The influence of upper-body mechanics, anthropometry and isokinetic strength on performance in wrist-spin cricket bowling

Abstract

Delivering a cricket ball with a wrist-spin (WS) bowling technique is considered one of the game’s most difficult skills. Limited biomechanical information exists for WS bowlers across skill levels. The purpose of this study was to compare biomechanical, isokinetic strength and anthropometric measures between elite (12) and pathway bowlers (eight). Data were collected using a motion analysis system, dynamometer and a level-two anthropometrist. A regression analysis identified that performance was best explained by increased wrist radial deviation torque and longitudinal axis rotational moments at the shoulder and wrist. From back foot impact (BFI) to ball release (BR), elite bowlers rotated their trunks less, experienced less trunk deceleration resulting in a more front-on position and increased pelvis rotation angular velocity. They also displayed an increased shoulder internal rotation moment as the upper arm moved from external into internal rotation and was a major contributor in the subsequent differences observed in the distal segments of the bowling limb. Anthropometric differences were observed at the wrist and finger joints and may be used to form the basis for talent identification programmes. This study highlights the important contribution to bowling performance of the musculature responsible for producing long axis rotations of the bowling limb.

Full text

The inuence of upper-body mechanics, anthropometry and isokinetic strength on
performance in wrist-spin cricket bowling
Wayne Spratford
a,b
, Bruce Elliott
c
, Marc Portus
d
, Nicholas Brown
e
and Jacqueline Alderson
c,f
a
University of Canberra Research Institute for Sport and Exercise (UCRISE), University of Canberra, ACT, Canberra, Australia��;
b
Discipline of Sport and
Exercise Science, Faculty of Health, University of Canberra, ACT, Canberra, Australia;
c
School of Sport Science, Exercise and Health, The University of
Western Australia, Perth, Australia;
d
Praxis Performance Group, Canberra, Australia�;
e
Australian Institute of Sport, University of Canberra Research
Institute for Sport and Exercise (UCRISE) Movement Science, Canberra, Australia;
f
Sports Performance Research Institute, Auckland University of
Technology, Auckland, New Zealand
ABSTRACT
Delivering a cricket ball with a wrist-spin (WS) bowling technique is considered one of the game’s most
dicult skills. Limited biomechanical information exists for WS bowlers across skill levels. The purpose of
this study was to compare biomechanical, isokinetic strength and anthropometric measures between
elite (12) and pathway bowlers (eight). Data were collected using a motion analysis system, dynamometer
and a level-two anthropometrist. A regression analysis identied that performance was best explained by
increased wrist radial deviation torque and longitudinal axis rotational moments at the shoulder and
wrist. From back foot impact (BFI) to ball release (BR), elite bowlers rotated their trunks less, experienced
less trunk deceleration resulting in a more front-on position and increased pelvis rotation angular
velocity. They also displayed an increased shoulder internal rotation moment as the upper arm moved
from external into internal rotation and was a major contributor in the subsequent dierences observed
in the distal segments of the bowling limb. Anthropometric dierences were observed at the wrist and
nger joints and may be used to form the basis for talent identication�programmes�. This study highlights
the important contribution to bowling performance of the musculature responsible for producing long
axis rotations of the bowling limb.
ARTICLE HISTORY
Accepted 16 August 2019
KEYWORDS
Spin bowling; revolutions;
strength; wrist spin;
biomechanics
Introduction
Delivering a cricket ball with a wrist-spin (WS) bowling techni-
que has long been considered one of the most dicult arts to
master in the game of cricket (Bradman, 1969; Philpott, 1995;
Tyson, 1994; Wilkins, 1991; Woolmer, Noakes, & Moett, 2008).
A bowler must release the ball out of the ulnar or the fth
phalangeal side of the hand while coordinating internal rota-
tion of the humerus, pronation of the radioulnar joint, and
extension and radial deviation of the wrist (Philpott, 1995;
Spratford, Portus, Wixted, Leadbetter, & James, 2014;
Woolmer et al., 2008) but without the control of the second
phalangeal which remains in ball contact at BR in nger-spin
(FS) (Beach, Ferdinands, & Sinclair, 2016; Spratford & Davison,
2010). An example of the release prole is shown in Figure 1.
This type of technique enables the bowler to deliver the ball
with considerably more revolutions in comparison with the
more common FS bowler (Beach, Ferdinands, & Sinclair, 2014;
Beach et al., 2016; Cork, Justham, & West, 2012; Spratford et al.,
2017). Greater revolutions also allow the bowler to take advan-
tage of the increase in the Magnus force arising from ball spin
and the subsequent increases in lateral and vertical deviation
during ight (Robinson & Robinson, 2013), and the potential
increase of lateral deviation (side-spin) after the ball bounces
(Beach et al., 2014; Woolmer et al., 2008). The anti-clockwise
horizontal axis revolutions placed on the ball by the right
handed WS bowler causes the ball to “drift into” the right
handed batsman during ight after reaching its zenith height
(Justham, Cork, & West, 2010; Robinson & Robinson, 2013) and
deviate away to the o-side of the batsman, in the direction of
the revolutions after bounce (Beach et al., 2014; Bradman, 1969;
Philpott, 1995; Wilkins, 1991; Woolmer et al., 2008). A ball
deviating, or in this case “spinning”, away from a batsman has
been shown to be a more dicult task for a batsman to
perceive and then intercept and subsequently control their
shot (Diaz, Cooper, Rothkopf, & Hayhoe, 2013; Sarpeshkar,
Mann, Spratford, & Abernethy, 2017; Welchman, Tuck, &
Harris, 2004). For this reason, coaching based literature has
stated that if a bowler can master this technique and bowl
with control, they have long been considered a match winner
(Woolmer et al., 2008).
To date very little research has focussed on WS bowlers, with
a single study focusing on the kinematic dierences between
FS and WS bowlers, showing large dierences in techniques
throughout the delivery cycle as well as in ball kinematics
(Beach et al., 2016). Researchers have also identied that ball
kinematic dierences exist between elite level WS bowlers and
lower level pathway bowlers. With ball velocity, revolutions and
a velocity/revolution index all capable of distinguishing perfor-
mance between groups and therefore used as outcome mea-
sures (Spratford et al., 2017). The majority of literature that
explores WS bowling exists within the subjectively based
CONTACT Wayne Spratford Wayne.spratford@canberra.edu.au University of Canberra Research Institute for Sport and Exercise (UCRISE), University of
Canberra, ACT, Canberra, 2601 Australia
JOURNAL OF SPORTS SCIENCES
https://doi.org/10.1080/02640414.2019.1696265
© 2019 Informa UK Limited, trading as Taylor & Francis Group

coaching literature (Bradman, 1969; Philpott, 1995; Woolmer
et al., 2008). This has focussed on the need for the body (trunk)
to be rotating forward at ball release (BR) and for the “wrist to
be turned on release”, which is a combination of radioulnar
pronation, radial deviation and wrist extension. This late move-
ment at the wrist also forms the basis for the bowling action’s
naming convention; however, there is limited evidence as to
what role, if any, these biomechanical movements play in
bowling performance. Without this knowledge, it is dicult to
understand the dierences that may occur between bowlers as
they progress through the development pathway. It is assumed
that segmental rotations of the arm around the long axis of the
bowling limb will be important within the kinetic chain, similar
to that observed in racquet based sports such as tennis and
squash (Marshall & Elliott, 2000; Martin, Kulpa, Delamarche, &
Bideau, 2013; Reid, Giblin, & Whiteside, 2015).
It is also not understood if upper-limb isokinetic strength or
anthropometric dierences exist between bowlers of dierent
skill levels or if these factors are linked to performance. In both
FS and fast bowling, the bowler takes advantage of the elbow
extension that naturally occurs late within the kinetic chain
(Marshall & Elliott, 2000; Marshall & Ferdinands, 2003;
Spratford, Elliott, Portus, Brown, & Alderson, 2018; Wixted,
Portus, Spratford, & James, 2011). However, given that a WS
bowler is internally rotating at the shoulder and pronating at
the elbow through BR, the passive extension of the forearm
relative to the upper-arm does not occur and as such may
require more strength to deliver the ball, although no evidence
of this exists. Quantifying the role of upper-limb isokinetic
strength and anthropometric variables will facilitate a better
understanding of WS technique and functional muscle strength
dierences between skill levels, as well as identifying anthro-
pometric variables that may be used in talent identication
�programmes.
Therefore, the aims of this study were to examine upper-
body bowling mechanics, anthropometry and isokinetic
strength across skill levels in WS bowlers. How these variables
inuence performance will be assessed and performance will
be measured by a ball velocity/revolution index, shown to
dierentiate skill level (Spratford et al., 2017). It is hypothesised
that elite bowlers will display higher joint moments, segment
angular velocities and greater isokinetic strength of the bowl-
ing limb compared to pathway WS bowlers. It is also hypothe-
sised that the mechanics of the distal arm will predict bowling
performance across all elite and pathway bowlers examined in
this study.
Methods
Twenty male WS bowlers were invited by the national spin
bowling coach to participate in this study. Participants were
assigned to one of two groups based on the level of cricket
previously played, (1) pathway (up to 1
st
Class) or (2) elite (1
st
class and above). This cohort represented the entire population
within Australia for this level of spin bowler and included three
players who had played Test cricket and one that had played
International 1-day cricket. The physical characteristics of the
participants are outlined in Table 1.
Ethics approval was granted and written informed consent
was obtained for each participant before the commencement
of the study, in accordance with the requirements of the
Human Research Ethics Committees of the Australian Institute
of Sport (AIS) and The University of Western Australia.
Bowling data collection took place in an indoor motion
capture laboratory that was purpose built for cricket analysis
and contained a permanent articial pitch.
Retro-reective markers were axed to each participant’s
head, torso, upper-limbs and ball according to a customised
marker set and model. The set consisted of numerous single
and three marker clusters attached to either semi-rigid plastic
or lightweight aluminium bases. Participant specic static trials
were collected to dene joint centres. The shoulder joint centre
was calculated using a regression equation based on anatomi-
cal landmarks as well as height and weight (Campbell, Lloyd,
Alderson, & Elliott, 2009). The elbow joint centre was estimated
using a pointer method based on the location of the lateral and
medial aspects of the humeral epicondyles (Chin, Lloyd,
Alderson, Elliott, & Mills, 2010) and the wrist joint centre as
the midpoint of markers placed on the styloid processes of
the radius and ulna (Lloyd, Alderson, & Elliott, 2000). Three
hemispherical markers comprised of ultralight foam (<0.1g)
axed in locations that did not impede the bowler’s preferred
grip on the ball were used to calculate ball velocities and
revolutions after BR (Spratford et al., 2017; Whiteside, Chin, &
Middleton, 2012). Marker trajectories were tracked using a 22
camera (MX 13 and 40) Vicon MX motion analysis system
Figure 1. The point of ball release for a right handed WS bowler. Note the ball
leaving the ulnar side of the hand.
Table 1. Mean (± standard deviations) age and physical characteristics of
participants.
Group N Age (yrs) Height (cm) Mass (kg)
Pathway WS 12 19.6 ± 3.6 179.6 ± 6.9 71.0 ± 8.0
Elite WS 8 29.6 ± 7.8 180.2 ± 4.2 71.8 ± 8.0
2W. SPRATFORD ET AL.

(Oxford Metrics, Oxford, UK) operating at 250 Hz for two full
strides before the delivery phase and 3 m of ball ight post-
release.
Participants warmed up as per their normal pre-game rou-
tine and then bowled six overs with a timed two-minute break
between each to replicate match conditions. Participants were
asked to nominate where their usual deliveries would pass
a right-handed batsman based on a clear target placed where
the batsman would normally stand that consisted of a series of
20 cm x 20 cm grids. A valid delivery was one that struck the
target on the nominated grid, the one directly above, under-
neath or immediately next to the nominated grid to the o-side
(to the batsman’s right).
Two-dimensional (2D) data from each of the 22 cameras
were captured for each marker and reconstructed into three-
dimensional (3D) marker trajectories and labelled using Vicon
Nexus software (Oxford Metrics, Oxford, UK). Trajectories were
ltered using a quintic spline Woltring lter at a mean square
error (MSE) of 20 after a residual analysis and visual inspection
of the data (Winter, 2005). Data were then modelled using the
University of Western Australia’s (UWA) upper-body and ball
model (Campbell et al., 2009; Chin, Elliott, Alderson, Lloyd, &
Foster, 2009; Chin et al., 2010; Whiteside et al., 2012). Joint
moments were determined using standard inverse dynamic
analysis starting from the hand and ball segment (treated as
a single segment) and owing to the shoulder joint of the
bowling arm with all segment inertial characteristics taken
from de Leva (1996)�and the weight of the cricket ball assumed
to be 156g. Moments were expressed in a non-orthogonal joint
coordinate system to allow functional meaning and reduce
potential cross talk introduced from kinematic measures
(Middleton, 2011; Schache & Baker, 2007). The ball was
assumed to have minimal inertia but with a point mass, and
force applied by a simple F = ma calculation. The ball created
a reaction force only when it was in contact with the hand. Ball-
hand contact was determined when the ball’s origin was less
than a distance calculated between it and a marker placed on
the distal intermediate carpal as calculated by the length the
fourth phalangeal and radius of the ball. Longer distances
indicated the ball had been released.
Selected biomechanical data for the pelvis, thorax, shoulder,
elbow and wrist considered critical to bowling performance
were included. Joint angles were determined relative to their
adjoining segments, with 0°indicating alignment between
segment coordinate systems. Segment angles (pelvis and
thorax) were measured relative to the global coordinate system
with 0°indicating a pelvis or thorax orthogonal with its den-
ing vector. Therefore, a pelvic and thorax rotation angle of 90°
indicated the bowler was standing front-on or orthogonal with
the pitch. Data were reported at BR and the peaks between
back foot impact (BFI)-BR for the pelvis and thorax and upper
arm horizontal (UAH)-BR for the bowling limb. All variables
were normalised to 101 points using a cubic spline approach
in a custom MATLAB program�(Mathworks Inc; Natick, MA) with
representative mean data being calculated from the rst six
valid WS trials based on the accuracy target information.
Selected anthropometric lengths, breadths and girths were
taken from the bowling limb as well as sitting height by an
accredited ISAK Level 2 anthropometrist (Marfell-Jones, Olds,
Stewart, & Lindsay Carter, 2007). All variables were measured in
triplicate with the criterion being the median (Pyne, Duthie,
Saunders, Petersen, & Portus, 2006). A combination of equip-
ment was used that included a large sliding calliper (British
Indicators Ltd), vernier callipers (Holtain, Cresswell, Dyfed, UK)
and a exible steel tape (Lufkin Executive, Thinline W 606 PM,
Cooper Industries, Lexington, SC, USA). Lengths were measured
for the arm (acromiale-radiale), forearm (radiale–stylion) and
hand (midstylion–dactylion). Breadths were measured for the
wrist (biacromial), distal humerus, transverse chest, and ante-
rior posterior chest depth (Marfell-Jones et al., 2007). Girths
were measured for the upper arm, forearm and wrist. An
assessment of the test-retest reliability was established with
a mean technical error of measurement (TEM) being 1.4% (ran-
ging from 1.2–1.7%).
Active range of motion of the bowling limb was assessed by
the same experienced clinicians using a bi-level inclinometer
and goniometer (US Neurologicals, Poulsbo, Washington,
United States) using previously validated methods (Gerhardt,
Cocchiarella, & Lea, 2002). Measures consisted of: exion, exten-
sion, abduction, internal and external rotation of the shoulder
with the shoulder at 90° of abduction and elbow exed at 90°,
extension and “carry angle” of the elbow, exion, extension,
radial and ulnar deviation of the wrist and exion and extension
of the metacarpophalangeal joint (MCP4). All variables were
measured in triplicate with the mean value used as the criter-
ion. An assessment of the test-retest reliability was established
with a mean TEM being 1.0%.
Bowlers undertook ve isokinetic strength tests on their
bowling arm using a HUMAC NORM (CSMI2009, version 9.5.2)
dynamometer at angular velocities of 60°.s
−1
and 180°.s
−1
. To
reduce the likelihood of fatigue one minutes rest was provided
between each test. Prior to testing, participants warmed-up
using a rowing ergometer for a period of 5 minutes at sub
maximal intensity (Claiborne, Armstrong, Gandhi, & Pincivero,
2006) and were allowed to perform personally selected
stretches if desired. Measures of the shoulder consisted of;
exion/abduction-extension/adduction in the prone position,
internal and external rotation in the prone position with the
shoulder at 90° of abduction and elbow exed at 90° and
normalised to gravity. At the radioulnar joint, pronation and
supination and at the wrist, exion, extension, radial and ulnar
deviation were taken in the sitting position and not corrected
for gravity. All measures were normalised to body weight.
The relationship (x) between ball velocity (V) (m.s
−1
) and
revolutions or angular velocity (rev.s
−1
) around the horizontal
axis of the global coordinate system (ω) was calculated using
the following equation to produce a velocity/revolution index
score. This allowed for a single measure of performance, that
had the ability to include variables that have been shown to be
important for spin bowlers and dierentiate bowling perfor-
mance levels (Spratford et al., 2018,2017).
x¼Vþωð Þ þ V2
ω
A Shapiro-Wilk test was performed to ensure data was normally
distributed prior to independent group t-tests being performed
to establish dierences between elite and pathway bowlers for
measured variables. A partial Bonferroni correction was
JOURNAL OF SPORTS SCIENCES 3

adopted due to the multiple comparisons being made with an
amended alpha level set at α ≤ 0.01. Eect sizes (ES) were
calculated to functionally dierentiate between groups, with
levels of, 0.2, 0.5 and 0.8 representing small, moderate and
large eect sizes respectively (Cohen, 1992). A multiple step-
wise regression analysis was then performed, collapsed across
elite and pathway groups using variables that were shown to
be signicant (p< 0.01) and/or had large eect sizes (ES ≥0.8)
using the velocity/revolution index identied as the dependant
variable. This enabled the measure or combination of measures
that best predicted performance to be identied.
Results
Elite bowlers displayed lower peak pelvic forward rotation
(p= 0.007), while not signicantly dierent, pelvic forward rota-
tion at BR displayed a large eect size (ES = 1.22) with elite
bowlers again exhibiting a more front-on pelvis at BR. Thorax
forward rotation followed a similar pattern with elite bowlers
having a lower peak forward rotation (p= 0.001) indicating
a more front-on position. Elite bowlers showed signicantly
greater amounts of radioulnar pronation at BR (p= <0.001),
peak radioulnar pronation (p= 0.003) and while not statistically
signicant, a greater level of peak wrist radial deviation as
evidenced by a large ES (ES = 0.80). Wrist extension values at
BR were signicantly higher for elite bowlers (p= 0.009) and
a large ES was recorded when comparing between group peak
wrist extension values (ES = 1.29) (Table 2).
Elite bowlers displayed increased levels of radioulnar prona-
tion angular velocity at BR (p= 0.007), peak radioulnar pronation
angular velocity (p= 0.004), wrist radial deviation angular velo-
city at BR, peak wrist radial deviation angular velocity (p= 0.010
and p= 0.010), wrist extension angular velocity at BR (p= 0.004)
and peak wrist extension angular velocity (p= 0.005). At BR large
ESs were also observed with elite bowlers displaying increased
pelvic forward rotation (ES = 1.27) and increased shoulder inter-
nal rotation (ES = 1.78) (Table 2).
Elite bowlers displayed increased shoulder extension (p
= 0.007) and internal rotation moments at BR (p= 0.010), with
a large ES returned for peak shoulder internal rotation
(ES = 0.80) when compared with pathway bowlers. At the
elbow joint large ESs were reported with elite bowlers
experiencing increased exion moment (ES = 1.50) and radio-
ulnar pronation moment at BR (p= 0.82), as well as peak
radioulnar pronation moment (p= 1.03). Large ESs were also
observed at the wrist joint with elite bowlers also displaying
increased peak wrist extension (ES = 0.85) and peak wrist radial
deviation (ES = 1.10) moments (Table 3).
Elite bowlers had increased hand length when compared to
pathway bowlers (midstylion-dactylion) (ES = 1.00). At the wrist
and nger joints, compared to pathway bowlers, elite bowlers
had larger range of motion for absolute radial deviation
(ES = 1.19), radial deviation full range (ES = 1.02), as well MCP4
absolute exion (ES = 1.18), MCP4 absolute extension (ES = 1.48)
and MCP4 total range of motion (ES = 1. 46) (Table 3).
Elite bowlers produced greater isokinetic torque at the
shoulder, elbow and wrist joints in comparison with pathway
bowlers. Signicant dierences were seen for peak wrist exten-
sion torque at 60°.s
−1
(p= 0.001) and peak wrist ulnar deviation
torque at 60°.s
−1
(p= 0.001), while large ESs were returned for
peak shoulder extension/adduction torque at 60°.s
−1
(ES = 1.13), peak wrist radial deviation torque at 60°.s
−1
(ES = 1.33) and 180 °.s
−1
(ES = 1.32) and peak wrist ulnar
deviation torque at 180°.s
−1
(ES = 1.26) (Table 4).
The stepwise multiple regression analysis in which the velo-
city (elite 19.5 ± 0.7 and pathway 18.6 ± 0.7 m.s
−1
) and revolu-
tions (elite 38.7 ± 4.9 and pathway 34.3 ± 4.0 rev.s
−1
) were
collapsed to form the overall index all participants (65.1 ± 4.3)
and used as the criterion variable revealed that peak isokinetic
radial deviation torque, peak shoulder internal rotation
moment, shoulder extension moment at BR and peak prona-
tion moment were the best predictors of WS performance,
explaining 82% of variance when treating the data as a single
cohort (r= 0.90; r
2
= 0.82; f= 15.63, p= < 0.001).
Discussion
The purpose of this study was to examine the dierences in
upper-body bowling mechanics, anthropometry and isokinetic
strength across skill levels in WS bowlers and to assess the
inuence these variables have on bowling performance, as
measured by a velocity/revolution index.
For the WS bowler to deliver the ball with the desired
amount of revolutions and velocity, the upper-body
Table 2. Means (± standard deviations) for selected angular displacement and velocity parameters for elite and pathway WS bowlers.
Variable Pathway Elite p-value Effect size (ES)
Pelvis forward rotation (°) (BR) 114.2 (6.1) 105.9 (7.4) 0.042 1.22
#
Pelvis forward rotation peak (°) (UAH-BR) 176.5 (8.3) 157.6 (11.4) 0.007* 1.90
#
Thorax forward rotation peak (°) (UAH-BR) 170.0 (9.4) 157.6 (8.7) 0.001* 1.37
#
Radioulnar pronation (°) (BR) 62.1 (18.0) 89.2 (9.9) <0.001* 1.87
#
Radioulnar pronation peak (°) (UAH-BR) 67.7 (28.4) 101.3 (14.9) 0.003* 1.48
#
Radial deviation peak (°) (UAH-BR) 20.6 (9.9) 28.5 (9.8) 0.123 0.80
#
Wrist extension (°) (BR) −22.5 (15.2) −37.9 (7.8) 0.009* 1.27
#
Wrist extension peak (°) (UAH-BR) −29.7 (6.3) −37.9 (6.4) 0.026 1.29
#
Pelvis forward rotation (°.s
−1
) (BR) 111.5 (33.0) 162.7 (28.9) 0.026 1.27
#
Shoulder internal rotation (°.s
−1
) (BR) 499.1 (147.7) 655.8 (116.5) 0.041 1.78
#
Radioulnar pronation (°.s
−1
) (BR) −368.9 (126.5) −490.4 (35.7) 0.007* 1.31
#
Radioulnar pronation peak (°.s
−1
) (UAH-BR) −509.6 (112.3) −645.3 (63.9) 0.004* 1.49
#
Radial deviation (°.s
−1
) (BR) −415.7 (142.9) −608.4 (119.5) 0.010* 1.46
#
Radial deviation peak (°.s
−1
) (UAH-BR) −549.4 (198.8) −746.3 (132.0) 0.010* 1.46
#
Wrist extension (°.s
−1
) (BR) −201.0 (131.2) −396.1 (85.5) 0.004* 1.76
#
Wrist extension peak (°.s
−1
) (UAH-BR) −819.2 (150.2) −1056.7 (146.0) 0.005* 1.60
#
*Significant ≤0.010 and
#
Large ES ≥0.80
4W. SPRATFORD ET AL.

undergoes a series of complex and rapid movements
between BFI and BR. Collapsing data across both groups
allowed variables to be identied that are critical to the
performance of high level WS bowling. Our results indicated
that performance is best explained by peak isokinetic wrist
radial deviation torque, peak shoulder internal rotation
moment, shoulder extension moment at BR and the peak
pronation moment.
At BR the pelvis and thorax segments are also rotating
forward, and the shoulder is internally rotating and extending,
the elbow and radioulnar joints exing and pronating, the wrist
extending and moving into ulnar deviation. At this point sev-
eral kinematic and kinetic dierences were observed between
elite and pathway bowlers. Elite bowlers exhibited lower pelvis
and thorax forward rotation displacements at BR, peak forward
thorax rotation and increased pelvis forward rotation angular
velocity at BR. The coaching literature makes mention of the
need to be side-on at BFI and for the shoulders to turn towards
the batsman but is vague on the specic body position at BR
(Philpott, 1995; Tyson, 1994; Woolmer et al., 2008). Results show
that elite bowlers deliver the ball with a pelvis 21° more front-
on, rotate both their thorax and pelvis forward to a lesser
degree between BFI and BR and have an increased pelvis rota-
tion angular velocity at BR, indicating that elite bowlers decele-
rate the pelvis more slowly in the BFI-BR phase. Increases in
trunk rotational angular velocity have also been reported
within the baseball literature when comparing professional
with non-professional players, and has been shown to be
responsible for velocity dependant moments in the distal seg-
ments of the upper-limb, although it must be noted that this
was only observed at the pelvis, and not the thorax for elite WS
bowlers (Aguinaldo, Buttermore, & Chambers, 2007; Hirashima,
Yamane, Nakamura, & Ohtsuki, 2008).
Elite bowlers also exhibited a higher peak shoulder internal
rotation moment and shoulder internal rotation and extension
moments at BR. As mentioned above, trunk angular velocities
have been linked to velocity dependant torques in the distal
segments of the limb. Although the timing of the peak internal
rotation moment suggests that elite bowlers may rely more on
this in driving the subsequent dierences observed in the distal
segments of the kinetic chain rather than increases seen at the
pelvis. This is due mainly to the peak internal rotation moment
occurring early within the UAH-BR phase, as the humerus
moves from peak external rotation into internal rotation, as
reported in baseball throwing (Chu, Fleisig, Simpson, &
Andrews, 2009; Fleisig, Barrentine, Escamilla, & Andrews, 1996;
Fleisig, Chu, Webber, & Andrews, 2008; Vogelpohl & Kollock,
2015). The importance of shoulder internal rotation has been
regularly cited in the literature as critical to maximising the
resultant distal segment velocity in other throwing and hitting
activities (Marshall & Elliott, 2000; Martin et al., 2013; Naito &
Maruyama, 2008; Naito, Takagi, Norimasa, Hashimoto, &
Maruyama, 2014; Reid et al., 2015).
Elite bowlers displayed a similar pattern to throwing and
hitting in the distal bowling limb, with a greater pronation
moment and angular velocity at BR. They also recorded
increased wrist ulnar deviation and extension moments at the
wrist, along with the corresponding angular velocities and
increased displacements of the wrist joint. The general biome-
chanical movements at the distal limb indicate that the WS
bowler must forcefully ex and pronate at the elbow and radio-
ulnar as well as extend and then deviate at the wrist joint up to
and through the point of BR. This supports the theory that elite
bowlers make better use of the degrees of freedom (DoF)
within the proximal to distal linkage system, which is heavily
inuenced by the longitudinal rotations of the upper arm and
Table 3. Group means (± standard deviations) and comparison for selected joint moments, anthropometry and ranges of motion for elite and pathway WS bowlers.
Variable Pathway Elite p-value Effect size (ES)
Shoulder extension moment (Nm/kg) (BR) −0.334 (0.139) −0.531 (0.120) 0.007* 1.53
#
Shoulder internal rot moment (Nm/kg) (BR) 0.018 (0.076) 0.109 (0.047) 0.010* 1.44
#
Shoulder internal rot peak moment (Nm/kg) (UAH-BR) 0.555 (0.177) 0.694 (0.170) 0.129 0.80
#
Elbow flexion moment (Nm/kg) (BR) 0.028 (0.063) 0.134 (0.078) 0.020 1.50
#
Pronation moment (Nm/kg) (BR) −0.030 (0.048) 0.004 (0.033) 0.097 0.82
#
Pronation peak moment (Nm/kg) (UAH-BR) 0.411 (0.186) 0.621 (0.221) 0.090 1.03
#
Wrist extension peak moment (Nm/kg) (UAH-BR) −0.634 (0.192) −0.925 (0.442) 0.102 0.85
#
Radial deviation peak moment (Nm/kg) (UAH-BR) −0.989 (0.297) −1.323 (0.308) 0.031 1.10
#
Midstylion–dactylion length (cm) (hand) 19.4 (0.9) 20.3 (0.9) 0.044 1.00
#
Radial deviation maximum (°) 27.8 (11.2) 42.3 (13.1) 0.057 1.19
#
Ulnar/radial deviation range (°) 76.3 (9.0) 86.0 (10.0) 0.117 1.02
#
MCP4 flexion maximum (°) 94.5 (9.2) 105.1 (9.7) 0.036 1.18
#
MCP4 extension maximum (°) 16.8 (5.5) 26.1 (7.0) 0.029 1.48
#
MCP4 flexion/extension range (°) 111.2 (11.1) 131.2 (15.8) 0.029 1.46
#
*Significant ≤0.010 and
#
Large ES ≥0.80
Table 4. Group means (± standard deviations) for selected peak isokinetic parameters for elite and pathway wrist-spin bowlers.
Variable Pathway Elite p-value Effect size
Shoulder 60°.s-
1
180°.s-
1
60°.s-
1
180°.s-
1
60°.s-
1
180°.s-
1
60°.s-
1
180°.s-
1
Ext/adduction (Nm/kg) 1.15 (0.08) 1.01 (0.10) 1.33 (0.21) 1.00 (0.35) 0.078 0.951 1.13
#
0.03
Wrist
Extension (Nm/kg) 0.11 (0.04) 0.10 (0.03) 0.18 (0.05) 0.16 (0.05) 0.001* 0.031 1.55
#
0.73
Radial deviation (Nm/kg) 0.20 (0.04) 0.16 (0.04) 0.26 (0.05) 0.22 (0.05) 0.048 0.050 1.33
#
1.32
#
Ulnar deviation (Nm/kg) 0.18 (0.07) 0.16 (0.03) 0.31 (0.09) 0.23 (0.06) 0.010* 0.034 1.61
#
1.26
#
*Significant ≤0.010 and
#
Large ES ≥0.8
JOURNAL OF SPORTS SCIENCES 5

forearm, similar to that reported for other overhead striking
activities such as the tennis serve and squash forehand
(Marshall & Elliott, 2000; Martin et al., 2013; Reid et al., 2015).
While injury surveillance data over a 10 year period in
Australia reveals that injury prevalence in spin bowlers is at
4%, a level described within the research as “acceptable”, it fails
to dierentiate between bowler type (FS or WS) (Orchard,
James, Alcott, Carter, & Farhart, 2002; Orchard, James, &
Portus, 2006). It does however report that the greatest propor-
tion of injuries to spin bowlers were to the shoulder tendon
(non-dened). The biomechanical results highlight the impor-
tance of the musculature responsible for developing shoulder
internal rotation as well as exion and pronation at the radio-
ulnar joint, it also highlights the potential loading placed on the
shoulder joint during WS bowling. Coupled without taking
advantage of the centrifugal force to extend the forearm seg-
ment (Wixted et al., 2011), as well as less assistance from the
forward rotating trunk, WS bowlers in comparison with FS and
fast bowlers must rely heavily on manipulating the shoulder in
order to get the body to the appropriate BR position. Research
has also linked decreased range of motion of the bowling arm
in comparison with the non-bowling arm in WS bowlers which
has been shown to increase the chances of subsequent injuries
(Chauhan & Gregory, 2003). During throwing type activities, the
glenohumeral joint must resist large distraction and translation
forces using the rotator cu muscles (teres minor, infraspinatus,
supraspinatus and subscapularis) and the internal rotators (pec-
toralis major, lattisimus dorsi, anterior deltoid and teres major)
and as such can be prone to injury (Fleisig, Andrews, Dilman, &
Escamilla, 1995; Fleisig et al., 1996; Polster et al., 2013). As
previously highlighted, there are major biomechanical dier-
ences between over-head throwing activities, such as baseball
and WS bowling, however, as previously recommended, the
factors that relate to shoulder injuries in WS bowlers warrants
further investigation (Gregory, Batt, & Angus Wallace, 2002).
The anthropometry screen revealed that large ES dierences
were reported for elite compared with pathway WS bowlers.
Hand length, radial deviation full range and total frontal plane
wrist range of motion (radial and ulnar deviation), as well as
MCP4 exion, extension and range of motion all dierentiated
WS bowling level. This provides valuable information that the
size of the hand and increased range of motion in the frontal
plane (radial and ulnar deviation) of the wrist may be of impor-
tance and used to form the basis of talent identication mea-
sures. From an applied coaching perspective, a larger hand
makes the ball easier to hold and ulnar deviation during bowl-
ing has been shown from this research to be important in
discriminating skill level. Functional measurements of the pha-
langeal joints were outside the scope of this current research,
however, high-speed footage taken during testing revealed
that some elite players used their fourth phalange in a similar
way to a FS bowler in an endeavour to control the ball at BR,
a point briey mentioned in the coaching text (Bradman, 1969).
It is possible that this has been reected in the anthropometry
screen, and as such warrant’s further investigation. An example
of this technique is shown in Figure 2.
Isokinetic strength dierences were reported at the shoulder
and wrist joints, with elite bowlers having greater shoulder
extension and adduction torque, wrist extension torque as
well as radial and ulnar deviation torque. There are examples
within the literature of increases in torque production at the
shoulder (adduction and extension) and wrist (extension) being
positively correlated to throwing speed, however this is the rst
to show that elite WS bowlers exhibit such strength proles
(Bartlett, Mitchell, Storey, & Simons, 1989; Pedegana, Elsner,
Roberts, Lang, & Farewell, 1982). It also highlights the potential
benet in strength interventions targeting the musculature
responsible for these movements for WS bowlers.
Conclusion
The results of this study suggest that elite bowlers rotate their
pelvis and thorax to a lesser degree and exhibit less decelera-
tion between BFI-BR resulting in a more front-on position and
increased pelvis rotation angular velocity at BR when com-
pared with pathway bowlers. Importantly, elite bowlers dis-
play higher peak shoulder internal rotation moments as the
upper arm moves from external rotation into internal rotation,
which is thought to be responsible for driving subsequent
dierences in the distal segments of the kinetic chain
between bowling groups. The anthropometry screen high-
lighted variances at the wrist, hand and the fourth phalange,
which may be used to form the basis for talent identication
�programmes. The regression analysis for performance further
reinforced that the WS bowling technique relies heavily on
the bowling limb strength and peak isokinetic strength at the
shoulder, elbow and wrist. While it is recommended that
strength interventions aimed at improving the musculature
responsible for the movement at the shoulder (internal rota-
tion and extension), radioulnar joint (pronation) and wrist
joints (radial deviation) be implemented, it should not come
at a cost to the young bowler learning the correct technique
for this complex skill.
The data from this study were collected within a laboratory
environment, allowing for sophisticated methodologies to be
implemented. This, however, came at the detriment to task
representation due to the absence of a batsman and the clinical
environment. There was also no attempt to quantify the small
amount of pelvis and thorax counter-rotation that occurred
Figure 2. An elite WS bowlers using their fourth phalange through the point of BR.
6W. SPRATFORD ET AL.

between BFI and BR, a movement that has been linked to
injuries in fast bowlers. It is recommended that future research
investigate this area.
Acknowledgments
The authors would like to thank Cricket Australia for support in providing
participants for this project.
Disclosure statement
No potential conict of interest was reported by the authors.�
ORCID
Wayne Spratford http://orcid.org/0000-0002-6207-8829
Jacqueline Alderson http://orcid.org/0000-0002-8866-0913
References
Aguinaldo, A., Buttermore, J., & Chambers, H. (2007). Eects of upper trunk
rotation on shoulder joint torque among baseball pitchers of various
levels. Journal of Applied Biomechanics,23, 42–51.�
Bartlett, R., Mitchell, D., Storey, D., & Simons, B. (1989). Measurement of
upper extremity torque production and its relationship to throwing
speed in the competitive athlete. American Journal of Sports Medicine,
17(1), 89–91.
Beach, A., Ferdinands, R., & Sinclair, P. (2014). Three-dimensional linear and
angular kinematics of a spinning cricket ball. Sports Technology,7(1/2),
12–25.
Beach, A., Ferdinands, R., & Sinclair, P. (2016). The kinematic dierences
between o-spin and leg-spin bowling in cricket. Sports Biomechanics,
15(3), 295–313.
Bradman, D. (1969). The art of cricket (5th ed.). London, England: Hodder
and Stoughton.
Campbell, A., Lloyd, A., Alderson, J., & Elliott, B. (2009). MRI development
and validation of two new predictive methods of glenohumeral joint
centre location identication and comparison with established
techniques. Journal of Biomechanics,42, 1527–1532.
Chauhan, R., & Gregory, P. L. (2003). Dierence in range of shoulder rotation
in wrist spin and nger spin bowlers. Paper presented at the BASEM.�
Chin, A., Elliott, B., Alderson, J., Lloyd, D., & Foster, D. (2009). The o-break
and “doosra”. Kinematic variations of elite and sub-elte bowlers in
creating ball spin in cricket bowling. Sports Biomechanics,8(3), 187–198.
Chin, A., Lloyd, A., Alderson, J., Elliott, B., & Mills, P. (2010). A marker-based
mean nite helical axis model to determine elbow rotation axes and
kinematics in vivo. Journal of Applied Biomechanics,26, 16.
Chu, Y., Fleisig, G., Simpson, K., & Andrews, J. (2009). Biomechanical com-
parison between elite female and male baseball pitchers. Journal of
Applied Biomechanics,25, 22–31.
Claiborne, T., Armstrong, C., Gandhi, V., & Pincivero, D. (2006). Relationship
between hip and knee strength and knee valgus during a single leg
squat. Journal of Applied Biomechanics,20, 41–50.
Cohen, J. (1992). A power primer. Psychological Bulletin,112(1), 155–159.
Cork, A., Justham, L., & West, A. (2012). Three-dimensional vision analysis to
measure the release characteristics of elite bowlers in cricket. Journal of
Sports Engineering and Technology, Part P,227, 1–12.
Diaz, G., Cooper, J., Rothkopf, C., & Hayhoe, M. (2013). Saccades to future
ball location reveal memory-based prediction in virtual-reality intercep-
tion task. Journal of Vision,3(1), 1–14.
Fleisig, G., Andrews, J., Dilman, C., & Escamilla, R. (1995). Kinetics of baseball
pitching with implications about injury mechanisms. The American
Journal of Sports Medicine,23(2), 233–239.
Fleisig, G., Barrentine, S., Escamilla, R., & Andrews, J. (1996). Biomechanics of
overhand throwing with implications for injuries. Sports Medicine,21,
421–437.
Fleisig, G., Chu, Y., Webber, A., & Andrews, J. (2008). Variability in baseball
pitching biomechanics among various levels of competition. Sports
Biomechanics,8(1), 10–21.
Gerhardt, J., Cocchiarella, L., & Lea, R. (2002). The practical guide to range of
motion assessment (1st ed.). American Medical Association.�
Gregory, P., Batt, M., & Angus Wallace, W. (2002). Comparing injuries of spin
bowling with fast bowling in young cricketers. Clinical Journal of Sports
Medicine,12, 107–112.
Hirashima, M., Yamane, K., Nakamura, Y., & Ohtsuki, T. (2008). Kinetic chain
of overarm throwing in terms of joint rotations revealed by induced
acceleration analysis. Journal of Biomechanics,41(18), 2874–2883.
Justham, L., Cork, A., & West, A. (2010). Comparative study of the perfor-
mances during match play of an elite-level spin bowler and an elite-level
pace bowler in cricket. Journal of Sports Engineering and Technology, Part
P,224, 237–247.�
Lloyd, D., Alderson, J., & Elliott, B. (2000). An upper limb kinematic model for
the examination of cricket bowling: A case study of Muttiah
Muralitharan. Journal of Sports Sciences,18, 975–982.
Marfell-Jones, M., Olds, T., Stewart, A., & Lindsay Carter, J. (2007).
International standards for anthropmetric assesment. International
Society for the Advancenet of Kinanthropometry.�
Marshall, R., & Elliott, B. (2000). Long-axis rotation: The missing link in
proximal-to- distal segmental sequencing. Journal of Sport Sciences,18,
247–254.
Marshall, R., & Ferdinands, R. (2003). The eect of a exed elbow on bowling
speed in cricket. Sports Biomechanics,2(1), 65–72.
Martin, C., Kulpa, R., Delamarche, P., & Bideau, B. (2013). Professional tennis
players’ serve: Correlation between segmental angular momentums and
ball velocity. Sports Biomechanics,12(1), 2–14.
Middleton, K. (2011). Mechanical strategies for the devlopment of ball release
velocity in cricket fast bowling: The interaction, variation and simulation of
execution and outcome parameters (Doctor of Philosophy). University of
Western Australia,
Naito, K., & Maruyama, T. (2008). Contributions of the muscular torques and
motion-dependant torques to generate rapid elbow extension during
overhand baseball pitching. Sports Engineering,11, 47–56.
Naito, K., Takagi, H., Norimasa, Y., Hashimoto, S., & Maruyama, T. (2014).
Intersegmental dynamics of 3D upper arm and forearm longitudinal axis
rotations during baseball pitching. Human Movement Science,38,
116–132.
Orchard, J., James, T., Alcott, E., Carter, S., & Farhart, P. (2002). Injuries in
Australian cricket at rst class level 1995/96 to 2000/2001. British Journal
of Sports Medicine,36, 270–275.
Orchard, J., James, T., & Portus, M. (2006). Injuries to elite male cricketers in
Australia over a 10 year period. Journal of Science and Medicine in Sport,
9, 459–468.
Pedegana, L., Elsner, R., Roberts, D., Lang, J., & Farewell, V. (1982).
Relationship of upper extremity strength to throwing speed. American
Journal of Sports Medicine,10(6), 352–354.
Philpott, P. (1995). The art of wrist-spin bowling. Ramsbury, Marlborough:
Crowood Press.
Polster, J., Bullen, J., Obuchowski, N., Bryan, J., Solo, L., & Schickendantz, M.
(2013). Relationship between humeral torsion and injury in professional
baseball pitchers. The American Journal of Sports Medicine,41(9),
2015–2021.
Pyne, D., Duthie, G., Saunders, P., Petersen, C., & Portus, M. (2006).
Anthropometric and strength correlates of fast bowling speed in junior
and senior cricketers. Journal of Strength and Conditioning Research,20
(3), 620–626.
Reid, M., Giblin, G., & Whiteside, D. (2015). A kinematic comparison of the
overhand throw and tennis serve in tennis players: How similar are they
really? Journal of Sport Sciences,33(7), 713–723.
Robinson, G., & Robinson, I. (2013). The motion of an arbitrarily rotating
spherical projectile and its application to ball games. Physica Scripta,88
(1), 018101–018117.
Sarpeshkar, V., Mann, D., Spratford, W., & Abernethy, B. (2017). The inuence
of ball-swing on the timing and coordination of a natural interceptive
task. Human Movement Science,54, 82–90.
Schache, A., & Baker, R. (2007). On the expression of joint moments during
gait. Gait and Posture,25, 440–452.
JOURNAL OF SPORTS SCIENCES 7

Spratford, W., & Davison, J. (2010). Measurement of ball ight characteristics
in nger-spin bowling. Paper presented at the Conference of Science,
Medicine and Coaching in Cricket. Gold Coast, Australia.
Spratford, W., Elliott, B., Portus, M., Brown, N., & Alderson, J. (2018).
Illegal bowling actions contribute to performance in cricket
nger-spin bowlers. Scandinavian Journal of Medicine & Science in
Sports,28, 1–9.
Spratford, W., Portus, M., Wixted, A., Leadbetter, R., & James, D. (2014). Peak
outward acceleration and ball release in cricket. Journal of Sports
Sciences,33(7), 754–760.
Spratford, W., Whiteside, D., Elliott, B., Portus, M., Brown, N., & Alderson, J.
(2017). Does performance level aect initial ball ight kinematics in
nger and wrist-spin cricket bowlers? Journal of Sports Sciences.
doi:10.1080/02640414.2017.1329547
Tyson, F. (1994). The cricket coaching manual (2nd ed.). Melboune, Australia:
Nelson in association with the Victorian Cricket Association.
Vogelpohl, R., & Kollock, R. (2015). Isokinetic rotator cu functional ratios
and the development of shoulder injury in collegiate baseball pitchers.
International Journal of Athletic Therapy and Training,20(3), 46–52.
Welchman, A., Tuck, V., & Harris, J. (2004). Human observers are biased in
judging the angular approach of a projectile. Vision Research,44(17),
2027–2042.
Whiteside, D., Chin, A., & Middleton, K. (2012). The validation of a
three-dimensional ball rotation model. Journal of Sports Engineering
and Technology,227, 8.
Wilkins, B. (1991). The bowler’s art. London, England: A & C Black Ltd.
Winter, D. (2005). Biomechanics and motor control of human movement (3rd
ed.). New Jersey, NJ: John Wiley & Sons.
Wixted, A., Portus, M., Spratford, W., & James, D. (2011). Detection of throwing
in cricket using wearable sensors. Sports Technology,4, 134–140.
Woolmer, B., Noakes, T., & Moett, H. (2008). Bob Woolmer’s art and science
of cricket (1st ed.). Sydney: New Holland Publishers Ltd.
8W. SPRATFORD ET AL.