Full Terms & Conditions of access and use can be found athttp://tandfonline.com/action/journalInformation?journalCode=rjsp20

Download by: [Ruhr-Universitat Bochum Universitaetsbibliothek] Date: 09 June 2016, At: 01:12

Journal of Sports Sciences

ISSN: 0264-0414 (Print) 1466-447X (Online) Journal homepage: http://tandfonline.com/loi/rjsp20

Technical performance and match-to-matchvariation in elite football teams

Hongyou Liu, Miguel-Angel Gómez, Bruno Gonçalves & Jaime Sampaio

To cite this article: Hongyou Liu, Miguel-Angel Gómez, Bruno Gonçalves & Jaime Sampaio(2016) Technical performance and match-to-match variation in elite football teams, Journal ofSports Sciences, 34:6, 509-518, DOI: 10.1080/02640414.2015.1117121

To link to this article: http://dx.doi.org/10.1080/02640414.2015.1117121

Published online: 27 Nov 2015.

Submit your article to this journal

Article views: 1128

View related articles

View Crossmark data

Citing articles: 2 View citing articles

Technical performance and match-to-match variation in elite football teamsHongyou Liu a,b,c, Miguel-Angel Gómez b, Bruno Gonçalvesc and Jaime Sampaio c

aSchool of Physical Education and Sports Science, South China Normal University, Guangzhou, China; bFaculty of Physical Activity and SportSciences, Technical University of Madrid, Madrid, Spain; cCreativeLab, Research Centre for Sports Sciences, Health and Human Development,University of Trás-os-Montes e Alto Douro, Vila Real, Portugal

ABSTRACTRecent research suggests that match-to-match variation adds important information to performancedescriptors in team sports, as it helps measure how players fine-tune their tactical behaviours andtechnical actions to the extreme dynamical environments. The current study aims to identify thedifferences in technical performance of players from strong and weak teams and to explore match-to-match variation of players’ technical match performance. Performance data of all the 380 matches ofseason 2012–2013 in the Spanish First Division Professional Football League were analysed. Twenty-oneperformance-related match actions and events were chosen as variables in the analyses. Players’technical performance profiles were established by unifying count values of each action or event ofeach player per match into the same scale. Means of these count values of players from Top3 andBottom3 teams were compared and plotted into radar charts. Coefficient of variation of each matchaction or event within a player was calculated to represent his match-to-match variation of technicalperformance. Differences in the variation of technical performances of players across different matchcontexts (team and opposition strength, match outcome and match location) were compared. All thecomparisons were achieved by the magnitude-based inferences. Results showed that technical perfor-mances differed between players of strong and weak teams from different perspectives across differentfield positions. Furthermore, the variation of the players’ technical performance is affected by the matchcontext, with effects from team and opposition strength greater than effects from match location andmatch outcome.

ARTICLE HISTORYAccepted 3 November 2015

KEYWORDSNotational analysis; matchanalysis; performanceindicators; situationalvariables; soccer

Introduction

Performance in football is the result of dynamic interactions ofphysical, technical and tactical actions and movements fromall competing players (Bangsbo, 1994; Bradley et al., 2011). Theavailable research on football players focuses primarily onanalysing their physical performance or a combination ofphysical and few technical parameters, whilst studies focusingon the technical and tactical performance are less frequent(Rampinini, Impellizzeri, Castagna, Coutts, & Wisloff, 2009;Russell, Rees, & Kingsley, 2013). However, the technical actionscan be better predictors of success in football compared topure physical parameters (Bush, Barnes, Archer, Hogg, &Bradley, 2015a; Castellano, Casamichana, & Lago, 2012; Lago-Peñas, Lago-Ballesteros, & Rey, 2011; Lago-Peñas, Lago-Ballesteros, Dellal, & Gómez, 2010; Rampinini et al., 2009;Russell et al., 2013). Thus, the development of technical per-formance profiles can be an important task to reveal newtrends in football performance and, ultimately, contribute toimprove task representativeness in practice sessions and toimprove the process of selecting the most appropriate playersto each match scenario.

Although there are already several methods available aim-ing to develop meaningful profiles (Butterworth, O’Donoghue,& Cropley, 2013; Eugster, 2012; Hughes, Evans, & Wells, 2001;James, Mellalieu, & Jones, 2005; Liu, Yi, Gimenez, Gómez, &

Lago-Peñas, 2015c; O’Donoghue, 2005, 2013), performanceprofiling is still required to be further explored. The mainsports profiling techniques were discussed by O’Donoghue(2013). The basic principle of these techniques is to combinea set of valid and reliable performance-related variables withina given sport to properly describe a certain performance/performer by using normative match data (O’Donoghue,2013). In fact, combining different types of variables alto-gether makes the profiling technique an appropriate proce-dure to evaluate the technical performance of football players(Eugster, 2012; Liu et al., 2015c). However, the current profilingtechnique is more focused on representing the typical perfor-mance and the spread of the performance of a single perfor-mer/player (O’Donoghue, 2013). The usefulness of profiles islikely to be improved when they are extended to present andcompare the technical performance of various football playersconsidering their field position and their team quality (Bushet al., 2015a; Rampinini et al., 2009).

Another important issue that has been given limited atten-tion is the concept and interpretation of the match-to-matchvariation of the football player’s technical performance(Kempton, Sullivan, Bilsborough, Cordy, & Coutts, 2015).Actually, football technical actions are motor skills that varydue to both internal and external influences (Kempton et al.,2015). From a more traditional point of view, human

CONTACT Hongyou Liu szu.youyou@hotmail.com

JOURNAL OF SPORTS SCIENCES, 2016VOL. 34, NO. 6, 509–518http://dx.doi.org/10.1080/02640414.2015.1117121

© 2015 Taylor & Francis

Dow

nloa

ded

by [

Ruh

r-U

nive

rsita

t Boc

hum

Uni

vers

itaet

sbib

lioth

ek]

at 0

1:12

09

June

201

6

movement variation has always been associated with errors intechnical executions; thus, it was often treated as noise fromthe motor system that should be reduced or eliminated (Kudo& Ohtsuki, 2008; Seifert, Button, & Davids, 2013; Stergiou, Yu, &Kyvelidou, 2013). However, more contemporary approachesare suggesting that variation can also be functional and ben-eficial, because it is very likely the result of fine-tuned adjust-ments to very dynamic performing environments (Kudo &Ohtsuki, 2008; Seifert et al., 2013; Stergiou et al., 2013).Therefore, this functional variation might be focused on howthe subject’s behaviour is flexible and adaptive, rather thanstereotyped and predictable, when facing complex environ-mental constrains in chaotic systems (Seifert et al., 2013).

The football games, in particular, can be considered ascomplex, self-organised, unstable, unpredictable and highlydynamic systems in which players from competing teams tryto keep stability of their own attacking, organising anddefending balance and to destabilise the balance of the oppo-sition (Davids, Araujo, Correia, & Vilar, 2013; Garganta, 2009;Vilar, Araujo, Davids, & Button, 2012). Several external matchconditions of this sport, such as match location, standard ofcompetition, team and opposition strength, match outcome,are suggested as very important variables that influence indi-vidual and team behaviours (Eccles, Ward, & Woodman, 2009;Gómez, Lago-Peñas, & Pollard, 2013; Mackenzie & Cushion,2013; Rampinini et al., 2009; Sarmento et al., 2014). In fact,players have to adapt physiologically and psychologically tothese different match scenarios (Eccles et al., 2009), leading todifferent changes in variation of actions and patterns (Vilaret al., 2012). For example, when playing at home, players aremore familiar with the match facility and environment (Pollard,2008) and may be in more positive psycho-physiological states(Poulter, 2009); thus, the variation of their home performancemay be expected to be smaller than their away performance.Accordingly, due to a more active psychological state (Kinrade,Jackson, & Ashford, 2015; Rampinini, Coutts, Castagna, Sassi, &Impellizzeri, 2007) and a presumably strategic intention ofrefraining risk taking (Almeida, Ferreira, & Volossovitch,2014), the variation of performance is likely to increase whenplayers facing a strong opposition than when facing a weakone. It also seems important to understand if the variation ofoffensive and defensive variables is affected differently by thematch contexts. Theoretically, it might be expected thatdefensive variables are less dependent on open and complexskills, therefore, can be less affected by environmental con-straints and become more stable. In sum, information pro-vided by analysing the variation of players’ performancesincorporating possible influences of different environmentalconstrains (match contextual variables) can be a very impor-tant contribution to a better understanding of football matchperformance.

Therefore, the aims of the current study were (1) to identifymatch performance profiles of players; (2) to compare thebetween-player differences in performance of strong andweak teams according to players’ specific field positions; (3)to explore the within-player match-to-match variation of tech-nical match performance taking consideration of four contex-tual variables (i.e. team and opposition strength, matchoutcome and match location).

Method

Sample and subject

Performance data of all the 380 matches of season 2012–2013in the Spanish First Division Professional Football League (LaLiga BBVA) were collected. Match data of goalkeepers wereexcluded because of the specificity of this position.Meanwhile, only the outfield players who started and playedat least one entire match were selected, which finally limitedthe subjects to 409 players (age: 27.0 ± 3.8 years, height:179.7 ± 9.7 cm, weight: 74.2 ± 5.8 kg, retrieved from theofficial website of Spanish Professional Football League:www.lfp.es, on the date 22 June 2013) who generated 5288full match observations. Uncompleted match observationsgenerated by players who were substituted out or were sentoff by red cards and by players who were substituted in wereexcluded. Players were divided into the following five differentcategories of positions (Bush et al., 2015a, 2015b): fullback(N = 79 players, n = 1289 full match observations), centraldefender (N = 74, n = 1393), wide midfielder (N = 77, n = 676),central midfielder (N = 120, n = 1398) and forward (N = 59,n = 532).

Data source and reliability

Data used in the study were made available by OPTASportsdata Spain Company (Madrid). Detailed information onthe process by which OPTA Sportsdata are collected, pro-cessed and output can be found elsewhere (Liu, Hopkins,Gómez, & Molinuevo, 2013). The tracking system (OPTAClient System) was tested to have an acceptable inter-operatorreliability (intra-class correlation coefficients ranged from 0.88to 1.00 and standardised typical error varied from 0.00 to 0.37when using the system to code match actions of individualoutfield players) (Liu et al., 2013). The Company maintainedthe anonymity of players and teams following European DataProtection Law. Institutional ethics committee approval wasobtained from the local university.

Variables

Based on the data accessibility, an initial selection of variableswas conducted by consulting two UEFA licensed professionalcoaches and a professional performance analyst of football.After that, a review of the available related literature(Castellano et al., 2012; Lago-Peñas & Lago-Ballesteros, 2011;Lago-Peñas et al., 2010, 2011; Liu, Gómez, Lago-Peñas, &Sampaio, 2015a; Liu, Hopkins, & Gomez, 2015b; Liu et al.,2013) was achieved to help chose the variables. Dependingon the selection and review, 21 performance related matchactions and events were finally chosen as variables in theanalyses. Operational definitions of these actions and eventsare as follows (Liu et al., 2013, 2015a, 2015b):

● Assist: the final pass or cross leading to the recipient ofthe ball scoring a goal.

● Shot: an attempt to score a goal, made with any (legal)part of the body, either on or off target.

510 H. LIU ET AL.

Dow

nloa

ded

by [

Ruh

r-U

nive

rsita

t Boc

hum

Uni

vers

itaet

sbib

lioth

ek]

at 0

1:12

09

June

201

6

● Shot on Target: an attempt to goal which required inter-vention to stop it going in or resulted in a goal/shotwhich would go in without being diverted.

● Ball Touch: a sum of count values of all actions andevents where a player touches the ball.

● Pass: an intentional played ball from a player to histeammate.

● Pass Accuracy (%): a ratio calculated from successfulpasses divided by all passes.

● Through Ball: a pass made by a player that split the lastline of defence and plays a teammate through on goal.

● Key Pass: the final pass or cross leading to the recipientof the ball having an attempt at goal without scoring.

● Cross: any ball sent by a player into the oppositionteam’s area from a wide position.

● Successful Dribble: a dribble is an attempt by a player tobeat an opponent in possession of the ball. A successfuldribble means the player beats the defender whileretaining possession; unsuccessful ones are where thedribbler is tackled. OPTA also log attempted dribbleswhere the player overruns the ball.

● Foul Drawn: where a player is fouled by an opponent.● Aerial Duel Won: two players competing for a ball in the

air, for it to be an aerial duel both players must jump andchallenge each other in the air and have both feet offthe ground. The player who wins the duel gets the AerialDuel Won, and the player who does not gets an AerialDuel Lost.

● Dispossessed: when a player is tackled without attempt-ing to dribble past his opponent.

● Turnover: loss of possession to opponent players due toa mistake/poor control.

● Offside: a player being caught in an offside positionresulting in a free kick to the opposing team.

● Tackle: act of gaining possession from an oppositionplayer who is in possession of the ball.

● Interception: a player intercepts a pass with some move-ment or reading of the play.

● Clearance: attempt made by a player to get the ball outof the danger zone, when there is pressure (from oppo-nents) on him to clear the ball.

● Shot Block: a player blocks a goal attempt headingroughly on target toward goal, where there are otherdefending players or a goalkeeper behind the blocker.

● Foul Committed: any infringement committed by a playerthat is penalised as foul play by the referee.

● Yellow Card: where a player is booked by the referee dueto illegal actions.

Meanwhile, the following match contextual variables werealso analysed:

(1) Strength of the team and opposition. The three top-ranked teams in the final league classification weredefined as strong teams/oppositions, as they qualifieddirectly to the UEFA Champions League of the nextseason; and the three bottom ranked teams weredefined as weak teams/oppositions, as they relegateddirectly to the second league of the next season.

(2) Match outcome (win, loss and draw).(3) Match location (home and away).

Statistical analysis

The first phase of the study was to set up the performanceprofiles of all the players using the profiling technique. Themain profiling techniques include (O’Donoghue, 2013) (1)using median and 95% confidence intervals (James et al.,2005) to represent and compare performances of differentperformers; and (2) using median and quartiles (O’Donoghue,2005) to represent typical performance and its spread of asingle performer. However, based on a large sample size, thecurrent study can establish profiles that allow to combine theadvantages of these two techniques (Liu et al., 2015c). Countvalues of the 21 performance-related match actions andevents of all players were transformed into standardisedscore (Z-Score, Z) and were unified into the same scale usingthe formula “T = 20Z + 50” (Barriopedro & Muniesa, 2012).Performance profiles of players from strong teams and weakteams were compared and plotted into radar charts.Comparisons were conducted in overall players and in playersof different positions (i.e. fullback, central defender, wide mid-fielder, central midfielder and forward), respectively.

The second phase was to explore the variation of matchperformance of the players. The within-player match-to-matchvariation was expressed by the coefficient of variation (CV) ofeach match action or event (Bush et al., 2015b; Kempton et al.,2015; Spencer, Losnegard, Hallen, & Hopkins, 2014). The differ-ences of variation of match performance between players, (a)from strong teams (Top3) and weak teams (Bottom3), (b)when playing against strong oppositions (Top3) and weakoppositions (Bottom3), (c) in games won and games lost/drawn, (d) when playing away and playing at home, werecompared separately. In order to calculate the within-playerCV (standard deviation/mean), only the players who played atleast two entire matches (ranged from 2 to 37 match observa-tions) were selected in each match context. When the mean ofthe count of an action or event was 0 (e.g. 0 yellow card in 10matches), the CV of this single action or event was treated as amissing sample.

The transform and the unification of the Z-Score wereperformed in the data package of IBM SPSS Statistics forWindows, Version 20.0 (Armonk, NY: IBM Corp.). The radarcharts were drawn in the Microsoft Excel 2007 (Redmond,Washington: Microsoft). Comparisons of performance pro-file and variation of match performance were achieved bythe spreadsheet developed by Hopkins (2007). Nonclinicalmagnitude-based inferences were employed and wereevaluated by using the smallest worthwhile difference.The smallest worthwhile difference was calculated by 0.2times the standardisation estimated from between-subjectstandard deviation. For the comparison of players’ varia-tion of match performance, there was only one indepen-dent inference (CV), so the 90% confidence intervals wereused to make the inferences. While for the comparison onthe performance profile of players, there were 21 indepen-dent inferences (21 performance-related match actions and

JOURNAL OF SPORTS SCIENCES 511

Dow

nloa

ded

by [

Ruh

r-U

nive

rsita

t Boc

hum

Uni

vers

itaet

sbib

lioth

ek]

at 0

1:12

09

June

201

6

events), in order to reduce the inflation of Type 1 error,only differences clear with 99% confidence intervals wereevaluated. Magnitudes of clear differences were assessedas follows: <0.20, trivial; 0.20–0.60, small; 0.61–1.20, mod-erate; 1.21–2.0, large; >2.0, very large (Batterham &Hopkins, 2006). Differences were deemed clear if the con-fidence intervals for the difference in the means did notinclude substantial positive and negative values (±0.2*stan-dardisation) simultaneously (Hopkins, Marshall, Batterham,& Hanin, 2009).

Results

The comparison on the difference in the mean counts ofeach match action or event between players from strong(Top3) and weak (Bottom3) teams is presented in theperformance profiles (Table 1 and Figure 1). All the differ-ences were clear when comparing all players together;players from strong teams achieved more assists, shotson target, ball touches, passes, through balls, successfuldribbles and higher pass accuracy, but fewer clearances,while differences in other actions and events were alltrivial (Figure 1a). Comparisons of players from differentpositions showed different results. Full backs from strongteams made more assists, shots, shots on target, balltouches, passes, through balls, key passes, successful drib-bles, fouls drawn, dispossesses, turnovers, offsides andbetter pass accuracy, but fewer clearances and yellowcards than their counterparts from weak teams(Figure 1b). Central defenders from strong teams mademore shots, ball touches, passes, successful dribbles, turn-overs, higher pass accuracy, but fewer interceptions andclearances comparing to their peers from weak teams(Figure 1c). Wide midfielders from Top3 teams achievedmore assists, shots, shots on target, ball touches, passes,through balls, aerial duels won and yellow cards, but fewercrosses than their counterparts from Bottom3 teams

(Figure 1d). Top3 teams’ central midfielders made moreassists, ball touches, passes, through balls, successful drib-bles, higher pass accuracy, but fewer shots than Bottom3teams’ counterparts (Figure 1e). Forwards from strongteams achieved more assists, shots on target, ball touches,passes, through balls, successful dribbles and better passaccuracy, but fewer aerial duels won, offsides, clearancesand fouls committed comparing to forwards from weakteams (Figure 1f).

The comparison on the difference of within-player match-to-match variation of technical performance under differentmatch contexts is presented in Table 2 and Figure 2. Shot, shoton target and assist were the three most unstable technicalmatch actions that showed substantial differences in variationin all the four match contexts (Figure 2). Players from weakteams showed higher variation in most of the attacking-related match actions (assist, shot, shot on target, ball touch,pass, pass accuracy, through ball, key pass, successful dribble,dispossess and turnover) and lower variation in defending-related match actions and events (interception, clearanceand yellow card) than their counterparts from strong teams(Figure 2a). Meanwhile, the variation of most attacking-relatedmatch actions (assist, shot, shot on target, ball touch, throughball, key pass, foul drawn, aerial duel won and offside) wassmaller when players played against weak oppositions thanwhen playing against strong oppositions, while differences inthe variation of most of the defending-related match actionswere trivial (Figure 2b). Most of the differences in the variationof match actions between players in games won and gameslost/drawn were trivial. There was a trend showing that thevariation in technical performance of players in victory gameswas smaller than players in lost or drawn games (Figure 2c).Similarly, there was a trend showing that players’ variation oftechnical performance was smaller when playing at homethan when playing away, even though many of the observeddifferences between the home and away variation were trivial(see Figure 2d).

Table 1. Descriptive statistics of match performance profiles of players from Top3 and Bottom3 teams (results are counts, except for Pass Accuracy).

All Players(n = 1583) Full Back (n = 382)

Central Defender(n = 415)

Wide Midfielder(n = 183)

Central Midfielder(n = 402) Forward (n = 201)

Variable Top3 Bottom3 Top3 Bottom3 Top3 Bottom3 Top3 Bottom3 Top3 Bottom3 Top3 Bottom3

Assist 0.2 ± 0.4 0.1 ± 0.2 0.1 ± 0.3 0.0 ± 0.2 0.0 ± 0.2 0.0 ± 0.1 0.3 ± 0.5 0.1 ± 0.3 0.2 ± 0.4 0.1 ± 0.3 0.3 ± 0.6 0.1 ± 0.4Shot 1.4 ± 1.9 1.1 ± 1.4 0.6 ± 0.9 0.4 ± 0.7 0.6 ± 0.8 0.4 ± 0.7 4.0 ± 3.2 2.0 ± 1.4 0.9 ± 1.1 1.2 ± 1.3 3.3 ± 2.0 3.1 ± 2.0Shot on target 0.5 ± 1.0 0.4 ± 0.7 0.2 ± 0.4 0.1 ± 0.3 0.2 ± 0.4 0.2 ± 0.4 1.5 ± 1.6 0.7 ± 0.8 0.3 ± 0.6 0.3 ± 0.6 1.5 ± 1.3 1.1 ± 1.0Ball touch 71 ± 27 53 ± 16 79 ± 24 59 ± 15 58 ± 21 48 ± 15 63 ± 18 53 ± 13 86 ± 30 59 ± 18 61 ± 24 41 ± 10Passes 52 ± 26 34 ± 15 50 ± 22 30 ± 12 44 ± 21 32 ± 14 41 ± 15 31 ± 10 70 ± 30 45 ± 18 44 ± 23 26 ± 8Pass Accuracy (%) 82 ± 11 75 ± 12 83 ± 10 73 ± 12 81 ± 14 75 ± 13 78 ± 9 78 ± 10 85 ± 9 77 ± 11 82 ± 9 70 ± 11Through ball 0.5 ± 1.0 0.1 ± 0.4 0.2 ± 0.5 0.0 ± 0.1 0.1 ± 0.3 0.1 ± 0.3 0.9 ± 1.1 0.1 ± 0.4 0.8 ± 1.2 0.2 ± 0.5 0.9 ± 1.2 0.2 ± 0.5Key pass 1.0 ± 1.3 0.9 ± 1.2 0.9 ± 1.0 0.6 ± 0.8 0.2 ± 0.5 0.2 ± 0.4 2.1 ± 2.0 1.6 ± 1.4 1.3 ± 1.3 1.2 ± 1.4 1.4 ± 1.1 1.4 ± 1.6Crosses 1.8 ± 2.6 2.1 ± 3.2 2.9 ± 2.7 2.5 ± 2.3 0.1 ± 0.4 0.1 ± 0.3 4.5 ± 4.4 6.3 ± 4.4 1.8 ± 2.4 1.6 ± 2.9 1.3 ± 1.5 1.7 ± 2.5Successful dribble 0.9 ± 1.4 0.6 ± 0.9 0.8 ± 1.1 0.4 ± 0.8 0.3 ± 0.5 0.1 ± 0.4 1.5 ± 1.6 1.2 ± 1.2 0.9 ± 1.2 0.6 ± 0.9 2.0 ± 2.0 1.0 ± 1.2Foul drawn 1.2 ± 1.3 1.2 ± 1.3 1.2 ± 1.1 0.8 ± 1.0 0.6 ± 0.9 0.5 ± 0.8 1.9 ± 1.6 1.9 ± 1.5 1.3 ± 1.2 1.5 ± 1.3 1.9 ± 1.7 2.1 ± 1.4Aerial duel Won 1.5 ± 1.7 1.7 ± 1.8 1.2 ± 1.4 1.3 ± 1.2 2.2 ± 1.9 2.0 ± 1.6 1.2 ± 1.7 0.7 ± 1.0 1.4 ± 1.6 1.5 ± 1.8 0.9 ± 1.3 3.5 ± 2.7Dispossessed 1.1 ± 1.3 0.9 ± 1.2 1.0 ± 1.0 0.5 ± 0.7 0.2 ± 0.5 0.2 ± 0.4 1.6 ± 1.4 1.6 ± 1.3 1.0 ± 1.2 1.1 ± 1.1 2.3 ± 1.7 2.2 ± 1.4Turnover 0.9 ± 1.2 0.7 ± 1.0 0.8 ± 0.9 0.6 ± 0.7 0.3 ± 0.5 0.2 ± 0.5 1.8 ± 1.5 1.4 ± 1.2 0.9 ± 1.0 0.7 ± 0.8 1.9 ± 1.6 1.7 ± 1.5Offside 0.3 ± 0.7 0.2 ± 0.6 0.1 ± 0.4 0.1 ± 0.3 0.1 ± 0.3 0.0 ± 0.2 0.6 ± 1.0 0.4 ± 0.7 0.1 ± 0.4 0.1 ± 0.3 0.9 ± 1.1 1.2 ± 1.2Tackle 2.1 ± 1.8 2.1 ± 1.7 2.7 ± 2.0 2.8 ± 1.8 2.0 ± 1.5 2.0 ± 1.6 1.7 ± 1.7 1.3 ± 1.2 2.7 ± 2.1 2.5 ± 1.9 0.9 ± 1.0 0.6 ± 0.9Interception 1.7 ± 1.6 1.7 ± 1.7 2.2 ± 1.7 2.1 ± 1.7 2.1 ± 1.5 2.5 ± 1.9 0.6 ± 0.8 0.8 ± 1.0 1.9 ± 1.7 1.7 ± 1.5 0.5 ± 0.8 0.3 ± 0.6Clearance 2.7 ± 3.3 3.7 ± 3.9 2.5 ± 2.0 4.0 ± 2.6 6.3 ± 3.7 8.0 ± 4.2 0.6 ± 1.0 0.7 ± 0.9 1.5 ± 1.8 1.5 ± 1.7 0.3 ± 0.8 0.9 ± 1.5Shot block 0.3 ± 0.6 0.3 ± 0.6 0.3 ± 0.5 0.3 ± 0.5 0.5 ± 0.8 0.6 ± 0.9 0.1 ± 0.2 0.1 ± 0.3 0.3 ± 0.6 0.3 ± 0.6 0.1 ± 0.3 0.1 ± 0.3Foul committed 1.3 ± 1.3 1.3 ± 1.3 1.3 ± 1.4 1.2 ± 1.1 1.0 ± 1.1 1.2 ± 1.1 1.1 ± 1.2 1.0 ± 1.0 1.6 ± 1.4 1.7 ± 1.6 1.2 ± 1.3 1.7 ± 1.5Yellow card 0.2 ± 0.4 0.3 ± 0.4 0.1 ± 0.4 0.3 ± 0.5 0.2 ± 0.4 0.3 ± 0.4 0.2 ± 0.4 0.1 ± 0.3 0.3 ± 0.4 0.3 ± 0.5 0.2 ± 0.4 0.2 ± 0.4

512 H. LIU ET AL.

Dow

nloa

ded

by [

Ruh

r-U

nive

rsita

t Boc

hum

Uni

vers

itaet

sbib

lioth

ek]

at 0

1:12

09

June

201

6

Discussion

Technical match performance profiles of outfield players fromSpanish First Division Professional Football League have beenset up by using a profiling technique, and the within-playermatch-to-match variation of technical performance wasexplored by considering effects of team and oppositionstrength, match outcome and match location.

Results of our profiles showed that players from Top3teams achieved more assists, shots on target, ball touches,passes, through balls, successful dribbles and higher pass

accuracy, but fewer clearances than those from Bottom3teams. Similar findings were reported in a previous study onItalian Serie A league (Rampinini et al., 2009), which showedthat “involvement with the ball, short pass, successful shortpass, tackle, dribbling, shot and shot on target” (Rampininiet al., 2009, p. 232) were the technical parameters that differ-entiated top and bottom teams. Within the established per-formance profiles, comparisons on the technical matchperformances can be achieved and displayed in a morestraightforward and illustrative way (Liu et al., 2015c). Fromthe radar charts plotted by unified standardised score, the

Figure 1. Comparison of the performance profiles of players from Top3 and Bottom3 teams.Letters in parentheses denote the magnitude: t = trivial; s = small; m = moderate; l = large. Asterisks indicate the likelihood for the magnitude of the true differencein means as follows: *possible; **likely; ***very likely; ****most likely. SoT = shot on target; BT = ball touch; PA = pass accuracy (%); TB = through balls; KP = keypass; SD = successful dribble; FD = foul drawn; ADW = aerial duel won; SB = shot block; FC = foul committed; YC = yellow card. (a) All Players, (b) Full Back, (c)Central Defender, (d) Wide Midfielder, (e) Central Midfielder, (f) Forward.

JOURNAL OF SPORTS SCIENCES 513

Dow

nloa

ded

by [

Ruh

r-U

nive

rsita

t Boc

hum

Uni

vers

itaet

sbib

lioth

ek]

at 0

1:12

09

June

201

6

Table2.

Descriptivestatisticsof

match-to-match

variatio

n(CV)

ofplayers’performance

(arbitraryun

it).

Overall(n)

Top3

(n)

Bottom

3(n)

Strong

oppo

sitio

n(n)

Weakop

positio

n(n)

Win

(n)

Loss/Draw

(n)

Hom

e(n)

Away

(n)

Assist

2.9±1.2(184)

2.4±1.1(40)

2.9±1.0(14)

1.8±0.5(32)

1.6±0.6(52)

2.1±0.9(129)

3.0±0.8(102)

2.3±0.9(129)

2.6±0.8(114)

Shot

1.2±0.7(352)

1.1±0.8(54)

1.4±0.7(33)

1.2±0.6(161)

1.1±0.6(172)

1.1±0.7(266)

1.2±0.6(316)

1.1±0.6(312)

1.3±0.7(295)

Shot

ontarget

2.0±1.1(297)

1.7±1.1(46)

2.3±1.3(23)

1.5±0.6(95)

1.3±0.6(111)

1.6±0.9(216)

2.0±1.0(241)

1.7±0.9(249)

1.9±0.9(216)

Balltouch

0.2±0.1(364)

0.2±0.1(56)

0.2±0.1(33)

0.2±0.1(207)

0.2±0.1(207)

0.2±0.1(299)

0.2±0.1(335)

0.2±0.1(325)

0.2±0.1(325)

Passes

0.3±0.1(364)

0.3±0.1(56)

0.3±0.1(33)

0.3±0.2(207)

0.3±0.2(207)

0.3±0.1(299)

0.3±0.1(335)

0.3±0.1(325)

0.3±0.1(325)

Pass

accuracy

0.1±0.1(364)

0.1±0.1(56)

0.1±0.0(33)

0.1±0.1(207)

0.1±0.1(207)

0.1±0.1(299)

0.1±0.1(335)

0.1±0.1(325)

0.1±0.1(325)

Throug

hball

2.4±1.2(232)

2.1±1.3(46)

2.5±0.8(15)

1.6±0.6(67)

1.4±0.5(77)

1.8±0.8(153)

2.2±1.0(186)

1.9±0.9(170)

2.1±0.9(171)

Keypass

1.5±0.9(352)

1.3±0.8(56)

1.5±0.8(31)

1.3±0.6(137)

1.1±0.6(155)

1.3±0.7(254)

1.4±0.8(306)

1.3±0.7(299)

1.5±0.8(283)

Crosses

1.4±1.1(323)

1.4±1.2(52)

1.5±1.2(29)

0.9±0.6(137)

0.9±0.6(151)

1.1±0.8(232)

1.3±0.9(286)

1.2±0.9(279)

1.2±0.8(261)

Successful

dribble

1.7±0.9(327)

1.4±0.7(54)

1.7±0.9(28)

1.3±0.6(129)

1.3±0.6(138)

1.4±0.7(237)

1.6±0.8(280)

1.5±0.7(267)

1.5±0.8(265)

Foul

draw

n1.0±0.5(357)

1.1±0.5(55)

1.1±0.6(31)

1.1±0.6(174)

0.9±0.6(189)

1.0±0.6(285)

1.0±0.5(322)

1.1±0.5(309)

1.0±0.6(315)

Aeriald

uelW

on1.0±0.6(355)

1.1±0.7(53)

1.0±0.4(32)

1.0±0.6(181)

0.8±0.5(191)

1.0±0.5(285)

1.0±0.5(323)

0.9±0.5(311)

1.0±0.5(310)

Dispo

ssessed

1.3±0.9(340)

1.2±0.8(53)

1.6±1.1(32)

1.1±0.6(155)

1.1±0.6(155)

1.2±0.7(247)

1.3±0.9(305)

1.3±0.8(290)

1.3±0.8(285)

Turnover

1.3±0.7(356)

1.2±0.5(55)

1.5±0.8(33)

1.1±0.6(161)

1.1±0.6(163)

1.2±0.7(269)

1.3±0.7(317)

1.3±0.7(305)

1.3±0.7(304)

Offside

2.6±1.4(223)

2.5±1.3(41)

3.2±1.7(17)

1.6±0.5(58)

1.3±0.7(68)

1.9±1.0(131)

2.2±1.1(168)

2.1±1.0(169)

2.0±0.9(152)

Tackle

0.8±0.3(362)

0.8±0.3(55)

0.8±0.3(33)

0.8±0.4(202)

0.7±0.4(205)

0.7±0.4(297)

0.8±0.4(332)

0.8±0.3(323)

0.8±0.4(323)

Interceptio

n0.9±0.5(352)

1.0±0.5(56)

0.9±0.4(32)

0.8±0.5(197)

0.8±0.5(189)

0.9±0.5(289)

0.9±0.5(322)

0.9±0.5(314)

0.9±0.5(311)

Clearance

1.0±0.7(344)

1.2±0.8(53)

0.8±0.4(32)

0.8±0.5(189)

0.7±0.5(187)

0.8±0.6(277)

1.0±0.6(316)

0.9±0.6(300)

0.9±0.6(303)

Shot

block

2.0±1.0(270)

2.2±1.0(40)

2.1±1.0(28)

1.4±0.5(125)

1.4±0.6(114)

1.8±0.8(200)

1.9±0.9(240)

1.9±0.8(222)

1.8±0.8(226)

Foul

committed

0.9±0.4(362)

1.0±0.6(55)

1.1±0.4(32)

0.9±0.5(202)

0.9±0.5(202)

0.9±0.4(293)

0.9±0.4(329)

0.9±0.4(323)

0.9±0.5(319)

Yellow

card

1.9±0.8(314)

2.0±0.9(44)

1.8±0.9(27)

1.4±0.5(118)

1.4±0.6(109)

1.8±0.8(220)

1.8±0.7(275)

1.8±0.7(241)

1.8±0.7(257)

514 H. LIU ET AL.

Dow

nloa

ded

by [

Ruh

r-U

nive

rsita

t Boc

hum

Uni

vers

itaet

sbib

lioth

ek]

at 0

1:12

09

June

201

6

typical performance (mean or median) in each match action orevent of any single player and the spread (lower and upperquartiles) of performance can be expressed as how manystandard deviations it is away from the mean (50 in thechart) of all the players (Liu et al., 2015c).

Furthermore, our profiles identified differences in the tech-nical demands of players from different positions. Defenders(full backs and central defenders) from strong teams mademore attacking- and passing-related actions but less defend-ing-related actions and events than their counterparts fromweak teams. A prior study on English Premier League showed

that, in recent years, central defender’s offensive contributionhad evolved by providing additional pass options when theteam was in possession of ball (Bush et al., 2015a). While ourresults indicate that defenders, both central defenders and fullbacks, from strong teams involve more in the organisation ofattacking, not only by providing extra passing lines, but alsoby invading the attacking third, contributing key passes,assists, even attempting directly to goal. Wide midfieldersfrom Top3 teams were observed having higher frequenciesin most of the attacking- and passing-related actions exceptcrossing. And they also received more yellow cards than the

Figure 2. Effect sizes of differences in mean of CV of each match action or event of (a) players from Bottom3 and Top3 teams; (b) players playing against Bottom3and Top3 teams; (c) players in lost/drawn games and in victory games; (d) players playing away and playing at home.Asterisks indicate the likelihood for the magnitude of the true differences in mean as follows: *possible; **likely; ***very likely; ****most likely. Asterisks located inthe trivial area denote for trivial differences.

JOURNAL OF SPORTS SCIENCES 515

Dow

nloa

ded

by [

Ruh

r-U

nive

rsita

t Boc

hum

Uni

vers

itaet

sbib

lioth

ek]

at 0

1:12

09

June

201

6

players from Bottom3 teams. This finding demonstrates thatalthough wide midfielders from weak teams had fewer oppor-tunities in participating offensive phase, they employed morestrategies of sending the ball from wide position into theopponent’s area (cross). This is probably because weak teamsare usually less developed and worse prepared in the offen-sive organisation (Bourbousson, Seve, & McGarry, 2010). Topteams’ central midfielders made more passing- and organis-ing-related actions (assist, ball touch, pass, pass accuracy,through ball, successful dribble), but less shots than bottomteams’ counterparts. This result may indicate that good centralmidfielders focused more on their “primary responsibility”: toorganise the offensive process by proper ball controls andpasses, rather than to invade too much into the opponent’sarea (Goncalves, Figueira, Macas, & Sampaio, 2014). Forwardsfrom strong teams were superior in most of the attacking-related match actions (more assists, shots on target, balltouches, passes, through balls, successful dribbles and higherpass accuracy, but less offsides), which can be easily explainedby that they could have more attacking opportunities withinbetter squads. Attention should be paid that there was apossibly trivial (at 90%CI, unclear at 99%) difference in thevariable of shot between forwards from top teams and frombottom teams, but a likely substantial difference (at 99%CI) inshot on target, which probably indicates that the ability ofseizing the opportunities and driving the ball on goal is thedeterminant of being a good forward. This can be supportedby previous studies on the success of football teams thatshowed the effectiveness and quality of the shots rather thenumber and quantity determine the result of football matches(Lago-Peñas et al., 2010; Liu et al., 2015a, 2015b; Yue, Broich, &Mester, 2014). Another concern is that the forwards from low-level teams achieved more aerial duels won, clearances andfouls committed than those from high-level teams, indicatingthat they had higher heading ability and participated more inthe defensive phases. Attackers that are good at heading theball are not only used for struggling scoring opportunitiesfrom crosses and high balls, but also used for defending setplays (i.e. corners and free-kicks). Hence, good heading abilityof forwards may not be a weakness but actually a benefitespecially for weaker teams who will face more of theseevents (crosses, high balls, corners and free kicks).

Analyses on the match-to-match variation of individualfootball players’ technical match actions bring us novel andfurther understanding of the match performance when con-trolling for the effects of contextual variables (i.e. team andopposition strength, match outcome and match location). Arecent study of Bush et al. (2015b) identified limited influencefrom match contexts (game location, outcome and oppositionstrength) on technical parameters match-to-match variation.However, our results identified that shot, shot on target andassist were three technical match actions that displayed highmatch-to-match variation in all the four match contexts, whichindicates that these variables are sensitive measures, thus,caution should be paid when using them to assess and inter-pret players’ match performance (Kempton et al., 2015). Whatis more, we found that the effects of the team and oppositionstrength on the performance of individual players were similarto the effects on the team performance (Shafizadeh, Taylor, &

Lago-Peñas, 2013). Our results showed that players from weakteams were more variable in attacking-related match actionsbut more stable in defending-related match actions compar-ing to the players from strong teams, which may reveal thatweak teams are less developed in attacking and organisingand focus more on defending. Previous studies showed thebetter teams were always linked to higher ratio of ball posses-sion in matches (Bradley, Lago-Peñas, Rey, & Sampaio, 2014;Jones, James, & Mellalieu, 2004), which explained why weakteams were inconsistent in attacking but predictable indefense. In addition, when we take the effect of oppositionstrength into account, we find that players were more consis-tent in attacking-related match actions when playing againstweak oppositions than when playing against strong ones. Thisfinding reinforces the idea that low-level teams are normallyless tactically developed and put more efforts in defendingbut with limited defensive intensities (Bourbousson et al.,2010), which lead to their opponents’ attacking morepredictable.

Moreover, when accounting for the effect of match out-come, we found that players showed more stable performancein variables of assist, shot on target, through ball, successfuldribble, offside and clearance in victory games than in lost ordrawn games. This result is related to the available researchfrom the perspective of team performance, because most ofthese variables were found to be associated to the winning offootball match(Castellano et al., 2012; Lago-Peñas et al., 2010,2011; Liu & Gómez, 2014; Liu et al., 2015a, 2015b). Similarly,Shafizadeh and colleagues (Shafizadeh et al., 2013) reportedthat the best teams of a championship achieved most persis-tent and consistent performance through the competition.

Previously, match location was considered an importantfactor influencing both offensive and defensive performancesof football (Almeida et al., 2014; Lago-Peñas & Lago-Ballesteros, 2011; Mackenzie & Cushion, 2013; Sarmentoet al., 2014). However, results from Bush and colleagues(Bush et al., 2015b) showed that there were no meaningfuldifferences between the match-to-match variation of players’home and away technical performances. By contrast, the pre-sent research identified that assist, shot and shot on targetwere three match actions that showed greater variation whenplaying away than when playing at home. These three vari-ables are all goal-scoring related. Given the fact that teams aremore likely to score at home than away (Poulter, 2009), thisresult is reasonably self-explanatory. In addition, in line withour hypothesis, we observed that there was a trend showingthat players’ technical performance was more stable whenplaying at home than when playing away. As referred in theintroduction, this is probably the result of positive psycho-physiological states and familiarity of the match facility andenvironment when playing at home pitch (Pollard, 2008).

Differently from earlier studies which indicated that foot-ball players’ “technical performances are variable per seregardless of context” (Bush et al., 2015b), current resultsseem to suggest that match-to-match variation of players’technical performances is affected by different match con-texts. And we find that the effects of team and oppositionstrength are greater than the effects of match location andmatch outcome on the variation of players’ technical match

516 H. LIU ET AL.

Dow

nloa

ded

by [

Ruh

r-U

nive

rsita

t Boc

hum

Uni

vers

itaet

sbib

lioth

ek]

at 0

1:12

09

June

201

6

performance, which implies that team and oppositionstrength should be attributed to higher level of environ-mental constrains than match location and match outcomewhen analysing football players’ match behaviours andactions (Eccles et al., 2009).

Practical applications and limitations

Established performance profiles can be used as baseline refer-ences for training and pre-match preparations and post-matchevaluations. For example, performance data of one player inany single match can be integrated into the database and becompared to the previous performance of his own or of others;thus, his competition status, strengths and weaknesses can beidentified, and ultimately, training sessions and match prepara-tions can be specifically modified according to the identifiedinformation. The profiles can also be used for talent identifica-tion, development and player selection in the transfer market.For instance, as referred in the discussion, good central mid-fielders should focus more on organising the offensive processby proper ball controls and passes, rather than invading toomuch into the opponent’s area. Top teams should choosedefenders who should not only have the ability to accomplishproperly their defending work but also the ability to getinvolved in the organising and attacking. Furthermore, coachesand analysts should be cautious when using some perfor-mance-related variables such as shot, shot on target and assistto assess and interpret players’ match performance, becausethey are sensitive measures with high match-to-match varia-tions. Meanwhile, these variables are so imperative for a for-ward that they should not be completely discounted bycoaches. Finally, among the four contextual variables analysed,the effects from strengths of team and opposition should bepaid more attention than the effects from match location andmatch outcome when analysing player’s match performance.

There are several limitations in the current research thatshould be considered in further studies concerning player’stechnical performance. First, except for the team andopposition strengths, match location and match outcome,the competition period (e.g. beginning, middle and end ofseason) is also an important contextual factor that has notbeen explored by the present study. Second, there is a lackof information about tactics and other in game variables(such as possession and time spent in areas of the pitch)that could play a major role in explaining offensive anddefensive performance outcomes. Finally, the currentresearch employed the performance data of entire matcheswhich cannot show the dynamics within matches, futurestudies can split the data into per half or per 15 minutesand consider the score-line influences, in order to digdeeper the within-match variation of performance.

Acknowledgements

Authors would like to express their appreciations to the OPTA SportsdataSpain Company (Madrid) for their data supports. Comments and sugges-tions from the anonymous reviewers which improved the quality of thispaper are appreciated as well.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

The first author was funded by the China Scholarship Council (CSC) fromthe Ministry of Education of P.R. China.

ORCID

Hongyou Liu http://orcid.org/0000-0003-4341-672XMiguel Gómez http://orcid.org/0000-0002-9585-3158Jaime Sampaio http://orcid.org/0000-0003-2335-9991

References

Almeida, C. H., Ferreira, A. P., & Volossovitch, A. (2014). Effects of matchlocation, match status and quality of opposition on regaining posses-sion in UEFA champions league. Journal of Human Kinetics, 41(1), 203–214. doi:10.2478/hukin-2014-0048

Bangsbo, J. (1994). The physiology of soccer: With special reference tointense intermittent exercise. Acta Physiologica Scandinavica, 151(suppl.619), 1–156.

Barriopedro, M. I. F., & Muniesa, C. (2012). Analisis de Datos en Las Cienciasde La Actividad Fisica y Del Deporte (Data Analysis in the Sciences ofPhysical Activity and Sport). Madrid: Piramide.

Batterham, A. M., & Hopkins, W. G. (2006). Making meaningful inferencesabout magnitudes. International Journal of Sports Physiology andPerformance, 1(1), 50–57.

Bourbousson, J., Seve, C., & McGarry, T. (2010). Space-time coordinationdynamics in basketball: Part 2. The interaction between the two teams.Journal of Sports Sciences, 28(3), 349–358. doi:10.1080/02640410903503640

Bradley, P. S., Carling, C., Archer, D., Roberts, J., Dodds, A., Di Mascio, M., &Krustrup, P. (2011). The effect of playing formation on high-intensityrunning and technical profiles in English FA Premier League soccermatches. Journal of Sports Sciences, 29(8), 821–830. doi:10.1080/02640414.2011.561868

Bradley, P. S., Lago-Peñas, C., Rey, E., & Sampaio, J. (2014). The influence ofsituational variables on ball possession in the English Premier League.Journal of Sports Sciences, 32(20), 1867–1873. doi:10.1080/02640414.2014.887850

Bush, M., Barnes, C., Archer, D. T., Hogg, B., & Bradley, P. S. (2015a).Evolution of match performance parameters for various playing posi-tions in the English Premier League. Human Movement Science, 39, 1–11. doi:10.1016/j.humov.2014.10.003

Bush, M. D., Archer, D. T., Hogg, R., & Bradley, P. S. (2015b). Factorsinfluencing physical and technical variability in the English PremierLeague. International Journal of Sports Physiology and Performance, 10,865–872. doi:10.1123/ijspp.2014-0484

Butterworth, A., O’Donoghue, P., & Cropley, B. (2013). Performance profil-ing in sports coaching: A review. International Journal of PerformanceAnalysis in Sport, 13(3), 572–593.

Castellano, J., Casamichana, D., & Lago, C. (2012). The use of match statisticsthat discriminate between successful and unsuccessful soccer teams.Journal of Human Kinetics, 31, 139–147. doi:10.2478/v10078-012-0015-7

Davids, K., Araujo, D., Correia, V., & Vilar, L. (2013). How small-sided andconditioned games enhance acquisition of movement and decision-making skills. Exercise and Sport Sciences Reviews, 41(3), 154–161.doi:10.1097/JES.0b013e318292f3ec

Eccles, D. W., Ward, P., & Woodman, T. (2009). Competition-specific pre-paration and expert performance. Psychology of Sport and Exercise, 10(1), 96–107. doi:10.1016/j.psychsport.2008.01.006

Eugster, M. J. A. (2012). Performance profiles based on archetypal athletes.International Journal of Performance Analysis in Sport, 12(1), 166–187.

Garganta, J. (2009). Trends of tactical performance analysis in team sports:Bridging the gap between research, training and competition. RevistaPortuguesa De Ciências Do Desporto, 9(1), 81–89.

JOURNAL OF SPORTS SCIENCES 517

Dow

nloa

ded

by [

Ruh

r-U

nive

rsita

t Boc

hum

Uni

vers

itaet

sbib

lioth

ek]

at 0

1:12

09

June

201

6

Gómez, M. A., Lago-Peñas, C., & Pollard, R. (2013). Situational variables. InT. McGarry, P. O’Donoghue, & J. Sampaio (Eds.), Routledge handbook ofsports performance analysis (pp. 259–269). London: Routledge.

Goncalves, B. V., Figueira, B. E., Macas, V., & Sampaio, J. (2014). Effect ofplayer position on movement behaviour, physical and physiologicalperformances during an 11-a-side football game. Journal of SportsSciences, 32(2), 191–199. doi:10.1080/02640414.2013.816761

Hopkins, W. G. (2007). A spreadhseet to compare means of two groups.Retrieved 12 January, 2015, from http://www.sportsci.org/2007/inbrief.htm#xcl2

Hopkins, W. G., Marshall, S. W., Batterham, A. M., & Hanin, J. (2009).Progressive statistics for studies in sports medicine and exercisescience. Medicine & Science in Sports & Exercise, 41(1), 3–13.doi:10.1249/MSS.0b013e31818cb278

Hughes, M. D., Evans, S., & Wells, J. (2001). Establishing normative profilesin performance analysis. International Journal of Performance Analysis inSport, 1(1), 1–26.

James, N., Mellalieu, S. D., & Jones, N. M. P. (2005). The development ofposition-specific performance indicators in professional rugby union.Journal of Sports Sciences, 23(1), 63–72. doi:10.1080/02640410410001730106

Jones, P. D., James, N., & Mellalieu, S. D. (2004). Possession as a perfor-mance indicator in soccer. International Journal of Performance Analysisin Sport, 4(1), 98–102.

Kempton, T., Sullivan, C., Bilsborough, J. C., Cordy, J., & Coutts, A. J. (2015).Match-to-match variation in physical activity and technical skill mea-sures in professional Australian Football. Journal of Science andMedicine in Sport, 18(1), 109–113. doi:10.1016/j.jsams.2013.12.006

Kinrade, N. P., Jackson, R. C., & Ashford, K. J. (2015). Reinvestment, taskcomplexity and decision making under pressure in basketball.Psychology of Sport and Exercise, 20, 11–19. doi:10.1016/j.psychsport.2015.03.007

Kudo, K., & Ohtsuki, T. (2008). Adaptive variability in skilled human move-ments. Information and Media Technologies, 3(2), 409–420.

Lago-Peñas, C., & Lago-Ballesteros, J. (2011). Game location and teamquality effects on performance profiles in professional soccer. Journalof Sports Science and Medicine, 10(3), 465–471.

Lago-Peñas, C., Lago-Ballesteros, J., Dellal, A., & Gómez, M. (2010). Game-related statistics that discriminated winning, drawing and losing teamsfrom the Spanish Soccer League. Journal of Sports Science and Medicine,9(2), 288–293.

Lago-Peñas, C., Lago-Ballesteros, J., & Rey, E. (2011). Differences in perfor-mance indicators between winning and losing teams in the UEFAChampions League. Journal of Human Kinetics, 27, 137–148.doi:10.2478/v10078-011-0011-3

Liu, H., & Gómez, M. A. (2014, November). Relationships between matchperformance indicators and match outcome in 2014 Brazil FIFA world cup.Paper presented at the VIII Congreso Internacional de la AsociaciónEspañola de Ciencias del Deporte, Caceres.

Liu, H., Gómez, M. A., Lago-Peñas, C., & Sampaio, J. (2015a). Match statisticsrelated to winning in the group stage of 2014 Brazil FIFA World Cup.Journal of Sports Sciences, 33(12), 1205–1213. doi:10.1080/02640414.2015.1022578

Liu, H., Hopkins, W., Gómez, M. A., & Molinuevo, J. S. (2013). Inter-operatorreliability of live football match statistics from OPTA Sportsdata.International Journal of Performance Analysis in Sport, 13(3), 803–821.

Liu, H., Hopkins, W. G., & Gomez, M.-A. (2015b). Modelling relationshipsbetween match events and match outcome in elite football. EuropeanJournal of Sport Science, 1–10. doi:10.1080/17461391.2015.1042527

Liu, H., Yi, Q., Gimenez, J. V., Gómez, M. A., & Lago-Peñas, C. (2015c).Performance profiles of football teams in the UEFA ChampionsLeague considering situational efficiency. International Journal ofPerformance Analysis in Sport, 15(1), 371–390.

Mackenzie, R., & Cushion, C. (2013). Performance analysis in football: Acritical review and implications for future research. Journal of SportsSciences, 31(6), 639–676. doi:10.1080/02640414.2012.746720

O’Donoghue, P. (2005). Normative profiles of sports performance.International Journal of Performance Analysis in Sport, 5(1), 104–119.

O’Donoghue, P. (2013). Sports performance profiling. In T. McGarry, P.O’Donoghue, & J. Sampaio (Eds.), Routledge Handbook of sports perfor-mance analysis (pp. 127–139). London: Routledge.

Pollard, R. (2008). Home advantage in football: A current review of anunsolved puzzle. The Open Sports Sciences Journal, 1(1), 12–14.doi:10.2174/1875399X00801010012

Poulter, D. R. (2009). Home advantage and player nationality in interna-tional club football. Journal of Sports Sciences, 27(8), 797–805.doi:10.1080/02640410902893364

Rampinini, E., Coutts, A. J., Castagna, C., Sassi, R., & Impellizzeri, F. M. (2007).Variation in top level soccer match performance. International Journal ofSports Medicine, 28(12), 1018–1024. doi:10.1055/s-2007-965158

Rampinini, E., Impellizzeri, F. M., Castagna, C., Coutts, A. J., & Wisloff, U.(2009). Technical performance during soccer matches of the ItalianSerie A league: Effect of fatigue and competitive level. Journal ofScience and Medicine in Sport, 12(1), 227–233. doi:10.1016/j.jsams.2007.10.002

Russell, M., Rees, G., & Kingsley, M. I. C. (2013). Technical demands ofsoccer match play in the English championship. Journal of Strengthand Conditioning Research, 27(10), 2869–2873. doi:10.1519/Jsc.0b013e318280cc13

Sarmento, H., Marcelino, R., Anguera, M. T., Campanico, J., Matos, N., &Leitao, J. C. (2014). Match analysis in football: A systematic review.Journal of Sports Sciences, 32(20), 1831–1843. doi:10.1080/02640414.2014.898852

Seifert, L., Button, C., & Davids, K. (2013). Key properties of expert move-ment systems in sport an ecological dynamics perspective. SportsMedicine, 43(3), 167–178. doi:10.1007/s40279-012-0011-z

Shafizadeh, M., Taylor, M., & Lago-Peñas, C. (2013). Performance consis-tency of international soccer teams in Euro 2012: A Time series analysis.Journal of Human Kinetics, 38, 213–226. doi:10.2478/hukin-2013-0061

Spencer, M., Losnegard, T., Hallen, J., & Hopkins, W. G. (2014). Variabilityand predictability of performance times of elite cross-country skiers.International Journal of Sports Physiology and Performance, 9(1), 5–11.doi:10.1123/Ijspp.2012-0382

Stergiou, N., Yu, Y., & Kyvelidou, A. (2013). A perspective on human move-ment variability with applications in infancy motor development.Kinesiology Review, 2, 93–102.

Vilar, L., Araujo, D., Davids, K., & Button, C. (2012). The role of ecologicaldynamics in analysing performance in team sports. Sports Medicine, 42(1), 1–10. doi:10.2165/11596520-000000000-00000

Yue, Z. Y., Broich, H., & Mester, J. (2014). Statistical analysis for the soccermatches of the first Bundesliga. International Journal of Sports Scienceand Coaching, 9(3), 553–560. doi:10.1260/1747-9541.9.3.553

518 H. LIU ET AL.

Dow

nloa

ded

by [

Ruh

r-U

nive

rsita

t Boc

hum

Uni

vers

itaet

sbib

lioth

ek]

at 0

1:12

09

June

201

6