Dating today: love lives decided algorithmically

Love in MidtownAs I sat down to write this, I overheard two colleagues in the break room discussing the trials of modern dating.

“How do people meet today?” asked one, almost rhetorically.

With absolutely no hesitation, the other fired back what I — only a couple of cubicles over — had anticipated.  “Online.”

Once heavily stigmatized as the course for only the most desperate of singles, online dating has in the last decade become a widely accepted, fairly standard means of finding romance.  Thanks to websites like eHarmony and, along with younger, more mobile-friendly adaptations, singles can now fish from the widest pool of potential mates our species has ever known.  And if we aren’t capitalizing on this unprecedented opportunity ourselves, we know at least a handful of people who are — with an imaginably infinite variety of motives and equally vast array of outcomes.  Indeed, according to a national survey conducted by the Pew Research Center in 2013, as many as one in ten Americans have used a mobile dating app or website.  Perhaps more significantly, the survey found that a whopping 5% of all committed relationships in this country began via something as simple as a “like,” “swipe,” or “match.”  This figure doubled from 2005 to 2013, indicating that the online dating industry, which reportedly made as much as $2 billion last year, will go on surging as the accessibility of these platforms increases and our attitudes towards them continue to shift.


If someone were to ask me 5 years ago — or even 10 months ago — whether or not I’d consider jumping into the online dating game, I imagine I would have laughed and quickly changed the subject.  The mere thought of eHarmony’s mawkish television ads set to “This Will Be” would’ve been enough to elicit a very real response from the pit of my stomach.  These days, however, I too count myself among the millions of online daters, and it was here in this crowded, noisy digital space where I met my first boyfriend.

Online dating is far simpler and dare I say, cooler than it was in the mid to late 1990s.  As opposed to the first websites with their exhaustive questionnaires and profiles to complete, mobile apps like Tinder and Hinge allow users to check out potential matches moments after signing up for the service — an immediacy that appeals to their large millennial base.  If the Ashley Madison hack illustrated anything other than the most obvious fact — that being, online privacy can never be 100% guaranteed — it is that there’s a site for just about anyone looking for romance, from a twenty-something who can’t be bothered to set up an OkCupid profile to a forty-something in earnest pursuit of extramarital bliss.


Photo courtesy of The Grade

But however varied, every one of these platforms — at least to some extent — relies on a series of algorithms to bring strangers together, and it is not just dating sites running on formulas.  The back end of many mobile and web platforms we use daily filter what we ultimately see, and though there are certainly more choices to be had when it comes to deciding where to eat, what to watch, and even whom to love, our choices have arguably in this way been partially made for us.  In his new book Modern Romance, actor and comedian Aziz Ansari turns to leading social scientists to understand how best to navigate this strange age of choice.  For the purposes of today’s post, I too consulted a number of scientists, starting with Rashied Amini, a systems engineer at NASA’s Jet Propulsion Laboratory.

Why I decided to put my faith in a soon-to-be physics PhD when it comes to matters as messy and seemingly unscientific as those pertaining to the heart may confound readers.  But upon learning that Amini was developing Nanaya, an app that could predict a person’s 7-year romantic outlook, I — only recently single — was intrigued.  And it was his impressive scientific resume that inspired my confidence.  One of Amini’s projects for NASA involved formulating ways of building bases on the Moon and Mars.  In addition, he counted himself among a team of researchers developing optimal mission architecture for redirecting an asteroid.  How could I not take this man seriously?  Seriously.  If Amini could theorize about constructing bases on Mars, surely he could take a crack at human relationships down here on Earth.

“Getting the probability of finding someone has already been done…but it’s not very interesting nor does it actually express the real odds,” says Amini.  To calculate the odds, he tells me, one must employ complex systems modeling, a technique Amini used while dreaming up those lunar and martian bases for NASA.  “[F]inding the real odds takes sociological modeling,” Amini continues.  “For instance, a person who has been in the same place for 5 years is going to be meeting fewer people than someone who just moved there.  A person who is extroverted and open to experience will be meeting more people than someone who is not.”

But what good is Nanaya for those of us who aren’t looking for love?  For users who believe they’ve already found it?  Amini has them covered, too, using something called stochastic options analysis to determine if two people should stay together.   At this point, I admit he’s lost me.  Stochastic processes?  This is the randomness of how things actually work in real life, Amini explains.

So, what if by stochastic options analysis, a couple learns their relationship is doomed to fail?

“Mathematically, it’s important to state we don’t predict relationship success or failure, but whether or not someone will be better off single or in a current relationship,” says Amini.  That being said, when the app is released at the end of the year, this service will not be provided.  Not until Amini has enough data and control subjects, that is.  Still, after an hour and half of walking me through the Nanaya prototype, Amini sent over a 22-page “Romantic Options Report,” which interestingly enough, included an appraisal of my last relationship.  And though I should probably mention that it bore no influence over my decision to ultimately go solo, there was a certain affirmation in knowing that all 5 graphs indicated I’d be happier single than with my former boyfriend.

It is in part out of concern that users will take this portion of his love analysis too seriously that Amini is withholding it.  “The bottom line,” he tells me, “is that Nanaya is a simulation, and no simulation is the same as reality,” adding, “the last thing we’d ever want to do would provide a basis for a decision that would hurt anyone.”

Probability of Finding a Match for User

But in my case, the proverbial box had been opened and my interest piqued.  Better off single, eh, Nanaya?  Well, if I were indeed destined to be alone, I could only guess that the pages detailing my romantic outlook would be bleak, if not abysmal.  Not so, suggested 4 other graphs titled, “Probability of Finding a Match,” according to which the chances of me meeting someone — “any match,” it specifies — peak at around the fourth quarter (see graph).  I interpret this to mean that in as little as one year’s time I’ll be perfectly content with your average Joe.  As for my ideal match, he remains elusive until roughly the 24th quarter, so as for my happiness, I’ll be waiting in line for possibly another 6 years.

However dismal my prospects, the written assessment located on the final pages of the report somewhat buoyed my spirits.  “You fare on the higher end with respect to romantic opportunity,” it reads, “primarily due to general open-mindedness to most traits.”  But if luck in love is where my open-mindedness meets romantic opportunity, where exactly should I be spending my free time?  Coffee shops?  Bookstores?  Where are these matches I’m supposed to meet one to six years from now?

When Nanaya launches, users with similar questions will be pleased to know they’ll receive the demographics of their ideal matches, along with a helpful map illustrating where they’ll most likely find them.  As for me, my report suggests forgetting the possibility of a real life meet cute and settling for the far more likely scenario of finding someone online.

Fortunately, for the sake of my sanity and that of other online daters, this overwhelming supply of prospects has been significantly narrowed by a series of algorithms, as well our brains which naturally sift through the options with their own set of filters.  Sites like eHarmony try to operate in coordination with both.  Specifically, it combines a “[u]ser’s stated matching requirements (i.e. self-select criteria); research conducted on married couples which has examined the elements which predict relationship satisfaction (i.e. compatibility models); and predictive analytics based on machine learning algorithms applied to user view and communication behaviors on the site (i.e. affinity models),” says Steve Carter, the site’s Vice President of Matching.

With the help of all these formulas, one might reasonably conclude that online dating removes some of the complexity of meeting offline, but Sheena Iyengar, Columbia professor and author of The Art of Choosing, is not so convinced.  Indeed, it is these very filters that Iyengar views as being so problematic and possibly detrimental to online daters.  “Due to the search and filter format of online dating, we end up holding more strictly to our criteria than we would in a more organic, real-world setting,” she tells me.  “And as a result, we can miss out on potentially great fits in other areas.”  For instance, a woman seeking men taller than five-foot-ten may be more willing to overlook this criterion if she were to meet a certain special someone offline than if she did on.


What might surprise readers is where online daters are willing to make compromises.  “Data shows that even when dating sites take into account personality, by far the biggest predictor of whether you will click ‘yes’ on somebody has to do with their photo,” says Iyengar.  So, a person’s picture matters — perhaps more than anything else — validating the millions of happy Tinder users who often equate finding singles on the photo-driven app to spotting someone attractive in a bar or gym; Northwestern psychology professor Eli J. Finkel agrees, hailing it as “may be the best option that has ever existed.”

But whether you sign up for Tinder, Hinge, eHarmony or all three, Iyengar advises having a strategy.  “[H]aving more choice only helps if you’re thoughtful about your choosing process and understand what sort of an outcome you’re seeking.  If you go into a too much choice process blindly, then on average it will leave people feeling confused, distracted, and frustrated.”

Swarthmore psychology professor Barry Schwartz echoed Iyengar’s sentiments when we spoke over e-mail.  “This is both the blessing and the curse of online sites,” writes Schwartz.  “With so many options available, standards inevitably go up, so that people are rarely satisfied.  Why ‘settle’ when there are thousands of more possibilities to examine?”

Of course, the problems associated with choice are not the only obstacles online daters may face in pursuit of romance.  Unlike more conventional ways of meeting people, most of them are connecting with complete strangers, whose intentions might vary considerably.  In some extreme examples highlighted by The New York Times, swindlers used dating platforms to specifically target elderly women; one seventy-something recalls sending her online interest nearly $300,000 before realizing she was being scammed.  And shockingly, some 6,000 people between July 1 and Dec. 31, 2014 made similar complaints of confidence fraud.


Photo courtesy of The Grade

Far more common, however, are the digital equivalents of catcalling — unsolicited messages, photos and/or unsavory propositions.  And perhaps unsurprisingly, female users bear the brunt of it.  Mike Maxim, the Chief Technology Officer at OkCupid, told Scientific American that policing this kind of misbehavior was a continuous battle,  and Steve Carter agreed, reporting that eHarmony must shutdown 300 accounts daily.

To better address the problem, Cliff Lerner of Snap Interactive developed the mobile dating app The Grade.  As the name suggests, The Grade holds users accountable for how they comport themselves by doling out letter grades, and it makes no difference whether it’s an A+ or F; their grade is posted on their profile.  According to the company’s website, if a person does drop down to a D or failing grade, he or she will be given a warning, as well as a set of instructions on how to raise the grade.  However, those who choose not to comply are summarily kicked out of The Grade’s dating community.

“We are not aware of any dating site that has an algorithm to grade users in various categories, and make the grade available for other singles to view, while also expelling users who don’t meet community standards,” says Lerner, a 10-year veteran of the online dating industry.  He got started with his first dating site in 2005 and went on to launch AreYouInterested?, which, according to Lerner, has over 70 million Facebook installs.  It would seem he knows what works and in this case, what doesn’t.  And online daters have welcomed his new approach.

“The Grade has over 70,000 downloads in just a few months with about 25% of them in our core New York demographic,” says Lerner.  “One of our key metrics, retention, is the highest we’ve seen on any app we’ve launched, with users in our core demographic logging in 8 times on their first day and 10 on their second.”

So, thanks to industry innovators like Lerner, online dating continues to evolve, but are users any closer to finding love?  Throughout this piece, I’ve been careful when and where I use that word.  Obviously, not everyone who installs a dating app or registers with a site is looking for love online.  But for those users who are, has this technology made their quest any easier?

When I pose this question to Barry Schwartz, he leaves me with something that I doubt I’ll very soon forget.  “I think the problem here is that serious relationships must be built, rather than discovered,” he tells me.  “People are not willing to put in the time.”

I think Schwartz has cut to the heart of it — pun intended.  The mechanism through which we meet people today may have changed — even improved by some estimations — but the process of developing a meaningful and loving relationship remains unchanged.  Quite simply, it’s hard work.