The Current Use of GPS, Its Potential, and Limitations in Soccer

Liam Hennessy, PhD1,2 and Ian Jeffreys, PhD, FNSCA, CSCS*D, RSCC*D3 1Setanta College, Thurles, Co Tipperary, Ireland; 2Department of Sport and Exercise, University of South Wales, Pontypridd, Wales; and 3Department of Sport and Exercise, University of South Wales, Pontypridd, Wales

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (http://journals.lww.com/nsca-scj).

ABSTRACT multitude of fitness components GLOBAL POSITIONING SYSTEM needed to perform at a high level. DEFINED IN SOCCER, GLOBAL POSITION- Added to this are the logistical chal- GPS is a satellite-based navigation sys- ING SYSTEM (GPS) MONITORING lenges of developing these capacities tem and is made up of over 30 orbiting OF PLAYER WORKLOADS IS NOW in conjunction with the extensive tech- satellites. The system was originally EXTENSIVELY USED ACROSS ALL put into orbit by the U.S. Department nical and tactical training a player will LEVELS OF THE SPORT. TO MAKE of Defense for military use in the early need to undertake, on a daily basis. BETTER USE OF THIS TECHNOL- 1970s. Later, in the 1980s, it was made This challenge is further exacerbated OGY IT IS IMPORTANT TO APPRE- available for civilian use. The orbiting by the increasing length of playing sea- CIATE HOW IT WORKS. FURTHER, satellites transmit unique signals that sons and an increasing density of fix- WHEN THE LIMITATIONS OF GPS allow GPS devices to locate the satel- tures within these seasons (42). USE ARE APPRECIATED AND THE lites precise location. Essentially, the RATIONALE OF USE IS AGREED Partly in response to these develop- GPS receiver calculates the distance AND ARTICULATED, THEN THE ments, there has been an increased to each satellite by the amount of time POTENTIAL OF GPS MONITORING focus on the overall monitoring of it takes to receive a transmitted signal CAN BE EFFECTIVELY REALISED player workloads in an attempt to with the result that the user’s location TO BETTER MANAGE PLAYERS understand the stress placed on the can then be precisely determined. PERFORMANCE, WORKLOAD AND player during training and match play, Once the receiver links with a number of orbiting satellites, a GPS receiver WELFARE. (SEE VIDEO, SUPPLE- with the goal of maximizing perfor- can provide a reasonably exact position MENTARY DIGITAL CONTENT, mance and minimizing the risk of and speed of the receiver device. There NUMBER 1, WHICH SUMMARIZES injury (30,70). To facilitate this, there are 2 main communication systems GPS USE, LIMITATIONS, AND has been a great increase in the use associated with GPS technology. The POTENTIAL IN SOCCER, HTTP:// of global positioning system (GPS) direct satellite to receiver device sys- LINKS.LWW.COM/SCJ/A238). technology (31). However, although this technology is now used by many tem of communication described clubs at both amateur and professional above is known as nondifferential, INTRODUCTION levels, it is important to appreciate the whereas the use of ground-based cor- ith over 265 million partici- limitations of this technology. Further- rection techniques to enhance the pants, soccer is the most more, it is important to understand quality of location data gathered using Wpopular sport in the world how the technology works and how GPS receivers is known as a differential across both sexes and all age groups related metrics are derived. Finally, it system (44,45). The important point to (25). The sport provides one of the is equally important to appreciate the KEY WORDS: greatest challenges to the strength potential of GPS and related technol- technology; global positioning and conditioning coach with the ogy in supporting the goals to improve system; monitoring; potential; limita- player performance and ultimately, to Address correspondence to Dr. Liam Hen- tions; soccer nessy. [email protected]. ensure better player welfare.

Copyright Ó National Strength and Conditioning Association Strength and Conditioning Journal | www.nsca-scj.com 83 The Current Use of GPS

note is that although the nondifferen- currently derived from integrated GPS cumulative concept, be they physical tial system can provide useful data, and MEMS technology (Table 1). or otherwise. The quantification and there will always be a degree of error There is no doubt that GPS technology use of these workload data within which the practitioner needs to be has facilitated a greater understanding a broader concept of workload is a desir- aware of (44). Some of this can come of the physical activity demands of dif- able goal for all involved in sport but it is for the sampling frequency of the ferent team formations and positional also a challenge for the support staff of device and others from factors that profiling during match play, as well as a team of players. affect the signal, as the signal from training, and when integrated with the satellites can be influenced by the internal responses such as heart rate RELIABILITY AND VALIDITY atmosphere and obstructions such as (HR) monitoring, a more holistic tall buildings (45). Differential GPS workload evaluation may be attained As with all measuring devices in sport, can help to reduce this error. Differen- (1,14,53,61,63,68). it is important to assess both reliability tial GPS uses stationary receivers and validity of GPS and related devices. placed at known locations on the EARLY WORKLOAD MONITORING GPS devices are classified by the rate at IN SPORT ground. These fixed-position receivers which they sample per second. In the are then used to facilitate better com- Foster et al. (26,27) established the val- early years of GPS use in sports, the munication with orbiting satellites (45). idity of using “rating of perceived exer- devices used a sampling rate of 1 Hz tion” or RPE as a tool to monitor or 1 sample per second (3,43). When EVOLUTION OF GLOBAL physical workload. Both team and indi- such devices were assessed for reliability POSITIONING SYSTEM vidual sports have since used this and validity, studies reported that these TECHNOLOGY method to establish exercise workload devices were somewhat unreliable and One of the first studies using GPS tech- (34). Studies also reported the use of not valid especially for short high- nology was published by Ishii et al. (38). session RPE and duration as a valid intensity sprint runs (18,39,41,43,46). This group used mobile GPS units to instrument for tracking workload Jennings et al. (39), for example, used track referees during match play in both within anaerobic sports and within GPS devices sampling at 1 and 5 Hz rugby union and soccer. The study a resistance training exercise environ- and concluded that GPS systems may compared video tracking, manual ment (49,62,66). In parallel with this be limited for the assessment of short recording, and GPS tracking of the ref- growth and interest in sport-related high-speed straight-line running and ef- eree and found that there was little dif- workload monitoring, methods were forts involving change of direction. The ference among all 3 methods for total investigated that could provide objec- authors did note that an increased sam- distance covered in the match. Since the tive data on workload. An important ple rate improved validity and reliability original study of Ishii et al. (38), GPS factor that assisted in this development of GPS devices. Other authors sup- technology has become more sophisti- was when GPS selective availability ported this finding when assessing 1- cated in the metrics it can provide and was removed in 1999, making it avail- and 5-Hz GPS devices (18,41). has been integrated with other micro- able for use within a sporting environ- Johnston et al. (41) reported that GPS electromechanical sensors (MEMS) ment (45). Subsequently, RPE and error was found to increase along with such as accelerometers and gyroscopes. session duration workload quantifica- the velocity of movement and noted that This has been facilitated by team sport tion were supplemented with GPS- more caution should be exercised when governing bodies that now allow play- derived workload metrics and these analyzing movement demands .20 ers to use GPS tracking devices during measures became increasingly popular km$h21. Coutts and Duffield (18) re- match play (24,57). Recently, the world (3,34,35,39). ported an acceptable level of accuracy governing body for soccer (FIFA) With a view to establish a broader and reliability for total distance during amended their rules to allow the use definition of what workload means in high-intensity, intermittent exercise but of GPS technology in competitive a sporting environment, Quarrie et al. not for higher intensity activities. Subse- match play (26). This has created the (57) defined “load” within a team sport quently, the use of 10-Hz devices became opportunity for the quantification of perspective as “the total stressors and more common in team sport and Varley several physical workload metrics asso- demands applied to the players.” This et al. (69) reported that devices using ciated with elite competitive soccer included sport-related and non–sport- such higher sampling rates were 2–3 match play (Table 1). related stress inputs. Furthermore, the times more accurate than the 5-Hz de- The integration of MEMS technology authors noted that the relevance of load vices. The same authors then concluded now allows for the quantification of to athlete performance, well-being, and that newer GPS devices may provide an metrics related to body impact, adding injury risk should be considered from an acceptable tool for the measurement of to the concept of workload monitoring acute and cumulative perspective. Thus, constant velocity, acceleration, and (1,7,14,15,20,29,64,71). Metrics such as taken together, workload should be deceleration during straight-line running dynamic stress load and step balance considered from the totality of all stres- and have sufficient sensitivity for detect- are examples of impact-related metrics sors, as an acute and chronic or ing changes in movement velocity in

84 VOLUME 40 | NUMBER 3 | JUNE 2018 Table 1 Metrics listed and definitions used are derived and informed from a number of sources (19,64,73,74)

Metrics commonly determined using GPS technology Definition Total distance This is a measurement of the total distance traveled during the playing period by a player. This is commonly measured in kilometers. Meters per minute This is expressed as total meters in a given minute and as an average during a specified period. Average speed This is the total distance covered by a player and divided by the total playing duration in hours. This is frequently measured in km per hour. Maximal speed The maximal speed reached for a one second sample period. Speed zones or thresholds Classified by Cummins et al. (19) as: Walking (0–7 km$h21) Running at low speed (7–13 km$h21) Running at medium speed (13–18 km$h21) Running at high speed (18–21 km$h21) Sprinting (.21 km$h21) Acceleration Acceleration activity is measured as the change in GPS speed data using established statistical methods. To count as an acceleration, the increase in speed must take place for at least half a second with maximum acceleration in the period at least 0.5 m/s2. The acceleration finishes when the player stops accelerating. Frequently, the number of accelerations is reported through specific zones. Deceleration Deceleration is the decrease in speed that takes place for at least half a second for an activity to be counted as deceleration. Also, the maximum deceleration in the period must be at least 0.5 m/s2. The classification of decelerations by zone is based on the maximum deceleration in the period. Frequently, the number of decelerations is reported through specific zones. Exertion index This is calculated differently between brands. Exertion index according to Wisbey et al. (74) is based on the sum of a weighted instantaneous speed, a weighted accumulated speed over 10 seconds, and a weighted accumulated speed over 60 seconds. Exertion index per minute This is a measure of game intensity and is determined by dividing exertion index by playing time. Longest continuous time above a specified speed This is a measurement of the longest period the player stays above this speed, without dropping below this speed. Time is recorded even when the player enters a higher speed zone. This provides an indication of the longest continuous effort at varying speeds. Time at steady state This is measured as any time at a speed above 8 km$h21 where the player’s velocity does not alter by more than 1.5 km$h21 within a 1- second sample period. This gives an indication of time spent at continual running speeds. Efficiency This is a measure of the work requirements for game involvement. It is measured by dividing the exertion index by total number of possessions.

(continued)

Strength and Conditioning Journal | www.nsca-scj.com 85 The Current Use of GPS

Table 1 (continued)

Energy expenditure The energy expenditure metric gives the total energy associated with running only, including accelerating and decelerating activity measured in kcal. It is based on the level of activity and scaled by the weight of the player in kilograms, set in the player profile screen. This metric and that of HMLD are derived from the work of Osgnach et al. (55). High metabolic load distance (HMLD) HMLD is the distance (meters) traveled by a player when his/her metabolic power (energy consumption per kilogram per second) is above the value of 25.5 W/Kg. Metrics derived from integrated accelerometer Impacts Impacts are frequently identified as maximum accelerometer magnitude values above 2g in a 0.1-s period and reported as maximum values and cumulative over a specific period. Dynamic stress load (DSL) Dynamic stress load is the total of the weighted impacts, which is based on accelerometer values of magnitude above 2g. It weights the impacts using an approach similar to that used in the speed intensity or heart rate exertion calculations, with the key concept being that an impact of 4g is more than twice as hard on the body as an impact of 2g. Statsports (www.statsports.com) note that impacts are a mixture of collisions and step impacts while running. For team sports such as soccer, these peak impacts may increase as a player fatigues during match-play. Total load Total load gives the total of the forces on the player over the entire activity period based on accelerometer data alone. It uses the magnitude of the accelerometer values taken in 3 directions, sampled 100 times per second. The total is scaled by 1,000 to give manageable values. Step balance Using the accelerometer data, sampled at 100 Hz in each of the 3 directions, step balance is defined as the average peak impact on each step for the left and right feet.

GPS 5 global positioning system. team sport. As noted earlier, before we They compared the accelerometer- accelerometers incorporate gravity com- can confidently rely on these data, the derived metrics for impact and pensation formula in their algorithms, validity and reliability of such devices speed with a dual-beam timing system the usefulness of any accelerometer- need to be established. To date, it does as well as a video-capture system when derived algorithm is questionable (2). seem that practitioners can have confi- professional team sport players com- pleted 3 3 40 m maximal-sprint efforts Although GPS provides a potentially dence that 10-Hz GPS devices have, in useful tool for the strength and condi- on an indoor synthetic running track. general, acceptable levels of reliability tioning coach to measure several work- The accelerometer data were recorded and validity for many metrics measured. load metrics (Table 1), it is important to as the raw output and filtered using a vari- Early studies on accelerometer use in note that there is still concern regarding ety of filtering techniques. The authors team sports reported acceptable reliabil- the precision of certain GPS-derived ity levels both within and between devi- reported that the accelerometer could metrics (6,9,10). For example, the ces (8). However, a recent work from not accurately measure average acceler- dynamic stress load is calculated in Alexander et al. (2) cautioned that accel- ation values during high-speed running, the GPS software as a weighted score erometer use may not be accurate during and significantly overestimated average using a combination of accelerometer maximal acceleration and high-speed acceleration values during both 0–10 m and GPS data along with impacts running. The authors assessed a com- and 10–20 m segments, regardless of the (64). This external load metric is in- monly used 100-Hz triaxial accelerome- filtering technique used. The authors tended to reflect the total body load ter integrated within a wearable device. concluded that until GPS-integrated impact weighted to dominant and

86 VOLUME 40 | NUMBER 3 | JUNE 2018 nondominant limbs over the duration of of physical training, practice, and and blocking (10). GPS technology the physical activity. This metric, match play derived from GPS and now features triaxial accelerometers although appealing as a workload met- accelerometer technology. Note that integrated into the GPS unit ric, has not been validated in field stud- the table is not an exhaustive one, (8,35,48). This integration of technolo- ies. Yet, the potential of using this metric rather an example of metrics used in gies has allowed for the quantification hasbeendemonstratedinthatithas team sports (19,33,64,73,74). Also note of additional important impact metrics been associated with risk of injury and that some metrics have not been as- associated with contact sports espe- maybeusedinaninjurypredictive sessed regarding validity and many cially (1,3,8,9,13,54,56). Frequently, model (42). Nevertheless, although the metrics are defined differently depend- accelerometer values are reported as potential of using such metrics is evi- ing on the unit’s brand (48). G-force and less commonly as meters dent, more applied research is required per second2 and these metrics are now in establishing the reliability and validity TOTAL DISTANCE reported by many popular GPS/accel- of these metrics. Over the past 10 years, the demands of erometer devices (33,64). This addi- amateur and professional match play tional workload quantification WORKLOAD DEMANDS have been well established using GPS inevitably adds to the broader under- CONVENIENTLY DESCRIBED technology (11,49,58,60,67,68). Studies standing of workload. For example, it Clearly, the previous approach to report that players can cover a distance allows for greater insight into position- match analysis when using notational of between 8,500 and 13,000 m using specific match play preparation with or time-motion analysis was a labor- various locomotion intensities during threshold levels being determined that intensive one and required several match play (20,58). Research has also are appropriate for the player and his hours of analysis before presenting shown that the physical activity profile or her position (15). feedback (43). The time efficiency of can be different depending on playing using GPS is a very welcome technol- The usefulness of accelerometer- position (47), age of player (11), and related metrics has been evident in ogy addition to the team support staff. team tactical formation (67). This now allows for the immediate recent studies. For example, Cormack et al. (15) assessed the relationship compilation and presentation of rele- SPEED ZONES vant metrics, thereby facilitating more between accelerometer-derived impact The speed activity profile of players load per minute and neuromuscular immediate feedback to coaching staff may be categorized into zones typi- and players. In fact, it has been argued fatigue in elite team sport players. cally using 5 absolute speed zones as The authors reported an association that when considering the alternative described in Table 1 (19). Although practices for player tracking and sub- between fatigue and vertical axis these zones are commonly used, they impact, thus providing a potential use sequent feedback, some errors may be are not standardized even within an acceptable trade-off for both the and support for the application of a given sport (19,48). Cummins et al. accelerometer load–derived data in time-efficient and ease of use of GPS (19) reported that speed zone criteria devices (43,53). player workload and recovery moni- often varied widely within and toring. More recently and within soc- USES OF GLOBAL POSITIONING between sports. As a result, recent cer, Abade et al. (1) described training SYSTEM–DERIVED METRICS studies have proposed a more individ- session body impact data for elite age Quantifying player workload requires ualized approach to speed thresholds grade players during a 9-week period analysis of all the activities a player based on a systematic and detailed pro- of training. Although the authors undertakes during a training and com- filing of acceleration and maximum noted that their results showed high petitive period (57). To facilitate this, velocity testing of players (50,52,54). variability between training sessions GPS metrics are used during a variety This development in the individualized across the age grades, the use of of physical activity sessions or units. application of GPS technology and its accelerometer-derived data added to For example, GPS-derived data are integration with field testing facilitates the overall description of the players’ used to quantify competitive and non- a more individualized profiling that has workload during this phase of the competitive match play, training- the potential to better inform a player’s season. related activities such as small-sided specific workload. games, technical and tactical prepara- tion units, on-field team fitness rou- IMPACT METRICS DIFFERENT TEAM FORMATIONS tines, on-field specific individual Until recently, GPS devices did not Tactical performance plays an impor- fitness, or rehabilitation routines. allow for the quantification of impor- tant role in soccer and different team tant sport-related events such as the formations are used by teams both as GLOBAL POSITIONING SYSTEM– impact associated with the take-off a standard formation and as adjust- DERIVED METRICS when jumping and landing, as well as ments during the game (22,67). These Table 1 describes some of the com- the impact occurring when falling to formations can significantly affect the monly used measurements or metrics the ground or when tackling, jostling, game-related tasks players will need to

Strength and Conditioning Journal | www.nsca-scj.com 87 The Current Use of GPS

achieve and subsequently, the physical the position-specific demands of soccer data presented 35–100% less workload capacities they will need to be able to players. The authors noted the benefits values than the middle-week data. perform these tasks. Only recently the of such position-specific monitoring Understanding the weekly workload impact of different formations on over- and concluded that training activities variations according to the competi- all workload and specific metrics has should focus on the level of physical tion and the developmental ages of been reported (67). Using GPS, Tier- and physiological demands based on the players can, according to the au- ney et al. examined 5 common playing the individual player profile during thors, contribute to optimizing both formations used in elite-level under 18 match play. They elaborated that many short- and mid-term planning. In addi- and under 21 team match play over the training drills can elicit similar internal tion, Stevens et al. (65) describe the course of a season (67). The metrics loads (HR-based) compared with actual workload tapering approach that is described included total distance, match play providing an optimal phys- either intentionally or unintentionally high-speed running, high metabolic iological training stress. For greater used in the week leading into match load distance, high-speed accelera- effectiveness, the authors noted that play. The authors draw attention to tions, and decelerations. The authors each team should have its own physio- the fact that nonstarters may actually reported that all positions within logical and mechanical demands asso- not be challenged sufficiently in phys- a 3-5-2 formation elicited higher total ciated with soccer match play during ical work during match week. This has distances (10,528 6 565 m, P 5 0.05), official games, measured with the same implications for players training pro- greater high-speed running distance technology that these players are mon- grammes and match readiness prepa- (642 6 215 m, P 5 0.001), and greater itored with during training sessions. ration. Thus, monitoring of all players high metabolic load distance (2025 6 can be facilitated by GPS technology 304 m, P 5 0.001) than all other WEEKLY WORKLOAD DESCRIBED use on an on-going basis. formations. Also, the authors reported above average acceleration and Before the common use of GPS in deceleration profiles (34 6 7, P 5 monitoring workload, it is arguable PLAYER WORK:RECOVERY 0.036 and 57 6 10, P 5 0.006, respec- whether precise workload monitoring MONITORING—INTEGRATED WITH tively) for a 3-5-2 formation. The au- in the immediate days during match OTHER KEY METRICS thors concluded that all positional week was a standard feature of team Integrating GPS- and MEMS-derived physical characteristics are influenced sports (37). However, since the intro- metrics with other standardized phys- by the demands of playing different duction of affordable and reliable GPS ical and physiological testing as well as formations. technology in professional soccer and psychological data may provide a more other team sports, a more thorough complete profile of the stress status of Tactical and technical factors have and complete workload profile is now the player. For example, Watson et al. been reported to be related to overall a feature of in-season on-field training (70) reported that in female adolescent team success (22,23). Di Salvo et al. and practice (60). This was a positive soccer players, lower mood scores and (22) reported that although physical development in player workload man- higher acute training load as derived performance metrics impact on match agement. The potential of such moni- from GPS workload tracking are fac- outcome, overall, technical and tactical toring is reflected in the specific daily tors associated with an increased injury effectiveness of the team rather than workloads leading into a competitive risk. The authors note that monitoring high levels of physical performance match that are now monitored by soc- well-being and training load may facil- are more important in determining cer teams. itate intervention that may reduce the success in soccer. Therefore, it is evi- risk of in-season injury and illness. Such dent that systems which assist in the Coutinho et al. (17), for example, metrics could also be further comple- quantification of technical and tactical described the time-motion and physi- mented using other internal system approaches and events be integrated ological performance profiles of age- variables such as HR monitoring and with physical workload analysis for grade soccer players during a typical HR variability, the latter reflecting a more holistic representation of week of a competitive season. The data a metric of the status of the autonomic performance. were captured using a 15-Hz GPS device and divided into postmatch nervous system. In addition, metrics (session after the match), prematch reflecting other internal or biological POSITION-SPECIFIC DEMANDS (session before the match), and middle systems including neuromuscular, en- The use of GPS and related technology week (average of remaining sessions). docrinal, and immunological systems has allowed for a convenient and cumu- Differences in the various training/ coupled with psychological (i.e., mood lative approach to player positional practice days were reported for the dif- or well-being status) as well as lifestyle profiling in terms of workload (67,68). ferent age grades. The older age group profiling (i.e., sleep duration and For example, Torren˜o et al. (68) used an displayed relatively higher body im- quality) hold the potential to attain integrated 5-Hz GPS tracking device pacts with impacts above 10 G being a more complete monitoring profile and a 100-Hz accelerometer to quantify reported. Furthermore, the prematch (31,32,42,70).

88 VOLUME 40 | NUMBER 3 | JUNE 2018 INJURY RISK–REDUCTION POTENTIAL Within soccer, a theoretical predictive player response to such workloads Clearly, the full potential of GPS and model of injury derived from GPS met- (35). This may be due to a number of related technology is yet to be compre- rics has been proposed (42). The author factors and Halson (35) notes that re- hensively investigated and realized. examined the incidence of intrinsic sources in the form of time, money, or This is arguably true from an injury injury, defined as an injury not related the human resources needed to collect, risk–reduction perspective when GPS- to collision or contact, and several GPS- process, and analyze the data are all derived data are combined with histor- related metrics. The study demonstrates issues that may limit the use of player ical and prospective injury-related data that key GPS metrics aggregated over monitoring systems. An important as well as other psychoemotional re- a weekly workload can offer workload point made by Halson (35) is that there sponses such as mood (13,28,70). Nev- and temporal guidelines that assist in are no guarantees that monitoring ertheless, studies have investigated the better workload-recovery practices that training load will result in successful association between physical workload seek to reduce injury risk. There are performances. In addition, a lack of using data derived from GPS technol- some practical limitations to the study knowledge or experience with moni- ogy and from retrospective and pro- and one is the limitation of the applica- toring techniques or the inability to spective injury data surveillance tion of such predictive models, given interpret derived data can result in an (13,28,30,59). Coughlan et al. (16) sug- that contact is an inevitable part of the impediment to implement a practical gested that GPS data used in conjunc- game. Other limitations relate to the and sustainable monitoring system. In tion with video analysis can help clarify rigorous process required in establishing conjunction, the reality may be that the mechanism of many injuries. This an association between injury risk fac- many teams do not have a clear ratio- integration of technologies can no tors and injury risk–reduction practice nale identifying why monitoring is doubt be of assistance to the support (4). Nonetheless, the research is of occurring, what is to be monitored, team in understanding the factors that interest as it is among the first to report how often monitoring will occur, and may be associated with injury during the potential of such modeling techni- how the data are interpreted and pre- match play and training. Added to this ques in soccer. Other related wearable sented to the coaching staff and play- potential, Gabbett and Ullah (28) re- technologies offer potential for more in- ers. Halson (35) also notes that the ported that higher distances and high- depth and critical workload evaluation ability and opportunity to implement speed running are variables or metrics as well as injury rehabilitation workload change and provide feedback is critical associated with injury in elite sports. In monitoring (19). These may include to a successful monitoring system and addition, Colby et al. (13) examined the wearable electromyographic (EMG) if this does not occur, any attempts at association between overall physical technology. Together, GPS and EMG monitoring are not sustainable. workload using GPS and wearable technologies highlight the accelerometer-derived measurements potential benefit of such technological POTENTIAL TO TRACK SPORT OR and injury risk in elite team sport players integration in establishing workloads GAME CHANGES OVER TIME during a season. The authors reported that are on the one hand, tolerable to As previously discussed, studies have that from an injury risk perspective, the player, but also allow for thresholds described the changing nature of the their findings supported consideration to be determined that may assist in game of soccer at the professional level of several GPS/accelerometer running reducing the risk of injury and in assist- (5,22). To do so, the authors used a mul- load variables. They noted that cumu- ing in the return to play program after ticamera, stadium-based, computerized lative weekly loads could be closely injury (19,51). tracking system (Prozone Sports Lim- monitored with 3-weekly loads most ited, Leeds, United Kingdom). Such indicative of a greater injury risk across tracking systems are becoming popular ACTUAL USE OF GLOBAL especially at the top tier of professional both pre-season and in-season phases. POSITIONING SYSTEM AND More recently, Windt et al. (72) high- MICROELECTROMECHANICAL socccer and have been shown to be lighted the apparent training-load injury SENSORS TECHNOLOGY IN valid and reliable (21). The initial setup paradox within team sport players. The INJURY RISK–REDUCTION cost, the fixed nature of the system MANAGEMENT authors reported that additional pre- within a given stadium, the absence To what extent such integration of of such a system in all physical and season training sessions were associated injury data and GPS/MEMS-derived technical training areas, and the lack with a reduction in the odds of injury in workload metrics are recorded and of impact-related outputs are all factors the subsequent training week and fur- analyzed in team sport is not clear that may limit stadium-based system ther, with a lower percentage of games (35). Having the technology alone does use. As evidenced earlier, all physical missed due to injury during the in- not ensure that such data are recorded, workload regardless of where it occurs season. The authors concluded that analyzed, and used to inform better needs to be recorded and reported to maximizing participation in pre-season player management practice. Also, it quantify the total workload of a player. training may protect elite team players is clear that not all team sport coaches Therefore, the portability associated against in-season injury. will regularly monitor workload and with GPS tracking is a positive feature

Strength and Conditioning Journal | www.nsca-scj.com 89 The Current Use of GPS

of this system as it allows locomotor this can be a common problem when tracking (33,64). For example, local metrics from most all outdoor physical using GPS tracking (13,20,45). Colby positioning systems which use accel- exercise to be recorded and re- et al. (13) report that, on occasions, erometer, gyroscope, and magne- ported (45). GPS data were deemed unreliable tometer capability are currently Although no long-term comparative because of an intermittent signal where promoted to provide several locomo- GPS-based studies have been re- insufficient connecting satellites were tor and impact metrics in an indoor ported in soccer, GPS-based technol- detected. On such occasions, the au- environment (64). However, more ogy has been shown to be useful in thors ensured that workload data were research is required to establish the tracking the changing nature of other predicted by calculating individual device’s accuracy, especially during team sports, notably Australian Rules player averages for drills completed. high-speed running efforts (2). Football (73,74). Wisbey et al. exam- This does demonstrate the need to have To ensure better reception and conse- ined the demands of elite-level Aus- alternative methods to record work that quently more accurate data capture, tralianRulesFootballbasedonGPS is actually completed when satellite sig- differential GPS is sometimes though data collection with the cooperation nal reception is problematic. not always used (13,20). More recent of 16 clubs (73,74). Interestingly, the Many practitioners may use a combi- advances of GPS and wireless com- comparison of data collected on dif- nation of methods to report overall munication technology are now in ferent brands of GPS devices showed workload; yet, few studies report use (64). These advances are based similar work rates but differences in a combination of GPS data and other on augmented satellite signal recep- accelerations and surges. Thus, the methods of physical activity workload tion using ultra-wideband positional accuracy of the data is questioned, reporting such as session RPE and beacons placed strategically within given the use of different GPS brands. duration (13). For example, Coutinho a stadium or training/competition On a positive note, the changes in cer- et al. (17) provided no description or arena, which are reported to augment tain metrics do facilitate an analysis of report of any indoor-related work satellite signal reception (64). Peer- the collective changes occurring over completed by the subjects in their reviewed research reporting the accu- several seasons in the sport. Therefore, study. Other studies may also not racy and reliability of such ultra- the capability to track changes in match report all physical workload due to wideband augmented technology is play workload over several seasons this limitation (60,65). As a result, yet to be completed when using such when using GPS tracking systems is there is need to use more traditional device systems. well supported. workload-reporting formats such as In addition to the factors previously With respect to the use of different RPE and session duration (26) not on- noted, other factors are considered to tracking systems, some issues deserve ly for indoor aerobic and anaerobic impact on the precision and accuracy attention if data from both a fixed cam- exercise but also for resistance training of GPS tracking (43,44,48,51). For era system and a GPS system are to be (49,66). To assist in indoor workload example, Malone et al. (48) note that combined (5,21). The fixed system dif- quantification, the advent of com- in addition to device sampling rate, fers in the manner in which data are monly available and reliable linear positioning and fitting of devices captured and in which it categorizes position transducers now allows for and data filtering methods can affect metrics such as speed and accelera- a more precise quantification of the measures obtained from GPS and tion/deceleration zones (4,19,22). Such indoor resistance training–related MEMS devices. The authors encour- standardization issues are important to workload (36). Some players may age researchers to report device address before a valid representation of pursue physical recreation pastimes brand or model, sampling frequency, total workload can be quantified using and again, this is a further physical number of satellites locked onto, hor- a combination of systems. On this workload that might be included in izontal dilution of precision and soft- point, the author is unaware of any all physical workload-monitoring ware/firmware versions, in any published research which compares systems. published research. An understand- locomotor and other work-related Another limitation of GPS technol- ing of the impact of such factors on metrics derived from stadium tracking ogy is that, in its popular format, it a unique GPS session is also impor- systems and GPS-derived metrics and can only track outdoor physical tant for the practitioner or scientist a comparative study investigating this activity. Clearly, there is a need to who will eventually be responsible would be useful and beneficial. monitor all physical activities includ- for reporting and interpreting GPS- ing indoor activities if a more com- and MEMS-derived data. LIMITATIONS OF GLOBAL plete representation of overall POSITIONING SYSTEM workload is to be reported. GPS LIMITATIONS IN GLOBAL TECHNOLOGY manufacturing and distribution com- POSITIONING SYSTEM USE Surprisingly, few studies report issues in panies are working to overcome A GPS-related limitation with respect signal reception, whereas in practice, this limitation associated with GPS to use is that few studies have used

90 VOLUME 40 | NUMBER 3 | JUNE 2018 GPS technology to inform tactical and THE NEED FOR CAUTION could affect the estimates of technical performance criteria (67). Buchheit et al. (10) caution the coach accelerations, decelerations, and meta- Arguably, this may also reflect the lack and user when comparing data collected bolic power during short shuttle running of integration with other technologies with different brands or units or when and change of direction activities, with that describe tactical and technical el- data are analyzed with different software consequent underestimation of total ements of match play. In the absence of versions (Table 2). This, according to energy expenditure (EE). an integration of GPS and tactical/ Buchheit et al., is problematic especially With reference to EE, Brown et al. (9) technical event description, other tech- when dealing with historical data col- reported that GPS tracking using the nologies are often used to complement lected on a large number of players metabolic power model of EE does GPS-derived metrics so that a more (10). Researchers have reported differ- not accurately estimate EE in field- comprehensive evaluation of perfor- ences between devices even of the same sport movements or over an exercise mance workload can be attained. For model, suggesting that it is prudent that session consisting of mixed locomotor example, Nic (53) notes that most if not aplayerbemonitoredwiththesame all professional soccer teams use some device (18). By contrast, research from activities interspersed with recovery pe- form of video feedback from a tactical/ Castellano et al. (12) found small varia- riods. However, GPS tracking, is able to technical perspective. This concept has tions between devices, with a coefficient provide a reasonably accurate estima- been advanced with virtual reality bio- ofvariationof1.3and0.7%inrunsof15 tion of EE during continuous jogging feedback technologies being used in and 30 m, respectively. These research- and running (9). Nagahara et al. (52) professional team settings (www. ers concluded that it is not always nec- caution that although GPS tracking beyondsports.nl). It is possible that essary to monitor players with the same technology has merit, the more tradi- the integration of GPS with such tech- device. In a more recent study, Beato tional means of assessing sprint velocity nology has the potential to create et al. (6) reported that external load var- should be preferred compared with a more holistic and broader under- iables such as total distance and high- GPS-derived velocities. The authors re- standing of the physical, technical, speed running distance may be under- ported both overestimation and under- and tactical demands of the sport. estimated by GPS tracking even when estimation for both 5- and 20-Hz GPS The limitation for such use may not using a 10-Hz sampling rate. This under- tracking systems when measuring max- actually be a technical one, but rather estimation according to the authors in- imum velocity during a running sprint. a more human limitation, where creases in short-distance tracks as well as The above evidence is not intended to coaching and performance knowledge, during on-field–based scenarios includ- be biased toward negating the positive as well as analytical and communica- ing change of direction activities. The impact of GPS-derived data. It merely tion skills are insufficient in deriving authors state that these limitations attempts to caution the practitioner as meaningful interpretation from such should be considered even in sport sci- to the limitations of GPS tracking as it integrated data. ence research, especially because they is currently used (Table 2).

Table 2 Issues concerning GPS use in sport and recommendations to offset limitations

Issue concerning GPS Recommendation to resolve Systems using a sampling rate of 1 or 5 Hz may be less reliable and valid Select device with a sampling frequency of 10 Hz especially where high-speed running is concerned (20,40,43,61). (40,43,61). Comparing metrics derived from different brand devices may be Use only the same GPS brand (51). problematic (10,47). Changes in software may create practical problems in using historical Ensure compatibility, should software options data as the new software may not be completely compatible with the change (48). older software (10,48). Energy expenditure calculations derived from GPS systems may not be Use established systems for assessing athlete valid for intermittent team sport activities but may be appropriate for maximum velocity and energy expenditure (2,9,52). steady-state jogging or running (6). GPS may underestimate high-speed running especially in short-distance Use metrics that have established validity (10). Wear tracks and during change of direction (2,6). the same GPS tracking unit (61).

These are derived and informed from a number of sources.

GPS 5 global positioning system.

Strength and Conditioning Journal | www.nsca-scj.com 91 The Current Use of GPS

SUMMARY monitoring only outdoor exercise Liam Hennessy The introduction of GPS technology and its current incapability to moni- is Academic to monitor the physical demands of tor resistance training–related exer- Director at Se- match play and all outdoor-related cise is a limitation. Here, other tanta College and physical practice, training, and gen- MEMS technologies can play an visiting Professor eral activity in the early 2000s repre- important role, provided their validity at the University sented a significant step change in the and reliability are established and of South Wales. way in which team sport physical acceptable, and provided that manu- workload was recorded and assessed. facturers and service providers work In the mid-2000s, various studies closely with the end user in estab- examined the reliability and validity Ian Jeffreys is of the devices used to track external lishing the limitations of use as well as Professor of workload, and many reported reli- the beneficial uses of integrated Strength and ability and validity concerns. This systems. Conditioning at was related chiefly to the limited sam- Clearly, there is a need to identify reli- the University of pling frequencies for 1- and 5-Hz de- able, valid, and meaningful metrics South Wales. vices. The advances in sampling that are not only external workload frequency since then to include and GPS-derived but that also include a 10-Hz sampling frequency resulted internal and psychological metrics, as in greater confidence in the precision well as technical and tactical of GPS devices to track player phys- elements. There is also a need to inte- ical work-related metrics. Although grate all workload-related monitoring there are still concerns regarding the REFERENCES precision of both GPS tracking and systems so that unnecessary and irrel- evant data can be minimized. Future 1. Abade E, Gonc¸alves B, Leite N, and MEMS devices in tracking high- Sampaio J. Time motion and physiological speed activities as well as change of research should focus on identifying profile of football training sessions direction and other more novel met- the training and match play metrics performed by under 15, under 17 and rics, as derived from algorithms cal- that may allow for a valid injury pre- under 19 elite Portuguese players. Int J culated from accelerometer data, dictive model, thus facilitating Sports Physiol Perform 9: 463–470, 2014. GPS technology has made a clear a reduced risk of injury and individu- 2. Alexander J, Hopkinson T, Wundersitz D, statement as to its usefulness in player alized workload thresholds based on Serpell B, Mara J, and Ball N. Validity of a wearable accelerometer device to workload monitoring and to a lesser optimal work:recovery processes. measure average acceleration values extent, in overall performance and The use of GPS and related technol- during high-speed running. J Strength welfare management. As GPS tech- ogies, when the rationale of use is Cond Res 30: 3007–3013, 2016. nology and its software/firmware ad- agreed and articulated, can offer 3. Aughey R. Applications of GPS vances and integration with MEMS technologies to field sports. Int J Sports positive outcomes to both support technology has progressed, it is clear Physiol Perform 6: 295–310, 2011. staff and players. This is predicated that the practitioner or analyst 4. Bahr R. Why screening tests to predict recording, analyzing, and presenting on the end users of such technology injury do not work— and probably never these metrics needs to be both well appreciating the potential and will.: A critical review. B J Sports Med 50: informed of their limitations and limitations of such devices and soft- 776–780, 2016. potential. Furthermore, support staff ware. Finally, if the potential of GPS 5. Barnes C, Archer DT, Hogg B, Bush M, and should ensure a clear rationale for and related technologies is to be Bradley PS. The evolution of physical and technology use and application realized, there are many challenges technical performance parameters in the English Premier League. Int J Sports Med within the club or team setting. This, to all stakeholders. Nonetheless, 35: 1095–1100, 2014. then, is more likely to facilitate a pos- with the skills and competencies to 6. Beato M, Bartolini D, Ghia G, and Zamparo itive environment that allows for de- report, present, interpret, and apply P. Accuracy of a 10 Hz GPS unit in cisions about player recovery, the metrics derived from such devi- measuring shuttle velocity performed at practice, and training workload to ces, the potential for better player different speeds and distances (5–20 m). be implemented. welfare and performance manage- J Hum Kinetics 54: 15–22, 2016. 7. Boyd L, Ball K, and Aughey R. Quantifying Although the future for GPS technol- ment in such technologies can be realized. external load in Australian football matches ogy looks promising, there is a need and training using accelerometers. Int J to embrace a broader and more holis- Conflicts of Interest and Source of Funding: Sports Physiol Perform 8: 44–51, 2013. tic approach to workload monitoring. The authors report no conflicts of interest 8. Boyd L, Ball K, and Aughey R. The reliability The limitation of GPS technology in and no source of funding. of MinimaxX accelerometers for measuring

92 VOLUME 40 | NUMBER 3 | JUNE 2018 physical activity in Australian football. Int J 20. Dellaserra C, Gao Y, and Ransdell L. Use soccer training. Int J Sports Physiol Sports Physiol Perform 6: 311–321, 2011. of integrated technology in team sports. Perform 10: 860–864, 2015. 9. Brown D, Dwyer D, Robertson S, and J Strength Cond Res 28: 556–573, 2014. 34. Haddad M, Padulo J, and Chamari K. The Gastin P. Metabolic power method: 21. Di Salvo V, Collins A, McNeill B, and usefulness of session rating of perceived Underestimation of energy expenditure in Cardinale M. Validation of prozone: A new exertion for monitoring training load despite field-sport movements using a global video-based performance analysis system. several influences on perceived exertion. Int J positioning system tracking system. Int J Int J Perf Anal Sport 6: 108–119, 2006. Sports Physiol Perform 9: 882–883, 2014. Sports Physiol Perform 11: 1067–1073, 22. Di Salvo V, Gregson W, Atkinson G, 35. Halson S. Monitoring training load to 2016. Tordoff P, and Drust B. Analysis of high understand fatigue in athletes. Sports Med 10. Buchheit M, Haddad H, Simpson B, Palazzi intensity activity in Premier League soccer. 44: 139–147, 2014. D, Bourdon P, and Salvo V. Monitoring Int J Sports Med 30: 205–212, 2009. 36. Hansen K, Cronin J, and Newton M. The accelerations with GPS in football: Time to 23. Di Salvo V, Baron R, Tschan H, Calderon reliability of linear position transducer and slow down? Int J Sports Physiol Perform 9: Montero F, Bachl N, and Pigozzi F. force plate measurement of explosive 442–445, 2014. Performance characteristics according to force–time variables during a loaded jump 11. Buchheit M, Mendez-Villanueva A, playing position in elite soccer. Int J Med squat in elite athletes. J Strength Cond Res Simpson B, and Bourdon P. Repeated- Sports 28: 222–227, 2007. 25: 1447–1456, 2011. sprint sequences during youth soccer 37. Hennessy L. Evolution of a strength and matches. Int J Sports Med 31: 709–716, 24. FIFA. FIFA and IFAB to develop global conditioning programme. Presentation at 2010. standard for electronic performance. Available at: http://www.fifa.com/about- annual UK Strength and Conditioning 12. Castellano J, Casamichana D, Calleja- fifa/news/y52015/m510/news5fifa-and- Conference. Stirling, Scotland, 2011. Gonza´lez J, San Roma´n J, and Ostojic SM. ifab-to-develop-global-standard-for- 38. Ishii T, Masuda T, Kurogi H, Yabuno S, and Reliability and accuracy of 10 GPS devices electronic-performance-an-2709918.html: Yamamoto H. The distance covered of for short-distance exercise. J Sci Med Accessed April 23, 2017. soccer and rugby referees during the Sport 10: 233–234, 2011. match using a mobile “GPS”. In: ISBS- 25. FIFA. FIFA Big Count.Zu¨rich, Switzerland: 13. Colby M, Dawson B, Heasman J, Rogalski Conference Proceedings Archive 1: 323– FIFA Communications Division, Information B, and Gabbett T. Accelerometer and 325, 2002. Services, 2007. GPS-derived running loads and injury risk 39. Jennings D, Cormack S, Coutts A, Boyd L, in elite Australian footballers. J Strength 26. Foster C, Florhaug J, Franklin J, Gottschall and Aughey R. The validity and reliability of Cond Res 28: 2244–2252, 2014. L, Hrovatin L, and Parker S. A new GPS units for measuring distance in team approach to monitoring exercise training. 14. Cormack S, Mooney M, Morgan W, and sport specific running patterns. Int J Sports J Strength Cond Res 15: 109, 2001. McGuigan M. Influence of neuromuscular Physiol Perform 5: 328–341, 2010. fatigue on accelerometer load in elite 27. Foster C. Monitoring training in athletes with 40. Johnston RJ, Watsford ML, Kelly SJ, Pine Australian football players. Int J Sports reference to overtraining syndrome. Med Sci MJ, and Spurrs RW. Validity and interunit Physiol Perform 8: 373–378, 2013. Sports Exerc 30: 1164–1168, 1998. reliability of 10 Hz and 15 Hz GPS units for 15. Cormack S, Smith R, Mooney M, Young W, 28. Gabbett T and Ullah S. Relationship assessing athlete movement demands. and O’Brien B. Accelerometer load as between running loads and soft-tissue J Strength Cond Res 28: 1649–1655, a measure of activity profile in different injury in elite team sport athletes. J Strength 2014. standards of netball match play. Int J Cond Res 26: 953–960, 2012. 41. Johnston R, Watsford M, Pine M, Spurrs R, Sports Physiol Perform 9: 283–291, 29. Gabbett T, Jenkins D, and Abernethy B. Murphy A, and Pruyn E. The validity and 2014. Physical demands of professional rugby reliability of 5-Hz global positioning system 16. Coughlan GF, Green BS, Pook PT, Toolan league training and competition using units to measure team sport movement E, and O’Connor SP. Physical game micro-technology. J Sci Med Sport 15: 80– demands. J Strength Cond Res 26: 758– demands in elite rugby union: A global 86, 2012. 765, 2012. positioning system analysis and possible 42. Kampakis S. Predictive modelling of implications for rehabilitation. J Orthop 30. Gabbett T. The development and football injuries. Doctor of Philosophy Sports Phys Ther 41: 600–605, 2011. application of an injury prediction model for noncontact, soft-tissue injuries in elite dissertation. Univ Coll Lond, 2016. 17. Coutinho D, Gonc¸alves B, Figueira B, Abade collision sport athletes. J Strength Cond 43. Kelly VG, Scott MT, and Scott TJ. The E, Marcelino R, and Sampaio J. Typical Res 24: 2593–2603, 2010. validity and reliability of global positioning weekly workload of under 15, under 17, and systems in team sport: A brief review. under 19 elite Portuguese football players. 31. Gabbett T. The training—injury prevention J Aust Strength Cond 22: 186–190, 2014. J Sports Sci 33: 1229–1237, 2015. paradox: Should athletes be training smarter and harder? Br J Sports Med 50: 44. Langley RB. Dilution of precision. GPS 18. Coutts A and Duffield R. Validity and 273–280, 2016. World 10: 52–59, 1999. reliability of GPS devices for measuring movement demands of team sports. J Sci 32. Galambos S. Psychological predictors of 45. Larsson P. Global positioning system and Med Sport 13: 133–135, 2010. injury among elite athletes. Commentary. sport-specific testing. Sports Med 33: Br J Sports Med 39: 351–354, 2005. 19. Cummins C, Orr R, O’Connor H, and West 1093–1101, 2003. C. Global positioning systems (GPS) and 33. Gaudino P, Iaia F, Strudwick A, Hawkins R, 46. MacLeod H, Morris J, Nevill A, and microtechnology sensors in team sports: A Alberti G, and Atkinson G. Factors Sunderland C. The validity of a non- systematic review. Sports Med 43: 1025– influencing perception of effort (session differential global positioning system for 1042, 2013. rating of perceived exertion) during elite assessing player movement patterns in

Strength and Conditioning Journal | www.nsca-scj.com 93 The Current Use of GPS

field hockey. J Sports Sci 27: 121–128, 56. Paul GM, David B, and Clare LM. The session rating of perceived exertion method. 2009. physical and physiological demands of J Strength Cond Res 18: 796–802, 2004. 47. Mallo J, Mena E, Nevado F, and Paredes V. basketball training and competition. Int J 67. Tierney P, Young A, Clarke N, and Duncan Physical demands of top-class soccer Sports Physiol Perform 5: 75–86, 2010. M. Match play demands of 11 versus 11 friendly matches in relation to a playing 57. Quarrie K, Raftery M, Blackie J, Cook C, professional football using Global position using global positioning system Fuller C, and Gabbett T. Managing player Positioning System tracking: Variations technology. J Hum Kinetics 47: 179–188, load in professional rugby union: A review across common playing formations. Hum 2015. of current knowledge and practices. Br J Mov Sci 49: 1–8, 2016. Sports Med 51: 421–427, 2016. 48. Malone J, Lovell R, Varley M, and Coutts A. 68. Torren˜o N, Munguı´a-Izquierdo D, Coutts A, Unpacking the black box: Applications and 58. Rampinini E, Coutts A, Castagna C, Sassi de Villarreal E, Asian-Clemente J, and considerations for using GPS devices in R, and Impellizzeri F. Variation in top level Suarez-Arrones L. Relationship between sport. Int J Sports Physiol Perform 12: soccer match performance. Int J Sports external and internal loads of professional 1–30, 2016. Med 28: 1018–1024, 2007. soccer players during full matches in 49. McGuigan MR, Egan AD, and Foster C. 59. Rogalski B, Dawson B, Heasman J, and official games using global positioning Salivary cortisol responses and perceived Gabbett T. Training and game loads and systems and heart-rate technology. Int J exertion during high intensity and low injury risk in elite Australian footballers. Sports Physiol Perform 11: 940–946, intensity bouts of resistance exercise. J Sci Med Sport 16: 499–503, 2013. 2016. J Sports Sci Med 1: 8–15, 2004. 60. Scott BR, Lockie RG, Davies SJ, Clark AC, 69. Varley M, Fairweather I, and Aughey R. 50. McLellan C, Lovell D, and Gass G. Lynch DM, and de Jonge XAJ. The physical Validity and reliability of GPS for measuring Performance analysis of elite rugby league demands of professional soccer players during instantaneous velocity during acceleration, match play using global positioning in-season field-based training and match-play. deceleration, and constant motion. systems. J Strength Cond Res 25: 1703– J Aust Strength Cond 22: 7–15, 2014. J Sports Sci 30: 121–127, 2012. 1710, 2011. 61. Scott M, Scott T, and Kelly V. The validity 70. Watson A, Brickson S, Brooks A, and Dunn 51. Mendez-Villanueva A, Buchheit M, and reliability of global positioning systems W. Subjective well-being and training load Simpson B, Peltola E, and Bourdon P. in team sport. J Strength Cond Res 30: predict in-season injury and illness risk in Does on-field sprinting performance in 1470–1490, 2016. female youth soccer players. Br J Sports young soccer players depend on how fast 62. Scott T, Black C, Quinn J, and Coutts A. Med 51: 194–199, 2017. they can run or how fast they do run? Validity and reliability of the session-RPE 71. Wellman A, Coad S, Goulet G, and J Strength Cond Res 25: 2634–2638, method for quantifying training in Australian McLellan C. Quantification of 2011. football. J Strength Cond Res 27: 270– accelerometer derived impacts associated 52. Nagahara R, Botter A, Rejc E, Koido M, 276, 2013. with competitive games in NCAA Division I Shimizu T, and Samozino P. Concurrent 63. Serrano J, Shahidian S, Sampaio J, and college football players. J Strength Cond validity of GPS for deriving mechanical Leite N. The importance of sports Res 31: 330–338, 2017. properties of sprint acceleration. Int J Sports performance factors and training contents 72. Windt J, Gabbett T, Ferris D, and Khan K. Physiol Perform 12: 129–132, 2017. from the perspective of futsal coaches. Training load–injury paradox: Is greater 53. Nic J. The role of notational analysis in J Hum Kinetics 38: 151–160, 2013. preseason participation associated with soccer coaching. Int J Sports Sci Coach 1: 64. Statsports.com: STATSports technologies lower in-season injury risk in elite rugby 185–198, 2006. LTD. Version 1.2. November 19, league players? Br J Sports Med 51: 645– 650, 2016. 54. Nu´ n˜ez-Sa´nchez FJ, Toscano-Bendala FJ, 2012Available at: http://statsports.com. Campos-Va´zquez MA, and Suarez-Arrones Accessed April 23, 2017. 73. Wisbey B, Rattray B, and Pyne DB. LJ. Individualized speed threshold to 65. Stevens T, De Ruiter C, Van Maurik D, Van Quantifying changes in AFL player game analyze the game running demands in Lierop C, Savelsbergh G, and Been P. demands using GPS tracking: 2008 AFL soccer players using GPS technology. Measured and estimated energy cost of season. Special Report: Florey (ACT): RETOS 32: 130–133, 2017. constant and shuttle running in soccer FitSense Australia, 2008. 55. Osgnach C, Poser S, Bernardini R, Rinaldo players. Med Sci Sports Exerc 47: 1219– 74. Wisbey B, Montgomery P, Pyne D, and R, and Di Prampero P. Energy cost and 1224, 2015. Rattray B. Quantifying movement demands metabolic power in elite soccer. Med Sci 66. Sweet T, Foster C, McGuigan M, and Brice G. of AFL football using GPS tracking. J Sci Sports Exerc 42: 170–178, 2010. Quantitation of resistance training using the Med Sport 13: 531–536, 2010.

94 VOLUME 40 | NUMBER 3 | JUNE 2018