New Goalkeeping Ability Index for Football Goalkeepers: Using a Regression Formula for Predicting the Probability of Goalkeeping Failure
Yusuke Hirashima, Kensuke Suzuki, Masaaki Koido, Takeshi Asai and Masao Nakayama
[Received April 14, 2023; Accepted October 24, 2023]
Currently, the most commonly used goalkeeping ability metric for goalkeepers is Save percentage (SV%); however, SV% is not an effective evaluation index of goalkeeping ability because it does not account for goalkeeping-related difficulty. Recently, a regression equation was constructed to predict the probability of goalkeeping failure; with this equation, it has been possible to quantify goalkeeping-related difficulty. The purpose of the present study was to develop a new goalkeeping ability index for goalkeepers, which would take into account the degree of goalkeeping-related difficulty using a logistic regression model for predicting the probability of goalkeeping failure. To this end, we propose two new evaluation metrics: Saving Indicator (SVI) and Saving Score (SVS). Using SVI, SVS, and SV%, we evaluated the performance of goalkeepers who competed in the 2018 FIFA World Cup in Russia. Comparison of the three evaluation indexes revealed that both SVI and SVS considered the degree of goalkeeping-related difficulty but SV% did not. In future research, it will be necessary to evaluate an entire football tournament and season using SVI and SVS and to consider in detail the characteristics of each new goalkeeping evaluation index and how to specifically apply them.
Keywords: soccer, goalkeeper, objective rating, logistic regression, game performance analysis
[Football Science Vol. 20, 62-70, 2023]
|