Facebook knows a break up is coming before you do

OH, DEAR. You might be heading for splitsville and you didn't even know it. But Facebook did. Here's the troubling formula behind whether your relationship will succeed or fail.

A graphical representation of one person's network neighbourhood on Facebook.

FACEBOOK can predict when you're going to break up.

Yes, apparently the fate of your relationship is not written in the stars but in your social circle.

Cornell University researcher Jon Kleinberg and Facebook senior engineer Lars Backstrom proved as much when they presented their co-written research paper at a social computing conference in February.

The researchers took the datasets of 1.3 million Facebook users listed as being in a relationship, and found that the more well connected their mutual friends were, the more likely they were to break up.

This theory is described as dispersion.

Couples with high dispersion have mutual friends who are not well connected.

Couples with low dispersion have mutual friends who are well connected.

Therefore the Facebook theory suggests if you and your partner share the same social circle on Facebook (low dispersion), you're less likely to have your own lives and therefore the relationship is more likely to implode.

A healthy relationship, according to Facebook, is one where both partners have connections to a lot of different groups of people, even if those friendships aren't particularly strong.

"Instead of embededness, we propose that the link between and an individual u and v his or her partner should display a 'dispersed' structure: the mutual neighbours of uand v are not well connected to one another and hence u and vact jointly as the only intermediaries between these different parts of the network," the researchers wrote in the study.

In a nutshell, get your own damn lives and friends.

Of course, this algorithm might not take into account the fact that some couples don't take their social circles on Facebook particularly seriously and therefore might look like they don't have as wide group of friends when they actually do.

Probably because they are out living their lives.

A graphical representation of one person's network neighbourhood on Facebook.

HOW IT WORKS

Take the above diagram, it represents the social network of a single Facebook user who sits at the centre of the network. The cluster at the top is the person's co-workers, the dots on the right represent that users' university friends, the dot in the lower left of the graphic with links to two different clusters but sits kind of separate from the two groups is the person's spouse.

"A spouse or romantic partner is a bridge between a person's different social worlds," Mr. Kleinberg explained in an interview.

The researchers correctly identified a user's spouse 60 per cent of the time, that's better than a one in two chance.

When the algorithm failed, it meant the relationship was in trouble.

And if that isn't the most depressing thing we've learned this week, we don't know what is.