Modelling age-related changes in executive functions of soccer players

V.w.l. Chin, Adam Beavan, Job Fransen, Jan Mayer, Robert Kohn, Louise Ryan, Scott A. SissonarXiv 2021 · DOI · arXiv

Abstract

The widespread popularity of soccer across the globe has turned it into a multi-billion dollar industry. As a result, most professional clubs actively engage in talent identification and development programmes. Contemporary research has generally supported the use of executive functions - a class of neuropsychological processes responsible for cognitive behaviours - in predicting a soccer player's future success. However, studies on the developmental evolution of executive functions have yielded differing results in their structural form (such as inverted U-shapes, or otherwise). This article presents the first analysis of changes in the domain-generic and domain-specific executive functions based on longitudinal data measured on elite German soccer players. Results obtained from a latent variable model show that these executive functions experience noticeable growth from late childhood until pre-adolescence, but remain fairly stable in later growth stages. As a consequence, our results suggest that the role of executive functions in facilitating talent identification may have been overly emphasised.

Introduction: The Cognitive Dimension of Soccer

In the high-stakes world of professional soccer, identifying and nurturing talent is a multi-million dollar endeavor. While physical scouts have traditionally focused on speed, strength, and technical skill, modern clubs are increasingly turning to neuropsychology to gain a competitive edge. At the heart of this shift are executive functions (EF)—the cognitive processes responsible for decision-making, goal-oriented behavior, and motor inhibition.

The prevailing theory suggests that elite players possess superior EFs, allowing them to process complex on-field information faster than their peers. However, the developmental trajectory of these skills—how they evolve as a player matures—has been a subject of debate. Most existing research is cross-sectional, comparing different players at a single point in time. This study provides a critical advancement by analyzing longitudinal data, tracking the same elite youth players over several years to model how these "brain games" actually develop.

Methodology: A Latent Variable Approach

To analyze the developmental trajectories of 304 elite players, the researchers employed a latent variable model. This approach allowed them to link unobserved cognitive traits to a battery of ten observed performance variables.

The core of the analysis used a piecewise linear spline model with "knots" (change points) at ages 12, 15, and 18. This structure was chosen to capture the potentially different rates of cognitive growth across childhood, pre-adolescence, adolescence, and early adulthood. The model was implemented using Bayesian inference via Markov chain Monte Carlo (MCMC) methods, which provided a robust framework for handling missing data—a common occurrence in professional sports academies due to player transfers or injuries.

Crucially, the researchers modeled two distinct latent curves: one for domain-generic executive functions (such as basic reaction times and fluid intelligence) and one for domain-specific functions (technical-cognitive tasks performed in a soccer context).

Experimental Setup and Datasets

The study tracked 304 male soccer players from an elite German Bundesliga club between 2016 and 2019. The test battery was designed to differentiate between abstract cognitive skills and those applied directly to soccer movements.

Overview of the cognitive and motor tests: (a) Determination test, (b) Pre-cued choice response time task, (c) Helix test, and (d) Footbonaut.

The assessment tools included:

  • Generic Tests: Determination tests for perceptual-motor speed and stop-signal tasks for response inhibition.
  • Soccer-Specific Tests:
    • The Helix: A 180-degree curved screen where players must track multiple moving avatars, simulating the visual scanning required during a match.
  • The Footbonaut: A high-tech training cage where players receive a ball from a dispenser and must pass it into a randomly highlighted target gate as quickly and accurately as possible.

Key Findings: The Developmental Plateau

The results revealed a clear pattern of cognitive development that challenges the industry's heavy reliance on late-stage testing:

  • The 10-15 Growth Window: The most dramatic cognitive gains occur between late childhood and pre-adolescence. For domain-generic EFs, the growth rate peaked between ages 10 and 12 (posterior mean rate of 28.59) before slowing significantly.
  • The Plateau Paradox: Domain-specific EFs (measured by the Helix and Footbonaut) showed their highest growth during pre-adolescence (ages 12–15). However, after age 15, this growth dropped to nearly zero (0.11 and 0.07). This suggests that by mid-adolescence, these specific cognitive-technical skills have largely stabilized.
  • Independence of Skills: Perhaps the most surprising finding was that the growth rates of generic and specific EFs were largely independent. A player who excels at a computer-based reaction test is not necessarily superior in soccer-specific tasks like the Footbonaut.
  • Positional and Seasonal Effects: Goalkeepers generally performed worse on Footbonaut response times, likely because their training does not emphasize the same passing/receiving loops as outfield players. Additionally, players showed sharper response times in post-season assessments, suggesting an acute cognitive benefit from active participation in the sport.

Implications for Talent Identification

The observed plateau after age 15 suggests that the role of executive functions in talent identification for older adolescents may be overemphasized. If these traits have already stabilized by age 17 or 18, they may not be effective predictors of future professional success at that stage. Instead, the "window of opportunity" for both identifying and developing these cognitive foundations appears to be much earlier, between the ages of 10 and 15.

Conclusion and Future Directions

While this study provides a robust longitudinal look at EF development, the authors acknowledge that even sophisticated tests like the Footbonaut are still somewhat artificial. They lack the full "perception-action coupling" of a real match, where players must react to unpredictable human opponents. Future research should aim to measure executive functions during actual match play and explore more flexible non-parametric modeling techniques to further refine our understanding of the "soccer brain."

Summary updated

Abstract

The widespread popularity of soccer across the globe has turned it into a multi-billion dollar industry. As a result, most professional clubs actively engage in talent identification and development programmes. Contemporary research has generally supported the use of executive functions - a class of neuropsychological processes responsible for cognitive behaviours - in predicting a soccer player's future success. However, studies on the developmental evolution of executive functions have yielded differing results in their structural form (such as inverted U-shapes, or otherwise). This article presents the first analysis of changes in the domain-generic and domain-specific executive functions based on longitudinal data measured on elite German soccer players. Results obtained from a latent variable model show that these executive functions experience noticeable growth from late childhood until pre-adolescence, but remain fairly stable in later growth stages. As a consequence, our results suggest that the role of executive functions in facilitating talent identification may have been overly emphasised.
BibTeX
@misc{chin2021modelling,
  author = {V.W.L. Chin and Adam Beavan and Job Fransen and Jan Mayer and Robert Kohn and Louise Ryan and Scott A. Sisson},
  title = {{Modelling age-related changes in executive functions of soccer players}},
  year = {2021},
  eprint = {2105.01226},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2105.01226},
}

Loading related content...

Figure