science of us

What Does It Mean to Quantify Desirability?

Photo: Sam Diephuis/Getty Images

Recently, some relatives of mine were telling a story about a California wine tour they’d been on, one where the guide had kicked things off by announcing that there were only really two types of wine. Did anyone, he asked the group, want to guess what they were? Red and white. Nope. Dry and sweet. Wrong. Good and bad? Closer, but no — the two types of wine were the kind you liked and the kind you didn’t.

It’s a little funny how much of a revelation this was — both to them, in the moment, and to me, in hearing the retelling. On the one hand, well, duh, of course it’s the most logical organizing principle. An expensive bottle of something or other may win fancy wine awards and still taste like garbage on your tongue. You like what you like, and that’s really the only distinction that matters. It’s true for wine, but also really for anything. On the other hand, though, it feels a little … messy. I like fruity reds may not be a universally true statement — of course there are going to be some out there that you think are gross — but at least it’s a start. It gives you something to grab on to.

And that matters, especially when you zoom out beyond this wine example. So much of what we do, and what we like, and what we choose, is driven by an ever-present, unfilled need for structure in a chaotic world. Humans like certainty, and we like patterns that help us achieve some semblance of it.

Which is why I have some complicated feelings about a recent study in the journal Science Advances about how online daters tend to pursue people slightly out of their league. You may have already read about it, but the main takeaway is that on average, people send the most messages to users who are 25 percent more desirable than they are. It’s a very precise number, 25 percent. Fascinatingly precise, and also a little confusingly so: Can you really, with that level of objectivity, quantify desirability?

On the most literal level, yes, because, that’s what the study authors did, using messages received on a single online dating site as their metric. If each message is an expression of desire — to meet someone or at least flirt with them a little, if not necessarily to date them — then more messages in a person’s inbox must mean more desire directed their way. (Using PageRank, a system used by Google to evaluate the importance of a website, the researchers also assigned differing weights to each message based on the desirability of the person sending it.)

According to study co-author Mark Newman, a professor of statistical physics at the University of Michigan, one thing their data didn’t capture was the intention behind this sort of romantic striving. “You can imagine two possible explanations. One is that everybody’s being ambitious,” he says, “but another possibility is that people just mis-estimate where they stand. Maybe I’m a five, but I think I’m a seven, so I’m sending messages to people who are sevens.”

It’s likely an impossible question to untangle, says biological anthropologist Helen Fisher, a researcher at the Kinsey Institute and a scientific adviser for Match. For one thing, there’s no overarching pattern of how people view their own value as a partner, she explains — we all have different levels of self-esteem, different personalities, different strengths and weaknesses when it comes to relationships.

More importantly, though, once you exit the parameters of desirability set by a dating site, it becomes less clear whether you can say that any of us really has a so-called “mate value.” Researchers — and regular people, for that matter — often like to talk about dating as a marketplace. “There are buyers and sellers,” says David Frederick, a psychology professor at Chapman University who studies attraction. “There are some people who have highly desirable traits and are going to be more in demand, so they’re going to be able to be choosier about who they mate with. And then there are people who are going to be lower in demand and generally higher in supply, and those people are going to be buyers in the market.”

To use an extreme example, there are more people trying to snag, I don’t know, Drake, than there are people Drake’s trying to be with. But once you get away from the outliers, it’s also fair to say that we’re all shopping for slightly different things on this market. All the characteristics that make a person romantically or sexually appealing can be divided into two categories, Frederick explains. There are “consensually valued traits,” or things that everyone more or less agrees are desirable: good looks, intelligence, a sense of humor. And then there are “idiosyncratic preferences,” or your individual turn-ons — you’re a sucker for biting sarcasm, say, or green eyes, or an unusually deep knowledge of World War II history.

These two categories don’t always work in tandem. Sometimes, a person’s idiosyncratic preferences run counter to, and override, the typical consensually valued traits — maybe you’re really into big noses, say, even though they may not be part of what we would consider a stereotypically beautiful face. Or you find something irresistibly tragic about people who can’t seem to hold down a job, even though professional achievement and financial stability are two things that would typically make someone more of a catch.

And whatever your personal collection of likes comprises, they don’t always play out as anticipated once a collection of likes becomes a living, breathing human. If you think a little more critically about the true utility of Tinder, OkCupid, and their ilk, “None of these are dating sites,” Fisher says. “They’re introducing sites.”

Compatibility is a complex, often unknowable alchemy. And of the several bazillion factors that go into determining whether two people click, many can only be sussed out by having them actually interact with each other in person. That’s step two, where the real assessment of another person’s desirability happens. Evolutionary biologist Justin Garcia, who like Fisher is a researcher at the Kinsey Institute and a scientific adviser to Match, uses the analogy of fingerprints: “Every person you try and court, they’re also a unique fingerprint, so imagine you’re pushing two fingers together — every time, not only is the other finger different, but the dynamic interaction of you and the other person is going to be different.”

And then, to make it both more accurate and more complicated, imagine that each of these fingerprints is in a constant state of evolution, the loops and lines morphing slightly but noticeably over time. The way they fit together will change, too. It’s not just that someone’s desirability can vary from person to person — it can vary from year to year, or even minute to minute, within each individual. (And within that single minute, you may be scrolling through your online matches and find yourself drawn to two very different people for two very different purposes: Research has shown that we tend to have different standards for short-term flings or hookups than we do for potential long-term mates.)

That’s a lot of change. Really, it’s a lot of chaos. The search for a partner, in general, is often defined by chaos — making the first move, whether online or in person, means placing your immediate romantic future in the hands of a stranger, with all the unpredictable whims that strangers can have.

So the idea of measuring something as abstract as “desirability,” while restricting and degrading and a little bit nonsense, can also be comforting, in a way: Here is order. Here is a rulebook. Here, maybe, is relief, that in at least one regard the dating world looks the way you’re wired to want it to look. “The brain is extremely well-built to rank ourselves and our opportunities in relation to everybody else,” Fisher says. “This is a study that’s quantifying that. It enables us to make more concrete the game we’re playing and how we should play it.”

“The only real algorithm,” she adds, “is your own brain.” But it’s easy to see why another one is easier to trust.

What Does It Mean to Quantify Desirability?