This study investigated whether a multi-dimensional assessment could assist with talent identification in junior Australian football (AF). Participants were recruited from an elite under 18 (U18) AF competition and classified into two groups; talent identified (State U18 Academy representatives; n = 42; 17.6 ± 0.4 y) and non-talent identified (non-State U18 Academy representatives; n = 42; 17.4 ± 0.5 y). Both groups completed a multi-dimensional assessment, which consisted of physical (standing height, dynamic vertical jump height and 20 m multistage fitness test), technical (kicking and handballing tests) and perceptual-cognitive (video decision-making task) performance outcome tests. A multivariate analysis of variance tested the main effect of status on the test criterions, whilst a receiver operating characteristic curve assessed the discrimination provided from the full assessment. The talent identified players outperformed their non-talent identified peers in each test (P < 0.05). The receiver operating characteristic curve reflected near perfect discrimination (AUC = 95.4%), correctly classifying 95% and 86% of the talent identified and non-talent identified participants, respectively. When compared to single assessment approaches, this multi-dimensional assessment reflects a more comprehensive means of talent identification in AF. This study further highlights the importance of assessing multi-dimensional performance qualities when identifying talented team sports.

This study examined whether skill tests were predictive of status in junior Australian football. Players were recruited from the 2013 under 18 (U18) West Australian Football League competition and classified into two groups: elite (state U18 squad representative; n = 25; 17.9 ± 0.5 years) and subelite (nonstate U18 squad representative; n = 25; 17.3 ± 0.6 years). Both groups completed the Australian football kicking (AFK) and Australian football handballing (AFHB) tests, assessing kicking accuracy/ball speed and handballing accuracy on dominant and nondominant sides. A multivariate analysis of variance (MANOVA) modelled the main effect of “status”, whilst logistic regression models were built for the predictive analysis using the same test parameters. Between-group differences were noted across all parameters, with the combination of kicking accuracy and ball speed on the dominant and nondominant sides being the best predictor of status for the AFK test (wi = 0.25, AUC = 89.4%) and the combination of accuracy on the dominant and nondominant sides being the best predictor of status for the AFHB test (wi = 0.80, AUC = 88.4%). The AFK and AFHB tests are predictive of status, suggesting that their use is warranted as a means of talent identification in junior Australian football.