Tinder’s match system used to be based on the Elo rating system, originally designed to rank chess players. It’s a system of ranking for zero-sum games, where when you win, the other person loses equally. An analogy – in a piece of cake divided equally, if you get a piece, your friend does not get the piece.
Profiles on a similar range of attractiveness get banded together under the Elo system. For instance, let’s assume you have a really great profile, leading to a high Elo score. Your profile is floated across people with a similarly high score so that there are high chances of a match. This, in their judgment, mimics real life where people of a similar level of attractiveness tend to be drawn to one another.
While Tinder no longer relies on this algorithm alone, the consensus amongst people remains that in order to match with people you’d like, you want to present yourself in a way that would appeal to them. In this way, most dating apps are a network version of the real world, where your persona, your looks, your beliefs, and a certain undefinable ‘X factor’ work together to make you seem attractive to someone.
This isn’t the only system around, though. Apps like Match.com, founded by Gary Kremen and Peng T. Ong, collect information about your values system and interests and go a step further by letting you specify what you want (or don’t want) in your future partner. For instance, if you don’t like the idea of dating a pet owner, you can specify it as a dealbreaker, and the app won’t bring those users into the mix of people you might like.
Just like Coca-Cola has kept its formula secret for decades, apps like Bumble are similarly secretive about their algorithm. They don’t reveal about how people are presented to you apart from by using their filter functions (which let you specify distance, and factors like religion, political beliefs, and other parameters for the premium version) and those who might appeal to you. They also make sure not to show you inactive users accounts, which makes good sense.