- Email: [email protected]
Dissecting the survivors – how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues
+Model
ARTICLE IN PRESS
SCISPO-3080; No. of Pages 9 Science & Sports (2017) xxx, xxx—xxx
Disponible en ligne sur
ScienceDirect www.sciencedirect.com
ORIGINAL ARTICLE
Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues Comment l’âge du club contribue-t-il à maintenir dans les ligues de football européennes la première position de l’équipe P.R. Mourao Department of Economics, Economics & Management School, University of Minho, 4700 Braga, Portugal Received 25 August 2015; accepted 21 January 2017
KEYWORDS Soccer; Survival; Cox proportional hazard rates
Summary Objectives. — To test the hypothesis that the age of a club is a major determinant of the higher survival rates exhibited by European soccer teams. News. — We studied 185 teams from the 2006/2007 season to the 2012/2013 season. We used Cox proportional hazards regression models. We also controlled for the age effect with variables suggested by the literature. These variables included transfer flows, the presence of the team at UEFA competitions (Champions League or Europa League), stadium attendance, and the percentage of national players. Prospect. — We observed that the age of the club is an important determinant in explaining the higher survival rates of the teams belonging to older clubs. More nuance was discovered when the teams were analyzed by professional league. Among English teams, the survival rate at the top is positively affected by having higher percentages of national players, higher transfer inflows, participation in UEFA competitions and higher stadium attendance. Stadium attendance is an important factor in keeping a team at the top, independent of the European league (with the exception of the French competition). French teams’ survival tends to benefit from the acquisition of players (proxied by higher transfer outflows). Although significant, the age effect nevertheless does not prevent young clubs (those founded within the last 50 years)
E-mail addresses: [email protected], paulo r [email protected]. http://dx.doi.org/10.1016/j.scispo.2017.01.008 0765-1597/© 2017 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
2
P.R. Mourao from achieving first place. Young clubs that manage transfers well, have impressive stadium attendance, and are stimulated by UEFA participation tend to experience additional positive effects on their survival rate at the top. Conclusion. — The most robust results indicate that the age effect is a statistically significant determinant. Older clubs tend to benefit from four age-related dimensions: the relevance of assets, more efficient decision-making processes, monopolistic market power and the control of decisions by the executive board. © 2017 Elsevier Masson SAS. All rights reserved.
MOTS CLÉS Football ; Survie ; Taux de risque proportionnel
Résumé Objectifs. — Tester l’hypothèse que l’âge du club est un déterminant des taux de survie plus élevés des équipes européennes de football. Actualités. — Nous avons étudié 185 équipes de la saison 2006/2007 à la saison 2012/2013. Nous avons utilisé des modèles de régression des risques de Cox. Nous avons également contrôlé pour l’effet de l’âge avec les variables suggérées par la littérature. Ces variables comprenaient les flux de transfert, la présence de l’équipe lors des compétitions de l’UEFA (Ligue des Champions ou Europa League), la fréquentation des stades, et le pourcentage de joueurs nationaux. Perspectives. — Nous avons observé que l’âge du club est un déterminant important pour expliquer les taux de survie plus élevés des équipes appartenant à des clubs plus âgés. Plus de nuance a été découverte lorsque les équipes ont été analysées par chaque ligue professionnelle. Parmi les équipes franc ¸aises, le taux de survie au sommet est affecté positivement pour avoir des pourcentages plus élevés de joueurs nationaux, la hausse des entrées de transfert, la participation à des compétitions de l’UEFA et une meilleure fréquentation du stade. La fréquentation du stade est un facteur important dans le maintien de l’équipe au sommet, indépendant de la ligue européenne (à l’exception de la compétition franc ¸aise). La survie des équipes franc ¸aises tend à bénéficier de l’acquisition de joueurs (représentée par la hausse des sorties de transfert). Conclusion. — Le plus robuste des résultats indique l’effet de l’âge est un déterminant statistiquement significatif. Les clubs plus âgés bénéficient de quatre dimensions liées à l’âge : la pertinence de l’actif, les processus de prise de décision plus efficaces, un pouvoir de marché monopolistique et un plus grand contrôle des décisions par le conseil exécutif. eserv´ es. © 2017 Elsevier Masson SAS. Tous droits r´
1. Introduction During the 2011/2012 season, Udinese, an Italian soccer team, reached the leading position of the competitive Serie A in the 14th match. Because the Italian league hosts 38 matches, this achievement was significant for a team that did not belong to one of the exclusive Big Ones of Italian calcio (Juventus, which won that season’s title, Inter Milan and A.C. Milan). Udinese was founded in 1896 and has a mean stadium attendance of 17,000 (approximately 50% of the stadium’s capacity). Therefore, reaching first place in the 14th match of the 2011/2012 season was one of the greatest achievements that Udinese’s fans had experienced in a decade. Similar achievements were accomplished by teams such as Auxerre (first place at the 29th match of the French League in the 2009/2010 season), TSG 1899 Hoffenheim (first place at the 21st match of the German League in the 2008/2009 season), and Leixoes (first place at the 10th match of the Portuguese League in the 2008/2009 season). These teams’ budgets are much lower than those of the usual champions of their national leagues. Nevertheless, the supporters of Auxerre (founded at 1905), Hoffenheim (founded at 1899) and Leixoes (founded at 1907) had the opportunity to celebrate as their teams overtook first place at matches
played in the middle or at the end of the season. Although these teams did not succeed in keeping first place until the season’s end, they reached the top of the standings, which is rare for teams with similar histories and endowments. Why is it so difficult for most soccer teams to maintain first place? After the first third of the season, the top 3 ranks are typically held exclusively by the ‘‘Big Ones’’, whose teams are difficult to dislodge. However, if there is no written rule forbidding the occupation of first place by a ‘‘not-Big’’ team, there must be a robust reason explaining the low probability that a team with a small stadium, a modest budget and a humble history will be able to maintain first place. The search for this explanation is the purpose of this study. The literature review focuses on one major explanation —– the age hypothesis. Older clubs are more likely to maintain the top places because of factors such as the value of their assets, their organizational structure, their monopolistic power and the number of generations of fans and supporters. Several dimensions control the age hypothesis:
• the transfer cycle; • the presence of the team at UEFA competitions; • stadium attendance.
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
Age of a club and ability to maintain the top position in soccer leagues These dimensions will also be considered as determinants of the survival rate of a soccer team in first place. The remainder of this paper is structured as follows: • Section 2 presents a review of the literature, introducing the major lines of research on the topic and the variables that will be used in the empirical analyses; • Section 3 describes our methodology, including the Cox proportional hazards regression model, the variables of our database, the major results of the analyses and the main implications of the findings; • Section 4 concludes this work.
2. Literature review Most athletes desire to reach first place during a sports competition. However, the real achievement is to still be in first place at the end of the competition. However, to achieve this goal, the competing agents (individual players or teams) use strategies during the competition that, depending on a varied set of factors, can help them achieve first place as soon as possible or only near the end of the competition. This is true for soccer teams competing in their national leagues. During a typical season, various teams may spend some time in first place; however, only one will be the national champion. Usually, during the initial matches, some of the less valuable teams have a greater chance of being in the top positions; in the middle of the competition, only a limited number of the teams with the largest budgets have the opportunity of being at the top [1]. Several authors have studied how budget differences contribute to a soccer team’s success in achieving the best standings. Hone [2] argues that high-budget teams have players that are better prepared to manage the pressure of being in the top position and who are also the most competitive and the most talented. This assumption is also the basis of Mourao’s study [3]. Mourao [4] and Peeters and Szymanski [5] found that the values related to the inflows and to the outflows of player transfers were correlated with the budget values of each soccer team. The values of players’ transfers can also act as proxies to signal the strategy used by that team’s managers, whether to renew the team (increasing the players’ hires, which generates significant outflows when compared to the exit of new players) or to maintain the team’s status quo by only selling a limited number of players (increasing the inflows over the outflows).
2.1. Advantages and disadvantages of being an old club However, this paper focuses on testing the age hypothesis. According to the age hypothesis, older clubs (not older squads) with more robust organizations, more significant assets, and more fans are more likely to maintain the top positions. The advantages (and the disadvantages) of older firms or corporations described in the literature can be adapted to the case of older clubs. Clubs accumulate intangible values over time; therefore, older clubs have an accumulated value created by multiple generations of supporters, fans and investors that is more difficult to compute than the values presented on the balance sheets.
3
As with older organizations, the characteristics of the older clubs can be divided into four dimensions: asset accumulation [6,7], clear definition of decision-making processes [6,8], rigidity of the organizational structure [9], and monopolistic market power [10,11]. As Pfaffermayr et al. [7] argue, older organizations (clubs, firms, or companies) tend to have more diversified assets and more valuable assets than do younger organizations. The reasons are various, from the desire to develop assets such as real estate or tangible investments over time [6] to the need to expand the radius of action by increasing the assets’ value at certain critical moments of the organization’s growth. Older organizations also tend to better define the process of decision-making. In the case of sports clubs, it is most efficient for decisions to be made by the president, the executive office or the team of managers [6]. This diminishes the length of time needed to reach a decision and prevents misinterpretation on the internal rules. However, though the decision process is more efficient, over time older organizations begin to exhibit more rigid organizational structures. It becomes difficult to integrate new functions, new departments or new internal structures, as Loderer and Waelchli [9] reported. This can be a serious handicap in times of severe exogenous challenges. Finally, older organizations are usually associated with monopolistic market power. In the sports context, older clubs tend to benefit from dominant positions when trading players, changing competition rules, collecting revenues and hiring promising young players [10,11].
3. Empirical section—the Cox proportional hazards regression, data, sources and results 3.1. The Cox proportional hazards regression—the rationale behind our choice To analyze the survival rate of a professional soccer team in the first position of its national league, it is important to choose a suitable empirical method. We adopted the Cox proportional hazards regression. The Cox regression has been applied in various fields (for an extensive discussion, see, among others, Klein and Moeschberger [12]). This type of regression allows researchers to determine which characteristics of the subject (in this paper, the characteristics of each soccer team) influence the survival of the subject in some risky situation (in this case, the maintenance of a first-place position in the national league). As Danacica and Babucea [13] demonstrate, the Cox regression also allows researchers to estimate the statistical significance of this influence and the direction (i.e., if the analyzed characteristic increases or decreases the individual’s ability to survive). The simplest version of the Cox proportional hazards regression is shown in equation (1):
i (t) = exi ˇ ∗ 0 (t) , i = 1, 2, . . ., n
(1)
Equation (1) shows that, with n units being observed, the hazard rate (the probability of maintaining the first position/ascending to a top position after t periods, i.e., i (t))
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
4
P.R. Mourao
depends on the vector of the considered variables (x), the vector of the regression coefficients [ˇ = (ˇ1, ˇ2 , . . ., ˇk )], and on the baseline hazard (0 (t) [i.e., the probability of maintaining the first position/exiting a lower position by ascending to the top when all of the explanatory variables are equal to 0]). The regression coefficients can be interpreted as expressing the relationship between the proportional change that is expected in the hazard and the changes in the explanatory variables. They are usually estimated using maximum likelihood. In this model, the baseline hazard does not depend on the explanatory variables; it depends only on t. Furthermore, the exponential component comprises the explanatory variables but not the number of time periods. Finally, the explanatory variables are time-independent (typically called the ‘‘proportional hazard assumption’’). There are several advantages to using the Cox proportional hazards regression. It does not require the probability distribution of status changes over time to have a known parametric form. It also provides valuable information about cases of transitions not studied in previous literature. However, a strong assumption must hold: the proportional hazards assumption states that the difference in log hazard associated with a change in each covariate is not timedependent. The tests suggested by Grambsch and Therneau [14] and by Hosmer and Lemeshow [15] were performed on our data, and their results indicated that this assumption held (full details are available by request).
3.2. Data and sources To study the top-position survival rates of European soccer teams, we developed a database that included the explanatory variables suggested in the literature. We collected all of the data related to the 185 teams (and their clubs) that played in the six professional soccer leagues under study1 . We determined the number of matches in which each of these teams remained in the first position for each season (to construct the variable related to the survival time at the top). Additionally, we also computed the age of each club on the 31st of December of each observed season using the information available at each club’s official websites.
3.3. Controlling the age effect However, other variables must be considered to control the age effect (our main explanatory variable). Following the literature, we have to discuss these variables: • the percentage of national players: Several authors (e.g., Kleven et al. [16]) argue that a higher percentage of national players better represents the national conditions of a league. Because most of the national players have played for other clubs, these players exhibit a more profound knowledge of the national teams’ competitors. Additionally, the national players, having been educated
1
English, French, German, Italian, Portuguese, and Spanish professional soccer leagues.
in the country’s youth teams, tend to better understand the weather conditions of the country and the competition’s characteristics. Based on Marquez and Martin [17], we expect this variable to control the age of the club, although there are claims that this variable has a positive effect on the survival rate. Other authors (Kleven et al. [16]), however, argue for a negative effect; • the presence of the team at UEFA competitions: other authors (Marquez and Martin [17]) argue that the presence of a team at international competitions (such as the UEFA Champions League or the UEFA Europa League) influences the quality of that team’s matches at national competitions. Heron and Cupples [18], for example, comment that a team’s presence at international competitions exerts additional pressure on the players and can be physically taxing, diminishing the probability that the team will survive at the top. As suggested by Marquez and Martin [17] and Heron and Cupples [18], this dimension can influence the performance of a team in the national leagues because of the additional physical effort required and the prestige effect; • the relationship between transfer inflows and outflows: according to Mourao [4] and Peeters and Szymanski [5], these dimensions are related to a team’s budget and to the strategic renewal of the squad; • stadium attendance: there is a significant body of literature on the determinants of stadium attendance [19]. One of the correlated dimensions is the team’s success record —– the most successful teams tend to receive higher attendance. However, there is an open debate about the causality of this relation. Some authors [20,21] question whether higher attendance leads to more impressive victories and to a more consistent incidence of victory. The authors [20,21] suggest that these values can stimulate a professional soccer team to maintain the top position. Therefore, for these controls, we collected data from http://www.transfermarkt.com/for the following variables (for each season): ‘‘percentage of national players’’, ‘‘transfers inflows’’, ‘‘transfers outflows’’, ‘‘presence at UEFA Champions League at the beginning of the season’’, ‘‘presence at UEFA Europa League at the beginning of the season’’, and ‘‘stadium attendance’’. Tables 1—7 show the descriptive statistics of our database. Table 1 displays the statistics for the whole sample. The mean age of the European professional soccer teams in our sample is 92 years (ranging from 40 to 161 years). The mean number of matches that a professional soccer team retained the top position is 3.8. There were several cases in which teams maintained the top position the entire season (justifying the maximum value of 38). The mean values for the percentage of national players and for the log ratio between transfer inflows and transfer outflows (used for minimizing heteroscedasticity concerns) are, respectively, 53.8% and 0.31. Most of the soccer teams in our sample do not have the opportunity to play in UEFA competitions, resulting in mean values of 0.156 (Champions League) and 0.134 (Europa League) for this variable. Tables 1—7 present the values of our sample by national league. Table 8 displays the main results after estimating equation (1) using the appropriate proportional hazards
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
Age of a club and ability to maintain the top position in soccer leagues Table 1
5
Descriptive statistics (whole sample).
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of Journeys at the first place Age of the sports club % of national players Transfers inflows over transfers outflows (log) Champions League Europa League Stadium attendances (log)
3.826 92.449 53.824 0.311 0.156 0.134 9.937
9.639 44.798 15.121 1.588 0.364 0.341 0.830
0 40 14.3 −5.257 0 0 7.107
38 161 97.2 17.783 1 1 11.296
798 798 798 798 798 798 798
Table 2
Descriptive statistics (England).
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of Journeys at the first place Age of the sports club % of national players Transfers inflows over transfers outflows (log) Champions League Europa League Stadium attendances (log)
3.811 119.0 40.668 0.853 0.171 0.143 10.384
9.863 16.302 11.905 1.306 0.378 0.351 0.375
0 102 14.3 −1.632 0 0 9.666
38 153 56.7 5.013 1 1 11.234
140 140 140 140 140 140 140
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of Journeys at the first place Age of the sports club % of national players Transfers Inflows over Transfers Outflows (log) Champions League Europa League Stadium attendances (log)
4.378 111 58.791 −0.017 0.150 0.128 9.801
10.577 9.319 9.849 1.146 0.358 0.336 0.494
0 98 36.8 −3.401 0 0 8.754
38 150 80 4.784 1 1 11.236
140 140 140 140 140 140 140
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of Journeys at the first place Age of the sports club % of national players Transfers inflows over transfers outflows (log) Champions League Europa League Stadium attendances (log)
4.992 104.2 50.503 0.686 0.158 0.103 10.553
10.217 22.01 9.944 1.428 0.366 0.305 0.397
0 70 31.4 −3.076 0 0 9.695
38 161 93.8 6.386 1 1 11.296
126 126 126 126 126 126 126
Table 3
Table 4
Descriptive statistics (France).
Descriptive statistics (Germany).
method — the Cox regression. The event of interest was a professional soccer team maintaining first place in the national league. This model was applied to all of the professional soccer teams from the following six male leagues: the English Premier League, the French Ligue 1, the
German Bundesliga, the Italian Serie A, the Portuguese Super Liga, and the Spanish La Liga. Data were collected from the 2007/2008 season through the 2013/2014 season (because of data availability). Therefore, Table 8 presents the relative rates of the risk of being in the first position (the
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
6
P.R. Mourao Table 5
Descriptive statistics (Italy).
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of Journeys at the first place Age of the sports club % of national players Transfers inflows over transfers Outflows (log) Champions League Europa League Stadium attendances (log)
3.178 111.97 59.975 0.396 0.135 0.100 9.903
9.242 14.389 14.945 1.903 0.343 0.301 0.545
0 100 22.6 −2.988 0 0 8.675
38 154 91.4 17.781 1 1 10.998
140 140 140 140 140 140 140
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of journeys at the first place Age of the sports club % of national players Transfers inflows over transfers outflows (log) Champions League Europa League Stadium attendances (log)
3.571 92.785 45.269 −1.088 0.161 0.241 8.685
8.758 19.267 12.487 1.632 0.369 0.429 1.099
0 40 20 −5.257 0 0 7.107
33 125 77.4 5.011 1 1 11.218
112 112 112 112 112 112 112
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of journeys at the first place Age of the sports club % of national players Transfers inflows over transfers outflows (log) Champions League Europa League Stadium attendances (log)
3.072 89.071 65.761 0.089 0.171 0.107 10.107
8.934 24.565 13.591 1.570 0.378 0.310 0.548
0 75 40 −3.178 0 0 8.992
38 120 97.2 4.381 1 1 11.279
140 140 140 140 140 140 140
Table 6
Table 7
Descriptive statistics (Portugal).
Descriptive statistics (Spain).
estimated coefficients of the exponent, ˇ) derived from the Cox regressions. In accordance with the literature, the following explanatory variables were used: the age of the club on the 31st of December of each observed year, the percentage of national players, the ratio of transfer inflows to transfer outflows, the presence of the team at the UEFA Champions League or the UEFA Europa League (at the beginning of the season), and the value of each club’s stadium attendance at official matches of the professional male soccer team. We expected that older clubs would be characterized by higher rates of survival at the first position of the national soccer league. The literature also suggested that the presence of the teams at UEFA competitions and stadium attendance would positively increase those rates of survival. Therefore, we also controlled the age effect using these dimensions and the remaining variables discussed in the literature review (although the transfer cycle and the percentage of national players are not clearly associated with positive
effects on a team’s ability to remain in first place for a long time).
4. Results 4.1. An overall perspective Table 8 (columns for ‘‘Whole Sample’’) shows that the likelihood of a European soccer team surviving at the top is highest among the oldest teams. Another significant coefficient relates to the ratio of inflow from player transfers to the outflow from hires, whose increase tends to reduce a team’s ability to be at the top at the final match. Therefore, the ratio of inflow to outflow tends to be smaller for successful teams than for those that experienced a greater need to sell players, which introduced the need to renew the squad. As expected, a higher mean stadium attendance tends to provide additional resources for any European soccer team
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model
ARTICLE IN PRESS
SCISPO-3080; No. of Pages 9
Age of a club and ability to maintain the top position in soccer leagues Table 8
7
Cox regression analysis.
Variables Age of the sports club % of national players Transfers Inflows over Transfers Outflows (log) Champions League
Whole sample ***
1.028 (0.006) 0.967 (0.327) 0.967*** (0.007)
0.824 (0.273) 0.713 Europa League (0.351) Stadium attendances 1.193*** (0.092) 185 Obs. teams 1st-placed teams a 79 R2 0.187 LogL −147.9 Variables Age of the sports club % of national players Transfers Inflows over Transfers Outflows (log) Champions League
England ***
1.028 (0.005) 0.967 (0.317) 0.884* (0.066)
0.715 (0.351) 1.198*** (0.078) 185 79 0.181 −132.2
***
1.028 (0.005) 0.967 (0.397) 0.890 (0.070)
1.093 (0.013) 1.059*** (0.008) 1.264* (0.174)
1.195*** (0.074) 185 79 0.186 −138.0
1.331*** (0.552) 0.344 (0.221) 1.012*** (0.001) 47 14 0.473 −13.651
Germany 1.349 (0.429) 0.923 (0.135) 1.546*** (0.638)
0.992 (0.152) 1.002 Europa League (0.182) Stadium attendances 1.805*** (0.271) 27 Obs. Teams 1st-placed Teams a 17 R2 0.261 LogL −13.621
***
1.354 (0.418) 0.914 (0.145) 1.558*** (0.628)
1.092*** (0.026) 0.810 (0.188) 0.743*** (0.051)
1.093*** (0.026) 0.811 (0.189) 0.742*** (0.050)
1.013*** (0.001) 47 14 0.451 −14.561
1.144*** (0.071) 0.341 (0.501) 0.057 (0.043) 34 14 0.199 −15.421
0.045 (0.081) 34 14 0.201 −15.412
0.088 (0.098) 34 14 0.213 −15.422
1.103*** (0.040) 0.942** (0.028) 0.407*** (0.101)
0.972 (0.009) 0.901 (0.712) 0.992 (0.861)
0.912 (0.009) 0.899 (0.713) 0.991 (0.863)
0.911 (0.009) 0.898 (0.714) 0.990 (0.868)
1.102*** (0.039) 0.936*** (0.016) 0.466 (0.338)
1.101*** (0.041) 0.943** (0.027) 0.406*** (0.103)
1.701*** (0.238)
1.702*** (0.239)
1.923*** (0.281) 30 12 0.104 −12.623
1.924*** (0.282) 30 12 0.105 −12.624
0.564 (0.236) 0.721 (0.853) 4.807 (16.20) 20 9 0.122 −15.031
0.660 (0.248)
1.818*** (0.251) 27 17 0.263 −13.712
1.702*** (0.241) 1.001 (0.261) 1.923*** (0.281) 30 12 0.102 −12.615
0.991 (0.215)
1.812*** (0.241) 27 17 0.269 −13.671
0.997 (0.009) 1.141*** (0.045) 2.576*** (0.452) 1.002 (0.005) 0.997 (0.010) 1.004*** (0.001) 27 13 0.141 −11.132
% of national players Transfers inflows over transfers outflows (log)
Obs. teams 1st-placed teamsa R2 LogL
1.094*** (0.027) 0.861 (0.169) 0.749*** (0.062)
1.388 (0.402) 0.901 (0.121) 1.581*** (0.611)
Age of the sports club
Stadium attendances
0.332 (0.221) 1.011*** (0.001) 47 14 0.412 −12.521
1.093 (0.012) 1.058*** (0.008) 1.263 (0.221)
Portugal ***
Spain
Europa League
1.093 (0.013) 1.059*** (0.008) 1.264* (0.174)
***
Italy
Variables
Champions League
France ***
***
***
10.159* (12.07) 20 9 0.118 −14.919
0.997 (0.008) 1.142*** (0.044) 2.571*** (0.402)
0.996 (0.005) 1.141*** (0.043) 2.572*** (0.401)
0.996 (0.011) 1.004*** (0.001) 27 13 0.142 −11.133
1.005*** (0.001) 27 13 0.143 −11.134
10.160* (12.069) 20 9 0.117 −14.821
Values are hazard rates of exiting lower ranks by ascending to the first rank. Robust standard errors between parentheses. a Number of different teams ranked at the 1st place of the league since the first journey of the season 2007/2008 to the last journey of the season 2013/2014. * Significance level P < 0.1. ** Significance level P < 0.05. *** Significance level P < 0.01.
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
8
P.R. Mourao
to maintain the top classification in its national league. Our results (‘‘Whole sample’’ column of Table 8 suggest that a team exhibiting 1% more attendance during the season than a competing team is 1.2 times more likely to survive at the highest position than the other (less attended) team. Actually, as Table 8 shows and depending on the presence of additional covariates, this estimated coefficient ranges between 1.193 and 1.198. In the sample of 185 European teams, the percentage of national (or foreign players) did not significantly affect survival. The same observation was made of the variables related to the presence of the team at the Champions League or the Europa League. Overall, therefore, the presence of a European team at the UEFA does not particularly stimulate or burden the squads, whose rates of survival in their country’s championships are more influenced by the organizational structure (a factor related to the age of the club), to transfers, or to attendance at the home stadiums.
4.2. National cases Among the English teams, the age of the club tends to increase a team’s ability to survive at the top, with a factor (1.093) higher than the estimated coefficient for the whole sample (1.028). The stadium attendance also positively contributed to the ability to manage the pressure of being in the top position among the English teams. However, Table 8 reveals two interesting issues. The first relates to the positive estimated coefficients of the ratio of transfer inflows to transfer outflows, suggesting that, among English teams, more frequent rotation of the players promotes the ability to remain in the top positions. The second issue relates to the positive influence of having a larger proportion of English players on a team’s survival rate at the top. This effect may be due to the better assimilation of the managers’ instructions or to the transfer of highly profitable players from foreign origins. Finally, the presence of an English team at the Champions League tends to provide a stimulus for keeping the team at the highest position of the Premier League. Therefore, for English teams, the Champions League increases a team’s stamina, which compensates for the associated costs (e.g., players’ fatigue and the weight of mediocre international results). The age of the club also positively affects French teams. As with English teams, the French teams also gain additional stimulus from playing in the Champions League. However, the survival rates at the top significantly drop among teams that sell more players than they hire, indicating an interpretation in the opposite direction of that made for the English teams and also showing that the transfer variable had the opposite effect. The German, Italian and Spanish teams seem are not significantly influenced by the age of the club when fighting to maintain first place. More significant for the German and Spanish teams are the effects exerted by the value of transfer inflows and higher stadium attendance (also important for Italian teams). The age of the club also influenced Portuguese teams’ ability to remain in the top position. These teams were also significantly influenced by stadium attendance.
4.3. Synthesis and discussion of the results Obtaining the first position in its league is always a positive experience for any soccer team. However, the ability to maintain that top position is subject to a variety of factors. As our review of literature revealed, these factors include the strength of the club’s organizational structure [22] and the value invested in the team (usually positively correlated with the age of the club), the management of the squad’s players (related to the transfer cycle) and the external signs of confidence manifested in stadium attendance. In this paper, we focused on the age of the club. Older clubs survive because, generation after generation, their directors and managers have demonstrated the capacity to renew, usually by modernizing the buildings owned by the club and by enlarging the number of stakeholders, supporters, investors and fans. From a financial perspective, older clubs are expected to have more valuable assets; from a sports competitiveness perspective, older clubs are expected to exhibit more knowledge about trading players and coaches, hiring managers and managing publicity or the revenues from TV rights. As a result, older clubs tend to exhibit stronger structures that contribute to their teams’ ability to maintain the top positions in a sports league. Clubs without these strong structures are expected to encounter more difficulty in maintaining the top position after reaching first place, even if the teams of these clubs have the opportunity of participating at UEFA competitions, as alternatively Heron and Cupples [18] discussed. Our results confirm this hypothesis for European soccer teams, especially for English, French, and Portuguese teams. In their respective professional leagues, older clubs tend to survive longer in the top position than younger clubs. Even when controlling for the other two dimensions (‘‘management of the squad’’ and ‘‘external signs of confidence’’), we found that the age hypothesis holds for the professional soccer teams of England, France, and Portugal, which renew the works of authors like Peeters and Szymanski [5], Poli et al. [22], Scelles et al. [20] or Mourao [3]. For teams from Germany, Italy and Spain, this effect was not as significant. In these cases, the management of the squad and the external signs exerted more influence. Mourao [4] explains that some younger clubs have huge financial assets because they have received large investments from particular individuals (mostly from emerging economies) and can therefore compete with the older clubs for the topranked positions.
5. Conclusions and further work This study examined the ability of European soccer teams to remain in first place in their respective national leagues. We studied 185 teams from the 2006/2007 season to the 2012/2013 season. Using Cox proportional hazards regression models, we concluded that the age of the club is an important determinant in explaining the higher survival rates of the teams belonging to older clubs. However, we also controlled for the age effect with variables suggested by the literature. These variables included transfer flows, the presence of the team at UEFA competitions (Champions League or
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
Age of a club and ability to maintain the top position in soccer leagues Europa League), stadium attendance, and the percentage of national players. The most robust results indicate that the age effect is a statistically significant determinant. Older clubs tend to benefit from four age-related dimensions: the relevance of assets, more efficient decision-making processes, monopolistic market power and the control of decisions by the executive board. More nuances were discovered when the teams were analyzed by professional league. Among English teams, the survival rate at the top is positively affected by having higher percentages of national players, higher transfer inflows, participation in UEFA competitions and higher stadium attendance. Stadium attendance is an important factor in keeping a team at the top, independent of the European league (with the exception of the French competition). French teams’ survival tends to benefit from the acquisition of players (proxied by higher transfer outflows). Although significant, the age effect nevertheless does not prevent young clubs (those founded within the last 50 years) from achieving first place. Young clubs that manage transfers well, have impressive stadium attendance, and are stimulated by UEFA participation tend to experience additional positive effects on their survival rate at the top. Based on this study’s findings, we identify three opportunities for future research. The first opportunity relates to the extension of this database to national leagues not included in this study. The second opportunity is the inclusion of greater detail in the four dimensions used to study the age effect (asset structure, organizational process, monopolistic power, and control of the decisions). We intend to compile convenient proxies for analyzing these four dimensions. Finally, the third opportunity is to broaden our interpretation of the ‘‘top’’ (here, only the first position) to consider additional top positions, for example, the positions higher or equal than 3rd place.
Disclosure of interest The author declares that he has no competing interest.
References [1] Davis M, End C. Team Success, Productivity and Economic Impact. In: The Economics of Sport, Health and Happiness, chapter 7. Edward Elgar; 2011. [2] Hone P. ‘‘Limiting Player Lists In Sport: Who Really Wins?’’ Economics Series 2005 03. Deakin University, Faculty of Business and Law, School of Accounting, Economics and Finance; 2005. [3] Mourao P. Regional determinants of competitiveness: the case of European Soccer teams. Int J Sports Finance 2010;5:222—34. [4] Mourao P. The indebtedness of Portuguese soccer teams—looking for determinants. J Sports Science 2012;30(10):1025—35.
9
[5] Peeters T, Szymanski S. ‘‘Vertical restraints in soccer: Financial fair play and the English Premier League’’ Working Papers 2012028. University of Antwerp, Faculty of Applied Economics; 2012. [6] Dinopoulos E, Waldo D. Gradual product replacement, intangible-asset prices and Schumpeterian growth. J Econ Growth Springer 2005;10(2), 135-157, 06. [7] Pfaffermayr M, Stöckl M, Winner H. Capital structure, corporate taxation and firm age. Fiscal Stud Inst 2013;34(1) [109—135,03]. [8] Phua F. The antecedents of co-operative behaviour among project team members: an alternative perspective on an old issue. Constr Manage Econ Taylor Francis J 2004;22(10):1033—45. [9] Loderer C, Waelchli U. Firm age and performance, MPRA Paper 26450. Germany: University Library of Munich; 2010. [10] Shimomura K, Thisse J. ‘‘Competition Among the Big and the Small’’ Discussion Paper Series DP2012-03. Research Institute for Economics & Business Administration, Kobe University; 2012. [11] Huang Y, Huang D. Big vs. small under free trade: market size and size distribution of firms. Int Rev Econ Finance Elsevier 2014;34(C):175—89. [12] Klein JP, Moeschberger ML. Survival analysis: techniques for censored and truncated data. New York: Springer Verlag; 2005. [13] Danacica D, Babucea A. Using survival analysis and economics. scientific annals of the Alexandru loan. Cuza University of Labsi: Economic Sciences Series, No. 1; 2010. p. 2010. [14] Grambsch P, Therneau T. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika 1994;81:515—26. [15] Hosmer D, Lemeshow S. Applied logistic regression. New York: Wiley; 2000. [16] Kleven H, Landais C, Saez E. Taxation and International migration of superstars: evidence from the European football market. Am Econ Rev 2013;103(5):1892—924. [17] Marquez M, Martin H. The New Football Business: a challenge for elite followers. Goteborg University; 2000. p. 18 [International Business Master Thesis]. [18] Heron N, Cupple M. The health profile of football/soccer players in Northern Ireland—a review of the UEFA pre-participation medical screening procedure. BMC Sports Sci Med Rehab 2014;6(5). [19] Buraimo B. Stadium attendance and television audience demand in English league football. Manage Decis Econ 2008;29(6):513—23 [John Wiley & Sons, Ltd]. [20] Scelles N, Durand C, Bonnal L, Goyeau D, Andreff W. My team is in contention? Nice, I go to the stadium! Competitive intensity in the French football Ligue 1. Econ Bull Access Econ 2013;33(3):2365—78. [21] Ponzo M, Scoppa V. Does the home advantage depend on crowd support? Evidence from same-stadium derbies. Università della Calabria, Dipartimento di Economia, Statistica e Finanza (Ex Dipartimento di Economia e Statistica); 2014. Working Papers 201402. [22] Poli R, Ravenel L, Besson R. Squad analysis: who will win the Champions League? CIES Football Observatory Monthly Report; 2015.
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
ARTICLE IN PRESS
SCISPO-3080; No. of Pages 9 Science & Sports (2017) xxx, xxx—xxx
Disponible en ligne sur
ScienceDirect www.sciencedirect.com
ORIGINAL ARTICLE
Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues Comment l’âge du club contribue-t-il à maintenir dans les ligues de football européennes la première position de l’équipe P.R. Mourao Department of Economics, Economics & Management School, University of Minho, 4700 Braga, Portugal Received 25 August 2015; accepted 21 January 2017
KEYWORDS Soccer; Survival; Cox proportional hazard rates
Summary Objectives. — To test the hypothesis that the age of a club is a major determinant of the higher survival rates exhibited by European soccer teams. News. — We studied 185 teams from the 2006/2007 season to the 2012/2013 season. We used Cox proportional hazards regression models. We also controlled for the age effect with variables suggested by the literature. These variables included transfer flows, the presence of the team at UEFA competitions (Champions League or Europa League), stadium attendance, and the percentage of national players. Prospect. — We observed that the age of the club is an important determinant in explaining the higher survival rates of the teams belonging to older clubs. More nuance was discovered when the teams were analyzed by professional league. Among English teams, the survival rate at the top is positively affected by having higher percentages of national players, higher transfer inflows, participation in UEFA competitions and higher stadium attendance. Stadium attendance is an important factor in keeping a team at the top, independent of the European league (with the exception of the French competition). French teams’ survival tends to benefit from the acquisition of players (proxied by higher transfer outflows). Although significant, the age effect nevertheless does not prevent young clubs (those founded within the last 50 years)
E-mail addresses: [email protected], paulo r [email protected]. http://dx.doi.org/10.1016/j.scispo.2017.01.008 0765-1597/© 2017 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
2
P.R. Mourao from achieving first place. Young clubs that manage transfers well, have impressive stadium attendance, and are stimulated by UEFA participation tend to experience additional positive effects on their survival rate at the top. Conclusion. — The most robust results indicate that the age effect is a statistically significant determinant. Older clubs tend to benefit from four age-related dimensions: the relevance of assets, more efficient decision-making processes, monopolistic market power and the control of decisions by the executive board. © 2017 Elsevier Masson SAS. All rights reserved.
MOTS CLÉS Football ; Survie ; Taux de risque proportionnel
Résumé Objectifs. — Tester l’hypothèse que l’âge du club est un déterminant des taux de survie plus élevés des équipes européennes de football. Actualités. — Nous avons étudié 185 équipes de la saison 2006/2007 à la saison 2012/2013. Nous avons utilisé des modèles de régression des risques de Cox. Nous avons également contrôlé pour l’effet de l’âge avec les variables suggérées par la littérature. Ces variables comprenaient les flux de transfert, la présence de l’équipe lors des compétitions de l’UEFA (Ligue des Champions ou Europa League), la fréquentation des stades, et le pourcentage de joueurs nationaux. Perspectives. — Nous avons observé que l’âge du club est un déterminant important pour expliquer les taux de survie plus élevés des équipes appartenant à des clubs plus âgés. Plus de nuance a été découverte lorsque les équipes ont été analysées par chaque ligue professionnelle. Parmi les équipes franc ¸aises, le taux de survie au sommet est affecté positivement pour avoir des pourcentages plus élevés de joueurs nationaux, la hausse des entrées de transfert, la participation à des compétitions de l’UEFA et une meilleure fréquentation du stade. La fréquentation du stade est un facteur important dans le maintien de l’équipe au sommet, indépendant de la ligue européenne (à l’exception de la compétition franc ¸aise). La survie des équipes franc ¸aises tend à bénéficier de l’acquisition de joueurs (représentée par la hausse des sorties de transfert). Conclusion. — Le plus robuste des résultats indique l’effet de l’âge est un déterminant statistiquement significatif. Les clubs plus âgés bénéficient de quatre dimensions liées à l’âge : la pertinence de l’actif, les processus de prise de décision plus efficaces, un pouvoir de marché monopolistique et un plus grand contrôle des décisions par le conseil exécutif. eserv´ es. © 2017 Elsevier Masson SAS. Tous droits r´
1. Introduction During the 2011/2012 season, Udinese, an Italian soccer team, reached the leading position of the competitive Serie A in the 14th match. Because the Italian league hosts 38 matches, this achievement was significant for a team that did not belong to one of the exclusive Big Ones of Italian calcio (Juventus, which won that season’s title, Inter Milan and A.C. Milan). Udinese was founded in 1896 and has a mean stadium attendance of 17,000 (approximately 50% of the stadium’s capacity). Therefore, reaching first place in the 14th match of the 2011/2012 season was one of the greatest achievements that Udinese’s fans had experienced in a decade. Similar achievements were accomplished by teams such as Auxerre (first place at the 29th match of the French League in the 2009/2010 season), TSG 1899 Hoffenheim (first place at the 21st match of the German League in the 2008/2009 season), and Leixoes (first place at the 10th match of the Portuguese League in the 2008/2009 season). These teams’ budgets are much lower than those of the usual champions of their national leagues. Nevertheless, the supporters of Auxerre (founded at 1905), Hoffenheim (founded at 1899) and Leixoes (founded at 1907) had the opportunity to celebrate as their teams overtook first place at matches
played in the middle or at the end of the season. Although these teams did not succeed in keeping first place until the season’s end, they reached the top of the standings, which is rare for teams with similar histories and endowments. Why is it so difficult for most soccer teams to maintain first place? After the first third of the season, the top 3 ranks are typically held exclusively by the ‘‘Big Ones’’, whose teams are difficult to dislodge. However, if there is no written rule forbidding the occupation of first place by a ‘‘not-Big’’ team, there must be a robust reason explaining the low probability that a team with a small stadium, a modest budget and a humble history will be able to maintain first place. The search for this explanation is the purpose of this study. The literature review focuses on one major explanation —– the age hypothesis. Older clubs are more likely to maintain the top places because of factors such as the value of their assets, their organizational structure, their monopolistic power and the number of generations of fans and supporters. Several dimensions control the age hypothesis:
• the transfer cycle; • the presence of the team at UEFA competitions; • stadium attendance.
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
Age of a club and ability to maintain the top position in soccer leagues These dimensions will also be considered as determinants of the survival rate of a soccer team in first place. The remainder of this paper is structured as follows: • Section 2 presents a review of the literature, introducing the major lines of research on the topic and the variables that will be used in the empirical analyses; • Section 3 describes our methodology, including the Cox proportional hazards regression model, the variables of our database, the major results of the analyses and the main implications of the findings; • Section 4 concludes this work.
2. Literature review Most athletes desire to reach first place during a sports competition. However, the real achievement is to still be in first place at the end of the competition. However, to achieve this goal, the competing agents (individual players or teams) use strategies during the competition that, depending on a varied set of factors, can help them achieve first place as soon as possible or only near the end of the competition. This is true for soccer teams competing in their national leagues. During a typical season, various teams may spend some time in first place; however, only one will be the national champion. Usually, during the initial matches, some of the less valuable teams have a greater chance of being in the top positions; in the middle of the competition, only a limited number of the teams with the largest budgets have the opportunity of being at the top [1]. Several authors have studied how budget differences contribute to a soccer team’s success in achieving the best standings. Hone [2] argues that high-budget teams have players that are better prepared to manage the pressure of being in the top position and who are also the most competitive and the most talented. This assumption is also the basis of Mourao’s study [3]. Mourao [4] and Peeters and Szymanski [5] found that the values related to the inflows and to the outflows of player transfers were correlated with the budget values of each soccer team. The values of players’ transfers can also act as proxies to signal the strategy used by that team’s managers, whether to renew the team (increasing the players’ hires, which generates significant outflows when compared to the exit of new players) or to maintain the team’s status quo by only selling a limited number of players (increasing the inflows over the outflows).
2.1. Advantages and disadvantages of being an old club However, this paper focuses on testing the age hypothesis. According to the age hypothesis, older clubs (not older squads) with more robust organizations, more significant assets, and more fans are more likely to maintain the top positions. The advantages (and the disadvantages) of older firms or corporations described in the literature can be adapted to the case of older clubs. Clubs accumulate intangible values over time; therefore, older clubs have an accumulated value created by multiple generations of supporters, fans and investors that is more difficult to compute than the values presented on the balance sheets.
3
As with older organizations, the characteristics of the older clubs can be divided into four dimensions: asset accumulation [6,7], clear definition of decision-making processes [6,8], rigidity of the organizational structure [9], and monopolistic market power [10,11]. As Pfaffermayr et al. [7] argue, older organizations (clubs, firms, or companies) tend to have more diversified assets and more valuable assets than do younger organizations. The reasons are various, from the desire to develop assets such as real estate or tangible investments over time [6] to the need to expand the radius of action by increasing the assets’ value at certain critical moments of the organization’s growth. Older organizations also tend to better define the process of decision-making. In the case of sports clubs, it is most efficient for decisions to be made by the president, the executive office or the team of managers [6]. This diminishes the length of time needed to reach a decision and prevents misinterpretation on the internal rules. However, though the decision process is more efficient, over time older organizations begin to exhibit more rigid organizational structures. It becomes difficult to integrate new functions, new departments or new internal structures, as Loderer and Waelchli [9] reported. This can be a serious handicap in times of severe exogenous challenges. Finally, older organizations are usually associated with monopolistic market power. In the sports context, older clubs tend to benefit from dominant positions when trading players, changing competition rules, collecting revenues and hiring promising young players [10,11].
3. Empirical section—the Cox proportional hazards regression, data, sources and results 3.1. The Cox proportional hazards regression—the rationale behind our choice To analyze the survival rate of a professional soccer team in the first position of its national league, it is important to choose a suitable empirical method. We adopted the Cox proportional hazards regression. The Cox regression has been applied in various fields (for an extensive discussion, see, among others, Klein and Moeschberger [12]). This type of regression allows researchers to determine which characteristics of the subject (in this paper, the characteristics of each soccer team) influence the survival of the subject in some risky situation (in this case, the maintenance of a first-place position in the national league). As Danacica and Babucea [13] demonstrate, the Cox regression also allows researchers to estimate the statistical significance of this influence and the direction (i.e., if the analyzed characteristic increases or decreases the individual’s ability to survive). The simplest version of the Cox proportional hazards regression is shown in equation (1):
i (t) = exi ˇ ∗ 0 (t) , i = 1, 2, . . ., n
(1)
Equation (1) shows that, with n units being observed, the hazard rate (the probability of maintaining the first position/ascending to a top position after t periods, i.e., i (t))
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
4
P.R. Mourao
depends on the vector of the considered variables (x), the vector of the regression coefficients [ˇ = (ˇ1, ˇ2 , . . ., ˇk )], and on the baseline hazard (0 (t) [i.e., the probability of maintaining the first position/exiting a lower position by ascending to the top when all of the explanatory variables are equal to 0]). The regression coefficients can be interpreted as expressing the relationship between the proportional change that is expected in the hazard and the changes in the explanatory variables. They are usually estimated using maximum likelihood. In this model, the baseline hazard does not depend on the explanatory variables; it depends only on t. Furthermore, the exponential component comprises the explanatory variables but not the number of time periods. Finally, the explanatory variables are time-independent (typically called the ‘‘proportional hazard assumption’’). There are several advantages to using the Cox proportional hazards regression. It does not require the probability distribution of status changes over time to have a known parametric form. It also provides valuable information about cases of transitions not studied in previous literature. However, a strong assumption must hold: the proportional hazards assumption states that the difference in log hazard associated with a change in each covariate is not timedependent. The tests suggested by Grambsch and Therneau [14] and by Hosmer and Lemeshow [15] were performed on our data, and their results indicated that this assumption held (full details are available by request).
3.2. Data and sources To study the top-position survival rates of European soccer teams, we developed a database that included the explanatory variables suggested in the literature. We collected all of the data related to the 185 teams (and their clubs) that played in the six professional soccer leagues under study1 . We determined the number of matches in which each of these teams remained in the first position for each season (to construct the variable related to the survival time at the top). Additionally, we also computed the age of each club on the 31st of December of each observed season using the information available at each club’s official websites.
3.3. Controlling the age effect However, other variables must be considered to control the age effect (our main explanatory variable). Following the literature, we have to discuss these variables: • the percentage of national players: Several authors (e.g., Kleven et al. [16]) argue that a higher percentage of national players better represents the national conditions of a league. Because most of the national players have played for other clubs, these players exhibit a more profound knowledge of the national teams’ competitors. Additionally, the national players, having been educated
1
English, French, German, Italian, Portuguese, and Spanish professional soccer leagues.
in the country’s youth teams, tend to better understand the weather conditions of the country and the competition’s characteristics. Based on Marquez and Martin [17], we expect this variable to control the age of the club, although there are claims that this variable has a positive effect on the survival rate. Other authors (Kleven et al. [16]), however, argue for a negative effect; • the presence of the team at UEFA competitions: other authors (Marquez and Martin [17]) argue that the presence of a team at international competitions (such as the UEFA Champions League or the UEFA Europa League) influences the quality of that team’s matches at national competitions. Heron and Cupples [18], for example, comment that a team’s presence at international competitions exerts additional pressure on the players and can be physically taxing, diminishing the probability that the team will survive at the top. As suggested by Marquez and Martin [17] and Heron and Cupples [18], this dimension can influence the performance of a team in the national leagues because of the additional physical effort required and the prestige effect; • the relationship between transfer inflows and outflows: according to Mourao [4] and Peeters and Szymanski [5], these dimensions are related to a team’s budget and to the strategic renewal of the squad; • stadium attendance: there is a significant body of literature on the determinants of stadium attendance [19]. One of the correlated dimensions is the team’s success record —– the most successful teams tend to receive higher attendance. However, there is an open debate about the causality of this relation. Some authors [20,21] question whether higher attendance leads to more impressive victories and to a more consistent incidence of victory. The authors [20,21] suggest that these values can stimulate a professional soccer team to maintain the top position. Therefore, for these controls, we collected data from http://www.transfermarkt.com/for the following variables (for each season): ‘‘percentage of national players’’, ‘‘transfers inflows’’, ‘‘transfers outflows’’, ‘‘presence at UEFA Champions League at the beginning of the season’’, ‘‘presence at UEFA Europa League at the beginning of the season’’, and ‘‘stadium attendance’’. Tables 1—7 show the descriptive statistics of our database. Table 1 displays the statistics for the whole sample. The mean age of the European professional soccer teams in our sample is 92 years (ranging from 40 to 161 years). The mean number of matches that a professional soccer team retained the top position is 3.8. There were several cases in which teams maintained the top position the entire season (justifying the maximum value of 38). The mean values for the percentage of national players and for the log ratio between transfer inflows and transfer outflows (used for minimizing heteroscedasticity concerns) are, respectively, 53.8% and 0.31. Most of the soccer teams in our sample do not have the opportunity to play in UEFA competitions, resulting in mean values of 0.156 (Champions League) and 0.134 (Europa League) for this variable. Tables 1—7 present the values of our sample by national league. Table 8 displays the main results after estimating equation (1) using the appropriate proportional hazards
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
Age of a club and ability to maintain the top position in soccer leagues Table 1
5
Descriptive statistics (whole sample).
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of Journeys at the first place Age of the sports club % of national players Transfers inflows over transfers outflows (log) Champions League Europa League Stadium attendances (log)
3.826 92.449 53.824 0.311 0.156 0.134 9.937
9.639 44.798 15.121 1.588 0.364 0.341 0.830
0 40 14.3 −5.257 0 0 7.107
38 161 97.2 17.783 1 1 11.296
798 798 798 798 798 798 798
Table 2
Descriptive statistics (England).
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of Journeys at the first place Age of the sports club % of national players Transfers inflows over transfers outflows (log) Champions League Europa League Stadium attendances (log)
3.811 119.0 40.668 0.853 0.171 0.143 10.384
9.863 16.302 11.905 1.306 0.378 0.351 0.375
0 102 14.3 −1.632 0 0 9.666
38 153 56.7 5.013 1 1 11.234
140 140 140 140 140 140 140
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of Journeys at the first place Age of the sports club % of national players Transfers Inflows over Transfers Outflows (log) Champions League Europa League Stadium attendances (log)
4.378 111 58.791 −0.017 0.150 0.128 9.801
10.577 9.319 9.849 1.146 0.358 0.336 0.494
0 98 36.8 −3.401 0 0 8.754
38 150 80 4.784 1 1 11.236
140 140 140 140 140 140 140
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of Journeys at the first place Age of the sports club % of national players Transfers inflows over transfers outflows (log) Champions League Europa League Stadium attendances (log)
4.992 104.2 50.503 0.686 0.158 0.103 10.553
10.217 22.01 9.944 1.428 0.366 0.305 0.397
0 70 31.4 −3.076 0 0 9.695
38 161 93.8 6.386 1 1 11.296
126 126 126 126 126 126 126
Table 3
Table 4
Descriptive statistics (France).
Descriptive statistics (Germany).
method — the Cox regression. The event of interest was a professional soccer team maintaining first place in the national league. This model was applied to all of the professional soccer teams from the following six male leagues: the English Premier League, the French Ligue 1, the
German Bundesliga, the Italian Serie A, the Portuguese Super Liga, and the Spanish La Liga. Data were collected from the 2007/2008 season through the 2013/2014 season (because of data availability). Therefore, Table 8 presents the relative rates of the risk of being in the first position (the
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
6
P.R. Mourao Table 5
Descriptive statistics (Italy).
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of Journeys at the first place Age of the sports club % of national players Transfers inflows over transfers Outflows (log) Champions League Europa League Stadium attendances (log)
3.178 111.97 59.975 0.396 0.135 0.100 9.903
9.242 14.389 14.945 1.903 0.343 0.301 0.545
0 100 22.6 −2.988 0 0 8.675
38 154 91.4 17.781 1 1 10.998
140 140 140 140 140 140 140
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of journeys at the first place Age of the sports club % of national players Transfers inflows over transfers outflows (log) Champions League Europa League Stadium attendances (log)
3.571 92.785 45.269 −1.088 0.161 0.241 8.685
8.758 19.267 12.487 1.632 0.369 0.429 1.099
0 40 20 −5.257 0 0 7.107
33 125 77.4 5.011 1 1 11.218
112 112 112 112 112 112 112
Variable
Mean
Standard deviation
Minimum
Maximum
Number of observations
Number of journeys at the first place Age of the sports club % of national players Transfers inflows over transfers outflows (log) Champions League Europa League Stadium attendances (log)
3.072 89.071 65.761 0.089 0.171 0.107 10.107
8.934 24.565 13.591 1.570 0.378 0.310 0.548
0 75 40 −3.178 0 0 8.992
38 120 97.2 4.381 1 1 11.279
140 140 140 140 140 140 140
Table 6
Table 7
Descriptive statistics (Portugal).
Descriptive statistics (Spain).
estimated coefficients of the exponent, ˇ) derived from the Cox regressions. In accordance with the literature, the following explanatory variables were used: the age of the club on the 31st of December of each observed year, the percentage of national players, the ratio of transfer inflows to transfer outflows, the presence of the team at the UEFA Champions League or the UEFA Europa League (at the beginning of the season), and the value of each club’s stadium attendance at official matches of the professional male soccer team. We expected that older clubs would be characterized by higher rates of survival at the first position of the national soccer league. The literature also suggested that the presence of the teams at UEFA competitions and stadium attendance would positively increase those rates of survival. Therefore, we also controlled the age effect using these dimensions and the remaining variables discussed in the literature review (although the transfer cycle and the percentage of national players are not clearly associated with positive
effects on a team’s ability to remain in first place for a long time).
4. Results 4.1. An overall perspective Table 8 (columns for ‘‘Whole Sample’’) shows that the likelihood of a European soccer team surviving at the top is highest among the oldest teams. Another significant coefficient relates to the ratio of inflow from player transfers to the outflow from hires, whose increase tends to reduce a team’s ability to be at the top at the final match. Therefore, the ratio of inflow to outflow tends to be smaller for successful teams than for those that experienced a greater need to sell players, which introduced the need to renew the squad. As expected, a higher mean stadium attendance tends to provide additional resources for any European soccer team
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model
ARTICLE IN PRESS
SCISPO-3080; No. of Pages 9
Age of a club and ability to maintain the top position in soccer leagues Table 8
7
Cox regression analysis.
Variables Age of the sports club % of national players Transfers Inflows over Transfers Outflows (log) Champions League
Whole sample ***
1.028 (0.006) 0.967 (0.327) 0.967*** (0.007)
0.824 (0.273) 0.713 Europa League (0.351) Stadium attendances 1.193*** (0.092) 185 Obs. teams 1st-placed teams a 79 R2 0.187 LogL −147.9 Variables Age of the sports club % of national players Transfers Inflows over Transfers Outflows (log) Champions League
England ***
1.028 (0.005) 0.967 (0.317) 0.884* (0.066)
0.715 (0.351) 1.198*** (0.078) 185 79 0.181 −132.2
***
1.028 (0.005) 0.967 (0.397) 0.890 (0.070)
1.093 (0.013) 1.059*** (0.008) 1.264* (0.174)
1.195*** (0.074) 185 79 0.186 −138.0
1.331*** (0.552) 0.344 (0.221) 1.012*** (0.001) 47 14 0.473 −13.651
Germany 1.349 (0.429) 0.923 (0.135) 1.546*** (0.638)
0.992 (0.152) 1.002 Europa League (0.182) Stadium attendances 1.805*** (0.271) 27 Obs. Teams 1st-placed Teams a 17 R2 0.261 LogL −13.621
***
1.354 (0.418) 0.914 (0.145) 1.558*** (0.628)
1.092*** (0.026) 0.810 (0.188) 0.743*** (0.051)
1.093*** (0.026) 0.811 (0.189) 0.742*** (0.050)
1.013*** (0.001) 47 14 0.451 −14.561
1.144*** (0.071) 0.341 (0.501) 0.057 (0.043) 34 14 0.199 −15.421
0.045 (0.081) 34 14 0.201 −15.412
0.088 (0.098) 34 14 0.213 −15.422
1.103*** (0.040) 0.942** (0.028) 0.407*** (0.101)
0.972 (0.009) 0.901 (0.712) 0.992 (0.861)
0.912 (0.009) 0.899 (0.713) 0.991 (0.863)
0.911 (0.009) 0.898 (0.714) 0.990 (0.868)
1.102*** (0.039) 0.936*** (0.016) 0.466 (0.338)
1.101*** (0.041) 0.943** (0.027) 0.406*** (0.103)
1.701*** (0.238)
1.702*** (0.239)
1.923*** (0.281) 30 12 0.104 −12.623
1.924*** (0.282) 30 12 0.105 −12.624
0.564 (0.236) 0.721 (0.853) 4.807 (16.20) 20 9 0.122 −15.031
0.660 (0.248)
1.818*** (0.251) 27 17 0.263 −13.712
1.702*** (0.241) 1.001 (0.261) 1.923*** (0.281) 30 12 0.102 −12.615
0.991 (0.215)
1.812*** (0.241) 27 17 0.269 −13.671
0.997 (0.009) 1.141*** (0.045) 2.576*** (0.452) 1.002 (0.005) 0.997 (0.010) 1.004*** (0.001) 27 13 0.141 −11.132
% of national players Transfers inflows over transfers outflows (log)
Obs. teams 1st-placed teamsa R2 LogL
1.094*** (0.027) 0.861 (0.169) 0.749*** (0.062)
1.388 (0.402) 0.901 (0.121) 1.581*** (0.611)
Age of the sports club
Stadium attendances
0.332 (0.221) 1.011*** (0.001) 47 14 0.412 −12.521
1.093 (0.012) 1.058*** (0.008) 1.263 (0.221)
Portugal ***
Spain
Europa League
1.093 (0.013) 1.059*** (0.008) 1.264* (0.174)
***
Italy
Variables
Champions League
France ***
***
***
10.159* (12.07) 20 9 0.118 −14.919
0.997 (0.008) 1.142*** (0.044) 2.571*** (0.402)
0.996 (0.005) 1.141*** (0.043) 2.572*** (0.401)
0.996 (0.011) 1.004*** (0.001) 27 13 0.142 −11.133
1.005*** (0.001) 27 13 0.143 −11.134
10.160* (12.069) 20 9 0.117 −14.821
Values are hazard rates of exiting lower ranks by ascending to the first rank. Robust standard errors between parentheses. a Number of different teams ranked at the 1st place of the league since the first journey of the season 2007/2008 to the last journey of the season 2013/2014. * Significance level P < 0.1. ** Significance level P < 0.05. *** Significance level P < 0.01.
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
8
P.R. Mourao
to maintain the top classification in its national league. Our results (‘‘Whole sample’’ column of Table 8 suggest that a team exhibiting 1% more attendance during the season than a competing team is 1.2 times more likely to survive at the highest position than the other (less attended) team. Actually, as Table 8 shows and depending on the presence of additional covariates, this estimated coefficient ranges between 1.193 and 1.198. In the sample of 185 European teams, the percentage of national (or foreign players) did not significantly affect survival. The same observation was made of the variables related to the presence of the team at the Champions League or the Europa League. Overall, therefore, the presence of a European team at the UEFA does not particularly stimulate or burden the squads, whose rates of survival in their country’s championships are more influenced by the organizational structure (a factor related to the age of the club), to transfers, or to attendance at the home stadiums.
4.2. National cases Among the English teams, the age of the club tends to increase a team’s ability to survive at the top, with a factor (1.093) higher than the estimated coefficient for the whole sample (1.028). The stadium attendance also positively contributed to the ability to manage the pressure of being in the top position among the English teams. However, Table 8 reveals two interesting issues. The first relates to the positive estimated coefficients of the ratio of transfer inflows to transfer outflows, suggesting that, among English teams, more frequent rotation of the players promotes the ability to remain in the top positions. The second issue relates to the positive influence of having a larger proportion of English players on a team’s survival rate at the top. This effect may be due to the better assimilation of the managers’ instructions or to the transfer of highly profitable players from foreign origins. Finally, the presence of an English team at the Champions League tends to provide a stimulus for keeping the team at the highest position of the Premier League. Therefore, for English teams, the Champions League increases a team’s stamina, which compensates for the associated costs (e.g., players’ fatigue and the weight of mediocre international results). The age of the club also positively affects French teams. As with English teams, the French teams also gain additional stimulus from playing in the Champions League. However, the survival rates at the top significantly drop among teams that sell more players than they hire, indicating an interpretation in the opposite direction of that made for the English teams and also showing that the transfer variable had the opposite effect. The German, Italian and Spanish teams seem are not significantly influenced by the age of the club when fighting to maintain first place. More significant for the German and Spanish teams are the effects exerted by the value of transfer inflows and higher stadium attendance (also important for Italian teams). The age of the club also influenced Portuguese teams’ ability to remain in the top position. These teams were also significantly influenced by stadium attendance.
4.3. Synthesis and discussion of the results Obtaining the first position in its league is always a positive experience for any soccer team. However, the ability to maintain that top position is subject to a variety of factors. As our review of literature revealed, these factors include the strength of the club’s organizational structure [22] and the value invested in the team (usually positively correlated with the age of the club), the management of the squad’s players (related to the transfer cycle) and the external signs of confidence manifested in stadium attendance. In this paper, we focused on the age of the club. Older clubs survive because, generation after generation, their directors and managers have demonstrated the capacity to renew, usually by modernizing the buildings owned by the club and by enlarging the number of stakeholders, supporters, investors and fans. From a financial perspective, older clubs are expected to have more valuable assets; from a sports competitiveness perspective, older clubs are expected to exhibit more knowledge about trading players and coaches, hiring managers and managing publicity or the revenues from TV rights. As a result, older clubs tend to exhibit stronger structures that contribute to their teams’ ability to maintain the top positions in a sports league. Clubs without these strong structures are expected to encounter more difficulty in maintaining the top position after reaching first place, even if the teams of these clubs have the opportunity of participating at UEFA competitions, as alternatively Heron and Cupples [18] discussed. Our results confirm this hypothesis for European soccer teams, especially for English, French, and Portuguese teams. In their respective professional leagues, older clubs tend to survive longer in the top position than younger clubs. Even when controlling for the other two dimensions (‘‘management of the squad’’ and ‘‘external signs of confidence’’), we found that the age hypothesis holds for the professional soccer teams of England, France, and Portugal, which renew the works of authors like Peeters and Szymanski [5], Poli et al. [22], Scelles et al. [20] or Mourao [3]. For teams from Germany, Italy and Spain, this effect was not as significant. In these cases, the management of the squad and the external signs exerted more influence. Mourao [4] explains that some younger clubs have huge financial assets because they have received large investments from particular individuals (mostly from emerging economies) and can therefore compete with the older clubs for the topranked positions.
5. Conclusions and further work This study examined the ability of European soccer teams to remain in first place in their respective national leagues. We studied 185 teams from the 2006/2007 season to the 2012/2013 season. Using Cox proportional hazards regression models, we concluded that the age of the club is an important determinant in explaining the higher survival rates of the teams belonging to older clubs. However, we also controlled for the age effect with variables suggested by the literature. These variables included transfer flows, the presence of the team at UEFA competitions (Champions League or
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008
+Model SCISPO-3080; No. of Pages 9
ARTICLE IN PRESS
Age of a club and ability to maintain the top position in soccer leagues Europa League), stadium attendance, and the percentage of national players. The most robust results indicate that the age effect is a statistically significant determinant. Older clubs tend to benefit from four age-related dimensions: the relevance of assets, more efficient decision-making processes, monopolistic market power and the control of decisions by the executive board. More nuances were discovered when the teams were analyzed by professional league. Among English teams, the survival rate at the top is positively affected by having higher percentages of national players, higher transfer inflows, participation in UEFA competitions and higher stadium attendance. Stadium attendance is an important factor in keeping a team at the top, independent of the European league (with the exception of the French competition). French teams’ survival tends to benefit from the acquisition of players (proxied by higher transfer outflows). Although significant, the age effect nevertheless does not prevent young clubs (those founded within the last 50 years) from achieving first place. Young clubs that manage transfers well, have impressive stadium attendance, and are stimulated by UEFA participation tend to experience additional positive effects on their survival rate at the top. Based on this study’s findings, we identify three opportunities for future research. The first opportunity relates to the extension of this database to national leagues not included in this study. The second opportunity is the inclusion of greater detail in the four dimensions used to study the age effect (asset structure, organizational process, monopolistic power, and control of the decisions). We intend to compile convenient proxies for analyzing these four dimensions. Finally, the third opportunity is to broaden our interpretation of the ‘‘top’’ (here, only the first position) to consider additional top positions, for example, the positions higher or equal than 3rd place.
Disclosure of interest The author declares that he has no competing interest.
References [1] Davis M, End C. Team Success, Productivity and Economic Impact. In: The Economics of Sport, Health and Happiness, chapter 7. Edward Elgar; 2011. [2] Hone P. ‘‘Limiting Player Lists In Sport: Who Really Wins?’’ Economics Series 2005 03. Deakin University, Faculty of Business and Law, School of Accounting, Economics and Finance; 2005. [3] Mourao P. Regional determinants of competitiveness: the case of European Soccer teams. Int J Sports Finance 2010;5:222—34. [4] Mourao P. The indebtedness of Portuguese soccer teams—looking for determinants. J Sports Science 2012;30(10):1025—35.
9
[5] Peeters T, Szymanski S. ‘‘Vertical restraints in soccer: Financial fair play and the English Premier League’’ Working Papers 2012028. University of Antwerp, Faculty of Applied Economics; 2012. [6] Dinopoulos E, Waldo D. Gradual product replacement, intangible-asset prices and Schumpeterian growth. J Econ Growth Springer 2005;10(2), 135-157, 06. [7] Pfaffermayr M, Stöckl M, Winner H. Capital structure, corporate taxation and firm age. Fiscal Stud Inst 2013;34(1) [109—135,03]. [8] Phua F. The antecedents of co-operative behaviour among project team members: an alternative perspective on an old issue. Constr Manage Econ Taylor Francis J 2004;22(10):1033—45. [9] Loderer C, Waelchli U. Firm age and performance, MPRA Paper 26450. Germany: University Library of Munich; 2010. [10] Shimomura K, Thisse J. ‘‘Competition Among the Big and the Small’’ Discussion Paper Series DP2012-03. Research Institute for Economics & Business Administration, Kobe University; 2012. [11] Huang Y, Huang D. Big vs. small under free trade: market size and size distribution of firms. Int Rev Econ Finance Elsevier 2014;34(C):175—89. [12] Klein JP, Moeschberger ML. Survival analysis: techniques for censored and truncated data. New York: Springer Verlag; 2005. [13] Danacica D, Babucea A. Using survival analysis and economics. scientific annals of the Alexandru loan. Cuza University of Labsi: Economic Sciences Series, No. 1; 2010. p. 2010. [14] Grambsch P, Therneau T. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika 1994;81:515—26. [15] Hosmer D, Lemeshow S. Applied logistic regression. New York: Wiley; 2000. [16] Kleven H, Landais C, Saez E. Taxation and International migration of superstars: evidence from the European football market. Am Econ Rev 2013;103(5):1892—924. [17] Marquez M, Martin H. The New Football Business: a challenge for elite followers. Goteborg University; 2000. p. 18 [International Business Master Thesis]. [18] Heron N, Cupple M. The health profile of football/soccer players in Northern Ireland—a review of the UEFA pre-participation medical screening procedure. BMC Sports Sci Med Rehab 2014;6(5). [19] Buraimo B. Stadium attendance and television audience demand in English league football. Manage Decis Econ 2008;29(6):513—23 [John Wiley & Sons, Ltd]. [20] Scelles N, Durand C, Bonnal L, Goyeau D, Andreff W. My team is in contention? Nice, I go to the stadium! Competitive intensity in the French football Ligue 1. Econ Bull Access Econ 2013;33(3):2365—78. [21] Ponzo M, Scoppa V. Does the home advantage depend on crowd support? Evidence from same-stadium derbies. Università della Calabria, Dipartimento di Economia, Statistica e Finanza (Ex Dipartimento di Economia e Statistica); 2014. Working Papers 201402. [22] Poli R, Ravenel L, Besson R. Squad analysis: who will win the Champions League? CIES Football Observatory Monthly Report; 2015.
Please cite this article in press as: Mourao PR. Dissecting the survivors — how the age of a club contributes to the ability of a team to maintain the top position in the European soccer leagues. Sci sports (2017), http://dx.doi.org/10.1016/j.scispo.2017.01.008