DIFFERENCES AND CHANGES IN THE PHYSICAL

CHARACTERISTICS OF PROFESSIONAL AND AMATEUR

RUGBY UNION PLAYERS
DANIEL J. SMART, WILL G. HOPKINS, AND NICHOLAS D. GILL
Sport Performance Research Institute New Zealand, AUT University, Auckland, New Zealand

ABSTRACT training loads because of regular high-intensity matches and

Smart, DJ, Hopkins, WG, and Gill, ND. Differences and changes greater travel involved in the Super Rugby competition.
in the physical characteristics of professional and amateur rugby KEY WORDS body composition, strength, speed, repeated
union players. J Strength Cond Res 27(11): 3033–3044, 2013 sprint ability, mixed model
—Numerous studies have highlighted differences between play-
ing levels and positions in rugby union; however, few studies INTRODUCTION

R
have investigated longitudinal progressions of body composition
ugby union is a field-based team sport that requires
and physical performance. Between-player differences and
a diverse range of physical attributes to tolerate
within-player changes in body composition, strength, power, a large amount of physical contact and numerous
speed, and repeated sprint ability, from 1,161 New Zealand maximal sprints (12). Since the introduction of
rugby union players from 2004 to 2007, were estimated using professionalism in 1995, the characteristics of speed, strength,
a mixed modeling procedure. Props had the highest mass, per- power, and body composition of players has evolved rapidly,
cent body fat, strength, and slowest speed times compared with and as a consequence, the speed and physicality of matches
the other positions, whereas outside backs had the fastest has increased (12,35). The measurement of players’ physical
speed time and lowest percent body fat. For most measures, characteristics has highlighted position-specific attributes.
there were small-to-moderate differences (range, 1.1–14%) Forwards are involved in more rucks, mauls, line outs, and
between players selected and not selected for provincial teams scrums, which requires greater mass, height, strength, and
and small-to-large differences (range, 1.8–15%) between pro- power to be successful (9,11). In contrast, the backs primary
vincial and Super Rugby (professional) players. The faster 20-m role in beating the opposition in open play requires a combi-
sprint times in international compared with Super Rugby players nation of speed, acceleration, and agility (12,28,29). Differen-
ces between playing levels have also been reported. For
was small in magnitude for both the forwards (1.9%) and backs
example, senior club players possess greater height, mass,
(2.2%). The average annual improvements were small to mod-
speed, strength, and aerobic fitness compared with their
erate for strength (range, 2.1–15%) and small for repeated
lower-level age group counterparts (29).
sprint ability within the lower playing levels (;1.5%). Small in- Few longitudinal studies exist exploring the progression of
creases occurred in lower body strength (;7.0%) as players individual’s and teams’ physical characteristics. Olds (26)
moved from Super Rugby to provincial competition. Small de- used historical data to track the evolution of physique in
creases in sprint time (;1.6%) and small increases in strength male rugby union players from 1905 to 1999. It was shown
(;6.3%) occurred as players moved from Super Rugby to mid- that the body mass index had increased at a rate of 3–4 times
year international competition. The differences between levels in faster in rugby union players during the last 25 years com-
performance provide level-specific characteristics from Super pared with the rest of the century. In addition, a more recent
Rugby and below, but international players may be selected study showed rapid increases in mass (;10%) since the
because of greater skill and experience. Changes in physical inception of the professional era (30).
performance between competitions may be a result of reduced To track more specific individual player changes, Duthie
et al. (11) modeled the fat-free mass of professional Super
Rugby players over a period of 5 years, which included the
Super Rugby (the Southern Hemisphere premier profes-
Address correspondence to Dr. Daniel J. Smart, daniel.smart@hpsnz.org.nz. sional competition played across Australia, New Zealand
27(11)/3033–3044 and South Africa) and regional club competitions at different
Journal of Strength and Conditioning Research times of the year. It was reported that there was a decrease in
Ó 2013 National Strength and Conditioning Association the proportion of the fat-free mass of players that occurred

VOLUME 27 | NUMBER 11 | NOVEMBER 2013 | 3033

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 Physical Characteristics in Rugby

primarily during the club competition. More recently, Argus coaches, on all regional representative and professional play-
et al. (2) tracked the changes in strength and power over the ers in New Zealand. Data were only entered into the database
period of a Super Rugby season, reporting small increases if the test was performed using the stipulated New Zealand
(8.5%) in box squat 1 repetition maximum (1RM) and small Rugby Union testing procedures (see below). Although these
decreases (3.4%) in jump squat peak power. procedures were provided to standardize the testing, the
Traditionally, athletes undergo fitness assessments that are interpretation of the protocols and test performance (nutri-
used to determine the current level of fitness, motivate tion, hydration, and time of day) may be different between
individuals, and to assist in future program prescription. individuals. Nonetheless, players were typically tested by the
Typically, these test data are used for the short term to same individual strength and conditioning coach,
evaluate the success of a preseason program or a current thus minimizing the variation associated with multiple testers
phase of conditioning. Previous longitudinal studies have (20). No familiarization trials were performed as players were
shown progressions of physique and the effect different familiar with testing protocols and had tested on numerous
phases of a year have upon measures of body composition. occasions before the period of this study.
However, the long-term monitoring of individual players’ Data on body composition, strength, power, speed, and
physical performance, such as strength, power, and speed, repeated sprint ability from the beginning of the Super
has received less attention and requires further investigation. Rugby preseason 2004 (1 December, 2003) through to the
Because of a lack of knowledge in the area of long-term conclusion of the international end of year tour (World Cup)
athletic development in rugby union, there is a need to 2007 (25 November, 2007) was downloaded from the
understand the between-player differences and within-player database. Each of the 4 years included in the analysis were
changes over long periods of time. divided into specific phases of the year, which included the
Therefore, the primary purpose of this study was to preseason training periods and the competition phases of the
analyze performance test data of New Zealand rugby union respective competitions. The professional Super Rugby
players from 2004 to 2007, to determine differences between season running from the beginning of December the pre-
playing positions, playing level, and year of fitness test; and vious year through to May; the midyear international
changes within players as they moved between different competition from May to August; the semiprofessional
competitions played during the year. It was hypothesized national provincial competition (in which Super Rugby
that (a) the differences between playing positions would be players also participate) from July to October; and the
similar to those already reported in literature; (b) players of international end of year competition from October to the
higher playing levels would perform better than players of end of November.
lower levels; (c) players would show small annual improve- Informed consent for each player was obtained through
ments in physical performance; and (d) players would the player registration form each player must sign at the
perform better during higher-level competitions. beginning of each rugby season. The form stipulates that any
data collected from the player may be used at the discretion
METHODS of the New Zealand Rugby Union for research or data
analysis purposes. The study was approved by the Auckland
Experimental Approach to the Problem
University of Technology Ethics Committee.
A large battery of fitness tests that are deemed important for
the physical preparation of rugby union players are per- Procedure
formed on all registered players in New Zealand at pro- Body Composition. Anthropometric measurements included
vincial level and above. Performance test data that have been body mass and sum of 8 skinfolds (bicep, triceps, subscapular,
determined through these fitness tests, performed under abdominal, supraspinale, iliac crest, front thigh, and medial
standardized testing protocols, are entered onto a national calf). Body mass was measured on calibrated scales, and each
database. With this large amount of data available, the New skinfold site was located and measured as per the Interna-
Zealand Rugby Union was able to commission research into tional Society for the Advancement of Kinanthropometry
the progressions and differences between annual means, (ISAK) guidelines (25). Percentage body fat was calculated
playing level, and competition phases. To deal with the large from estimated body density (36) using the equation derived
amount of repeated measures on players, the differences from Siri (31). Fat-free mass was calculated from the player’s
between and changes within players were determined body mass and calculated body fat (fat-free mass = body
through appropriate statistical modeling. mass 2 (body mass 3 percentage body fat/100)) (32). Data
were only included if all technical errors of measurement were
Data Source
below the upper limits recommended by Perini et al. (27).
Performance test data for 1,161 players were downloaded
from the Performance Profiler Database (NZRU Version 7; Strength and Power. 1RM was calculated for a series of
Profiler Corporation, Dunedin, New Zealand). The Perfor- resistance training exercises from a 2–6 repetition maximum
mance Profiler Database contains results from performance lift using the formula derived by Landers (22). The strength
tests, conducted by various strength and conditioning exercises included bench press, box squat, back squat, and
the TM

3034 Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 the TM

Journal of Strength and Conditioning Research | www.nsca.com

chin-ups, whereas the power clean was used to indicate full 2 efforts was performed after at least 2-minute rest from the
body power. Each exercise was assessed for correct tech- previous repetition. The CV for speed has been shown to be
nique by a trained strength coach and only repetitions per- 0.9–3% (34).
formed unassisted with correct technique were recorded.
When performing the bench press, the feet were to remain Repeated Sprint Ability. Repeated sprint ability was tested
in contact with the floor and the buttocks and lower back using the Rugby-Specific Repeated-Speed (RS2) test. The
had to remain in contact with the bench throughout the lift. first component of the test is a measure of speed and is
During the lift, the bar was to be lowered to the chest (with performed as described above. The speed component allows
elbows at approximately 908 and not bouncing off the chest) a comparison of effort to be made between the sprints per-
and returned to the start position where elbows were to be formed during the repeated sprint component of the RS2
fully extended but not locked. Each player used a self- test. Five minutes after completing the speed component,
selected hand position. The back squat required the player the repeated sprint component was performed. The repeated
to descend in a controlled manner until the top of the thighs sprint component consists of 3 sets of 3 or 4 individual
were parallel with the floor before returning to the standing sprints performed maximally at set time intervals. Each set
position. The box squat was performed in a similar manner of sprints is separated by periods of standardized work
however the player was instructed to pause briefly in the where the players jog with a weighted bag (PowerBag; SPSS,
seated position on a box where the thighs were parallel with Christchurch, New Zealand) over their shoulders and per-
the floor. Players used a self-selected foot position, and form down and ups (get down off feet into a prone position
powerlifting belts were not used during the lifts. When per- on the ground—chest and chin was required to touch the
forming the chin-ups, a reverse underhand grip (palms facing ground—and then return to feet), also at set time intervals
toward face) was used. Players were instructed to start from (Figure 1). Players repeated sprints were measured using
a stationary position with arms fully extended and complete electronic timing gates over the same distance as speed
a repetition with the chin moving over the bar (2,5). The (30 m for backs and 20 m for forwards and half backs); how-
power clean required the player to set up in a crotched ever, only the time to complete the total distance was recorded.
position over the bar on the floor with fully extended arms. Two groups of 3 forwards (6 total) or 4 backs (8 total)
From this position, the player was instructed to thrust were able to perform the repeated sprint component at
upward in a triple extension movement, pulling the barbell 1 time. The master timer started a stopwatch and sent
upward into the catch position on the front of the shoulders individual players off at 10-second intervals. The master
with elbows forward (4). Between repetitions, the player timer started the players by counting down the time left
must have stayed connected with the bar. The coefficients before the start of the next repetition of work from 5-seconds
of variation (CV) for similar strength testing protocols within (i.e., 5; 4; 3; 2; 1; GO!). During the periods of standardized
professional rugby union players have been shown to be work, the master timer was required to countdown for all the
approximately 4.5% (2). players, as in some instances, all subjects were performing
some form of work at the same time.
Speed. All sprints (both speed and repeated sprints) were
performed on grass; however, a synthetic grass mat covering Forwards. The forwards (including the half backs) were
1.5 m behind and 3.5 m in front of the first timing gate required to sprint 4 times over 20 m, sprinting through the
(securely pegged at each corner) was laid to assist with timing gates, decelerating to a cone a further 10 m away.
traction. All sprints were performed in footwear that was They then jogged back toward the timing gates (total
appropriate for the conditions and those used during rugby distance decelerating/jogging = 20 m), and upon reaching
matches (moulded soles for firm and hard ground, football the gates walked back to the start line outside the running
boots for softer ground). The players were instructed to lane (total distance walking = 20 m). Each sprint repetition
sprint maximally for every repetition within the lane formed was performed on a 30-second turnaround. After completing
by the Swift (Swift Performance Equipment, NSW, Austral- 4 sprints (at master time 2 minutes), the player moved to the
ia) or Smart Speed (Fusion Sport, Queensland, Australia) side of the running lane and performed the standardized
electronic timing gates, which was approximately 2-m wide. work. The player was required to pick up the 30-kg Power-
Players started each sprint with their foot on a line 50 cm Bag and place it on their shoulder or behind their neck. On
from the light beam of the first timing gate, in a stationary the GO command, they had 10 seconds to carry it 20 m and
upright position, with no rocking back or forth before drop the PowerBag at the end. On a 10-second turnaround,
starting. the player performed a down and up (get down off feet into
Each player performed 2 repetitions over 20 m for a prone position on the ground—chest must touch the
forwards and half backs and 30 m for backs. For each PowerBag—and then return to feet) before picking up the
repetition, the time to complete the total distance (20 or 30 PowerBag and jogging with it for 20 m back to the start line.
m) and the time to cover the first 10 m of each sprint was The players were to keep the bag on their shoulders before
recorded, with the fastest overall time recorded. Each of the repeating the up and back shuttle. The standardized work

VOLUME 27 | NUMBER 11 | NOVEMBER 2013 | 3035

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 Physical Characteristics in Rugby

Figure 1. The repeated sprint component of the Rugby-Specific Repeated-Speed (RS2) test (A). Three sets of repeated sprinting, jogging, and walking 20 m
for the forwards (B) and 30 m for the backs (C) are performed at set time intervals and interspersed with PowerBag shuttles using a 30-kg weighted PowerBag,
also performed at set time intervals (D).

sequence was performed a total of 5 times. After a 20-second Backs. The backs (not including the half backs) were
rest (at master time 4 minutes), the player repeated the pre- required to sprint 3 times over 30 m, sprinting through the
vious sprint and standardized work protocol. Upon comple- timing gates decelerating to a cone a further 15 m away.
tion of the second period of standardized work, the players They then jogged back toward the timing gates (total
completed the final set of 4 sprints. A total of 12 sprints was distance decelerating/jogging = 30 m) and upon reaching
therefore performed over a period of 9 minutes 30 seconds. the gates walked back to the start line outside the running
the TM

3036 Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 the TM

Journal of Strength and Conditioning Research | www.nsca.com

lane (total distance walking = 30 m). Each sprint repetition confidence intervals, with a 99% level chosen to reduce
was performed on a 40-second turnaround. Upon complet- the error rate for clear outcomes. Differences between and
ing 3 sprints (at master time 2 minutes), the player moved to changes within players were standardized and assessed
the side of the running lane and performed the standardized for magnitude using a modified Cohen scale: ,0.2 = trivial,
work shuttles. The PowerBag sequence explained in the for- 0.2–0.59 = small, 0.6–1.19 = moderate, 1.2–1.99 = large,
wards protocol was performed 4 times. After a 30-second .2.0 = very large (21).
rest (at master time 4 minutes), the player repeated the pre-
vious sprint and standardized work protocol. After complet- RESULTS
ing the second period of standardized work, the player Differences Between Positions
completed the final set of 3 sprints. A total of 9 sprints were The mean performance and anthropometrical outcomes for
therefore performed over a period of 9 minutes 20 seconds. each positional group are summarized in Table 1. Mean
Performance Variables for the RS2 Test. Meas- performance in speed, body composition, strength, and
ures of performance derived from the RS2 test were: the power showed the expected differences between positional
mean time per sprint, fatigue (calculated as a percent change groups. Outside backs had the fastest 10, 20, and 30 m sprint
in sprint time predicted from the linearized change derived time, and their differences between other positions ranged
from all sprints performed), and mean of 12 vs. 20 m for from small (0.02 seconds with centers over 10 m) to very
forwards and mean of 9 vs. 30 m for backs (the percent large (0.32 seconds with props over 20 m). Trends showed
difference between the mean time per sprint and the corre- a linear decrease in speed time as the positional number
sponding sprint time performed during the first component increased (props jersey number 1, through to fullback [out-
of the RS2 test). The CV’s for the performance variables of side back] jersey number 15). Props were the heaviest and
the RS2 have been shown to be 0.8–2.3% (33,34). had the largest skinfold thickness, percent body fat, and fat-
free mass. In a similar trend to speed, skinfold thickness and
Statistical Analyses percent body fat linearly decreased as positional number
Players were identified and grouped according to their increased. Differences between positions in strength and
playing position (props, hookers, locks, loose forwards, inside power were highly varied and dependent upon the exercise
backs, centers, outside backs) and competitive level at time of performed. Trivial to small differences in bench press (9 kg),
test (in order of highest to lowest playing level: international, box squat (0 kg), and back squat (9 kg) 1RM were found
Super Rugby, provincial, and players not selected for pro- between the front row positions (props and hookers). How-
vincial). The mixed modeling procedure (Proc Mixed) in the ever, when the front row positions were compared with
Statistical Analysis System (Version 9.2; SAS Institute, Cary, other positions, differences were small to large. Differences
NC, USA) was used to determine the differences between between the inside backs (half backs) and the forwards in the
positions, playing level, and year of fitness test. The differ- forwards RS2 test ranged from small (0.9% with loose for-
ences between levels were adjusted to 2007, the last year of wards for mean of 12 vs. fastest) to very large (0.33 seconds
the data and year with the most data. The fixed effects were with props for mean of 12 sprints). Small differences in the
the highest playing level achieved in each year and the backs RS2 test occurred between the inside backs and out-
interaction of this variable with the year of test (as a linear side backs (all 3 RS2 test performance variables), centers and
numeric variable, to estimate a different progression within inside backs (backs fatigue), and centers and outside backs
each level). The random effects were the identity of the (mean of 9 sprints).
player (to estimate consistent differences between players),
the interaction of player identity and year (to estimate within- Differences Between Playing Levels
player variation between years), and the residual (represent- The percent differences between playing levels for forwards
ing variation within player for any tests repeated within and backs are shown in Table 2. The differences between
a year). A similar model was used in the analysis of the Super Rugby and provincial players were dependent upon
changes within players between different competition periods the positional group. The forwards had small (7.7%; 99%
during the same year. The fixed effects were the year of confidence limits 6 8.2% for back squat 1RM) to moderate
competition (as a nominal variable, to adjust for any (13.3%; 64.8% for bench press 1RM) differences in all
consistent changes between years) and the level of compe- strength and power measures, whereas the backs had small
tition at the time of the test (to estimate consistent changes as (1.9 6 1.5% for 30-m sprint) to moderate (4.5%; 61.7% for
players move between competition periods within the same 20-m sprint) differences in all speed measures. Few perfor-
year). The random effects were the same as the previous mance variables showed clear differences between Super
model. Mechanistic inferences about magnitudes of effects Rugby and international players. The only difference of sub-
were based on acceptable uncertainty in the effect estimates stantial magnitude within both the forwards and backs
(21). All results were expressed as single values to represent between international and Super Rugby players was the
the mean within-athlete changes and between-athlete differ- small difference in 20-m sprint (1.9%; 62.1% for forwards
ences over the 4 year period. Uncertainty was defined by and 2.2%; 62.2% for backs).

VOLUME 27 | NUMBER 11 | NOVEMBER 2013 | 3037

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 3038

Physical Characteristics in Rugby

Journal of Strength and Conditioning Research
the

TABLE 1. Mean 6 coefficient of variation (%) of physical performance and anthropometric tests in rugby union players, calculated with equal contribution from
level of player (amateur, semiprofessional, and professional) and year of test (2004–2007), separated into positional groups.

Props, Hookers, Locks, Loose forwards, Insides, Centers, Outside backs,

n = 143 n = 85 n = 121 n = 207 n = 212 n = 58 n = 172

10-m sprint (s) 1.85 6 4.7 1.81 6 4.1 1.79 6 4.7 1.76 6 4.5 1.72 6 4.0 1.70 6 4.0 1.68 6 4.4
20-m sprint (s) 3.21 6 4.4 3.14 6 3.7 3.13 6 4.2 3.06 6 4.4 2.96 6 3.5 2.95 6 4.6 2.89 6 3.3
30-m sprint (s) 4.14 6 4.1 4.12 6 4.2 4.11 6 3.9
Body mass (kg) 113.5 6 8.1 104.9 6 6.4 109.4 6 7.6 101.6 6 7.9 88.8 6 9.2 94.1 6 6.3 89.2 6 9.0
Skinfold thickness† (mm) 114 6 26 102 6 26 88 6 29 84 6 30 74 6 35 74 6 26 65 6 25
6 6 6 6 6 6 6
TM

Percent body fat (%) 16.1 26 14.5 25 12.7 29 12.1 28 10.7 32 10.6 24 9.4 23
Fat-free mass (kg) 94.4 6 7.9 88.5 6 6.2 95.0 6 6.4 88.9 6 7.0 78.8 6 8.2 83.9 6 6.3 80.8 6 8.6
Bench press 1RM (kg) 133 6 18 124 6 17 121 6 17 119 6 16 111 6 16 113 6 15 109 6 16
Box squat 1RM (kg) 185 6 19 185 6 25 157 6 21 169 6 26 155 6 20 163 6 23 157 6 20
Back squat 1RM (kg) 184 6 19 175 6 20 141 6 21 161 6 21 141 6 20 151 6 17 145 6 24
Chin-ups 1RM (kg) 140 6 10 137 6 9 139 6 11 132 6 11 123 6 11 127 6 9 123 6 11
Power-clean 1RM (kg) 102 6 14 101 6 14 103 6 17 98 6 18 91 6 16 93 6 15 91 6 20
Mean of 12 sprintsz (s) 3.44 6 5.5 3.35 6 4.0 3.31 6 4.5 3.23 6 4.6 3.11 6 3.7*
Mean of 9 sprintsz (s) 4.30 6 4.0 4.33 6 4.9 4.25 6 4.0
Forwards fatigue§k (%) 3.9 6 5.1 4.6 6 3.9 3.0 6 4.1 3.8 6 4.7 2.3 6 3.6*
Backs fatigue§k (%) 3.1 6 3.7 4.9 6 5.1 4.1 6 4.6
Mean of 12 vs. 20 mk¶ (%) 7.4 6 4.1 7.0 6 3.1 6.4 6 2.9 5.9 6 3.3 4.8 6 3.1*
Mean of 9 vs. 30 mk¶ (%) 4.8 6 2.8 5.9 6 3.8 5.9 6 3.4

*Halfbacks performed forwards protocol.

†Sum of 8 skinfolds.
zMean time for sprints performed in Rugby-Specific Repeated-Speed test.
§Fatigue for the Rugby-Specific Repeated-Speed test.
kData expressed as mean 6 SD.
¶The percent difference in time between the mean time and sprint time in the Rugby-Specific Repeated-Speed test.

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 TABLE 2. Mean differences (%); 699% confidence limits in physical performance and anthropometry between rugby union players of different playing levels in
the 2007 year within forwards and backs positional groups.*

Forwards Backs

Provincial, not Super Rugby, International, International, Provincial, not Super Rugby, International, International,
selected provincial Super Rugby provincial selected provincial Super Rugby provincial

10-m sprint 0.8; 61.2† 20.7; 61.4† 20.6; 62.3 22.1; 62.3z 22.2; 61.2z 23.7; 61.4§ 21.9; 62.0z 23.4; 62.0§
20-m sprint 1.1; 61.2z 0.1; 61.4† 21.9; 62.1z 22.1; 62.2z 22.4; 61.7§ 24.5; 61.7§ 22.2; 62.2z 24.3; 62.2§
30-m sprint 22.1; 61.4z 21.9; 61.5z 22.8; 62.1§ 22.6; 62.2§
Body mass 2.3; 61.6z 3.6; 61.9z 0.2; 62.0 1.4; 62.4† 2.3; 61.6z 2.9; 61.9z 20.5; 62.0† 0.1; 62.4
Skinfold 22.7; 65.9† 27.7; 66.4 20.9; 68.2 25.9; 68.7z 29.7; 65.5z 211.3; 66.2z 26.5; 67.7z 28.2; 68.7z
thickness
Percent body 23.6; 65.6† 28.8; 66.1z 21.2; 67.8 26.6; 68.3z 29.6; 65.3z 212.6; 65.8z 26.1; 67.4z 29.1; 68.1z
fat
Fat-free mass 3.1; 61.4z 5.2; 61.7§ 0.0; 61.8† 2.1; 62.1z 3.1; 61.5z 4.1; 61.7 0.0; 61.7† 1.0; 62.1†
Bench press 8.5; 63.9z 13.3; 64.8§ 0.4; 66.6 4.9; 67.4z 11.3; 64.4§ 13.0; 65.3§ 20.4; 67.2 1.2; 67.7
1RM
Box squat 6.2; 69.6z 10.5; 610.6z 7.5; 618.0 11.8; 619.4 8.9; 69.9z 1.0; 69.2 21.8; 615.4 28.9; 614.6
1RM
Back squat 11.1; 67.4z 7.7; 68.2z 1.4; 612.5 21.7; 612.1 14.0; 67.9§ 6.7; 68.9z 20.2; 616.1 26.6; 614.5
1RM
Chin-ups 1RM 8.2; 63.1§ 10.3; 63.5§ 6.3; 67.7§ 8.3; 67.9§ 4.9; 63.4z 5.6; 64.0z 0.7; 66.8 1.4; 67.0
Power-clean 7.2; 65.9z 12.0; 66.5§ 7.7; 69.6z 12.5; 610.3§ 9.9; 67.7z 15.2; 68.6§ 1.5; 611.2 6.4; 611.9

Journal of Strength and Conditioning Research

the
1RM
Mean of 12 0.2; 61.8 20.7; 62.0 21.6; 62.8 22.6; 62.8z 23.5; 64.1§ 27.0; 64.3k 23.1; 66.0 26.7; 65.9k
sprints
Mean of 9 21.7; 61.1z 21.8; 61.9z 23.2; 62.5§ 23.3; 62.6§
VOLUME 27 | NUMBER 11 | NOVEMBER 2013 |

sprints
Forwards 21.8; 61.5z 22.0; 61.6z 1.1; 62.3 0.9; 62.3 23.2; 63.1§ 24.2; 63.3§ 20.8; 64.8 21.8; 64.7
fatigue
Backs fatigue 20.1; 61.9 20.4; 62.0 2.4; 63.1z 2.0; 63.0
Mean of 12 vs. 20.4; 61.2 20.7; 61.3† 0.0; 61.9 20.2; 61.9 23.0; 62.6§ 24.4; 62.8k 0.6; 64.1 20.8; 64.0
20 m
Mean of 9 vs. 0.7; 61.4 0.2; 61.6 0.4; 62.2 20.1; 62.2
30 m

*A negative value indicates a greater value for the lower playing level (stated second).
†Trivial difference.
zSmall difference.
§Moderate difference.

TM
kLarge difference; all other differences were unclear.

| www.nsca.com
3039

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 3040

Physical Characteristics in Rugby

Journal of Strength and Conditioning Research
the

TABLE 3. Average annual changes (%); 699% confidence limits in physical performance or anthropometry in rugby union players between 2004 and 2007,
within different playing levels for forwards and backs.

Forwards Backs

Not selected Provincial Super Rugby International Not selected Provincial Super Rugby International

10-m sprint* 20.1; 61.0† 1.4; 61.1z 20.6; 61.1† 20.9; 61.9 1.0; 61.1z 1.4; 61.1z 0.4; 61.0† 21.4; 61.6z
20-m sprint* 20.6; 61.1† 1.5; 61.1z 0.1; 61.0† 20.9; 61.8† 1.5; 61.9z 1.3; 62.0z 0.4; 61.5 21.0; 61.8
30-m sprint* 1.1; 61.5z 0.8; 61.1z 1.2; 61.0z 21.9; 61.6z
Body mass 1.1; 60.8† 1.0; 60.7† 0.7; 60.6† 0.5; 60.8† 0.2; 60.7† 0.9; 60.8† 0.9; 60.6† 0.4; 60.7†
Skinfold thickness 2.2; 63.0† 3.2; 63.1† 0.4; 62.2† 20.2; 63.3† 0.8; 63.0† 1.5; 63.3† 1.2; 62.5† 21.2; 62.9†
TM

Percent body fat 2.5; 62.9† 3.5; 63.0† 1.0; 62.1† 0.1; 63.1† 1.3; 62.9† 2.0; 63.1† 1.2; 62.4† 20.8; 62.8†
Fat-free mass 0.9; 60.7† 0.6; 60.7† 0.6; 60.5† 0.4; 60.7† 0.4; 60.6† 0.8; 60.7† 0.7; 60.5† 0.4; 60.6†
Bench press 1RM 4.6; 62.8z 5.0; 62.3z 4.6; 62.4z 3.5; 65.0z 4.7; 62.9z 8.4; 63.2z 5.7; 63.1z 5.3; 65.3z
Box squat 1RM 2.8; 610.5 1.4; 67.1 12.5; 65.3z 15.3; 620.1§ 0.4; 611.0 6.9; 68.4z 8.8; 67.4z 8.4; 618.4
Back squat 1RM 1.7; 66.1 5.3; 65.0z 5.2; 66.3z 3.4; 68.8 0.1; 66.3 5.0; 67.0z 10.3; 67.9z 9.3; 613.2z
Chin-ups 1RM 2.1; 62.2z 2.4; 62.1z 0.4; 62.3† 5.5; 66.6z 3.7; 62.6z 3.9; 62.9z 2.9; 62.7z 4.0; 65.7z
Power-clean 1RM 3.5; 65.1z 21.5; 64.5† 2.5; 64.5 3.6; 69.1 23.7; 66.9 2.1; 65.9 1.8; 64.7† 1.1; 68.5
Mean of 12 sprint 21.9; 61.7z 0.5; 61.4† 20.2; 61.3† 20.6; 62.5 1.7; 64.1 1.0; 64.0 20.4; 64.2 20.2; 65.8
Mean of 9 sprints 0.4; 61.9 0.4; 61.4 1.1; 61.2z 21.5; 62.2z
Forwards fatiguek 21.3; 61.6z 21.3; 61.3z 0.1; 61.1 0.1; 62.1 2.0; 63.3 0.0; 63.2 21.6; 63.5 20.2; 64.9
Backs fatiguek 21.5; 62.2z 21.5; 61.7z 0.5; 61.7 2.8; 63.2§
Mean of 12 vs. 20 m¶ 21.5; 61.2z 20.7; 61.0† 0.3; 60.9† 0.4; 61.7 1.3; 62.7 20.8; 62.6 21.3; 62.9 1.2; 64.1
Mean of 9 vs. 30 m¶ 21.9; 61.6z 21.0; 61.2 20.4; 61.1 1.0; 62.1

*Negative value indicates a decrease in speed time (got faster).

†Trivial change.
zSmall change.
§Moderate change; all other changes were unclear.
kNegative value indicates a reduction in RS2 fatigue (greater repeated sprint ability).
¶Negative value indicates a mean time closer to sprint time (greater repeated sprint ability).

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 the TM

Journal of Strength and Conditioning Research | www.nsca.com

Small decreases in all the speed

times (range, 1.4–1.8%) and
TABLE 4. Mean within-athlete changes (%); 699% confidence limits in physical
performance and anthropometry in rugby union players as they move from one small increases in chin-ups
competition period to another competition period within the same year.* (3.3%; 62.6%) and power-
clean (7.3%; 64.8%) 1RM
Super Rugby to Super Rugby to Provincial to occurred as players moved
international† provincial internationalz from Super Rugby to the inter-
10-m sprint 21.8; 60.9§ 20.7; 60.5k 0.3; 60.9 national midyear competition.
20-m sprint 21.4; 60.8§ 0.0; 60.5k 0.0; 61.2
30-m sprint 21.5; 61.7§ 20.4; 60.8k 21.1; 62.2
Body mass 0.0; 60.3k 0.1; 60.2k 20.2; 60.4k DISCUSSION
Skinfold thickness 21.0; 62.3k 21.0; 61.2k 21.8; 63.2k
Percent body fat 21.3; 62.1k 20.8; 61.1k 22.1; 62.9k Previous studies of the physical
Fat-free mass 0.1; 60.4k 0.3; 60.2k 20.1; 60.5k characteristics of players, not
Bench press 1RM 1.0; 64.1 0.5; 61.1k only in rugby union but also
Box squat 1RM 8.3; 614.1 8.0; 63.0§ rugby league, have typically
Back squat 1RM 22.7; 610.0 5.9; 63.9§ had small sample sizes that
Chin-ups 1RM 3.3; 64.1§ 1.3; 61.2k
Power-clean 1RM 7.3; 67.5§ 2.8; 62.9k are tested on 1 or 2 occasions
Mean of 12 sprints 20.6; 61.4 0.0; 60.6k (9,11,14,17,19,23,24,28,29). This
Mean of 9 sprints 21.6; 61.7§ 20.1; 60.9 study is the first to use a
Forwards fatigue 1.8; 61.8§ 0.4; 60.8k sample size of such a large
Backs fatigue 2.4; 62.0¶ 0.8; 61.0k magnitude (over 1,100) with
Mean of 12 vs. 1.2; 61.2§ 0.5; 60.5k
20 m numerous repeated measure-
Mean of 9 vs. 30 m 0.6; 61.5 0.6; 60.8k ments in a wide variety of
physical characteristics over
*A negative value indicates a decrease in the variable from the preceding competition.
†Midyear competition.
a long period of time. The sta-
zEnd of year competition. tistical power that the large
§Small change.
kTrivial change.
number of observations has
¶Moderate change; all other changes were unclear. provided, enabled the use of
narrow confidence intervals
(99%), increasing the certainty
of the clear outcomes reported.
Differences Between Years Forwards were generally heavier, had greater skinfold
The mean annual changes in performance within each level thickness, percentage body fat, and fat-free mass than backs,
for forwards and backs is shown in Table 3. All anthropo- with props the heaviest and the strongest. The outside backs
metrical measures showed trivial changes across all levels in were the fastest over all measured sprint distances, and the
both forwards and backs. All levels and positions showed half backs and first five eights showed the lowest fatigue in
small increases in bench press 1RM, and all except Super both the forwards and backs RS2 tests, respectively. The find-
Rugby forwards showed small increases in chin-ups 1RM. ings are consistent with those studies that have previously
Similarly, all levels, except the players not selected, showed investigated the anthropometrical characteristics in rugby
small-to-moderate increases in lower body strength (either players (6–9,11,19,23,24,28–30). Props are generally the largest
back squat 1RM, box squat 1RM or both). The lower-level of the positions to contest the ruck, maul, and scrum situa-
players (not selected and provincial) appeared to improve tions to win or maintain possession of the ball. The inside
more in the RS2 test (both forwards and backs) compared backs are generally the smallest to be mobile around the field
with the higher-level players. Specifically, small decreases and agile around the scrum, ruck, and maul (12,19,28). Inter-
were observed in both forwards and backs fatigue (range, estingly, the players in this study were more than 10-kg
1.3–1.5%) and small decreases were observed in mean of heavier than Senior A players (29) and up to 15-kg heavier
12 vs. 20 m and mean of 9 vs. 30 m (range, 0.7–1.9%), indi- than players in the early studies on the United States national
cating a better maintenance of maximal sprint performance. team (24). The mass of players are more comparable to recent
studies published on professional Super 12 players (2,11). The
Changes Within Players large difference highlights the rate at which the physique of
Mean changes within player’s performance as they moved rugby players is increasing, which may be a result of greater
from one competition period to another during the year is training loads and enhanced nutritional and recovery strate-
displayed in Table 4. Small increases occurred in box squat gies that has accompanied professionalism.
(8.0%; 63.0%) and back squat (5.9%; 63.9%) 1RM as players Speed characteristics of the players in this study are
moved from Super Rugby to the provincial competition. similar to those previously reported in rugby union. Indeed,

VOLUME 27 | NUMBER 11 | NOVEMBER 2013 | 3041

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 Physical Characteristics in Rugby

backs have been shown to be faster over distances greater a Super 12 squad. The data from this study indicate minimal
than 30 m than forwards, with outside backs the fastest year to year changes in body composition, but does not
(8,28,29). Differences in strength assessments on rugby play- discount substantial longitudinal changes from 2004 to
ers, however, make direct comparisons with other studies 2007; following previous trends of players becoming heavier
difficult. Nonetheless, when compared with a small sample with greater fat free mass (26,30).
of Super 14 players and professional rugby league players, There were small to moderate annual increases in both
the players in this study have lower power-clean 1RM; props upper and lower body strength. Interestingly, the Super
are the only comparable position in bench press 1RM; and Rugby and international players (all professional players)
the hookers, props and loose forwards have greater back were able to increase strength up to 15% per year, indicating
squat and box squat 1RM (1,3,4). changes can still be made at higher playing levels. The
Few studies have specifically investigated the repeated moderate increases may be because of the improving pro-
sprint ability of rugby players. Furthermore, the novel fessionalism of the players. Increased emphasis on physical
protocol used in this study makes it difficult to compare enhancement strategies such as specific training techniques,
with studies of other team sports. Because of the importance nutrition, and recovery may have allowed the attainment of
of the anaerobic energy supply for the repeated sprints greater resistance training volumes and decreased injury
performed during matches, further research is required rates, allowing players to train more regularly (13,16).
specifically using tests similar in nature to the RS2 test The greater improvement in the RS2 test by the lower-
employed in this study (12). The RS2 test replicates the dis- level players may be because of their lower physical capacity
tances and work to rest ratios of matches, and although is as a result of a younger training age (16). Indeed, there were
more time consuming to implement than a multistage shut- small to large differences between Super Rugby and provin-
tle test, can be used to assesses specific qualities of rugby cial players and between international and provincial players
union performance (12). in all 3 RS2 test variables. The law of diminishing returns
It seems that as playing level increases, players are faster, suggests that the rate of improvement in a player’s fitness is
heavier, have greater fat-free mass, lower skinfold thickness inversely proportional to their initial level of fitness (10).
and percent body fat and have greater strength and power. Therefore, higher-level players in this study with greater
These results reinforce findings from previous studies repeated sprint ability have limited scope for improvements
illustrating level-specific fitness and physique characteristics in performance compared with provincial players and those
that can distinguish between semiprofessional and profes- not selected.
sional players in rugby and rugby league (3,29). The differ- Rugby in New Zealand is unique in that a player may play
ences between playing levels could be because of a greater up to 4 different competitions during 4 distinct periods
training history and an increased requirement of strength within a year. Because of this unique situation, very little
and size within the professional game (3,12). research has investigated the players change in physical
The increased physical capacity as playing level increases performance between respective competitions. This study is
only seems to occur up to a certain level. A large number of the first that has specifically investigated the changes in
trivial differences occurred between Super Rugby and physical performance as players moved from different
international players in mass and strength in both the competitions.
forwards and backs. The lack of consistent differences As the players moved from the Super Rugby to provincial
indicates that selection into the higher international level competition, they had small increases in lower-body
squad may not be determined by physical attributes as much strength. Furthermore, as players moved from Super Rugby
as lower levels. Alternatively, international players may be to the international midyear competition, they had small
selected because of greater skill and experience compared increases in chin-ups and power-clean 1RM and small
with Super Rugby players (15). The international players decreases in sprint time over all distances. The results show
also compete within the Super Rugby competition, and that the Super Rugby competition does not allow a player
the time between the conclusion of the Super Rugby com- to achieve optimal performance in aspects of strength,
petition and the beginning of the international midyear com- power, and speed compared with other competition
petition is negligible. Therefore, a combination of inadequate periods during the year. The Super Rugby competition
preparation time and a greater number of high-intensity has potentially higher match intensity than the provincial
matches may minimize the international players’ ability to competition and is performed over 3 continents (New
train to make increased physical gains over other Super Zealand, Australia, and South Africa), requiring a large
Rugby players. amount of travel between the weekly matches. Further-
There were trivial increases in the average annual change more, conditioning strategies employed by team manage-
of all anthropometrical variables over the period in which ment will be specific to the requirements of their own
data were obtained. Similarly, Duthie et al. (11) reported team and their schedule exclusively within the Super
trivial changes in mass over three years and small increases Rugby competition. A combination of these factors may
in sum of skinfolds in the third year of being involved in contribute to the decrease in training load that has been
the TM

3042 Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 the TM

Journal of Strength and Conditioning Research | www.nsca.com

shown from preseason to in-season in Super Rugby players 11. Duthie, GM, Hooper, SL, Hopkins, WG, Livingstone, S, and
(2). The reduction in training load may therefore not pro- Pyne, DB. Anthropometry profiles of elite rugby players: quantifying
changes in lean mass. Br J Sport Med 40: 202–207, 2006.
vide adequate stimulus for the achievement of high levels of
12. Duthie, GM, Pyne, DB, and Hooper, SL. Applied physiology and
physical performance during the competition compared game analysis of rugby union. Sports Med 33: 973–991, 2003.
with other competition periods. However, increasing train- 13. Gabbett, TJ. Changes in physiological and anthropometric
ing load in-season to achieve higher levels of physical per- characteristics of rugby league players during a competitive season.
formance may not be appropriate. Increased training loads J Strength Cond Res 19: 400–408, 2005.
have been related to higher injury rates in rugby league 14. Gabbett, TJ. A comparison of physiological and anthropometric
characteristics among playing positions in junior rugby league
players (18), whereas it is unknown whether increases in
players. Br J Sport Med 39: 675–680, 2005.
training load, in conjunction with high game loads, will
15. Gabbett, TJ. Influence of physiological characteristics on selection in
allow the players to recover effectively for optimal perfor- a semi-professional first grade rugby league team: a case study.
mance during matches (2). J Sport Sci 20: 399–405, 2002.
16. Gabbett, TJ. Physiological and anthropometric characteristics of
PRACTICAL APPLICATIONS junior rugby league players over a competitive season. J Strength
Cond Res 19: 764–771, 2005.
This study is the first to use longitudinal data to describe the 17. Gabbett, TJ. Physiological characteristics of junior and senior rugby
between-player differences and within-player changes in league players. Br J Sport Med 36: 334–339, 2002.
physical performance in rugby union players. The data have 18. Gabbett, TJ and Domrow, N. Relationships between training load,
provided normative profiles for different positions, multiple injury, and fitness in sub-elite collision sport athletes. J Sport Sci 25:
1507–1519, 2007.
playing levels, and the expected long-term improvements in
19. Holway, FE and Garavaglia, R. Kinanthropometry of Group I rugby
physical performance. Data such as these will inform
players in Buenos Aires, Argentina. J Sport Sci 27: 1211–1220, 2009.
coaches, so that they have better understanding of what
20. Hopkins, WG. Measures of reliability in sports medicine and
should be expected and what could be achieved with a rugby science. Sports Med 30: 1–15, 2000.
player over a long period of time. The novel aspect of the 21. Hopkins, WG, Marshall, SW, Batterham, AM, and Hanin, J.
study was the physical changes within players from one Progressive statistics for studies in sports medicine and exercise
competition to another. The information should help science. Med Sci Sport Exerc 41: 3–12, 2009.
coaches develop strategies, such as more specific periodiza- 22. Landers, J. Maximum based on reps. Strength Cond J 6: 60–61, 1985.
tion and recovery, to improve the decrements in perfor- 23. Maud, PJ. Physiological and anthropometric parameters that
describe a rugby union team. Br J Sport Med 17: 16–23, 1983.
mance during specific times of the year.
24. Maud, PJ and Shultz, BB. The US national rugby team:
a physiological and anthropometric assessment. Phys Sportsmed 12:
89–99, 1984.
REFERENCES
25. Norton, K, Whittingham, N, Carter, L, Kerr, D, Gore, C, and
1. Argus, CK, Gill, N, Keogh, J, Hopkins, WG, and Beaven, CM. Marfell-Jones, M. Measurement techniques in anthropometry. In:
Effects of a short-term pre-season training programme on the body Anthropometrica. K. Norton and T. Olds, eds. Sydney, Australia:
composition and anaerobic performance of professional rugby University of New South Wales Press Ltd, 2004. pp. 25–75.
union players. J Sport Sci 28: 679–686, 2010.
26. Olds, T. The evolution of physique in male rugby union players in
2. Argus, CK, Gill, ND, Keogh, JWL, Hopkins, WG, and Beaven, CM.
the twentieth century. J Sport Sci 19: 253–262, 2001.
Changes in strength, power, and steroid hormones during
a professional rugby union competition. J Strength Cond Res 23: 27. Perini, TA, de Oliveira, GL, dos Santos Ornellas, J, and de
1583–1592, 2009. Oliveira, FP. Technical error of measurement in anthropometry
(English version). Brazilian Journal of Sports Medicine 11: 86–90,
3. Baker, DG. Comparison of upper-body strength and power between
2005.
professional and college-aged rugby league players. J Strength Cond
Res 15: 30–35, 2001. 28. Quarrie, KL, Handcock, P, Toomey, MJ, and Waller, AE. The
New Zealand Rugby Injury and Performance Project. IV.
4. Baker, DG and Nance, S. The relation between strength and
power in professional rugby league players. J Strength Cond Res 13: Anthropometric and physical performance comparisons between
224–229, 1999. positional categories of senior A rugby players. Br J Sport Med 30:
53–56, 1996.
5. Beaven, CM, Gill, ND, Ingram, JR, and Hopkins, WG. Acute
salivary hormone responses to complex exercise bouts. J Strength 29. Quarrie, KL, Handcock, P, Waller, AE, Chalmers, DJ, Toomey, MJ,
Cond Res 25: 1072–1078, 2011. and Wilson, BD. The New Zealand Rugby Injury and Performance
Project. III. Anthropometric and physical performance
6. Bell, W. Body composition of rugby union football players. Br J characteristics of players. Br J Sport Med 29: 263–270, 1995.
Sport Med 13: 19–23, 1979.
30. Quarrie, KL and Hopkins, WG. Changes in player characteristics
7. Bell, W. Body composition and maximal aerobic power of rugby and match activities in Bledisloe Cup rugby union from 1972 to
union forwards. J Sport Med Phys Fitness 20: 447–451, 1980. 2004. J Sport Sci 25: 895–903, 2007.
8. Carlson, B, Carter, J, Patterson, P, Petti, K, Orfanos, S, and Noffal, G. 31. Siri, WE. Body volume measurement by gas dilution. In: Techniques
Physique and motor performance characteristics of US national for Measuring Body Composition. J. Brozek and A. Henschel, eds.
rugby players. J Sport Sci 12: 403–412, 1994. Washington, DC: National Academy of Sciences, National Research
9. Casagrande, G and Viviani, F. Somatotype of Italian rugby players. Council, 1961. pp. 108–117.
J Sport Med Phys Fit 33: 65–69, 1993. 32. Slater, GJ, Duthie, GM, Pyne, DB, and Hopkins, WG. Validation of
10. Duthie, GM. A framework for the physical development of elite a skinfold based index for tracking proportional changes in lean
rugby union players. Int J Sport Physiol Perform 1: 2–13, 2006. mass. Br J Sport Med 40: 208–213, 2006.

VOLUME 27 | NUMBER 11 | NOVEMBER 2013 | 3043

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
 Physical Characteristics in Rugby

33. Smart, DJ. Physical profiling of rugby union players: implications 35. van Rooyen, M, Rock, K, Prim, S, and Lambert, M. The
for talent development. PhD thesis, AUT University, Auckland, quantification of contacts with impact during professional rugby
2011. matches. Int J Perform Anal Sport 8: 113–126, 2008.
34. Smart, DJ. The Reliability of a Rugby Union Specific Test: With Reference 36. Withers, RT, Laforgia, J, Heymsfield, S, Wang, Z, and Pillans, R. Two,
to the Relationship between Performance and the Concentration of three, and four-compartment chemical models of body composition
Testosterone and Cortisol. Unpublished Master’s thesis, WINTEC, analysis. In: Anthropometrica. K. Norton and T. Olds, eds. Sydney,
Hamilton, New Zealand, 2005. Australia: University of New South Wales Press Ltd, 2004. pp. 199–231.

the TM

3044 Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.