Evaluation of the match external load in soccer

“The Evaluation of the Match External Load in Soccer: Methods Comparison” by Castagna C et al.
International Journal of Sports Physiology and Performance
© 2016 Human Kinetics, Inc.
Note. This article will be published in a forthcoming issue of the
International Journal of Sports Physiology and Performance. The
article appears here in its accepted, peer-reviewed form, as it was
provided by the submitting author. It has not been copyedited,
proofread, or formatted by the publisher.
Section: Original Investigation
Article Title: The Evaluation of the Match External Load in Soccer: Methods Comparison
Authors: Carlo Castagna1,2
, Matthew Varley3
, Susana Cristina Póvoas Araújo4
and Stefano
D’Ottavio2
Affiliations: 1
Fitness training and biomechanics laboratory, Italian Football Federation
(FIGC), Technical Department, Coverciano (Florence), Italy; 2
University of Rome Tor
Vergata, Rome, Italy; 3
College of Sport and Exercise Science, Victoria University,
Melbourne VIC, Australia; 4
Research Center in Sports Sciences, Health Sciences and Human
Development, CIDESD, University Institute of Maia, ISMAI, Maia, Portugal.
Journal: International Journal of Sports Physiology and Performance
Acceptance Date: August 2, 2016
©2016 Human Kinetics, Inc.
DOI: http://dx.doi.org/10.1123/ijspp.2016-0160

Title of the Article:
The Evaluation of the Match External Load in Soccer: Methods Comparison
Submission Type:
Original Investigation
Authors:
Carlo Castagna1,2, Matthew Varley3, Susana Cristina Póvoas Araújo4 and Stefano
D’Ottavio2
Authors’ Affiliations:
1) Fitness training and biomechanics laboratory, Italian Football Federation
(FIGC), Technical Department, Coverciano (Florence), Italy;
2) University of Rome Tor Vergata, Rome, Italy;
3) College of Sport and Exercise Science, Victoria University, Footscray PO Box
14428, Melbourne VIC 8001, Australia;
4) Research Center in Sports Sciences, Health Sciences and Human Development,
CIDESD, University Institute of Maia, ISMAI, Maia, Portugal.
Contact Details for the Corresponding Author:
Carlo Castagna PhD, via Sparapani 30, 60131, Ancona, Italy;
tel: +39 071-2866532, @mail: castagnac@libero.it
Preferred Running Head: Match External Load in Soccer
Abstract Word Count: 220 words
Text-Only Word Count: 2903 words
Number of Figures and Tables: 1 table and 2 figure
References number: 22 citations
DownloadedbyETSUon09/18/16,Volume0,ArticleNumber0

Abstract
Purpose: The aim of this study was to test the interchangeability of two match-analysis
approaches for external-load detection considering arbitrary selected speeds and metabolic
power (MP) thresholds in male top-class level soccer. Methods: Data analyses were
performed considering match physical performance of 120 team data (1200 player cases) of
randomly selected Spanish, German and English first division championship matches (2013-
14 season). Match analysis was performed with a validated semi-automated multi-camera
system operating at 25 Hz. Results: During a match players covered 10673±348m of which
1778±208m and 2759±241m were performed at High-Intensity using the speed (≥16 km·h-1
,
HI) and metabolic power notations (≥20 watt·kg-1
, MPHI). High-intensity notations were
nearly perfect associated (r=0.93, p<0.0001). A huge method bias (980.63± 87.82m. d=11.67)
was found when considering MPHI and HI. Very large correlations were found between
match total distance covered and MPHI (r=0.84, p<0.0001) and HI (r=0.74, p<0.0001).
Players high-intensity decelerations (≥-2 m·s2
) coverage was very largely associated with
MPHI (r=0.73, p<0.0001). Conclusions: The results of this study showed that the speed and
MP methods are highly interchangeable at relative (magnitude rank) but not absolute
(measure magnitude) level. The two physical match analysis methods can be independently
used to track match external-load in elite level players. However match-analyst decisions
must be based on single method use in order to avoid bias in external-load determination.
Key word: High-Intensity, Association Football, Match Analysis, High-intensity, Metabolic
Power

Introduction
In modern soccer training control and regulation is regarded as a relevant
methodological procedure to optimize training adaptations to maximize match performance 1-
3
. Training progress is the result of the interplay of external and internal loads imposed on
players during training sessions 2
. Although physiological adaptations are mediated by
internal load functional variation the doses necessary for obtaining them are practically seized
monitoring training external load 4
. The recent exponential advancement of match analysis
systems such as multi-camera and Global Position System Technology has enabled the
evaluation of player’s external load during specific training in elite and sub-elite competitive
and recreational soccer 5,6
. Besides the replicability and accuracy of match analysis hardware
of vital importance is the validity and reliability of the variables used to describe player’
activities constituting the back bone of external load evaluation. External training load is
usually assessed evaluating distances and time performed in arbitrary selected speed
categories 1,4
. This method approach has been used for practical issues as the data
interpretation is straightforward for training (i.e. sprinting, speed-endurance training) and not
requiring for definition consideration of acceleration calculations that would need the use of
often not sustainable devices 7,8
. Despite the interest in information obtained through the
speed method, if acceleration is not considered the actual nature of soccer specific training
results is underestimated 1,4
. Recently a metabolic approach (MP) was proposed to provide an
instantaneous picture of soccer specific activities7
. This method considers acceleration and
speed to profile individual distances and time spent by players at arbitrary chosen estimated
power thresholds 4,7
.
The metabolic approach assumes that the energy produced by a player during actual
match-play is a direct result of the product of the running cost from acceleration and the
corresponding instantaneous speed 7
. The relative cost of accelerations is evidenced from the

steady state O2 of running at an inclination that corresponds to the supposed body inclination
of the player during the acceleration bout. The theoretical framework of the MP relies on the
assumed constant energetic cost of running across players and on the estimation of the
energetic cost of acceleration from a mainly aerobic exercise performed in a stationary to
quasi stationary status 7,9
. Furthermore, a strong relationship between acceleration and body
inclination is postulated when player can variate acceleration ratio with no significant
variation on body posture 7
. Additionally, MP estimates the energetic cost of accelerations
from incline running performed by endurance up-hill runners in an laboratory set-up
questioning the contextual validity of this assumption for soccer 9,10
. Despite these theoretical
and practical incongruences that pose questions about the internal validity of this novel
approach, several papers were recently published a-critically considering this issue 10
.
Furthermore match-analysis systems software provide by defaults MP variables with the aim
to profile player game performance during soccer-specific training. However the actual
superiority of the MP versus the classic arbitrary speed zone approach has yet to be examined
in detail, posing doubt about the effectiveness of this interesting novel approach.
While the MP approach may provide a more detailed tracking of player’s game
activity, there is limited research regarding the validity of this method 4,7
. Furthermore in the
paper that firstly proposed the MP approach no objective evaluation was provided by authors
regarding the actual superiority (i.e. statistical verification) and interchangeability of the
metabolic approach with the classic speed method. Indeed in the Osgnach et al.7
study only
descriptive statistics were reported (i.e. means and standard deviations). Thus, information
regarding the objective difference and or association of the MP approach over the speed
threshold method is unknown. Additionally being the MP approach based on acceleration
calculation the associated error using common video-tracking systems may provide a large
measurement bias 7
. This information has huge practical implications as a growing number of

match analysis systems are reporting players activity with the MP variables in the attempt to
characterise match and training external-load.
Therefore the aim of this descriptive study was to examine the association between
classic speed and MP approach in tracking the external-load of elite level players during
highly competitive official matches. It was hypothesized that there would be a large
association between the two match analysis approaches.
Methods
Subjects
Match physical performance was assessed in 1200 male outfield soccer-players (age
24.5±0.8, height 176.4±4.5 cm, body mass 74.6 kg) playing in the first division
Championships of Germany (Bundesliga 1), England (English Premier League) and Spain
(Liga BBVA) during the 2013-14 season. Written informed consent was obtained from
players organisation to treat anonymously the collected data for research purpose. This study
design was approved by the Institutional Research Board before commencement of this
study.
Design
With the aim to examine methods interchangeability a descriptive correlation design
was considered. Match physical-performance was evaluated in professional top-class players
competing during European first-division championships matches. This provided population
specificity and internal validity of this research design. Association between selected
arbitrary speeds and MP zones was performed tracking players match activities with a
validated multi-camera semi-automatic system (operating at 25 hz) 11,12
. Games were
randomly selected from a proprietary match database in order to warrant external validity.

Methodology
Sixty official matches were considered for calculations, they were randomly selected
in reason of 20 from each national league (Germany, England and Spain) database. Players
were tracked for the entire duration of the game using a multicamera semi-automatic system
(Prozone Sport, Leeds, UK) working at 25 hz 11,12
. Home and away teams data were available
for each of the 60 games considered. Players were tracked for match physical-performance
using either speed and MP arbitrary selected intensity zones. In order to examine match
tempo of the most competitive leagues of the world only physical match variables tracking
high-intensity activities were considered. They were as follows;
 Distance covered at High-Intensity (speed ≥ than 16 km·h-1
, HI);
 Distance covered at High-Intensity Running (speed ≥18.97≤21.99 km·h-1
, HIR);
 Distance covered at Very High-Intensity Running (speed>21.99 km·h-1
, VHIR);
 Distance covered with High-Intensity Accelerations (acceleration ≥2 m·s2
, DAcHI);
 Distance covered with High-Intensity Decelerations (deceleration ≤-2 m·s2
, HIDec);
 Distance covered with Very High-Intensity Deceleration (deceleration ≤ -3 m·s2
,
VHIDec);
 Distance covered with Very High-Intensity Acceleration (acceleration ≥ 3 m·s2
,
VHIAcc);
 Distance covered at High-Intensity (MP≥20 watt·kg-1
, MPHI).
To characterize global match coverage the total distance (TD) covered and the
average MP (AMP) were calculated. Data were processed with proprietary software (K-
SportOnline, K-Sport, Montelabbate, PU, Italy) and then analysed with commercially
available spreadsheets (Excel, Microsoft, USA) and with a dedicated statistical package
(Statistica 10, Statsoft, USA).

Statistical Analysis
Data are presented as mean ± standard deviation and confidence interval (95%CI).
Assumption of normality was verified using the Shapiro-Wilk W-test. Variables association
was assessed using Pearson’s product-moment correlation coefficients. Qualitative magnitude
of associations was reported according to Hopkins (2002) as follows: 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, nearly perfect r
> 0.9 and perfect r= 1. Partial correlations were used for path analysis when necessary.
Differences between variables were assessed with paired t-tests using a Bonferroni correction
to account for comparison number. The Cohen’s d was used to assess effect-size (ES) 13
.
According to Hopkins et al. 14
ES of above 4, between 4 and 2, between 2 and 1.2, between
1.2 and 0.6, between 0.6 and 0.2 and 0.2 and 0 were considered as huge, very large, large,
moderate, small, and trivial respectively. In order to provide normative cues for metrics
changes the Smallest Worthwhile Change (SWC) was considered according to Hopkins et al.
14
Measure agreement was assessed with Bland and Altman plots with bias test against the
zero difference hypothesis for significance. Significance was set at p 0.05. Preliminary
power calculation showed that to obtain the a statistical power of 80% would be necessary
400 cases. The final power obtained with this study design was higher than 90% (1200
cases).
Results
Descriptive statistics and SWC for the considered variables depicting match physical-
performance are reported in table 1. Data showed high inter-match (CV>10%) variations for
all the considered variables but not for TD, AMP and MPHI. Bland and Altman plot statistic
showed a significant measurement bias of 980.63± 87.82m (95%CI 964.75  996.51,
p<0.0001) between MPHI and HI distances. Upper and lower 95% limits of agreement were
1152.77 (95%CI 1125.56  1179.97) and 808.50m (95%CI 781.29  835.70) respectively

(Figure 1). Plot of MPHI and HI difference vs mean showed the existence of a moderate data
heteroscedasticity (r= 0.38, p<0.0001, 95%CI 0.22  0.53) suggesting the likelihood of a
systematic measurement error as variables magnitude increased.
A nearly perfect association was found between MPHI and HI (r=0.93, p<0.0001,
95%CI 0.91 0.95, Fig. 2). The MPHI was very largely associated with TD (r=0.84,
p<0.0001, 95%CI 0.78  0.89) HIDec (r=0.73, p<0.0001, 95%CI 0.63  0.80) and HIR
(r=0.87, p<0.0001 95%CI 0.82  0.91). The HI showed very large correlation with TD
(r=0.74, p<0.0001, CI95% 0.64  0.81) and AMP (r=0.73, p<0.0001, 95%CI 0.63  0.80).
The AMP was very largely associated with TD (r=0.85, p<0.0001, 95%CI 0.79  0.89) and
deceleration categories VHIDec (r=0.72, p<0.0001, 95%CI 0.62 0.79) and , HIDec (r=0.76,
p<0.0001, 95%CI 0.68  0.83).
Discussion
This is the first study to assess the associations between two data analysis methods for
assessing external load in soccer, specifically; speed and metabolic power of arbitrarily
chosen activity categories. The main finding of this descriptive comparative study was the
almost perfect association (r=0.93) between the distance covered at a high-intensity speed
(≥16 km·h-1
) with that accumulated at high-intensity using the MP approach (≥20 watt·kg-1
).
However, practical very large absolute differences (i.e. 52%) in variable magnitude were
detected between HI and MPHI distance covered. These results confirm the original work
hypothesis for relative but not absolute interchangeability of the two external load approaches
here considered.
Match and training high-intensity is considered as a relevant construct in modern
soccer with a number of papers providing direct or indirect evidence 1,15
. This study used two
methods considering arbitrary categories to depict match high-intensity activity according to

the methods of Osgnach et al. 7
. However, in the original paper the authors failed to report
any quantitative basic statistic to compare methods and as a result information about measure
interchangeability was not provided. The speed (≥16 km·h-1
) and MP (≥20 watt·kg-1
)
thresholds used in this study were assumed to represent the average speed and corresponding
relative metabolic power at the estimated maximal aerobic-speed in professional soccer
players 7
. This enabled direct comparison between the two considered methods for external-
load detection to evaluate their concurrent (i.e. magnitude association) and convergent
construct (i.e. measure agreement) validity. The results of this study showed that speed and
MP approach possess an almost perfect relative (i.e. r= 0.93) but a poor absolute
interchangeability (measurement bias of 981m). This suggest that the two external-load
methods are sensible in depicting player’s high-intensity activity reported as distance covered
in arbitrary match high-intensity categories thresholds. However, the reported huge difference
in absolute values between the HI and MPHI underlines the difference in the informing
criteria of the two methods.
The lower HI distance compared to MPHI was likely the result of the inability of the
speed threshold method to consider the accelerative phases of high-intensity efforts thus
underestimating actual match HI demands. Indeed using the speed approach HI distance
accumulates only when players exceed the set velocity threshold (≥16 km·h-
1), thus
neglecting the preparatory phase involving high-intensity accelerative efforts 4
. Given these
limitations in the speed approach the metabolic power approach was introduced in the
attempt to account for instantaneous acceleration gradients during actual match-play 7
.
Despite the theoretical interest provided by the MP approach its essential use of
acceleration data provide some concern about the validity of this novel method 4,10
. Indeed
the acceleration data per se have an inherent systematic error consequence of filtering and
sampling rate 7
. The reported error in acceleration data collection and the required technology

for valid measurement limits the practical interest of the MP approach 10
. Additionally the
reported iso-power phenomenon considered as the likelihood of producing the same MP
values with an unpredictable variation in instantaneous acceleration and speed data, may
produce construct derived artefacts 7
. Indeed consideration for high-intensity bouts can be
done when actually they are in a lower intensity phenomenological domain.
Data difference of supposed convergent constructs was reported as a practical
representative of measurement variability (causal and random changes) 16
. Explorative data
analysis showed large association between MPHI and HI absolute difference with distance
covered (r=-0.62, p<0.0001, from -0.72 to -0.50,) and time spent at DAcHI (r= -0.70,
p<0.0001, from -0.78 to -0.60). This finding may represent an effect of speed maintenance
over the MPHI production during the highly competitive matches considered in this study.
The very large (r= 0.87, p<0.0001, 95%CI 0.82  0.91) association between HIR and MPHI
may partially confirm this assumption. Furthermore when controlling for HIR the association
between MPHI and HI changed from almost perfect into the lower range of the very large
correlation categories (i.e. from 0.93 to 0.74). Additionally HI showed to be only moderately
correlated with acceleration and deceleration performed during the match by players (r from
0.44 to 0.51, p<0.0001) confirming the lack of sensitivity of the speed threshold approach in
accurately tracking player’s match high-intensity activities. Although more thorough analyses
are necessary, the results of this study support the idea that the effect of movement speed on
MP is more pronounced than the supposed added advantage of considering acceleration as a
variable informing energy cost. Further studies providing more detailed information about the
real genesis of MP are warranted.
The data reported in this study were captured with a semi-automated multi-camera
system previously tested for validity and reliability and used for disseminating match
performance in most of the relevant reports published in soccer performance profiling

5,11,12,17
. Methods comparison was carried-out studying match physical performance in
players participating in the most important championships of the world such us English
Premier League (United Kingdom), Liga BBVA (Spain) and Bundesliga 1 (Germany). Given
the number of matches considered and the number of cases processed, this data possesses a
great internal validity as it is representative of elite level soccer. Match analysis data showed
to be in line with those provided by previous descriptive studies using similar or the same
video capture system11,18-21
. Indeed Osgnach et al.7
introducing the MP approach studied a
population of Italian professional soccer players that were reported to cover a match total
distance of 10,950 ± 1044 m with 1996m and 2839m covered at HI and MPHI respectively.
This study was the first to report random match data pooled from different European top class
leagues. Given that, the data reported may be considered as reference for match-analysts and
coaches interested in physical match activities (Table 1).
Data variability is a key factor in match analysis to determine the probability of causal
match to match changes in physical performance and to detect team or individual fitness and
or technical-tactical differences. In this study absolute inter-teams variability (i.e. CV%) were
below 5% for TD, AMP showing high measurement stability for these global match-activity
variables 22
(Table 1).
Higher CV% values were found in this study for the variables addressing match high-
intensity (i.e. C>10%) with MPHI showing a CV lower than 10%. These findings are in line
with those reported by other authors for English Premier League championship matches
confirming the high variability of high-intensity activities in top class soccer 18,19
. In this
context acceleration and decelerations showed the highest variability with CV above 20 and
25% respectively. In order to detect casual changes the SWC may result of great practical
interest. 14
According to this study data changes in the team mean from 2.3 and 1.8% may be
regarded as causal match performance variations for HI and MPHI respectively.

Practical Applications
Match and training external-load may provide key information for training
individualisation. Particularly at the elite level absolute performance variables may better
depict the intra individual and team differences. Thus, the use of valid and reliable methods
to profile match and training internal load result are vital. Given this study data the speed and
metabolic approaches showed similar sensitivity in depicting player and team profiles of the
supposed high-intensity constructs. Indeed the reported almost perfect association between
HI and MPHI provided evidence of construct converge (i.e concurrent validity). However
despite relative method interchangeability the reported very large absolute measure
differences (d > 4, ES) suggest accurate a priori choice of the data analysis method. The use
of the MP approach to provide concurrent consideration of speed and acceleration requires
further examination into the issue of population validity of the equation used for acceleration
derived energy-cost calculation and criterion validity. Additionally the provided equation
should be refined in order to account for collisions, tackles, jumps and non-orthodox
exercise-modes like sideward and backward running 1,10
.
Conclusions
The scientific limitations of the MP approach should be acknowledged by
practitioners when using this measure for reporting purposes. Moreover, recent validation
studies suggest caution in using MP when evaluating the external load in soccer players
during pre-planned shuttle running10
. Finally, the mismatch between the considerable higher
estimated anaerobic contribution to soccer match physical-performance and actual
physiological measures question the internal validity of the MP approach7,10
.

Acknowledgments
No financial support was provided for the completion of this study. The authors declare no
conflict of interest with the finding reported in this study.

References
1. Stølen T, Chamari K, Castagna C, Wisløff U. Physiology of Soccer: An Update.
Sports Med. 2005;35:501-536.
2. Impellizzeri FM, Rampinini E, Coutts AJ, Sassi A, Marcora SM. Use of RPE-based
training load in soccer. Medicine and science in sports and exercise. 2004;36:1042-
1047.
3. Impellizzeri FM, Rampinini E, Marcora SM. Physiological assessment of aerobic
training in soccer. J Sports Sci. 2005;23:583 – 592.
4. Manzi V, Impellizzeri F, Castagna C. Aerobic fitness ecological validity in elite
soccer players: a metabolic power approach. Journal of strength and conditioning
research / National Strength & Conditioning Association. 2014;28:914-919.
5. Buchheit M, Allen A, Poon TK, Modonutti M, Gregson W, Di Salvo V. Integrating
different tracking systems in football: multiple camera semi-automatic system, local
position measurement and GPS technologies. Journal of Sports Sciences.
2014;32:1844-1857.
6. Randers MB, Nielsen JJ, Bangsbo J, Krustrup P. Physiological response and activity
profile in recreational small-sided football: no effect of the number of players.
Scandinavian Journal of Medicine & Science in Sports. 2014;24 Suppl 1:130-137.
7. Osgnach C, Poser S, Bernardini R, Rinaldo R, di Prampero PE. Energy cost and
metabolic power in elite soccer: a new match analysis approach. Med Sci Sports
Exerc. 2010;42:170-178.
8. 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. 2008;38:839-862.
9. Minetti AE, Moia C, Roi GS, Susta D, Ferretti G. Energy cost of walking and running
at extreme uphill and downhill slopes. J Appl Physiol. 2002;93:1039-1046.
10. Stevens TG, De Ruiter CJ, Van Maurik D, Van Lierop CJ, Savelsbergh GJ, Beek PJ.
Measured and estimated energy cost of constant and shuttle running in soccer players.
Medicine and science in sports and exercise. 2015;47:1219-1224.
11. Sarmento H, Marcelino R, Anguera MT, CampaniCo J, Matos N, LeitAo JC. Match
analysis in football: a systematic review. Journal of Sports Sciences. 2014;32:1831-
1843.
12. Di Salvo V, Collins A, McNeill B, Cardinale M. Validation of Prozone:A new video-
based performance analysis system International Journal of Performance Analysis in
Sport. 2006;6:108-119.
13. Cohen J. Statistical power analysis for the behavioral sciences. Hillsdale, NJ:
Lawrence Erlbaum Associates, 1988.

14. Hopkins WG, Marshall SW, Batterham AM, Hanin J. Progressive statistics for studies
in sports medicine and exercise science. Medicine and science in sports and exercise.
2009;41:3-13.
15. Impellizzeri FM and Marcora SM. Test validation in sport physiology: lessons learned
from clinimetrics. International Journal of Sports Physiology and Performance.
2009;4:269-277.
16. Bland JM and Altman DG. Statistical methods for assessing agreement between two
methods of clinical measurement. Lancet. 1986;1:307-310.
17. Randers MB, Mujika I, Hewitt A, Santisteban J, Bischoff R, Solano R, Zubillaga A,
Peltola E, Krustrup P, Mohr M. Application of four different football match analysis
systems: a comparative study. Journal of Sports Sciences. 2010;28:171-182.
18. Di Salvo V, Gregson W, Atkinson G, Tordoff P, Drust B. Analysis of high intensity
activity in Premier League soccer. International journal of sports medicine.
2009;30:205-212.
19. Gregson W, Drust B, Atkinson G, Salvo VD. Match-to-match variability of high-
speed activities in premier league soccer. International journal of sports medicine.
2010;31:237-242.
20. Rampinini E, Coutts AJ, Castagna C, Sassi R, Impellizzeri FM. Variation in top level
soccer match performance. Int J Sports Med. 2007;28:1018-1024.
21. Rampinini E, Impellizzeri FM, Castagna C, Coutts AJ, Wisloff U. Technical
performance during soccer matches of the Italian Serie A league: effect of fatigue and
competitive level. Journal of science and medicine in sport / Sports Medicine
Australia. 2009;12:227-233.
22. Atkinson G and Nevill AM. Statistical methods for assessing measurement error
(reliability) in variables relevant to sports medicine. Sports Medicine. 1998;26:217-
238.

Figure 1. Bland and Altman Plot of the MPHI and HI variables.

Figure 2. Relationship (r=0.93) between match distances performed at high-intensity using
the metabolic power (MPHI) and speed threshold (HI) notations.
R² = 0.8734
2000
2200
2400
2600
2800
3000
3200
3400
3600
1200 1400 1600 1800 2000 2200 2400
MPHI(m)
HI (m)

Table 1 Values of match activities considered in this study.
Variable TD AMP HI HIR VHIR MPHI HIAcc HIDec VHIDec VHIAcc
Mean 10672.79 10.65 1778.34 482.04 239.81 2758.97 635.93 611.91 209.30 214.81
SD 347.74 0.49 208.00 67.46 48.22 240.99 118.21 97.68 47.75 55.90
CV% 3.26 4.61 11.70 13.99 20.11 8.73 18.59 15.96 22.81 26.02
Min 9417.76 9.14 1156.22 298.91 127.86 2028.14 397.10 437.58 120.24 106.44
Max 11595.77 12.09 2310.62 656.80 335.05 3344.14 911.68 889.97 380.48 370.11
SWC 69.55 0.10 41.6 13.49 9.64 48.20 23.64 19.54 9.55 11.18

Evaluation of the match external load in soccer

More Related Content

What's hot (20)

Viewers also liked (11)

Similar to Evaluation of the match external load in soccer (20)

More from Fernando Farias (19)

Recently uploaded (20)

Evaluation of the match external load in soccer