‘Project Gaydar’ Meets Twitter, Users’ Secrets Revealed

A new program can supposedly predict sexual orientation from your Facebook friend list. We applied the same algorithm to Twitter to see what some well-known users might be hiding.

By (@JRRaphael)

September 21, 2009

Project GaydarI can tell all sorts of stuff about you from your social network friend lists. I can figure out your age, your religion, even your sexual orientation. I can predict whether you’re happy, fat, or a terrorist (or all three, although you really don’t hear about too many rotund and jolly mongers of terror).

To clarify, I can do all of these things — if you believe what a couple budding researchers from MIT have to say. The students conducted a study on what kind of info you can infer by simply analyzing friend lists on sites such as Facebook. Their program, called “Project Gaydar,” is starting to get an awful lot of attention — and it isn’t hard to see why.

The students claim that, by looking only at a man’s Facebook friends, they can reliably predict whether or not he is gay. (And no, “looking at a man’s Facebook friends” is not a euphemism for “fondling his schwanz.” But good guess.) The two say they developed an algorithm that compiles the genders and sexualities of a person’s online friends, then uses that data to determine which way he prefers to get busy. The underlying principle: You share the same traits as the majority of your friends, so if most of them are gay, you probably are, too.

Now, to be fair, the study didn’t actually confirm the accuracy of most of its predictions; it did so only with 10 male subjects who the students knew to be homosexual (via “firsthand” knowledge — interpret as you will). It also failed to accurately predict lesbianism in a small sample of homosexual women, thereby eliminating any usefulness the program might have held for our purposes.

So basically, it’s all a bunch of meaningless shit. But who said meaningless shit can’t be fun? We decided to apply the study’s principles to Twitter to make some predictions of our own, analyzing three high-profile users to search for secrets. Our findings:

Ryan Seacrest1. @RyanSeacrest

Followers: Primarily teen and preteen girls who frequently tweet about being “boy crazy.”

Conclusion: Ryan Seacrest is a teen or preteen girl who frequently feels “boy crazy.”

Guy Kawasaki2. @GuyKawasaki

Followers: Lots of spambots, self-promotional marketers, and “adult dating” services selling “easy ways to find pleasure online.”

Conclusion: Guy Kawasaki is a robot who pleasures himself while spamming out useless drivel.

Ashton Kutcher3. @aplusk

Followers: Lots of vacant-looking accounts with no photos or information; guys with shirtless profile pics; and users with dull updates no one in their right mind would want to read. Also a smattering of UFO-obsessed oddballs with usernames like “UFOhunter” and “alienguide.”

Conclusion: Ashton Kutcher is an empty-headed pretty boy with nothing interesting to say. Either that, or he’s an alien (with nothing interesting to say).

You know, there just might be something to this system after all.

Get fresh geek humor delivered daily: RSS | E-Mail | Twitter