Journal of Sports Sciences, October 2005; 23(10): 1003 – 1012

Evaluation of the ‘‘under-21 rule’’: Do young adult soccer players

benefit?

ROEL VAEYENS1, AARON COUTTS2 & RENAAT M. PHILIPPAERTS1

1
Department of Movement and Sports Sciences, Ghent University, Ghent, Belgium, and 2School of Leisure, Sport and
Tourism, University of Technology, Sydney, Lindfield, NSW, Australia

(Accepted 18 July 2004)

Abstract
Young adult soccer players often encounter difficulty progressing from youth competition to being regularly selected for
high-level, open-age competition. In an attempt to increase the playing opportunities of semi-professional and amateur
under-21 players in first teams, the Royal Belgian Football Association required national division teams to include at least
two players younger than 21 in their match selection (under-21 rule). Over four seasons, the following variables were
analysed across 2138 semi-professional and amateur soccer players aged 16 – 39 years: (1) the number of times a player was
selected to be in the first team squad; (2) the number of times a player was selected to play in the starting line-up; and (3) the
number of minutes played. Chi-square analysis demonstrated that second and third division teams had complied with the
new selection regulations. However, two-way analysis of variance of individual player data revealed no increase in the
number of playing minutes in the under-21 group. It would appear that many teams had complied with the new regulations
by selecting young players as substitutes. The results highlight the difficulties that talented young soccer players experience
progressing from youth to senior competition. We conclude that the rule imposed by the Royal Belgian Football Association
has failed to improve the playing opportunities of young adult soccer players. Alternative strategies for increasing playing
opportunities for young talented players are required.

Keywords: Soccer, sports regulation, development and talent selection

renewed interest in talent identification and devel-

Introduction
Many soccer clubs now function in a similar way to
global corporations. Since the financial success of
professional soccer clubs is largely reliant on their
on-field success, the value of high-quality players to
professional clubs is substantial. However, for many
soccer clubs the likelihood of recruiting high-quality
players is now reduced due to the inflated transfer
fees and associated wages necessary to obtain their
services. As a result, professional soccer clubs now
have considerable interest in identifying and devel-
oping their own young talented players.
The pursuit of excellence in young athletes can be
divided into four key stages: (1) detection, (2)
identification, (3) development and (4) selection
(Williams & Franks, 1998; Williams & Reilly,
2000a). Recently, there has been increased research
interest in the pursuit of excellence in young soccer
players, with most studies focusing on identification
and development of talented soccer players. The
opment processes has been suggested to be the
stimulus for the recent improvement in performance
and accompanying reduction in age of elite perfor-
mers. The major contributing factors to this
increased interest in player talent identification and
development are the ‘Bosman Ruling’, which pre-
cludes clubs from withholding a player’s registration
at the completion of his contract, and the escalation
of the transfer costs for soccer players (Williams &
Franks, 1998). These two factors have made it more
difficult for team managers to maintain continuity in
a club’s first team squad, which can subsequently
jeopardize the stability of team performances. As a
result of these factors and to maintain their sporting
and financial status, professional clubs are now more
aware of the importance of identifying and develop-
ing their own talented youth players with the ability
to play in the first team.
In many countries, science-based support systems
(e.g. psychological and social counselling, training

Correspondence: R. Vaeyens, Department of Movement and Sports Sciences, Faculty of Medicine and Health Sciences, Ghent University, Watersportlaan 2,
B-9000 Ghent, Belgium. E-mail: roel.vaeyens@ugent.be
ISSN 0264-0414 print/ISSN 1466-447X online ª 2005 Taylor & Francis
DOI: 10.1080/02640410400023266
1004 R. Vaeyens & R. Philippaerts
and fitness programmes, computer-based match
analysis) are now a fundamental feature in athlete
preparation. Despite the increase in scientific sup-
port for athletes, many existing talent identification
programmes for soccer players have a relatively poor
scientific foundation (Moore et al., 1998; Williams &
Reilly, 2000b).
Most research on the factors that influence player
selection has been limited to the relative age effect
(Barnsley, Thompson, & Legault, 1992; Brewer,
Balsom, & Davis, 1995; Dudink, 1994; Verhulst,
1992; for a review, see Musch & Grondin, 2001).
The results of these previous studies demonstrated a
consistent asymmetry in the birth-date distribution
of professional senior soccer players. In particular,
birth-date distributions were significantly biased
towards a higher number of players born during the
early part of the selection year. Similar studies of
youth soccer indicated that players born at the onset
of the selection year, beginning in the 6 – 8 year
group, were more likely to be identified as talented
and to be exposed subsequently to higher levels of
coaching (Barnsley et al., 1992; Baxter-Jones, 1995;
Brewer, Balsom, Davis, & Ekblom, 1992; Brewer et
al., 1995; Helsen, Starkes, & Van Winckel, 1998).
Eventually, these players were more likely to be
transferred to higher-level teams, to play for their
national side and to become involved in the game
professionally (Helsen et al., 1998). In contrast,
players born late in the selection year tended to drop
out as early as 12 years of age (Helsen et al., 1998).
Kuhn (2002) reported that young talented profes-
sional soccer players often find it difficult to make the
first team. The abandonment of the quota limiting
the number of foreign players in domestic European
Union football leagues has restructured European
soccer. The identification and selection of elite
senior players are producing migration patterns that
appear to be having a negative impact on indigenous
player development (Littlewood, Richardson, Lees,
& Peiser, 2001; Maguire & Pearton, 2000).
In Belgium, young adult professional, semi-pro-
fessional and amateur soccer players were given few
playing opportunities. In response to this trend, the
Royal Belgian Football Association (in collaboration
with FIFA, the International Football Association)
introduced the ‘‘under-21 rule’’ for the 1999 – 2000
competitive soccer season (i.e. the season commen-
cing in 1999 and finishing in the year 2000). This
directive required all second and third national
division teams to include at least two players younger
than 21 (on the day of the match) in their match
selection. A match selection consists of 15 players:
11 starting players and four substitutes, of whom
three can play. For the 2000 – 2001 season, this
selection rule was relaxed to require a minimum
match selection of two players younger than 21 years
on 1 January of the previous competitive soccer
season. Additionally, the under-21 rule was also
modified to allow players who were aged 21 years on
1 January to be considered as being under-21 years
for the whole of that soccer season. Finally, to assist
with the enforcement of the under-21 rule, a system
of fines was also introduced for infringements.
No previous studies have examined the influence
of these rule changes on actual selection practices.
The aims of the present study were: (1) to examine
whether semi-professional and amateur soccer teams
complied with the under-21 rule (Study 1); and (2)
to determine if the under-21 rule was effective in
increasing the playing opportunities of young adult
(under-21) soccer players (Study 2). The objective of
Study 1 was addressed in an examination of team
selection variables before and after the introduction
of the under-21 rule. Study 2 focused on the
development of the number of under-21 players as
well as their individual selection opportunities and
playing times.
Methods
Sample
The official match data for the second and third
national division league games of the last four
competitive seasons (1998 – 1999, 1999 – 2000,
2000 – 2001 and 2001 – 2002) were collected from
the official statistics provided by the Royal Belgian
Football Association. Data collected during the
1998 – 1999 season (season 1) were used as control
baseline data, since the rule had not been imple-
mented at that time. In total, 53 different soccer
teams selected 2138 different soccer players aged
16 – 39 years.
Study 1: Application of the under-21 rule
Procedures. In the second national division league, 18
teams played on 34 match days. In the third national
division league, 16 teams played on 30 match days.
For every match day, the number of under-21
players (in accordance with the rule) selected,
starting and playing for each team was recorded.
The number of selected players was taken to be the
under-21 players who were selected for a game.
These selected players did not have to participate in
the game to be included in this category. The
number of starting players was taken to be the
number of under-21 players selected to play in the
starting eleven. The number of playing players was
taken to be the number of under-21 players who
played at least one minute in the game (i.e. these
players may have been substituted into or out of the
game). With these factors known, it was possible to
 Evaluation of the ‘‘under-21 rule’’ in Belgian soccer
calculate the number of under-21 substitutes and
players who remained on the bench during the entire
game (bench 90 min).
Data analysis. Descriptive statistics for each team
variable were calculated for both divisions over four
consecutive seasons. Chi-square tests were used to
evaluate seasonal differences (i.e. differences be-
tween the consecutive competitive soccer seasons) in
the frequency-distribution pattern of selected under-
21 players per team and per match and to examine
changes in total ‘‘team’’ values for both leagues over
the course of the four soccer seasons.
Study 2: Effectiveness of the under-21 rule
Procedures. To evaluate the effectiveness of the
under-21 rule, the distribution of the numbers of
players aged under and over 21 years were com-
pared. In the present study, the players were
classified into these age groups according to their
age at the beginning of the respective calendar year
(i.e. 1 January of the previous competitive season).
This classification is different to that used in Study 1,
where players were classified as being aged under 21
or 21 + according to the definition of the under-21
rule. Consequently, the age groups of Studies 1 and
2 are not the same in the first two seasons.
Initial data analysis was performed to determine if
there was a relative increase of the proportion of
under-21 players with the introduction of the under-
21 rule. Therefore, for each season, all players
selected for at least one game were divided into
two age groups, under and over 21 years, based on
their age at the beginning of the respective year (i.e. 1
January of the previous competitive season).
To evaluate the effectiveness of the under-21 rule,
individual player data for the 1962 soccer players
who appeared in the second division and for the 1666
players who appeared in the third division over the
past four seasons were also examined. Therefore, the
individual season totals for selections and time
played were analysed. Selection was defined as a
player who was selected for a game either as a
starting player or as a substitute. Time played refers
to the number of minutes that a player played in a
match.
Data analysis. Chi-square analysis was used to
evaluate changes in the proportion of players in both
the second and third divisions during each compe-
titive season.
Descriptive statistics based on the individual data
were calculated for the under-21 players. In addition,
a two-way analysis of variance (ANOVA) was used to
identify significant interaction effects between age
group and season for duration of play. Post-hoc
1005
analyses were performed using a Tukey test. Statis-
tical significance was accepted at P 5 0.05 unless
otherwise indicated.
Results
Study 1
Before the implementation of the under-21 rule,
second division teams selected two or more
under-21 players in 67.8% of league matches
compared with 76.0% in the third division. In
season 2 (1999 – 2000), this proportion increased
to 84.0% for second division teams and to 93.7%
for third division teams. During the last two
competitive years analysed, more than 96% of
second division games and more than 99% of
third division games were played in accordance
with the under-21 rule.
Figures 1 (second division) and 2 (third division)
illustrate the changes in the proportion of selected
under-21 players within match selections over the
study period. Chi-square analysis revealed signifi-
cantly changed frequencies between season 1 and
season 2 (division 2: w25 = 165.89, P 5 0.001;
division 3: w25 = 114.15, P 5 0.001). For both
divisions, the prevalence of two selected under-21
players per team and per game increased in season
2. This trend was most pronounced in the second
division, with an increase from 7.2% to 33.2%. The
selection of less than two under-21 players per
match decreased as a result of this. However, the
curves for the second and third divisions show that
a somewhat different time course of change
occurred between the two divisions. In third
division teams, five or more under-21 players were
selected more often since the implementation of the
under-21-rule (Figure 2). In contrast, second
division teams tended to select three or more
under-21 players less often following the introduc-
tion of the under-21 rule (i.e. in season 2). The
selection of three or more under-21 players per
team increased from season 3 onwards (Figure 1).
Descriptive statistics for the team variables are
summarized in Table I. A yearly increase (with the
exception of season 4 in the second division) was
observed in all mean team variables in both divisions.
These changes appeared, however, to be more
pronounced in the third division.
The mean number of under-21 players selected by
a second division team on a match day basis
increased from 3.2 in season 1 to 4.1 in season 4.
In the third division, this value was doubled from 2.6
to 5.1 under-21 players within 4 years. In both
divisions, on average more than two under-21
players were in the starting line-up in seasons 3 and
4. The number of under-21 substitutes also in-
1006 R. Vaeyens & R. Philippaerts

Figure 1. Frequencies of selected under-21 players per team and per game: Division 2.

Figure 2. Frequencies of selected under-21 players per team and per game: Division 3.

creased to almost 2 per team per match in the second
division. Table I shows that the largest increase in
under-21 substitutes in second division teams
occurred during season 2. Over the 4-year period,
third division coaches also selected more young adult
substitutes per game. Compared with season 1, more
under-21 players per team played during season 3
and season 4 (second division: 3.1; third division:
4.0). Furthermore, there was a tendency for an
increase in the mean number of selected under-21
players who did not play.
Tables II (second division) and III (third division)
present the total values of the team variables by
season over the study period. For both divisions, chi-
square analysis revealed significantly changed fre-
quencies between most seasons. In the second
division, the implementation of the under-21 rule
resulted in significant changes between season 1 and
2 for the ‘‘selected’’ (w21 = 9.0, P 5 0.005), ‘‘sub-
stitute’’ (w21 = 34.5, P 5 0.001) and ‘‘bench
90 min’’ (w21 = 27.6, P 5 0.001) variables; no sig-
nificant changes were observed for the ‘‘starting’’
(w21 = 0.6, P = 0.449) and ‘‘playing’’ (w21 = 1.5,
P = 0.217) variables. The w21-values were highest
when season 1 was compared with season 3,
indicating that the greatest increase was seen after
season 2. In contrast, for the third division, there were
significant differences for all team variables immedi-
 Evaluation of the ‘‘under-21 rule’’ in Belgian soccer 1007

Table I. Number of under-21 players per competitive game and team: Divisions 2 and 3 of the Belgian League (mean + s)

Season 1 Season 2 Season 3 Season 4

Variable Division (1998 – 1999) (1999 – 2000) (2000 – 2001) (2001 – 2002)

Selected 2 3.2 + 2.3 3.5 + 2.5 4.4 + 2.2 4.1 + 2.1

3 2.6 + 1.5 3.5 + 2.2 4.7 + 2.0 5.1 + 2.1

Starting 2 1.8 + 1.7 1.7 + 1.9 2.5 + 1.7 2.1 + 1.7

3 1.0 + 1.0 1.7 + 1.8 2.3 + 1.7 2.7 + 1.7

Substitute 2 1.4 + 1.1 1.8 + 1.1 2.0 + 1.1 1.9 + 1.0

3 1.6 + 1.0 1.8 + 0.9 2.3 + 0.9 2.4 + 1.1

Playing 2 2.5 + 2.0 2.6 + 2.3 3.6 + 2.1 3.1 + 2.1

3 1.9 + 1.4 2.6 + 2.2 3.5 + 2.1 4.0 + 2.1

Bench 90 min 2 0.7 + 0.7 0.8 + 0.8 0.8 + 0.7 0.9 + 0.7
3 0.8 + 0.7 0.9 + 0.7 1.2 + 0.8 1.1 + 0.8

Table II. Total frequencies of the five team variables per season: Division 2

w21

Age Season 1 Season 2 Season 3 Season 4 Seasons Seasons Seasons Seasons Seasons Seasons
Variable group (1998 – 1999) (1999 – 2000) (2000 – 2001) (2001 – 2002) 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4

Selected Under-21 1947 2116 2707 2462 9.0** 166.3*** 83.9*** 98.2*** 38.0*** 14.0***
21 + 7233 7064 6473 6657

Starting Under-21 1075 1043 1512 1282 0.6 N.S. 91.4*** 23.7***106.3*** 31.6*** 22.0***
21 + 5657 5689 5220 5406

Substitute Under-21 872 1073 1195 1180 34.5*** 87.4*** 83.5*** 12.2*** 10.9** 0.0 N.S.
21 + 1576 1375 1253 1251

Playing Under-21 1533 1603 2211 1905 1.5 N.S.151.4*** 47.2***123.1*** 31.9*** 30.0***
21 + 6613 6584 6009 6307

Bench 90 Under-21 414 513 496 557 27.6*** 27.1*** 88.3*** 0.0 N.S.18.3*** 18.0***
21 + 620 480 464 350

** P 5 0.005; *** P 5 0.001.

ately after the introduction of the rule. The latter
results also demonstrate a more gradual increase of
the observed variables.
Study 2
Table IV shows that the 21 + group comprised the
greatest proportion of players in all 4 years examined
in both the second and third divisions. There were no
significant changes over the four seasons in the
distribution of both age groups for both divisions.
In the second division, there was a progressive but
non-significant (w23 = 1.518, P = 0.678) decrease in
the proportion under-21 players selected over the 4-
year period. The proportion of under-21 players
selected in the second division decreased from 43.5%
in year 1 to only 40.0% in year 4. In contrast, the
relative numbers of under-21 players in the third
division increased compared with season 1
(w23 = 6.809, P = 0.078).
The mean number ( + s) of under-21 ‘‘selections’’
in the second division decreased from 13.6 + 10.5 in
season 1 to 12.8 + 11.0 in season 4. There was also a
concomitant decrease in the average playing time of
these players from 732 + 840 to 621 + 846 min.
The total playing time for the under-21 group was
approximately 1000 min less than for the 21 + group
(Table V). Statistical analysis revealed no significant
interaction effect between season and age group.
Results for the third division showed similar trends.
Discussion
This was the first study to examine the influence of
age-based selection regulations on selection strate-
gies in open-age divisional soccer competition. The
1008 R. Vaeyens & R. Philippaerts

Table III. Total frequencies of the five team variables per season: Division 3

w21

Variable Age Season 1 Season 2 Season 3 Season 4 Seasons Seasons Seasons Seasons Seasons Seasons
group (1998 – 1999) (1999 – 2000) (2000 – 2001) (2001 – 2002) 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4

Selected Under-21 1248 1699 2221 2451 86.2*** 364.0*** 526.2*** 98.6*** 192.2***15.6***
21 + 5944 5500 4948 4743

Starting Under-21 485 819 1114 1311 97.6*** 294.8*** 457.7*** 56.7*** 142.4***19.7***
21 + 4795 4461 4144 3969

Substitute Under-21 763 880 1107 1140 13.8*** 124.2*** 147.8*** 55.9*** 72.2*** 1.1 N.S.
21 + 1149 1039 804 774

Playing Under-21 887 1249 1665 1903 70.4*** 303.8*** 474.4*** 85.2*** 187.3***19.9***
21 + 5494 5181 4670 4474

Bench 90 Under-21 361 450 556 548 31.0*** 81.8*** 84.0*** 11.4** 12.4*** 0.0 N.S.
21 + 450 319 278 269

** P 5 0.005; *** P 5 0.001.

Table IV. Number of players (and percentage) per age group and competitive season: Divisions 2 and 3

Competitive season

Season 1 Season 2 Season 3 Season 4

Age group (1998 – 1999) (1999 – 2000) (2000 – 2001) (2001 – 2002)

Division 2 Under-21 211 (43.5%) 218 (43.3%) 208 (42.2%) 192 (40.0%)
21 + 274 (56.5%) 286 (56.7%) 285 (57.8%) 288 (60.0%)
Total 485 504 493 480

Division 3 Under-21 154 (38.8%) 201 (46.2%) 174 (43.0%) 201 (46.9%)
21 + 243 (61.2%) 234 (53.8%) 231 (57.0%) 228 (53.1%)
Total 397 435 405 429

Note: Division 2, w23 = 1.518 (P = 0.678); Division 3, w23 = 6.809 (P = 0.078).

Table V. Number of selections (Sel) and minutes played (Min) per player and per season: Divisions 2 and 3 (mean + s)

Effect

Season 1 Season 2 Season 3 Season 4 Age group

Division Age group (1998 – 1999) (1999 – 2000) (2000 – 2001) (2001 – 2002) Age group Season 6 Season

Min 2 Under-21 732.2 + 840.1 605.7 + 804.0 672.4 + 830.6 620.5 + 845.9 F = 574.1 F = 0.4 F = 0.7
2 21 + 1637.4 + 942.2 1646.9 + 1019.2 1625.6 + 908.6 1669.0 + 923.0 *** N.S. N.S.
3 Under-21 610.1 + 742.7 615.0 + 800.6 596.4 + 750.0 600.8 + 766.9 F = 543.6 F = 0.2 F = 0.4
3 21 + 1559.3 + 855.3 1495.0 + 850.3 1593.2 + 839.3 1545.6 + 861.8 *** N.S. N.S.

Sel 2 Under-21 13.6 + 10.5 12.5 + 10.4 13.0 + 10.9 12.8 + 11.0
2 21 + 23.1 + 9.6 22.6 + 10.5 22.7 + 9.3 23.1 + 9.5
3 Under-21 12.9 + 9.8 12.1 + 10.0 12.8 + 10.2 12.2 + 9.7
3 21 + 21.4 + 8.6 20.4 + 8.7 21.4 + 8.2 20.8 + 8.4

*** P 5 0.001.

intention of the under-21 rule was to increase the present results suggest that the introduction of the
playing opportunities and the professional develop- new regulations did not have the desired outcome on
ment of young, local soccer players. However, the playing opportunities for under-21 players.
 Evaluation of the ‘‘under-21 rule’’ in Belgian soccer
Study 1: Application of the under-21 rule
The under-21 rule was the first regulation requiring
young players to be selected by clubs in the Belgian
second and third divisions. Prior to this (season
1998 – 1999), an average of more than two young
players (i.e. under-21 years) were selected for every
match (Table I). However, there was large variation
in the selection policies of many teams (Table I) and
a considerable number of squads did not select any
young players for competitive matches (Figures 1
and 2). Following the introduction of the under-21
rule, the number of under-21 players being selected
increased, showing that teams had attempted to
comply with the new regulations. During the final
two seasons examined, more than 96% of matches
complied with the under-21 rule (Figures 1 and 2).
The introduction of the under-21 rule resulted in
changes in selection patterns by both second and
third division teams (Tables II and III). Notably,
however, only the proportion of ‘‘two’’ selected
under-21 players increased during the second
competitive year (1999 – 2000) in the second division
(Figure 1). This suggests that the teams in this
division were only interested in meeting the mini-
mum requirements of the new selection rule and not
the underlying intention of increasing markedly the
participation of young players. Further support for
this argument is the finding that there were no
significant changes in the team variables for ‘‘start-
ing’’ and ‘‘playing’’ under-21 players between season
1 and 2 among second division teams (Table II). A
possible explanation for this result may be that there
was a late communication of the alteration in
regulations to the soccer teams (approximately 1
month before the transfer deadline). This late
communication may not have allowed sufficient time
for second division teams to recruit local under-21
players with the appropriate skill and this may have
consequently reduced their ability to make appro-
priate team selection adjustments.
The significant increase in the ‘‘selection’’, ‘‘sub-
stitute’’ and ‘‘bench 90 min’’ variables during season
2 suggests that these teams selected under-21 players
as ‘‘substitutes’’ so they could comply with the new
rule (Table II). This strategy allowed teams to avoid
the underlying intention of the under-21 rule, as the
‘‘starting’’ teams still consisted mostly of players
aged 21 + . Interestingly, previous investigators have
reported similar strategies by sporting teams in
response to changes in sports regulations (Helsen,
Starkes, & Van Winckel, 2000). For example, a shift
in the start of the selection year (from August to
January) for youth soccer did not eliminate the
relative age effect. Helsen et al. found that a change
in the start of the selection year in youth soccer
players significantly biased the birth-date distribu-
1009
tions of the players towards the early part of the new
selection year. These findings are in line with the
present results, in that teams either deliberately or
inadvertently adjusted selection procedures to re-
duce the impact of changes in regulations on
performance.
The most marked increase of measured variables
for under-21 players competing in the second
division teams was seen in season 3. This result
suggests that the under-21 rule was adopted more
completely by the second division teams during the
third season (i.e. 1 year after the introduction of the
new rule). The reduction seen during season 4 may
indicate that this rule had been successfully imple-
mented and that teams had reached a saturation
point of being able to select under-21 players for
competitive matches.
The results show that almost all third division
teams immediately applied the rule upon its intro-
duction (Figure 2). It is most likely that the threat of
financial penalties may have had a greater effect on
these teams than the more financially sound second
division teams. Additionally, the lower initial mean
team values for the third division may also explain
why the most marked increase occurred during
season 2. In contrast to the second division, third
division team values increased consistently over the
period of observation (Tables I and III). It is also
probable that the relaxation in the definition of an
under-21 player from season 3 may have led to the
higher ‘‘team values’’ in the last two competitive
seasons.
Study 2: Effectiveness of the under-21 rule
The present results show that the proportion of
under-21 players selected to play in competitive
matches increased, although not significantly so, in
the third division (Table IV). This result, in
combination with the finding of an almost even
selection ratio of young (under-21) to old (21 + )
players in season 4 [second division: 0.4:0.6; third
division: 0.47:0.53 (see Table IV)] and the high level
of compliance (96%) with the under-21 rule may
lead one to the conclusion that the under-21 rule had
been successfully implemented. However, this is not
the case; other data show that although there was an
increase in selection rates for younger players, the
opportunities for them to play did not increase.
Closer inspection of the combined selection data
from the two divisions shows that 36% of under-21
players were selected five times or less during a
competitive year, while 48% of under-21 players did
not play in more than five league matches. These
results suggest that despite high selection rates in the
group of under-21 players, individuals were not
being selected consistently and that these players
1010 R. Vaeyens & R. Philippaerts
were rarely provided with playing opportunities when
selected. In addition, the large standard deviations in
the variables of ‘‘selected’’ and ‘‘playing’’ under-21
players per team and per game (Table I) highlight the
large variability in the playing opportunities for the
younger age group. The irregular distribution of
‘‘selections’’ and ‘‘playing time’’ in the under-21 age
group can be explained by the rotation system that
many teams used to comply with the rule. Under this
rotation system, the coaches often selected different
under-21 players as ‘‘substitutes’’ on a rotating
schedule, thus distributing the playing opportunities
over a larger number of under-21 players. For
instance, this was clearly evident in season 2, where
players who were 20 years at the beginning of the
season were rarely selected after their twenty-first
birthday as under the original definition of the
under-21 rule they were not considered to be
‘‘under-21’’ after their twenty-first birthday.
In general, young adult semi-professional and
‘‘amateur’’ players did not have many playing
opportunities. The present results are in line with
previous literature for professional senior soccer
(Kuhn, 2002; Littlewood et al., 2001), which
suggests that young talented professional soccer
players often experience difficulties in making first
team appearances. Additionally, immediately after
the introduction of the under-21 rule, foreign players
represented a small but important proportion of
under-21 players (mostly in the second division).
The influx in young foreign players further reduced
the playing opportunities for the indigenous youth in
this study. This result is in accordance with previous
studies which reported that an increase in foreign
players can reduce the opportunities to develop
indigenous sporting talent (Kuhn, 2002; Littlewood
et al., 2001; Maguire & Pearton, 2000). For example,
Kuhn (2002) recently argued that a pronounced
increase in mediocre foreign players precluded
indigenous talented individuals from integrating into
first team soccer.
The present results show that although the
selection rates of under-21 players increased after
the introduction of the under-21 rule, many teams
avoided playing younger players by selecting young
players as substitutes (especially substitute goal-
keepers). Notably, after the introduction of the new
rule, the proportion of under-21 goalkeepers in-
creased by nearly 50%. However, these youngsters
never played competitively, as almost 80% of them
were selected as one of the substitutes.
The thesis that younger players benefited from the
new soccer rule could not be supported by the
individual data. On average, neither the selections
nor the time played by an under-21 player increased
over the 4-year period (Table IV). This suggests that
soccer federations must remain vigilant of the effects
of introducing new rules, follow up the effectiveness
of new rules and reconsider new directives. Recently,
the German Regionalliga (i.e. the third and lower
divisions in German football) implemented a similar
regulation (three under-24 players and one under-21
player per game). Similarly, French senior clubs are
restricted to a maximum of 19 contract players over
the age of 21 but may have an unlimited number of
under-21s (Maguire & Pearton, 2000). To date, the
effectiveness of these rules in improving participation
by young soccer players in competitive matches has
not been examined. Nevertheless, the present results
demonstrate that imposing rules is not always the
most effective method for solving such problems.
An examination of the reasons underlying the
selection (or not) of players aged under-21 years is
warranted. One explanation is that the most gifted
youngsters had already been transferred to better
sides and had been given the opportunity to play in
those first teams. However, further analysis revealed
that less than 10 under-21 players were transferred
annually from lower division teams to the highest
national division in Belgium or abroad. Interestingly,
this analysis also demonstrated that these transferred
players were still not regularly selected to play in
these leagues. This suggests that other factors may
explain the limited playing opportunities for young
players in these divisions.
A second possible explanation for the reduced
playing opportunities within the under-21 group in
this study may originate from the ability of the
coaching and management personnel within clubs to
foster their existing young talent. Many soccer teams
in this study appeared either to lack the ability to
develop young native players or were reticent to
develop local youth talent to the level that allows
under-21 players to be integrated in the first team.
To fulfil the selection quota of two under-21 players,
many clubs transferred under-21 players from higher
to lower divisions. By adopting this recruitment
approach, young players were effectively transferred
down playing divisions (e.g. from the second to the
third division) and this allowed clubs to avoid the
underlying intention of the under-21 rule. This
transfer strategy seems to be in direct contrast to
the intended spirit of the under-21 rule, which was
developed to increase playing opportunities in
higher-level teams for young players.
A third possible reason for the reduced playing
opportunities of the younger soccer players is that
coaches are somewhat reluctant to select young
soccer players for first team matches due to their lack
of match experience. Even in the second and third
divisions, clubs are dependent upon playing success
for increased revenue. For this reason, and some-
times despite the longer-term benefits of developing
younger players, a manager will select players who
 Evaluation of the ‘‘under-21 rule’’ in Belgian soccer
will maximize revenues for that club in the short
term. When a choice has to be made between a
younger and an older player, often the latter will be
selected on account of his or her experience. There is
a contradiction here, because playing competitive
matches is the only way youngsters can develop their
soccer skills and gain experience. However, this
developmental process takes time and in the mean-
time young players may make poor decisions during
matches that can influence the outcome of a game.
Consequently, most coaches and team managements
prefer to play safe and apply short-term selection
policies. Most team managements evaluate the
success of their coach based on the results and
league position of the first team.
In summary, there appear to be several plausible
explanations for the limited playing opportunities
afforded to young soccer players. These include:
most talented young players are transferred to
higher leagues at an earlier age; the poor standard
of player management in clubs; and coaches are
reticent to play youth players due to their
inexperience in competitive matches. It is probably
a combination of the latter two that underlies the
lack of selection of players aged under-21 in the
Belgian leagues.
Conclusions
In the final two seasons examined in this study,
almost all second and third division games were
played in conformity with the under-21 rule. How-
ever, despite this, the present results show that the
underlying intention of the under-21-rule was not
embraced by the teams in these two divisions. It
appears that these teams were able to alter their
selection strategies so that the under-21 players were
most likely to be selected as substitutes and rarely
participated in competition. Indeed, the present
results reveal that the number of selections and
minutes played by these younger athletes decreased
slightly after the introduction of the under-21 rule.
Many teams were able to evade the underlying
intention of the new rule while still conforming with
it by adopting a rotation system whereby under-21
players were selected as substitutes on a rotating
basis. Additionally, the present results show that
young players were only selected as substitutes in the
first team and only for brief periods. The combina-
tion of these selection strategies allowed the teams to
meet the under-21 rule requirements without in-
creasing the first team playing opportunities of young
players.
Soccer federations and clubs have an important
role to play in the progresion of players from youth to
senior competition. A change in the approach of
soccer teams is required so that young talented
1011
players can optimize their performance in a stimulat-
ing environment. We suggest that soccer clubs,
leagues and national teams will only improve if more
playing opportunities are afforded to young players,
and there is guaranteed investment in qualitative
talent identification and development programmes
for young indigenous players. Our results highlight
the difficulties that young players experience in trying
to progress from youth to senior competition. It is in
this period of player development that soccer loses
many talented individuals. New guidelines need to
be formulated that will improve the playing oppor-
tunities of young adult soccer players.
Acknowledgements
The authors would like to thank the Administration
of the Royal Belgian Football Association for
providing the data.
References
Barnsley, R. H., Thompson, A. H., & Legault, P. (1992). Family
planning: Football style. The relative age effect in football.
International Review of Sport Sociology, 27, 77 – 87.
Baxter-Jones, A. D. G. (1995). Growth and development of young
athletes: Should competition levels be age related? Sports
Medicine, 20, 59 – 64.
Brewer, J., Balsom, P., & Davis, J. (1995). Seasonal birth
distribution amongst European soccer players. Sports Exercise
and Injury, 1, 154 – 157.
Brewer, J., Balsom, P., Davis, J., & Ekblom, B. (1992). The
influence of birth date and physical development on the
selection of a male junior international soccer squad. Journal
of Sports Sciences, 10, 561 – 562.
Dudink, A. (1994). Birth date and sporting success. Nature, 368,
592.
Helsen, W. F., Starkes, J. L., & Van Winckel, J. (1998). The
influence of relative age on success and drop out in male soccer
players. American Journal of Human Biology, 10, 791 – 798.
Helsen, W. F., Starkes, J. L., & Van Winckel, J. (2000). Effect of a
change in selection year on success in male soccer players.
American Journal of Human Biology, 12, 729 – 735.
Kuhn, W. (2002). Changes in professional soccer in Germany
since 1990. In W. Spinks, T. Reilly, & A. Murphy (Eds.),
Science and Football IV: Proceedings of the Fourth World Congress
of Science and Football (pp. 421 – 430). London: Routledge.
Littlewood, M., Richardson, D., Lees, A., & Peiser, B. (2001).
Migration patterns in top level English football. Insight – The FA
Coaches Association Journal, 3(4), 40 – 41.
Maguire, J., & Pearton, R. (2000). The impact of elite labour
migration on the identification, selection and development of
European soccer players. Journal of Sports Sciences, 18, 759 –
769.
Moore, P. M., Collins, D. J., Burwitz, L., Tebbenham, D.,
Abbott, A., & Arnold, J. (1998). Identification and develop-
ment of talent in selected UK sports. Journal of Sports Sciences,
16, 23.
Musch, J., & Grondin, S. (2001). Unequal competition as an
impediment to personal development: A review of the relative
age effect in sport. Developmental Review, 21(2), 147 – 167.
1012 R. Vaeyens & R. Philippaerts
Verhulst, J. (1992). Seasonal birth distribution of West European
soccer players: A possible explanation. Medical Hypotheses, 38,
346 – 348.
Williams, A. M., & Franks, A. (1998). Talent identification in
soccer. Sports Exercise and Injury, 4, 159 – 165.
Williams, A. M., & Reilly, T. (2000a). Talent identification and
development in soccer. Journal of Sports Sciences, 18, 657 – 667.
Williams, A. M., & Reilly, T. (2000b). Searching for the stars.
Journal of Sports Sciences, 18, 655 – 656.