Do you have a good all-around view? Evaluation of a decision-making skills diagnostic tool using 360° videos and head-mounted displays in elite youth soccer

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.

Keywords: expertise; football (soccer); perceptual-cognitive skills; talent identification and development; virtual reality.

Figure 1

Schematic illustration of a 6 vs. 6 soccer sequence. The central midfield player (No. 6; highlighted in yellow and a member of the “white team”) represents the position of the 360° camera.

Figure 2

Decision-making accuracy (in %) for each player separated for the four teams (n = 4 × 12; players are ordered by their ranked performance in the diagnostic within their team).

Figure 3

Receiver operating characteristic (ROC) curve for the prognostic validity of the YA players’ decision-making results.

Figure 4

Decision-making accuracy (in %) for each YA player separated for the achieved two adult performance levels (the players are ordered by their ranked performance in the diagnostic within their adult performance group) and the mean performance of all YA players.