GPS and Injury Prevention in Professional Soccer : The Journal of Strength & Conditioning Research

Secondary Logo

Journal Logo

Original Research

GPS and Injury Prevention in Professional Soccer

Ehrmann, Fabian E.1; Duncan, Craig S.2; Sindhusake, Doungkamol3; Franzsen, William N.4; Greene, David A.1

Author Information
Journal of Strength and Conditioning Research: February 2016 - Volume 30 - Issue 2 - p 360-367
doi: 10.1519/JSC.0000000000001093
  • Free



The use of Global Positioning Software (GPS) technology to measure players' training loads has become prevalent in professional soccer (36). The major focus of GPS use has been to identify the activity profiles of players in training or trial matches (36). However, in recent times, there has been increasing interest in using GPS to maintain the balance between training stress and recovery, thereby maximizing performance potential while minimizing the risk of overtraining (25,27). To date, no study has examined the link between GPS variables and noncontact soft tissue injuries in professional soccer players.

Players in European competitions miss on average 37 days in a 300-day season through injury (12), which presents a significant financial burden on professional soccer clubs and can severely damage their chances of success (1,21). The epidemiology of injuries in professional soccer has been well documented, with studies reporting that elite soccer players are at high risk of injury during training and even more so during match play (12,22). Most injuries sustained are soft tissue injuries, particularly strains to the thigh region (11), with some figures showing that up to 59% of injuries sustained were noncontact (22).

Furthermore, recent research has suggested a general increase in injury risk (2), or increased injury risks during match play (9), when players have less than 4 days to recover between matches, indicating that residual fatigue can impact injury rates significantly. To prevent fatigue-related injuries, it is essential to monitor the training load of players (23,25). A number of methods have been used to assess the effect of training on fatigue, including the monitoring of heart rate (23), heart rate variability (30) and hormonal levels (28). Moreover, session-RPE (13) has been used extensively and is identified as a reliable and valid measure of internal training load (24,33). Increases in session-RPE–based training loads have demonstrated a relationship with increased injury risk in elite youth soccer (5) and other sports such as Australian Football (31) and Rugby League (15,16).

In recent years, soccer teams have made use of GPS tracking devices to provide an objective measure of external training load (36). GPS units have been shown to be sufficiently reliable and accurate to quantify training workloads in team sports (8,34) and have been successfully used in Australian Rules Football and Rugby League for several years. Most studies using GPS in soccer have focussed on movement patterns during match and training activities. However, the relationship between GPS variables and other training load measurements has been explored. Total distance, distances measured at various intensities, and Body Load have been shown to correlate significantly with subjective session-RPE–based training loads (6,20,32).

However, the potential link between GPS training load measurements and injury risk in soccer is still unclear. Gabbett and Ullah (19) established a correlation between injury risk and high-intensity running efforts during training sessions in rugby league, whereas Nielsen et al. (29) noted that novice runners were more susceptible to injuries when weekly total distance increased by more than 30%, although these results were statistically nonsignificant.

At present, there are no studies to our knowledge that have investigated the relationship between GPS variables and soft tissue injury in professional soccer over a full season. Thus, the aim of this study was to examine the link between GPS variables and noncontact soft tissue injuries in professional soccer players. We hypothesize that GPS will be effective in identifying potential injury risk and that specific variables will be identified to enhance future practice in the applied setting.


Experimental Approach to the Problem

In this observational study, 19 elite soccer players' external training load data were monitored using 5 Hz GPS units (SPI-Pro GPSports, Canberra, Australia). Data were collected for 1 entire competitive season, spanning 37 weeks, from July 18, 2011 to March 30, 2012 (11 weeks preseason, 26 weeks in-season). Injuries were recorded and data retrospectively analyzed to identify potential risk factors.


Nineteen full time professional male soccer players (25.7 ± 5.1 years; age range 18–33 years; 181.0 ± 4.8 cm; 77.3 ± 4.2 kg) competing in the Australian Hyundai A-League participated in this study. Written informed consent was obtained from all subjects and the Australian Catholic University's ethics committee approved the study.

Data Collection

Players wore their allocated GPS units for every training session of the year, including recovery sessions and extra training, which did not include the whole squad. Exceptions were players who were injured or away on national team duty. Three pregame training sessions conducted away from the team's training ground were not recorded because of logistical difficulties. Players also wore GPS units during every preseason game and warm-up (n = 8), albeit, due to Fédération Internationale de Football Association (FIFA) regulations not during competitive matches. GPS data were downloaded immediately after every session or game and analyzed using GPSports Team AMS software (GPSports). Sessions were split individually and session splits are shown in Table 1.

Table 1:
All session splits, as well as split start and finish.

The GPS variables collected are shown in Table 2. The selection of these variables allows the comparative study of players' workloads with other studies that describe running patterns in soccer using computerized tracking systems (3,4). New body load was included to give a measure of accelerations and decelerations.

Table 2:
GPS variables measured in this study, their units, and zones.*

Injuries were documented according to FIFA outlines (14), but only noncontact soft tissue injuries that prohibited the injured player from participating in at least 1 match were included to target injuries, which have a higher possibility of being prevented. If an injury could not be pinpointed to a game or training session (e.g., if a player complained about pain occurring in prior weeks), it was excluded from the analysis. Players were assessed for fatigue and signs of overtraining before every training session and match to minimize injury risks. These assessments included the use of questionnaires (23,25), hormonal markers (28), and Heart Rate Variability (30).

Competitive Game Data

During the season, the team competed in 28 matches in which they were not allowed to wear GPS units. Australian soccer clubs and stadiums do not have computerized tracking systems, such as AMISCO Pro or ProZone, which monitor players' in-game workloads. Therefore, game data in this study were not measured but derived from game data collected preseason. Players who completed at least 1 full half in a preseason game were categorized according to position and positional averages were then calculated for in-game workloads. These averages were extrapolated according to playing time to derive players' competitive in-game workloads.

Although not specific to match location, quality of opposition, phase of the season or game status, 8 matches against high class opposition and played under varying conditions were included in these calculations. Results closely matched previously published data for soccer players in the A-League (35,36).

Data for youth/reserve team matches in which players included in this study participated were derived with the same method.

Statistical Analyses

All GPS variables were retrospectively analyzed by averaging individual players' data across 1 and 4 weeks leading up to an injury (injury block), averaging values across 1 and 4 weeks before the injury block (preinjury block) as well as averaging values from the beginning of the season to the point of injury (season average). This approach served to simplify the collected data and reflected the practical nature of this investigation, as moving averages are commonly used in elite sports to track changes in athletes' training patterns. A player's individual blocks were compared with each other, to determine workload differences before injuries occurred (Figure 1).

Figure 1:
Comparisons conducted between individuals' workload blocks.

A significant difference in a GPS measure between injury and preinjury blocks as well as injury blocks and season averages was assessed using analysis of variance with repeated measures design.

To reflect the practical nature of the collected data, Cohen's effect sizes (where η2 = sum of square [between the preinjury and injury blocks or injury blocks and season averages]/total sum of square) (26) were calculated. Effect sizes were classified according to Cohen's rule of thumb (0.02 = small, 0.13 = medium, 0.26 = large (7)). Per cent changes of magnitude were also calculated.

SPSS statistics version 19 for Windows (IBM) was used to analyze the data and a level of significance was set at 0.05 for all hypothesis testing.


A total of 16 injuries to 11 players fulfilled the criteria in this study and were subsequently analyzed. Three of these injuries were sustained in preseason, whereas 11 injuries occurred during competition matches. Injuries were spread evenly across the season, with a slight peak midway through the season (6 of 16 injuries occurred in December and January). Hamstring strains were the most common injury included in this study (n = 4), followed by calf strains (n = 3), ankle sprains (n = 3), and groin strains (n = 2).

Injured players completed significantly higher meters per minute in the injury block compared with the Season Average for both 1- and 4-week blocks (p = 0.008 for both comparisons) as shown in Figures 2 and 3.

Figure 2:
Significant difference (p = 0.008) between meters per minute (m·min−1) before injury and the season average (data are mean ± SE for 1-week blocks).
Figure 3:
Significant difference (p = 0.008) between meters per minute (m·min−1) before injury and the season average (data are mean ± SE for 4-week blocks).

Injured players' mean new body load was significantly reduced in injury blocks compared with the Season Average for both 1- and 4-week blocks (p = 0.006 and p = 0.01, respectively) as shown in Figures 4 and 5.

Figure 4:
Significant difference (p = 0.006) between new body load (arbitrary units) before injury and the season average (data are mean ± SE for 1-week blocks).
Figure 5:
Significant difference (p = 0.01) between new body load (arbitrary units) before injury and the season average (data are mean ± SE for 4-week blocks).

No other GPS variables compared between injury blocks and the Season Average yielded significant results. Similarly, comparisons between preinjury blocks and injury blocks yielded no significant results. Mean ± SE for all block comparisons conducted in this study, along with their p-values and effect sizes are shown in Tables 3 and 4.

Table 3:
Mean (SE) of the measurements for 1- and 4-week blocks leading up to and including the day of injury compared with the seasonal average for both 1- and 4-week blocks, with p values and effect sizes.*†
Table 4:
Mean (SE) of the measurements for 1- and 4-week blocks leading up to and including the day of injury compared with the 1- and 4-week blocks before that, with p values and effect sizes.*

Large effect sizes (effect size >0.26) were observed for total distance, new body load and meters per minute in both 1- and 4-week blocks, as well as for very–high-intensity running distance in 4-week blocks.

Per cent changes in magnitude for all block comparisons are shown in Table 5.

Table 5:
Per cent change in magnitude (SE) in all variables for both 1- and 4-week blocks.*†


This study identified 2 GPS variables as potential noncontact soft tissue injury predictors: average meters per minute and average new body load. Average meters per minute increased significantly by 9.6 and 7.4% from the season average blocks to injury blocks for 1 and 4 weeks, respectively, indicating a general increase in intensity during sessions leading up to injuries. These findings are similar to those made by Gabbett (15), who established a link between increased training intensity and injury occurrences in rugby league. Similarly, previous studies have established a correlation between high training loads and injury occurrences in team sports (16,31) using Foster's Session RPE method—a product of training duration (volume) and rate of perceived exertion (intensity). Furthermore, research has found that it is an increase in training load and not necessarily a high load itself that increases players' injury risks (31).

To date, no link has been established between session RPE training load and meters per minute during sessions in team sports. However, it is not surprising that average meters per minute increased before the occurrence of injury as it is the only variable that demonstrated a real measure of session intensity with respect to duration. Increases in meters per minute across a period of sessions reveal an increase in intensity, even if the overall training load remains relatively constant, as indicated by nonsignificant changes of total distance or high-intensity running distances. This may be the case during periods of the season when matches are played more frequently, allowing for less recovery time between matches and potentially increasing players' injury risks (2). Absolute training load may remain constant or only change slightly, as coaches try to allow for recovery by shortening sessions, not decreasing intensity. The result is an increase in average intensity across sessions, potentially leaving players unable to adapt or recover adequately.

Similarly, increased meters per minute during sessions could also be explained by coaches shortening breaks in between training drills, thereby increasing training density and potentially saving valuable training time. This could explain meters per minute increasing while high-intensity running distances remain constant or even decrease slightly. Such decreases in rest time in between drills are risk factors themselves and will lead to increases in meters per minute across the entire squad.

Although coaches shortening rest periods during training sessions is conceivable as a contributing factor toward injury occurrences during the collection period of this study, it was not an aspect considered at the outset of data collection. However, no conscious efforts by coaches to increase training density at certain stages of the season were observed.

Results also demonstrated that players averaged significantly less new body load in blocks leading up to an injury compared with their seasonal averages, 15.4 and 9.0% for 1- and 4-week blocks, respectively. This initially seems in contrast to previous research suggesting that increased training loads lead to injuries. However, recent studies have highlighted that some modes of training in team sports can leave players underprepared for the demands of competition play in Rugby League (17) and women's soccer (18). Eleven of 16 injuries investigated in this study occurred in match play. A period of relative undertraining during the week(s) leading up to a match, as signified by a lower average new body load, could leave players unable to perform the high number of accelerations and decelerations required during competition play. Variations in match intensity during a season, depending on opposition or score, e.g., could exaggerate such an “underpreparedness effect” (10).

A particular focus of this research into differences between training and match play intensities has been repeated bouts of very–high-intensity work or sprints. New body load is the only variable included in this study, which reflects these bouts, as new body load incorporates every force measured by the triaxial accelerometer sampling at 100 Hz within the GPS unit. The actual GPS variables included in this study may not be sensitive enough to measure any underpreparedness, as they do not include many energy demanding actions performed in soccer such as jumping, slide tackling, and accelerating and decelerating. Investigating accelerations and decelerations measured by displacement of the GPS unit itself, as opposed to by the accelerometer within the unit, could confirm these findings. However, accelerations and decelerations were not included in this study due to concerns over the accuracy and reliability of these variables measured by the 5 Hz GPS units used (34).

Another possible explanation for players averaging a lower new body load in the weeks leading up to an injury is that they may feel pain, soreness, or fatigue already and therefore “take it easy” during training sessions before breaking down due to the high-intensity demands during a game.

The concept of players playing when in pain or already injured, without necessarily informing physiotherapists, doctors, or coaches at their clubs, is not new and for a variety of reasons commonplace in soccer (25). This can lead to players purposely pacing themselves in training sessions, avoiding strenuous changes of direction and high-intensity accelerations, while running more at a steady pace to keep the overall distance covered high. In an intense competition match, the opportunities to adequately pace oneself to cover for an already existing problem are severely limited, and players can break down as a result of this.

The theory of players purposely pacing themselves best explains the paradox of having one variable—meters per minute—considerably higher leading up to injuries, whereas another—new body load—is considerably lower. However, every effort was made to minimize injury risks for the players participating in this study. Players were monitored for fatigue before every training session and game through the use of several tests. If doubts still persisted, players were then assessed and cleared for full training by physiotherapists. This made it unlikely that players trained while injured or in an injury susceptible state. It also explains the relatively low number of soft tissue injury instances in this study.

An increase in training density, i.e., a decrease in rest times while drill times remain constant throughout a training session, could also explain these training load patterns leading up to injuries. Decreased rest times in between training drills might therefore be considered a contributing factor toward soft tissue injury occurrences.

Natural variations in training load occurring over the course of the season create an obstacle when the findings of this study are applied in a practical setting. The question of how many times players showed significantly higher meters per minute or significantly lower new body load in blocks compared with their respective season averages without getting injured warrants further investigation. In the high-pressure environment of professional soccer, coaches are often inclined to risk players when results are needed. It is of importance to know how great the injury risk is to a player when he plays or trains when his GPS values are doubtful and to be able to assess whether he can be risked or not. Further risk assessment of similar data is needed to help maintain a balance between a player's health and safety and a club's urge to succeed. However, the lack of existing definite injury prediction models should not deter clubs from using motion tracking techniques to monitor players' workloads as, at the elite level, a set of GPS units plus a full time analyst, will only cost a fraction of an injured player's income per week.

The need for an individual approach to injury prevention using GPS technology is apparent and it is important to distinguish between nonmodifiable risk factors, such as age or injury history, and modifiable risk factors. The patterns in GPS data leading up to injuries observed in this study can justifiably be considered in the latter category.

This study was limited by FIFA and FFA regulations prohibiting the use of GPS units in competition play and 28 of 250 sessions were predicted and not measured. Because of the lack of infrastructure in Australian stadiums, soccer teams cannot use automated camera tracking systems and it is important to conduct similar studies in countries that do use such systems. GPS systems have also developed markedly since the commencement of this study, with units now offering higher sampling rates and analysis software providing more and maybe more relevant variables, which warrant investigation. A more sophisticated approach to the statistical analysis in this study could also help refine the results presented.

In conclusion, this study identified 2 GPS variables that were related to noncontact soft tissue injuries in professional soccer. A significant increase in meters per minute across 1- or 4-week blocks, as well as a significant decrease in new body load across 1- and 4-week blocks predisposed soccer players to injury. Many questions still remain and further research is required to investigate if additional GPS variables have a relationship with noncontact soft tissue injuries in professional soccer players.

Practical Applications

This study has found 2 variables that deserve particular attention when monitoring soccer players with GPS units for the purpose of injury prevention. These are average meters per minute, when significantly increased compared with the Season Average (9.6 and 7.4% for 1- and 4-week blocks, respectively), and average new body load, when significantly decreased compared with the season average (15.4 and 9.0% for 1- and 4-week blocks, respectively). Both variables have been found to be predictors of the noncontact soft tissue injuries investigated in this study and should therefore be considered modifiable risk factors to soccer players' soft tissue injuries. Meters per minute and new body load warrant extra consideration when training sessions are retrospectively analyzed and future training sessions are planned.


The authors thank Wai-Leng Wong for her contribution to this work.


1. Arnason A, Sigurdsson SB, Gudmundsson A, Holme I, Engebretsen L, Bahr R. Physical fitness, injuries, and team performance in soccer. Med Sci Sports Exerc 36: 278–285, 2004.
2. Bengtsson H, Ekstrand J, Walden M, Hagglund M. Muscle injury rates in professional football increase with fixture congestion: An 11-year follow-up of the UEFA champions league injury study. Br J Sports Med 48: 566–567, 2014.
3. Bradley PS, Di Mascio M, Peart D, Olsen P, Sheldon B. High-intensity activity profiles of elite soccer players at different performance levels. J Strength Cond Res 24: 2343–2351, 2010.
4. Bradley PS, Sheldon W, Wooster B, Olsen P, Boanas P, Krustrup P. High-intensity running in English FA Premier League soccer matches. J Sports Sci 27: 159–168, 2009.
5. Brink MS, Visscher C, Arends S, Zwerver J, Post WJ, Lemmink KA. Monitoring stress and recovery: New insights for the prevention of injuries and illnesses in elite youth soccer players. Br J Sports Med 44: 809–815, 2010.
6. Casamichana D, Castellano J, Calleja-Gonzalez J, San Roman J, Castagna C. Relationship between indicators of training load in soccer players. J Strength Cond Res 27: 369–374, 2013.
7. Cohen J. Statistical Power Analysis for the Behavioral Sciences. Hillsdale, NJ: L. Erlbaum Associates, 1988.
8. Coutts AJ, Duffield R. Validity and reliability of GPS devices for measuring movement demands of team sports. J Sci Med Sport 13: 133–135, 2010.
9. Dellal A, Lago-Penas C, Rey E, Chamari K, Orhant E. The effects of a congested fixture period on physical performance, technical activity and injury rate during matches in a professional soccer team. Br J Sports Med 49: 390–394, 2013.
10. Di Salvo V, Gregson W, Atkinson G, Tordoff P, Drust B. Analysis of high intensity activity in premier league soccer. Int J Sports Med 30: 205–212, 2009.
11. Ekstrand J, Hagglund M, Walden M. Epidemiology of muscle injuries in professional football (Soccer). Am J Sports Med 39: 1226–1232, 2011.
12. Ekstrand J, Hagglund M, Walden M. Injury incidence and injury patterns in professional football: The UEFA injury study. Br J Sports Med 45: 553–558, 2011.
13. Foster C. Monitoring training in athletes with reference to overtraining syndrome. Med Sci Sports Exerc 30: 1164–1168, 1998.
14. Fuller CW, Ekstrand J, Junge A, Andersen TE, Bahr R, Dvorak J, Hagglund M, McCrory P, Meeuwisse WH. Consensus statement on injury definitions and data collection procedures in studies of football (soccer) injuries. Br J Sports Med 40: 193–201, 2006.
15. Gabbett TJ. Influence of training and match intensity on injuries in rugby league. J Sports Sci 22: 409–417, 2004.
16. Gabbett TJ, Jenkins DG. Relationship between training load and injury in professional rugby league players. J Sci Med Sport 14: 204–209, 2011.
17. Gabbett TJ, Jenkins DG, Abernethy B. Physical demands of professional rugby league training and competition using microtechnology. J Sci Med Sport 15: 80–86, 2012.
18. Gabbett TJ, Mulvey MJ. Time-motion analysis of small-sided training games and competition in elite women soccer players. J Strength Cond Res 22: 543–552, 2008.
19. Gabbett TJ, Ullah S. Relationship between running loads and soft-tissue injury in elite team sport athletes. J Strength Cond Res 26: 953–960, 2012.
20. Gomez-Piriz PT, Jimenez-Reyes P, Ruiz-Ruiz C. Relation between total body load and session rating of perceived exertion in professional soccer players. J Strength Cond Res 25: 2100–2103, 2011.
21. Hagglund M, Walden M, Magnusson H, Kristenson K, Bengtsson H, Ekstrand J. Injuries affect team performance negatively in professional football: An 11-year follow-up of the UEFA champions league injury study. Br J Sports Med 47: 738–742, 2013.
22. Hawkins RD, Fuller CW. A prospective epidemiological study of injuries in four English professional football clubs. Br J Sports Med 33: 196–203, 1999.
23. Hooper SL, Mackinnon LT. Monitoring overtraining in athletes. Recommendations. Sports Med 20: 321–327, 1995.
24. Impellizzeri FM, Rampinini E, Coutts AJ, Sassi A, Marcora SM. Use of RPE-based training load in soccer. Med Sci Sports Exerc 36: 1042–1047, 2004.
25. Kellmann M. Preventing overtraining in athletes in high-intensity sports and stress/recovery monitoring. Scand J Med Sci Sports 20(Suppl.2): 95–102, 2010.
26. Kraemer HC, Thiemann S. How Many Subjects?: Statistical Power Analysis in Research. Newbury Park, CA: Sage Publications, 1987.
27. Meeusen R, Duclos M, Foster C, Fry A, Gleeson M, Nieman D, Raglin J, Rietjens G, Steinacker J, Urhausen A. Prevention, diagnosis, and treatment of the overtraining syndrome: Joint consensus statement of the European College of Sport Science and the American College of sports Medicine. Med Sci Sports Exerc 45: 186–205, 2013.
28. Mortatti AL, Moreira A, Aoki MS, Crewther BT, Castagna C, de Arruda AF, Filho JM. Effect of competition on salivary cortisol, immunoglobulin A, and upper respiratory tract infections in elite young soccer players. J Strength Cond Res 26: 1396–1401, 2012.
29. Nielsen RO, Cederholm P, Buist I, Sorensen H, Lind M, Rasmussen S. Can GPS Be used to detect deleterious progression in training volume among runners? J Strength Cond Res 27: 1471–1478, 2013.
30. Plews DJ, Laursen PB, Stanley J, Kilding AE, Buchheit M. Training adaptation and heart rate variability in elite endurance athletes: Opening the door to effective monitoring. Sports Med 43: 773–781, 2013.
31. Rogalski B, Dawson B, Heasman J, Gabbett TJ. Training and game loads and injury risk in elite Australian footballers. J Sci Med Sport 16: 499–503, 2013.
32. Scott BR, Lockie RG, Knight TJ, Clark AC, Janse de Jonge XA. A comparison of methods to quantify the in-season training load of professional soccer players. Int J Sport Physiol Perform 8: 195–202, 2013.
33. Scott TJ, Black CR, Quinn J, Coutts AJ. Validity and reliability of the session-RPE method for quantifying training in Australian football: A comparison of the CR10 and CR100 scales. J Strength Cond Res 27: 270–276, 2013.
34. Varley MC, Fairweather IH, Aughey RJ. Validity and reliability of GPS for measuring instantaneous velocity during acceleration, deceleration, and constant motion. J Sports Sci 30: 121–127, 2012.
35. Varley MC, Gabbett T, Aughey RJ. Activity profiles of professional soccer, rugby league and Australian football match play. J Sports Sci 32: 1858–1866, 2013.
36. Wehbe GM, Hartwig TB, Duncan CS. Movement analysis of Australian national league soccer players using global positioning system technology. J Strength Cond Res 28: 834–842, 2014.

Athlete monitoring; training load; fatigue

Copyright © 2016 by the National Strength & Conditioning Association.