Your standard for saying A is more X than B is a bit strange. For me, all it means is roughly that when picking two individuals from A and B the one from A is m

Your standard for saying A is more X than B is a bit strange. For me, all it means is roughly that when picking two individuals from A and B the one from A is more likely to have greater X (assuming here the distributions are not weird). This is in itself useful information, at least in principle. To take the political correct example, let's say that I am a bissexual man. A priori, the probability that I will be attacked will be greater when dating a man. Is this useful? I don't know. That will depend how strong the difference is in the first place and how fast I can update my estimation based on new information. How can you be sure than men are intrinsically more violent than women anyway? Couldn't it just be because of the way men are raised (i.e. the patriarchy)?