Tinder wants to make things clear as to how his platform classifies and shows people possible matches, so today he published a blog post on the subject, but still kept things rather vague. The Elo score of the company was a "hot topic" a few years ago, according to the blog post, but the ranking function has now been depreciated.
The idea behind Elo's score was that Tinder would classify people by their attractiveness. Elo scores are also used to classify chess players, but in the context of Tinder, the more people who have hit correctly (or liked) a person's profile, the higher their assigned score. Your card would be delivered to other people with a similar score, thus keeping the most desirable people interacting with each other. In Tinder, where profiles have a relatively limited range, a person's appearance often fuels much of the desire to mate, so people speculated that these ratings kept people warm talking to each other and leaving undesirable people wallow with a low rating. Tinder has grown out of the Elo score
Tinder, unlike other applications, only requires users to enter their age, distance and gender preferences. It does not take into account the compatibility score, like the sister company OkCupid, nor does it offer filters based on height, religion or ethnic origin, as a large part of its competence.
"Our algorithm is designed to be open," says the company. "Today, we do not trust Elo, although it is still important to consider both parties who like the profiles to form a coincidence."
Tinder adjusts the possible matches that a user sees each time someone acts on their profile. He says. The company reorders the possible profiles of coincidence of this user within 24 hours after the actions taken. That's as concrete as Tinder gets in his blog post, but it sounds a lot like Tinder is based on something similar to the Gale-Shapley algorithm, or the algorithm that Hinge has said he uses. This algorithm identifies patterns around "likes". If I like a guy, and so does another woman on the platform, she and I could have the same taste. If she liked someone on the platform that I have not seen yet, Tinder could show me that profile with the hope that I like it too.
Of course, Tinder is also the best money generator of Match Group, so it offers users the option to completely omit any of these algorithm classifications with a purchase from the application. That may be in the form of Super I like it, which automatically moves a letter to the top of a person's stack of profiles (and visually indicates that it was Super I liked), or an increase in profile, which Tinder says brings a profile closer to the top of the stack of profiles of many other users for 30 minutes.
Tinder feels like a free application for everyone, where everyone really exists, but as the platform grows, you should sort the profiles in a somewhat personalized way, or otherwise find a match would be impossible. While Elo scores worried many users, it is likely that the experiences of the hot people will improve, and if they slipped to the end of Tinder, they would probably have seen people with lower scores. Tinder, and all dating applications, need to create matches and produce dates for people to stay connected, so it has an incentive to show people to other users that they would like to have up to date.