We have been swiping furiously for nearly two decades—and are still no closer to the secret of finding true love. We are now mad at the algorithm and ready to break up with dating apps. What went wrong? In part one of this two-part series, we look at the history of dating apps—which gave us hope only to break our hearts.
First, the origins of swiping right
Online dating first arrived in our lives as old-fashioned websites in the 1990s—Match.com in the US and Shaadi.com for us sanskari Indians. Yet by the early noughties, there were signs of trouble. Studies showed that we fickle humans didn’t do well with greater choice:
In a classic example of choice overload, Iyengar and Lepper presented grocery store shoppers with a tasting booth containing either six or 24 flavours of gourmet jam. Despite being drawn to the booth with more options, shoppers were the most likely to make a purchase when given fewer choices.
Likewise, when online daters are given more profiles to examine, they are known to spend more time searching, to be less selective by considering options that do not meet their preferences, and to make poorer choices that do not fit what they are looking for in a partner.
The flaw was built into the very USP of online dating. How do we solve for a feature that behaves like a bug?
Enter, compatibility matching: Dating apps tried to narrow the choices by creating algorithms that would match compatibility. The first was developed and patented by eHarmony in 2000. It was pretty traditional in its approach. Users took a compatibility test—with a whopping 450 questions. The machine offered them a limited set of matches based on criteria established by psychologists.
Search and match: OK Cupid took this one step further—by combining search with matching—which meant “the algorithm functioned as more of a decision aid by empowering users to seek out potential partners for themselves while also offering suggestions to narrow the field.” Its algorithm used “match percentages” to assess compatibility—and weighted each question by its reported level of importance. On paper, technology would give users the power to define and find their Mr or Ms Right
But, but, but: The alluring theory fell apart when tested by the reality of human nature:
The problem with these early matching systems is that they assumed users knew precisely what they desired in a partner. However, people’s stated preferences for an ideal mate do not always align with what they find attractive in person. This is further complicated by the fact that online dating often encourages users to prioritise qualities (e.g., height, income) that are poor indicators of what it will be like to interact with someone in the flesh.
Naturally, we tried to solve the problem by throwing more tech at it.
Say hello to dating apps!
iPhone launched in 2007—and Grindr in 2009. We officially entered the era of ‘swiping’ for love (or at least sex). Grindr, for example, used the system of “collaborative filtering”—relying on user behaviour rather than stated preferences to offer you matches:
For example, imagine a hypothetical scenario where Tyrone is attracted to Carlos. If others who like Carlos also show an interest in Zach, then Zach will be presented to Tyrone as a possible match. This strategy is used to suggest products on Amazon and movies on Netflix, but on dating apps, recommendations must be reciprocal to minimise rejection. In other words, matching algorithms must consider not only whether one person is likely to find another attractive but also whether that interest will be well received.
Or to put it more simply: “If a lot of 25-year-old women who like Star Wars keep tapping ‘like’ on one man’s profile, the algorithm might show him to more sci-fi fans in that age range.” Then came Tinder in 2012—and a more Darwinian approach to “collaborative filtering.”
The Elo system: Tinder embraced the social hierarchy of dating—the idea that some people are in the “same league” and not others . And mapped it on to a system used to calculate skill level of chess players. Invented by physicist Arpad Elo, the system gave greater weight to someone who defeated a Grandmaster than an amateur player. In the dating world, it worked like this:
You rose in the ranks based on how many people swiped right on (“liked”) you, but that was weighted based on who the swiper was. The more right swipes that person had, the more their right swipe on you meant for your score.
Tinder would then serve people with similar scores to each other more often, assuming that people whom the crowd had similar opinions of would be in approximately the same tier of what they called “desirability.”
Basically, we all went back to high school—with all the heartache and rejection it entailed.
Point to note: FYI: Bumble presents itself as a more ‘progressive’ alternative by allowing the woman to make the first move. But it too uses the brutal Elo criteria. In 2019, Tinder claimed to have dumped the Elo system. Its new methodology is more like that used by the supposedly kinder and gentler Hinge—which isn’t quite as kind as we will see.
The Gale-Shapley algorithm: Founded in the same year at Tinder, Hinge positioned itself as the app for folks looking for relationships—not just a hook-up. It uses an algorithm developed by two Nobel prize-winning economists back in 1962. It is also known as the propose-and-reject algorithm—and is aimed at creating “stable pairs” :
The Gale-Shapley algorithm solves the problem of creating stable matches between two groups when both sides prefer some partners over others (e.g., in the case of college admissions, marriage). Matches are stable if there are no two people who would rather be with each other than the partner they have been recommended. For instance, by matching Ravi with Ava, one can be confident that there is no one else in the dating pool they would prefer who would also be interested in them in return.
This is a nerdy way of saying that anyone else you may have preferred to your partner is unlikely to have returned the interest. So both of you end up with a choice that is “as good as it gets.” It’s another way of matching you with someone in your “league”—but isn’t quite so mean about it.
But, but, but: The idea of matching people who are similarly “desirable” doesn’t work because of a simple reason: People are often “aspirational” in who they want to date. They swipe right on who they want—not necessarily who they can “get.” As an academic who specialises in relationships and tech bluntly puts it: “If you’re thoughtful about it, you realise that the algorithm is telling you something about your own desirability when it’s deciding who to show you.”
FYI: A more shameless version is literally called The League:
The League—an exclusive dating app that requires you to apply using your LinkedIn—shows profiles to more people depending on how well their profile fits the most popular preferences. The people who like you are arranged into a “heart queue,” in order of how likely the algorithm thinks it is that you will like them back.
The dating heartbreak algorithm
Most of the data shows that online dating isn’t any more or less successful than any other kind. And there is no greater benefit in opting for a relationship-oriented Hinge than a swipe-driven Tinder. The former may have more information on who you are, but that doesn’t mean they’re any better at finding you a match. There are a number of good reasons why dating apps often prove to be duds.
One: There is an inherent mismatch between how we make connections and how we search for partners on a screen. A 2002 study flagged this core disconnect:
We suggest that online dating frequently fails to meet user expectations-because people, unlike many commodities available for purchase online, are experience goods: Daters wish to screen potential romantic partners by experiential attributes (such as sense of humour or rapport), but online dating websites force them to screen by searchable attributes (such as income or religion). We demonstrate that people spend too much time searching for options online for too little payoff in offline dates—in part because users desire information about experiential attributes, but online dating websites contain primarily searchable attributes.
That’s why we’re often disappointed by people who tick all the boxes online—but do nothing for us in person. A 2017 study tried and failed to build a machine learning algorithm that could predict romantic desire using principles of relationship science. While it could indicate how desirable a person was, it failed entirely to “anticipate which people would hit it off in person.”
Two: Dating apps are geared toward keeping us hooked—always looking and swiping. But the longer we stay on the app, the less attractive our options become. Here’s how an OkCupid product exec explains it:
Hypothetically, if you were to swipe on enough thousands of people, you could go through everyone. [You’re] going through people one at a time … you’re talking about a line of people and we put the best options up front. It actually means that every time you swipe, the next choice should be a little bit worse of an option.
So, the longer you’re on an app, the worse the options get. You’ll see Tinder, Bumble, OkCupid, we all do recycling. If you’ve passed on someone, eventually, someone you’ve said “no” to is a much better option than someone who’s 1,000 or 10,000 people down the line.
But it feels pretty bleak from a user’s point of view: time spent starts to feel like time wasted.
Three: Dating apps are built to monetise our desire. They do it by keeping those we want most tantalisingly out of reach. Example: On Hinge, you can only contact people who are “Most Compatible”—and other “Standout” users who are “most desirable”by sending them a Rose:
Hinge users talk about the best matches being trapped in “Rose Jail,” where they’re sequestered to coerce you into paying. These ultra-appealing maybe-matches can show up in the app’s main feed eventually, but Hinge’s goes out of its way to suggest you might never see your Standouts again if you don’t shower them in $3.33 roses.
A premium version of Hinge makes your profile more visible, helps you “get recommended sooner” and gives you “access to people more your type.” Price of that privilege: $149.99 for six months. Tinder does something similar with Boosts and Super Likes.
Point to note: A 2023 study found that dating apps use more attractive (popular) users to “generate more revenue by boosting users’ engagement (through more likes and messages sent)”—and selling premium perks as listed above. OTOH, other users claim to have been “shadow banned” for not being quite as attractive as others—where the platform limits the reach of your account.
Quote to note: Here’s how the New York Times describes the state of existential despair over dating apps:
Some users now view the apps not as wingmen but as gatekeepers obstructing their romantic prospects. Katie Nguyen, 33, a recruiter in Los Angeles, said she had diminishing confidence that Bumble, Hinge or Raya were actually trying to connect her with a long-term partner. She wonders whether her feed appears bleak because there are truly no eligible options, or because the apps have a vested interest in keeping her swiping.
Hack that app! Inevitably, all this angst has led to frustration and jugaad. There are entire Reddit communities and how-to video guides to help you game that algorithm. Most of the tricks are just plain absurd—more a measure of desperation than effectiveness:
A 30-year-old app developer admitted to deleting unique details from his dating app profile, saying that the more generic and "meaningless" he made his profile, the more matches he seemed to get. On Reddit, a straight man wondered if changing his Tinder preferences to men and women then to gain a few extra likes, then switching it back to just women a few days later would "inflate his location on the algorithm," resulting in more matches.
Or as one former Match expert explains it: “It’s like trying to give your dating life a colonic. Like, how do I clean this out?” Except it often doesn’t work—or rarely with any consistency.
The bottomline: Unsurprisingly, all this angst has resulted in a collapsing market for dating apps. Can more tech—i.e AI—save them? Or will Gen Z spell their demise? We look at what’s next in part two tomorrow.
Reading list
This Harvard Data Science Review column is best on the history of dating apps—and evolution of the tech. Jumpstart offers a good overview of the main apps. BBC News looks at what dating app algorithms tell us about human desire. Vox and Gizmodo are best in analysing how dating apps like Tinder and Hinge work. New York Times looks at the increasing number of people trying to “hack” dating apps—and Business Insider argues these efforts are mostly futile. Nikkei Asia explains why dating apps became popular in India.
