Introduction

The relative age effect (RAE) refers to the over-representation of athletes born before a specific date for age grouping in various sports comparing to those born after this date [1]. Athletes born relatively late in the selection year potentially suffer a disadvantage during the selection process [2]. In sports where the selection cut-off date is January 1st, the number of athletes born between January and March may be several times higher than the number of athletes born between October and December [3]. Thus the RAE is widespread among young male athletes (age 15–18 years) performing competitively in soccer, athletics, and basketball [1,3–6]. Many studies have examined the wide spread of the RAE among soccer players of different ages and levels of competition [7–10]. Williams et al. showed that in a cohort of FIFA U17 World Cup players approximately 40% were born in the first quarter of the year while only 16% were born in the last three months of the year [11]. Rada et al. further demonstrated that the number of players born in the first month of the calendar year is twice as many of those born in the last month of the year, while the RAE is also widely prevalent in second-tier players [12]. More recent studies also showed that the RAE is evident in elite German [13] and Scottish soccer players [14]. Although, the prevalence of the RAE in older soccer players is not as high when compared to younger players [15].

The main negative factor of the RAE is the "discrimination" of late-born athletes likely related to them being less physically developed (as they are younger) and therefore considered less mature. For example, a study by Romann et al. showed that the difference in 60-meter sprint performance between “early-” and “late-born” athletes aged 8 to 15 years born in one year ranged from 5–10% [16]. However, maturity and the RAE are two independent phenomena, and among early-born and late-born athletes, the ratio of early-, on-time and late-maturing is the same [17], i.e., both early- and late-born athletes can be early- or late-maturing.

Recently, there has also been growing interest in analyzing the relationship between a soccer player’s date of birth and the corresponding market value, although the available data are conflicting and limited. Furley et al. found that in the top 100 soccer players, there were more “early-born” players and these players had a higher average market value [18]. Contrastingly, Doyle et al. analyzed the top 1000 professional UEFA U19 Youth League players and found that the market value of “early-born” players does not exceed the market value of “late-born” players [19]. Thus, it is practically interesting to examine the RAE and its association with the market value in soccer players. This study aimed to analyze the relationship between the player’s birth quarter and their market value and to assess the prevalence of the RAE depending on the competitive level and playing position of European soccer players.

Materials and methods

The materials used in this article are based on squad data from the top divisions of 54 European countries downloaded from www.transfermarkt.com. Data were publicly available as of September 10^th, 2020. The website Transfermarkt (www.transfermarkt.com) has been recently used as a data source for elite soccer studies [2,19–22]. Transfermarkt has information on male soccer players only; female players are not included. This source has previously proven to provide reliable match performance indicators and has been described as a good predictor of real market values [23]. This database has also been used for studies on the RAE. For example, Doyle et al. used the top 1000 soccer player Transfermarkt value in the 2013–2014 season to report that early-born players were worth more [19].

For data mining, software was used to exploit the PHP Simple HTML DOM Parser library. All publicly accessible pages have been analyzed page by page to obtain information on the player’s date of birth, position on the field, and market value. The collected data was organized into a MySQL database, which was then used to design queries to select and interpret the required data. Local ethics committee approval was not required since publicly available data was utilized. Overall, the analysis included data on 18,429 soccer players from 731 top-division teams playing in 54 European First Leagues from respective countries (there is no First League in the Principality of Liechtenstein). These 54 countries have been divided into four groups based on the level of representation of the country’s soccer clubs in the 2018/2019 UEFA Champions League [24], as this appears to be the most objective method of classifying European club soccer and this classification was made by European soccer authorities (UEFA). Group 1 included six countries that had the maximum representation of seven clubs in the Champions League, respectively (Spain, England, Germany, Italy, France, and Russia). Group 2 included nine countries with five teams in the Champions League (Portugal, Belgium, Ukraine, Turkey, Netherlands, Austria, Czech Republic, Greece, and Switzerland). Group 3 included 35 countries, four teams in Champions League (Denmark, Croatia, Cyprus, Serbia, Scotland, Belarus, Sweden, Norway, Kazakhstan, Poland, Azerbaijan, Israel, Bulgaria, Romania, Slovakia, Slovenia, Hungary, Albania, North Macedonia, Bosnia and Herzegovina, Moldova, Ireland, Finland, Georgia, Malta, Iceland, Wales, Northern Ireland, Montenegro, Estonia, Faroe Islands, Luxembourg, Armenia, Latvia, and Lithuania). Group 4 includes four countries with one to three teams in the Champions League (San Marino, Andorra, Kosovo, Gibraltar). The top 50 European teams according to the UEFA ranking 2019/2020 [25] were analyzed separately as a group of "best teams" (BT) and was the group of the highest competitive level.

All players were divided into four groups according to their month of birth:

• Players born in the first quarter of the year (January, February, March) (Q1, early-born)
• Players born in the second quarter of the year (April, May, June) (Q2)
• Players born in the third quarter
• of the year (July, August, September) (Q3)
• Players born in the fourth quarter of the year (October, November, December) (Q4, late-born).

The RAE was defined as a higher relative sample size in the first quarter compared to other quarters by date of birth.

Players were also grouped by position, Goal-keepers (n = 2075), defenders (n = 5884), midfielders (n = 7826), and forwards (n = 2644).

Results

18,429 players competed in 731 top division teams in 54 European countries. Of these, 30.9% (5,688) were born in Q1 and 19.6% (3,609) in Q4 (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Distribution of players in all European championships by the quarter of birth.

https://doi.org/10.1371/journal.pone.0283390.g001

When analyzing the prevalence of the RAE, a significant predominance of soccer players born in Q1 over those born in Q4 was revealed in all groups except for Group 4, which included countries with the lowest representation in the Champions League (Fig 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Distribution of players by quarter of birth in Groups 1–4 and BT.

https://doi.org/10.1371/journal.pone.0283390.g002

In all groups except Group 4, the proportion of players born in Q1 was 30% or more (Table 1). The highest percentage of players born in Q4 (23%) was observed in Group 4. In all other groups, this did not exceed 20% (Table 2). The highest percentage of early-born players was observed in Group 1 and in the BT group (33%) which represented the most competitive groups.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Distribution of “early-born” players in various groups.

https://doi.org/10.1371/journal.pone.0283390.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Distribution of “late-born” players in various groups.

https://doi.org/10.1371/journal.pone.0283390.t002

The strongest RAE was observed in defenders, where 31.3% (1844 players) were born in Q1 and 19% (1116 players) were born in Q4. The weakest RAE was observed in the forwards, where 29.5% (781 players) were born in Q1 and 20.4% (539 players)–in Q4. The RAE was observed in every analyzed group across all positions. Analysis of the dates of birth of players playing in different positions showed that for each position there was a majority of “early-born” players (Fig 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Relative Age Effect by playing position (%).

https://doi.org/10.1371/journal.pone.0283390.g003

The mean market value was the highest in players born in Q4 (€2,200,056) and the lowest in players born in Q3 (€1,815,839). The mean market value for players born in Q1 was €1,991,380, and €2,081,436 in Q2. (Fig 4). In the most competitive groups (Groups 1, 2, and BT), the highest mean market value was observed for players born in Q4 in BT (€14,589,015).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. The association between a player’s quarter of birth and market value.

https://doi.org/10.1371/journal.pone.0283390.g004

Forwards and midfielders had the highest market value in all groups (Fig 5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Relationship between the player’s playing position and transfer fee.

https://doi.org/10.1371/journal.pone.0283390.g005

Comparison between groups

When comparing the dates of birth of players from different groups, there was a correlation between the level of competition and the number of “late-born” players. As the level of participation increases, the number of players born in Q4 decreases.

Table 3 shows the differences between the number of “late-born” players in different groups. In all cases where the difference was statistically significant (p <0.05), there were more “late-born” players in the less elite groups.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Differences between the number of ‘late-born’ players in various groups (p values)—Chi-square.

https://doi.org/10.1371/journal.pone.0283390.t003

Table 4 shows the differences in the number of "early-born" players in different groups. In all cases where the difference was statistically significant (p <0.05), there were more “early-born” players in the more elite groups.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Differences between the number of “early-born” players in various groups (p values)—Chi-square.

https://doi.org/10.1371/journal.pone.0283390.t004

Market value did not differ significantly between players born in Q1-4, including between “early-born” (Q1) and “late-born” (Q4) players (p> 0.05). Also, there was no significant difference in market value between players of different playing positions, born in different quarters. (Table 5). When analyzing the market value of all soccer players, it was found that forwards born in the Q4 were the most expensive players.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Differences between market value across Q1-4—Kruskal-Wallis test.

https://doi.org/10.1371/journal.pone.0283390.t005

Mean market value differed significantly across various groups for all players, goalkeepers, defenders, midfielders, and forwards (p = 0.001) between each group. As expected, the highest player price was observed in group BT, while the lowest was observed in Group 4 (Table 6).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Differences between market value across groups—Kruskal-Wallis test.

https://doi.org/10.1371/journal.pone.0283390.t006

Post hoc pairwise comparisons were also performed using the Dunn test with Holm adjustment which showed significant differences in market value between Groups 1 and 4, 4 and BT, and 3 and BT (Table 7).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 7. Differences between market value between groups (p values)—Dunn test with Holm adjustment.

https://doi.org/10.1371/journal.pone.0283390.t007

Discussion

The study aimed to examine the prevalence of the RAE and any differences in the level of competition and playing position. This study also analyzed the relationship between player’s birth quarter and market value across 18,429 soccer players from 731 top division teams playing in 54 European Championships. Overall, the RAE was widely prevalent in the most competitive senior soccer teams and leagues in Europe regardless of playing position. The number of players born in Q1 decreased as the level of competition decreased, where the highest percentage of these players was observed in clubs that are among the top 50 ranked in UEFA listings or competed in the most prestigious European championships. The RAE was less pronounced in teams that participated in the lower competitive championships. There was no difference in market value between players of the same playing position and level of competition born in different quarters. Although, the most expensive soccer players in the European championships were late-born forwards, players from various groups differed in their market value.

The RAE has been previously described in various groups of soccer players with a different distribution pattern depending on several factors, including competitive level, age, playing position, and nationality [26,27]. According to multiple studies in soccer, the RAE is most pronounced in elite young male soccer players (under 18 years), and it can significantly influence the future career of these young players [28–30]. Furthermore, in German soccer, which is regarded as one of the most competitive in the world, “early-born” players have a greater probability of developing into professional player [31]. While in French soccer, also considered one of the leading soccer environments in the world, “late-born” players are more likely to drop out than “early-born” players [32]. These results are consistent with our findings.

The wide spread of the RAE in elite European youth and adult soccer over the past decade may be secondary to the increase in the popularity of soccer and, the differences in the level of competition, as well as the processes and timing of the initial selection. Considering that currently, the primary selection in the leading European soccer academies occurs before 10 years old, it becomes evident that a difference of even a few months of age can provide a significant advantage in terms of physical performance. Therefore, it is not surprising that relatively older and more biologically mature soccer players will have an advantage in youth soccer [33]. Gil et al. revealed statistically significant differences in anthropometry and physical performance in older pre-pubertal soccer players with an average age of 9.75 ± 0.30 years when compared to their younger counterparts [34]. Therefore, it may be reasonable to conclude that only “late-born” athletes with a more biologically mature status are likely to have the opportunity to compete with “early-born” athletes. This finding is supported by Müller et al. who analyzed the prevalence of the RAE and the degree of biological maturity in 222 male soccer participants from the UEFA European Under-9 Championship. Müller et al. results showed that the primary selection process in young international soccer appears to be related to biological maturity status and relative age. Furthermore, “late-born” children seem more likely to be selected for highly competitive soccer organizations when a more biological mature status is evident [35].

Under the existing selection system, late-maturing children may be subjected to certain discrimination. During the initial selection, the first wave of talented athletes to drop-out occurs, generally by athletes who may be temporarily physically and psychologically less developed [36–38]. The second wave of drop-out then seems to occur during the growth spurt (12 to 14 years old), when the elimination of players from many highly competitive sports organizations can take place due to the influence of different rates of biological maturation [36–38]. Young soccer players with normal or delayed biological maturation are thus discriminated against, and the advantage is towards early-maturing athletes, regardless of date of birth. This statement is well illustrated by Malina et al. who analyzed the degree of biological maturation of elite young Portuguese Academy soccer players aged 11 years old (pre-puberty), 13 to 14 years old, and 16 to 17 years old. In these age groups, the number of soccer players with different degrees of maturation varied significantly. According to this data, at 11 to 12 years old, the proportion of early- and late-maturing boys was similar (21%), while in 13 to 14 years old players it was 38% and 7% respectively, and in 15 to 16 years old players it was 65% and 2% respectively. The results of this study show that late-maturing boys are systematically excluded from elite youth soccer, and as chronological age and sport specialization increase, preference is given to “on-time” and early-maturing boys [39]. In contrast, it should be noted that not all young soccer players born in Q1 have a physical advantage over their younger peers, however, even in these cases coaches tended to rate players born in Q1 higher [40]. Therefore, it may be assumed that the RAE contributes to the exclusion of late-born players from highly competitive youth soccer, not only due to lower physical performance but also due to factors such as behavioral variables, coaches perceptions and the training environment [41].

The absence of the RAE in the least competitive championships (Group 4) may be due to the low level of competition in these countries in the initial selection and a large number of “local” soccer players in adult teams. Although, the market values in the most competitive groups (Groups 1, 2, and BT) were highest for players born in Q4, as the Kruskal-Wallis test showed that market value did not differ significantly between players from different positions born in Q1-Q4, as well as between "early-born" and "late-born" players. The similar market value of “early-” and “late-born” soccer players in adulthood confirms this notion, at least, in part dictates the need to ensure measures to reduce the severity of the RAE in childhood and youth soccer. Concurrently, it must be considered that despite a large number of studies reporting the negative impact of the RAE on the development programs of soccer players in many countries, its impact in European countries with a high level of soccer development from 2001 to 2011 did not decrease and thus the RAE can still be considered prevalent in all European countries [19].

The study by Salinero et al. conducted on adult soccer players also revealed the re-distribution of “early-born” players in the most competitive championships and persisted when players were classified by playing position. In the majority of groups, there were more “early-born” goal-keepers and defenders than in any other position. This may be explained by certain anthropometric requirements for soccer players in these positions (height and large muscle mass), which can lead to drop-out or changing to other playing positions at various stages of selection during specific time periods. In our study, however, this group of championships corresponds to Group 1. Like Salinero et al. we found that the percentage of players born in Q1 (33.3%) was higher than players born in other quarters of the year, with the highest percentage in defensive and midfield positions (34.0% and 34.2%, respectively) and the RAE was also most significant in these positions [42].

The prevalence of the RAE in elite adult midfield players may be associated with both a high competitive level and the requirements of modern soccer. Midfield players are required to perform large quantities of physical activity at high speed, which in youth soccer depends on a well-developed aerobic capacity and the degree of biological maturation [43]. Although there were no differences reported in the intra-group market value between “early-” and “late-born” players, while the most expensive soccer players were forwards born in Q4, which is also consistent with Romann et al. in older age group players. In the study by Romann et al. based on data from Transfermarkt.com, the prevalence of the RAE was studied in 2,000 of the most expensive soccer players aged 19 to 23 years old. In all age categories, the elevated prevalence of soccer players born in Q1 was also revealed. However, this tendency was not evident when examining by playing position. An important result of this study was the correlation between the change in the market value of “early-” and “late-born” soccer players and the increase of age. In the 19 years old age group, the most expensive soccer players were born in Q1, and players aged 21 to 23 years old were born in Q4 [44].

In the study by Doyle et al. no data indicated the different values of the most successful young soccer players. The authors investigated the relationship between a player’s market value and date of birth using data on the top 1,000 professionals and UEFA U19 Youth League players. They found no difference between the market value of “early-born” players and “late-born” [19]. In support of Doyle et al. our results were unable to report any differences in market value between players born in different quarters of the year within their respective country groups and competitive level. This trend was observed across all groups examined. Interestingly, despite the lack of statistically significant results, the total value of players born in Q4 exceeded the total value of players born in other quarters of the year. Similar results were obtained by Salinero et al. who analyzed the birth quarter of players from the top five European leagues (England, Italy, Germany, France, and Spain). They found that players born in Q1 were over-represented compared to players born in the other three quarters [42]. Furthermore, in the study by Fumarco et al. they reported possible reasons for the greater success of “late-born” athletes at the adult elite level in the National Hockey League [45]. The first explanation provided was based on psychological stability, where “late-born” players were more psychologically stable and motivated, competing with older players. This is considered the ’underdog’ hypothesis, in which relatively younger players are believed to benefit from more competitive play with their older counterparts [45]. The underdog phenomenon might also be present in soccer, where Cumming et al. showed that late-maturing players appear to possess a psychological advantage in academy soccer [37]. However, given the published data on the influence of biological maturation status and the possibility of selection into highly competitive sports organizations and maintaining this selection, this theory, at least in relation to children and elite youth soccer players, should be considered with caution [35,39,46]. The second explanation regarding biological status assumes that the selected “late-born” players are more talented, which allows them to compete with the older more developed (physically and psychologically) players [45].

In this regard, the main task of coaches and scientists should be to develop and implement measures to reduce the discriminatory influence of the RAE and the biological maturation status when screening and selecting less mature and “late-born” soccer players. Furthermore, the development of physiological and psychological assessments, considering the chronological and biological age of soccer players may allow the identification of talent to be un-biased.

The limitations of our study included the data collection procedure as this was performed from an open, online database. Even though Transfermarkt.com has previously shown to provide a reliable estimate of the market value of players [23], it is still limited. For example, it does not include female soccer players, thus, estimation of RAE among female players cannot be performed based on Transfermarkt data. Among these are the vast number of evaluations of top league players compared to minor league players, and the criteria used for player evaluation. The dependence on a high number of assessments, which require several matches to provide an adjusted market value (between 6 and 12 months) is a limitation [47,48]. Furthermore, the cross-sectional design is also a limitation.

Therefore, it can be summarized that “late-born” soccer players are not less successful, but are potentially exposed to selection bias at different stages in highly competitive sports organizations, such as soccer, where the most important physical qualities are strength and speed. In this regard, it is understandable why the “late-born” soccer players can be graded lower in adolescence (when the status of biological maturation is playing a significant role) compared with “early-born” players. Future studies should aim to examine the influence of factors such as player injury history on the market value as well as measures aiming to mitigate the discriminatory effects of selection bias based on the RAE including education of relevant stakeholders in soccer academies.

Conclusions

The RAE is currently prevalent in all of the most competitive senior soccer leagues and teams in Europe regardless of playing position. There are no significant differences in market value between players of the same playing position and competitive level when born in different quarters. The most expensive soccer players in the European championships are forwards born in Q4. These data may indicate the under-representation of “late-born” soccer players in youth, and then consequently in adult soccer, is not associated with lower talent, but with other factors, possibly based on physiological characteristics and socio-cultural factors. Further measures are needed to mitigate the discriminatory effects of selection bias based on the RAE.