Hodges and Lee
31
note several potential explanations for the detrimental effects
of explicit instructions on performance and learning. First, explicit instructions may lead to
a narrow focus of attention on specific aspects of the task rather than the whole task.
Relatedly, explicit instructions may discourage exploration, thus limiting exposure to the
task dynamics and important sources of intrinsic information available for the regulation of
performance. Explicit information may also impede learning about task dynamics and
critical sources of intrinsic information by increasing the information processing load and
the demands on attention. Finally, explicit instructions may force the learner to concentrate
only on finding the correct solution to the motor problem, leading to a stereotyped
repetition of the process used to solve the problem and increasing the chances the learner
will get stuck in an attractor.
De-emphasis on goals
The aforementioned discussion of the potential negative side effects associated
with explicit instructions leads naturally into a discussion about the potential negative side
effects of focusing on goal attainment given the focus on means rather than ends during
play. A surprisingly large body of literature has focused on the negative effects of providing
learners with specific goals, although most of that literature is in educational and
experimental psychology. Nevertheless, one can interpret some classic phenomena in
motor learning relative to the potential drawbacks of providing specific goals. Most obvious
is the implicit learning phenomenon
33
, which is also closely related to incidental learning
34
.
Both phenomena highlight that a large degree of what we learn and remember is learned
without conscious awareness or as a side effect of pursuing explicit goals. The negative
effects of frequent Knowledge of Results (KR) on motor learning is another example of the
potential downside of focusing too intently on goal accomplishment, if we acknowledge
that KR provides a constant reminder of the learning goal. Researchers have attributed the
negative effects of frequent KR to the learner’s failure to process the critical sources of
intrinsic feedback needed to sustain performance in the absence of KR
35
. Thus, the
learning processes subverted by frequent KR are similar to those subverted when
instructors provide learners with explicit instructions about how to move.
The most extensive evidence that a focus on specific goals can hurt learning is in
the educational psychology literature. This is particularly true in the literature on
mathematical learning, where scholars generally classify mathematics-related problems as
transformational problems, with an initial state, a goal state, and legal problem-solving
operators. Sweller
36
has argued convincingly that some forms of problem solving
encouraged by specific goals interfere with learning because they interfere with
development of the schemas presumed to support problem solving
36
. He questions the
common assumption that practice on a large number of conventional problems is the best
way to acquire schemas and develop problem-solving skill. Sweller
36
bases his
conclusions on numerous experiments in which he gave participants a variety of problems
that they could solve either by means-ends analysis or by inducing a rule based on the
problem structure. One can characterize means-ends analysis as trial and error learning,
in which the learner attempts to solve the problem by generating multiple solutions until
they find a successful solution. Sweller
36
found that while participants had little trouble
BJMB
Brazilian Journal of Motor Behavior
Anderson,
Steel
2022
VOL.16
N.2
185 of 193
Special issue:
The role of practice in motor learning
solving the problems using a trial and error approach, they usually did not induce the
relevant rules. Essentially, conventional, goal-directed search heuristics like means-ends
analysis typically prevented problem solvers from learning essential aspects of the
problem’s structure
36
.
In contrast to the negative effects of goal-directed means-ends analysis on
learning, Sweller
36
found nonspecific goals led to rapid learning of essential characteristics
of the problem structure. He attributes the detrimental effects of specific goals and the
beneficial effects of nonspecific goals to the related mechanisms of selective attention and
limited information processing capacity. First, he reasons that solving a problem and
acquiring schemas require largely unrelated cognitive processes. That is, to solve a
problem by means-ends analysis, a problem solver must attend to differences between the
current problem state and the goal state but can ignore previously used problem-solving
operators and relations between problem states, except to prevent retracing steps during
solution. In contrast, schema acquisition requires recognizing a problem state as belonging
to a particular category of problems that require particular moves. Thus, attention to
previously visited problem states and moves associated with those states is important for
schema acquisition. Second, means-ends problem solving strategies can impose a high
cognitive load during complex problem solving because the learner must keep so many
pieces of information in mind, including the goal state, the current problem state, the
relationship between the goal and current state, and relations between problem-solving
operators. Consequently, this high cognitive load may leave limited processing capacity
available for schema acquisition even if the learner solves the problem.
A series of experiments by Sweller and Levine
37
highlight the logic behind
Sweller’s ideas. The authors gave participants learning maze puzzles to solve and tracked
the speed at which the solution was achieved and the participants’ understanding of the
rules leading to the problem solution. They gave one group of participants an explicit goal
and another group of participants a non-specific goal. For example, in one experiment,
participants had to trace their way out of a maze while blindfolded. In the specific goal
group, they placed one of the participant’s hands on the start location and the other hand
on the goal location, while participants in the non-specific goal group had one hand placed
on the start location and were simply told them to find the end location. Intriguingly,
participants in the non-specific goal group solved the problem in fewer moves and were
much more likely to report the rule underlying the solution to the problem, which actually
required moving away from the goal location before moving toward it. Participants in the
specific goal group frequently followed paths toward the goal location, i.e., paths that
minimized the distance between the current location and the goal location and
encountered dead ends. One can imagine how difficult it would be to avoid such a strategy
as a member of the specific goal group; that is, proximity to the goal must have been
extremely alluring.
In summary, a large body of literature supports the counterintuitive assertion that
providing more information about the problem structure by introducing specific goals can
subvert learning about the remainder of the problem space. Learners’ preoccupation with
accomplishing the goal appears to interfere with important processes underlying the
development of advanced problem solving capacities. Sweller has described the potential
downsides of specific goals in a particularly lucid and convincing way. Interestingly, one
cannot help but notice how similar his findings are to those of Tolman
38
many decades
BJMB
Brazilian Journal of Motor Behavior
Anderson,
Steel
2022
VOL.16
N.2
186 of 193
Special issue:
The role of practice in motor learning
earlier. Tolman coined the term latent learning to refer to the learning acquired by rats
allowed to roam freely within mazes without specific goals to pursue. Relative to the focus
of this paper, it is interesting to note that Bruner
14
has referred to play as latent learning.
The detrimental effects of specific goals on attentional processes is also curiously similar
to the inattentional blindness phenomenon reported in the literature, wherein the tracking
of an explicit goal interferes with the detection of salient information in the environment
39
.
Repetition without repetition
The previous discussion has interesting implications for Bernstein’s
17
idea of
practice and learning as a form of repetition without repetition. Bernstein’s ideas about
practice popularized the notion that learning was a form of problem solving characterized
by a search through the problem space. Bernstein stated, “The process of practice toward
the achievement of new motor habits essentially consists in the gradual success of a
search for optimal motor solutions to the appropriate problems” (p. 362)
17
. With respect to
the idea of repetition without repetition, he explained, “Repetitions of a movement or action
are necessary in order to solve a motor problem many times (better and better) and to find
the best ways of solving it” (p. 176)
17
. According to Bernstein, what learners repeat during
practice is not specific solutions to the motor problem but the process of solving the
problem again and again and discovering better and better solutions, hence inspiring the
contemporary description of skill acquisition as exploration, discovery and selection.
The discussion so far has implications for Bernstein’s ideas for a least two
reasons. First, the idea of practice and learning as a search for solutions to task problems
conveys the notion that learning is about acquiring something already in existence. But as
Pachecho et al.
23
, have noted, skill acquisition is less an “acquisitionof something that
can be selected and more a “transformation” of the learner’s ability to solve the problem.
Secondly, Sweller’s
36
analysis of the downsides of specific goals reveals that learning and
problem solving are not necessarily the same thing. The learner might solve the problem
successfully but learn nothing useful about the problem structure in the process. This is far
more likely to occur if the learner uses a trial and error approach to generating the “correct”
solution to the problem and therefore does not learn the rules by which they can generate
solutions. A deeper understanding of the problem space and the rules underlying it can
lead to the generation of appropriate solutions not only to the current problem but also to
problems with a similar structure, thus providing a basis for transfer of learning (see for
example Harlow’s 1949 notion of learning to learn
40
). As Langer
27
notes, from a mindful
perspective, a person’s response to a particular problem is not an attempt to make the
best choice from among existing options, that is to take an available heuristic that seems
to work best, but to create options. An understanding of rules, especially rules linking
causes and effects, seems far more likely to promote the capacity to create options, to shift
the learner from a trial and error learner to one who adapts to problems via hypothesis and
insight
40
.
We do not intend to imply Bernstein’s ideas about practice and learning are
misguided. Quite the opposite. We want to highlight the need to evaluate Bernstein’s ideas
relative to the broader literature on learning and problem solving. One encounters the
phrase repetition without repetition frequently in the motor learning literature, however its
meaning seems to vary from author to author. From our perspective, and consistent with
the broader literature on learning and problem solving, learning can be viewed as a search