Consider a state with twelve districts: six 75-25 A-B, six 75-25 B-A. Votes are even, seats are even; the state isn't gerrymandered. So it isn't the case that

Consider a state with twelve districts: six 75-25 A-B, six 75-25 B-A. Votes are even, seats are even; the state isn't gerrymandered. So it isn't the case that a state where every district is split 75-25 is necessarily heavily gerrymandered. On the other hand, if all twelve districts are 75-25 A-B then it seems like B has three fewer seats than it deserves, which does suggest that that definition of an efficiency gap is missing something.