Rankings influence decision-making across many domains, from search engines to hiring processes. But how can we determine whether a ranking is fair? Many fairness frameworks rely on the proportional representation of protected groups, but these approaches often overlook the stochastic nature of ranking processes and the constraints of finite candidate pools. In this talk, I discuss hyperFAIR, a framework for assessing and enforcing fairness in rankings based on the hypergeometric distribution. This approach models selection from a finite pool, capturing real-world constraints that conventional probabilistic models, such as binomial sampling, fail to reflect. I will discuss how this perspective complements common fairness assumptions and present methods to detect and correct unfair rankings efficiently.