Corrected November 9, 2015—see last page.
Journal of Motor Learning and Development, 2013, 1, 89-95
© 2013 Human Kinetics, Inc.
www.JMLD-Journal.com
ORIGINAL RESEARCH
Overarm Throwing Variability as a Function
of Trunk Action
M. A. Urbin, David Stodden, and Glenn Fleisig
Individual body segment actions evolve during throwing skill development. Maximal trunk involvement is
typically the last feature of the movement pattern to fully develop. The current study examined developmental
levels of trunk action and the associated variability in the throwing motion. The throwing motions of children
and adolescents were analyzed via motion capture and trunk actions were classified as exhibiting no rotation
(n = 7), blocked rotation (n = 6), or differentiated rotation (n = 11). Results indicated nonrotators exhibited
greater variability than blocked-rotators in maximum humeral external rotation and humeral horizontal adduction angles at ball release; nonrotators also demonstrated greater variability than differentiated-rotators on these
parameters, in addition to forward trunk tilt and elbow extension angle at ball release. Nonrotators produced
more variable peak upper torso and humeral horizontal adduction angular velocities, as well as peak upper
torso linear velocity, relative to differentiated-rotators. Blocked-rotators produced more variable peak pelvis,
upper torso, and humeral horizontal adduction angular velocities, as well peak pelvis linear velocity, relative
to differentiated-rotators. Nonrotators were less consistent relative to blocked- and differentiated-rotators in
the time that elapsed from peak pelvis angular velocity to ball release. These results indicate that greater trunk
involvement is associated with more consistent movement production.
Keywords: coordination, motion capture, motor development, motor control
The overarm throw is a kinetic chain of events that
complies with the summation of speed principle (Bunn,
1972; Putnam, 1991). This principle holds that each body
segment makes a contribution to the throwing motion that
is not independent of other segments (Neal, Snyder, &
Kroonenberg, 1991). In essence, angular momentum can
be transferred through the skeletal linkage to the projectile
when the timing of body segment rotations is effective.
Though kinetic chain principles can be exploited when
only two body segments are involved (Alexander, 1991;
Chowdhary & Challis, 2001), a thrower’s ability to
exploit these principles with all of the appropriate body
segments is intimately tied to motor skill development
(Alexander, 1991; Langendorfer & Roberton, 2002;
Southard, 2002, 2009; Stodden, Langendorfer, Fleisig,
& Andrews, 2006a,b).
Researchers have examined trends in throwing skill
development using the component approach (Roberton
& Halverson, 1984), which categorizes the action of individual segments in the overall spatiotemporal patterning
of the throw. For example, the action of the trunk can be
categorized as having no rotation or anterior-posterior
Urbin is with the Dept. of Kinesiology, Auburn University,
Auburn, AL. Stodden is with the Dept. of Physical Education
and Athletic Training, University of South Carolina, Columbia,
SC. Fleisig is with the American Sports Medicine Institute,
Birmingham, AL.
sway, indicating a less developmentally advanced action.
More developmentally advanced trunk actions can be
characterized as having blocked or differentiated rotation
of the pelvis and upper torso. Accordingly, these categorizations reflect a continuum of kinetic chain exploitation, providing insight into throwing skill development.
Given the various combinations of segmental actions
that can potentially appear during skill development, the
component approach captures variations that may not be
accounted for using a whole-body approach (Roberton,
1978). This approach has been used in analyses of both
children and adult throwers (Roberton, Halverson, Langendorfer, & Williams, 1979; Williams, Haywood, & van
Sant, 1998) and accounts for 85% of the variance in ball
speed (Roberton & Konczak, 2001). Kinematic analyses
have indicated that individual body segment actions
classified with the component approach discriminate
quantitative differences in the associated action (Stodden et al., 2006a, 2006b), suggesting that the component
approach is a valid method for evaluating throwing skill
development.
Longitudinal inquiring into throwing skill development using the component approach indicates there is
considerable interindividual variation in the acquisition
of a developmentally advanced movement pattern (Langendorfer & Roberton, 2002). Stated another way, there
is no one common pathway observed in the progression
of skill development. Nevertheless, the most developmentally advanced component trunk action tends to be
89
90 Urbin, Stodden, and Fleisig
the last feature to appear and is thought to constrain the
component action of more distal body segments. The
kinetic chain is an open system of skeletally linked body
segments with the trunk serving as the fixed base about
which the more distal segments rotate (Southard, 2009).
Therefore, from a theoretical standpoint, the action of the
trunk should also have some impact on spatiotemporal
fluctuations occurring late in the throwing motion.
Variability in the context of overarm throwing skill
development is not well understood. Previous research
has examined kinematic variability between baseball
pitchers of different competition levels and found that
youth pitchers tend to exhibit the greatest variability
(Fleisig, Chu, Weber, & Andrews, 2009). A similar
investigation of throwing techniques used by handball
players with different training experience demonstrated
that the magnitude of variability associated with the jump
throw and standing throw without a run-up was similar
among skill groups (Wagner, Pfusterschmied, Klous,
Serge, & Müller, 2012). However, the low-skilled group
was found to exhibit more variability during the standing
throw with a run-up.
These studies provide meaningful insight into variability and its association with competition level and
training experience, but they do not directly explain
variability as it relates to differences in coordination that
emerge as a function of skill development. The inclusion
of individuals competing at different levels and with different training experience is generally a reflection of skill,
but throwing skill development is not directly captured by
these criteria. This view is supported by the trunk action
exhibited by youth pitchers in the Fleisig et al. (2009)
study. Though youth pitchers may be considered lesser
skilled relative to higher competition levels (e.g., high
school, collegiate, professional), the sample of youth
league pitchers exhibited the most developmentally
advanced trunk action (i.e., differentiated rotation). This
observation underscores the arbitrary nature of competition level and experience as indexes of throwing skill
development. A deeper understanding of skill-related
trends in performance variability may be ascertained by
establishing a link between variability in the throwing
motion and valid measures of skill development. Understanding whether intertrial variation is influenced by the
amount of trunk involvement is theoretically appealing
for two reasons. First, a more skillful throwing pattern is
characterized by more trunk involvement, whereas, a less
skillful pattern is characterized by less trunk involvement.
Second, the action of the trunk is thought to constrain the
component action of more distal body segments.
A sound theoretical framework in which to interpret
trends in variability is as necessary as establishing such
trends. For many years, developmentalists have argued
that variability around a transition is a prerequisite for
acquiring more evolved motor repertoires (Thelen &
Corbetta, 1994). According to this view, a fully developed
throwing motion is acquired through repeated production
of different movement characteristics and experimenting
with different musculoskeletal organizations (Goldfield,
Kay, & Warren, 1993). However, in the context of a highspeed, whole-body movement such as maximal-effort
overarm throwing, exploration may be limited to early
movement characteristics involving the lower-extremities
and trunk. Though the sequence of body segment rotations is not as sequentially distinct in a less developed
throwing motion, the action of the trunk precedes the
respective actions of more distal body segments. Since
the trunk action can affect the inertial constraints of distal
segments, exploration of different lower-extremity and
trunk configurations also may influence the variability
of these segmental trajectories. To address these possibilities, the current investigation analyzed the throwing
motions of children and adolescents exhibiting different component trunk actions. It was hypothesized that
throwers with greater trunk involvement would exhibit
significantly less variability in positional, peak velocity,
and temporal kinematic parameters relative to throwers
with less trunk involvement.
Methods
Kinematic data were collected from the throwing
motions of 24 children and adolescents in an indoor
laboratory at the American Sports Medicine Institute,
in Birmingham, Alabama. Each participant and his/her
legal guardian provided informed assent and consent
before participation. Next, 1.0-cm diameter reflective
markers were placed bilaterally on the distal end of
the third metatarsal, lateral malleolus, lateral femoral
epicondyle, greater trochanter of the femur, lateral tip
of the acromion, lateral humeral epicondyle, and ulnar
styloid (Figure 1). A reflective marker was also placed
on the radial styloid of the throwing hand. Participants
were then guided through a dynamic stretch routine of the
upper and lower extremities followed by a self-selected
number of warm-up throws. Participants were instructed
to throw “as hard as you can” for each throw to a framed
net (2 m length × 2 m width) approximately 18 m away.
Reducing the distance between the net and some of the
children participants was necessary to provide a more
developmentally appropriate task constraint.
Twelve adolescents with pitching experience (M
age =13.3 ± 1.0 years; all male) threw baseballs (mass =
150 g, diameter = 6.3 cm) from a mound. The remaining
participants (M age =7.2 ± 2.4 years; 8 female, 5 male)
lacked experience throwing a baseball from a mound.
Therefore, these participants threw tennis balls (mass = 63
g, diameter = 6.0 cm) from flat ground. Previous research
has indicated there are minimal differences in kinematics
characteristics of the throwing motion when throwing
balls of different masses (Fleisig et al., 2006) and throwing from a mound or flat ground (Fleisig, Bolt, Fortenbaugh, Wilk, & Andrews, 2011). In addition, according to
motor learning principles, task novelty is associated with
increased performance variability (Fitts & Posner, 1967).
Therefore, it was necessary to scale task constraints to
ensure that ball mass and throwing surface were typical
for adolescents and developmentally appropriate for
Overarm Throwing Variability 91
Figure 1 — Reflective marker set-up for motion capture.
children. These conditions were held constant across all
of the throws analyzed for each participant.
Six synchronized cameras (Motion Analysis, Corp
Santa Rosa, CA) tracked the reflections of markers at
a frequency of 240 Hz. The orientation of each camera
was optimized to capture the entire throwing motion.
Raw coordinate data were digitally filtered with a
13.4-Hz Butterworth low-pass filter. Positional and
peak velocity kinematic parameters were calculated
based on previously described procedures (Escamilla,
Fleisig, Barrentine, Zheng, & Andrews, 1998). Ball
speed was recorded from a radar gun (JUGS Inc. Tulatin,
OH). These parameters were chosen because they are
performance-related features of the throwing motion
(Barrentine, Fleisig, Whiteside, Escamilla, & Andrews,
1998; Fleisig et al., 2009; Stodden, Fleisig, McLean, &
Andrews, 2005; Stodden, Fleisig, McLean, Lyman, &
Andrews, 2001). Three of the four positional kinematic
parameters were referenced at the time of ball release:
forward trunk tilt relative to the horizontal plane, elbow
extension angle, and humeral horizontal adduction angle.
The remaining positional parameter was the maximum
angle of humeral external rotation. Ball speed and eight
velocity parameters were examined, including peak
angular and linear velocities of the pelvis and upper
torso, as well as peak angular velocities of humeral
horizontal adduction, humeral internal rotation, and
elbow extension associated with the throwing arm. In
addition, the time between peak pelvis angular velocity
and ball release was calculated.
The trunk action for each throw was analyzed qualitatively via sagittal- and rear-view digital video recordings (60 Hz) of the throwing motion. Trunk action was
scored by two independent raters with the developmental
sequence for the component trunk action of the overarm throw (Roberton & Halverson, 1984). Level 1 was
recorded if no trunk twist preceded the arm movement,
or if there was flexion/extension of the trunk (i.e., nonrotator). Level 2 was recorded if the pelvis and upper
torso simultaneously rotated in the throwing direction
as one unit (i.e., blocked-rotator). Level 3 was recorded
if the pelvis rotated in the throwing direction while the
upper torso simultaneously rotated away from the throwing direction (i.e., differentiated-rotator). Interrater and
intrarater objectivity was established using the proportion of agreement adjusted for chance (Safrit & Wood,
1995). The kappa coefficient for interrater and intrarater
objectivity was k = 0.92 and k = 0.96, respectively. Participants were grouped according to trunk action for data
analysis: nonrotators (n = 7, M age = 6.4 ± 2.1 years; 7
females; height = 120.8 ± 15.3 cm; mass = 26.2 ± 7.6
kg); blocked-rotators (n = 6, M age = 8.1 ± 2.6 years; 5
males, 1 female; height = 131.5 ± 12.1 cm; mass = 28.1
± 5.7 kg); differentiated-rotators (n = 11, M age = 13.3
± 1.0 years; 11 males; height = 169.1 ± 10.2 cm; mass =
54.9 ± 9.1 kg). Qualitative analyses indicated that trunk
developmental level remained constant across each of
the five throws for all participants.
For each participant, the standard deviation of each
parameter was calculated to provide an index of the
overall magnitude of variability and entered into analysis.
The use of five trials is consistent with previous research
examining variability between throwers on similar
kinematic parameters (Fleisig et al., 2009). Multivariate
analysis of variance (MANOVA) was used to evaluate
variability between groups on kinematic positional and
velocity parameters. Correlation matrices of parameters
contained in each MANOVA indicated correlations were
between .2 and .7. Thus, the assumption of multicollinearity was not violated. Pillai’s trace was the test statistic
used because it is the most conservative test, and it is
robust when sample sizes are small (Tabachnick & Fidell,
2001). Bonferroni (two-tail) post hoc tests were used to
correct for multiple comparisons. Analysis of variance
(ANOVA) was used to examine variability in the overall
time between stride-foot contact and ball release. Significance for all analyses was set at the .05 level.
Results
The MANOVA for kinematic positional parameters
was significant [F(8, 38) = 4.030, p = .002, η2 = .459].
Descriptive statistics for positional parameters are contained in Table 1. All four parameters contributed significantly to the model: forward trunk tilt angle at ball release
92 Urbin, Stodden, and Fleisig
[F(2, 21) = 4.472, p = .024, η2 = .299]; elbow angle at ball
release [F(2, 21) = 11.718, p < .001, η2 = .527]; humeral
horizontal adduction at ball release [F(2, 21) = 33.264, p
<. 001, η2 = .760]; and maximum humeral external rotation [F(2, 21) = 21.668, p <. 001, η2 = .674]. Bonferroni
post hoc test indicated: nonrotators were more variable
than differentiated-rotators in forward trunk tilt angle at
ball release (p = .022), elbow angle at ball release (p <
.001), humeral horizontal adduction angle at ball release
(p < .001), and maximum humeral external rotation (p <
.001). In addition, nonrotators were more variable than
blocked-rotators in humeral horizontal adduction at ball
release (p < .001) and maximum humeral external rotation (p = .001).
The MANOVA for kinematic peak velocity parameters was significant [F(16, 30) = 8.269, p < .001, η2
= .815]. Descriptive statistics for peak velocity parameters are contained in Table 2. Post hoc tests indicated
that the following parameters contributed significantly
to the model: peak pelvis angular velocity [F(2, 21) =
10.514, p = .001, η2 = .500]; peak pelvis linear velocity
[F(2, 21) = 3.820, p = .038, η2 = .267]; peak upper torso
angular velocity [F(2, 21) = 16.044, p < .001, η2 = .604];
peak upper torso linear velocity [F(2, 21) = 4.835, p =
.019, η2 = .315]; and peak humeral horizontal adduction
angular velocity [F(2, 21) = 11.582, p < .001, η2 = .524].
Bonferroni post hoc test indicated: blocked-rotators were
more variable than differentiated-rotators (p = .001) in
peak pelvis angular velocity; differentiated-rotators were
less variable than nonrotators (p < .001) and blockedrotators (p = .022) in peak upper torso angular velocity;
differentiated-rotators were less variable than nonrotators
(p = .001) and blocked-rotators (p = .015) in peak humeral
horizontal adduction angular velocity; blocked-rotators
were more variable than differentiated-rotators (p = .042)
in peak pelvis linear velocity; nonrotators were more
variable than differentiated-rotators (p = .016) in peak
upper torso linear velocity.
The ANOVA for the overall time between peak pelvis
angular velocity and ball release was significant [F(2, 19)
= 11.814, p < .001, η2 = .554]. Descriptive statistics for
this interval are contained in Table 3. Tukey post hoc tests
indicated nonrotators were more variable than blockedrotators (p = .008) and differentiated-rotators (p < .001).
Table 1 Mean and Mean Standard Deviation for Positional Parameters by Trunk Action
No rotation
Blocked rotation
Differentiated rotation
Φ
Forward trunk tilt (°)
17.4 ± 6.4*
5.0 ± 4.4
32.9 ± 2.5
.70
Elbow extension angle (°)
91.7 ± 20.7*
84.5 ± 11.9
37.3 ± 3.0
.99
Horizontal add (°)
35.4 ± 15.6*†
32.1 ± 4.7
10.2 ± 1.3
1.0
External rot (°)
142.6 ± 14.2*†
148.4 ± 5.2
179.9 ± 2.4
1.0
Positional
* Significant difference relative to differentiated rotation.
† Significant difference relative to blocked rotation.
Table 2 Mean and Mean Standard Deviation for Peak Velocity Kinematic Parameters by Trunk Action
No rotation
Blocked rotation
Differentiated rotation
Φ
Pelvis (m/s)
0.5 ± 0.1
1.1 ± 0.2*
2.0 ± 0.1
.63
Upper torso (m/s)
0.9 ± 0.3*
1.2 ± 0.2
3.2 ± 0.1
.74
Pelvis (°/s)
202.5 ± 48.1
473.7 ± 70.0*
606.5 ± 26.0
.98
Upper torso (°/s)
391.0 ± 93.5*
780.5 ± 64.6*
1037.6 ± 28.2
1.0
Humeral hor. adduction (°/s)
383.9 ± 129.7*
532.7 ± 105.6*
551.3 ± 41.5
.99
Elbow extension (°/s)
1281.9 ± 204.9
1573.3 ± 131.3
2224.8 ± 119.6
.16
Humeral internal rotation (°/s)
2293.7 ± 484.3
3580.5 ± 360.0
6928.8 ± 517.8
.11
8.2 ± 0.9
13.1 ± 1.0
28.8 ± 0.5
.37
Peak velocity
Ball (m/s)
* Significant difference relative to differentiated rotation.
† Significant difference relative to blocked rotation.
Overarm Throwing Variability 93
Table 3 Mean and Mean Standard Deviation for the Time Elapsing Between Peak Pelvis Angular
Velocity and Ball Release by Trunk Action
No rotation
Blocked rotation
Differentiated rotation
111 ± 57*†
100 ± 19
94 ± 8
Temporal
Peak pelvis angular velocity—ball release (msec)
* Significant difference relative to differentiated rotation.
† Significant difference relative to blocked rotation.
Discussion
The purpose of this study was to examine the magnitude
of variability in various kinematic features of the throwing
motion as a function of trunk involvement. The findings
indicate that greater trunk involvement is associated with
more consistent movement production, supporting the
overall hypothesis. For clarity, the differences observed
between groups in the throwing parameters analyzed
will be discussed in the order they appear during the
throwing motion.
Pelvis and upper torso linear velocities are the initial
source of momentum in the throwing motion that can
augment the angular velocities of the pelvis and upper
torso if the timing of each rotation is effective. Results
of the current study indicate that differentiated-rotators
were more consistent in producing these peak velocities.
Nonrotators exhibited greater variability in peak linear
and angular velocities of the upper torso. Blocked-rotators
were more variable in generating peak linear and angular
velocities of the pelvis, as well as upper torso angular
velocity. The coexistence of increased variability in the
linear and angular velocities of the same trunk segment in
each respective form of trunk rotation may be accounted
for by the physics associated with the throwing motion
and theory surrounding motor skill development.
Differentiation of the trunk action is due to the
initial rotation of the pelvis and the combined mass
of the upper torso and throwing arm resisting angular
motion in throwing direction. The forceful rotation of
the pelvis in the throwing direction as the upper torso
rotates away causes the trunk musculature to be eccentrically loaded. This dynamic loading promotes storage
and recovery of elastic energy as the upper torso rotates
in the throwing direction (Stodden et al., 2006a). The
results of the current study suggest this form of trunk
rotation is associated with more consistent movement
production. A thrower’s ability to exhibit differentiated trunk rotation is a developmental phenomenon,
and instability in coordination signals a transition into
a more evolved movement pattern (Goldfield et al.,
1993; Thelen & Corbetta, 1994). Previous research
has demonstrated that performing throws at different
percentages of maximum effort influences the stability
of throwing coordination (Southard, 2002; 2009) and
ball speed (Urbin, Stodden, Boros, & Shannon, 2012).
Accordingly, the intent to throw at a specific percentage
of maximum may act as a control parameter that induces
change in the movement pattern. Participants in the
current study were instructed to throw at maximum
effort. The increased variability in each primitive form
of trunk rotation, therefore, may reflect an exploration
of different strategies to increase trunk involvement and,
in turn, maximize ball speed. If this view is accurate,
the results of the current study indicate there may be
an ordered, serial transition from no rotation to blocked
rotation followed by blocked rotation to differentiated
rotation. In essence, the transition from no rotation to
blocked rotation may be characterized by increased
variability in linear and angular velocities of the upper
torso. Similarly, the transition from blocked rotation to
differentiated rotation may be characterized by increased
variability in the linear and angular velocities of the
pelvis. Differentiated trunk rotation is achieved when
the performer acquires the ability to forcefully rotate
the pelvis allowing the upper torso to lag behind. The
variability blocked-rotators exhibit in peak pelvis linear
and angular velocities also may explain the variability
that results in peak upper torso angular velocity later
in the kinetic chain as upper torso movement becomes
more dependent on pelvis movement.
Just as the angular acceleration of the pelvis causes
the upper torso to lag behind when trunk rotation is
differentiated, lag is also observed in the action of the
humerus relative to the upper torso (Hong, Cheung, &
Roberts, 2001). As such, kinematic characteristics of the
upper extremity segments are influenced by the angular
velocity of the upper torso and the motion-dependent
inertia of upper extremity mass (Stodden et al., 2006b).
Differentiated-rotators were significantly more consistent
in producing peak humeral horizontal adduction angular
velocity relative to both nonrotators and blocked-rotators.
However, nonrotators were more variable in the humeral
horizontal adduction angle at the time of ball release
relative to both blocked- and differentiated-rotators. The
fact that blocked-rotators were not more variable than
differentiated-rotators in this regard, despite exhibiting
more variability in the peak velocity of this joint action,
is a curious finding. A similar finding was observed in
elbow extension angle at ball release. Though nonrotators
exhibited greater variability in this angle at ball release,
there were no group differences in the variability of peak
elbow extension angular velocity. Collectively, these
findings suggest that variations in rotational velocities of
94 Urbin, Stodden, and Fleisig
the trunk induced variability in the velocity of this joint
action without affecting variations in its position at the
time of ball release.
Differentiated-rotators were more consistent in
achieving maximum humeral external rotation compared
with both nonrotators and blocked-rotators. Relatively
greater kinetic energy is transferred to the upper extremity when the trunk action is maximized. Greater energy
transfer contributes to the extreme range of motion at the
glenohumeral joint in elite level throwers (≥ 180°; Werner,
Fleisig, Dillman, & Andrews, 1993) and is produced via
high inertial loading characteristics of the forearm when
the elbow is flexed at approximately 90° (Escamilla &
Andrews, 2009). Thus, the passive mechanisms stemming
from more trunk involvement may lead to more consistent
maximum humeral external rotation.
Nonrotators were more variable in forward trunk
tilt angle at the time of ball release. Increased ball speed
is associated with greater trunk tilt in elite-level throwers (Stodden et al., 2005). Though it could be argued
that negative acceleration of the pelvis upon stride-foot
contact causes the trunk to flex passively, there is likely
some active shortening of the trunk musculature leading
to more tilt. Minimal data are available on trunk muscle
activity patterns during the throwing motion. However,
the existing data indicates that the rectus abdominis musculature reaches its peak activity just before ball release
(Hirashima, Kadota, Sakurai, Kudo, & Ohtsuki, 2002). It
is, therefore, possible that the variability of forward trunk
tilt at the time of ball release in the differentiated-rotators
is not exclusively associated with the passive mechanisms
attributed to the previously discussed parameters.
Finally, nonrotators were significantly more variable
than both blocked- and differentiated-rotators in the overall time from peak pelvis angular velocity to ball release.
This finding suggests that when the trunk is more involved
in the throwing motion, more energy is transferred up the
kinetic chain, resulting in greater temporally stability
of upper torso and distal joint rotations. These findings
appear compatible with previous research that indicates
less-developed throwers exhibit greater variability in the
timing of distal-joint lag (Southard, 2002; 2009).
A limitation of the current study was the lack of
sample homogeneity in terms of age and gender. It is
important to note, however, that there is no evidence to
suggest that body segment mass (i.e., due to age or gender)
is associated with overarm throwing variability. Moreover,
motor development is age-related not age-determined
(Clark, 1994). Though there are differences in the rate
of throwing skill development (Halverson, Roberton, &
Langendorfer, 1982; Malina, Bouchard, & Bar-Or, 2004),
the movement pattern that characterizes a fully developed
throwing motion has not been shown to be different
between males and females. Nevertheless, the respective associations between the variability of the throwing
motion and developmental phenomena (i.e., trunk action),
throwing experience, maturation, and/or gender cannot
be definitively addressed by these data. Recruiting a sufficient number of participants that are matched on such
characteristics and exhibit one of the three developmental
levels of trunk action is a logistical challenge that future
research can improve upon in light of the current findings. An additional limitation relates to differences in task
constraints (i.e., ball mass/circumference and throwing
surface) between groups. Though these task constraints
were scaled to be developmentally appropriate, it is possible that the cumulative effect of these differences may
have impacted performance variability between groups.
In conclusion, the findings of the current study
indicate that maximal trunk involvement is associated
with more consistent movement production. Trunk
involvement occurs on a continuum that may not be fully
captured by the three ordinal classifications used in the
current study. However, these classifications represent
the general developmental progression of a thrower’s
ability to dynamically integrate the trunk in the throwing motion. The enhanced consistency associated with
a more developmentally advanced trunk action appears
to be linked to the passive transfer of momentum stemming from greater trunk rotation. From a performance
perspective, focusing on the development of trunk action
may be advantageous for promoting the integration and
consistency of distal body segment actions that occur late
in the kinetic chain (Stodden, 2006).
References
Alexander, R.M. (1991). Optimum timing of muscle activation
for simple models of throwing. Journal of Theoretical Biology, 150, 349–372. doi:10.1016/S0022-5193(05)80434-5
Barrentine, S.W., Fleisig, G.S., Whiteside, J.A., Escamilla, R.F.,
& Andrews, J.R. (1998). Biomechanics of windmill softball pitching with implications about injury mechanisms
at the shoulder and elbow. The Journal of Orthopaedic
and Sports Physical Therapy, 28, 405–414. doi:10.2519/
jospt.1998.28.6.405
Bunn, J.W. (1972). Scientific principles of coaching (2nd ed.).
Englewood Cliffs, NJ: Prentice-Hall.
Chowdhary, A.G., & Challis, J.H. (2001). The biomechanics of
an overarm throwing task: A simulation model examination
of optimal timing of muscle activations. Journal of Theoretical Biology, 211, 39–53. doi:10.1006/jtbi.2001.2332
Clark, J.E. (1994). Motor development. In V.S. Ramachandran (Ed.), Encyclopedia of human behavior (Vol. 3, pp.
245–255). NY: Academic Press.
Escamilla, R.F., & Andrews, J.R. (2009). Shoulder muscle
recruitment patterns and related biomechanics during
upper extremity sports. Sports Medicine (Auckland, N.Z.),
39, 569–590. doi:10.2165/00007256-200939070-00004
Escamilla, R.F., Fleisig, G.S., Barrentine, S.W., Zheng, N., &
Andrews, J.R. (1998). Kinematic comparisons of throwing different types of baseball pitches. Journal of Applied
Biomechanics, 14, 1–23.
Fitts, P.M., & Posner, M.I. (1967). Human performance. Belmont, CA: Brooks/Cole.
Fleisig, G.S., Bolt, B., Fortenbaugh, D., Wilk, K.E., & Andrews,
J.R. (2011). Biomechanical Comparison of Baseball Pitching and Long-Toss: Implications for Training and Rehabilitation. The Journal of Orthopaedic and Sports Physical
Therapy, 41, 296–303. doi:10.2519/jospt.2011.3568
Overarm Throwing Variability 95
Fleisig, G.S., Chu, Y., Weber, A., & Andrews, J.R. (2009). Variability in baseball pitching biomechanics among various
levels of competition. Sports Biomechanics, 8, 10–21.
doi:10.1080/14763140802629958
Fleisig, G.S., Phillips, R., Shatley, A., Loftice, J., Dun, S., Drake,
S., . . . Andrews, J.R. (2006). Kinematics and kinetics of
youth baseball pitching with standard and lightweight balls.
Sports Engineering, 9, 155–163. doi:10.1007/BF02844117
Goldfield, E.C., Kay, B.A., & Warren, W.H. (1993). Infant
bouncing: The assembly and tuning of action systems.
Child Development, 64, 1128–1142. doi:10.2307/1131330
Halverson, L.E., Roberton, M.A., & Langendorfer, S.L. (1982).
Development of the overarm throw: Movement and ball
velocity changes by seventh grade. Research Quarterly
for Exercise and Sport, 53, 198–205. doi:10.1080/02701
367.1982.10609340
Hirashima, M., Kadota, H.K., Sakurai, S., Kudo, K., & Ohtsuki,
T. (2002). Sequential muscle activity and its functional
role in the upper extremity and trunk during overarm
throwing. Journal of Sports Sciences, 20, 301–310.
doi:10.1080/026404102753576071
Hong, D., Cheung, T.K., & Roberts, E.M. (2001). A threedimensional, six-segment chain analysis of forceful
overarm throwing. Journal of Electromyography and Kinesiology, 11, 95–112. doi:10.1016/S1050-6411(00)00045-6
Langendorfer, S.J., & Roberton, M.A. (2002). Individual pathways in the development of forceful throwing. Research
Quarterly for Exercise and Sport, 73, 245–256. doi:10.1
080/02701367.2002.10609018
Malina, R.M., Bouchard, C., & Bar-Or, O. (2004). Growth,
maturation and physical activity (2nd ed.). Champaign,
IL: Human Kinetics.
Neal, R.J., Snyder, C.W., & Kroonenberg, P.M. (1991).
Individual differences and segment interactions in
throwing. Human Movement Science, 10, 653–676.
doi:10.1016/0167-9457(91)90021-O
Putnam, C.A. (1991). A segment interaction analysis of
proximal-to-distal sequential segment motion patterns.
Medicine and Science in Sports and Exercise, 23, 130–144.
doi:10.1249/00005768-199101000-00019
Roberton, M.A. (1978). Stability of stage categorizations in
motor development. In D. Landers & R. Crisitina (Eds.),
Psychology of motor behavior and sport (pp. 494–505).
Champaign, IL: Human Kinetics.
Roberton, M.A., & Halverson, L.E. (1984). Developing children—their changing movement: A guide for teachers.
Philadelphia, PA: Lea & Febiger.
Roberton, M.A., Halverson, L.E., Langendorfer, S., & Williams,
K. (1979). Longitudinal changes in children’s overarm
throw ball velocities. Research Quarterly for Exercise and
Sport, 50, 256–264.
Roberton, M.A., & Konczak, J. (2001). Predicting children’s
overarm throw ball velocities from their developmental
levels in throwing. Research Quarterly for Exercise and
Sport, 72, 91–103. doi:10.1080/02701367.2001.10608939
Safrit, M.J., & Wood, T.M. (1995). Introduction to measurement
in physical education and exercise science (3rd ed.). St.
Louis, MO: Mosby-Year Book.
Southard, D. (2002). Change in throwing pattern: Critical values
for control parameter of velocity. Research Quarterly for
Exercise and Sport, 73, 396–407. doi:10.1080/02701367
.2002.10609039
Southard, D. (2009). Throwing pattern: Changes in timing of
joint lag according to age between and within skill level.
Research Quarterly for Exercise and Sport, 80, 213–222.
doi:10.1080/02701367.2009.10599555
Stodden, D.F. (2006). Facilitating the acquisition of complex
ballistic motor skills: Promoting proximal or distal system
perturbations? Journal of Human Movement Studies, 51,
197–220.
Stodden, D.F., Fleisig, G.S., McLean, S.P., & Andrews, J.R.
(2005). Relations of biomechanical factors to baseball
pitching velocity: Within pitcher variation. Journal of
Applied Biomechanics, 21, 44–56.
Stodden, D.F., Fleisig, G.S., McLean, S.P., Lyman, S.L., &
Andrews, J.R. (2001). Relationship of trunk kinematics
to pitched ball velocity. Journal of Applied Biomechanics,
17, 164–172.
Stodden, D.F., Langendorfer, S.J., Fleisig, G.S., & Andrews, J.R.
(2006a). Kinematic constraints associated with the acquisition of overarm throwing Part I: Step and trunk actions.
Research Quarterly for Exercise and Sport, 77, 417–427.
Stodden, D.F., Langendorfer, S.J., Fleisig, G.S., & Andrews, J.R.
(2006b). Kinematic constraints associated with the acquisition of overarm throwing Part II: Upper extremity actions.
Research Quarterly for Exercise and Sport, 77, 428–436.
Tabachnick, B.G., & Fidell, L.S. (2001). Using multivariate
statistics (4th ed.). New York, NY: Harper & Row.
Thelen, E., & Corbetta, D. (1994). Exploration and selection in
the early acquisition of skill. International Review of Neurobiology, 37, 75–102. doi:10.1016/S0074-7742(08)60240-8
Urbin, M.A., Stodden, D.F., Boros, R.L., & Shannon, D.M.
(2012). Examining impulse-variability in overarm throwing. Motor Control, 16, 19–30.
Wagner, H., Pfusterschmied, P., Klous, M., Serge, P., & Müller,
E. (2012). Movement variability and skill level of various throwing techniques. Human Movement Science, 31,
78–90. doi:10.1016/j.humov.2011.05.005
Werner, S.L., Fleisig, G.S., Dillman, C.J., & Andrews, J.R.
(1993). Biomechanics of the elbow during baseball pitching. The Journal of Orthopaedic and Sports Physical
Therapy, 17, 274–278. doi:10.2519/jospt.1993.17.6.274
Williams, K., Haywood, K., & van Sant, A. (1998). Changes
in throwing by older adults: A longitudinal investigation.
Research Quarterly for Exercise and Sport, 69, 1–10. doi
:10.1080/02701367.1998.10607661
Erratum: Urbin, Stodden, & Fleisig (2013)
In Table 2, nine values were incorrectly listed. The first two rows (Pelvis, Upper torso) and the last row (Ball)
listed incorrect values in three columns (No rotation, Blocked rotation, Differentiated rotation). The correct values
have now been placed. In Table 3, the unit was incorrectly listed as m/s and should be msec. Additionally, the
three values contained erroneous zeros. All items have been corrected.