Anything Can Happen in Women's Tennis, or Can It? An Empirical Investigation Into Bias in Sports Journalism
Abstract
The claim that "anything is possible in women's sports" frequently employed by both sports journalists and general audiences highlights the widespread perception of a seemingly uncontested truth about female athletes and their (in)ability to perform consistently at peak levels in comparison to male athletes. We focus on this treatment of female athletes in the world of women's tennis and contest the "common sense" and "experience" justifications of the unpredictability in women's sports with actual data to reveal clear media bias.
Utilising a database of the Association of Tennis Professionals and Women's Tennis Association tournaments dating back to the late 1960s and covering approximately 225,000 fully described matches, we examine the "anything can happen in women's tennis" assumption through logistic regression, focusing on the effect of rank differential on the winning probability in the match while controlling for other factors (tournament type and stage, court surface, age differential, and elite players). The results are rather shocking.
The women's matches do not show higher instability or lower predictability at all, but rather the contrary-the men's matches show lower dependence on the rank difference. The results are robust as checked for data sets of the year 2000 onwards and those including only special events such as Grand Slams.