For it to be zero-sum it requires a shift in responsibility from someone onto someone else. The cases are independent of one another and do not happen simultaneously or co-happen. They are the two very different results of what people studied think.

Case A: The woman drinks more. She is held to be more responsible for being raped than a sober woman would. The man is unaffected in terms of responsibility.

Case B: The man drinks more. He is held to be less responsible for the act than a sober man would. The woman is unaffected in terms of responsibility.

Should it be a zero-sum game, there needs to be a link from the man's total responsibility to the woman's total responsibility. There is not.

Case Z: The woman is drunk. The woman is held to be more responsible, ergo the man is less responsible.

Case Z would be a zero-sum example. The others are not.

Responsibility is pretty damn separate to criminal culpability if only in matters of discreteness - responsibility can be in terms of degree but you are not partially culpable in the eyes of the law, you are guilty or not guilty. We are not merely dancing around synonyms if there is a clear difference in properties like that.

My 'final conjecture' is a complaint of how ridiculous the number game works out to.