Facebook knows when you'll break-up

16:06, Nov 04 2013
LOVE PREDICTIONS: "If the NSA can use Facebook to track terrorists by mapping sociograms and pinpointing the busy content generation of fanatics prior to an attack, then why should it not be able to tell you that, statistically, today is the day your loved one leaves you?"

All of a sudden, love not money is the theme of the tech world.

In October, Spike Jonze's latest film Her premiered in New York, offering viewers the chance to see Joaquin Phoenix swept off his feet by a futuristic update of Siri - an instantly available personal assistant and spirit guide.

Last week, two researchers from Cornell and Facebook revealed that not only can statisticians, to a chilling level of accuracy, identify romance through Facebook behaviour, they can also detect the alarm bells of a failed relationship many months in advance.

If you've been in a long-term relationship over the last few years, as I have, you basically miss it. But the speed with which people are being trained to date in different ways - from geolocating apps like Grindr, to the lurid exposures of Snapchat and the ballistic Facebook activity of singles trawling through accidentally public profiles for points of connection, (or Neill Straussian trance themes), is intimidating.

I was snapped out of my bubble yesterday when a young single man who has been lurking on dating sites for the better part of the last year showed me Tinder.

Tinder doesn't really make sense until an addict demonstrates it to you.


This young man has for a while been quietly using a variety of online sites to masquerade under embellished identities at forthright first dates, but the zeal with which he thumbed through local Tinder profiles was unique.

To use the app you flick through photos of strangers, approving or disapproving them, whittling down your dating circle. Those you've rejected will no longer see your profile -- you are effectively erased from their second life and they from yours.

Like Bejewelled Blitz and Angry Birds, Tinder is one of those smart phone apps that through a wily combination of colour and movement seems to have made the act of poking and flicking at a small glass rectangle bewitching. The app is mindlessly addictive. After seeing a few photos rejected or approved I could already feel myself being pulled in.

Within a few minutes it has trained you to date in an entirely new way.

Within a few minutes the old impulse to scan a crowd for a pretty face is redirected towards a card deck of strangers who can be discarded or collected during every idle moment.

Approaching dating through sophisticated online apps is of course not that new. The battles between sites like eHarmony, True.com and Perfectmatch to refine and market a 'love algorithm' occupied the attention of many an active dater through the noughties.

eHarmony's founder Neil Clark Warren patented a love matching algorithm based on 29 personal attributes, arrived at through regression analysis of 5000 married couples.

Perfectmatch relies more heavily on the Myers-Briggs personality test, a classification tool developed in the 1940s and built around Carl Jung's theory of archetypes.

True.com uses a 99-point test that new users must fill in to find an ideal match (and to help refine True.com's regression formulae).

But social networks go one better by teaching you to express your personality, make friends and fall in love in ways that can be measured and monitored continuously.

The much discussed research paper released in late October, Romantic Partnerships and the Dispersion of Social Ties: A Network Analysis of Relationship Status on Facebook, suggests there are already some fairly accurate statistical measures of love on Facebook. Basically, if two people are connected closely on Facebook, and their mutual friends are also very interconnected, then they're very likely to be in love. If they say they're in a relationship, but a snapshot of their social activity shows they haven't managed to introduce their friendship groups to each other, then the relationship is probably stuffed.

Facebook was originally set up as a match-making site for college kids, so it's not surprising that statisticians of intimacy should have a field day here.

As the paper's authors state: 'As people use on-line social networks to manage increasingly rich aspects of their lives, the structures of their on-line network neighborhoods have come to reflect these functions, and the complexity that goes with them.'

The ability to predict relationship breakdown is a lucrative tool, and failing lovers are sure to start receiving targeted ads about couples' therapy in the near future. But what ultimately happens when the social programs we use end up having a better ability to predict the course of our lives, in a totally dispassionate way, than we do ourselves? When social networks are able to 'manage' the 'rich aspects' of our lives with greater attention than even than our closest friends and family?

Will our smart phones begin casually warning us against the fire that consumes when, for instance, we are caught staring for minutes on end into the radiant headshot of our beloved, like Narcissus teetering on the edge of his silvery lake?

Will jealous lovers pay extra for more granular Facebook metrics so that hard stats may confirm the infidelity they feel is being cruelly pursued behind their back? (But not beyond the unblinking gaze of Facebook's Santa Clara server farm).

That creepy feeling you might be having now is the dawning recognition that the Quantified Self movement has discovered love, and that the answer to all these questions is 'yes'. 

If the NSA can use Facebook to track terrorists by mapping sociograms and pinpointing the busy content generation of fanatics prior to an attack, then why should it not be able to tell you that, statistically, today is the day your loved one leaves you?

After all, love and terrorism are about as rational as each other. And to the dilettante user of Tinder, almost indistinguishable.