Free Love, eHarmony, Matchmaking pseudoscience


Nice article in today’s New York Times by Shaila Dewan:

When Misty Terrell turned 28, she happened to see an ad for a special deal on the dating site eHarmony­ and decided it was time to get serious about her love life. Terrell felt pretty optimistic. The site claims responsibility for 542 marriages a day through its “scientific approach” to finding soul mates: an exhaustive questionnaire, the trademarked “29 dimensions of compatibility” algorithm and its clinical labs where psychologists spend hours analyzing couple interactions. For this sort of comprehensive matchmaking, the company charges $60 a month, which is far more than most dating sites, but perhaps something of a bargain when it comes to finding true love. Terrell signed up to receive five potential matches a day for six months.

I checked out eHarmony for myself, curious about their personality tests. I wasn’t impressed; mostly very basic and obvious compatibilities, and unlike many carefully-constructed personality tests, no attempt to detect the more obvious liars or people with either inflated or depressed self-esteem. But as screening, they do tend to keep you from matching the most unlikely sorts, which does save some time.

Her first encounters, however, were not all that great. One guy’s mother chauffeured them to dinner; another date took her to the Chili’s where his ex-girlfriend worked. So Terrell tweaked her settings to encourage better potential matches. She unchecked the box for sci-fi fans but still remained unimpressed by the selections. “It’s kind of like, Whom am I not getting introduced to?” she says.

So she applies her own prejudices (however much some sci-fi fans might resemble Comicbook Man from the Simpsons, most do not) and tries to outguess the algorithm. That didn’t work.

Nowhere are the middleman’s limitations more evident than dating websites. Consider, for instance, that they don’t even do the thing we perhaps most want them to do: vet potential matches for truthfulness. As a result, you almost have to assume that the lovelorn are lying about their height, weight and income; the entire online dating market, despite its immense popularity, is a giant buyer-beware zone. Some dating sites have tried to address this, writes Paul Oyer, the author of “Everything I Ever Needed to Know About Economics I Learned From Online Dating,” including a Korean site that checks national registration forms, diplomas and proof of employment. Oyer suggests that more and more companies will compete in this heavily vetted space. But it’s difficult to see that type of scrutiny — in which our profiles are written by some third parties in white coats, after a weigh-in and a background check — going over in the United States, where privacy concerns are paramount. In the meantime, that buyer-beware zone is likely to continue.

Vetting backgrounds is something one should always do, but only after finding a promising candidate. This is not time-consuming; and since you are of course cautious and setting up any first meeting in a public place, you are free to turn on your heels and leave if the lying was too blatant. The real question is whether the candidates the service brings you are worthy of your time in going through them. People who sign up on dating sites tend to be one of two types: those who are busy and have a limited social network that has few people they find attractive, and those who are found unattractive by most and are trying online out of desperation. The first type is a good pool to look in, the second, you wish to avoid contacting — luckily most will show their flaws even in limited interaction online, or on the first date (like the man who had his mother chaffeur the date!)

The good news is that the more seemingly useless brokers are, somewhat counterintuitively, the more valuable they can be in signaling our interest — what Oyer might call the “money to burn” move. If anyone can wink at you free on a dating website, or for that matter beam in a job résumé, their actions don’t mean much. On the other hand, if someone fills out hundreds of questions and pays $60 a month — or in the case of a job applicant, researches a company and writes a detailed proposal — it signals a much deeper interest.

So, on some level, an expensive broker does nothing more than indicate the level of your game. Mikolaj Jan Piskorski, a Harvard Business School professor and author of “A Social Strategy,” examined hundreds of thousands of interactions on dating sites and found that the profiles people view on eHarmony­ are very similar to the profiles people view on other sites. The vaunted matching algorithm, he says, doesn’t really do that much that you can’t do for yourself. And as much as we may appreciate having our choices limited, if only to save us from being overwhelmed, from a purely economic standpoint, there is no benefit to limiting your own options, even if it means getting sucked into a time-consuming rabbit hole.

The site’s clients, therefore, are at least motivated and can follow through on complex, multi-step actions, including putting up some cash. Which does winnow out lots of losers and dangerous sorts. Your future stalker, however, is undeterred and works hard for the opportunity to meet you!

What is more valuable, Piskorski says, is that eHarmony­ limits its other members’ choices. In other words, it reduces the competition and makes the market smaller. That means that people whose highly visible characteristics might otherwise disqualify them from consideration (short men, older women) are more likely to get a fair hearing on the site. In one paper, Piskorski and his co-author, Hanna Halaburda, went so far as to theorize that a broker could make selections completely at random and still benefit you, simply by limiting the options on both sides of the transaction. “Suppose the broker was clueless,” Piskorski says. “All that broker did was restrict choice, just match people randomly. It’s what you worry that the broker is doing. Would people still pay for that? Yes.”

And this is particularly important, Piskorski says, for people in a hurry. “Our entire economy has been built on the idea that more competition is better,” Piskorski says. “It drives innovation and reduces prices. But if everyone competes with everyone else, no one actually wins. Then it is better to restrict competition.” As much as consumers may be intoxicated by the prospect of the democratizing force of the Internet, or the notion that everything should be free, most of them simply don’t have the patience to put up with it. You may make more money by selling your house on your own, but if time is a factor, an agent can sell it faster. If you have all the time in the world to date and don’t mind doing it, you don’t need to pay eHarmony­. But if you feel that time is running out and you want to meet other people who want a serious relationship, you should.

This is very poorly expressed. Limiting choices at random is not helpful; limiting undesirable others’ ability to see you and waste your time is. This is why Tinder has done well: it allows women, otherwise very hard to lure into a dating app where men can harass them, to be seen and spoken to only by men they have already selected as suitable. Since this greatly increases the number of quality women on the site, it also serves the interest of male clients.

In the end, dating sites are another tool to meet people. They can save time and energy over real-world meeting places, and perhaps allow you to consider more carefully personality over such factors as height and immediate sex appeal (which, as I argue in the book, are not useful guiding factors for long-term partner choice.) If you have unusually specific needs (say, your partner must be Jewish and you live in a town with few Jews), they can be invaluable. But for most it will still require a lot of patience and care.

And the article goes on to note that Misty Terrell met her future husband on eHarmony just as her subscription was running out.

PS — At a reader’s suggestion, I joined okCupid and answered 100+ questions. What was interesting is that 1) There were no real attachment type questions; and 2) There were intelligence testing questions requiring some thought. This means that at least smart people can seek out smart, competent people willing to sit through a lot of puzzles.

All dating sites have a business model dilemma: the more questions they ask and the more difficult and intrusive the sign-up process, the fewer customers they will have signing up. Most people try these things on a lark then get sucked in by the real people they are presented with to commit more deeply. okCupid is probably smaller as a result, but may have a higher quality customer. And still they rely on unreliable self-reporting and don’t really go after the most critical factor, attachment type.

One comment

Leave a Reply