I don’t see the Court accepting the efficiency gap as an actual objective measure of much of anything. It’s interesting but has it’s own problems. If you look at the data that Katz, Grofman, et al used as well as what the Brennan Center used and follow their methodology, some flaws pop out.
To begin with, they ignore 3rd party votes and “no votes”. That means (effectively) that they default to a presumption that 3rd parties don’t count and don’t deserve representation.
They also apparently don’t account for races where a major party candidate (either D or R) ran unopposed. Those races skew the numbers significantly because those districts will always have an efficiency gap of 49.999%. When you start calculating your state-wide averages, those 49.999% numbers skew the calculations.
Just as an example, for the 9 House races here in MA for 2016, 4 candidates ran unopposed. A 5th ran against 3rd party candidates only. (all 5 of those happened to have been Democrats.) So of the 9 races, only 4 ran as a Democrat facing a Republican.
I ran the numbers using the same methodology the Brennan Center used and got the same results — a 2 seat skew for Democrats in MA.
But if I run the same process and only look at the 4 races that had candidates from both major parties, the skew disappears entirely.
Another problem with the uncontested races is the “I’ll vote for him/her because they are the only person running!” factor. If you look at the 4 races in MA where a Democrat ran against a Republican, the Democrat won by roughly a 70%-30% vote split in each race. But when you look at the districts where they ran unopposed, the winning candidate took a higher percentage of vote over the “no vote” ballots. Those came up roughly 85%-15%. That would seem to indicate that some people who would have voted for a Republican if one had been on the ballot, voted for the Democrat because they were the only name listed. Neither Katz/Gorfman, et al nor the Brennan Center account for that (and I don’t know of any way for anyone to do so.)
In MA for 2016, 3rd Party candidates and “no vote” ballots accounted for 17% of all ballots cast in House races. I don’t see how the Supreme Court can logically accept a method for determining partisan gerrymandering that purposely excludes almost 20% of the actual voters.