It's crazy because all the overt caution messages we get, or at least that I got, involved stranger danger. It's crazy because making connections with people is supposed to afford some protection. But even if we never go out with men we don't know, how likely is it that we can actually get to know all of the men around us well enough to have a clue if they are the kind of men who turn out to be rapists, before we let ourselves be alone in a room with them?
The comments on the SR article talked a bit about this stat, that most rapists are known to their victim. They raised the very good point that it doesn't say how well known. Rapists are going to run the gamut, from close family members to first dates to someone whose name you know but haven't ever talked to. (To total strangers.) Not terribly comforting.
ETA: It's also crazy because our good Buffista guys would never dream of doing that sort of thing, but to a certain extent, it is logical for the women around them (who don't know them as well as we do) to put them in the Schoedinger's Rapist category.
This part I understand, and I think it's quite reasonable. It's not a personal call on anything internal to the guy in the Schrodinger position. It's just that at the outset, there's nothing to distinguish threats from not-so-much threats. It's the woman's decision what cues and clues will make her feel comfortable, and that article was very well done in providing a framework for a guy to think about that.