Effect of goalkeepers' offensive participation on team performance in the women Spanish La Liga: a multinomial logistic regression analysis

Biol Sport. 2024 Jan;41(1):29-39. doi: 10.5114/biolsport.2024.125592. Epub 2023 May 25.

Abstract

This study aimed to examine the effect of goalkeeper distribution on offensive team performance, during the 2018/2019 and 2019-2020 seasons of the Women Spanish La Liga. A total of 10,868 distributions, during 376 matches were analyzed by systematic observation. Two UEFA PRO coaches designed an ad hoc observation instrument "GOALDFOOT" and one observer coded the data after a training process. An intra-observer reliability kappa index of 0.94 was established. Results show how the offensive effectiveness of the goalkeepers was similar to outfield players, with 0.4% of possessions ending in a goal, 2.2% ending in an attempt on goal, with 79.4% ending unsuccessfully. The goalkeeper lost possession from their distribution 32.5% of the time. Multivariate analysis identified several predictors of goalkeepers' distributions. The results show that teams classified in the middle zone of the final classification of the regular league had 1.2 times more probability of being successful compared with the lowest ranked teams (p < 0.05). Goalkeeper's distribution beginning during Open play after a transition, represented an increase success rate of almost 3 times compared to being performed from a free kick (p < 0.05). Passes from outfield players to a goalkeeper made from distant zones to the own goal, decreased the probability of success (p < 0.001). The pitch location of the distribution outcome near to the opponent goal offered the best probability of success. In conclusion, the most effective offensive sequences occur with dynamic transitions initiated with short passes. This information can provide coaches and players with insights to improve the offensive performance of goalkeepers.

Keywords: Female football; Goalkeeper’s distribution; Key performance indicators; Match analysis; Multinomial logistic regression; Observational methodology.