Thinking about online dating? These tips will up your odds:

1) Include a photo in your profile.
2) Be flexible about your age, height, and weight preferences.
3) Stop smoking cigarettes.

These aren’t just rumors or rules of thumb—this advice has been verified by plenty of evidence and complicated statistics. Elizabeth Bruch, a professor in the Center for the Study of Complex Systems and the Department of Sociology, has been working with colleagues in LSA to figure out exactly how people find romance online. They’ve seen some telling patterns in how people choose partners.

Bruch and her colleagues examined romantic encounters in an online dating service—more than one million interactions among nearly 2,000 people in New York and New Jersey. The researchers found that people make the process less overwhelming by deciding on dates in two simple steps: screening matches quickly for obvious flops, and then investing more time to closely consider the rest.

First, people weed out prospective partners with “deal-breakers”—characteristics so unappealing that the elimination step becomes a breeze. As you may expect, a profile with no photo is a no-go. Age is also a big deal-breaker, especially among younger women who specify a range. Women tend to have height requirements: Those seeking men show a clear preference for partners who are about six inches taller than themselves. Men, too, prefer pairing with relatively shorter—and thinner—women. A smoker’s profile gets fewer clicks by a factor of 10.

“People tend to invoke their deal-breakers earlier in the mate-choice process,” Bruch says, and “a bigger constellation of attributes” enters the next phase of finding a date. After the elimination round, people continue their search with a more discerning eye. They apply more of a sliding scale to decide whether to send a cutie a message. As long as a partner is the right age or height, factors such as education level, income, and favorite movies are more happily compromised.

Big Data, Big Decisions

Bruch and her colleagues found some intriguing patterns, but also some wildly anomalous data points. Amid the majority of profiles that receive a mere handful of messages, one standout case is a dating profile the researchers call the “movie star.” This 23-year-old New York City woman interested in sexual relationships received more than 5,000 messages during the month the research team pulled their data, but responded to just a few messages each week.

Your inbox might collect fewer messages than that, but don’t lose hope. While your locale may have fewer fish in the sea than a populous place like New York City, Bruch says, “People in, say, Ann Arbor might invoke fewer deal-breakers than people in New York simply because they’re facing a much smaller set of choices.” But that doesn’t mean relationships are better in one place than another.

"One of our striking findings is that as cities get larger, the number of incoming messages goes up quite dramatically," Bruch says, “but the number of quality conversations among cities is almost identical. In smaller places, people don’t need to work as hard to manage their options. People in New York are winnowing like crazy in the beginning. In smaller cities, they’re winnowing, too—just not to the same extent."



The same deal-breaker strategy people use in online dating pops up in many other big decisions. It’s just easier to narrow down the options as a first step, and then evaluate the rest using more refined criteria—especially when you’ve got limited time and resources.

Bruch first encountered the math she needed for the online dating study by seeing how marketing researchers observe the ways people decide among products in major purchases. “I discovered that choice models in marketing were incredibly sophisticated in describing human behavior,” Bruch says. “I quickly realized that online dating data, and actually a lot of online interaction data, are incredibly rich in terms of seeing, for example, the criteria people use to search, who they looked at, who they decided to pass up, who they decided to consider, and of course who was willing to consider them. It was a rich source of data for understanding decision making.”

Also applicable: applying to colleges, finding a job, and choosing a place to live. “It would be great to apply our decision-making models to a housing search website,” Bruch says.

While dating data help Bruch piece together how people make decisions in the online “meet” market, they don’t say anything about whether the decisions turn out well. That’s a big study in human behavior that Bruch has some plans to pursue. As of now, the data can’t show how many message threads lead to a date in real life, let alone if those dates result in a happy relationship or devastating breakup.

That’ll be up to you.


Illustrations by Julia Lubas