“Imagine that 100 percent of gay men wear pink shirts all the time, and 10 percent of straight men wear pink shirts all the time. Even though all gay men wear pink shirts, there would still be twice as many straight men wearing pink shirts. So, even in this extreme example, people who rely on pink shirts as a stereotypic cue to assume men are gay will be wrong two-thirds of the time,” Cox said in a statement.

Previous studies have come out claiming that gaydar is very real, and perhaps even based on cues as innate as facial shape. So a single study can’t debunk all of those. It’s possible there’s some detectable difference that’s common, if not inherent or universal, in people of different sexual orientations. But in his own team’s studies, Cox found that any facial gaydar people seemed to have could be attributed to something pretty trivial: Photo quality.

Gay men tended to use nicer photos on their dating profiles, and those are the pictures that end up being used in studies like these.