The main purpose of the present study was to build a parsimonious model to predict the probability of winning in rink hockey from different situational variables and evaluate each predictor`s contribution to the match outcome. A sample of 238 matches played during the last season in the Spanish first division (OkLiga) was analysed. The best predictive logistic model for match outcome was selected through all possible regression methods. The entire model included five categorical predictor variables (match location, team level, opponent level, scoring first, and match status at halftime) and one binary outcome variable (match outcome). The final model selected excluded the scoring first predictor and had a sensitivity and specificity greater than 80% for a cut-off point of .413. This model was applied to predict winning a match in 32 frequent situations determined from a two-step cluster analysis. The predictor with the highest contribution to the match outcome was match status at halftime, followed by the opponent`s level, team level, and match location. Our findings may help rink hockey coaches and practitioners to recognise the contribution of situational variables on the match outcome to tailor their game plans and design more aggressive game plans, improving game understanding.
© Copyright 2021 International Journal of Performance Analysis in Sport. Taylor & Francis. Kaikki oikeudet pidätetään. / Kääntänyt https://www.DeepL.com/Translator