Relative age effect in lower categories of international basketball

Abstract

To be able to value the relative age effect in the male and female World Championships played between 2005 and 2010 in the U17 categories (athletes 17 years or younger), U19 (athletes 19 or younger) and U21 (athletes 21 years or younger) a sample of 954 players has been selected. The variables registered were their dates of birth, the category of the competition, gender, height and official statistics of each player obtained from the International Basketball Federation (FIBA). A clear relative age effect was found (in both male and female categories) fading with age, being higher in the U17 category, slightly less but also significant in the U19, and no significant effect found in U21. This effect persists when the different specific positions were analysed in the male categories, being clearer in the positions that require more physical strength. In female categories the results do not back the existence of the relative age effect. Also, differences were found in height in the male category with regard to the players’ year-quarter of birth, but its interpretation is not consistent with the relative age effect. In the female category no differences were found in height. Finally, the performance difference of the players in the male and female categories hardly varies with regard to the year-quarter of birth.

Keywords

basketball performance analysis relative age effect World Championship young athletes

Introduction

The term Relative Age Effect (RAE) has been used to determine the effect of the influence of the date of birth in the performance of the person. The first studies that investigated this effect were done in the education environment (Armstrong, 1966; Freyman, 1965).

Investigations made in the education area by Russell and Startup (1986) based their study in the relevance of being born at the start or at the end of the academic year, concluding that pupils born at the start had an academic advantage over the rest until 18 years of age, but after this age, the ones born at the end of the year gave better performance.

Grondin et al. (1984) were the first to do a study of the RAE in sport, finding an unequal distribution in the dates of birth in the players in different levels of Canadian ice hockey and volleyball, concluding that many players were born near to the cut date. These studies have been revised, updated and confirmed (Gibbs et al., 2011; Nolan and Howell, 2010).

Musch and Hay (1999) investigated the age effect in an intercultural sample (Germany, Japan, Brazil and Australia) and concluded that the cut date in football is the main factor that effects relative age in professional football.

Other studies that analyzed the RAE in the formation ages of sport groups included the essays by Helsen et al. (1998), whose results indicated that youth football players born between August and October (the first part of the year for the selection) are more probable to be identified as talented and to be exposed to higher levels of training, whilst the players born at the end of the year tended to drop out at 12 years of age. Also, an essay with young footballers by Ashworth and Heyndels (2007) proved that players born after the cut-off date earned higher wages.

There are studies that found that age bias is not only prevalent in the minor leagues but also carries over into the professional leagues. The results of the Helsen et al. (2005) investigation show an excessive representation of players born in the first three months of the year (from January to March) for all youth National Teams in the under 15 years of age (U15), U16, U17 and U18, and the same being true for the UEFA Sub-16 and the Meridian Cup. The players with a relatively older age are more likely to be identified as talents because of the probable physical advantages that they have over the other, younger players. Continuing with football at professional level, Jullien et al. (2008) concluded that coaches tend to select players born in the first four months of the year.

Carling et al. (2009) investigated whether the maturity, the anthropometric profile and the valuation of physical state varied in the distribution of the date of birth in the elite. This study suggested that the relative age of the athlete does not always relate to a significant advantage in physical components.

Delorme and Raspaud (2009) found clear differences in the relative age effect in French athletes aged between seven and 18 years, in both male and female categories. They also studied the height of the players, finding that the ones born in the first two terms of the year were taller. Also, Delorme et al. (2011) investigated the relative age effect as a factor for abandoning sport in basketball players, finding a higher index of leavers in players born at the end of the year. Nolan and Howell (2010) found that age bias is not only prevalent in the minor leagues but also carries over into the National Hockey League (NHL).

Other studies found that relative age effect is only prevalent in lower categories, diminishing over time and not being present in professional sport. In basketball, the investigation by Esteva et al. (2006) determined a strong tendency to select players born in the first three months of the year compared to those born towards the end of the year. This tendency loses its strength as the players go through the categories, until arriving at professional basketball. This is explained because in the first stages of the sport, the players are selected only because of their advanced maturity or other indicators such as height. This way, a great quantity of possible future talents are lost and other players have more opportunities of getting to become professional players only because they were born in the first three months of the year. Gibbs et al. (2011) found that the relative age effect is moderate for the average Canadian National Hockey League player and reverses when examining the most elite professional players.

Lidor et al. (2010) studied the effects in a small country (Israel) of the relative age and the place of birth of the players in various sports, in which basketball was included. No significant relative age effect or the effect of the place of birth was found.

Baker et al. (2010) revised the possible causes of the RAE and suggested some solutions. The most used explanation for justifying the RAE is the process of maturation of the athletes: the athletes born nearer the cut-off date have higher levels of performance than the younger ones (Barnsley and Thompson, 1988; Malina, 1994; Malina et al., 2004). The solutions suggested tend to be related to the variation of the age, which means that the RAE changes but is persistent (Helsen et al., 2000; Musch and Hay, 1999; Simmons and Paull, 2001). Other solutions suggest an enormous administrative complex, such as the one by Barnsley and Thompson (1988) who say that the selection of participants should adjust to a certain distribution or control of the average age in all types of teams Helsen et al. (1998, 2000).

The objective of this present study is to check whether the relative age effect does exist in the World Basketball Championship U17, U19 and U21 male and female categories, to investigate if the relative age effect exists in the different specific positions and also try to find differences in height and in performance between players depending on their date of birth.

Material and methods

Sample

The athlete populations were selected from the last male and female editions of the Basketball World Championships U17 played in 2010, U19 played in 2011 and U21 played in 2005 in the male category, and in 2007 in the female category. The total number of athletes in the sample is 954, of which 472 are from the male category (143 player in the U17, 191 in U19 and 138 in U21) and 482 from the female category (144 players in the U17, 194 in U19 and 144 in U21).

The International Basketball Amateur Federation (FIBA) defines the international rules of basketball and is responsible for controlling and regulating all international competitions. In the World Championship in the U17, U19 and U21 categories, the participation of the athletes must be that age or younger. The criteria selection used indicates that at least 10 participating athletes are born in the same year. Applying this rule, in the analysis, in the male category a 15-year-old athlete was excluded in the U17, another aged 16 in the U19 and eight athletes in the U21 (seven players were 18 years old and one was 17). In the female category five players aged 15 and two aged 17 were excluded in the U17, seven players aged 16 and two aged 15 in the U19 and four players aged 17 and two aged 16 in the U21.

Procedure

The following variables were studied: gender, category, season, team, classification, position and date of birth. Also the height of the players was registered and the relative variables of the performance of the players (games played; minutes played; converted field goals, tries, and the percentage of effectiveness; two point field goals, tries, and the percentage of effectiveness; three point field goals, tries, and the percentage of effectiveness; free goals scored, tried, and the percentage of effectiveness; defensive rebounds; offensive and total of rebounds; assistances; personal faults; recuperations; stolen; blocked; points; points per game; rebounds per game and game assistance).

The information was collected from the FIBA website (http://www.fiba.com/). Afterwards the variable term was generated, dividing the dates of birth of the players into four terms (quarters) that start on the 1st of January and end on the 31st of December of the same year. This way the athletes born from the 1st of January until the 31st of March form the first term (Q1), the players born from the 1st of April until the 30th of June form the second term (Q2), the players born from the 1st of July until the 30th of September belong to the third quarter (Q3) and finally the players born from the 1st of October until the 31st of December make up the fourth term (Q4).

Analysis of information

To determine the RAE, the Chi-squared test was used to determine if the distribution of the dates of birth differ significantly from the theory distribution that is hoped to be found (in which the probability of finding athletes born in whichever term of the year is the same).

To compare between the height of the players and the relative performance variables for each term of birth, an ANOVA was used and the post-hoc analysis used the Tukey range test, a previous guarantee of normality (Kolmogorov–Smirnov test) and an equality (Levene test). In the case that the requirements of application were not met, the use of ANOVA is rejected and the Kruskal–Wallis test is used. In this study the statistic tests are considered significant when p<0.05.

Results

Table 1 shows the term distribution of the birth dates of all players of the basketball world championships in lower categories.

Table 1.

Terms of the date of births of all players in the Basketball World Championships U17, U19 and U21.

Gender		Overall		U17		U19		U21
		Observed	Expected	Observed	Expected	Observed	Expected	Observed	Expected
Males	Q1	176	118	66	35.8	68	47.8	42	34.5
	Q2	126	118	38	35.8	49	47.8	39	34.5
	Q3	102	118	26	35.8	47	47.8	28	34.5
	Q4	68	118	12	35.8	27	47.8	29	34.5
	Total	472		142		191		138
	χ²	52.41		43.66		17.65		4.32
	Sig.	0.000		0.000		0.001		0.229
Females	Q1	159	120.5	50	36	71	48.5	38	36
	Q2	144	120.5	40	36	57	48.5	47	36
	Q3	102	120.5	31	36	38	48.5	33	36
	Q4	77	120.5	23	36	28	48.5	26	36
	Total	482		144		194		144
	χ²	35.43		11.28		22.87		6.50
	Sig.	0.000		0.010		0.000		0.090

The global distribution observed is different than expected in both male (χ²=52.41; d.f.=3; p<0.001; Figure 1), and female categories (χ²=45.43; d.f.=3; p<0.001; Figure 2).

Figure 1.

Distribution of the dates of birth in function of age in the male categories.

Figure 2.

Distribution of the dates of birth in function of age in the female categories.

When analyzing the categories in the U17 competition, the distribution observed is different than the uniform expected in the male (χ²=43.66; d.f.=3; p<0.001) and female categories (χ²=11.28; d.f.=3; p<0.011). The same happened in the U19 category in both the male (χ²=17.65; d.f.=3; p<0.001) and female categories (χ²=22.87; d.f.=3; p<0.001). Finally in the U21 category the distribution observed does not differ from the expected in either male (χ²=4.32; d.f.=3; p<0.229) or female categories(χ²=6.50; d.f.=3; p<0.091).

Specific positions

In the analysis of the specific positions (Table 2) a different distribution was found in the birth terms than the one expected in all of the positions (p<0.05) in the male category, being more distinct in the positions that require higher physical form (centre and power forward). In the female category the distribution in terms are different in the point guard, shooting guard and small forward (p<0.05).

Table 2.

Birth terms of players in the Basketball World Championships in predicting performance in specific positions.

Gender		Point-Guard		Shooting-Guard		Small-Forward		Power-Forward		Centre
		Obs.	Exp.	Obs.	Exp.	Obs.	Exp.	Obs.	Exp.	Obs.	Exp.
Males	Q1	21	12.8	24	11.8	22	15.3	19	11.8	20	18
	Q2	6	12.8	7	11.8	18	15.3	12	11.8	27	18
	Q3	18	12.8	10	11.8	14	15.3	11	11.8	15	18
	Q4	6	12.8	6	11.8	7	15.3	5	11.8	10	18
	Total	51		47		61		47		72
	χ²	14.65		17.77		8.05		8.40		8.78
	Significance	0.002		0.000		0.045		0.038		0.032
Females	Q1	19	10	24	13.5	17	13.3	11	11.5	22	20.3
	Q2	11	10	9	13.5	21	13.3	13	11.5	23	20.3
	Q3	4	10	13	13.5	5	13.3	10	11.5	22	20.3
	Q4	6	10	8	13.5	10	13.3	12	11.5	14	20.3
	Total	40		54		53		46		81
	χ²	13.40		11.93		11.53		0.44		2.61
	Significance	0.004		0.008		0.009		0.933		0.457

Height of players

When comparing height in the four terms, significant differences were found (p<0.008) in the male category, although no differences were found using the post-hoc analysis between the first and fourth term. Two homogeneous subsets were defined, the first formed by the heights of the first (195 cm), third (195.41 cm) and fourth terms (197.86 cm); the second is formed by the second (198.45 cm) third (195.41 cm) and fourth terms (197.86 cm). In the female categories no significant differences were found.

Performance

Male category

In the U17 category after the application of the Kruskal–Wallis test, significant differences were found (p<0.017) in the percentage of three point field throws, showing better percentages in players born in the first (22.5%) or in the second term (22.8%) of the year than players born in the third (20.0%) and fourth terms (21.4%).

In the U19 category significant differences were found (p<0.036) using the Kruskal–Wallis test in the points obtained per game with values of 6.7% for the first term, 7.3% for the second term, 5.2% for the third term and 6.8% for the last term. In the U21 category no significant differences were found in either of the variables studied.

Female category

No significant differences were found in the performance by terms in the U17 category. In the U19 category differences were found in the percentage of field throws (p<0.005) with lower percentages in the first (33.1%) and second terms (34.8%) in relation to the third (41.7%) and fourth terms (39.2%). In the percentage of the two points (p<0.005), the same performance and values were found in all four terms respectively, 35.2%; 36.9%; 46.6% and 41.1%. The assistances (p<0.013) show values of 8.9%, 4.7%, 7.1% and 7.1% for the four terms respectively, showing lower values in the second term. The assistances per game (p<0.013) show values of 1.17%, 0.64%, 0.95% and 0.94% for the four terms, and also show lower values in the second term. Finally, in the U21 category no significant values were found in the performance with regard to the term of birth.

Discussion

In the present study, the existence of the relative age effect has been confirmed in the Basketball World Championships in U17 and U19. In the U21 championship no significant differences were found. The same behaviour was found in the male and female categories. Also, the effects of relative age persist when talking about specific positions, proving to be more distinct in the positions that require more height in the male category. In the female category the effects of the relative age are more significant in the positions that require less height. However, the height of the players does not show expected values if the relative age effect exists. No significant differences were found in the height of the players analysed. Finally, small variations have been found in the performance of the players with regard to the effect of the term of birth in both male and female categories, although these variations do not coincide with the existence of the relative age effect.

The existence of the relative age effect in basketball has been documented by various authors (Delorme and Raspaud, 2009; Delorme et al., 2010; Esteva et al., 2006) in male categories. Most investigations analyse male sport and few have been done in female categories and even fewer in basketball. Delorme et al. (2010) found a relative age effect in female basketball and Roman and Fuchslocher (2011) found it in football in 2011. These studies, as with the present one, disagree with the ones done by Delorme and Raspaud (2009) who found no relative age effect in female basketball. This effect may have been found in other young athletes in other sport disciplines (Ashworth and Heyndels, 2007; Helsen et al., 1998, 2005).

As in this study, the investigations done in German football by Schorer et al. (2009b) or in handball by Gutiérrez et al. (2012) found a slight fall in the relative age effect as the age of the athletes increases.

Schorer et al. (2009b) documented a relative age effect in different specific positions in German football, proving results that correspond with those obtained in the present study in the male categories, as the positions that require a higher physical form are taken more often by athletes born in the first months of the year. The results found in female categories do not support the idea of the relative age effect as the players born in the first months of the year tend to take base point, shooting guard and small guard positions, which are less dependent on biologic maturity. However, in football, Roman and Fuchslocher (2011) found a stronger relative age effect in goalkeeping and defence positions than in midfield and upfront positions.

In the present study no relative age effect was found in function of height or performance of the players. Schorer et al. (2009a) documented similar results in handball and determined that the cause of RAE is not related to either height, weight or technical abilities, as no differences were found between relatively older and younger players. Neither anthropometric factors nor physical performance in young footballers found a RAE (Carling et al., 2009; Hirose, 2009).

Conclusions

The relative age effect exists and is significant in the Basketball World Championships in both male and female U17 and U19 players. This effect diminishes as the age of the athletes increases, and disappears in U21.

In specific positions the effect of relative age is also significant, being in the male category clearer in the centres, power forwards and small forwards and less in point guards and shooting guards. In the female category the results do not support the existence of the relative age effect.

In the male category, differences were found in the height of the players with regard to the term of birth, although these differences do not coincide with those expected if the relative age effect exists. However, no significant differences were found in height in the female categories.

The performance of the players with regard to the birth term shows minimal differences in the male category in the U17, as the players born in the first two terms obtained better percentages in the three point field throws than the ones born in the last terms of the year. In the U19, the points obtained are fewer in players that belong to the third term than the rest of the terms. In the female category some differences were found in the U19 but they do not support the existence of the RAE and no differences were found in either U17 or U21.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Armstrong

(1966) A comparison of the performance of summer and autumn-born children at eleven and sixteen. British Journal of Educational Psychology 36: 72–76.

Ashworth

Heyndels

(2007) Selection bias and peer effects in team sports: The effect of age grouping on earnings of German soccer players. Journal of Sports Economics 8: 355–377.

Baker

Schorer

Cobley

(2010) Relative age effects: An inevitable consequence of elite sport? Sportwissenschaft 40: 26–30.

Barnsley

Thompson

(1988) Birthdate and success in minor hockey: The key to the NHL. Canadian Journal of Behavioural Science 20: 167–176.

Carling

le Gall

Reilly

. (2009) Do anthropometric and fitness characteristics vary according to birth date distribution in elite youth academy soccer players? Scandinavian Journal of Medicine & Science in Sports 19(1): 3–9.

Delorme

Raspaud

(2009) Is there an influence of relative age on participation in non-physical sports activities? The example of shooting sports. Journal of Sports Sciences 27(10): 1035–1042.

Delorme

Boiché

Raspaud

(2010) Relative age and dropout in French male soccer. Journal of Sport Sciences 28(7): 717–722.

Delorme

Chalabaev

Raspaud

(2011) Relative age is associated with sport dropout: Evidence from youth categories of French basketball. Scandinavian Journal of Medicine & Science in Sports 21(1): 120–128.

Esteva

Drobnic

Puigdellivol

. (2006) Fecha de nacimiento y éxito en el baloncesto profesional. Medicina de l’esport 41(149): 25–30.

10.

FIBA.com (2012) FIBA Archive. Aviable at: http://archive.fiba.com (accesed 13 May 2012).

11.

Freyman

(1965) Further evidence on the effect of date of birth on subsequent school performance. Educational Research 8: 58–64.

12.

Gibbs

Jarvis

Dufur

(2011) The rise of the underdog? The relative age effect reversal among Canadian-born NHL hockey players: A reply to Nolan & Howell. International Review for the Sociology of Sport. Epub ahead of print 22 August 2011. DOI: 1012690211414343.

13.

Grondin

Deshaies

Nault

(1984) Trimestres de naissance et participation au hockey et au volleyball. La Revue Québécoise de l’Activité Physique 2: 97–103.

14.

Gutiérrez

Saavedra

Contreras

. (2012) Influencia del año de nacimiento de una jugadora en las posibilidades de ser captada como talento en el balonmano femenino internacional. Apunts 108(2): 46–52.

15.

Helsen

Starkes

Van Winckel

(1998) The influence of relative age on success and dropout in male soccer players. American Journal of Human Biology 10(6): 791–798.

16.

Helsen

Starkes

Van Winckel

(2000) Effect of a change in selection year on success in male soccer players. American Journal of Human Biology 12: 729–735.

17.

Helsen

Van Winckel

Williams

(2005) The relative age effect in youth soccer across Europe. Journal of Sports Sciences 23: 629–636.

18.

Hirose

(2009) Relationships among birth-month distribution, skeletal age and anthropometric characteristics in adolescent elite soccer players. Journal of Sports Sciences 27(11): 1159–1166.

19.

Jullien

Turpin

Carling

(2008) Influence of birth date on the career of French professional soccer players. Science & Sports 23: 149–155.

20.

Lidor

Côte

Arnon

. (2010) Relative age and birthplace effects in division 1 players – Do they exist in a small country? Talent Development and Excellence 2(2): 181–192.

21.

Malina

(1994) Physical growth and biological maturation of young athletes. Exercise and Sport Sciences Reviews 22: 389–434.

22.

Malina

Bouchard

Bar-Or

(2004) Growth, Maturation, and Physical Activity. Champaign, IL: Human Kinetics.

23.

Musch

Hay

(1999) The relative age effect in soccer: Cross-cultural evidence for a systematic discrimination against children born in late competition year. Sociology of Sport Journal 16: 54–64.

24.

Nolan

Howell

(2010) Hockey success and birth date: The relative age effect revisited. International Review for the Sociology of Sport 45(4): 507–512.

25.

Roman

Fuchslocher

(2011) Influence of the selection level, age and playing position on relative age effects in Swiss Women’s Soccer. Talent Development and Excellence 3(2): 239–247.

26.

Russell

RJH

Startup

(1986) Month of birth and academic achievement. Personality and Individual Differences 7(6): 839–846.

27.

Schorer

Baker

Büsch

. (2009a) Relative age, talent identification and youth skill development: Do relatively younger athletes have superior technical skills? Talent Development and Excellence 1: 45–56.

28.

Schorer

Cobley

Büsch

. (2009b) Influences of competition level, gender, player nationality, career stage and playing position on relative age effects. Scandinavian Journal of Medicine & Science in Sports 19: 720–730.

29.

Simmons

Paull

(2001) Season-of-birth bias in association football. Journal of Sports Sciences 8: 677–686.