See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/243969087

Global Positioning Systems (GPS) and Microtechnology Sensors in Team

Sports: A Systematic Review

Article  in  Sports Medicine · June 2013

DOI: 10.1007/s40279-013-0069-2 · Source: PubMed

CITATIONS READS

442 9,683

4 authors, including:

Cloe Cummins Rhonda Orr

University of New England (Australia) The University of Sydney
46 PUBLICATIONS   599 CITATIONS    98 PUBLICATIONS   2,512 CITATIONS   

SEE PROFILE SEE PROFILE

Helen O'Connor
The University of Sydney
161 PUBLICATIONS   3,444 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Food, Mood & Mind Study View project

Training Load , Injury Characteristics, and Performance in University Rugby Union View project

All content following this page was uploaded by Rhonda Orr on 22 May 2014.

The user has requested enhancement of the downloaded file.

Sports Med (2013) 43:1025–1042
DOI 10.1007/s40279-013-0069-2

SYSTEMATIC REVIEW

Global Positioning Systems (GPS) and Microtechnology Sensors

in Team Sports: A Systematic Review
Cloe Cummins • Rhonda Orr • Helen O’Connor •

Cameron West

Published online: 28 June 2013

Ó Springer International Publishing Switzerland 2013

Abstract limited representation in other team sports: cricket (n = 3),

Background Use of Global positioning system (GPS) hockey (n = 3), lacrosse (n = 1), and netball (n = 1). Of
technology in team sport permits measurement of player the included manuscripts, 34 (97 %) detailed work rate
position, velocity, and movement patterns. GPS provides patterns such as distance, relative distance, speed, and
scope for better understanding of the specific and positional accelerations, with only five (14.3 %) reporting on impact
physiological demands of team sport and can be used to variables. Activity profiles characterizing positional play
design training programs that adequately prepare athletes and competitive levels were also described. Work rate
for competition with the aim of optimizing on-field patterns were typically categoriszed into six speed zones,
performance. ranging from 0 to 36.0 km
h-1, with descriptors ranging
Objective The objective of this study was to conduct a from walking to sprinting used to identify the type of
systematic review of the depth and scope of reported GPS activity mainly performed in each zone. With the exception
and microtechnology measures used within individual of cricket, no standardized speed zones or definitions were
sports in order to present the contemporary and emerging observed within or between sports. Furthermore, speed
themes of GPS application within team sports. zone criteria often varied widely within (e.g. zone 3 of AFL
Methods A systematic review of the application of GPS ranged from 7 to 16 km
h-1) and between sports (e.g. zone 3
technology in team sports was conducted. We systemati- of soccer ranged from 3.0 to \13 km
h-1 code). Activity
cally searched electronic databases from earliest record to descriptors for a zone also varied widely between sports
June 2012. Permutations of key words included GPS; male (e.g. zone 4 definitions ranged from jog, run, high velocity,
and female; age 12–50 years; able-bodied; and recreational to high-intensity run). Most manuscripts focused on the
to elite competitive team sports. demands of higher intensity efforts (running and sprint)
Results The 35 manuscripts meeting the eligibility crite- required by players. Body loads and impacts, also sum-
ria included 1,276 participants (age 11.2–31.5 years; 95 % marized into six zones, showed small variations in
males; 53.8 % elite adult athletes). The majority of descriptions, with zone criteria based upon grading systems
manuscripts reported on GPS use in various football codes: provided by GPS manufacturers.
Australian football league (AFL; n = 8), soccer (n = 7), Conclusion This systematic review highlights that GPS
rugby union (n = 6), and rugby league (n = 6), with technology has been used more often across a range of
football codes than across other team sports. Work rate
pattern activities are most often reported, whilst impact
Electronic supplementary material The online version of this data, which require the use of microtechnology sensors
article (doi:10.1007/s40279-013-0069-2) contains supplementary
material, which is available to authorized users. such as accelerometers, are least reported. There is a lack
of consistency in the definition of speed zones and activity
C. Cummins (&)
R. Orr
H. O’Connor
C. West descriptors, both within and across team sports, thus
Faculty of Health Sciences, The University of Sydney,
underscoring the difficulties encountered in meaningful
75 East St, Lidcombe, Sydney, NSW 2141, Australia
e-mail: [email protected]; comparisons of the physiological demands both within and
[email protected] between team sports. A consensus on definitions of speed
1026 C. Cummins et al.

zones and activity descriptors within sports would facilitate acceleration/deceleration, related changes of direction and
direct comparison of the demands within the same sport. impacts from both the player-to-player collision and con-
Meta-analysis from systematic review would also be sup- tact with the ground (foot strikes and falls) [7]. GPS
ported. Standardization of speed zones between sports may technology allows for further analysis of speed and impact
not be feasible due to disparities in work rate pattern characteristics by classification into six activity bands
activities. known as speed and impact zones, respectively. Zone one
indicates the lowest, whilst zone six indicates the highest
level of effort or impact [8]. These details can be utilized to
1 Introduction analyze training or game settings and compare player
performance or session intensity.
Global positioning system (GPS) is a satellite-based navi- There is an abundance of literature examining the
gational technology originally devised for military pur- validity and reliability of GPS for the measurement of
poses. A number of informative, technical reviews on GPS movement in the football codes, hockey, and cricket. The
[1–3] have previously been published outlining how this gold standard criterion method used to investigate GPS
technology enables three-dimensional movement of an validity for distance is to measure a course with a trundle
individual or group to be tracked over time in air-, aquatic-, wheel or tape measure and, for speed, use of timing gates at
or land-based environments. The recent development of the start and finish [9] or a speed gun [10]. GPS devices are
portable GPS units has permitted wider application of this currently manufactured with 1-, 5-, and 10-Hz sampling
technology in a variety of settings, including sport, thus rates (the speed at which the unit gathers data). The liter-
providing an additional means for describing and under- ature suggests that GPS with a higher frequency rate pro-
standing the spatial context of physical activity. First uti- vides greater validity for measurement of distance. When
lized for athlete tracking in 1997 [4], GPS technology is comparing the precision of distance acquisition between a
now increasingly used in team sport settings to provide 1- and a 5-Hz GPS, the standard error of a standing start
sports scientists, coaches, and trainers with comprehensive 10-m sprint was 32.4 and 30.9 %, respectively [11]. By
and real-time analysis of on-field player performance dur- contrast, a 10-Hz GPS demonstrated a 10.9 % standard
ing competition or training. error over a 15-m sprint [9]. Recently, it has been reported
GPS technology has been used extensively in rugby that GPS devices at 1 Hz may be unable to record move-
league, rugby union, Australian football league (AFL), ments taking\1 s to complete [12]. The newer 10-Hz units
cricket, hockey, and soccer, with only limited research are capable of measuring the smallest worthwhile change
available in netball, hockey, and lacrosse. Current literature in acceleration and deceleration, whereas the 5-Hz unit is
provides an array of information on the activity profile of unable to do so [10]. The greater errors associated with
field sport athletes. By measuring player movements, GPS measurement of distance with the 1- and 5-Hz versus the
can be used to objectively quantify levels of exertion and 10-Hz GPS devices indicate that the sampling rate may be
physical stress on individual athletes, examine competition limiting the accuracy of distance measurements and
performances, assess different positional workloads, velocity.
establish training intensities, and monitor changes in player The speed of a movement impacts the validity of the
physiologic demands [5]. Player movement patterns and GPS-measured distance. The earliest validation of a GPS
activity profiles (external loads) can be used in addition to device (GPS 45, Garmin) showed various walk and run
tactical information and physiological responses (internal velocities (2–20 km
h-1) were highly correlated (r = 0.99)
load) to characterize competitive match play [5]. From its with a chronometer [4]. A more recent study [13] shows
introduction, GPS was used to measure basic components GPS distance measurement error to be lowest during
of player movement patterns, speed, and distance travelled walking (*1.8 ms-1; standard error of estimate [SEE]
and the number of accelerations and decelerations. The 0.7 %) and highest during running (*6 ms-1; SEE 5.6 %).
integration of GPS with a triaxial accelerometer enables the Similarly Johnston et al. [12] reported that GPS is capable
capture of information on work rate patterns and physical of measuring work rate patterns performed at velocities
loads. The triaxial accelerometer measures a composite \20 km
h-1; however, recommended caution when ana-
vector magnitude (expressed as a G-force) by recording the lyzing work rate patterns at velocities [20 km
h-1. These
sum of accelerations measured in three axes (X, Y, and results indicate that movement velocity impacts upon
Z planes) [6]. In addition, the number and intensity of accuracy, with GPS reported as a valid method for mea-
physical contacts and collisions between athletes and surement of distance travelled at low to moderate but not
objects or surfaces can be quantified by body load and high speeds.
impact measures. Body load (measured as G-force) is the The validity of distance measures improves with longer
collation of all forces imposed on an athlete, including duration activities [11]; for example, the coefficient of
GPS Technology in Team Sports 1027

variation (CV) diminished from 32.4 to 9.0 % for sprint further research on the validity and reliability for this
distances of 10 and 40 m, respectively. The CV was purpose is warranted.
further reduced to 3.8 % for a range of velocities com- Although many studies have examined the GPS appli-
pleted over a 140-m modified team-sport running circuit cation within specific team sports, a comprehensive docu-
[11]. mentation of movement patterns and physiological
The factors of sampling frequency, distance, and speed, demands of team sports has not been undertaken. Fur-
which affect GPS validity, similarly affect the reliability of thermore, the rapidly evolving utilization of GPS in the
GPS. The impact of sampling frequency still remains football codes, through the application of player profiles to
unclear; for example, the CV of a linear soccer task has investigate collision impacts, match fatigue, and injury
been reported as 4.4–4.5 % for a 1-Hz GPS and 4.6–5.3 % risk, may inform all team sports.
for a 5-Hz [14]. However, another study [11] reported the Therefore, the aim of this study was to conduct a sys-
CV of a 10-m sprint as 77 and 39 % for 1 and 5 Hz, tematic review to collate the considerable depth and scope
respectively. More recently, a higher sampling rate of of reported GPS measures within individual sports in order
10 Hz has demonstrated improved reliability during the to present the contemporary and emerging themes of GPS
constant velocity and acceleration or deceleration phase application within team sports.
(CV\5.3 and\6 %, respectively) [10]. Whilst the data are
currently ambiguous and may be explained through the use
of different GPS manufacturers and models [9], it would 2 Methods
seem that an increased sample rate appears to improve the
reliability of GPS measures. 2.1 Design
The reliability of GPS decreases with the increased
velocity of movement. The CV of walking for a 5-Hz GPS Studies investigating the use of GPS systems to monitor the
was 1.4–2.6 %, whilst the CV of sprinting over a 20-m position, speed, or distance of an athlete during outdoor
distance was 19.7–30 %. Similarly, CVs of 30.8 and training or competition settings within team sports were
77.2 %, respectively, for walking and sprinting over a eligible for inclusion in the review. A systematic search of
10-m distance were noted with a 1-Hz GPS [11]. The electronic databases (MEDLINE, SPORTdiscus, CINAHL,
reliability of GPS devices is also negatively affected by Web of Science, Scopus, EMBASE, Cochrane, and Google
movements requiring changes in direction. The CV for Scholar) was performed from the earliest record to June
gradual and tight change of direction movements at walk- 2012. All study designs (e.g. randomized controlled trials,
ing pace has been reported as 11.5 and 15.2 %, respectively cohort cross over, observational) were included. The search
[11]. The tight change of direction movements may dem- strategy combined terms covering the topics of device
onstrate a decreased reliability due to the increased number (GPS), population (male and female, 12–50 years, able
of speed changes performed [11]. bodied, recreational to elite), and activity (sport, team
The re-test reliability between GPS devices is consis- sport, exercise, training, physical activity, rugby, AFL,
tent. Waldron et al. [6] examined the re-test reliability football, rugby union, rugby league, soccer, basketball,
between GPS units, finding random errors between two hockey, and cricket). Studies were included if athletes were
tests ranging from 0.56 to 1.64 km
h-1 and small mean monitored using GPS systems and at least one of the fol-
biases (-0.01 to -0.14 km
h-1) for all sprint intervals. lowing GPS outcomes was reported: distance, velocity/
Overall, studies conclude that GPS devices have an speed, acceleration, impacts, or body load.
acceptable level of validity and reliability for assessing
movement patterns at lower speeds and over increased 2.2 Selection of Studies
distance efforts. The decreased reliability of GPS units to
accurately measure movement patterns during short- After eliminating duplicates, search results were screened
duration, high-speed, straight-line running, and efforts independently by two researchers (CC, CW) against eli-
requiring changes in direction may limit both accuracy gibility criteria. References that could not be eliminated by
and reliability for assessing these aspects in team sports. title or abstract were retrieved and independently evaluated
However, GPS units with increased sampling frequency for inclusion. Reviewers were not masked to the title or
demonstrate improved reliability and validity and can be authors of the publications. Disagreements were resolved
utilized in the monitoring of physical activity in situations by discussion or via a third researcher (RO). Abstracts and
such as team sports, so long as caution is taken when conference papers from annual meetings were not included
interpreting individual sprints and rapid changes in both due to not meeting the rigour of outcome measures. In
direction and velocity. Use of GPS for quantifying cases where journal articles contained insufficient infor-
impacts in collision sports is still relatively new and so mation, attempts were made to contact authors to obtain
1028 C. Cummins et al.

missing details. Reference lists of all retrieved papers were sion is displayed in Fig. 1. One author was contacted and
manually searched for other potentially eligible papers. provided mean and SD of data presented graphically in the
Papers from all languages were included, but were exclu- manuscript as median and range [16].
ded if a translation could not be made.
3.2 Study Characteristics
2.3 Data Extraction
There were a total of 1,276 participants in the included
Data relating to the participant characteristics (age, sex, manuscripts, with individual study numbers ranging from 3
height, mass, and level of competition/athletic calibre), GPS to 179. Manuscripts mainly described young (age
unit (brand, model, speed, and distance recording fre- 11.2–31.5 years) males (95.2 %), with 32 studies exclu-
quency), and work rate patterns were extracted by two sively recruiting male, three exclusively female, and none
researchers (CC, RO). Work rate patterns included speed of mixed sex (Table 1). Of the 35 papers, 18 (51.4 %)
zone identification (zones 1–6), time spent in zones (min), report on elite adult athletes, whilst eight (22.9 %) report
and speed (km
h-1). Distance covered was extracted as on children/adolescent athletes.
either total distance (m) over the competition or training The team sports included AFL (n = 8) [17–24], soccer
session, distances covered in each zone, or average maximal (n = 7) [16, 25–30], rugby union (n = 6) [31–36], rugby
or total sprint distance (m). Sprint duration (min) over the league (n = 6) [5, 8, 37–40], cricket (n = 3) [41–43],
competition or training session was also extracted in addition hockey (n = 3) [44–46], netball (n = 1) [47], and lacrosse
to the number of acceleration and decelerations (m
s-2), (n = 1) [48]. There was a strong bias towards description
impact (G), and body load (G). Studies reporting speed in of football codes (AFL, rugby union, rugby league, and
m
min-1 were converted to km
h-1. Authors were contacted soccer), with the majority of all studies (26 of 35) [5, 8, 17–
to provide data if reported only in graphical form. When data 24, 29–32, 35, 37–45, 47, 48] conducted in Australian
were unavailable, two researchers (CC, RO) estimated mean sports settings. Papers demonstrated varied use of GPS,
and standard deviation (SD) in duplicate using a ruler. with 25 [5, 8, 16–20, 22, 24–28, 33–38, 40–43, 45, 46]
studies examining competition match variables, eight [21,
2.4 Assessment of Methodological Quality 23, 29–32, 47, 48] examining training variables, and three
reporting on both training and competition variables [32,
The quality of included studies was independently assessed 39, 44].
by two researchers (CC and RO) using the modified
assessment scale of Downs and Black [15]. Of the 27 cri- 3.3 Methodological Quality
teria, 12 that logically applied to the study designs included
in this review were used. No studies were eliminated and The scores for the assessment of methodological quality
no additional subgroup analysis was undertaken on the ranged from 6 to 10 across the 12 items assessed (See
basis of methodological quality. A meta-analysis was not Electronic Supplementary Material, T1). The main quality
performed, as study designs were heterogeneous and issue was the provision of exact P values. Three [16, 18,
unable to be pooled. 43] studies failed to provide a clear description of partici-
pants. One criterion required that a hypothesis, aim, or
2.5 Statistics objective was stated; only four studies identified an
hypothesis [5, 8, 20, 40].
All data are presented as mean or mean (SD) unless
otherwise specified. 3.4 Analysis of Physiological Outcomes

All 35 manuscripts, apart from one [8], reported on various

3 Results GPS variables of movement (n = 34; 97 %) [5, 16–48],
with only five manuscripts (14.3 %) [8, 33, 34, 36, 39]
3.1 Identification and Selection of Studies evaluating impact variables.

The original search netted 7,333 studies. After the removal 3.4.1 Work Rate Patterns
of duplicates and screening, 43 studies remained. One of
these was unable to be translated and was excluded; seven Movement variables included distance covered (m), mean
studies were eliminated based on the eligibility criteria. and maximum velocity (km
h-1), acceleration and decel-
The flow of the papers from potentially relevant to inclu- eration (m
s-2), impact (G), and body load (G).
GPS Technology in Team Sports 1029

Fig. 1 Study selection flow

chart. GPS global positioning
system

Table 1 Characteristics of studies

Sport Sample size % Male Age range (years) Athletic calibre Reference

AFL 348 100 16.2–28.9 Elite, subelite, junior [17–24]

Rugby Union 188 100 14.0–28.6 Elite, semi-professional, junior [31–36]
Rugby League 287 100 16.9–31.5 Elite [5, 8, 37–40]
Soccer 286 95.8 11.2–20.0 Elite, semi-professional, elite under 19, [16, 25–30]
elite junior, junior
Cricket 96 100 19.3–24.9 Elite [41–43]
Hockey 55 70.9 18.0–29.7 Elite [44–46]
Netball 9 0 18.0–27.2 State league 1 [47]
Lacrosse 7 100 18.5–21.3 Collegiate division 1 [48]
Overall 1,276 95.2 11.2–31.5
AFL Australian Football League

3.4.1.1 Total Distance Total distance travelled (Fig. 2) Athletic calibre and age impacted upon total distance.
was, historically, one of the first GPS variables to be Elite [33] and junior [32] rugby union forwards covered
monitored and thus the most commonly reported variable 5,853 and 3,511 m, respectively, with elite players cover-
in the studies included in this review. Cricket players ing 40 % further distance. Similarly in soccer, under
covered the greatest distance per competition game, with a 18 years players covered 8,867 m [27], an additional
fast bowler in One Day International cricket covering 27.2 % distance compared with the 6,459 m observed in
15,903 m [41–43], followed by elite AFL players covering under 13 years players [27].
a total distance of 12,939 m [17–20, 22, 24]. Soccer con- Positional differences were evident within each sport,
sistently demonstrated a higher distance travelled per game for example, backs in both rugby league [37, 39, 40] and
than rugby union and rugby league. rugby union [32–34] covered greater distances than
1030 C. Cummins et al.

Fig. 2 Total distance covered

during competitive matches or
training [16, 19, 20, 22, 27–34,
36–40, 42–47]. Values
expressed as mean (SD). AFL
Australian Football League, BR
back row, FR front row, HU hit
up, ODI one day international,
Rep representative, TS talent
squad, Ux under x years, WR
wide running, 2v2 2 versus 2
players, 4v4 4 versus 4 players,
6v6 6 versus 6 players
GPS Technology in Team Sports 1031

forwards; however, the margin of difference was less in [16, 25–30], cricket [41–43], hockey [44–46], and lacrosse
rugby union. McLellan et al. [40], reported that elite [48].
rugby league backs covered 5,747 m or 16.9 % greater
distance than forwards (4,774 m). By comparison, in 3.4.1.4 Impact, Body Load, and Collision Only five
rugby union, backs covered 7.6 % more distance than manuscripts (14.3 %) reported on impact variables
forwards (6,471 and 5,853 m, respectively) [33]. Similarly (Table 3 and Electronic Supplementary Material, T2) in
in AFL, Brewer and colleagues [19] found that mid- rugby league [8, 39] and rugby union [33, 34, 36]. Impact
fielders and small forwards/backs covered greater total intensity was generally graded into six impact zones based
distances than other positions, with ruckmen recording the upon the grading system provided by GPS manufacturers
lowest distances. or in accordance with work by Cunniffe et al. [36]. Small
variation in impact range was observed between football
3.4.1.2 Relative Distance Relative distance or total dis- codes, with three manuscripts [8, 33, 36] reporting on six
tance travelled per minute of competition game time impact zones and one manuscript [39] only reporting on
(m
min-1) may provide a more accurate reflection of match four. Cunniffe et al. [36] monitored impacts in only two
intensity than total distance covered, as it takes into rugby union players, where the forward experienced 3.8
account the event time. Additionally, distance is only a times more frequent impacts during a single competition
measure of volume, whilst relative distance is a measure of game than the back. By comparison, the back showed
intensity. Only eight [16, 22, 35–39, 45] manuscripts slightly greater loads than the forward (52 %) in the first
reported on relative distance (Fig. 3). Although few papers half, whereas the forward recorded a significantly higher
address this, positional differences were apparent when body load in the second half of competition games than the
examining the distance covered per minute, showing sim- back (66 %). Similarly in rugby league [8], the forwards
ilarities to total distance. A difference of 7.2 % relative sustained 63 more impacts (7.3 %) than the backs.
distance was evident between elite rugby union forwards
and backs (66.7 vs. 71.9 m
min-1, respectively) [36]. The 3.4.1.5 Acceleration and Deceleration Acceleration is
impact of age on distance travelled per minute was also another measure of high-intensity exertions in team sport.
observed in soccer, where under 16 years players covered Only five studies quantified acceleration across three team
115.2 m
min-1, an additional 10 % relative distance sports (AFL [17, 21], rugby union [33, 35], and hockey
compared with that of under 13 years players [44]) in the contexts of training drills [21, 44] and matches
(103.7 m
min-1) [16]. [17, 33, 35]. Manuscripts investigated the physical
demands of accelerations of rugby union playing positions
3.4.1.3 Work Rate Patterns by Speed Zones Work rate [33] and compared the physical efforts in regular compe-
patterns were further categorized into six speed zones, tition matches against the physical efforts of finals [17] and
ranging from 0 to 36 km
h-1, and each of these zones was international matches [35]. The latter was the only paper to
linked with a description of the activity (e.g. walking quantify deceleration demands [35]. The limited avail-
through to sprinting). There were large variations in the ability of acceleration and deceleration data can mostly be
range of speeds for each zone, both within and across attributed to the inability of GPS to accurately measure
sports. The greatest variations were observed between these variables, until the introduction of 10-Hz GPS units.
sports, for example, the zone 4 speeds reported were
hockey [44–46] and soccer [16, 25–30] (7–18 km
h-1), 3.5 Sport-Specific Summary of Global Positioning
cricket [41–43] (13–14 km
h-1), rugby union [31–36] System (GPS) Studies in Team Sports
(12–21 km
h-1), rugby league [5, 8, 37–40]
(14.1–20 km
h-1), and AFL [17–24] (14–20 km
h-1). 3.5.1 Australian Football League (AFL)
Furthermore, zone 4 definitions ranged from jog, stride,
run, high velocity, to high-intensity run, both within and The eight AFL manuscripts included 348 male participants.
between the sports. Cricket, where all manuscripts were The studies included elite (n = 292; aged 18.7–28.9 years),
from a singular authorship [41–43], was the only sport sub-elite (n = 26; aged 19.1–25.9 years), and elite junior
where speed zones and descriptors were consistently (n = 30; aged 15.9–17.2 years) athletes. All AFL studies
applied across studies. In AFL papers, zone 3 speeds varied examined the work rate patterns of high-intensity running
from 7 to 16 km
h-1 (Table 2). The majority of papers (30 or sprinting in addition to distance covered at various
[85.7 %]) concentrated upon the high-intensity efforts and speeds (see Electronic Supplementary Material, T3). Two
distances covered in zones 5 and 6. Analysis of the higher manuscripts reported distance measures [22, 23]. Piggot
intensity zones was apparent in AFL [17, 18, 20, 21, 24], and colleagues [23] examined training load and injury
rugby league [5, 37, 38, 40], rugby union [32–36], soccer levels throughout 15 weeks of pre-season training. Weekly
1032 C. Cummins et al.

Fig. 3 Relative distance (distance travelled per minute) in competitive matches or training [16, 22, 34–39]. Values expressed as mean (SD). AFL
Australian Football League, Dom dominant, HU hit up, Int international, Sub substitute, Ux under x years WR wide running
Table 2 Zone classification of work rate patterns in team sports
References Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6
Speed Description Speed Description Speed Description Speed Description Speed Description Speed Sprinting
(km
h-1)a (km
h-1)a (km
h-1)a (km
h-1)a (km
h-1)a (km
h-1)a

Australian Football League (AFL)

Farrow et al. [21] 7.2–14.4 Moderate [14.4 High velocity [20.0 Sprint
velocity
Brewer et al. [19] [15.0 High- 20.0–23.0 Higher-speed [23.0 Sprint
intensity run running
GPS Technology in Team Sports

Aughey and 15.2–18.0 Jog [18.0 [25.0

Falloon [18]
Coutts et al. [20] 0–0.7 Standing 0.7–7.0 Walk 7.0–14.4 Jog 14.4–20.0 Run 18.0–24.9 Run 24.9–36.0 Sprint
Wisbey et al. [24] 0–0.8 8.0–12.0 12.0–16.0 16.0–18.0 18.0–21.6 [18.0 Sprint/
maximum
intensity
Aughey [17] 7.2–10.8 14.4–18.0 [18.0 Sprinting
Rugby Union
Hartwig et al. 0–1.0 Stationary 1.0–7.0 Walk 7.0–12.0 Jog 12.0–21.0 Stride 81.0–95.0 % Sprinting [21.0 Sprint
[31] Vmax
Cunniffe et al. 0–6.0 Standing, 6.0–12.0 Jogging 12.0–14.0 Cruising 14.0–18.0 Striding
[36] walking
Hartwig et al. 0–1.0 Stationary 1.0–7.0 Walking 7.0–12.0 Jogging 12.0–21.0 Striding 18.0–20.0 High-intensity [21.0 Sprint
[32] running
Venter et al. [34] 0–1.0 Standing \20 % Walking 20–50 % Jogging 51–80 % Striding 96–100 % Maximum
Vmax Vmax Vmax Vmax sprint
Higham [35] 0–7.2 7.2–12.6 12.6–18 18.0–21.6 [21.6
Suárez-Arrones 0.1–5.9 Standing, 6.0–11.9 Jogging 12.0–13.9 Cruising 14.0–17.9 Striding 18.0–19.9 High-intensity [20 Sprinting
et al. [33] walking running
Rugby League
McLellan et al. 0–6.0 Standing/ 6.1–12 Jogging 12.1–14.0 Cruising 14.1–18.0 Striding 18.1–20.0 High-intensity [20.1 Sprint
[5, 40] walking running
Austin and Kelly 0–12 Standing, 12–14 Cruising 14–18 Striding 18–20 High- 20–24 Sprinting [24 High-
[37] walking, or intensity intensity
jogging running sprinting
Duffield et al. \14.4 Low-speed [14.5 High-speed [20 Very-high
[38] activity running speed
running
Gabbett et al. 0–3.6 Low speed [3.6 High speed
[39]
Soccer
Barbero Alvarez 0.0–0.4 Standing/stop 0.5–3.0 Walk 3.1–8.0 Low-intensity 8.1–13.0 Medium 13.1–18.0 High-intensity [18.0 Sprint
et al. [26] running or intensity running
trotting running
Hill-Haas [29] 0–6.9 7.0–9.9 10.0–12.9 13.0–15.9 16.0–17.9 [19.1 Sprinting
1033
Table 2 continued
1034

References Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6

Speed Description Speed Description Speed Description Speed Description Speed Description Speed Sprinting
(km
h-1)a (km
h-1)a (km
h-1)a (km
h-1)a (km
h-1)a (km
h-1)a

Hill-Haas et al. 0.0–0.4 Standing/stop 7.9–12.9 Jogging 13.0–17.9 Cruising Sprinting

[30]
Castagna et al. 0–0.4 Standing 0.4–3.0 Walking 3.0–8.0 Jogging 8.0–13.0 Medium- 13.0–18.0 High-intensity [18.0 Sprinting
[25] intensity running
running
Casamichana 0–3.9 Stationary– 4.0–6.9 Jogging 7.0–12.9 Quick 13.0–17.9 High-intensity [18.0 Sprinting
Castellano [28] walking running running
Bucheit et al. [27] \13 Low-intensity 13.1–16.0 High- 16.1–19.0 Very high- [20.0 Very high
running intensity intensity intensity
running running
Harley et al. [16] Standing Walking Jogging Running High-speed
running
Cricket
Petersen et al. Standing/ 7.2–12.6 Jogging 12.6–14.4 Running 14.4–18.0 Striding
[41–43] walking
Hockey
Gabbett [44] 0–3.6 Low intensity 3.6–10.8 Moderate 10.8–18 Moderate 18–25.2 High intensity
intensity intensity
Macutkiewicz 0–0.6 Standing Walking 6.1–11.0 Jogging 11.1–15.0 Running 15.1–19.0 Fast running
and Sunderland
[46]
Jennings et al. 0.36–15 Lower speed [15 Hig-speed
[45] activity running
Lacrosse
Duffield et al. \7.0 Low-intensity 7.0–14.4 Moderate [14.5 High-intensity
[48] activity intensity run
activity
Vmax maximum velocity
a
Except where otherwise stated
C. Cummins et al.
Table 3 Game impacts and body loads in team sports [8, 33, 34, 36]a
References Sample, age Variable Zone 1, light Zone 2, light– Zone 3, mod– Zone 4, heavy Zone 5, very Zone 6, severe Total
GPS Technology in Team Sports

(years) (5–6 G) mod (6–6.5 G) heavy (6.5–7 G) (7–8 G) heavy ([10 G)

[mean (8–10 G)
(SD)]
B F B F B F B F B F B F B F

Rugby Union
Cunniffe n = 2, 5 Impacts [n, 349 563 328 398 55 143 38 101 24 56 4 13 798 1,274
et al. [36] (3.6) mean]
Body load 31,402 119,103
(G) [n, mean]
Body load/min 376 1,426
(G) [n, mean]
Suarez- n = 9, 25.9 Impacts [n, 382 501.6 326 341.3 54.3 161.6 29.8 143.1 35.2 66.6 6.3 (4) 10.4 (5) 883 1,225
Arrones (4) mean (SD)] (129) (106) (173) (219) (28.9) (107) (9) (122) (26) (48)
et al. [33]
Venter [34] n = 17, Impacts [n, 8 12.16 683.4 474.33
18.5 (0.5) mean (SD)] (4.58) (3.18) (295.04) (295.04)
Rugby Leagueb
McLellan n = 17, Impacts [n, 214 215 154 146 334 392 50 47 (24) 26 29 20 (5) 21 (8) 795 (145) 858 (125)
et al. [8] 19.0 (1.3) mean (SD)] (126) (80) (105) (68) (195) (151) (31) (14) (14)
Impacts [n/min, 10 (2) 11 (2)
mean (SD)]c
a
All studies reported on elite subjects in match situations
b
See also Electronic Supplementary Material T2
c
Impacts [n/min, mean (SD)] for zones 4–6 were 1.2 (0.5) for B and 1.2 (0.6) for F
B outside backs (centres and wingers), F forwards, mod moderate, SD standard deviation
1035
1036 C. Cummins et al.

distances averaged 11.3 km, with weekly recorded peaks 3.5.3 Soccer
between 15.4 and 15.8 km, and a week 5 nadir of 6.0 km.
Mooney et al. [22], investigated the relationship between The seven studies in soccer included 286 participants from
physical capacities and match performance, reporting the six male (n = 274) and one female (n = 12) teams. The
total relative distance of 139.0 ± 11.1 m
min-1, whereas studies reported on junior (n = 28; aged 11.2–16.2 years),
the relative distance at high intensity was 40.6 ± elite junior (n = 232; aged 13.0–18.0 years), elite under 19
9.6 m
min-1. (n = 16; aged 15.7–16.9 years), and semi professional
The studies in AFL demonstrate the wider use of GPS (n = 10; aged 22.0–23.0 years) athletes. In regard to work
utilization for monitoring the physical demands of playing rate patterns, the soccer studies were similar to those in
positions during competition game quarters and the finals as AFL, with a focus upon sprint time and sprint distance
well as in training situations. Six manuscripts reported com- measures (see Electronic Supplementary Material, T5). As
petition [17–20, 22, 24] and two [21, 23] appraised training with other football codes, the soccer manuscripts reported
data. Activity profiles from training showed that open drills on both training and game data, with five manuscripts
proved more physically demanding than closed [21], with reporting exclusively on game [16, 25–28] and two
*20 % more moderate-velocity efforts in open drills. exclusively on training variables [29, 30].

3.5.2 Rugby League and Rugby Union 3.5.4 Cricket

Rugby league (n = 6) [5, 8, 37–40] and rugby union The three cricket manuscripts reported on 96 male, elite
(n = 6) [31–36] manuscripts were grouped together due to (n = 42; aged 19.3–24.9 years), State (n = 42; aged
the similar nature of the codes (see Electronic Supple- 23.8–30.4 years), and international (n = 12; aged
mentary Material, T4). These manuscripts included 470 26.2–33.4 years) participants. The studies examined the
junior to elite male athletes; rugby league (elite: n = 276, work rate patterns of speed, distance, sprint distance, and
aged 16.9–31.5 years; amateur: n = 11, aged 18–22.6 the number of high-intensity efforts at the higher speeds
years) and rugby union (junior: n = 135, aged 14–18 years; (see Electronic Supplementary Material, T6). In contrast to
semi-professional: n = 26, aged 18–29.9 years; elite: the football codes, the cricket manuscripts only report on
n = 21, aged 18.5–28.6 years). Two manuscripts did not competition-based GPS variables [41–43]. The observed
report on movement demand variables. Hartwig and co- consistency of speed zone definitions and descriptors in
authors [31] compared distance in match performance of these studies is the result of singular authorship.
junior rugby union players (schoolboys, talent, and repre-
sentative squads) (Fig. 2). McLellan et al. [8] explored the 3.5.5 Hockey, Lacrosse, and Netball
intensity, number, and distribution of impacts associated
with collisions in rugby league competition matches. Five manuscripts reported on 71 participants and covered the
Interestingly, recent research within these two codes has remaining team sports of hockey (n = 55 elite; aged
progressively further investigated the demands of impacts 18.0–29.7 years) [44–46], lacrosse (n = 7; National division,
and collisions on players using GPS (Table 3). aged 18.5–21.3 years) [48], and netball (n = 9; elite State
The work-to-rest ratio by positional play evaluated in one level, aged 18–27.2 years) [47]. Because of the limited
study [5] indicated that rugby league forwards completed number of manuscripts in these sports, they are summarized
slightly more work than backs (1:7 and 1:6, respectively). together in the Electronic Supplementary Material, T7 . Three
Conversely, a study in rugby union [36] indicated that backs [44, 46, 47] of the four papers report exclusively on female
produced marginally more work than forwards (1:5.7 and athletes (n = 48 participants; 67.6 %) in netball and hockey.
1:5.8, respectively). Similar to AFL, all six papers from these GPS capability was used in elite hockey to assess the
football codes reported on movement demand variables. influence of international competitions [45] and playing
However, by comparison, rugby union papers focused lar- position [46] and compared game-based training versus
gely on distance travelled over the game as opposed to sprint competition [44] on player activity profiles. Two studies
distance and time within each speed zone. The football codes utilized GPS to evaluate the efficacy of distinct interven-
demonstrate an increased utilization for monitoring both tions on exercise performance. They examined part-body
training and game situations, demonstrating similarities with pre-cooling procedures on physiological responses to
AFL. Nine manuscripts reported exclusively on competition lacrosse training in the heat [48] and the effect of com-
matches [5, 8, 33–38, 40], one exclusively on training [31] pression garments on physiological demand and sprint
and two on combined training and game-based data [32, 39] performance in a simulated game-specific circuit for net-
(see Electronic Supplementary Material, T4). ball [47] (see Electronic Supplementary Material, T7).
GPS Technology in Team Sports 1037

4 Discussion rugby league, whilst zone 6 ranged from[18 to 36 km
h-1

in AFL). Activity descriptors also varied, with zone 2
To our knowledge, this is the first systematic review descriptions ranging from jogging and walking in rugby
evaluating the application of GPS technology to classify union to jogging and cruising in rugby league.
work rate patterns and physical impact demands across Abt and Lovell [50] suggested that, as players differ in
team sports. The review clearly demonstrates that existing the speed at which they begin to run at high intensity, there
GPS classification definitions vary widely both within and is a need for individualized high-intensity speed thresholds
between sports. Lack of uniformity in movement demand and an absolute maximum intensity threshold of
descriptors and classification of speed zones limits com- 15 km
h-1. Waldron and colleagues [6] highlight that
parisons within the same team sport. Consensus on a players are influenced not only by the field position they
consistent GPS movement, physical demand, and external possess, but also by the possibility that positional playing
work criteria for individual team sports may help facilitate groups display different speed characteristics (e.g., in
comparisons within individual sporting disciplines and rugby league, forwards are slower than backs and adju-
levels of participation. The literature for GPS use in team stables) [5, 37, 39]. The difference in field position and
sports remains limited and is strongly biased towards adult speed characteristics prevents the development of an arbi-
male athletes participating in football codes (AFL, rugby trary sprint zone classification, whilst the suggestion of
league, rugby union, and soccer). The majority of the data individualized speed zones poses both a logical and a
have been collected in competition game situations, with logistical problem in speed zone determination [6].
limited training data available, perhaps attributable to the
fact that many professional teams are reluctant for such 4.2 Distance and Relative Distance
information to be placed in the public domain.
Reporting of the GPS parameter total distance covered per
4.1 Zone Classifications game or training session provides for little comparison
between team sports. Comparing distance travelled in
Clarity and consistency of speed zones and movement sports may only be meaningful when reference to the time
pattern definitions within a team sport facilitates precise spent in activity or game play is made; for example, elite
comparison and analysis of performance between players, AFL players covered 12,939 m per game in comparison
teams, levels of competition, and seasons. Furthermore, with 4,302 m by elite rugby union backs [36], suggesting
with speed and impacts divided into the six activity bands, AFL players [20] covered 67 % greater distance. However,
the upper zones (4–6) provide more insightful information when the time spent in activity was accounted for, AFL
about the physical stressors and demands experienced by players recording a relative distance of 139 m
min-1 [22]
players. It is important to note that GPS reliability is in fact covered 48 % more ground than rugby union backs,
reduced with increased movement intensity [20 km
h-1 with a relative distance of 71.9 m
min-1 [36]. Thus, the
[3]. It is likely this uncertainty is due to the rapid changes time on field is most certainly a key metric to include,
in velocity that are apparent in movements performed at enabling qualification of distance covered not only between
higher speeds [11]. Although GPS reliability is reduced but also within sports.
with increased movement intensity, validity is improved The significance of using the relative distancemetric
through increased distance; for example, the standard of was identified in a recent review [9]. Although cricket fast
error is reduced by 67 % when comparing sprinting over bowlers appeared to run greater distances than AFL
40- and 10-m distances [11]. However, a 5-Hz non-dif- players, the calculated relative distances revealed a higher
ferential GPS has been reported by Portas et al. [14] to be a work intensity in AFL (63 and 109 m
min-1 for cricket
reliable method (CV 55 %) of monitoring overall distance and AFL, respectively).The use of estimated game time
during multiple sprint team sports-specific movements. (i.e. via use of the game and rule durations) presents a
However, due to the variability of exercise activity, GPS limitation to this approach. Only a minority of included
devices, sample rates, and statistical methods applied, manuscripts have reported the actual player match time
direct comparison of GPS validity within team sports is (time on field including stoppages). Data should be
difficult [9]. reported relative to precise field time and not include time
Movement demand classifications across all sports are on the bench (e.g. from injury or substitution) in order to
derived from the manufacturer or modified from rugby minimize the margin of error when reporting distance per
union [49]. Classifications from rugby union potentially minute.
fail to adequately reflect the demands of other football Stoppages account for 33 % of the overall match time in
codes (e.g., zone 1 speed criteria ranged from 0 to senior elite rugby league matches, with an average of 48 s
5.9 km
h-1 in rugby union and from 0 to 12 km
h-1 in per stoppage [51]. The majority of manuscripts to date
1038 C. Cummins et al.

have failed to account for player time on the field in has not been fully examined. GPS devices do not have the
competition or training sessions. Reporting of distance capacity to measure the forces upon the body without
relative to match time is strongly recommended. Increased integrated accelerometer sensing devices [53]. When
ease and ability of GPS devices in allowing users to take combined with a triaxial accelerometer, information relat-
account of various stoppages in play (i.e. injury or inter- ing to the physical loads and impacts can be measured
change) could further facilitate this reporting. through the acceleration and deceleration forces experi-
enced by the player. Although the work rate patterns such
4.3 Positional Play as speed and distance are automatically detected and
reported by GPS technology, the analysis of loads experi-
Some studies that inform this review have compared the enced through tackles or collisions is relatively complex.
physical demands by player position. The work-to-rest One GPS unit (Catapult SportsTM) equipped with tackle
ratio by positional play was investigated in one rugby detection technology, allows for the detection of tackles;
league [5] and one rugby union [36] manuscript. Rugby however, it is limited by the fact that the technology does
league forward playing positions completed more work not distinguish between the types of tackles, such as
than back playing positions (1:7 and 1:6, respectively), whether it is a multiple player tackle, or the position of
whilst rugby union showed backs to produce slightly more contact to the player. Complete analysis of tackle and
work than forwards (1:5.7 and 1:5.8, respectively). The contact events still requires manual pairing of GPS loads
distance in maximal speed by positional play has also been and video footage. Measurement of attack- and defence-
explored in rugby league [5], where backs attained a 20 % specific impact events would allow for the evaluation and
higher maximal sprint speed than forwards (30.6 and quantification of the loads sustained through individual
24.5 km
h-1, respectively). It should be noted that the tackles, the cumulative load experienced through a game,
findings might be attributed to the field position and indi- training session, or season and the forces experienced
vidual sprint characteristics of positional groups [6]. Sprint during injury.
performance appears to be influenced through field posi-
tion, whereby outside backs are presented with large areas 4.5 Global and Sport-Specific Use of GPS
of open space in which to develop locomotion, in com-
parison with forwards, who are more often in closer The available literature on GPS use in team sports dem-
proximity (10 m) [6] to the opposition and involved in onstrates a strong bias to Australian studies (20 of 27
increased contacts [52], which diminishes their opportunity manuscripts, 74.1 %), indicating that its use, certainly for
to generate higher speeds. research, has proliferated. This could be due to a number
The GPS metric of distance has also been explored in of the publications emanating from the same research
regards to positional play in hockey [44]. The distance groups. It may also be that the application of GPS to team
covered at high-intensity running showed midfielders sports favors popular sports in Australia, such as AFL,
covered more ground per match (77 m) than strikers rugby league, rugby union, and cricket, or that local or
(46 m) and defenders (52 m), due to greater game researcher interest in the technology has encouraged
involvement moving the ball up and down the field as the greater use in these disciplines. Limited research within
link between defensive and attacking players. Positional soccer may possibly be attributed to the soccer federation
play data enable the provision of position-specific training not permitting the use of GPS within professional com-
programs and targets such as distance and maximal petition matches. However, it is clear that GPS technol-
speeds. Further positional data allows for player perfor- ogy has a more logical application to certain team sports
mance to be evaluated against positional-specific estab- than others. Sports in which there is movement across a
lished benchmarks. substantial distance of varied speed and player-to-player
contact or collisions likely lends itself to monitoring by
4.4 Specific Application and Usefulness of GPS GPS compared with other team sports such as water polo
Metrics or volleyball. Use of GPS in stadiums with high walls and
curved roofs may provide unreliable data because fewer
Although GPS metrics provide quantitative data on a range satellites are available to triangulate signals from devices.
of physical demand variables in team sports, the relative GPS is also less reliable for indoor sports and for those
load on players may be difficult to compare across sports where short distances and changes in direction are pre-
due to differences in the interchange of players, game dominant [54]. Although this technology is less able to
durations, and conditions. Specific metrics may have attain satellite connections when used indoors and there-
greater application in some team sports than others. The fore unable to provide the parameters of distance or
accuracy of GPS for evaluating impact and collision data speed, newly developed GPS units equipped with new
GPS Technology in Team Sports 1039

indoor functions are currently being assessed within within the demands of open and closed drills, where open
indoor sports such as basketball to examine the player drills prove more physically demanding. However, these
load and heart rate zones throughout training and game were just drills from one club and one manuscript [21],
situations. such that more research is needed to adequately describe
the demands of different training drills using GPS. This
4.6 Use of GPS Across Different Age Groups detailed understanding of training drills and match
and Levels of Athletic Calibre demands enables for the provision of individually tailored
training programs that more accurately reflect competition
The varying age of athletes from junior/prepubescent to demands and ensures athletes reach optimal training tar-
post pubescent adults may require the addition of age- gets. It may be of value to report the time of year in which
appropriate speed zones so as to more accurately reflect the training has occurred, as sports are periodized, with ath-
work rate patterns of both pre and post pubescent athletes. letes training less and at lower intensities during the season
Speed zone classifications defined and designed around or when recovering from injury, than during the pre season.
post pubescent adult athletes may not be suitable for, or The major benefit of GPS technology within competition
adequately reflect the competition or training demands of, settings is the lightweight, unintrusive nature of the mon-
prepubescent athletes due to the inherent physiologic, itoring units with the ability to provide real-time movement
biomechanical, and metabolic differences during exercise. demand information (e.g. distance and speed) to coaches
Compared with adults, children and young adolescents and training staff.
exhibit a lower energy reserve between submaximal and In competition settings, the provision of realistic impact
maximal aerobic exercise; that is, an increased cost of and load data has eluded sports scientists for some time.
locomotion—for a given running speed, they will work at a Currently, the only mechanism available for in-depth
higher percent of their maximal aerobic capacity [55]. analysis of tackle and collision events is to manually label
Metabolic differences between children/prepubescent ado- impact data through cross referencing video footage with
lescents and adults during exercise are due to three phys- the GPS and accelerometer measurements. This is a time-
iological factors. Children/prepubescent adolescents have consuming process, which needs to take place for each
lower walking/running economy due to shorter legs individual player, making it impossible to provide coaches
requiring a higher stride frequency and smaller stride with real-time tackle or impact information [49].
length [56]. They also display less efficient running
mechanics (higher peak ground reaction forces, greater 4.8 Use of GPS with Injuries
braking forces, greater vertical movements) [55, 57] and
less efficient co-contraction of antagonist muscles where The use of GPS metrics and training practices should
there is a less than optimal neuromuscular control to syn- represent a balance between developing the individual for
chronize action between muscle groups [58]. Furthermore, the specific and positional demands of team sports, and the
they have a mass–speed imbalance, where the lighter child maximum training load that can be sustained by the indi-
cannot effectively match their load to an imposed speed of vidual before a marked increase in the likelihood of injury
movement [59, 60]. Energy costs associated with exercise [62]. Piggott et al. [23] observed that with professional
decline steadily throughout childhood and into late ado- athletes, 42 % of illnesses and 40 % of injuries could be
lescence [61]. explained by a preceding spike in training load [23]. The
The varying calibre of athletes from non-elite to elite optimal amount of training should not exceed an individ-
may also require the addition of reduced speed zone ual’s exercise tolerance and capability for recovery. A
brackets, reflective of the athlete’s differing ability and recently developed injury-prediction model showed that
skill level. Although these data were informed by a rela- rugby league players exceeding a training threshold were at
tively small body of literature [32] in rugby union, it was significantly higher risk of soft tissue injuries [63].
seen that semi-professional [34] forward players covered Accordingly, GPS technology can be used to inform and
17.3 m during sprinting, whilst junior forward players regulate training loads for individual athletes at crucial
covered 28.8 % less distance with 12.3 m of sprinting. parts during the competitive season. Further research into
this area is warranted.
4.7 Use of GPS Within Training and Competition
4.9 Use of GPS and Fatigue
The football codes demonstrate an increase in GPS utili-
zation for monitoring both training and game situations. Research is emerging on the monitoring of fatigue through
AFL training data indicate activity profile differences GPS. Higham and colleagues [35] examined the movement
1040 C. Cummins et al.

patterns between full match and substitute players, 4.11 Limitations

showing that substitute players exhibited a substantially
greater work rate than players who played the entire A limitation of this review revolves around the comparison
match. Substitutes were seen to cover 24 % greater dis- of speed zones and classifications due to the large vari-
tance per minute, and an additional 110 and 123 % ability of descriptors used. The validity and reliability of
greater distance at 5–6 and [6 m
s-1, respectively. The different GPS devices used and their many variables may
authors suggest that players pace their activity through- be a major contributor. There is a strong sex bias in the
out a match so as to mitigate fatigue later in the match studies to recruitment of male athletes, so the results may
and that the replacement of one or more players at a key not be generalizable to females. Speed zone classifications
stage of a match could potentially influence a match defined and designed around post pubescent adult athletes
result. may not be suitable for, or adequately reflect the compe-
The ability of GPS to monitor activity demands and tition or training demands of, prepubescent or adolescent
fatigue development during a soccer match has recently athletes due to the inherent physiologic differences.
been examined [64]. The game–fatigue index (the differ- Another limitation is that some manuscripts informing this
ence in high-intensity running between the first and last 15 review report on small sample sizes (fewer than ten) of
minutes of the game) was 49.9 ? 25.7 and 44.7 ? 27.2 %, athletes from within a team or training institution; whereby
for different GPS devices, and not significantly different results may not be representative of the entire population.
from either a video-based time–motion analysis system or a Additionally, there is a possible bias towards Australian
semi-automatic multiple-camera system. Whilst this study research (26 of 35 studies). The start of 2012 saw a surge in
proposes that GPS is a valid measure of fatigue develop- research in rugby league and rugby union. Although these
ment, it has also recently been reported [65] that pacing latest papers [8, 33, 34, 39] show an increased trend for
strategies are difficult to discern from time–motion data. analysing the contact events through quantification of
The use of the initial 15-min phase of the match as a cri- impact loads, the use of the resultant G force as a precise
terion for match-related fatigue is highly questionable, measure of player-to-player collisions is yet to be explored
thereby providing a stimulus for more robust analytic and determined.
approaches to research into the measurement of fatigue in
the field. 4.12 Future Directions

4.10 Comparison of GPS with Other Technologies The introduction of and improvements in GPS technology
Assessing Time–Motion Analysis have provided for a shift from descriptive studies of
movement patterns to comparisons between levels of
This review has not reported on the evaluation or com- competition, training, and match situation work rate pat-
parison of GPS with other technologies assessing move- terns of distance, speed, and speed zones. The use of GPS
ment patterns as it is beyond the scope of this review. No parameters to detect injury, fatigue, and overtraining could
‘gold standard’ method for determining movement patterns determine the timing of return to play, inform provision of
and workload in sport has been ascertained [64]. Two supplementary or complementary conditioning, or help to
studies [64, 66] have compared GPS with other technolo- better characterize training drills.
gies, such as video capture and semi-automated image The physiological demands and characteristics of junior
recognition systems, in assessing work rate patterns. These and subelite athletes should be investigated so as to define
studies observed that the absolute agreement between modified speed zones that more accurately reflect the age
measurements produced from each of the systems is not and skill level of these athletes. Further utilization of GPS
well established. Randers et al. [64] compared four match- technology in junior athletes may facilitate the develop-
analysis methods (two different GPS systems, video cap- ment of younger players. The GPS variables can also be
ture, and semi-automated multiple-camera image-recogni- used to formulate rule modifications, age limits, or training/
tion systems) during the same football match. The findings competition parameters of junior to elite athletes. In addi-
highlighted that although similar performance decrements tion, GPS measurements may inform coaches and trainers
were detected, between-system differences in absolute of methods for the introduction and preparation of younger
running distances covered were present. The authors cau- players to the impacts experienced at elite level. By mon-
tioned that differences in performance measurements itoring the physiological stress on each athlete, GPS tech-
should be taken into account when comparing results col- nologies allow sports scientists, coaches, and trainers to
lected by different methods of time–motion analysis optimize and individualize training and potentially dimin-
systems. ish the incidence of injury.
GPS Technology in Team Sports 1041

Finally, the use of live analysis will provide instanta- 10. Varley M, Aughey R. Validity and reliability of GPS for mea-
neous feedback to coaches of player work rate patterns in suring instantaneous velocity during acceleration, deceleration,
and constant motion. J Sports Sci. 2012;30(2):121–7.
order to monitor fatigue and allow for adjustments to 11. Jennings D, Cormack S, Coutts AJ, et al. The validity and reli-
player performance or effect of player rotations. Future ability of GPS units for measuring distance in team sport specific
research on impact classification zones, player impacts, and running patterns. Int J Sports Physiol Perform. 2010;5(3):
body loads within game and training situations is war- 328–41.
12. Johnston R, Watsford M, Pine M, et al. The validity and reli-
ranted. Quantification of these parameters to coaches in ability of 5-Hz global positioning system units to measure team
real-time could provide data for instantaneous modification sport movement demands. J Strength Cond Res. 2012;26(3):
to play and tactical substitutions based upon loads, impacts, 758–65.
and the physiological limits of players, as well as improve 13. Portas M, Rush C, Barnes C, et al. Method comparison of linear
distance and velocity measurements with global positioning
injury surveillance of athletes. satellite (GPS) and the timing gates techniques. J Sci Med Sport.
2009;4:381–93.
14. Portas M, Rush C, Barnes C, et al. The validity and reliability of
5 Conclusion 1-Hz and 5-Hz global positioning systems for linear, multidi-
rectional, and soccer-specific activities. Int J Sports Physiol
Perform. 2010;5:448–58.
GPS utilization within the sporting context has facilitated 15. Downs SH, Black N. The feasibility of creating a checklist for the
collection of metrics that describe the physical demands assessment of the methodological quality both of randomised and
and impacts on players, both at training and in competition non-randomised studies of health care interventions. J Epidemiol
Community Health. 1998;52(6):377–84.
settings. Use of the technology for this purpose is still 16. Harley JA, Barnes CA, Portas M, et al. Motion analysis of match-
relatively new and evolving. This review provides insight play in eltie U12 to U16 age-group soccer players. J Sports Sci.
into the depth and span of its use within team sports and 2010;28(13):1391–7.
some of the existing limitations with the metrics reported, 17. Aughey RJ. Increased high-intensity activity in elite Australian
Football finals matches. Int J Sports Physiol Perform.
and with high-speed running involving rapid directional 2011;6(3):367–79.
changes. 18. Aughey RJ, Falloon C. Real-time versus post game GPS data in
team sports. J Sci Med Sport. 2010;13(3):348–9.
Acknowledgments This manuscript contributes to CC’s PhD 19. Brewer C, Dawson B, Haesman J, et al. Movement pattern
qualification. No funding has been received for the preparation of this comparisons in elite (AFL) and sub-elite (WAFL) Australian
manuscript. The authors declare that there are no conflicts of interest Football games using GPS. J Sci Med Sport. 2010;13(6):618–23.
that are directly relevant to the context of this review. 20. Coutts AJ, Quinn J, Hocking J, et al. Match running performance
in elite Australian Rules football. J Sci Med Sport.
2010;13(5):543–8.
21. Farrow D, Pyne D, Gabbett T. Skill and physiological demands of
References open and closed training drills in Australian Football. Int J Sports
Sci Coach. 2008;3(4):489–99.
1. Larsson P. Global positioning system and sport-specific testing. 22. Mooney M, O’Brien B, Cormack S, et al. The relationship
Sports Med. 2003;33(15):1093–101. between physical capacity and match performance in elite Aus-
2. Schutz Y, Herre R. Assessment of speed of human locomotion tralian Football: a mediation approach. J Sci Med Sport.
using a differential satellite global positioning system. Med Sci 2011;14(5):447–52.
Sports Exerc. 2000;32:642–6. 23. Piggot B, Newton M, McGuian M. The relationship between
3. Gray AJ, Jenkins D, Andrews MH, et al. Validity and reliability training load and incidence of injury and illness over a pre season
of GPS for measuring distance travelled in field-based team at an Australian Football League club. J Aus Strength Cond.
sports. J Sports Sci. 2010;28(12):1319–25. 2009;17(3):4–17.
4. Schutz Y, Chambaz A. Could a satellite-based navigation system 24. Wisbey B, Montgomery PG, Pyne DB, et al. Quantifying
(GPS) be used to assess the physical activity of individuals on demands of AFL football using GPS tracking. J Sci Med Sport.
earth? Eur J Clin Nutr. 1997;51(5):338–9. 2010;13(5):531–6.
5. McLellan CP, Lovell DI, Cass GC. Performance analysis of elite 25. Castagna C, Impellizzeri F, Cecchini E, et al. Effects of inter-
Rugby League match play using global positioning systems. mittent-endurance fitness on match performance in young male
J Strength Cond Res. 2011;25(6):1703–10. soccer players. J Strength Cond Res. 2009;23(7):1954–9.
6. Waldron MT, Highton C, Worsolf J, et al. Movement and 26. Barbero Alvarez J, Lopez M, Barbero Alvarez V, et al. Heart rate
physiological match demands of elite Rugby League using por- and activity profile for young female soccer players. J Hum Sport
table global positioning systems. J Sports Sci. 2011;29(11): Exerc. 2008;3(2):1–11.
1223–30. 27. Bucheit M, Mendez-Villanueva A, Simpson BM, et al. Match
7. GPSports Systems. GPSports team analysis user manual. Version running performance and fitness in youth soccer. Int J Sports
1.5;2006. Med. 2010;31(11):818–25.
8. McLellan CP, Lovell DI, Gass GC. Biochemical and endocrine 28. Casamichana D, Castellano J. Heart rate and motion analysis by
responses to impact and collision during elite Rugby League GPS in beach soccer. J Sports Sci Med. 2010;9:98–103.
match play. J Strength Cond Res. 2011;25(6):1553–62. 29. Hill-Haas SV, Coutts AJ, Rowsell GJ, et al. Variability of acute
9. Aughey RJ. Applications of GPS technologies to field sports. Int J physiological responses and performance profiles of youth soccer
Sports Physiol Perform. 2011;6(3):295–310. players in small-sided games. J Sci Med Sport. 2008;11(5):487–90.
1042 C. Cummins et al.

30. Hill-Haas SV, Dawson BT, Coutts AJ, et al. Physiological 49. Docherty D, Wegner HA, Neary P. Time-motion analysis related
responses and time-motion characteristics of various small-sided to the physiological demands of rugby. J Hum Mov Stud.
soccer games in youth players. J Sports Sci. 2009;27(1):1–8. 1988;14:269–77.
31. Hartwig T, Naughton G, Searl J. Defining the volume and 50. Abt G, Lovell R. The use of individualized speed and intensity
intensity of sport participation in adolescent Rugby Union play- thresholds for determining the distance run at high-intensity in
ers. Inter J Sports Physiol Perform. 2008;3:94–106. professional soccer. J Sports Sci. 2009;27(9):893–8.
32. Hartwig T, Naughton G, Searl J. Motion analyses of adolescent 51. Sykes D, Ceri N, Lamb K, et al. An evaluation of the external
Rugby Union players: a comparison of training and game validity and reliability of a rugby league match simulation pro-
demands. J Strength Cond Res. 2011;25(4):966–72. tocol. J Sports Sci. 2012;31(1):48–57.
33. Suárez-Arrones JL, Portillo LJ, Gonzalez-Rave MJ, et al. Match 52. King D, Hume P, Milburn P, et al. Match and training injuries in
running performance in Spanish elite male Rugby Union using Rugby League: a review of published studies. Sports Med.
global positioning system. Isokinetic Exer Sci. 2012;20:77–83. 2010;40(2):163–78.
34. Venter R, Opperman E, Opperman S. The use of global posi- 53. Kelly D, Coughlan FG, Green SB, et al. Automatic detection of
tioning system (GPS) tracking devices to access movement collisions in elite level Rugby Union using a wearable sensing
demands and impacts in under-19 Rugby Union match play. Afr J device. Sports Eng. 2012;15:81–92.
Phys Health Ed Rec Dance. 2011;17(1):1–8. 54. Dogramaci SN, Watsford ML, Murphy AJ. The reliability and
35. Higham DG, Pyne DB, Anson JM, et al. Movement patterns in validity of subjective notational analysis in comparison to global
Rugby sevens: effects of tournament level, fatigue and substitute positioning system tracking to assess athlete movement patterns.
players. J Sci Med Sport. 2012;15(3):277–82. J Strength Cond Res. 2011;25(3):852–9.
36. Cunniffe B, Proctor W, Barker JS, et al. An evaluation of the 55. Ebbeling C, Hamill J, Freedson P, et al. An examination of
physiological demands of elite Rugby Union using global posi- efficiency during walking in children and adults. Pediatr Exerc
tioning system tracking system. J Strength Cond Res. 2009; Sci. 1992;4(1):36–49.
23(4):1195–203. 56. Rowland TW, Auchinachie JA, Keenan TJ, et al. Physiologic
37. Austin DJ, Kelly SJ. Positional differences in professional rugby responses to treadmill running in adult and prepubertal males. Int
league match play through the use of global positioning systems. J Sports Med. 1987;8(4):292–7.
J Strength Cond Res. 2013;27(1):14-9. 57. Grieve DW, Ruth GJ. The relationships between length of stride,
38. Duffield R, Murphy A, Snape A, et al. Post match changes in step frequency, time of swing and speed of walking for children
neuromuscular function and the relationship to match demands in and adults. Ergonomics. 1966;9(5):379–99.
amateur Rugby League matches. J Sci Med Sport. 2012;15(3): 58. Frost G, Dowling J, Dyson K, et al. Cocontraction in three age
238–43. groups of children during treadmill locomotion. J Electromyogr
39. Gabbett TJ, Jenkins DG, Abernethy B. Physical demands of Kinesiol. 1997;7(3):179–86.
professional rugby league training and competition using mic- 59. Davies CT. Metabolic cost of exercise and physical performance
rotechnology. J Sci Med Sport. 2012;15(1):80–6. in children with some observations on external loading. Eur J
40. McLellan CP, Lovell DI, Cass GC. Creatine kinase and endocrine Appl Physiol Occup Physiol. 1980;45(2–3):95–102.
responses of elite players pre, during and post rugby league match 60. Thorstensson A. Effects of moderate external loading on the
play. J Strength Cond Res. 2010;24(11):2908–19. aerobic demand of submaximal running in men and 10 year-old
41. Petersen CJ, Pyne DB, Dawson B, et al. Movement patterns in boys. Eur J Appl Physiol Occup Physiol. 1986;55(6):569–74.
cricket vary by both position and game format. J Sports Sci. 61. Ariens GA, Mechelen W, Kemper HC, et al. The longitudinal
2010;28(1):45–52. development of running economy in males and females aged
42. Petersen CJ, Pyne DB, Portus MR, et al. Comparison of player between 13 and 27 years: the Amsterdam Growth and Health
movement patterns between 1-day and test cricket. J Strength Study. Eur J Appl Physiol Occup Physiol. 1997;76(3):214–20.
Cond Res. 2011;25(5):1368–73. 62. Gabbett TJ, Domrow N. Relationships between training load,
43. Petersen CJ, Pyne DB, Portus MR, et al. Variability in movement injury, and fitness in sub-elite collision sport athletes. J Sports
pattern during one day internationals by a cricket fast bowler. Int Sci. 2007;25(13):1507–19.
J Sports Physiol Perform. 2009;4(2):278–81. 63. Gabbett TJ. The development and application of an injury pre-
44. Gabbett TJ. GPS analysis of elite women’s field hockey training diction model for noncontact, soft-tissue injuries in elite collision
and competition. J Strength Cond Res. 2010;24(5):1321–4. sport athletes. J Strength Cond Res. 2010;24(10):2593–603.
45. Jennings DH, Cormack SJ, Coutts AJ, et al. International field 64. Randers M, Mujikab I, Hewitt A, et al. Application of four dif-
hockey players perform more high speed running than national ferent football match analysis systems: a comparative study.
level counterparts. J Strength Cond Res. 2012;26(4):947–52. J Sports Sci. 2010;28(2):171–82.
46. Macutkiewicz D, Sunderland C. The use of GPS to evaluate 65. Lovell R, Barrett S, Portas M, et al. Re-examination of the post
activity profiles of elite women hockey players during match- half-time reduction in soccer work-rate. J Sci Med Sport.
play. J Sports Sci. 2011;29(9):967–73. 2013;16:250–4.
47. Higgins T, Naughton GA, Burgess D. Effects of wearing com- 66. Waldron M, Worsfold P, Twist C, et al. Concurrent validity and
pression garments on physiological and performance measures in test–retest reliability of a global positioning system (GPS) and
a simulated game-specific circuit for netball. J Sci Med Sport. timing gates to assess sprint performance variables. J Sports Sci.
2009;12(1):223–6. 2011;29(15):1613–9.
48. Duffield R, Steinbacher G, Fairchild TJ. The use of mixed-
method, part body pre-cooling procedures for team-sport athletes
in the heat. J Strength Cond Res. 2009;23(9):2524–32.

View publication stats