Comments on The Internet Predicts Your Gender

At work, I'm 76% female and 24% male, which I think is accurate enough. I thought it would be higher, considering how much time I spend at work looking up football news, but it doesn't seem to get into specifics about what Guardian site I've been on.

You'll always be 50/50 in my book, Paul.

More interesting than the overall female / male rating is the ratio reported for specific URLs. I can understand, for example, why scienceblogs.com might have a ratio of 1.44 M/F but a 0.79 M/F ratio for irs.gov is a surprise. Among other feminine sites are: apple.com (0.89) and my bank (0.87). NHL.com has a seemingly low M/F of 1.35.

Huh. I'm surprised my work browser history has me at 93% female, given how much time I spend on news sites, USS Mariner, istockphoto, etc. I guess, um, eonline.com and bravotv.com sort of sank me. The shame!

Likelihood of you being FEMALE is 93%
Likelihood of you being MALE is 7%

Not bad...

What I find odd is how it pulled up sites in my browser history that I think I've visited only a few times in the past two months, like linkedin or Perez Hilton (or the King5 website, which I visit so rarely that I can actually remember the distinct reason for going there), while not touching the sites I spend the most time on, like the AV Club or NewsNow: Arsenal or something. Odd.

I'm 97% female
3% male

And now my testicles are sad.

Abby: It's only selecting sites from this Quantcast Top 10K sites thing, I guess, which I now realize is why something like USS Mariner clearly wouldn't be on there.

Ahhh. I wonder what it would look like if it actually analyzed my browser history? Most of the sites I visit probably aren't on there. Would I be more female or less?

It does seem to miss out on things, maybe based on URL discrepancies. avclub.com, for instance, is in the top 10k sites, as is nwsource.com (the base domain for The Seattle Times, the P-I, and all their classified sites), which I go to a gajillion times a day but didn't pop up in my history.

I'm dumb. Never mind, nwsource.com is in there (skews male, too).

thestranger.com, however, is not. slog.thestranger.com must not register as part of the larger site. (thestranger.com is in the top 5k sites!)

Yeah, its limited site recognition can skew things.

It put me at 54% male (I'm a chick). Mostly because I was on mininova.org today; otherwise it would have had me at 50/50.

Which you'd find amusing if you knew me.

My results from work:
Likelihood of you being FEMALE is 38%
Likelihood of you being MALE is 62%

...but I'm female. Maybe it'll be different from home. It was surprising to learn that icanhascheezburger has a higher ratio of men visiting it than google.

Likelihood of you being FEMALE is 54%
Likelihood of you being MALE is 46%

Too close to call? (Mind, I was female the last time I looked.)

Site Male-Female Ratio
facebook.com 0.83
flickr.com 1.15
cafepress.com 0.8
icanhascheezburger.com 1.04
gutenberg.org 1.06

I'm a girl. I'm a hetero girl. I'm totally a girlie girl and I'll kick you in the ass with my 5 inch stilettos if you disagree. And yet:

Likelihood of you being FEMALE is 27%
Likelihood of you being MALE is 73%

Look, just because I go to Game FAQs and Urban Dictionary and the fucking Stranger doesn't mean I'm a guy. It means I cheat at video games and don't understand half the slang I read on Slog, which I have a compulsive need to read and post to. Stupid? Yes. A male? NO!

UPS skews male, but USPS skews female. That's amuses me.

Good thing I cleared my browser history just now.

I feel like I should win something for this...

Likelihood of you being FEMALE is 100%
Likelihood of you being MALE is 0%

i thought i was a female. too much shoe browsing i guess.

should be "it thought i ..." oops. well, that explains part of the problem. my typos...

But it doesn't calculate how much total time you spend on the sites....I spend an inordinate amount of time at Facebook (0.83) but only long enough on the Michael's website (0.47) to find out when they close. Also, I have wasted most of the last two days at Flickr (1.15) but that doesn't seem to skew my results at all.

Whatever. This at least makes more sense than that one site that could guess whether you were male or female based on your blog entries.

I'm 53% female? I guess it missed all that gay porn in my browser history.