Hernandez, Julia ORCID: https://orcid.org/0009-0001-2948-5174, Widmer, Chantal ORCID: https://orcid.org/0009-0006-6209-8770, Abbott, Shaun ORCID: https://orcid.org/0000-0002-6111-1033, Cobley, Stephen, Born, Dennis-Peter ORCID: https://orcid.org/0000-0002-1058-4367, Kern, Raphael, Tschopp, Markus ORCID: https://orcid.org/0009-0007-3446-2907, Taube, Wolfgang ORCID: https://orcid.org/0000-0002-8802-2065 and Romann, Michael ORCID: https://orcid.org/0000-0003-4139-2955 (2026) Improving longitudinal performance assessment of youth soccer players: 10 m sprint percentile curves adapted to biological age. Science and Medicine in Football. pp. 1-12.

Preview

Text Improving longitudinal performance assessment of youth soccer players 10 m sprint percentile curves adapted to biological age.pdf - Published Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (2MB) | Preview

Abstract

Monitoring athletic development in youth soccer is crucial for player evaluation, identifying training needs, and determining long-term progression. However, standard percentile assessments based on chronological age (CA) do not account for biological maturity or developmental variability. This study aimed to improve 10 m sprint performance assessment in youth soccer by integrating biological age (BA) into percentile modeling and applying linear mixed models (LMM) to capture individual development. The analysis was based on 10 m sprint data collected within the Swiss Football Association’s talent development program between 2017 and 2024, comprising 9476 observations for the Lambda Mu Sigma (LMS) method and 3983 for LMMs. BA was calculated using the Mirwald method as an estimation for peak height velocity. Empirical percentile curves (LMS) were generated for both CA and BA, while LMMs established longitudinal reference curves and enabled individual performance predictions using bootstrap resampling. In males, BA explained more variance in sprint performance than CA (R2 = 0.22 vs. 0.18), whereas no significant predictors were identified for females. Percentile curves based on BA elevated rankings of late-maturing players and lowered those of early-maturing players, suggesting better consideration of developmental differences. LMMs provided a more comprehensive modeling framework than LMS, by incorporating repeated measures and individual developmental trajectories. Integrating BA and LMMs longitudinal modeling could enhance the fair evaluation of youth soccer players. Findings support individualized, maturity-adjusted monitoring, offering practical value for longer-term performance diagnostic and evaluation. This statistical approach, applied to a large practice-oriented dataset, enables targeted and sustainable improvement of youth player development.

Item Type:	Article
Status:	Published
DOI:	10.1080/24733938.2026.2643531
School/Department:	School of Science, Technology and Health
URI:	https://ray.yorksj.ac.uk/id/eprint/14426

University Staff: Request a correction | RaY Editors: Update this record