Faculty + Staff

His research clears the way to finding his true love

Back in the day, love potions had numbers.
Today, they have algorithms.
In 2012, UCLA Ph.D. student Chris McKinlay spent most of his time working on his dissertation dealing with high-dimensional clustering. In his spare time, he answered questions on the dating website OkCupid.
McKinlay decided to apply the same high-level mathematics he was using in his doctoral research to understand how OkCupid really worked.
“I was interested in meeting someone," said McKinlay, who is now doing postdoctoral work at the University of Minnesota. “But it started off as an intellectual exercise. Once I realized how effective [the site’s method] was,  though, I decided to take the dating part more seriously.”
OkCupid matches singles through a series of questions: You state how you’d like someone else to answer and rate how important each question is to you. To view another user’s answers, you must first answer the same questions yourself.
McKinlay said most users start off answering the questions ad hoc, without regard to how each answer will affect one’s matches in a high-dimensional space.
UCLA alumnus Chris McKinlay used the knowledge he gained from working on his dissertation on high-dimensional clustering to find a more efficient way mathematically to find his true love on an online dating site. The result: He found Christine Tien Wang, an artist who earned her M.F.A. at UCLA in the same year McKinlay got his Ph.D. They plan to wed in 2015.
After a month of gathering data, McKinlay figured out which questions were most important and that responses clustered around a set of common beliefs. “When 20,000 women answer enough questions, you get seven different archetypes, or clusters,” he explained. Armed with his research, he adjusted his own profile to better match the archetypes that most appealed to him.
The results were immediate and dramatic.
“I started receiving several hundred profile views a day, a 20-fold increase,” he said. “And I started getting 10 unsolicited messages a day, when most men get zero.”
He found he was now a top match for a sizeable proportion of the OkCupid population. He started going on dates — lots of them.
“I went on one date a day for three months over the summer of 2012,” he said. “The first few dates were romantic, like hikes and dinners, but that became exhausting.” He switched over to 20-minute coffee dates, finding he could learn more about a person in a simpler environment.
After a while, he had to step back and figure out why he was going on so many dates. Was it just to see how long he could keep it going?
That’s when he met Christine Tien Wang,  another UCLA alumna with an M.F.A., an artist and prison activist.
“Her message was different. It was very direct and very real,” McKinlay said.
Said Wang: “The idea that women could be grouped into general clusters was interesting. As a cultural thinker and an artist, I thought about the clusters from an idea of gender performativity and psychoanalysis.”
McKinlay's research has ultimately led to a planned summer 2015 wedding and also to a book by McKinlay. "Optimal Cupid: Mastering the Hidden Logic of OkCupid" has attracted major media interest and inquiries from movie studios who think the project has a "(500) Days of Summer" feel.
“Social media is ubiquitous now, and we’re all data scientists,” McKinlay said. “The question is: Are we effective?”

This story appears in UCLA Magazine online.
Media Contact