To assess a coach’s subjective assessment of their athletes’ performances and whether the use of athlete-monitoring tools could improve on the coach’s prediction to identify performance changes.
Eight highly trained swimmers (7 male and 1 female, age 21.6 [2.0] y) recorded perceived fatigue, total quality recovery, and heart-rate variability over a 9-month period. Prior to each race of the swimmers’ main 2 events, the coach (n = 1) was presented with their previous race results and asked to predict their race time. All race results (n = 93) with aligning coach’s predictions were recorded and classified as a dichotomous outcome (0 = no change; 1 = performance decrement or improvement [change +/- > or < smallest meaningful change]). A generalized estimating equation was used to assess the coach's accuracy and the contribution of monitoring variables to the model fit. The probability from generalized estimating equation models was assessed with receiver operating characteristic curves to identify the model's accuracy from the area under the curve analysis.
The coach’s predictions had the highest diagnostic accuracy to identify both decrements (area under the curve: 0.93; 95% confidence interval, 0.88-0.99) and improvements (area under the curve: 0.89; 95% confidence interval, 0.83-0.96) in performance.
These findings highlight the high accuracy of a coach’s subjective assessment of performance. Furthermore, the findings provide a future benchmark for athlete-monitoring systems to be able to improve on a coach’s existing understanding of swimming performance.
