The IOC's SRY Test Is Crude, and Women's Sport Still Needs It

Between 2000 and 2025, roughly 50 to 60 athletes with differences in sex development reached the finals of women's events at global championships. That number comes from Dr. Stéphane Bermon's presentation at the September 2025 World Athletics panel in Tokyo. In a dataset of elite female finalists, DSD athletes are wildly over-represented relative to the roughly 2% base rate in the general population. That gap is the entire argument. The IOC's new SRY gene screen is imprecise, politically charged, and guaranteed to produce at least a few ugly misclassifications. I still think the IOC is right to implement it, because a women's category without a defined biological threshold is statistically undefined, and an undefined category protects no one.

I know. The stats guy defending a test with a known false-positive problem. Feels off-brand. But I've spent my career arguing that you need a measurement framework before you can evaluate outcomes. In baseball, we fought for decades to replace batting average with OPS+ and wRC+ because BA was a bad proxy for offensive production. Nobody argued we should stop measuring hitting entirely. The SRY screen is batting average: crude, incomplete, better than nothing.

What Happens When You Don't Draw the Line

The IOC's 2021 Framework on Fairness, Inclusion and Non-Discrimination explicitly opposed blanket sex testing. It was a principled document. It also left federations with no enforceable standard for the women's category. The result was chaos. Caster Semenya was barred. Imane Khelif was publicly attacked during competition. Dutee Chand fought a years-long legal battle. The absence of a clear rule didn't protect these athletes. It exposed them to ad hoc decisions, political pressure, and social media tribunals that were far more invasive than a cheek swab.

Think of it this way. In the NFL, the salary cap is an imperfect tool. Teams exploit loopholes constantly. But nobody argues the league should abolish the cap because it's imperfect. The cap defines the competitive structure. Remove it and you don't get fairness; you get the Yankees.

The women's category in sport exists because male puberty confers measurable, persistent advantages in speed, power, and oxygen-carrying capacity. If you refuse to define who qualifies for that category on biological grounds, you haven't protected inclusion. You've made the category meaningless.

The Honest Tension

Here's where I have to be straight about my own model's weakness. The SRY screen will flag women with complete androgen insensitivity syndrome, athletes who have XY chromosomes but develop female bodies and derive zero performance benefit from those chromosomes. The 1996 Atlanta Olympics proved this exact failure mode. Several women were misclassified. The IOC abandoned universal testing afterward.

That history is damning. And the IOC calling this policy "evidence-based" while releasing no public evidence is, charitably, a bad look. Megan Rapinoe's accusation that the timing reflects political pressure rather than scientific rigor deserves a serious answer the IOC hasn't provided.

But the question isn't whether the SRY screen is perfect. The question is whether it's better than the status quo. The status quo produced 50 to 60 DSD finalists in women's events, public humiliation of athletes on live television, and a patchwork of federation-level rules that varied by sport and by year. Kaillie Humphries, with 3 Olympic golds, called the announcement "a great day for women's sports." She's not wrong to want a defined playing field.

The IOC should publish its evidence. It should name the working group members. It should build an appeals process that accounts for CAIS and other conditions where the SRY gene is present but functionally irrelevant. Those are implementation problems, not reasons to abandon the principle.

Caster Semenya is right that this policy will disproportionately affect women from the global South. That pattern is real, and the IOC owes those athletes transparency and procedural fairness. But the solution to unequal enforcement is better enforcement, not no enforcement.

A category you can't define is a category you can't protect. The SRY screen is version 1.0. Make it better. Don't delete the file.