Skip to main content

Hacking OkCupid to find a perfect match

Hacking OkCupid to find a perfect match

Share this story

love-heart
love-heart

In the summer of 2012, Chris McKinlay was trying — and failing — to find love. Like 40 million Americans, McKinlay was using dating sites to search for prospective partners, but in nine months, he'd gone on only six dates. But unlike 40 million Americans, McKinlay was both brilliant at math, and as a PhD student at UCLA, had access to a supercomputer. He decided to look at his problem like a mathematician. McKinglay used statistical sampling and late night sessions on UCLA's supercomputer to break LA-based OkCupid members into distinct groups, identifying their likes and wants, before tailoring his profile to their tastes. Wired tells McKinlay's story, detailing how he became so invested in his quest for love that he started sleeping on his desk, and explaining what happened when the number of women he was matched up with jumped from a few hundred to several thousand.