Journal of Sports Sciences

ISSN: 0264-0414 (Print) 1466-447X (Online) Journal homepage: http://www.tandfonline.com/loi/rjsp20

The effect of elbow hyperextension on ball speed

in cricket fast bowling

P. J. Felton & M. A. King

To cite this article: P. J. Felton & M. A. King (2016): The effect of elbow hyperextension on ball
speed in cricket fast bowling, Journal of Sports Sciences, DOI: 10.1080/02640414.2015.1137340

To link to this article: http://dx.doi.org/10.1080/02640414.2015.1137340

Published online: 28 Jan 2016.

Submit your article to this journal

Article views: 47

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=rjsp20

Download by: [Ming Chuan University] Date: 23 March 2016, At: 21:42
 JOURNAL OF SPORTS SCIENCES, 2016
http://dx.doi.org/10.1080/02640414.2015.1137340

The effect of elbow hyperextension on ball speed in cricket fast bowling

P. J. Felton and M. A. King
School of Sport, Exercise and Health Sciences, Loughborough University, Leicestershire, UK

ABSTRACT ARTICLE HISTORY

This study investigates how elbow hyperextension affects ball release speed in fast bowling. A two- Accepted 25 December 2015
segment planar computer simulation model comprising an upper arm and forearm + hand was KEYWORDS
customised to an elite fast bowler. A constant torque was applied at the shoulder and elbow Simulation; modelling; pace
hyperextension was represented using a damped linear torsional spring at the elbow. The magnitude bowler; ball velocity
of the constant shoulder torque and the torsional spring parameters were determined by concurrently
matching three performances. Close agreement was found between the simulations and the perfor-
mances with an average difference of 3.8%. The simulation model with these parameter values was
then evaluated using one additional performance. Optimising ball speed by varying the torsional spring
Downloaded by [Ming Chuan University] at 21:42 23 March 2016

parameters found that elbow hyperextension increased ball release speed. Perturbing the elbow
torsional spring stiffness indicated that the increase in ball release speed was governed by the
magnitude of peak elbow hyperextension and the amount that the elbow recoils back towards a
straight arm after reaching peak elbow hyperextension. This finding provides a clear understanding that
a bowler who hyperextends at the elbow and recoils optimally will have an increase in ball speed
compared to a similar bowler who cannot hyperextend. A fast bowler with 20° of elbow hyperextension
and an optimal level of recoil will have increased ball speeds of around 5% over a bowler without
hyperextension.

Introduction extension is usually limited by the tension in the anterior

joint capsule and flexor muscles and to some extent in the
Fast bowling is a dynamic activity within cricket where the
anterior parts of the collateral ligament (Palastanga, Field, &
bowler utilises the speed at which they are able to deliver the
Soames, 2002). It is speculated that during fast bowling, the
ball towards the batsman. The fastest bowlers are capable of
load on the elbow can cause elbow hyperextension to occur
delivering the ball in excess of 40 m/s (90 mph) (Worthington,
(Ferdinands & Kersting, 2004; Portus et al., 2006) with peak
King, & Ranson, 2013). The bowling action can be thought of
hyperextension angles reported in excess of 20° (King &
as a series of segmental movements which ends with the
Yeadon, 2012). The results of these investigations into the
forward rotation of the bowling arm (Bartlett, Stockill, Elliott,
effect of elbow extension on ball speed suggest that there
& Burnett, 1996). Previous research has suggested that the
appears to be a relationship between elbow extension and
action of the bowling arm during this delivery period (the
ball release speed (Portus et al., 2006; Roca, Elliott, Alderson, &
phase between the upper arm being horizontal and ball
Foster, 2006). Further research by Middleton, Alderson, Elliott,
release) is the most important aspect for ball release speed
and Mills (2015) has suggested, however, that increased joint
with contributions of 40–50% to the final ball release speed
extension does not necessarily result in increased wrist velo-
coming from the angular displacement of the bowling arm
city but those bowlers who flex their elbow joint immediately
(Davis & Blanksby, 1976; Elliott, Foster, & Gray, 1986). There is
prior to ball release gain an advantage in increased ball
still a lack of understanding however, regarding the effect of
release velocity. A linear relationship was also found when
the elbow joint angle time history on ball release speed.
investigating how a flexion-extension offset from a straight
Research investigating the effect of elbow extension on ball
arm affects wrist velocity with a hyperextended arm being
release speed has been motivated by a law governing bowling
slowest and a flexed arm fastest. Non-constant elbow hyper-
in cricket which prohibits elbow extension exceeding 15°
extension time histories, however, were not investigated.
between the upper arm reaching horizontal and ball release
In order to investigate the effect of elbow hyperextension
(International Cricket Council [ICC], 2015). During these inves-
on ball release speed, a theoretical approach can be used. Two
tigations, elbow hyperextension has been witnessed during
forward dynamic simulation models have previously been
the bowling action in the joint angle-time history (Ferdinands
developed for cricket bowling (Ferdinands, Kersting, &
& Kersting, 2004; King & Yeadon, 2012; Portus, Rosemond, &
Marshall, 2008; Middleton et al., 2015). Ferdinands et al.
Rath, 2006). Elbow hyperextension occurs when the joint
(2008) developed a preliminary model which required kinetic
angle exceeds a straight position (180°) which is considered
inputs derived from inverse dynamics which could be manipu-
to be the anatomical range of motion (Alter, 2004). Elbow
lated to elicit kinematic effects. Middleton et al. (2015)

CONTACT Mark King [email protected] School of Sport, Exercise and Health Sciences, Loughborough University, Leicestershire LE11 3TU, UK
© 2016 Taylor & Francis
 2 P. J. FELTON AND M. A. KING

developed a 3D model which required the joint angle-time

histories to be input. The elbow joint-angle time history was
manipulated to investigate the effect differing joint angle-time
histories had on wrist speed. The validity of the results in this
model was compromised however, since the manipulations of
the input variables did not take into consideration the
mechanical properties of muscles.
Although elbow hyperextension is exempt from counting
towards the 15° extension limit since it is considered to be an
involuntary movement caused by the load on the elbow dur-
ing the bowling action (ICC, 2015), there is still a lack of
understanding of the effect of elbow hyperextension on ball
release speed as previous research has not distinguished
between extension and hyperextension. The aim of this
study was to determine the effect of elbow hyperextension
on ball speed in fast bowling through the use of a subject-
specific simulation model.
Downloaded by [Ming Chuan University] at 21:42 23 March 2016

Methods
A four-stage theoretical process was used to investigate the
effect of elbow hyperextension on ball release speed in fast
bowling (King & Yeadon, 2013). The model was developed,
customised to an elite bowler, evaluated by comparison with
the elite bowlers’ performance and then used to investigate
the effect of elbow hyperextension on ball release speed. Figure 1. Two-segment simulation model of the bowling arm. The torque
generator Ts opens the shoulder joint angle θs, and a torsional spring TE allows
hyperextension of the elbow (θE > 0).
Data collection
Performance data were collected from a member of the
England and Wales Cricket Board elite fast bowling group relative to the downwards vertical) and the elbow joint
(age 19 years; height 1.80 m; mass 82.4 kg) at the National angle (Figure 1). A quadratic function was fitted to the
Cricket Performance Centre in accordance with Loughborough time history of the horizontal and vertical displacement of
Universities Ethical Advisory Committee guidelines. Four max- the shoulder joint centre in the sagittal plane so that deri-
imal ball speed bowling trials of a good length were recorded vatives could be derived.
using an 18-camera (MX13) Vicon Motion Analysis System Ball release was determined as the first frame where the
(OMG Plc, Oxford, UK) operating at 300 Hz on a standard distance between the ball marker and wrist joint centre had
length indoor cricket pitch. Three pairs of 14 mm retro-reflec- increased more than 5 cm (Worthington et al., 2013). The
tive markers were attached across the wrist, elbow and coordinates of the reflective tape on the ball in the sagittal
shoulder joints on the bowling arm such that their mid-points plane were used to calculate the ball release velocity as the
coincided with the joint centres (King & Yeadon, 2012) and a average resultant velocity calculated over the first five frames
reflective patch (approximately 15 × 15 mm) was attached to after ball release. The average percentage increase between
the ball to enable ball release velocity and the instant of ball wrist and ball speed at ball release across the four trials was
release to be determined. also calculated in order to establish the general increase in ball
speed due to wrist flexion for the participant used in this
study.
Data processing
The four trials were manually labelled and initially processed
Simulation model
using the Vicon Nexus software with all trials tracked with-
out any marker loss. All marker trajectories were then fil- A two-segment planar simulation model of the bowling arm
tered using a recursive fourth-order low-pass Butterworth delivery period of fast bowling (Figure 1) was constructed
filter with a cut-off frequency of 30 Hz determined using a using AutolevTM (Kane & Levinson, 1985). The two segments
residual analysis (Winter, 1990). The three-dimensional wrist, represented the upper arm and lower arm + hand segments. A
elbow and shoulder joint centre-time histories were calcu- ball was included at the end of the forearm + hand segment.
lated from the pairs of markers across the wrist, elbow and A constant torque generator was employed at the shoulder
shoulder. The projection of the joint centres on the sagittal TS, which opened (extended) the shoulder joint angle θs. The
plane (vertical plane parallel to a line joining the two middle shoulder joint centre was driven horizontally using the displa-
stumps together) was then used to determine the orienta- cement time history from the performance data. The vertical
tion angle (the angle of the upper arm in the sagittal plane displacement of the shoulder joint centre was ignored since
 JOURNAL OF SPORTS SCIENCES 3

the performance data showed minimal movement (<0.025 m) matching parameters were fixed and a single simulation run
throughout the delivery period. (King & Yeadon, 2013). The trial with the best match was then
The torque at the elbow was modelled as a damped linear selected to provide the initial inputs to the simulation model
torsional spring which only acted when the elbow was in for all subsequent simulations.
hyperextension (Lundon, 2007):

kE θE  cE θ_ E ; θE  0 Simulations investigating the effect of elbow
TE ¼ hyperextension on ball release speed
0; θE < 0
Initially, to quantify the effect of elbow hyperextension on ball
where θE is the elbow joint angle, θ_ E is the elbow joint angular
release speed, the ball speed for the best matched simulation
velocity, kE is the torsional spring stiffness and cE is the tor-
was compared to the ball speed for a simulation with the one-
sional spring damping.
segment model where the same matched inputs were used.
Ball release was defined to have occurred once the upper
Secondly, to quantify the maximum effect of elbow hyperex-
arm had passed the vertical, and the calculated horizontal
tension on ball release speed, an optimisation was run where
projectile distance travelled by the ball to the predicted land-
the torsional spring parameters kE and cE were varied using
ing site matched the performance data. This was in order to
the Simulated Annealing algorithm (Corana et al., 1987) in
ensure the outcome of each simulation delivered a ball which
order to maximise ball release speed. A penalty was imposed
landed in the same place and was therefore comparable.
if peak elbow hyperextension exceeded an upper bound of
Input to the simulation model comprised the magnitude of
Downloaded by [Ming Chuan University] at 21:42 23 March 2016

25° based on previous research (King & Yeadon, 2012). Thirdly,

the constant shoulder torque TS, the torsional spring para-
to investigate the relationship between the magnitude of
meters kE and cE , the horizontal shoulder joint centre displace-
elbow hyperextension and ball release speed, simulations
ment time history, the segmental inertial parameters and the
were required with different elbow hyperextension angle-
initial orientation and angular velocity of the upper arm and
time histories. To achieve this, the elbow torsional spring
lower arm + hand segments. The output from the simulation
stiffness kE was perturbed to give a wide range of realistic
model comprised the shoulder and elbow joint angle time
elbow hyperextension time histories whilst the elbow torsional
histories as well as the tangential velocity of the wrist, which
spring damping term cE was set to the optimised value. The
was converted to ball speed using the average percentage
time where elbow hyperextension would start to occur within
increase from the performance data in order to incorporate
the delivery period was also varied. In order to do this, the first
the effect of wrist flexion.
part of the delivery period was simulated using the one-seg-
In order to quantify the effect of elbow hyperextension on
ment model before switching to the two-segment model at
ball release speed compared to a straight arm, a one-segment
different times in the delivery period with the output from the
planar simulation model was also constructed. The one-seg-
one-segment model input to the two-segment model.
ment model had the same inputs and outputs as the two-
Penalties were included to prevent unrealistic elbow hyperex-
segment model but the elbow joint was omitted and the
tension angle time histories where all simulations included in
single segment represented the upper arm + forearm + hand.
the final analysis incurred no penalties.

Parameter determination Results

The segmental inertia parameters were calculated using the The simulation model closely matched the movement of the
inertia model of Yeadon (1990) from 95 anthropometric mea- bowling arm during the delivery period with overall difference
surements taken from the elite fast bowler. A common set of values of 4.4%, 4.5% and 2.5% for the three matched trials
parameters consisting of the magnitude of the constant (Figure 2). An evaluation simulation, in which a fourth bowling
shoulder torque TS, and the torsional spring parameters at the trial was simulated using the matched parameters, returned a
elbow kE and cE , were determined concurrently for three max- RMS difference function value of 4.4% (Table 1).
imal speed bowling performances using the Simulated The best match (M3) simulated ball release speed to be
Annealing algorithm to minimise an objective function (Corana 85.8 mph (38.1 m/s). When the initial conditions for this match
et al., 1987). The objective function was the average of a cost were input into the one-segment model, ball release speed
function defined as a root mean square (RMS) score of the was simulated to be 82.5 mph (36.7 m/s). This indicated that
absolute differences between the simulation and recorded per- the bowler’s elbow hyperextension caused an increase in ball
formance for four variables: ball release speed, total time of release speed of 4% (Table 1).
simulation, maximum elbow hyperextension angle and elbow Optimising ball release speed by varying the torsional
extension angle at ball release. Each difference score was spring parameters found a solution with a peak elbow hyper-
weighted equally and 1° difference was considered to be extension of 25° (upper limit) and 5° of recoil. This optimal
equivalent to 1% difference (Yeadon & King, 2002). simulation had a ball release speed of 86.6 mph (38.5 m/s).
This equated to an increase of 5% compared to bowling with a
straight arm (Table 1). In the optimum solution, the damping
Evaluation of the model
parameter was equal to zero.
The robustness of the matching set of parameters was eval- Perturbing the spring stiffness and varying the start time of
uated using a fourth performance by the bowler, where the elbow hyperextension resulted in 7000 simulations with
 4 P. J. FELTON AND M. A. KING
Downloaded by [Ming Chuan University] at 21:42 23 March 2016

Figure 2. Comparison of key kinematic variables during the bowling delivery

period for Match 1; solid line – actual, dashed line – simulation.
Figure 3. The relationship between peak elbow hyperextension and gain in ball
speed (compared to a straight arm) for (a) “recoiling” and (b) “at peak” categories.

different elbow hyperextension characteristics. These simula-

tions were classified into one of two categories depending on
their elbow angle time history: “recoiling” – the elbow hyper-
extends and is recoiling at the instance of ball release; or “at
peak” – the elbow hyperextends and is at peak hyperexten-
sion at ball release. In both these categories, the ball release
speed was faster than bowling with a straight arm (Figure 3).
Further investigation of the simulations in the recoiling cate-
gory indicated that an optimal amount of recoil exists in order
to maximise ball release speed for each amount of peak elbow
hyperextension (Figure 4). For both the “recoiling” and “at
peak” groups, proximal to distal sequencing was evident
with the peak upper arm angular velocity occurring prior to
the lower arm peak angular velocity at ball release.
The relationship between the optimal recoil percentage and
peak elbow hyperextension to maximise ball speed was found
to be inversely hyperbolic (R2 = 0.7) (Figure 5). Rapid growth in Figure 4. The percentage gain in ball speed (compared to a straight arm) as a
the optimal recoil percentage only occurred when the peak function of peak elbow hyperextension and recoil percentage.
hyperextension was less than 1°. Therefore, for the majority of
hyperextensions seen within fast bowling (>1°), the optimal approaches 1°. The initial exponential phase (<1°) provides a 1%
recoil lies between 30% and 60% of the theoretical maximum, gain in ball release speed and thereafter each 1° of hyperexten-
becoming closer to 60% the closer the peak hyperextension sion leads to a gain in ball speed of 0.2% (Figure 6).

Table 1. The initial conditions, matching parameters and RMS differences for the matching and evaluation simulations.
Initial conditions Matching parameters Matched RMS differences
θS (°) θE (°) θ′S (°s−1) θ′E (°s−1) Ball speed (mph) Peak θE (°) θE at BR (°) Time (s) Ball speed (%) Peak θE (°) θE at BR (°) Time (%) Overall RMS
M1 109 4.3 1272 379 86.1 13.5 11.3 0.07 5.0 0.1 0.4 7.6 4.5
M2 93 0.5 1261 144 86.3 13.9 11.9 0.07 1.1 0.0 1.1 8.6 4.4
M3 93 0.1 1250 164 86.8 14.0 12.3 0.06 2.2 0.2 1.4 4.3 2.5
E 99 0.0 1187 203 86.9 14.3 12.9 0.06 4.2 0.5 2.5 7.4 4.4
Notes: Mi, match; E, evaluation; θS, shoulder angle; θE, elbow angle.
 JOURNAL OF SPORTS SCIENCES 5

Figure 5. The relationship between recoil percentage and peak elbow hyper-
extension for the fastest simulation for each torsional spring stiffness.
Figure 7. The work done at the shoulder between upper arm horizontal (UAH)
and ball release (BR) for (a) a straight arm and (b) a hyperextending elbow.

angular velocities of the elbow and shoulder act in the same

direction and as a consequence ball speed is increased. These
two mechanisms work against one another; as the elbow
Downloaded by [Ming Chuan University] at 21:42 23 March 2016

recoils, the increase in work done at ball release due to the

first mechanism is reduced (Figure 8). This creates a trade-off
between the two and an optimal recoil percentage exists for
each peak elbow hyperextension in order to maximise the
gain in ball speed (Figure 4).
The increase in ball speed of non-recoiling elbow hyperex-
tensions over a straight arm disagrees with the results found
Figure 6. The relationship between peak elbow hyperextension and percentage
gain in ball speed when optimal recoil occurs.
by Middleton et al. (2015), which indicated that an elbow with
a fixed offset in hyperextension bowled slower than a straight
arm. It is proposed that an increase in ball speed is always
Discussion possible with a flexion/extension and/or abduction/adduction
The ball release speed of the bowler in this study was
increased by 4% due to elbow hyperextension when com-
pared to a straight arm. Optimising ball speed by varying
the two parameters which govern the laxity of the elbow
joint showed that a larger peak elbow hyperextension was
better with the arm starting to recoil before ball release.
Perturbing the stiffness of the spring governing the amount
of elbow hyperextension possible within the simulation model
found that any amount of elbow hyperextension increases ball
release speed during fast bowling although the increase is
governed by the magnitude of peak elbow hyperextension
and the amount the elbow recoils. To optimise the increase in
ball speed, the larger the amount of elbow hyperextension the
better as long as the elbow recoil percentage is optimal. The
optimal recoil percentage to maximise the ball release speed
for each peak elbow hyperextension has an inverse hyperbolic
relationship.
The increase in ball speed caused by elbow hyperextension
is a consequence of two mechanisms within the bowling
delivery. Firstly, the simulations where the elbow reaches
peak hyperextension at ball release reveal that in order to
satisfy the ball release criteria (i.e., release the ball towards
the same landing point) the shoulder release angle has to
increase as the elbow hyperextension increases. This allows
the shoulder torque to be applied over a longer period which
increases the work done by the shoulder. As a result, the
upper arm has a faster angular velocity at ball release and as
Figure 8. The work done at the shoulder between upper arm horizontal (UAH)
a consequence a faster ball speed (Figure 7). Secondly, in the and ball release (BR) and the angular velocity of the elbow at BR for (a) a
simulations where the elbow is recoiling at ball release, the straight arm, (b) at peak and (c) recoiling elbow time histories.
 6 P. J. FELTON AND M. A. KING

offset as long as the orientation of the upper arm increases braking of the shoulder joint centre increases elbow hyperex-
the ability for the shoulder to do work as suggested by tension and/or changes the optimal amount of recoil.
Marshall and Ferdinands (2003). Previous research investigat- In summary, a two-dimensional simulation model capable of
ing the effect of flexion or extension of the elbow from upper recreating the kinematics of the bowling arm delivery period in
arm horizontal to ball release has found differing results where fast bowling showed that elbow hyperextension along with
both flexion (Middleton et al., 2015) and extension (Portus optimal recoil increased ball release speed. Although it may be
et al., 2006; Roca et al., 2006) have been shown to increase possible for bowlers who do not hyperextend to bowl faster
ball speed. The results in this study show that a greater than those who hyperextend due to other technique or strength
increase in ball speed is caused by the recoil (mechanism 2) parameters, a bowler who can hyperextend at the elbow and
than extension to peak hyperextension (mechanism 1) which recoil optimally will have an increase in ball speed compared to
agrees with Middleton et al. (2015). In reality, however, it is a similar bowler who cannot hyperextend. For example, a bow-
probable that the increase in ball speed caused by the second ler with an optimal recoil peak hyperextension of 20° will experi-
mechanism can be achieved by either flexion or extension ence an increase in ball speed of 5% over a bowler with a
depending on the orientation of the upper arm. If the flex- straight arm. At an elite level, in which fast bowlers are seen to
ion–extension axis is orientated such that flexion is away from bowl in excess of 90 mph (40 m/s), this equates to an increase of
the target and extension is towards, then the first mechanism 5 mph (2. m/s), which is a substantial increase in performance.
explained in this study is caused by flexion and the second
mechanism by extension. If, however, the flexion–extension
Disclosure statement
Downloaded by [Ming Chuan University] at 21:42 23 March 2016

axis is aligned so that extension is away from the target and

flexion is towards, then the roles are reversed and the first No potential conflict of interest was reported by the authors.
mechanism is caused by extension and the second by flexion.
The main application of this work is to give a clear under-
standing of how movements at the elbow effect ball release Funding
speed in fast bowling and the potential advantage individuals This project was funded by the England and Wales Cricket Board (ECB).
with the ability to hyperextend at the elbow have over those
bowlers who cannot hyperextend.
Although, this two-dimensional planar simulation model References
can offer an explanation as to how the kinematics of the Alter, M. J. (2004). Science of flexibility. Champaign, IL: Human Kinetics.
elbow joint can affect the mechanical system and increase Bartlett, R. M., Stockill, N. P., Elliott, B. C., & Burnett, A. F. (1996). The
ball speed, it is not exempt from limitations. The elbow joint biomechanics of fast bowling in men’s cricket: A review. Journal of
is a complex three-dimensional system which has been gen- Sports Sciences, 14, 403–424. doi:10.1080/02640419608727727
Corana, A., Marchesi, M., Martini, C., & Ridella, S. (1987). Minimizing multi-
eralised in two dimensions within this study. The degrees of modal functions of continuous variables with the simulated annealing
freedom which have been omitted would reduce the length of algorithm. ACM Transactions on Mathematical Software, 13, 262–280.
the forearm + hand and upper arm segments if the upper arm doi:10.1145/29380.29864
was rotated away from the plane and/or an abduction–adduc- Davis, K., & Blanksby, B. (1976). The segmental components of fast bowling
tion angle existed. This would create a trade-off between the in cricket. Australian Journal for Health, Physical Education and
Recreation, 71(Suppl), 6–8. doi:10.1080/07303084.2000.10605990
increase in angular velocity due to the reduced inertia of the Elliott, B. C., Foster, D. H., & Gray, S. (1986). Biomechanical and physical
arm and the decrease in the linear velocity of the wrist factors influencing fast bowling. Australian Journal of Science and
towards the target. It is speculated that a three-dimensional Medicine in Sport, 18, 16–21.
model would follow the same mechanics as the two-dimen- Ferdinands, R. E. D., & Kersting, U. G. (2004). Elbow angle extension and
sional model where the optimal solution maximises the ability implications for the legality of the bowling action in Cricket. In A.
McIntosh (Ed.), Proceedings of Australasian biomechanics conference
for work to be done at the shoulder before the joint moves (Vol. 5, pp. 26–27). Sydney: University of New South Wales Press.
back towards the target to optimise the trade-off between the Ferdinands, R. E., Kersting, U., & Marshall, R. N. (2008). A preliminary
two mechanisms. This suggests bowlers with large abduction forward solution model of cricket bowling. International Journal of
angles and more movement towards the target using flexion Sports Science and Engineering, 2, 211–215.
or extension are likely to benefit with increased ball speeds International Cricket Council. (2015). ICC regulations for the review of
bowlers reported with suspected illegal bowling actions. Dubai:
compared to those with straighter arms. International Cricket Council. Retrieved from http://www.icc-cricket.
In the future, the model could be developed to investigate com/cricket-rules-and-regulations/
whether the mechanics discussed in this study also hold true Kane, T. R., & Levinson, D. A. (1985). Dynamics, theory and applications.
for flexion and extension by adding an active torque generator New York: McGraw Hill.
at the elbow, as well as increasing the complexity of the King, M. A., & Yeadon, M. R. (2012). Quantifying elbow extension and
elbow hyperextension in cricket bowling: A case study of Jenny
shoulder torque profile. A hand segment could also be intro- Gunn. Journal of Sports Sciences, 30, 937–947. doi:10.1080/
duced to directly investigate the effects of the wrist joint on 02640414.2012.682082
elbow hyperextension and ball speed. Close agreement, how- King, M. A., & Yeadon, M. R. (2013). Advances in the development of whole
ever, was found between the matching simulations and the body computer simulation modelling of sports technique. Movement &
recorded performances indicating the model was capable of Sport Sciences-Science & Motricité. doi:10.1051/sm/2013048
Lundon, K. (2007). The effect of mechanical load on soft connective
reproducing the relevant parts of the bowling action for this tissues. In W. Hammer (Eds.), Functional soft-tissue examination and
study. In addition, the effect of varying the shoulder displace- treatment by manual methods (3rd ed., pp. 15–30). Boston, MA: Jones
ment profiles could also be investigated to determine whether and Bartlett.
 JOURNAL OF SPORTS SCIENCES 7

Marshall, R., & Ferdinands, R. (2003). Cricket: The effect of a flexed elbow medium bowlers. Journal of Sports Sciences, 24(11), 1127–1135.
on bowling speed in Cricket. Sports Biomechanics, 2, 65–71. doi:10.1080/02640410500497618
doi:10.1080/14763140308522808 Winter, D. A. (1990). Biomechanics and motor control of human movement.
Middleton, K. J., Alderson, J. A., Elliott, B. C., & Mills, P. M. (2015). The influence New York, NY: Wiley.
of elbow joint kinematics on wrist speed in cricket fast bowling. Journal of Worthington, P. J., King, M. A., & Ranson, C. A. (2013). Relationships
Sports Sciences, 33, 1622–1631. doi:10.1080/02640414.2014.1003586 between fast bowling technique and ball release speed in cricket.
Palastanga, N., Field, D., & Soames, R. (2002). Anatomy and human movement: Journal of Applied Biomechanics, 29, 78–84.
Structure and function (4th ed.). Oxford: Butterworth and Heinemann. Yeadon, M. R. (1990). The simulation of aerial movement—II. A mathema-
Portus, M. R., Rosemond, C. D., & Rath, D. A. (2006). Cricket: Fast bowling arm tical inertia model of the human body. Journal of Biomechanics, 23, 67–
actions and the illegal delivery law in men’s high performance cricket 74. doi:10.1016/0021-9290(90)90370-I
matches. Sports Biomechanics, 5, 215–230. doi:10.1080/14763140608522875 Yeadon, M. R., & King, M. A. (2002). Evaluation of a torque-driven
Roca, M., Elliott, B., Alderson, J., & Foster, D. (2006). The relationship simulation model of tumbling. Journal of Applied Biomechanics, 18,
between shoulder alignment and elbow joint angle in cricket fast- 195–206.
Downloaded by [Ming Chuan University] at 21:42 23 March 2016