Cracking the Code of Love: Navigating the Intricacies of Dating Algorithms
Understanding how compatibility systems work can play a crucial role in finding a companion
It’s pretty common knowledge current dating sites use sophisticated algorithms to aid in determining potentially good matches. With that in mind, understanding how they work might help you evaluate their efficacy and increase your chances of finding the right person
As someone who once used mind-numbingly boring statistics to develop algorithms, let me see if I can explain what they are, how they work, and how they affect your dating odds without making your eyes cross.
A short history of dating algorithms
Florian Schroeder explains that “The use of algorithms to predict romantic success is not entirely new. Mathematical approaches to matchmaking emerged in the 1960s, when researchers began creating algorithmic compatibility models based on psychological theories of what makes relationships work.” Primitive early algorithms primarily matched people on demographics and a few interests.
For example, Match.com began using what, at the time, were relatively primitive algorithms in 1995. Five years later, eHarmony joined in their use.
How modern dating matchmaking works
Algorithms, at heart, consist of an equation. The solution to the equation yields a compatibility index. These equations only work with good data. The programmer and statistician mantra remains, Garbage in, garbage out.
According to Woz U, a learning platform founded by Steve Wozniak, “—dating apps also use behavioral data such as a user’s swiping and messaging patterns to refine their matchmaking algorithms.” Swiping, poking, or emailing tells the AI that a match meets your approval (or doesn’t). The app then suggests similar matches.
Supposedly, this allows the matchmaking equation to improve the more activity you engage in onsite. Linked social media provides additional data points for some equations.
A competitor can get a jump start on their algorithm by stealing an existing algorithm, so, as Schroeder adds, “—the details of proprietary algorithms are closely guarded secrets—” In the end, all a person or organization can say is how the person thinks an equation works based on how that person would go about making matches.
Potential problems of matchmaking equations
Theoretically, the more data points in the equation, the better the match. In real life, however, computer time and computing power limit the number of data points in an equation. So, companies make educated choices about what to include.
In addition, many of the questions on your profile do not correlate strongly with matchmaking success. Having infinite data points that have nothing to do with attraction predicts no better than chance.
According to Eli Finkel and his research partners, “No compelling evidence supports matchmaking sites’ claims that mathematical algorithms work – that they foster romantic outcomes that are superior to those fostered by other means of pairing partners.” Finkel also mentions that unquantifiable something—call it chemistry—that causes sparks and can’t be reduced to a data point.
Algorithms develop racial and other biases if not monitored closely. According to Wired Magazine, “Dating apps like Tinder, Hinge, and Bumble use “collaborative filtering,” which generates recommendations based on majority opinion. Over time, those algorithms reduce human choice and marginalize certain types of profiles.”
Jevan Huston, a researcher at Washington University, further explains, “Dating platforms have the opportunity to disrupt particular social structures, but you lose those benefits when you have design features that allow you to remove people who are different than you.”
Poor matches also happen when the number of people who match your compatibility scores hits the bottom. In my experience, some match services that advertise as matching gays and lesbians, such as eHarmony and Match.com, have such small pools to choose from in any one area that the matches perform poorly.
According to the Pew Research Center, “Roughly seven in ten online daters believe it is very common for those who use these platforms to lie to try to appear more desirable.” This false information throws the matchmaking equations off and results in poor matches.
Now for the good news
On a positive note, the equations used by dating apps may be close enough to help find someone compatible. Even though some of the information used in dating equations may not be correlated with long-term happiness, their equations may still be effective.
In a research study done by Liesel Sharabi, “Findings suggested that whether algorithms actually worked mattered less than whether participants had the perception that they worked for finding a partner. Moreover, participants reported better first dates to the extent that they believed in the efficacy of the compatibility matching process.”
In the end, it’s clear dating sites and apps significantly reduce the time necessary to find everything from casual to long-term relationships. The only question remaining is how efficient they could get with the right algorithms put in place.
Who knows, maybe in the next few years, finding an ideal partner or merely the right person for right now won’t take days, weeks, or hours but faster than we can say “swipe right”?
Image Sources: Depositphotos
