Professional Documents
Culture Documents
DOI 10.1007/s40279-016-0588-8
SYSTEMATIC REVIEW
123
D. J. McNamara et al.
123
Workload, Performance and Injury in Elite Fast Bowlers
recently published systematic review regarding factors SPORTDiscus, MEDLINE, Scopus, and Web of Science.
associated with non-contact injuries in adult fast bowlers These databases were searched using the combinations of
[8]. This review will distinguish itself from the recent the following two major key terms in the title and abstract:
systematic review with a primary aim of demonstrating the (1) ‘cricket’ and (2) ‘bowl’, with the terms ‘pace’, ‘fast’,
extent to which workload is related to injury and/or per- ‘medium’, and ‘seam’ connected to ‘bowl’ using ‘OR’.
formance in elite cricket fast bowlers using observational These major two search categories were combined using
and cross-sectional studies. Secondary aims were first, to ‘AND’. To remain contemporary, the search was restricted
provide an update of the physical requirements of elite fast to full-length, peer-reviewed articles written in English
bowlers since the introduction of T20 cricket, and second, after 2005 through to April 2015. All full-length papers
to provide an update on the measures of workload within were searched for additional references not identified in the
cricket fast bowling. initial search.
123
D. J. McNamara et al.
Data needed to be collected in situ, i.e. in the field, not a Quality reporting was ranked independently then dis-
laboratory. To be included, articles were required to pro- agreements on scores were discussed to a point of con-
vide data on injuries, fast bowling performance, cricket sensus among the authors.
workload, or measures of cricket workload. Articles were
excluded if they were case studies, abstracts, review arti-
cles, published in a language other than English, or non- 3 Results
peer reviewed articles. Disagreements between the inde-
pendent reviewers regarding article inclusion or exclusion The flow diagram in Fig. 3 shows the number of papers
were discussed and resolved. found, excluded (with reasons), and included. A total of
751 studies were initially retrieved from the four databases.
2.3 Data Extraction Thirty-nine studies were reviewed in full, and a further 23
articles were excluded for reasons including the following:
Data extracted included the year and country of publica- outside the scope of research (n = 15), abstract only (n = 3),
tion, the nature of participants, key dependent variables and population (n = 1), entire data collection pre-2005 (n = 3),
study design, statistical treatment, major results, key out- and not peer reviewed (n = 1). The remaining 16 articles
comes, and limitations. Reporting of outcomes included were included in the systematic review. A further article
odds ratios, relative risk, means and confidence intervals, was included after being located by another source. In total,
and effect sizes. 17 articles were included in the systematic review.
123
Workload, Performance and Injury in Elite Fast Bowlers
Excluded
n = 527
● Duplicates n = 388
● Pre 2005 n = 121
● Conference abstract n = 17
● Review articles n=1
Title and abstracts screened
n = 224
Excluded
n = 185
● Outside scope of research n = 181
● Population n=4
Excluded
n = 23
● Outside scope of research n = 15
● Abstract only n=3
● Population n=1
● Data collection pre 2005 n=3
● Not peer reviewed n=1
Articles found in other searches meeting
inclusion criteria
n=1
underreporting of measures of variability in testing. Sixty- bowling workload was linked to injury likelihood.
three percent clearly explained the variability of testing Although it appeared that 50 overs of bowling in a single
procedures within their study. week was a threshold for injury likelihood in elite fast
bowlers [4], a sharp increase in acute bowling workload
3.4 Origin of Studies Reviewed relative to the fast bowler’s chronic workload (i.e. 4-week
rolling average bowling load) also increased risk of injury
Eighty-two percent of the articles included in the review [6]. Bowling workload history appeared to have a complex
originated from Australia, 12 % were from the UK, and 6 interaction with likelihood of injury type. Tendon injuries
% from South Africa (Table 1). appeared to be associated with high acute workload and
high previous season workload. High medium-term work-
3.5 Workload and Injury load (3-months) was identified as protective. Bone injury
likelihood was increased with high medium-term and low
From the six studies on fast bowling using injury as an career workload, while joint injury likelihood was associ-
outcome [4, 6, 13–16], four counted the balls bowled as a ated with high previous season and career workloads [16].
measure of workload [4, 6, 14, 16]. In each of these studies, The remaining studies using injury as an outcome either
123
Table 1 Summary of studies included in the systematic review
Study Nationality Participants Level of Variables Methods Statistics Results Limitations Quality
(n) competition (score,
123
%)
Pyne et al. Australian 72 24 first class Bowling performance; Senior and junior fast bowlers Cohen’s ES statistic for Senior fast bowlers had Variability in 15, 94
[22] Australian ball velocity in were profiled in bowling differences between greater ball velocity, 1-legged CMJ %
cricketers; 48 bowling. performance, anthropometric senior and junior muscle mass, and bench protocol
Australian Anthropometric measures and strength cricketers. Correlation press throw. The best
junior measures: height, measures. Data were coefficient for association predictors for ball
representative mass, skinfolds, observed as a comparison between bowling velocity velocity in senior fast
players body composition between both groups and as a and anthropometric/ bowlers were static
estimates. Strength predictor of bowling strength variables jumping and arm length.
testing: single leg performance (ball velocity) The best predictors for
jumps and bar throw within each age group ball velocity in junior fast
tests bowlers were static jump,
bench throw, body mass,
% body mass and height
Duffield Australian 6 First class Body mass, heart rate, Initial measures of repeated A one-way, repeated No decrement in bowling Limited to training 15, 94
et al. [17] Australian core temperature, sprint ability and peak measures ANOVA was velocity and accuracy environment %
cricketers capillary blood bowling speed were taken. A performed to determine across repeated bowling
lactate, pH, and second session of 2 9 6-over differences between spells. No differences in
glucose. Perceptual bowling spells was bowling spells. Cohen’s heart rate, core
measures or RPE and completed to measure ES statistic was used to temperature, lactate, pH,
muscle soreness. bowling performance (speed determine practical glucose, RPE, muscle
Repeated vertical and accuracy) and run-up importance of change. soreness, and vertical
jump efforts speed. Dependent variables Pearson’s correlation co- jump across repeated
were assessed during the efficient was used to bowling spells
repeated bowling spells determine the association
between measures
Orchard Australian 129 Professional Injury: acute Prospective cohort study over T tests to compare the Bowlers who performed Irregular scheduling 13, 81
et al. [4] cricketers noncontact or a 10 years of injury history in number of overs bowled. [50 overs in a match had of match play. %
across gradual-onset Australian fast bowlers. The Risk ratios were a significantly increased Definition of
Australia bowling mechanism study compared future injury calculated to compare injury incidence in the overuse injury
risk in bowlers of high and high and low workload next 21 days compared
low workload status groups. CIs were with those bowlers who
calculated using Taylor bowled \50 overs.
series expansions Bowlers who bowled[30
overs in the second
inning of a match had a
significantly increased
injury risk per over
bowled in the next 28
days
Johnstone UK 15 Professional Anthropometric data, A cross-sectional design was Descriptive results Small to moderate effects Sample sizes were 11, 69
and Ford cricketers aerobic fitness used to profile the physical expressed as mean ± SD were observed across all limited to 9 %
[23] from the testing (multistage and athletic profiles of and 95 % CI. Magnitude variables except for bowlers and 6
English first fitness test), professional cricketers. of difference interpreted medicine ball throw and batsmen.
class county flexibility (sit and Participants were classified as using Cohen’s ES. ES press-up tests, which Reliability and
competition reach test), speed either a batsman or a bowler thresholds: 0.2–0.6 showed large ESs validity of field
(quick single, run ‘a’ for comparison (small), 0.6–1.2 based testing
3), strength (moderate), 1.2–2.0
(medicine ball throw, (large), [2.0 (very large)
press ups, jump
testing)
D. J. McNamara et al.
Table 1 continued
Study Nationality Participants Level of Variables Methods Statistics Results Limitations Quality
(n) competition (score,
%)
Orchard Australian 205 Australian first Injury (excluding Prospective cohort risk factor Two-by-two v2 test. A risk Risk of calf strain was Survey records may 13, 81
et al. [13] class fast lumbar stress study from data collected ratio was calculated for strongly associated with have missing data %
bowlers fracture) between 1998 and 2009. Risk each injury category. CIs prior lumbar stress from injuries that
ratios were calculated to were calculated using fracture injury history occurred prior to
compare the seasonal Taylor series expansions data collection.
incidence of various injuries Players may have
between bowlers with a prior had asymptomatic
history of lumbar stress pars defects that
fracture and those with no went undetected
history of lumbar stress
fracture
Orchard Australian Not stated Australian first Injury Researchers compared injury Not stated When expressing injury 7, 44 %
et al. [5] class rates from 1998 to 2009. The rates relative to days of
cricketers researchers suggested injury play, injury rates
rates should be expressed normalised for T20
relative to days of play rather cricket
than hours of play as cricket
Workload, Performance and Injury in Elite Fast Bowlers
123
Table 1 continued
Study Nationality Participants Level of Variables Methods Statistics Results Limitations Quality
(n) competition (score,
123
%)
Lombard South 10 Elite South CK, visual analogue Physiological measures were ANOVA. Post hoc testing CK and muscle pain were Indoor bowling 10, 63
et al. [24] African African fast scale muscle pain assessed as a baseline was performed using elevated 1 h and 24 h surface may not %
bowlers ratings, ROM and followed by 2 days of rest. Tukey’s multiple post-bowling. No change reflect match
isometric maximal Physiological measures were comparison test was observed in maximal requirements
strength testing then assessed the day isometric strength. Ankle
following 8 overs of bowling ROM was reduced at 24
h post-bowling
Phillips Australian 32 Australian Ball velocity, bowling Bowlers were required to bowl Effects of skill level and Elite fast bowlers Potential overlap in 12, 75
et al. [25] national, elite accuracy 30 balls at varying lengths to delivery type on each performed better in speed skill level in %
emerging and targets. Bowling measures of accuracy measure were and accuracy. National developmental
junior fast accuracy and velocity were assessed independently and emerging fast groups
bowlers recorded using a multiple mixed bowlers were able to
design ANOVA with adapt to varying bowling
repeated measures. Post length changes within the
hoc multiple comparisons same session
were performed using
Bonferroni corrected
t tests. Pearson’s product
moment correlation
coefficient was used to
find relationships
between consistency,
accuracy, and velocity
McNamara Australian 26 National level Movement patterns Fast bowlers and non-fast ES statistics were used to Fast bowlers had greater Lack of control of 14, 88
et al. [2] under 19s and measured with GPS bowlers were assessed over a analyse differences competition and physical activity %
17s cricketers technology; 7-week preparation period between groups in preparation workloads. outside structured
endocrine markers and 10-day competition endocrine, perceptual Endocrine markers training sessions
(cortisol and period well-being, and suggested that fast
testosterone); neuromuscular function bowlers were in a more
perceptual well- measures. Non- catabolic state during
being; parametric ES statistics both preparation and
neuromuscular were used for movement competition, whilst
function patterns perceptual well-being
was lower in non-fast
bowlers
Ranson UK 76 International Time-loss injury; non- Time-loss and non-time-loss Not stated 23 time-loss and 97 non- 10, 63
et al. [15] cricketers time-loss injury injuries sustained during the time-loss injuries were %
performing in 2011 Cricket World Cup recorded across 5 teams
the Cricket were recorded by medical in the 2011 Cricket
World Cup staff World Cup
D. J. McNamara et al.
Table 1 continued
Study Nationality Participants Level of Variables Methods Statistics Results Limitations Quality
(n) competition (score,
%)
Hulin et al. Australian 28 Elite Australian Session RPE. Non- Data were collected over a The likelihood of Findings demonstrated that 14, 88
[6] fast bowlers contact injury 6-year period. Workloads sustaining injury was larger increases in acute %
resulting in loss of were the total number of balls analysed using a logistic workloads were
match-time or bowled per week (external regression model associated with increased
training session workload) and the session injury risk in elite cricket
RPE (internal workload). fast bowlers
1-week data (acute
workload), together with
4-week rolling average data
(chronic workload), were
calculated for external and
internal workloads. Training-
stress balance was assessed
as negative or positive based
on acute and chronic
workloads
McNamara Australian 12 Elite national Automated bowling Participants performed a series The relationship between The bowling detection Low-intensity 12, 75
Workload, Performance and Injury in Elite Fast Bowlers
et al. [21] and detection from GPS/ of bowling, throwing, and the MinimaxX data and algorithm was shown to bowling was not %
international microtechnology fielding activities. Sensitivity direct measures of balls be sensitive in both assessed
Australian fast unit and specificity of the bowled were assessed training (99.0 %) and
bowlers automated detection of fast using the Pearson’s competition (99.5 %).
bowling events was assessed product moment Specificity was 98.1 %
against manually recorded correlation coefficient. during training and 74.0
outputs MinimaxX and direct % during competition
measure data were
compared using
independent t tests
Orchard Australian 235 Australian Fast bowling injury Workload patterns were 95 % CIs of relative risks Bowlers who bowled [50 14, 88
et al. [14] professional assessed over a 15-year were calculated using overs in a 5-day period %
cricketers period. Injury rates were Taylor series expansions had increased risk when
assessed for from 5 to 26 to assess significance at compared with those who
days after each workload p \ 0.05 level bowled \50 overs
Orchard Australian 235 Australian Fast bowling injury Workload patterns were Multivariate analysis was Tendon injuries were Diagnostic 14, 88
et al. [16] professional assessed over a 15-year conducted using binary associated with high categories chosen %
cricketers period. Injury rates were logistic regression. A acute workload and high overlapping in
assessed for injury type and forward stepwise method previous season some injury types.
relationship to workload was used, with a p value workload. High medium- Missing workload
of \0.05 required for a term workload was in some fast
risk factor to be included protective. For bone bowlers
at each step injury, high medium-term participating in
and low career workload non-Australian
were risk factors. For teams
joint injury, high
previous season and
career workloads were
risk factors
ANOVA one-way analysis of variance, CI confidence interval, CK creatine kinase, CMJ counter movement jump, CV coefficient of variation, ES effect size, GPS global positioning system, ROM range of motion, RPE
rate of perceived exertion, SD standard deviation, T20 Twenty20
123
D. J. McNamara et al.
identified a need to change the injury definition for cricket [5, 15, 16]. These studies have observed the relationship
[5], observed injury rate across a single competition period between accumulated bowling workloads and injury like-
[15], or observed injury likelihood in fast bowlers with a lihood. Despite a moderate amount of available research on
history of spinal bone stress injury [13]. the mechanisms of fast bowling injury, the inconsistencies
in definitions, units of time, and outcomes preclude
3.6 Workload and Performance meaningful comparisons. Field-based fast bowling perfor-
mance research is mostly limited to very few studies, and
No articles directly linked workload and fast bowling as such, the impact of the introduction of T20 cricket is
performance. Compared with junior elite fast bowlers, elite difficult to assess.
adult fast bowlers were identified as having greater ability
to functionally adapt their actions to improve bowling 4.1 Workload and Injury
accuracy and tasks [25]. The age differences in these
groups may imply that a high career workload has a pos- Measures of workload in cricket fast bowling have been
itive influence on bowling performance. modified to be very specific to the sport and position
demands (i.e. counting balls bowled) [6, 14] and also via
3.7 Physical Requirements of Fast Bowling more commonly used workload monitoring methods [i.e.
session rate of perceived exertion (RPE) and GPS]
In one of the three studies on the physical requirements of [2, 6, 18, 20]. These workload measures are identified as
fast bowling, physiological measures of fatigue did not internal or external loads and are often reported as a
clearly identify change in physiological status during chronic (fitness) and acute (fatigue) load [6]. Acute load
cricket training and simulated match play [2]. Fast bowlers (often a 7-day workload) and chronic load (often a 28-day
were also able to maintain bowling performance during rolling average) [6] allow investigation and management of
these same cricket training and simulated match-play drills the complex nature of fast bowling preparedness and injury
[17], despite markers of muscle damage increasing over likelihood.
bowling spells of up to eight overs [24]. These findings Results from this review suggest that injury risk
suggest that elite fast bowlers were prepared for acute increases significantly for fast bowlers when acute load
bowling workloads. High-performing fast bowlers had exceeds chronic load, particularly in the week following
greater body mass and upper-body strength than junior this ‘spike’ in acute load [6]. When chronic loads of fast
representative fast bowlers [22]. bowlers were systematically increased, injury likelihood
decreased [6]. Complexity surrounds systematically
3.8 Measures of Workload increasing chronic loads in elite fast bowlers, particularly
for players competing in all forms of competition. Despite
In most (80 %) of the five studies using GPS and restrictions on match bowling loads in 50-over (ten-over
microtechnology, workload differences were identified restriction) and T20 (four-over restriction) cricket, appro-
between player roles, competition type, or competition and priate preparation for upcoming multi-day competitions
training [2, 18–20]. Fast bowlers had a greater workload with greater workload requirements may be compromised.
than other player types, and bowling in limited-over cricket Fast bowling generates at least some level of injury
generated greater sprinting per hour of play than multi-day resilience [3]. However, rules restricting substitutions in
cricket [2, 17–20]. Training load was often greater than the match play can expose fast bowlers to unplanned higher
competition requirements of cricket match play in elite bowling workloads in multi-day competition. Bowlers who
players [19]. bowled more than 50 overs in a match had a significantly
increased risk of injury in the following 21 days [4, 14].
Career workload, medium-term workload, and acute
4 Discussion workload all need to be considered in the management of
fast bowlers [16].
In this systematic review, 17 studies were available to The most commonly reported injuries in fast bowlers
update the applied fast bowling literature since the intro- include lumbar spine bone stress-related injuries. These are
duction of T20 cricket. Demands of cricket match play a concern to fast bowlers as they are difficult to predict,
have been re-defined via three varying formats played at have risk factors that can be unique to individuals, take
the elite level, often by the same athletes. A major com- considerable time to rehabilitate, and can lead to other soft
ponent of cricket match play is fast bowling. Fast bowling tissue injury [13].
has often been associated with greater training and match- The growth in T20 cricket across the world over the past
play bowling workloads [2, 18–20] and injury likelihood 10 years has changed the commercial landscape of cricket.
123
Workload, Performance and Injury in Elite Fast Bowlers
This commercial growth in T20 cricket has likely sur- settings, in spite of the muscle damage and soreness
passed the output of cricket research in peer-reviewed lit- associated with fast bowling [17, 24].
erature. The growth of T20 cricket has resulted in Cricket performance in this review was limited to
recommendations for injury definition changes in cricket training settings and cannot advance the understanding of
[5]. The range in match duration across cricket formats (4 h the relationship between workload and/or physical prepa-
to 5 days) has skewed injury rates reported relative to hours ration, and competition performance. A paucity of studies
of play [5]. Reporting injury relative to days of play has addressing cricket competition performance exists.
normalised these injury rates [5]. Re-defining injury has
also been considered as many injuries are likely to manifest 4.3 Physical Requirements of Fast Bowling
during preparation phases.
Match-play and scheduling variability has provided a Elite fast bowlers demonstrated the ability to maintain
unique situation for cricket researchers in comparison with bowling performance across repeated bowling spells in a
other team sports. T20 cricket has likely influenced the training setting [17]. The influence of environmental factors
preparation strategies used by coaches, support staff, and on performance and fatigue over repeated bowling spells
cricketers. Competition is a key contributor to the cumu- remains under-researched [17]. Investigations into the
lative load of fast bowlers and significantly influences the physical characteristics of fast bowlers are limited to com-
preparation for upcoming competition. The lower load of parisons of general strength and anthropometry assessments
one competition type might not adequately prepare the fast between senior and junior fast bowlers [22], and a descriptive
bowler for the higher load of upcoming competitions in study of a fast bowling cohort [23]. The findings of this
other formats. Therefore, there may be a reliance on review provide only limited recommendations to coaching
increasing training load to adequately prepare the fast staff on the physical requirements of elite fast bowling.
bowler. Rethinking the impact of accumulative load is
required for understanding unloading or overloading, using 4.4 Measures of Workload
current methodologies beyond ball counting and across
competitions of varying formats. In summary, measuring Typically, GPS devices have been used in cricket to identify
workload, injury, and performance in elite fast bowlers the different activity profiles across competition types.
across multi-day, 50-over, and T20 competitions is Shorter formats of cricket match play have been associated
complex. with greater playing intensity [20]. Compared with other
types of bowlers and field positions, fast bowlers not only
4.2 Workload and Performance experienced greater match-play demands [2, 18, 19], but also
had greater activity profiles during physical preparation [2].
Typically, successful team performance in cricket is mea- Greater match and physical preparation workloads in fast
sured by the number of runs scored in comparison to the bowlers differentially influenced the neuromuscular, endo-
opposition, and successful fast bowling in cricket is mea- crine, and perceptual fatigue responses compared with non-
sured via how many wickets the bowlers can take for their fast bowlers [2]. Fast bowlers had greater cortisol and lower
team and the number of runs to which they can restrict the testosterone concentrations than non-fast bowlers, while
opposition [26]. However, bowling velocity and accuracy perceived well-being was also greater in fast bowlers than
are key components of fast bowling success. Elite fast non-fast bowlers. Differences in neuromuscular function
bowlers have been found to bowl only a mean 5.6 km h-1 (measured via a countermovement jump) remained unclear
faster than junior participants, but importantly, perform [2]. Nonetheless, it is likely that fast bowlers were generally
with greater accuracy [25]. Success in fast bowling in well prepared for the workload requirements of fast bowling
professional teams may have other characteristics and and cricket match play.
predictors. Along with the GPS technology, the microtechnology
The length of cricket match play often requires fast (i.e. tri-axial accelerometer and gyroscope) housed within
bowlers to perform over extended periods. Fast bowlers the GPS unit has successfully been integrated to automate
must have the physical capacity to achieve this perfor- the monitoring of bowling workloads [21]. Automated
mance requirement and the injury resilience to prevent the bowling workload monitoring will likely increase accuracy
prevalent chronic injuries unique to fast bowling. Elite fast of fast bowling workload measures. It is postulated that
bowlers have greater muscle mass and strength character- variability in bowling intensity arises from various training
istics than junior participants and other team peers [22, 23]. and competition workloads. Including all balls bowled in
Typically, elite fast bowlers are physically capable of this workload definition may not provide an accurate
maintaining bowling performance over acute workload measure of the intensity of the workload [27]. In
123
D. J. McNamara et al.
conjunction with the current workload methodology of fast bowlers, and advance the understanding of varying
counting balls bowled, microtechnology may provide an match-play demands on injury likelihood.
opportunity to gain insight into bowling intensity via
Compliance with Ethical Standards
accelerometers and gyroscopes housed within wearable
tracking devices. The relatively large number of citations Funding No sources of funding were used to assist in the preparation
recorded in articles using GPS or wearable technology in of this article.
cricket may further highlight the direction and pace of
cricket workload monitoring trends. Conflict of interest Dean McNamara, Tim Gabbett, and Geraldine
Naughton declare that they have no conflicts of interest relevant to the
content of this review.
4.5 Limitations
123
Workload, Performance and Injury in Elite Fast Bowlers
17. Duffield R, Carney M, Karppinen S. Physiological responses and 23. Johnstone JA, Ford PA. Physiologic profile of professional
bowling performance during repeated spells of medium-fast cricketers. J Strength Cond Res. 2010;24(11):2900–7.
bowling. J Sports Sci. 2009;27(1):27–35. 24. Lombard WP, Muir GA, McKune AJ. Minimal changes in indi-
18. Petersen CJ, Pyne D, Dawson B, et al. Movement patterns in rect markers of muscle damage after an acute bout of indoor pre-
cricket vary by both position and game format. J Sports Sci. season fast bowling. S Afr J Res Sport Phys Educ R.
2010;28(1):45–52. 2012;34(2):105–14.
19. Petersen CJ, Pyne DB, Dawson BT, et al. Comparison of training 25. Phillips E, Portus M, Davids K, et al. Performance accuracy and
and game demands of national level cricketers. J Strength Cond functional variability in elite and developing fast bowlers. J Sci
Res. 2011;25(5):1306–11. Med Sport. 2012;15(2):182–8.
20. Petersen CJ, Pyne DB, Portus MR, et al. Comparison of player 26. Petersen C, Pyne DB, Portus MJ, et al. Analysis of Twenty/20
movement patterns between 1-day and test cricket. J Strength Cricket performance during the 2008 Indian Premier League. Int
Cond Res. 2011;25(5):1368–73. J Perform Anal Sport. 2008;8(3):63–9.
21. McNamara DJ, Gabbett TJ, Chapman P, et al. The validity of 27. McNamara DJ, Gabbett TJ, Naughton G, et al. How submarine
microsensors to automatically detect bowling events and counts and guided missile technology can help reduce injury and
in cricket fast bowlers. Int J Sports Physiol Perform. improve performance in cricket fast bowlers. Br J Sports Med.
2015;10(1):71–5. 2016;. doi:10.1136/bjsports-2015-095935.
22. Pyne DB, Duthie GM, Saunders PU, et al. Anthropometric and
strength correlates of fast bowling speed in junior and senior
cricketers. J Strength Cond Res. 2006;20(3):620–6.
123