Surprised no one's mentioned the shortest splitline aglorithm or maximizing minority influence. The first naturally cracks populated, D-leaning areas while the second would serve to maximally pack D-leaning non-White voters.
In both cases though those redistricting procedures can benefit Democrats depending on the situation--they are very high-variance.
Well, I mean sure. But it's not like a strict efficiency gap/proportionality criterion benefits Democrats
everywhere either (i.e., California.)
The name of the game is trading egregious pro-Republican gerrymanders in some states for some nationwide criteria that make the median map better for the GOP than it is now.
The
538 Atlas of Redistricting has Republicans at 180 safe seats under a national shortest splitline algorithm standard, 184 under a standard that maximizes the number of minority seats.