Does Facebook Know Your Future?

December 6th, 2013 by Andrea Bennett

Researchers are saying an algorithm can predict the success of your relationship.

Brought to you by Liberty Mutual's
The Responsibility Project

Hindsight may be 20/20, but the results of a new study suggest that when it comes to predicting the success of relationships, Facebook isn’t far off.

According to researchers from Cornell University, it turns out that the way in which your friends are dispersed – and connected with your significant other’s – may be a successful gauge of your future.

The research team of Jon Kleinberg, a computer scientist, and Lars Backstrom, a senior engineer, examined the social connections of 1.3 million people, selecting randomly from among all users who are at least 20 years old and list a spouse or relationship partner in their profile and have between 50 and 2,000 friends.

One big finding was that the total number of mutual friends two people share (“embeddedness” in social media-speak) is actually not a great predictor. Rather, it’s the idea of “dispersion,” or connections to friends from different parts of each other’s lives – your connection to your significant other’s friends from high school, their family, and the people they work with, for instance – that’s more telling.

“A spouse or romantic partner is a bridge between a person’s different social worlds,” Kleinberg told The New York Times. He said that the dispersion algorithm was able to correctly identify a user’s spouse 60 percent of the time. It also did well with people who declare themselves to be “in a relationship,” correctly identifying them a third of the time – a 1 in 3 chance compared with guessing 1 out of 50 friends. Interestingly, the research, which tracked users every two months for two years, also found that a couple in a declared relationship and without a high dispersion on the site are 50 percent more likely to break up over the next two months than a couple with a high dispersion.

Do you believe your future relationship status can be predicted using an algorithm? Weigh in.