However, rape (no matter what definition) is not evenly spread, some groups of women are more likely to become victims then others. Perpetrators pick their victims in a sort of rational way - some are more vulnerable.

So if you bias the group the right way, you can get large numbers.

No, it's actually commentary on the fact that women themselves often cannot clearly state "I was raped." The article in question was trying to show that lots more women have been raped than will admit to it or claim it. But you can also read it the other way: It casts light on how difficult it is to define rape.

Rape hinges on the detail of consent. Even women who have been fed alcohol until they were incapacitated and then taken to someone's room will go online and say things like "I had sex with a friend last night. I'm in a relationship. I know I need to drink less..." and get a chorus of people saying "Honey, that was rape."

People imagine rape to be some clear cut crime. In reality women are often very unclear in their own minds whether or not they were raped.

If the crime hinges on the detail of consent and the woman in question can't state clearly whether or not she wanted it, much less whether or not she consented, then you have bigger problems in determining data than in how surveyors ask their questions.

Sorry, no. It was an article I read a long time ago.

Rape stats are an inherently difficult thing to measure well. I could, in all earnestness, argue for either much higher or much lower figures than the reported figures.

That's a discussion I don't care to try to have on HN at all, much less on two hours sleep.