How do college coaches in the United States identify youth female and male soccer players?

Matthew Andrew, Sam Barraclough, Andrew O. Triggs, James H. Dugdale, Adam L. Kelly, Matthew J. Reeves

Abstract

Talent identification (TI) in soccer is a complex and multifactorial process within the context of collegiate sport in the United States, where coaches must assess performance-ready athletes often under strict regulatory and resource constraints. Despite the critical role college coaches play in bridging youth to professional soccer, little is known about their evaluative priorities during recruitment. This study examined how soccer coaches from female and male NCAA Division I and II programs perceived the importance of various player attributes and scouting methods in the TI process.

A total of 178 college soccer coaches completed a survey assessing perceptions across seven attribute categories (technical, physical, psychological, game intelligence, social, other, and coach-specific) and common scouting methods. Bootstrapped trimmed means, effect sizes, and inter-rater agreement (r_wg) were used to analyse the coaches’ ratings of importance across attributes and to assess for differences in coach perceptions within the female and male programs. Results showed coaches across female and male programs rated technical proficiency, coachability, decision-making, and work rate as critically important. College-specific soccer knowledge was the most highly valued coach attribute, while live match observation was the most preferred scouting method.

Sex-based differences were generally minimal, although emerged in perceptions of physical and social attributes, with coaches of female players placing greater emphasis on communication and agility. Results highlighted a shared prioritisation of technical and psychological qualities in college TI, with contextual differences influenced by sex and program structure. These insights support the development of more aligned and evidence-informed TI strategies in collegiate soccer environments.