Tinder has been sparking matches since 2012, and while many of us log into the app almost daily as a part of our social media round up, how many of us really know how it works?
The past decade has seen a complete turnaround in attitudes towards online dating. Before, the sorts of eHarmony and OKCupid might have been seen as something of a last resort, but with the number of 18 to 24-year olds signed up tripling since 2013, the majority of young people now think online dating is a good way to meet people. Gone are the days of lengthy profiles and long conversations, replaced by the incredibly visual and convenient ‘Hot or Not’ format. But how much of this never-ending carousel of potential is innocent, harmless fun, and how much are the people we see, match with, and date already decided for us?
When it comes to the apps I use and interact with, I am guilty of not thinking too much about what’s going on in the background, but when it comes to something as important as the people we are dating, maybe we should. This boils down to the algorithm our dating apps are using to pick who we see, and who we don’t. Tinder has always been incredibly tight lipped around the design of its algorithm, but we know it must be fairly simplistic given how little information we as users provide to the App. However, if a stranger set me up on a blind date with someone, knowing nothing about me except my gender preference and an age range I deem appropriate, then I’m not sure I would even go, let alone expect it to be a good match.
How http://datingranking.net/nl/lovoo-overzicht would you feel if you were being assigned a ‘desirability score’? It sounds a little like we’re Nosediving into a dystopian nightmare (see Bryce Dallas Howards’ Black Mirror descent into dating hell), but it’s exactly how Tinder used to work. It’s known as the ELO Rating System and it was originally designed to rank the skill level of chess players online, only now it’s being used to rank people by how attractive they are. The basics of how it works is that every time someone swipes right on your profile, your ELO score goes up, and every time someone swipes left it goes down. It also takes into account the score of the person swiping on you, so the higher their score, the more points you gain or lose from their decision. Tinder then presents profiles with similar ELO scores to each other, in the hopes that these were people you would be interested in and they would also be interested in you. In theory this should lead to the most successful matches, but in reality, it just kept users in a feedback loop unable to see anyone outside their designated tier.
This opens the door to a whole new demographic and consequently, it has led to a whole new breed of dating apps
In March this year however, Tinder released a statement that ELO scores and desirability rankings are “old news” and that its new algorithm uses “cutting edge technology” to “adjust your potential matches you see each and every time your profile is Liked or Noped”. While they still won’t release exactly what system they are using, it sounds a lot like the Gale-Shapley algorithm. This was created by two economists in 1962 who wanted to prove that any random group of people could be sorted into couples that would all have stable marriages. Tinder uses it to identify patterns in your user history and compares this to other people. If I have a similar history to another woman, swiping right on a large proportion of the same profiles, then I will be shown the profiles of people that she has liked, and I haven’t seen yet in the hopes that I might also like it.
Tinder updates your new potential matches every 24 hours which theoretically means that if you log into your Tinder every day, you should be presented with a stack of new profiles, with the first being the most compatible and each one getting slightly less so with every swipe. While this is an improvement on ELO, as a fairer, more human way of making connections, unfortunately any good this might have brought has been thwarted by the introduction of in-app purchases like Boost and Tinder Gold. The integration of microtransactions being built into the app has changed the motives behind the algorithm altogether. Both Boost and Gold are essentially ways to override it, by pushing your profile to the top of everyone else’s deck. Tinder has incentive to make these extras as effective as possible to keep people buying them, however, since their release, Tinder have been accused of sabotaging the free version of the app by holding back compatible matches until you pay for them.
So, what else is Tinder doing to filter through the estimated 50 million users?
As Tinder refuses to comment on their monetisation scheme or their match-making methods, there is no way we can confirm it, but if it is true then this will raise very serious ethical questions surrounding Duty of Care, and what effect Tinder is having on young people. By creating a false climate of restricted choice and even more limited connection, Tinder would be exploiting loneliness, frustration and low self-esteem all to turn a profit.