Predictors of selection into an elite level youth football academy: A longitudinal study

ABSTRACT

Talent identification in football is complex. Research regarding talent indicators that predict selection into professional academies is increasingly multi-disciplinary, though is typically cross-sectional, failing to consider whether the rate of change of those indicators influences selection. The aim of this study was to determine if longitudinal changes in talent indicators are associated with selection into a professional football academy. A total of 110 Dutch male football players (aged 8–12) playing in a youth development programme of a professional club were included in the study. Players were assessed regularly on their anthropometry, physical fitness, gross motor coordination, technical ability, and psychosocial capacities over four years. A subset of players was then selected into the academy. Cross-sectional and longitudinal analyses revealed the indicators that best predicted selection. The best indicator was the 30 m sprint speed, with prediction not improving when including other predictors or their rate of change. The individuals that the club ultimately selected at age 12 could have been predicted well above chance levels using the sprint speeds at age nine or ten. The relative consistency of the rate of improvement in indicators across participants meant that the rate at which they developed played little role in selections.

KEYWORDS:

• Talent identification
• talent development
• youth football
• youth soccer
• longitudinal analysis

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ A total of 16 players dropped out before the end of the training program (i.e., at transition to U13). A total of 201 individual measures of at least one talent indicator were available for the players who dropped out. Those players appeared to have missed more test opportunities than expected. Out of the 4,221 possible data values (21 talent indicators*201 assessments), only 972 values were available (23.1%). In comparison, in our wider sample, out of 33,894 possible values (21 talent indicators*1614 assessments), 23,611 values were available (69.7%). Large age gaps were also evident between the measurements for many of the players who dropped out of the program.

² The athletic skills parkour and the KTK do appear to measure different constructs. Results show mild-moderate correlations within each method (i.e., the KTK measures correlate with each other at a magnitude ranging between ∣r∣ = .32 and ∣r∣ = .38, and the ASP measures correlate together between ∣r∣ = .12 and ∣r∣ = .52), yet there are largely negligible correlation values between the KTK and ASP measures (i.e., between ∣r∣ = .005 and ∣r∣ = .12).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.