Alex is a real catch. Lean and lanky, he cuts a fine figure in a Savile Row suit or hoodie and jeans. He has a sort of genial magnetism that draws you in, and a certain charisma that lets him to move with ease in any social situation. But beyond his style and demeanor, he’s generous and friendly, and an extraordinarily good listener. I know because he still remembers stories I’ve casually told him over a decade ago.
Not only is he handsome, intelligent and witty, but he’s wealthy, too. It seems the world is at his feet, but while he has a strong career in the international art world, lovely homes in several countries and a wide circle of friends, he has yet to find his match in love and life.
“Maybe I should give up,” he tells me, in an uncharacteristically forlorn and pessimistic mood. Then, that wit kicks in and he jokes that he’s ready to embrace permanent bachelor status.
He has no problem attracting women, but he is, at heart, a true romantic, with a deep yearning to find someone who truly accepts and loves him. But with each breakup, his hopes and expectations grow for the next person, which more often than not, ends in a deeper pain when it fails to work out.
No matter the age, gender or astrology sign, finding true love is never easy. But lonely hearts, like Alex, shouldn’t give up looking for companionship. Help is coming from an unlikely source, as algorithms evolve to take dating to the next level.
Like most, Alex spent his 20s in a dating frenzy, not particularly eager to settle down. But then, one day, he told me he was ready to find the “One.” He radiated confidence as he painted a picture of his near future: he’d find her within two years, and get married in a ceremony held at his family vineyard, just shy of his 30th birthday.
With a plan in place, he set out to meet women on his own. But before long, he became frustrated with the passive nature of dating, waiting for fate to put the right person in his path. So he canvassed friends to set him up on blind dates. When those didn’t pan out, he ventured into upscale matchmaking, paying thousands for high-end services that yielded few enduring relationships.
Finally, after much resistance, he tried what millions had already discovered: online dating.
“I know a lot of people who found great relationships through sites like eHarmony and the like,” he tells me. “But I still feel like ‘ugh’ at the thought of doing online dating myself.” After I slightly prod him, he admits to being embarrassed.
“What would I tell our kids? That their mother and I met online? That doesn’t sound as good of a story as telling them we met at a game, or a bookstore, or something like that.” he says. “It’s hard to let go of that wish for a ‘meet cute’ scenario you see in a million movies or hear about from your friends.”
That sentiment is shared by millions, but it hasn’t stopped millions more from storming these sites. According to Census data, of the more 100 million people who are single in the U.S, in 2011, 40 million have tried online dating in 2012, Mashable reported.
It’s a big business: in 2012, revenues totaled short of $2 billion, making online dating the top source of paid content on the Internet. In fact, these millions of people will, on average, spend about $240 a year on sites, and today, Match and eHarmony take credit for about one-in-five marriages.
Of course, dating services have existed well before the Internet. When modern newspapers appeared in the late 17th century, “matrimonial agencies” began to place advertisements on behalf of men who paid them to find them wives, according to H.G. Cocks, a professor and lecturer at University of Nottingham, and author of the book “Classified: The Secret History of the Personal Column.”
And according to the New Yorker, when Lewis Altfest, an accountant from Queens, ran across a supercomputer at the 1964 World’s Fair, he believed the questionnaire it used to match people to foreign pen pals could be adapted to do the same for dating. So he called his friend, Robert Ross, then a programmer at IBM, to alter the system, and a year later, they launched “Project TACT,” short for “Technical Automated Compatibility Testing,” to become New York’s first computer dating service.
The Internet took the ideas of classifieds and TACT to cast a wider-reaching and more efficient net. The first generation of dating sites, like Matchmaker.com, created in 1986, were merely bulletin boards where lonely hearts posted advertisements that described what they were looking for. They were similar to classifieds, but easier to sift through, so browsing through hundreds or thousands of results became more convenient.
Then, dating sites shifted to sortable databases. With thousands of potential mates to sort and sift through, matches could be winnowed and drilled down by preferences. Looking for a non-smoker? Dating only men six feet or taller? Looking for a blonde under the age of 34? It was easy to “customize” a date to any heart’s liking.
But as people flocked online, combing through the matches became overwhelming, prompting sites to add layers of innovation to make sorting more efficient and targeting more accurate. Some sites, like JDate.com and ChristianMingle.com, specialized in matching like-minded people who sought cultural or religious preferences, while others leaned on algorithms and formulas that promised a more scientific and precise method to make matchmaking.
OkCupid, for example, famously used an algorithm to calculate and “match” the likelihood two people would have a successful relationship.
Alex thought the plethora of options would yield some success, but he tells me the opposite was true — some of his worst matches came from dating sites. He tried them all, from OkCupid to eHarmony to even JDate — there was always something “off” in online dating, he says. “It’s just not working, and I’ve tried everything.”
The problem with online dating is that it relies on the assumption that we’re open and honest about whom we purport to be and what we want — but that simply isn’t the case. According to the University of Wisconsin, four-in-five online daters admit to lying on profiles. Whether they add an extra inch to their height, shave off 10 pounds on their weight, post an old picture from a decade ago, or subtract a year off their age, people routinely fudge even basic facts.
Of course, there’s a reason we lie: online dating is extremely brutal, showing our ugliest biases. When Quartz analyzed data from Facebook dating app, “Are You Interested,” for example, it found that black men and women received the lowest response rate on the service, while white men and Asian women garnered the most, revealing entrenched stereotypes and fetishes in the idea of a “color-blind” society.
Now, in his late-30s, Alex has given up on online dating. “I’ve come close. I’ve dated beautiful, accomplished women, and I’ve even had real relationships with some of them,” he says. “And yet… nothing. Either I don’t feel it, or she doesn’t, or something is missing. What am I doing wrong?”
The problem may not be bad luck or unresolved issues, but the methodology of online dating itself. According to Scientific American, many scientists believe online dating, ultimately, fails to produce long-lasting relationships because the algorithms used fail to account for some important factors.
For example, psychological studies often report the strongest predictors of solid relationships are based on how a couple interacts with one another, and how they handle stress together — factors a formula can’t predict until people actually meet. While sites can gather fun data on personality, attitudes and habits, those traits are rarely the signposts of a successful relationship.
As my mother used to say, nothing shows more about a relationship than taking a trip together to an unfamiliar place.
But the problem isn’t merely in algorithms, according scientists at Boston University, Harvard and MIT, the very model of online dating itself is flawed. When researchers analyzed dissatisfied online daters, they found experience was, too often, like online shopping.
You compile a list of desired attributes in a mate, and then, click search to browse through a variety of matches. It’s like scanning boxes of cereal on a store shelf. The problem is that, too often, we begin to form a picture of a prospective date before we even meet them — and then, when we do, the reality simply doesn’t live up to the vivid pictures in our head.
The shopping model has limitations, too. While we know whether we want a toaster with two or four slots, it turns out we’re not especially good at gauging what we want in relationships. “With people, even when we have well-articulated stories about the kind of person we want to meet, we’re often wrong,” Michael Norton, who worked on the study at Harvard, noted. Social psychologists have long known that what we say, and what we actually do, are often two very different things — and nowhere is that truer than in love.
According to Northwestern University, while men and women tend to say they look for certain qualities in a potential mate — men emphasize looks, and women say money — in an actual setting, like speed dating, they often choose differently. What participants said they wanted beforehand, and what they actually said they liked, afterwards, had little in common, because desired traits — like socioeconomic status, income level or height — were often superseded by qualities that are actually important to us — like rapport, a sense of humor, or sexual compatibility.
Are they kind, a good listener, responsive? Do they have a nice laugh and seem easy to please? These factors aren’t easily measured by a questionnaire, and yet they’re crucial to building chemistry, and that undefinable “spark,” that makes us want to get together.
But if online dating emphasizes “compatibility” over ineffable, immeasurable factors like chemistry and interaction, a number of dating apps are going in the opposite direction. Instead of lengthy questionnaires to analyze compatibility, software like “Are You Interested” and “Tinder” rely on social networks to spark an interest, often based on just a short bio and a photo.
With Tinder, you simply swipe through photos of people nearby, and then message the ones you like. From there, it’s up to you to decide if there’s a connection, much less any compatibility. Other looks-centered services, like FaceMate, are gaining headlines, too, underscoring a “back to the drawing board” approach in dating, mating or simply hooking up.
Alex tried Tinder, but he hated it. “It might have been fun in my 20s, when dating was more of an adventure,” he says. “But now I show up to coffee and discover it’s with a 30-year-old with the sophistication of a ’80s Valley Girl, and I feel like I wasted an hour of my life.”
Sometimes, I think his exhaustion with dating stems from high expectations, but it mirrors a changing picture of marriage in the 20th century. Until recently, marriage was often seen as a social and economic arrangement, benefiting two partners and their families, so it was helpful to share similar goals and outlooks on life — or, at least, get along.
Then, marriage evolved to a companion-based model — someone to share our interests, raise a family with, and experience life together as we get older. In a sense, online dating, with its emphasis on shared outlooks and desired traits, reflects that picture: we look for someone who wants a similar lifestyle, and who we get along with.
But someone like Alex longs for something else entirely — a communion that comes from a deep emotional and even spiritual connection. A soul mate. But when each person has a different, deeply personal definition of what a soul mate is, one that’s not easily articulated, a tidy algorithm can’t possibly quantify it.
Not that scientists aren’t trying. They’re evolving a generation of algorithms to approach the science of human attraction and chemistry. Dating sites will soon take cues from Netflix and Amazon, which determine what we like and serve us similar items. Sites like Netflix or Amazon don’t make us sit down and fill out questionnaires — instead, their algorithms “learn” from what we watch, click on and browse, to recommend options based on our tastes.
We could tell Netflix we’re interested in arthouse dramas, but if we end up watching or renting the entire Meg Ryan canon through the site, it’ll recommend the cheesy romantic comedies our actions say we actually prefer.
Kang Zhao, an assistant professor of management sciences at University of Iowa, believes dating sites can do the same thing, and his team is creating an algorithm that learns from a person’s clicks and contact history to recommend dates of interest. The assumption is that our real behavior is a better sign of what we actually want, despite what we say or think we want.
By developing a model to combine our tastes and what we find attractive, Zhao’s algorithm improved the paltry 25 percent reciprocation rate in an unnamed popular dating service to 44 percent.
“Those combinations of taste and attractiveness do a better job of predicting successful connections than relying on information that clients enter into their profile, because what people put in their profile may not always be what they’re really interested in,” Zhao told Forbes. “They could be intentionally misleading, or may not know themselves well enough to know their own tastes in the opposite sex.”
There are signs, though, that Alex may be off the market by the time a “Netflix of dating” emerges to capture a piece of the online dating pie. Recently, he met someone he likely would’ve never met through an online site. For one thing, she’s older than he was looking for, and she’s a single mom, in a non-glamorous profession — someone he describes as “pretty in a girl-next-door way,” but not a stunning beauty.
They met in the gluten-free aisle at Whole Foods, where Alex had stopped by after the gym, both looking sweaty and dressed in ratty gym clothes. And yet, he liked her smile and her calm voice, and they both made one another laugh easily. He got her number and they met for coffee, which lead to a first date, and a second date, and then more.
After dating for some time, he discovered things about her that he never would have thought to put in a profile, or check off on a questionnaire: her easy self-acceptance and confidence, her humor, kindness, adventurous sense of fun, as well as the ability to be totally himself around her.
That’s the funny thing about love: when it works, you don’t always get what you want, but you discover what you truly need, and perhaps that’s the type of knowledge that no algorithm can really replace. ♦