The reliability and validity of mobalytics proving ground as a perceptual-motor skill assessment for esports

Abstract

This study aimed to investigate the test-retest reliability and discriminant validity of the Mobalytics Proving Ground™ assessment for League of Legends. Forty participants (age: 24.15 ± 3.68 y, sex: male = 31, female = 9) were a priori classified into two expertise groups: (1) esports players (age: 22.98 ± 3.64 y, sex: male = 18, female = 2), and (2) controls (age: 25.31 ± 3.42 y, sex: male = 13, female = 7). Participants completed three separate trials (60 s each) online. To assess test-retest reliability, variables displaying normal distributions were analysed using intraclass correlation coefficient (ICC) estimates for two-way mixed-effects models with 95% confidence intervals. The average ICC for all the independent variables in the esports group and control group were moderate (ICC esports = 0.53 and ICC control = 0.72). The average 95% confidence intervals for the independent variables in the esports and control group were ICC = 0.30–0.75 and ICC = 0.55–0.86, respectively. A Friedman test revealed an effect size of 0.11 in the esports group and 0.07 in the control group. In terms of discriminant validity, there were significant differences for 17 variables when comparing the best scores of each group. Overall, the Mobalytics Proving Ground™ assessment used in the current study can, to some extent, distinguish esports players from controls.