Ball Possession Strategies in Elite Soccer According to the Evolution of the Match‐Score: the Influence of Situational Variables

Journal of Human Kinetics volume 25 2010, 93‐100 93 Section III – Sport, Physical Education & Recreation

by Carlos Lago‐Peñas1, Alexandre Dellal2

In soccer, the ability to retain possession of the ball for prolonged periods of time has been suggested to be linked to success. The accuracy of this assertion was investigated by examining 380 matches involving Spanish League

First Division teams during the 2008‐2009 season. Possession of the ball, according to the status of the match (win‐ ning, drawing and losing), was recorded during the different matches using a multiple‐camera match analysis sys‐ tem (Gecasport®). The results suggest that the best classified teams maintained a higher percentage of ball posses‐ sion and that their pattern of play was more stable. The coefficient of variation, with respect to ball possession per match, was smaller for the best placed teams. Indeed, first placed F.C. Barcelona had the smallest coefficient of varia‐ tion for possession time (8.4%), while bottom placed Recreativo showed the highest values with 17.1%. Linear re‐ gression analysis showed that possession strategies were influenced by situation variables. Team possession was greater when losing than when winning (p<0.01) or drawing (p<0.01), home teams enjoyed greater possession than visiting teams (p<0.01), and playing against strong opposition was associated with a reduction in time spent in pos‐

session (p<0.01). The findings indicate that strategies in soccer are influenced by situational variables and that teams alter their playing style accordingly during the match. Key words: match analysis; possession strategies; soccer; team performance; tactical component

area of great interest in performance analysis Introduction (Hughes and Bartlett, 2002). In soccer, for a goal to be scored a team usually Performance analysis refers to the objective re‐ has to have possession of the ball. Although it might cording and examination of behavioral events that be anticipated that longer periods of possession occur during sporting competition (Carling et al., Sport, Physical Education & Recreation should predict goal scoring, support for this notion 2005; Dellal, et al., 2010). The main aim of analyzing is divided. Bate (1988), for example, found that the one’s own team’s performance is to identify higher number of possessions a team had, the strengths that can be further developed, and weak‐ greater the chance they had of entering the attacking nesses that might be improved. Understanding the third of the field and consequently more goal scoring differences between the patterns of play developed opportunities were created. On the basis of this by successful and unsuccessful teams, as well as finding, Bate (1988) rejected the notion of possession those of the same team in different matches, is an soccer and advocated a direct strategy. However, Hughes and Franks (2005), Grant et al. (1999), Hook

1 - Faculty of Education and Sports Sciences, University of Vigo, Pontevedra, Spain 2 - University of Sports Sciences, Strasbourg, France and National Center of Medicine and Science in Sport, Tunis, Tunisia

Authors submitted their contribution of the article to the editorial board. Accepted for pinting in Journal of Human Kinetics vol. 25/2010 on July 2010. 94 Ball Possession Strategies in Elite Soccer According to the Evolution of the Match‐Score: the Influence of Situational Variables

and Hughes (2001) and Bloomfield et al. (2005a) re‐ tion found in this study and those of Sasaki et al. ported that successful teams (e.g., European Cham‐ (1999) and Tucker et al. (2005) emphasize the need to pions League, World Champions, Europa Cup) validate the models developed across numerous ad‐ maintained possession for longer than unsuccessful ditional teams. Secondly, Lago and Martin (2007) did teams. In contrast, Stanhope (2001) found that time not incorporate into their study the quality of oppo‐

in possession of the ball was not indicative of success sition as an independent variable to explain the de‐ in the 1994 World Cup. terminants of possession of the ball in soccer. Many of the equivocal findings of previous stud‐ In this context, the aim of this investigation was ies examining the possession strategies of successful to provide a large‐scale study of elite professional and unsuccessful teams may originate from con‐ soccer teams and examine the effects of situational ceptual and methodological issues. For example, variables on ball possession strategies. It was hy‐ Jones et al. (2004) indicate that many have failed to pothesized that ball possession was influenced by demonstrate the reliability of the data gathering such variables as match location, match status, qual‐ system used. In addition, selecting matches played ity of opposition and by the level of the team. On the as part of individual tournaments means that the basis of these findings, it is hoped that the informa‐ chosen teams (successful and unsuccessful) are im‐ tion will be of value in contributing to more tactical balanced in term of the strength of opposition and knowledge for the prescription of specific exercises number of matches played. Such factors are likely to within the training regimen and for analyzing match influence a team´s performance and might explain performance. the discrepancies seen among studies. According to Taylor et al. (2008), effective Materials and Methods evaluation of soccer performance requires knowl‐ edge of the contextual factors that can potentially af‐ Study Design fect behaviour incidence and outcomes (Carling et al., 2005; Dellal, 2008). Existing literature (Jones et al., Although it might be anticipated that longer pe‐ 2004; Shaw and O´Donoghue, 2004) suggests that the riods of possession should predict increased goal variables match location (i.e., playing at home or scoring opportunities, support for this notion is di‐

Sport, Physical Education & Recreation Sport, Physical Education away) (Tucker et al., 2005), match status (i.e., vided. To verify this, all 380 matches of the Spanish whether the team was winning, losing or drawing) soccer League played throughout the 2008‐2009 sea‐ (Bloomfield et al., 2005a, 2005b; Lago and Martin, son were assessed. The dependent variable was the 2007; Taylor et al., 2008; Tucker et al, 2005) and the proportion of time (%) during matches in which the quality of the opposition (strong or weak) (Taylor et team had possession of the ball when the ball was in al., 2008), which all require consideration when play. Empirical evidence suggests that the variables evaluating soccer performances. Unfortunately, match location, match status (winning, drawing and these findings are still inconclusive given that most losing) and the quality of the opponent can affect of these studies were based on small sample sizes, soccer performances. These factors were included in and with the exception of Lago and Martin (2007) the study as independent variables.

and Taylor et al. (2008), the existing performance analysis literature has examined situation factors in‐ Match Sample dependently, thereby neglecting to account for the The examined sample consisted of 380 Spanish complex and dynamic nature of soccer performance Soccer League First Division matches played (MacGarry and Franks, 2003; Reed and O´Donoghue, throughout the 2008‐2009 season. The performance 2005; Taylor et al., 2008). of one team obviously impacts upon the second (i.e., However, the results of Lago and Martin (2007) the frequency and duration of possession is depend‐ and Taylor et al. (2008) have two limitations in their ent on the opposition). As a consequence, data were

findings. First, Taylor et al. (2008) adopted a fine‐ analyzed using one team from each match. The grained approach to soccer analysis by considering number of observations was therefore 380. The col‐ the performances of a single team over a sustained lected data (possession of the ball, especially ac‐ period (two seasons). An obvious limitation of case cording to the match status: winning, drawing and studies designs is that generalization of findings is losing during a match) were provided by a multiple‐ precluded. The contradictory effects of match loca‐ camera match analysis system (Gecasport®), a pri‐

Journal of Human Kinetics volume 25 2010, http://www.johk.awf.katowice.pl by C. Lago‐Peñas and A. Dellal 95

vate company dedicated to assessing the perform‐ of time each team was winning (Minutes Winning: ance of teams in the Spanish Soccer League. The ac‐ MW), drawing (Minutes Drawing: MD) and losing curacy of the Gecasport® System has been verified (Minutes Losing: ML) during a match were included. by Gómez et al. (2009). Written permission to analyze This means that the panel match status in the regres‐ data was provided by Gecasport®. Ethics approval sion model presents two coefficients from the com‐

for all experimental procedures was granted by our parison of Drawing to Losing and from the compari‐ institute’s Human Research Ethics Committee. son of Winning to Losing. Match location was re‐ corded as “home” or “away”, depending on whether Reliability Testing or not the sampled team was playing on its own ground or that of its opponent: 0 = playing at home, Reliability of the data was assessed through inter‐ 1= playing away (Match Location: ML). Quality of and intra‐ observer test procedures. Inter‐observer opposition was the distance in the end‐of‐season‐ reliability was assessed by the authors, coding five ranking between competing teams (Quality of Op‐ matches randomly selected, with data being com‐ position: QO). The 20 teams were divided into four pared with those provided by Gecasport®. Intra‐ob‐ groups according to their final league ranking server reliability was completed by the authors, (Team: TE). Group 1 contained the top five teams, coding five random matches selected from the data Group 2 contained those that finished 6th through sample. Following a six‐week period, to avoid any 10th, and Group 3, those that finished 11th through possible negative learning effects, the matches were 15th, and Group 4, those that finished 16th through recoded and the two data sets compared. The Kappa 20th. β1 is the intercept and, β2, β3, β4, β5, and β6, (K) values recorded from 0.93 to 0.98. The parame‐ the impact of each one of the independent variables. ters included in the inter‐ and intra‐ observer reli‐ ε ability were ball possession, the time each team was i is the disturbance term. The regression model losing, drawing or winning during a match, where used was therefore: the match was played (match location) and the dis‐ PO = β1 + β2 .MWi + β3 .MDi + β4 .MLi + β5 . QOi tance between end‐of‐season rankings of the com‐ + β6 . TEi+ εi [1] peting teams. When estimating the regression models, no evi‐

dence of heteroscedasticity in the residuals or multi‐ Statistical analysis colinearity among the regressors was found. More‐ All statistical analysis was performed using SPSS over, the Ramsey Regression Equation Specification for Windows, version 17.0 (SPSS Inc., Chicago, USA). Error Test (RESET) (Ramsey, 1969) revealed no The level of significance was set at p<0.05. To exam‐ specification problems. When interpreting the statis‐ ine the inter‐match variation in the teams’ ball pos‐ tical results, positive or negative coefficients indicate session times, Pearson coefficients of variation (CV) a greater or lower propensity to increase/decrease were calculated. The CV is defined as the relation‐ ball possession. ship between the standard deviation and the arith‐ S Results CV = metic mean: x . A standard multiple regres‐ The mean ball possession values for all 20 teams, Sport, Physical Education & Recreation sion was used to examine how much the ball posses‐ along with their CV were presented in the Table 1. sion (Possession: PO) was explained by the contex‐ Considerable differences can be seen between these tual variables (match status, match location, quality possession times, with successful teams showing a of opposition and by the level of the team). Posses‐ smaller CV during the season. For example, first sion was deemed to start when a player on the ana‐ placed F.C. Barcelona had the smallest CV for pos‐ lyzed team had sufficient control over the ball to en‐ session time (8.42%), while bottom‐placed Recreativo able a deliberate influence on its direction. Posses‐ presented the highest at 17.1%. In addition, the more sion continued until the ball either went out of play, successful teams showed higher mean possession an opposing player touched the ball or the referee percentages than the other teams. For example, the blew the whistle for an infringement (Jones et al., mean for F.C. Barcelona (65.29%) was 16.24% greater 2004). A possession of 50% means that a team pos‐ than those found for the bottom‐placed Recreativo sessed the ball for half the time the ball was in play. (48.05%). To measure the variable “match status”, the lengths

96 Ball Possession Strategies in Elite Soccer According to the Evolution of the Match‐Score: the Influence of Situational Variables

Table 1 Ball possession characteristics for the teams of Spanish league Division One for the 2008‐2009 season Ranking in Team Mean possession (%) CV league 1 FC. Barcelona 64.3 8.4 2 Real Madrid 53.2 11.5

3 Sevilla 54.6 12.2 4 Atlético de Madrid 52.8 10.8 5 Villarreal 50.2 12.0 6 Valencia 54.3 12.3 7 Deportivo 49.0 12.4 8 Málaga 44.9 12.1 9 Mallorca 47.3 13.4 10 Espanyol 48.3 12.7 11 Almería 49.4 14.0 12 Racing 46.7 14.4 13 Athletic 46.5 14.2 14 Sporting 44.6 16.1 15 Osasuna 49.0 15.6 16 Valladolid 50.4 15.7

17 Getafe 51.9 16.3 18 Betis 48.9 15.9 19 Numancia 46.2 16.8 20 Recreativo 48.1 17.2

The independent and interactive effects of situa‐ when winning (p<0.01) or drawing (p<0.01). For each tion variables on the teams’ ball possession were de‐ minute winning the team possession decreased by scribed in Table 2. The linear regression model in‐ 0.09% compared with each minute losing. For exam‐ volving the variables MW, MD, match location and ple, if team A were winning for the full 90 min, the Sport, Physical Education & Recreation Sport, Physical Education quality of opposition explained about 48% of the expected possession would be 8 percentage points variance in team ball possession. When all the pre‐ lower than the opponent’s possession (90 min losing dictor variables were equal to zero, possession was multiplied by 0.09). For each minute drawing the 55.22%. Possession was greater when losing than team possession decreased by 0.04% compared with each minute losing. Playing against strong opposi‐ Table 2 tion was associated with a reduction in time spent in The effects of match status, match location and possession of the ball (p<0.01). Each point of differ‐ quality of opposition on team possession ence in the end‐of‐season‐ranking between the teams Variables increases/decreases a corresponding team’s posses‐ Match status sion by 0.52%. Playing away reduced possession

Drawing ‐0.04* (0.01) time by 2.43%, in comparison with playing at home. Winning ‐0.09* (0.01) Moreover, the teams of Groups 2 and 3 showed, re‐ Match location ‐2.43* (0.47) spectively, 4.01% and 3.01% less ball possession than those of Group 1 (p<0.01), whereas no significant dif‐ Quality of opposition ‐0.52* (0.04) ferences were observed in ball possession between the teams of Group 1 and Group 4. Team Finally, to illustrate the findings, Tables 3 and 4 Group 1 ‐4.01* (0.68) provide different values for each variable in the re‐ Group 2 ‐3.01* (0.75) gression model to simulate ball possession for F.C Group 3 1.22 (0.87) Barcelona in a match against Espanyol (10th in the Intercept 55.24* (0.83) Number of observations end‐of‐season‐ranking). The predicted possession R2 380 0.48 differs considerably – by up to 9% ‐ according to Notes: Standard errors of the mean are in parentheses. match status and match location. *(p < 0.01).

Journal of Human Kinetics volume 25 2010, http://www.johk.awf.katowice.pl by C. Lago‐Peñas and A. Dellal 97

Table 3 Simulated ball possession for F.C. Barcelona, depending on match location, quality of opposition and match status in the game between F.C. Barcelona and Espanyol Minutes winning

0 15 30 45 60 75 90

0 62 61 60 58 56 55 53 15 61 60 58 57 55 54 30 60 59 56 56 54 Minutes 45 59 58 55 54 drawing 60 58 57 54 75 57 56 90 56 Note: The outcome measure is the percentage of time (%) in which F.C. Barcelona has the possession of the ball

Table 4 Simulated ball possession for F.C. Barcelona, depending on match location, quality of opposition and match status in the game between Espanyol and F.C. Barcelona. Minutes winning

0 15 30 45 60 75 90 0 59 59 57 55 54 52 51 15 58 57 55 54 52 52 30 57 56 54 53 51 Minutes 45 56 55 52 51 drawing 60 55 54 51 75 54 53 90 53

Note: The outcome measure is the percentage of time (%) in which F.C. Barcelona has the possession of the ball

match location and match status on soccer perform‐ Discussion ances (Bloomfield et al., 2005a, 2005b; Dellal, 2008; Jones et al., 2004; Lago, 2009; Taylor et al., 2008). The aim of the present work was to provide a However, the small sample sizes examined, the ex‐ large‐scale study of professional soccer teams and amination of situation variables independently, the identify the effects of match location, match status, limitation of case study designs and the evaluation quality of opposition and the level of the team on of matches in competitions in which the selected ball possession strategies. teams (successful and unsuccessful) were imbal‐ Existing performance analysis studies have pro‐ anced in terms of the strength of opposition and the vided inconclusive information regarding the rela‐ Sport, Physical Education & Recreation number of matches played, make it difficult to come tionship between ball possession and competition to any conclusion. success (Bate, 1988; Grant et al., 1999; Hook and The results of the present study indicated that Hughes, 2001; Hughes and Franks, 2005; Stanhope, ball possession was influenced by situational vari‐ 2001). Some authors have suggested the existence of ables, either independently or interactively. Losing‐ patterns of play involving ball possession shown by match status was associated with an increase in ball successful and unsuccessful teams (Bloomfield et al, possession. Similar to the findings reported by 2005a; Hughes and Franks, 2005), while others indi‐ Bloomfield et al. (2005a), Jones et al. (2004), Lago and cate that ball possession time is not a marker of suc‐ Martin (2007) and Sasaki et al. (1999), the present re‐ cess in a match (Bate, 1988; Stanhope, 2001). More‐ sults support that possession strategies are influ‐ over, the effects of situational variables on ball pos‐ enced by score‐lines. Possession was always found session strategies are not well‐known. Existing per‐ to be greater when losing than winning (p<0.01) or formance analysis has provided preliminary infor‐ drawing (p<0.01). This could be explained by mation on the effects of situation variables such as changes in tactics and the style of play adopted by

98 Ball Possession Strategies in Elite Soccer According to the Evolution of the Match‐Score: the Influence of Situational Variables

teams according to the status within the match tion on the technical aspects of performance within a (evolving score). When ahead, teams decreased their professional soccer team was non‐existent. Perhaps, possession, suggesting they preferred to play coun‐ the “strong‐weak” dichotomy used in this study may terattacking or direct play (that is, move the ball lack the necessary sensitivity to differentiate changes quickly to within scoring range, often using long in behavior incidence as a function of the quality of

passes or long balls downfield). However, when be‐ the opposition. Moreover, Taylor et al. (2008) hind, they increased their possession, suggesting adopted a fine‐grained approach to soccer analysis they preferred to “control” the match by dictating by considering a single team’s performances over a play. This style of play is indirect play (also calling sustained period (two seasons). This contrasts with possession style, which is slower than direct play previous soccer literature that has tended to aggre‐ and uses many short passes, while weaknesses in the gate performances of different teams during analy‐ opposition defense are sought). Consequently, the sis. tactical component directly influences the physical This study also revealed that the characteristic of and the technical demands within a match‐play, ac‐ ball possession in each match varies depending on cording to the findings of Dellal et al. (2010a, 2010b). the team analyzed. The top‐placed teams showed Playing away was characterized by a decrease in lower CVs for ball possession per match compared team possession. Visiting teams reduced possession to those who finished lower in the table. For exam‐ by 2.43 percentage points in comparison with home ple, F.C. Barcelona (the champion) showed the teams. These results provide evidence of home ad‐ smallest CV (8.4%), while Recreativo (bottom of the vantage in soccer. This in line with the findings of table) showed the greatest (17.1%). It would appear Bloomfield et al. (2005b), Jones et al. (2004), Lago and that the teams with the best performance are able to Martin (2007), Sasaki et al. (1999), Tucker et al. (2005) impose and maintain their pattern of play despite and Taylor et al. (2008). Despite the fact that home the alteration in variables over the match (e.g., advantage in soccer is a well‐known and well‐docu‐ evolving score) and between matches (e.g., playing mented fact (Bloomfield et al., 2005a; Lago and Mar‐ at home or away). Different teams appear to employ tin, 2007; Tucker et al., 2005), the precise causes and different strategies when ahead, level or behind, re‐ their simple or interactive effects on performance are flecting the individual styles of coaching and man‐

Sport, Physical Education & Recreation Sport, Physical Education still not clear. However, the most plausible explana‐ agement, the budget, the characteristics of the play‐ tions are: crowd effects, travel effects, familiarity, ers, team formation and philosophy of play based on referee bias, territoriality, specific tactics, rule factors, the tradition of the clubs. and psychological factors (Carling et al., 2005). The main findings of the present study empha‐ The top‐placed teams in the Spanish League for size the importance of accounting for match location, the 2008‐2009 season presented a higher percentage quality of opposition and match status during the ball possession per match than the less successful assessment of tactical aspects of soccer performance teams. The present results are similar to those re‐ (Carling et al., 2005; Taylor et al., 2008). Possession ported by Jones et al. (2004) and Lago and Martín strategies appear to be influenced by situational (2007), whose studies had designs similar to that of variables, either independently or interactively. The

the present work. They found that top teams re‐ characteristics of ball possession time in each match tained more possession than their opponents, sug‐ may differ for different teams. The top‐placed teams gesting that they prefer to control the match by dic‐ showed lower CVs for ball possession per match tating play. Bloomfield et al. (2005a), for example, compared to those who finished lower in the table. showed that the top three teams in the English Pre‐ The detailed evaluation of the influence of match lo‐ mier League in the 2003‐2004 season (Chelsea, Man‐ cation, quality of opposition, and match status (win‐ chester, and Arsenal), dominated possession against ning, drawing and losing) on soccer performance their opponents whether winning, losing or draw‐ within this study presents a number of implications

ing. Hughes and Franks (2005) suggested that be‐ for analysts and coaches. Existing recommendations cause successful teams (League champions, World suggest that the scouting of upcoming opposition champions, European champions) do not resort to should be carried out under circumstances that are direct play, there are patterns of play for successful reflective of the conditions under which the future and unsuccessful teams. Taylor et al. (2008), how‐ match will occur. However, such procedures are ever, found that the influence of quality of opposi‐ unlikely to be practical due to time and resource

Journal of Human Kinetics volume 25 2010, http://www.johk.awf.katowice.pl by C. Lago‐Peñas and A. Dellal 99

constraints. Consequently, by establishing the im‐ objective by factoring in the effects of situational pact of particular situational variables on perform‐ variables. Finally, if a notational analyst or coach has ance, teams can be observed, when possible, with identified that the technical, physical or tactical as‐ appropriate adjustments being made to analyses pects of performance are adversely influenced by based on knowledge of such effects. Similarly, post‐ specific situational variables, possible causes can be

match assessments of the technical, tactical, and examined and match preparation focused on reduc‐ physical aspects of performance can be made more ing such effects.

References Bate R. Football chance. Tactics and strategy. In: Science and Football V. Eds: Reilly T., Less A, Davies K. and Murphy W. 1988. London: E and FN Spon. 293‐301. Bloomfield JR, Polman RCJ, O’Donoghue PG. Effects of score‐line on team strategies in FA Premier League Soccer. J Sports Sci, 2005. 23: 192‐193. Bloomfield JR, Polman, RCJ, O’Donoghue PG. Effects of score‐line on intensity of play in midfield and forward players in the FA Premier League. J Sports Sci, 2005. 23: 191‐192. Carling C, Williams AM, Reilly T. Handbook of soccer match analysis: A systematic approach to improving

performance. Abingdon, UK: Routledge, 2005.

Dellal A, Chamari C, Wong DP, Ahmaidi S, Keller D, Barros MLR, Bisciotti GN, Carling C. Comparison of physical and technical performance in European professional soccer match‐play: The FA Premier League and La LIGA. Eur J Sport Sci, ahead to print. Dellal A, Wong DP, Moalla W, Chamari K. Physical and technical activity of soccer players in the French first division – with special reference to the playing position. Int Sport Med J, 2010. 11. Dellal A. (2008). Analyze of the soccer player physical activity and of its consequences in the training: special reference to the high intensities intermittent exercises and the small sided‐games. Thesis in the University of Strasbourg, Sport Sciences Department, France. Gómez M, Álvaro J, Barriopedro MI. Differences in playing actions between men and women in elite soccer teams. 2009. In: Current trends in performance analysis. World Congress of Performance Analysis of Sports VIII. Eds: Hökelmann A, Witte K, O´Donoghue P. 2009. Magdebourg: Shaker Verlag: 50‐51. Grant, AG, Williams AM, Reilly T. An analysis of the successful and unsuccessful teams in the 1998 World Cup. J Sports Sci 1999. 17: 827. Hook C, Hughes MD. Patterns of play leading to shots in Euro 2000. In: Pass.com. Ed: CPA (Center for Performance Analysis). 2001. Cardiff: UWIC. 295‐302.

Hughes MD, Bartlett RM. The use of performance indicators in performance analysis, J Sports Sci, 2002. 20: 739‐ Sport, Physical Education & Recreation 754. Hughes MD, Cooper S, Nevill A. Analysis procedures for non‐parametric data from performance analysis. Int J Perf An Sport, 2002. 2: 6‐20. Hughes MD, Franks I. Analysis of passing sequences, shots and goals in soccer. J Sport Sci, 2005. 23: 509‐514. Hughes M, Robertson K, Nicholson A. Comparison of patterns of play of successful and unsuccessful teams in the 1986 World Cup for soccer. In: Science and Football. Eds: Reilly T, Lees A, Davis K, Murphy WJ. 1988. London: E. and F.N. Spon. 363‐367. James N, Mellalieu SD, Holley C. Analysis of strategies in soccer as a function of European and domestic competition. Int J Per An Sport, 2002. 2: 85‐103. Jones P, James N, Mellalieu SD Possession as a Performance Indicator in Soccer. Int J Per An Sport, 2004. 4: 98‐ 102.

100 Ball Possession Strategies in Elite Soccer According to the Evolution of the Match‐Score: the Influence of Situational Variables

Lago C. Consequences of a busy soccer match schedule on team performance: empirical evidence from Spain. Int Sport Med J, 2009. 10: 86‐92. Lago C, Martin R. Determinants of possession of the ball in soccer. J Sports Sci, 2007. 25: 969‐974. McGarry T, Franks I. The science of match analysis. In Science and Soccer. Eds: Reilly, T., Clarys, J., Stibbe, A.

2003. London: Routledge. 265‐275 Ramsey JB. Test for specification errors in classical lineal least squares regression analysis. J Royal Sta Soc, 1969. 31: 350‐371. Reed D, O’Donoghue P. Development and application of computer‐based prediction methods. Int J Per An Sports, 2005. 5: 12‐28. Sala‐Garrido R, Liern V, Martinez A, Boscá J. Analysis and Evolution of Efficiency in the Spanish Soccer League (2000/01 ‐ 2007/08). Journal of Quantitative Analysis in Sport, 2009. 5: 34‐37 Sasaki Y, Nevill A, Reilly T. Home advantage: A case study of Ipswich Town football club during the 1996‐1997 season. J Sports Sci, 1999. 17: 831. Shaw J, O´Donoghue P. The effect of scoreline on work rate in amateur soccer. In P. O´Donoghue and M.D. Hughes (Eds.), Notational analysis of sport VI (pp. 84‐91).2004. Cardiff: UWIC.

Stanhope J. An investigation into possession with respect to time, in the soccer world cup 1994. In: Notational Analysis of Sport III. Ed: Hughes MD. 2001. Cardiff, UK: UWIC. 155‐162. Taylor JB, Mellalieu SD, James N, Shearer D. The influence of match location, qualify of opposition and match status on technical performance in professional association football. J Sports Sci, 2008. 26: 885‐895. Taylor S, Williams M. (2002) A Quantitative analysis of Brazil’s performances. Insight, 2002. 3: 28‐30.

Sport, Physical Education & Recreation Sport, Physical Education

Corresponding author Carlos Lago Peñas, University of Vigo, Faculty of Sports Sciences Campus Universitario s/n 36005. Pontevedra. Spain. phone: +34986801700 fax: +34986801701 e‐mail: [email protected]

Journal of Human Kinetics volume 25 2010, http://www.johk.awf.katowice.pl