Guessing Gender Based on Web History

Interesting in concept...

Using your browser URL history to estimate gender

...though it pegged me with an 83% likelihood of being female based on this history list:

Site - Male/Female Ratio

google.com - 0.98
myspace.com - 0.74
youtube.com - 1
wikipedia.org - 1.08
amazon.com - 0.9
facebook.com - 0.83
blogger.com - 1.06
flickr.com - 1.15
mlb.com - 1.33
time.com - 1.44
wamu.com - 0.85
snopes.com - 0.74
americanexpress.com - 0.98
linkedin.com - 0.94
amtrak.com - 0.75
godaddy.com - 1.17
metropcs.com - 0.77
lulu.com - 0.96
spinner.com - 0.8
abc.com - 0.47