ABSTRACT
Elite youth players' decision-making skills are considered important predictors of adult performance in soccer. The presentation of 360° videos in head-mounted displays offers new potential for the diagnostic of these skills in talent development programs. This study evaluated a new diagnostic tool using soccer-specific 360° videos for assessing decision-making skills in youth academy (YA) players. The evaluation consisted of players' subjective feedback as well as the analysis of diagnostic and prognostic validity. It was hypothesized that high-level YA players achieve better diagnostic results than regional-level players, and U19 outperform U17 players. Moreover, YA players' diagnostic results should be positively associated with future adult performance level. During the 2018/19 season, N = 48 youth players participated in the diagnostic procedures (split-half reliability r = .78). Participants were shown 54 videos which terminated when the central midfielder received a teammate's pass. Participants were then asked how to best continue playing. The subjective evaluation explored YA players' experiences with the diagnostic tool via quantitative ratings (e.g., "How exciting was the task?", "How involved did you feel in the game situation?") and additional interviews. Diagnostic validity was examined in a balanced cross-sectional 2 × 2-design (performance level x age group) and prognostic validity in a 3-year prospective design. Sensitivity and case-by-case analyses completed the evaluation. The YA players provided positive quantitative ratings regarding their experienced immersion into the environment. Players' qualitative feedback indicated general acceptance of the diagnostic tool as well as it offered recommendations for improvements. Confirming the diagnostic validity, ANOVA revealed significant main effects for performance level (p < .001, η2 = .29) and age group (p < .01, η2 = .14). Contributing to the prognostic validity, the diagnostic results discriminated between YA players achieving a higher and a lower adult performance level ("League 1-4" vs. "League 5 or below") in adulthood (p < .05; d = 0.80). A ROC curve and the AUC showed that the correct assignment to the adult performance levels is possible with a 71% probability. YA players with a high decision-making accuracy had a six times higher chance of playing in "League 1-4". The results demonstrated empirical evidence for the new diagnostic tool in terms of YA players' acceptance and validity coefficients exceeding effect sizes of former studies. The technology provides opportunities to test soccer-specific situations demanding an all-around view that were not testable in former experimental settings. Further technological advancements will enable the realization of improvements recommended by the players. Nonetheless, case-by-case analyses suggest caution in using such a diagnostic as a selection tool in talent development programs.
ABSTRACT
Talent research has recommended that multidimensional assessments of performance are needed to improve the identification and development of talented young athletes. However, factors such as the relative age effect may cloud our ability to assess factors related to performance. The aim of this study was to determine the extent of any relationship between soccer players' chronological and relative age, and objective and subjective performance assessments. Data for highly talented male soccer players selected into the German Soccer Associations' talent promotion program (N = 16,138) for U12 to U15 age groups (M age = 12.62 ± 1.04 years) were examined. Besides anthropometric assessments, players completed a battery of five motor tests that objectively assessed speed abilities and technical skills (specifically sprint, agility, dribbling, ball control, and juggling). In addition, coaches subjectively rated players on their kicking, tactical, and psychosocial skills, as well as providing holistic evaluations of each player's current and future performance levels. Correlation analyses were used to investigate the extent of any relationships between the chronological and relative age of players and their results for each of the assessments. A strong linear decrease in the frequency of later-born players confirmed the overrepresentation of early-born players in all age groups (0.92 ≤ |r| ≤ 0.95, each p < 0.001). From U12 to U15, significant (each p < 0.001) correlations were found between the chronological age of players and their height (|r| = 0.70), weight (|r| = 0.69), speed abilities (|r| = 0.38), and technical skills (|r| = 0.43). When evaluating each age group separately, small effects were found when correlating relative age with the anthropometric assessments (0.18 ≤ |r| ≤ 0.26), and only trivial effects with speed abilities and technical skills (0.01 ≤ |r| ≤ 0.06). Similarly, low correlations were found for the subjective evaluations of kicking, tactical, and psychosocial skills with chronological age across age groups (0.03 ≤ |r| ≤ 0.07), and with relative age in each age group (0.01 ≤ |r| ≤ 0.11). The results show a skewed distribution toward early-born players and-in reference to their relative age-advanced performance in late-born athletes. However, trends toward a better holistic rating of early-born players for current and future performance levels were found. Coaches should be aware of these effects during talent selection, but also when interpreting results from subjective and objective assessments of performance.
