Online Dating: How the Tinder Algorithm Works


_________________________

In short, for those who do not know the principle of Tinder (and almost all of today’s dating apps): Tinder shows in a very simplified form profiles of contact-friendly singles (or non-singles) in the surrounding area, which is based on the superficial ” or not “principle.

With a left-hand wipe, one sorts one, while with a right-wiping interest is expressed. If you find two matches well, match it and the contact goes in the next round. At Tinder, there is no big drum-talk.

But according to which criteria does Tinder user introduce each other? Does it make sense to distribute as many green hearts as possible to increase the chances of a match?

With Facebook, the Tinder profile is created automatically
Since 2012, the social network has been on the market and currently counts over two million users in Germany. To log in to Tinder, a Facebook account is required, which gives the app access to all the information posted by Facebook and automatically creates the profile.

Publicly visible in the Tinder account are then photos, name, age, occupation, location, common friends as well as details – they should coincide with those of another user. In addition, it is possible to add a short profile text and a favorite member, which is coupled with Spotify. The images can be deleted, supplemented and sorted manually.

If the photos are not enough for Tinder, he can link his Instagram account. 2015, the app has expanded with a few payment functions: With the appropriate package name can be changed, age hidden and the likes of the others can be viewed even before you have decided yourself.

Other in-app purchases, such as boosters and additional super-likes, are also available. From privacy perspective, the app is questionable, since identities are served almost on the silver tray, but it is not to go now. However, it is assumed that the number of fake profiles is limited.

The “Elo Score” provides information about the popularity of the users
The aim of the algorithm is to make as many matches as possible. This is the internal “Elo Score” decisively involved. It raises the popularity of users, which in turn is measured by numerous factors.

One part of the “Elo Score” is, for example, the so-called “Desirability Score”, which provides information about the placement of a user in the internal ranking of other users.

In addition, the “Elo Score” is fed not only with information, which left users of the app fed, but above all by it pushed. This means that anyone who provides more information to the broad masses is already better off, independently of the information itself.

In principle, one could assume that especially attractive people have a higher “Elo Score” with other users and thus presented to other particularly attractive ones. According to Tinder CEO Sean Rad, however, it is not to be quite the opposite of the Süddeutsche Zeitung. The factors taken into account are multifarious.

“Make yourself scarce and you’re the star!”
But what exactly does Tinder now provide with the information it has about a user and which are included in the score? What it does with the hard facts about their own age and interests is obvious: the search criteria are coordinated. But this alone is not an algorithm.

The right and left swipes are relevant to the time and location. For this, Tinder counts the green hearts (right-swipes) that a user forgives and then puts them in relation to the green hearts he receives.

In doing so, the awarded green heart gains its value on the basis of the assessment of the other users who return either many or few green hearts. It is, therefore, true that the less green heart someone forgives and the more he receives at the same time, the more valuable the individual heart is – to express it in a very pathetic way.

Value can be increased
So it does not bring much, simply liking, if not much comes back. And even with a one-to-one relationship, the potential of the algorithm is obviously not exhausted as if it were simply sparing with its right-handed swipes and increasing the value by creating a negative ratio.

So Tinder’s: “Make yourself rare and you’re the star.” Of course, users who are more generous with their Likes are not undermined, but they mostly receive suggestions that do not seem to be very selective. And so the circle closes.

Location and time of the Likes
Tinder also evaluates the location and the time at which many left or right swipes pass. On Saturday evening, at his favorite restaurant, user X may be able to distribute more generous, green hearts than Monday mornings at the office. However, Tinder has covered the details with regard to his attractiveness and the “Elo score”.

Of course, many more information and measurements are still flowing into the algorithm, such as the increased matching with people who share a certain interest in a limited time and a particular location. How many variants there are, can not be said.

Frequently change the user behavior
In summary, it can be said that the only internal “Elo Score” increases with retention. No matter when and to what extent: fewer left-handed swipes increase the attractiveness and thus the value.

It can also not hurt to vary its user behavior from time to time and change search criteria. The Tinder algorithm will perhaps retaliate with a wider range of proposals.


_________________________