Asking a dating exec how their matchmaking algorithm works is like asking Coca-Cola for its top-secret formula: they’ll tell you it’s a mystery, that it’s too hard to explain, that they simply cannot talk about it, Fight Club style.
Tinder says that it looks at things like app usage, profile details, and swipe history to find your matches. Bumble also says it looks at your swipe history, while Hinge didn’t say much, except that its “proprietary algorithm” was based on work done by two Nobel Prize-winning mathematicians.
But there’s a reason why these companies are so cagey about their code. Mathematician Cathy O’Neil says she thinks that if daters really knew how basic the algorithms are, they might not put so much blind faith into them. The apps, she suspects, run off of predictive algorithms. “They just take historical data,” she explains, and “look for patterns of success or failure.” In other words, “they extrapolate.”
In episode three of Land of the Giants: Dating Games, we dive into just how these algorithms work and speak with daters who are trying to hack the code that controls their love lives. Algorithms may seem “kind of magic,” O’Neil explains. But the truth? “We don’t want anyone to see how dumb they are.”
The result is users who go on a bunch of terrible dates — and some who come up with strategies to get more and better matches. Take Jeremy, a 30-year-old app developer in Philadelphia, who discovered that his algorithm works better for him when he scrubs any unique details from his profile. “It’s definitely discouraging to know that what’s generic and watered down kind of works best,” he says. “When I started getting a lot more matches on my profile, it was because I had things that were meaningless.”