Abstract
<jats:p>Background and Purpose Talent identification in youth sport has traditionally relied on isolated physical performance measures, which often fail to capture the complexity of long-term athletic potential. Growing evidence suggests that talent is a multidimensional construct influenced by physical, technical, and psychological factors; however, experimental validation of integrated scouting models remains limited. This study aimed to experimentally validate a multidimensional talent scouting model integrating physical, technical, and psychological indicators in youth athletes and to compare its classification accuracy with a traditional physical-based scouting approach. Materials and Methods A quasi-experimental design was employed involving 120 youth athletes (aged 13–16 years) from regional development programs. Physical performance (20 m sprint, countermovement jump, Yo-Yo IR1), technical skills (skill execution accuracy and decision-making), and psychological characteristics (motivation, resilience, competitive confidence) were assessed using standardized and validated instruments. Talent classification accuracy of the multidimensional model was compared with a traditional scouting model using multivariate and regression analyses. Results The multidimensional model demonstrated higher classification accuracy than the traditional model (78.3% vs. 62.5%), with greater sensitivity (81.0% vs. 59.1%) and specificity (75.6% vs. 65.4%). Physical (β = 0.41, p < 0.001), technical (β = 0.47, p < 0.001), and psychological indicators (β = 0.29, p = 0.003) all contributed significantly to talent classification. Significant interaction effects were observed among domains. Conclusions Integrating physical, technical, and psychological indicators provides a more accurate and holistic approach to youth talent identification. Multidimensional scouting models may enhance decision-making, reduce selection bias, and better support long-term athlete development.</jats:p>