Relationship Between Indicators of Training Load in Soccer Players : The Journal of Strength & Conditioning Research

Secondary Logo

Journal Logo

Original Research

Relationship Between Indicators of Training Load in Soccer Players

Casamichana, David1; Castellano, Julen1; Calleja-Gonzalez, Julio1; San Román, Jaime1; Castagna, Carlo2

Author Information
Journal of Strength and Conditioning Research 27(2):p 369-374, February 2013. | DOI: 10.1519/JSC.0b013e3182548af1
  • Free



To develop physical fitness and team skills, an extensive use of group training (i.e., specific training) drills is considered in soccer (13). Specific training in soccer assumes the form of small-sided games using different number of players, pitch dimensions, and game rules to promote the requested adaptations (21). Team-skill training load (TL) quantification is of importance when the objective is to evaluate magnitude compliance between planned and performed training drills. This enables TL to be modulated according to seasonal training aims. This assumes value as efficient training prescription is work load dependent (29).

In soccer, the individual training response (internal load) to a given imposed training program (external load) may result in being different among players, and consequently, training individualization may result problematic (12). Therefore, the development of valid methods for TL assessment is paramount in soccer because extreme training responses may result in training maladaptations and injuries (17,18).

With the aim to profile the internal load, a number of methods have been proposed using effort perception or heart-rate (HR) responses to training (3). Recently, the session–rating of perceived exertion method (sRPE) has been the object of studies that examined its validity assuming as construct HR methods (24), which has been correlated with other internal and external TL (8).

Despite the practical interest provided by these studies, a conclusive response as per sRPE method criterion validity is yet to be reported in soccer. Indeed, HR methods were based on theoretical construct and consequently cannot be considered as TL gold-standard criteria.

A viable procedure to assess the criterion validity of the supposed indicators of internal load may result in the comparison with variables representing the imposed TL. This assumes the training work load as reflection of the coach-imposed external load hypothesizing a cause-effect relationship (25).

In soccer, this procedure gains logical validity because specific training (i.e., small-sided games and ball-drills) may induce differences in accumulated external load because of the random intermittency of this training method (12). The profile of those activities performed during training such as distance accumulated in arbitrary categories was considered as criteria to track players' external load in team sports (5). This procedure is now made feasible in the training setup by the advancement in global position system (GPS) technology allowing individual tracking of external load to be a reality.

The external load during specific training in soccer can be objectively assessed (7,10,19,26,34,35) and easily analyzed with dedicated software (11,28). As a result, GPS systems offer a highly practical way of monitoring players' movements during training (20).

To the best of this study author's knowledge, no study has addressed the association between the external and internal load variables experienced during soccer training.

As a result, information gained studying the possible relationship between internal and external loads may have a great practical impact on the development of scientific coaching in soccer.

Therefore, the aim of this study was to examine the relationships of common indicators of internal TL with objective measures of the external TL in soccer (i.e., criterion validity). As a work hypothesis was assumed the association between internal and external load variables.


Experimental Approach to the Problem

In this study, a descriptive correlational design was used. Two popular indicators of internal load were used: the Edward and the sRPE method. The former is an HR-based method that assumes as individual TL the sum of time spent in arbitrary HR-zones weighted multiplying the accumulated time in each HR zone (in minutes) by a relative factor (50–60% HRmax = 1; 60–70% HRmax = 2; 70–80% HRmax = 3; 80–90% HRmax = 4; and 90–100% HRmax = 5). The sRPE, is obtained by multiplying the duration of each training session (in minutes) by the intensity assigned to that session on the RPE scale (15). Despite the practical interest of these indicators of individual response to TL, no study verified their sensibility in tracking variation in training work load (i.e., external load). Indeed, the published articles only assessed the validity of sRPE assessing the relationship with HR-based indicators of internal TL (1,2,16,24,29). Therefore, the association of internal load indicators and criteria of work load produced as a result of training is yet to be investigated.

In this study, players' training activities were monitored using GPS technology, and the resulting activity categories were assumed as constructs representing individual external load. Convergent construct validity of sRPE and Edwards methods was assessed examining their association with objective measures of training activities.


The participants were 28 semiprofessional soccer players (age 22.9 ± 4.2 years, height 177 ± 5 cm, body mass 73.6 ± 4.4 kg) of a Spanish Third Division team and possessing a mean of 12.5 years of playing experience in competitive soccer. All the players were notified of the research design and its requirements and the potential benefits and risks. Each player gave his written informed consent before the commencement of this study. The Ethics Committee of the University of the Basque Country (CEISH) gave its institutional approval before the procedures of this study took place.


The players' external load was monitored and quantified by means of portable GPS devices (MinimaxX, v.4.0, Catapult Innovations) operating at a sampling frequency of 10 Hz and incorporating a 100-Hz triaxial accelerometer. Each player wore a special harness that enabled this device to be fitted to the upper part of his back. The GPS devices were activated 15 minutes before the start of each training session, in accordance with the manufacturer's instructions. After recording, the data were downloaded to a PC and analyzed using the software package Logan Plus v.4.4 (Catapult Innovations, 2010). The reliability and validity of the devices used in this study were reported elsewhere (6,38).

Training HR was assessed using a short-range telemetry system (Polar Team System, Polar Electro Oy, Finland). Individual maximal HR (HRmax) was assessed before commencing the study using the Yo-Yo Intermittent Recovery Test Level 1 (27).

The sRPEs were obtained using the 10-point Borg scale by having players rating their training perceived effort 30 minutes after the end of training according to the procedures suggested by Foster (14).

A total of 44 training sessions were monitored between January and April of the 2009–2010 competitive season. The monitored training sessions took place at least 48 hours between each other and were all performed on the same outdoor artificial grass pitch and at similar times of day (20:00 PM). Each training session began with a 15-minute standard warm-up (running, stretching, and contact with the ball), followed by different drills (small-sided games, running exercises, technical and tactical drills). During the period of this study, no strength-training session was performed by the players.

Two to 3 sessions per week were monitored, with a mean duration of 90.4 ± 23.0 minutes per session. The mean number of GPS devices used per training session was 4.6 (±1.8), with a maximum of 9 players being monitored in any single session. All the observed sessions were designed by the team's head coach and fitness trainer, who supervised all the training sessions. Data analysis included all the activities performed during the training sessions including the recovery periods.

During rest periods, the players were allowed to drink fluids “at libitum.” The players were advised to maintain their normal diet, with special emphasis being placed on a high intake of water and carbohydrates.

The indicators of external load were as follows: (a) TD, total distance covered; (b) DHS, distance covered at high speed (≥18 km·h−1); (c) DSS, distance covered at sprint speed (≥21 km·h−1); (d) WRR, work:rest ratio, defined as the distance covered by the player at a speed ≥4 km·h–1 (period of activity or work) divided by the distance covered at a speed <3.9 km·h–1 (period of recovery or rest); (e) FEHS, frequency of efforts at high speed (≥18 km·h−1); and (f) FESS, frequency of efforts at sprint speed (≥21 km·h−1). A further indicator used was player load, obtained via accelerometry (4,9,31), combining the accelerations produced in 3 planes of body movement by means of a 100-Hz triaxial accelerometer. Player load is a new indicator of the external load, which showed to be highly correlated with both HR and blood-lactate levels (31) and possessing high both interdevice and intradevice reliabilities. This suggesting accelerometers as a viable tool to track activity changes during exercise (4,38). Player load was calculated using the following formula:

where aca is the acceleration in the anteroposterior or horizontal axis, act is the acceleration in the transverse or lateral axis, acv is the acceleration in the vertical axis, i is the current time, and t is time.

Familiarization with the used procedures and devices took place during the week preceding the study.

Statistical Analyses

The data are presented as mean ± SD. The homogeneity of variances was examined with Levene's test. Association between variables was assessed using Pearson correlation coefficients. Magnitude of correlation coefficients was considered as trivial (r < 0.1), small (0.1< r <0.3), moderate (0.3 < r <0.5), large (0.5 < r <0.7), very large (0.7 < r <0.9), and nearly perfect (r > 0.9) and perfect (r = 1) (22). All the statistical analyses were performed using SPSS 16.0 for Windows, with significance being set at p ≤ 0.05.


Mean values for Player load, sRPE, and Edward method were 789.2 ± 224.9, 462.4 ± 237.9, and 216.3 ± 72.6 arbitrary units (AU), respectively. During the training session, players' TD, DHS, and DSS were 6,385.4 ± 1,713.2 m; DHS: 191.3 ± 147.7 m; DSS: 62.6 ± 68.4 m, respectively. The mean training bouts for FEHS and FESS were 39.1 ± 30.1 and 4.8 ± 4.8 bouts, respectively. The mean value for WRR was 0.89 ± 0.41.

Large and very large correlations were reported between Player load and TD (r = 0.70, p < 0.01, Figure 1) and Edwards indicator (r = 0.72, p < 0.01, Figure 2).

Figure 1:
Relationship between the Edwards indicator and the total distance covered for the 210 recordings made (r = 0.72; p < 0.01). “AU” is arbitrary unit.
Figure 2:
Relationship between player load (determined by accelerometry) and the training load indicator obtained via the Edwards method for the 210 recordings made (r = 0.70; p < 0.01). “AU” is arbitrary unit.

The FEHS, DHS, and FESS were moderate to trivially correlated with Edwards values (r = 0.37, 0.25, and 0.18, respectively, p < 0.01, n = 210).

The sRPEs were very large associated with TD (r = 0.74, p < 0.01, Figure 3) and Player load (r = 0.76, p < 0.01, Figure 4).

Figure 3:
Relationship between the session-rating of perceived exertion indicator and the total distance covered for the 210 recordings made (r = 0.76; p < 0.01). “AU” is arbitrary unit.
Figure 4:
Relationship between player load (determined by accelerometry) and the session-rating of perceived exertion indicator for the 210 recordings made (r = 0.74; p < 0.01). “AU” is arbitrary unit.

The FEHS and WRR showed large and small correlations with sRPE (r = 0.64 and −0.29, respectively, p < 0.01, n = 210).

A large correlation (r = 0.57, p < 0.01) was found between sRPE and Edwards methods (Figure 5).

Figure 5:
Relationship between the session-rating of perceived exertion indicator and the training load indicator obtained via the Edwards method for the 210 recordings made (r = 0.57; p < 0.01). “AU” is arbitrary unit.


This is the first study that examined the relationships between indicators of external and internal load in soccer. The main finding of this study was the reported significant association between sRPE and Edwards methods with variables representing the activity performed by players during soccer training. This supports this study's work hypothesis.

The sRPE method has been considered as a viable method to track internal load using no cost and easily accessible procedures as the individual global perception of training effort and total training session time (8). Despite the practical interest of the sRPE method for scientific coaching, the studies that examined method validity considered HR-based indicators as construct of athletes' internal load only.

Borresen and Lambert (2) reported a very-large correlation (r = 0.84) between the sRPE indicator and that obtained via the Edwards method. Studying the intersubject variability in female soccer players TL, Alexiou and Coutts (1) found a similar correlation (r = 0.85) between Edwards and sRPE methods. This adds to what reported by Manzi et al. (29) that monitored eight professional basketball players over 40 training sessions reporting a very large association (r = 0.85). In young soccer player, a very large association but of lower magnitude (r = 0.71) was reported by Impellizzeri et al. (24). In this study, only a large association between sRPE and the Edwards methods (r = 0.57, p < 0.01) was reported.

These discrepancies could be because of the different type of training tasks used in these studies (24). However, the reported association between the 2 indicators of internal TL suggests at best method equivalence only.

The finding of this study showed that the sRPE method was significantly (p < 0.01) related with several indicators of external physical load during training. Indeed, the sRPE was related to TD, Players load, and frequency of efforts provided at high intensity during training. Interestingly, sRPE showed moderate to trivial association with variables representing the activities performed at high intensity and sprinting. This finding partially supports the notion of sRPE as a global measure of TL with limited influence of casual effort provided at high intensity. This confirms the internal validity of the procedures used in this study to assess effort perception (i.e., RPE timing).

The Edwards method uses the individual HR response to training to estimate the internal load of athletes. Although it is popular, the method was not addressed for construct validity in soccer. Furthermore, the strategy used in this method for accounting for differences in exercise strain considers linearity of the HR responses when usually they are nonlinear (30). Despite these limitations, the Edwards method has been used as indicator of TL in several articles. This study results revealed that several indicators of external TL were correlated with the Edwards method. However, similarly to what was found for sRPE, the variables that represented activities performed at high intensity showed small to trivial correlation with the Edwards method. This occurrence may be partly explained by the inherent limitation of HR monitoring in tracking exercise bouts that are eliciting intensities exceeding the individual maximal HR (37).

Team sports involve rapid and nonlinear accelerations and decelerations; consequently, a variable quantifying this momentary variations in work rate may result in being of great interest (36). Game accelerations was reported to be of importance in tracking soccer players' energetic expenditure because they can provide an instantaneous report of activity perturbation during the game (33). However, no study has provided evidence as per validity of acceleration measures during the game or training in soccer against accepted gold standards.

In the attempt to account for instantaneous variation in training activity, the Player load was considered in this study. The results showed that Player load was large to very large associated with both the indicators of internal load used in this study. Those findings are similar to those reported by Montgomery et al. (31) that showed high correlations between Player load and both HR and blood-lactate levels. Despite the Player load not being examined for criterion validity, the findings of this study provide evidence for the interest of this variable in monitoring training effort in soccer players. Furthermore, in the literature, there are studies that showed the validity and reliability of this variable obtained from the accelerometer in MinimaxX v.4.0 devices (4,38).

In the present analysis, Player load was large to very large associated with either indicator of external load (i.e., TD, r = 0.70; Figure 1) and internal load (Edwards and sRPE, r = 0.70 and 0.74, respectively; Figures 2 and 4). These finding may suggest that internal load is related to the volume of accelerations produced by the external load. Interestingly, sRPE showed to be very large associated with Player load supporting the assumption of sRPE as global indicator of exercise intensity (32).

Future studies should examine the validity, reliability, and sensitivity and responsiveness of Player load indicator in different training setup (Impellizzeri and Marcora [23]). In this regard, one of the main limitations of this study was that it did not examine the correlation between TL indicators at different points in the season or across different training sessions (with different content, different work load levels, or on different days of the week), these being factors that could influence TL and alter the relationships between indicators (24,29).

Practical Applications

The results of this study provide evidence for considering the sRPE as a global indicator of individual training response in soccer. Being easy to be performed and inexpensive compared with HR-based methods, sRPE should be regarded as a viable way to track internal load in training setup in soccer (24).

Interestingly, sRPE showed to be related to frequency of effort performed at high intensity during the training sessions compared with the Edward method that is based on HR record and analysis.

In this study, a novel measurement of external load considering accelerations was used. The very-large associations reported between Player load and indicators of internal load suggest the interest of accelerations monitoring in soccer. Training load analysts should take advantage of GPS technology and sRPE and or Edward methods for post hoc TL monitoring in soccer.


This study is part of the project entitled Avances Tecnológicos y Metodológicos en la Automatización de Estudios Observacionales en Deporte, funded by Spain's Dirección General de Investigación, Ministerio de Ciencia e Innovación (PSI2008-01179) in 2008–2011. In addition, the authors thank the Basque Country University (EHU/UPV) and the Department of Physical Education and Sport provided funding. No conflicts of interest exist for this research.


1. Alexiou H, Coutts A. A comparison of methods used for quantifying internal training load in women soccer players. Int J Sports Physiol Perform 3: 320–330, 2008.
2. Borresen J, Lambert MI. Quantifying training load: A comparison of subjective and objective methods. Int J Sports Physiol Perform 3: 16–30, 2008.
3. Borresen J, Lambert MI. The quantification of training load, the training response and the effect on performance. Sports Med 39: 779–795, 2009.
4. Boyd LK, Ball K, Aughey RJ. The reliability of MinimaxX accelerometers for measuring physical activity in Australian football. Int J Sports Physiol Perform 6: 311–321, 2011.
5. Carling C, Bloomfield J, Nelsen L, Reilly T. The role of motion analysis in elite soccer: Contemporary performance measurement techniques and work rate data. Sports Med 38: 839–962, 2008.
6. Castellano J, Casamichana D, Calleja-González J, San Román J, Ostojic S. Reliability and accuracy of 10 Hz GPS devices for short-distance exercise. J Sports Sci Med 10: 233–234, 2011.
7. Coutts A, Duffield R. Validity and reliability of GPS devices for measuring movement demands of team sports. J Sci Med Sport 13: 133–135, 2010.
8. Coutts A, Lovell T, Sirotic AC, Impellizzeri FM. Efficacy of using session-RPE to quantify training load in rugby league. 7th World Congress on Science and Football, Nagoya, Japan, May 26–30, 2011.
9. Cunniffe B, Proctor W, Baker J, Davies B. An evaluation of the physiological demands of elite rugby union using global positioning system tracking software. J Strength Cond Res 23: 1195–1203, 2009.
10. Duffield R, Reid M, Baker J, Spratford W. Accuracy and reliability of GPS devices for measurement of movement patterns in confined spaces for court–based sports. J Sci Med Sport 13: 523–525, 2010.
11. Edgecomb SJ, Norton KI. Comparison of global positioning and computer-based tracking systems for measuring player movement distance during Australian football. J Sci Med Sport 9: 25–32, 2006.
12. Flanagan T, Merrick E. Quantifying the work–load of soccer players. In: Science and Football IV. Spinks W., Reilly T., Murphy A., eds. London, United Kingdom: Routledge, 2002. pp. 341–349.
13. Ford PR, Yates I, Williams AM. An analysis of practice activities and instructional behaviours used by youth soccer coaches during practice: Exploring the link between science and application. J Sports Sci 28: 483–495, 2010.
14. Foster C. Monitoring training in athletes with reference to overtraining syndrome. Med Sci Sports Exerc 30: 1164–1168, 1998.
15. Foster C, Daines E, Hector L, Snyder AC, Welsh R. Athletic performance in relation to training load. Wis Med J 95: 370–374, 1996.
16. Foster C, Florhaug JA, Franklin J, Gottschal L, Hrovatin LA, Parker S, Doleshal P, Dodge C. A new approach to monitoring exercise training. J Strength Cond Res 15: 109–115, 2001.
17. Gabbet T. Reductions in pre-season training loads reduce training injury rates in rugby league players. Br J Sports Med 38: 743–749, 2004.
18. Gabbet T, Jenkins DG. Relationship between training load and injury in professional rugby league players. J Sci Med Sport 14: 204–209, 2011.
19. Gray AJ, Jenkins D, Andrews MH, Taaffe DR, Glover ML. Validity and reliability of GPS for measuring distance traveled in field-based team sports. J Sports Sci 28: 1319–1325, 2010.
20. Hartwig TB, Naughton G, Searl J. Motion analyses of adolescent rugby union players: A comparison of training and game demands. J Strength Cond Res 25: 966–972, 2011.
21. Hill-Haas S, Dawson B, Impellizzeri FM, Coutts A. Physiology of small-sided games training in football. Sports Med 41: 199–220, 2011.
22. Hopkins WG. A scale of magnitudes for effect statistics. 2002. Available at: Accessed March 7, 2009.
23. Impellizzeri FM, Marcora SM. Test validation in sport physiology: Lessons learned from clinimetrics. Int J Sports Physiol Perform 4: 269–277, 2009.
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. Impellizzeri FM, Rampinini E, Marcora SM. Physiological assessment of aerobic training in soccer. J Sports Sci 23: 583–592, 2005.
26. Jennings D, Cormack S, Coutts AJ, Boyd L, Aughey RJ. The validity and reliability of GPS units for measuring distance in team sport specific running patterns. Int J Sports Physiol Perform 5: 328–341, 2010.
27. Krustrup P, Mohr M, Amstrup T, Rysgaard T, Johansen J, Pedersen PK, Bangsbo J. The yo-yo intermittent recovery test: Physiological response, reliability, and validity. Med Sci Sports Exerc 35: 697–705, 2003.
28. MacLeod H, Morris J, Nevill A, Sunderland C. The validity of a non–differential global positioning system for assessing player movement patterns in field hockey. J Sports Sci 27: 121–128, 2009.
29. Manzi V, D'Ottavio S, Impellizzeri FM, Chaouachi A, Chamari K, Castagna C. Profile of weekly training load in elite male professional basketball players. J Strength Cond Res 24: 1399–1406, 2010.
30. Manzi V, Iellamo F, Impellizzeri F, D'Ottavio S, Castagna C. Relation between individualized training impulses and performance in distance runners. Med Sci Sports Exerc 41: 2090–2096, 2009.
31. Montgomery PG, Pyne DB, Minahan CL. The physical and physiological demands of basketball training and competition. Int J Sports Physiol Perform 5: 75–86, 2010.
32. Morgan WP. Psychological components of effort sense. Med Sci Sports Exerc 26: 1071–1077, 1994.
33. Osgnach C, Poser S, Bernardini R, Rinaldo R, Prampero PE. Energy cost and metabolic power in elite soccer: A new match analysis approach. Med Sci Sports Exerc 42: 170–178, 2010.
34. Petersen C, Pyne D, Portus M, Dawson B. Validity and reliability of GPS units to monitor cricket-specific movement patterns. Int J Sports Physiol Perform 4: 381–393, 2009.
35. Portas MD, Harley JA, Barnes CA, Rush CJ. The validity and reliability of 1-Hz and 5-Hz global positioning systems for linear, multidirectional, and soccer-specific activities. Int J Sports Physiol Perform 5: 448–458, 2010.
36. Sheppard JM, Young WB, Doyle TL, Sheppard TA, Newton RU. An evaluation of a new test of reactive agility and its relationship to sprint speed and change of direction speed. J Sci Med Sport 9: 342–349, 2006.
37. Stølen T, Chamari K, Castagna C, Wisløff U. Physiology of soccer: An update. Sports Med 35: 501–536, 2005.
38. Varley MC, Fairweather IA, Aughey RA. Validity and reliability of GPS for measuring instantaneous velocity during acceleration, deceleration, and constant motion. J Sports Sci 30: 121–127, 2012.

association football; training control; session-RPE; heart rate; GPS technology

© 2013 National Strength and Conditioning Association