Due to its unique scoring system, in tennis it is possible to win a match with fewer points or fewer games than the opponent. This scoring quirk has been called Quasi-Simpson paradox (QSP) and has been analyzed by Wright et al. (2013) for the years 1990-2011. This work follows up that of Wright et al. (2013) and extends it in that: i) the QSP is studied for different, and more recent, years (2012-2017), allowing a time comparison; ii) QSP is considered with respect to games as well as to points; iii) it considers both men and women; iv) the significance of the results is verified by means of statistical tests; v) it analyzes also the difference of won points, between the winner and the loser, when QSP occurs; vi) QSP is studied also through Monte Carlo simulations allowing to analyze QSP excluding any possible players` strategy. The simulative approach allows us to solve the seeming "Federer paradox".
© Copyright 2019 Journal of Sports Analytics. IOS Press. All rights reserved.
|Subjects:||tennis mathematical statistics simulation|
|Notations:||technical and natural sciences sport games|
|Tagging:||Quasi-Simpson Paradoxon Federer Paradoxon|
|Published in:||Journal of Sports Analytics|