Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=gspm20

Research in Sports MedicineAn International Journal

ISSN: 1543-8627 (Print) 1543-8635 (Online) Journal homepage: http://www.tandfonline.com/loi/gspm20

Variations of external load variables betweenmedium- and large-sided soccer games inprofessional players

Filipe Manuel Clemente, Hugo Sarmento, Alireza Rabbani, Cornelis M. I.(Niels) Van Der Linden, Mehdi Kargarfard & Israel Teoldo Costa

To cite this article: Filipe Manuel Clemente, Hugo Sarmento, Alireza Rabbani, Cornelis M. I.(Niels) Van Der Linden, Mehdi Kargarfard & Israel Teoldo Costa (2018): Variations of external loadvariables between medium- and large-sided soccer games in professional players, Research inSports Medicine, DOI: 10.1080/15438627.2018.1511560

To link to this article: https://doi.org/10.1080/15438627.2018.1511560

Published online: 21 Aug 2018.

Submit your article to this journal

View Crossmark data

Variations of external load variables between medium- andlarge-sided soccer games in professional playersFilipe Manuel Clemente a,b, Hugo Sarmento c, Alireza Rabbani d,Cornelis M. I. (Niels) Van Der Lindene, Mehdi Kargarfardd and Israel Teoldo Costaf

aSports Sciences Department, Instituto Politécnico de Viana do Castelo, Escola Superior de Desporto eLazer, Melgaço, Portugal; bApplied Mathematics department, Instituto de Telecomunicações, Delegação daCovilhã, Covilhã, Portugal; cResearch Unit for Sport and Physical Activity (CIDAF), Faculty of Sport Sciencesand Physical Education, University of Coimbra, Coimbra, Portugal; dDepartment of Exercise Physiology,Faculty of Sport Sciences, University of Isfahan, Isfahan, Iran; eDepartment of Sport Sciences, JOHAN sports,Noordwijk, The Netherlands; fCentre of Research and Studies in Soccer, Department of Physical Education,Universidade Federal de Viçosa, Viçosa, Brazil

ABSTRACTThis study compared the relative physical demands of official matchesand sided games (medium and large) in professional soccer players bymeans of a global positioning system. Twenty-three professional malesoccer players (24.63� 2.84 years old; 180.94� 6.49 cm;77.19� 6.46 kg; 52.99� 5.01 VO2max) participated in the study. Totaldistance, running distance, sprinting distance, number of sprints, andacceleration sumwere quantified per minute to compare the differentgames. Running distance in full match was greater than in 5vs5+GK(d = 2.303, moderate effect), 6vs6+GK (d = 1.719, moderate effect) and9vs9+GK (d = 1.084, minimum effect) sided games. Greater values forsprinting distance were found in the full match compared to 5vs5+GK(d = 3.673, strong effect), 6vs6+GK (d = 2.606,moderate effect) and 9vs9+GK (d = 1.903, moderate effect) sided games. However, the load wasgreater in the 5vs5+GK game compared to the 6vs6+GK (d = 1.323,moderate effect) and 9vs9+GK (d = 1.030, minimum effect) games andthe full match (d = 1.478, moderate effect). This study revealed thatmedium-sided games are not appropriate for simulating the sprintingconditions of official full matches. However, medium-sided games aremore intense than full matches in that accelerations are made moreoften in medium-sided games.

ARTICLE HISTORYReceived 21 December 2017Accepted 9 August 2018

KEYWORDSAssociation football; trainingload; Global positioningsystem (GPS); small-sidedgames; drill-based activities

Introduction

Research on match analyses in adult male football players has been the subject of growinginterest over the past five years (Sarmento et al., 2017). Nevertheless, a big challenge forresearchers is aligning new findings with the needs of coaching staff, through a moreinteractive relationship between all practitioners, to produce relevant information that canimprove players’ performance through constant adaptations of training design (Sarmentoet al., 2017).

CONTACT Filipe Manuel Clemente filipe.clemente5@gmail.com Complexo Desportivo e Lazer de Melgaço –Monte de Prado, 4960-320 Melgaço, Portugal

RESEARCH IN SPORTS MEDICINEhttps://doi.org/10.1080/15438627.2018.1511560

© 2018 Informa UK Limited, trading as Taylor & Francis Group

Training methods in football have evolved over the years (Selmi, Gonçalves, Ouergui,Sampaio, & Bouassida, 2018), from privileged exercises without a ball which developphysical capacities to new methods and exercises which simultaneously improve physicalcapacities along with technical and tactical skills in accordance with the modern demandsof the game (Mello et al., 2017; Selmi, Marzouki, Ouergui, BenKhalifa, & Bouassida, 2018). Inthis sense, small-sided games (SSGs), also referred to as skill-based conditioning games(Gabbett, 2006), game-based training (Gabbett, Jenkins, & Abernethy, 2009) or small-sidedand conditioned games (Davids, Araújo, Correia, & Vilar, 2013) are modified games playedon small fields with modified rules and which involve a fewer players than traditionalsoccer games (Hill-Haas, Dawson, Impellizzeri, & Coutts, 2011; Silva, Garganta, Santos, &Teoldo, 2014). SSGs integrate all the specific needs of soccer and represent a useful solutionto improve the efficiency of the training process (Beato, Jamil, & Devereux, 2018).

The evolution of technological devices (e.g. GPSs, heart rate monitors, etc.) helpcoaches and scientists study specific characteristics of different SSGs, and the progres-sion of SSG research in soccer has increased exponentially in the past few years (Ade,Harley, & Brandley, 2014; Halouani, Chtourou, Gabbett, Chaouachi, & Chamari, 2014; Hill-Haas et al., 2011). The scientific evidence suggests that physiological responses (e.g.perceived exertion, heart rate, and blood lactate concentration) and tactical and tech-nical skill requirements can be modified during SSGs by altering certain factors, such asthe number of players, the size of the pitch, the rules of the game, and coach encour-agement (Aguiar, Botelho, Lago, Maças, & Sampaio, 2012; Castelão, Garganta, Santos, &Teoldo, 2014; Rampinini et al., 2007; Silva et al., 2014).

Nonetheless, some limitations remain in the published studies. Specifically, there is alack of consistency in the design of SSGs, the age and level of ability of the players, thelevel of coach encouragement, and the rules (e.g. number and type of goals, number oftouches in the ball, etc.) used by researchers. Such inconsistencies make it difficult tocompare studies. Additionally, some potential weaknesses may be apparent in morerecently published research, such as the small sample sizes used in some studies(Beenham et al., 2017; Halouani et al., 2014; Silva, Vilar, Davids, Araújo, & Garganta,2016; Stevens, De Ruiter, Beek, & Savelsbergh, 2016). Additionally, the main body ofresearch in this area is conducted with young players (Clemente, Wong, Martins, &Mendes, 2014; Halouani, Chtourou, Dellal, Chaouachi, & Chamari, 2014; Halouani,Chtourou, Gabbett, et al., 2014), while studies conducted with senior players (McLeanet al., 2016; Ric et al., 2016; Stevens et al., 2016), and especially with professional players,are rare.

In this sense, a study that analyzes the training process of an adult professional teamby monitoring the most common types of sided games (medium and large) andcomparing their specific demands with the activity profile of official matches seems tohave emerged as a necessity. Based on this, the purpose of this study was to comparethe relative physical demands of official matches and sided-games (medium and large)in professional soccer players by means of a global positioning system (GPS).Comparisons between playing positions were also conducted

2 F. M. CLEMENTE ET AL.

Methods

Subjects

Twenty-three professional male soccer players (24.63� 2.84 years old; 180.94� 6.49 cm;77.19� 6.46 kg; 52.99� 5.01 VO2max) participated in the study. They were all membersof the same team that competes in Portuguese Second League. Participants includedfour wide defenders (WDs), four central defenders (CDs), nine central midfielders (CMs),three wide forwards (WFs) and three strikers (STs). The data was collected from dailyplayer external load monitoring in which player activities were routinely measuredduring training sessions and matches. Participants agreed in to participate in the data-collection process and signed a free informed consent about the purpose of the study.The procedures followed the Declaration of Helsinki for the study of human beings.

Experimental approach

In this study, medium-sided (5vs5+GK and 6vs6+GK) and large-sided (9vs9+GK) gameswere compared with official full matches in terms of the activity profiles of the players. Anonexperimental descriptive comparative design was used to inspect possible differ-ences between sided games and full matches over arbitrary categories normalized perexercise duration. Activity profiles were obtained using 10-Hz GPS technology whichenabled us to monitor players’ movements in a valid and reliable manner (Aughey, 2011;Malone, Lovell, Varley, & Coutts, 2017). Comparisons were conducted by analyzing maleprofessional soccer players for four weeks (August: Microcycles 9 to 12) during the2017–2018 competitive season. The training sessions were conducted between 10:00and 11:30 a.m. on natural turf with sunny weather conditions and similar temperatures(21.2� 2.3°C). The official matches were played in both home and away conditions,between 4:00 and 6:00 p.m. on natural turf with sunny conditions and similar tempera-tures (25.7� 1.7°C).

Data collection

A 10-Hz GPS (JOHAN Sports, Noordwijk, The Netherlands), consisting of a GPS sensor(10 Hz, including EGNOS correction), an accelerometer, a gyroscope, and a magnet-ometer (100 Hz, 3 axes, � 16 g) was used to monitor the players. The value of 10 Hzseems to be valid and reliable enough to measure position and speed in a sports setting(Scott, Scott, & Kelly, 2016). The GPS sensor used in this study was tested with a 2.5 �0.41% (error � deviation) reliability for in-line movements and change of direction(Nikolaidis, Clemente, van der Linden, Rosemann, & Knechtle, 2018). Each participantwore a body-tight vest to ensure valid (i.e. body-oriented) accelerometer data. A motiontracker was then placed in a pocket of the vest that was located at the dorsal region ofthe players. After each training session, motion data from the trackers were uploaded tothe JOHAN Sports online analysis platform. Normalized data per minute was treated toderive total distance (TD, m/min), running distance (RD, m/min at 14–20 km:h�1),sprinting distance (SD, m/min at >20 km:h�1), number of sprints (NS: n:min�1), andacceleration sum. Acceleration sum (AS) was calculated as the sum of the squared rates

RESEARCH IN SPORTS MEDICINE 3

of change in acceleration in n consecutive moments of the sided games along the threemovement axes, where ay represents acceleration along the forward-backward axis, axrepresents acceleration along the sideways axis, and az represents acceleration alongthe vertical axis (see Equation 1) (Boyd, Ball, & Aughey, 2011).

AS ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiaxn � axn�1ð Þ2 þ ayn � ayn�1ð Þ2 þ azn � azn�1ð Þ2

q(1)

Medium- and large-sided games

The most commonly-used SSGs played during the four-week period of the study werecompared with full matches. These SSGs were (i) 5vs5+GK, dimensions: 40 × 31 m,average area per player (excluding GKs): 124 m2, repetitions: 2, duration: 6 min; restingtime: 3 min; (ii) 6vs6+GK, dimensions: 45 × 32 m, average area per player (excludingGKs): 120 m2, repetitions: 3, duration: 7.5 min; resting time: 3 min; and (iii) 9vs9+GK,dimensions: 70 × 50 m, average area per player: 194 m2, repetitions: 2, duration: 11 min;resting time: 3 min. The format and size of the pitch were selected based on previousstudies of medium- and large-sided games (Owen, Wong, Paul, & Dellal, 2014). All theparticipants took part in the sided games during training sessions.

A ball was immediately replaced every time it crossed the sideline or end line duringsided games. There were four balls placed around the pitch to make this replacementprocess quick. Coaches gave verbal encouragement to players during the exercises. Allsided games were implemented immediately after a warm-up that constituted moderaterunning, dynamic stretching, balance and agility exercises, and accelerations. This pro-cess helped to ensure similar conditions across all sided games. The sided games wereplayed in the middle training days of the week (i.e. 3–4 days before an official match and2–3 days after an official match).

Official matches

Match activities were assessed from data collected over four official 11-a-side games.Each microcycle finished with an official match. The system of play used by the observedteam in all the full matches was 1-4-3-3, comprising 2 WDs, 2 CDs,3 CMs, 2 WFs, and 1ST. A player had to have played for a minimum of 45 minutes in a match for their data tobe included in the data analysis for any given match. Each game lasted ~93 minuteswith a 15-minute halftime break. The same warm-up protocol was conducted beforeeach game, including moderate running, dynamic stretching, mobility and balanceexercises, agility exercises, accelerations, and ball possession drills in a 5vs5 format.The playing area relative to each player was approximately 260 m2 across the fourgames, and the length-width ratio of the pitch was 1.45:1.

Statistical procedures

Data are presented as mean, standard deviation and 90% confidence intervals. One wayANOVA was executed to analyze the variance of distance covered, running distance,sprint distance, number of sprints and acceleration sum between sided games and full

4 F. M. CLEMENTE ET AL.

matches. The partial eta squared (η2p) has tested the effect size of ANOVA (ES). TheFerguson’s classification for the ES was used (Ferguson, 2009): no effect (ES < 0.04);minimum effect (0.04 < ES < 0.25); moderate effect (0.25 < ES < 0.64); and strong effect(ES > 0.64). The pairwise comparisons were tested with Cohen’s d to analyze the effectsize. The following classification to measure the magnitude of ES was used (Ferguson,2009): no effect (d < 0.41), minimum effect (0.41 < d < 1.15), moderate effect(1.15 < d < 2.70) and strong effect (d > 2.70). All statistical analysis was carried outusing SPSS statistical analysis software (SPSS version 23.0, Chicago, USA). The level ofstatistical significance was set at p � 0.05.

Results

Descriptive statistics (mean and 90% confidence intervals) of distance covered, runningdistance, sprinting distance, acceleration sum and number sprints per minute can be foundin Figure 1. One-way ANOVA tested the variance of dependent variables between sidedgames and full match. Results found minimum differences in total distance covered(p = 0.001; η2 = 0.173, minimum effect), moderate differences in running distance(p = 0.001; η2 = 0.484, moderate effect), acceleration sum (p = 0.001; η2 = 0.261, moderateeffect) and number of sprints (p = 0.001; η2 = 0.478, moderate effect); and strong differenceswere found on sprinting distance (p = 0.001; η2 = 0.716, strong effects).

Pairwise comparisons revealed no effect and minimum effect of difference in totaldistance covered between games. Nevertheless, running distance in full match wasgreater than in 5vs5+GK (d = 2.303, moderate effect), 6vs6+GK (d = 1.719, moderateeffect) and 9vs9+GK (d = 1.084, minimum effect). Greater values of sprinting distancewere found in full match comparing to 5vs5+GK (d = 3.673, strong effect), 6vs6+GK(d = 2.606, moderate effect) and 9vs9+GK (d = 1.903, moderate effect). Moreover, 9vs9+GKhad greater values of sprinting distance than 5vs5+GK (d = 1.681, moderate effect).Considering the number of sprints, full match had greater average than 5vs5+GK(d = 3.091, strong effect) and 6vs6+GK (d = 1.395, moderate effect); and 9vs9+GK hadgreater values than 5vs5+GK (d = 1.294, moderate effect). The load was greater in 5vs5+GK comparing to 6vs6+GK (d = 1.323, moderate effect), 9vs9+GK (d = 1.030, minimumeffect) and full match (d = 1.478, moderate effect).

Discussion

The purpose of the present study was to compare the relative physical demands of sidedgames (medium and large) with those of official matches in professional soccer players.The main findings of our study are (i) medium-to-large (5vs5+GK, 6vs6+GK, and 9vs9+GK) sided games and official matches demand different physical efforts from players interms of running, sprinting, loading, and positional roles; and (ii) large-sided games(9vs9+GK) simulate the official full match more accurately than other sided games interms of sprinting and loading demands.

Differences in total distance covered between full matches and sided games are inagreement with previous reports (Casamichana, Castellano, & Castagna, 2012; Owenet al., 2014). However, the magnitude of these differences was minimal compared to the

RESEARCH IN SPORTS MEDICINE 5

values reported by Owen et al. (2014), who showed that there was a large superiority oflarge-sided games over medium-sided games in terms of the distance covered byplayers (120.4 vs. 108.3 m.min−1). It could be that different skill levels of the players(elite vs. professional) or other possible influential factors (e.g. pitch dimensions, playingrules) might have been responsible for this discrepancy. Indeed, covering a greaterdistance during a match has been highlighted as a distinguishing factor between playerswith different skill levels (Mohr, Krustrup, & Bangsbo, 2003).

The results of present study show that, in terms of running performance (running at aspeed ~ 14–20 km· h−1), the 9vs9+GK sided game format is most similar to an officialmatch (minimal difference), as the medium-sided games (5vs5+GK and 6vs6+GK) requireless running. The importance of running is well recognized by the findings ofCasamichana et al. (2012) which show that higher-level competitions (i.e. internationallevel vs. domestic and national league level) demand greater running (striding) perfor-mance in elite women soccer players. In this sense, our results are consistent with arecent study (Owen et al., 2014) which showed that greater running activity (effect size:1.3) is required for large-sided games than for medium-sided games. Our results, how-ever, are not in agreement with those of Casamichana et al. (2012), who did not reportany significant differences between small-to-large-sided game formats in terms of theamount of running activity required. The size of the playing area used in the study ofCasamichana et al. (2012) for sided games (210 m2) was large and was kept constant,whereas the present study varied the size of the playing area (120–194 m2); theexperiment conducted by Owen et al. (2014) also used playing areas of various sizes(177–184 m2). This difference might be responsible for this disagreement. Indeed, as thesize of the playing area increases, players have less ball possession and need to run morein defensive (to close gaps and apply pressure) and offensive (to open spaces and buildup) moments of the game (Owen, Wong, McKenna, & Dellal, 2011).

Our data showed moderately greater playing loads in the 5vs5+GK game compared tothe other sided games (6vs6+GK and 9vs9+GK) and even compared to full matches. Thisresult supports recent findings (Beenham et al., 2017; Casamichana et al., 2012;Montgomery, Pyne, & Minahan, 2010) that showed higher playing loads in small-sidedgames compared to full matches. Indeed, it seems that playing soccer with a smallernumber of players may result in greater accelerometer-derived loads (Beenham et al.,2017; Casamichana et al., 2012).

Although playing load has been reported to be strongly related to other internal loadmeasures (e.g. Edwards’ training impulse and session rating of perceived exertion [sRPE]) andexternal loadmeasures (e.g. total distance) in soccer players (Casamichana, Castellano, Calleja-Gonzalez, San Román, & Castagna, 2013), this measure is mostly calculated based on accel-erometer data and is sensitive to minor high-intensity actions (e.g. accelerations, decelera-tions, jumping) in all planes (Beenham et al., 2017). When the number of players reduces inskill-based sided games, quick high-intensity actions (e.g. dribbles, shots, and tackles) areincreased, and players havemore involvementwith the ball (Owen et al., 2014). Such increasesin high-intensity actions might be the reason for the moderately greater playing loads in the5vs5+GK format compared to those observed in other sided games (i.e. 6vs6+GK, 9vs9+GK,and 11vs11+GK) in the present study. Nevertheless, these results are not consistent with thereports of Montgomery et al. (2010) in which playing load was greater in a full match than insmall-sided games. However, this discrepancymay be explained by the differences in terms of

6 F. M. CLEMENTE ET AL.

the sport that was studied (basketball vs. soccer) and by the use of different methodologicalapproaches.

Greater sprint performance (in terms of both the number of sprints and sprintingdistance) was observed in the full match compared to medium-sided (5vs5+GK, 6vs6+GK)and large-sided (9vs9+GK) games in the present study (with moderate-to-strong effects)(Figure 1). The 9vs9+GK game also demanded greater sprint performance than the 5vs5+GKformat. These results suggest that medium- and large-sided games cannot completelysimulate a full match. However, as the number of players and pitch size increase, theconditions become more similar to those of a real match scenario in terms of sprintperformance. These results are in agreement with the findings of previous studies that

50

60

70

80

90

100

110

5vs5+GK 6vs6+GK 9vs9+GK Official

Match

Dis

ta

nce

Co

ve

re

d (

m/m

in-1

)

0

2

4

6

8

10

12

14

16

5vs5+GK 6vs6+GK 9vs9+GK Official

Match

Ru

nn

ing

dis

ta

nce

(m

/m

in-1

)

0

1

2

3

4

5

6

5vs5+GK 6vs6+GK 9vs9+GK Official

Match

Sp

rin

tin

g d

ista

nce

(m

/m

in-1

)

0

0.1

0.2

0.3

0.4

0.5

5vs5+GK 6vs6+GK 9vs9+GK Official

Match

Nu

mb

er o

f s

prin

ts (

n/m

in-1

)

3

4

5

6

7

5vs5+GK 6vs6+GK 9vs9+GK Official Match

Acce

lera

tio

n s

um

(A

S)

Figure 1. Mean � 90% confidence intervals of distance covered, running distance, sprintingdistance, number of sprints and acceleration sum intensity in the sided games and official matches.

RESEARCH IN SPORTS MEDICINE 7

have reported a higher number of sprints (Casamichana et al., 2012; Gabbett & Mulvey,2008) and greater sprint distance (Casamichana et al., 2012; Owen et al., 2014) in fullmatches than in small- or medium-sided games.

The importance of sprint performance has been highlighted, as players in interna-tional-level competitions have shown greater sprint performances than players indomestic- and national-level games (Gabbett & Mulvey, 2008). Because players aremore involved with the ball and because space is limited in smaller rectangles (Owenet al., 2011), high-speed thresholds (~ > 20 km· h−1) are not usually reached, and solower sprint performance and greater acceleration/decelerations would be expected onsmaller pitches. Given that performing repeated sprints or high-speed runs in a fullmatch is of paramount importance (Dupont, Akakpo, & Berthoin, 2004; Gabbett &Mulvey, 2008), conducting friendly games (Casamichana et al., 2012) and adding high-speed runs to sided games (Harrison, Kinugasa, Gill, & Kilding, 2015; Vitale et al., 2018)during preparation phases are strongly recommended. Although previous studies (Hill-Haas et al., 2011; Owen et al., 2011) have suggested that playing smaller-sided games asopposed to larger-sided games more accurately replicates the intensity of a full match,these suggestions have been derived mostly from observed heart rate responses andnot from actual physical demands. Furthermore, lactate levels have been reported to behigher in larger-sided games than in smaller-sided games (Dellal et al., 2012), confirmingthe higher specificity of playing in larger rectangles with a greater number of players tosimulate an official match.

Conclusions

This study conducted in professional soccer players revealed that medium- to large-sided games and official matches put different demands on players in terms of running,sprinting, loading, and positional roles; large-sided games (9vs9+GK) simulate official fullmatches more accurately than the other sided-games that were studied (5vs5+GK and6vs6+GK).

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Filipe Manuel Clemente http://orcid.org/0000-0001-9813-2842Hugo Sarmento http://orcid.org/0000-0001-8681-0642Alireza Rabbani http://orcid.org/0000-0002-1500-0447

References

Ade, J. D, Harley, J. A, & Bradley, P. S. (2014). Physiological response, time–motion characteristics,and reproducibility of various speed-endurance drills in elite youth soccer players: small-sidedgames versus generic running. International Journal of Sports Physiology and Performance, 9(3),471–479.

8 F. M. CLEMENTE ET AL.

Aguiar, M., Botelho, G., Lago, C., Maças, V., & Sampaio, J. (2012). A review on the effects of soccersmall-sided games. Journal of Human Kinetics, 33, 103–113.

Aughey, R. J. (2011). Applications of GPS technologies to field sports. International Journal of SportsPhysiology and Performance, 6(3), 295–310.

Beato, M., Jamil, M., & Devereux, G. (2018). Reliability of internal and external load parameters inrecreational football (soccer) for health. Research in Sports Medicine, 26(2), 244–250.

Beenham, M., Barron, D. J., Fry, J., Hurst, H. H., Figueirdo, A., & Atkins, S. (2017). A comparison ofGPS workload demands in match play and small-sided games by the positional role in youthsoccer. Journal of Human Kinetics, 57, 1.

Boyd, L. J., Ball, K., & Aughey, R. J. (2011). The reliability of minimaxx accelerometers for measuringphysical activity in Australian football. International Journal of Sports Physiology andPerformance, 6(3), 311–321.

Casamichana, D., Castellano, J., Calleja-Gonzalez, J., San Román, J., & Castagna, C. (2013).Relationship between indicators of training load in soccer players. Journal of Strength andConditioning Research, 27(2), 369–374.

Casamichana, D., Castellano, J., & Castagna, C. (2012). Comparing the physical demands of friendlymatches and small-sided games in semiprofessional soccer players. Journal of Strength andConditioning Research, 1. doi:10.1519/JSC.0b013e31822a61cf

Castelão, D., Garganta, J., Santos, R., & Teoldo, I. (2014). Comparison of tactical behaviour andperformance of youth soccer players in 3v3 and 5v5 small-sided games. International Journal ofPerformance Analysis in Sport, 14(3), 801–813.

Clemente, F. M., Wong, D. P., Martins, F. M. L., & Mendes, R. S. (2014). Acute effects of the numberof players and scoring method on physiological, physical, and technical performance in small-sided soccer games. Research in Sports Medicine (Print), 22(4), 380–397.

Davids, K., Araújo, D., Correia, V., & Vilar, L. (2013). How small-sided and conditioned gamesenhance acquisition of movement and decision-making skills. Exercise and Sport SciencesReviews, 41(3), 154–161.

Dellal, A., Owen, A., Wong, D. P., Krustrup, P., van Exsel, M., & Mallo, J. (2012). Technical andphysical demands of small vs. large sided games in relation to playing position in elite soccer.Human Movement Science, 31(4), 957–969.

Dupont, G., Akakpo, K., & Berthoin, S. (2004). The effect of in-season, high-intensity interval trainingin soccer players. The Journal of Strength & Conditioning Research, 18(3), 584–589.

Ferguson, C. J. (2009). An effect size primer: A guide for clinicians and researchers. ProfessionalPsychology: Research and Practice, 40(5), 532–538.

Gabbett, T., Jenkins, D., & Abernethy, B. (2009). Game-based training for improving skill andphysical fitness in team sport athletes. International Journal of Sports Science and Coaching, 4(2), 273–283.

Gabbett, T. J. (2006). Skill-based conditioning games as an alternative to traditional conditioningfor rugby league players. The Journal of Strength and Conditioning Research, 20(2), 309.

Gabbett, T. J., & Mulvey, M. J. (2008). Time-motion analysis of small-sided training games andcompetition in elite women soccer players. Journal of Strength and Conditioning Research, 22(2),543–552.

Halouani, J., Chtourou, H., Dellal, A., Chaouachi, A., & Chamari, K. (2014, April). Physiologicalresponses according to rules changes during 3 vs. 3 small-sided games in youth soccer players:Stop-ball vs. small-goals rules. Journal of Sports Sciences, 37–41. doi:10.1080/02640414.2014.899707

Halouani, J., Chtourou, H., Gabbett, T., Chaouachi, A., & Chamari, K. (2014). Small-sided games inteam sports training. Journal of Strength and Conditioning Research, 28(12), 3594–3618.

Harrison, C. B., Kinugasa, T., Gill, N., & Kilding, A. E. (2015). Aerobic fitness for young athletes:Combining game-based and high-intensity interval training. International Journal of SportsMedicine, 94(11), 929–934.

Hill-Haas, S. V., Dawson, B., Impellizzeri, F. M., & Coutts, A. J. (2011). Physiology of small-sidedgames training in football. Sports Medicine, 41(3), 199–220.

RESEARCH IN SPORTS MEDICINE 9

Malone, J. J., Lovell, R., Varley, M. C., & Coutts, A. J. (2017). Unpacking the black box: Applicationsand considerations for using GPS devices in sport. International Journal of Sports Physiology andPerformance, 12(Suppl2), S2–18-S2–26.

McLean, S., Kerhervé, H., Naughton, M., Lovell, G., Gorman, A., & Solomon, C. (2016). The effect ofrecovery duration on technical proficiency during small sided games of football. Sports, 4(3), 39.

Mello, R., Mello, R., Gomes, D., Paz, G. A., Nasser, I., Miranda, H., & Salerno, V. P. (2017). Oxidativestress and antioxidant biomarker responses after a moderate-intensity soccer training session.Research in Sports Medicine, 25(3), 322–332.

Mohr, M., Krustrup, P., & Bangsbo, J. (2003). Match performance of high-standard soccer playerswith special reference to development of fatigue. Journal of Sports Sciences, 21(7), 519–528.

Montgomery, P. G., Pyne, D. B., & Minahan, C. L. (2010). The physical and physiological demands ofbasketball training and competition. International Journal of Sports Physiology and Performance,5, 75–86.

Nikolaidis, P. T., Clemente, F. M., van der Linden, C. M. I., Rosemann, T., & Knechtle, B. (2018).Validity and reliability of 10-Hz global positioning system to assess in-line movement andchange of direction. Frontiers in Physiology, 9. doi:10.3389/fphys.2018.00228

Owen, A. L., Wong, D. P., McKenna, M., & Dellal, A. (2011). Heart rate responses and technicalcomparison between small- vs. large-sided games in elite professional soccer. Journal ofStrength and Conditioning Research, 25(8), 2104–2110.

Owen, A. L., Wong, D. P., Paul, D., & Dellal, A. (2014). Physical and technical comparisons betweenvarious-sided games within professional soccer. International Journal of Sports Medicine, 35(4),286–292.

Rampinini, E., Impellizzeri, F. M., Castagna, C., Abt, G., Chamari, K., Sassi, A., & Marcora, S. M. (2007).Factors influencing physiological responses to small-sided soccer games. Journal of SportsSciences, 25(6), 659–666.

Ric, A., Hristovski, R., Gonçalves, B., Torres, L., Sampaio, J., & Torrents, C. (2016). Timescales forexploratory tactical behaviour in football small-sided games. Journal of Sports Sciences, 34(18),1723–1730.

Sarmento, H., Clemente, F. M., Araújo, D., Davids, K., McRobert, A., & Figueiredo, A. (2017). Whatperformance analysts need to know about research trends in association football (2012–2016):A systematic review. Sports Medicine, Ahead-Of-P. doi:10.1007/s40279-017-0836-6

Scott, M. T. U., Scott, T. J., & Kelly, V. G. (2016). The validity and reliability of global positioningsystems in team sport. Journal of Strength and Conditioning Research, 30(5), 1470–1490.

Selmi, O., Gonçalves, B., Ouergui, I., Sampaio, J., & Bouassida, A. (2018). Influence of well-beingvariables and recovery state in physical enjoyment of professional soccer players during small-sided games. Research in Sports Medicine, 26(2), 199–210.

Selmi, O., Marzouki, H., Ouergui, I., BenKhalifa, W., & Bouassida, A. (2018). Influence of intensetraining cycle and psychometric status on technical and physiological aspects performed duringthe small-sided games in soccer players. Research in Sports Medicine, Ahead-of-P, 1–12.doi:10.1080/15438627.2018.1492398

Silva, B., Garganta, J., Santos, R., & Teoldo, I. (2014). Comparing tactical behaviour of soccer playersin 3 vs. 3 and 6 vs. 6 small-sided games. Journal of Human Kinetics, 41, 1.

Silva, P., Vilar, L., Davids, K., Araújo, D., & Garganta, J. (2016). Sports teams as complex adaptivesystems: Manipulating player numbers shapes behaviours during football small-sided games.SpringerPlus, 5(1), 191.

Stevens, T. G. A., De Ruiter, C. J., Beek, P. J., & Savelsbergh, G. J. P. (2016). Validity and reliability of6-a-side small-sided game locomotor performance in assessing physical fitness in footballplayers. Journal of Sports Sciences, 34(6), 527–534.

Vitale, J. A., Povìa, V., Vitale, N. D., Bassani, T., Lombardi, G., Giacomelli, L., . . . La Torre, A. (2018). Theeffect of two different speed endurance training protocols on a multiple shuttle run perfor-mance in young elite male soccer players. Research in Sports Medicine, Ahead-of-P, 1–14.doi:10.1080/15438627.2018.1492402

10 F. M. CLEMENTE ET AL.