Modelling elite golf performance: predictors of hole score on the European Tour from 2017-2019

Investigations of golf performance utilising European Tour data are scarce, particularly when compared to the Professional Golfer’s Association (PGA) Tour in the United States of America. The European Tour differs from the PGA Tour owing to contextual factors such as frequent intercontinental travel and as such, understanding which factors explain performance is necessary. The study’s aims were to investigate changes in performance across time and to establish which variables explain average score on par three, par four, and par five holes. Using mixed linear modelling, performances of 249 individual players across three consecutive seasons (2017 n= 151; 2018 n= 165; 2019 n= 151) were analysed. Stepwise multiple linear regressions were also used to establish predictive models for score according to par for the hole. Drive distance (DD) (p= 0.01. Effect Size (ES)= 0.35) and scramble % (p= 0.01. ES= 3.19) significantly increased from 2017 to 2019. Drive accuracy (DA), greens in regulation, putts per round, and stroke average showed no change (p> 0.05). Stepwise regression could predict up to 57% of par three hole performance (2019= F(3, 157)= 71.524. p< 0.001. f2= 1.320. Adjusted R2= 0.569. Model= -1.172(GIR) + 0.045(Putts) + 0.001(DD) + 2.157), 81% of par four hole performance (2018= F(3, 161)= 234.432. p< 0.001. f2= 3.630. Adjusted R2= 0.810. Model= -1.345(GIR) + 0.063(Putts) - 0.087(DA) + 3.166), and 68% of par five performance (2019= F(4, 156)= 86.281. p< 0.001. f2= 2.125. Adjusted R2= 0.680. Model= -1.585(GIR) + 0.058(Putts) -0.002(DD) – 0.314(Scramble) + 4.917). This study suggests that on the European Tour, DD is increasing with no change to DA. The most important predictor variables appear to be GIR and putts per round as they entered all nine predictive models, in combination with either DD, which entered seven, DA which entered four, and/or scramble % which entered three.