I think I've "cracked" Minnesota and found a formula that determines the winner

[I spent the winter writing songs about getting better]

Sort of. I learned of it playing around with some scenarios of areas needed to cancel out another area, and this seems pretty solid:

-Add together the totals and margins of the two strongest R congressional districts. Currently that's 6 and 7. Take the R margin of victory in vote totals from these together. That's Value X.
-Add together the totals and margins of Minneapolis and St. Paul. Just the cities themselves, not districts 4 and 5. Take the DFL vote margin of victory from this. That's Value Y.

If Y > X, then the DFL wins.
If X > Y, then the Republican wins.

And of course it's been a very long time since X > Y.

[President Punxsutawney Phil]

Sort of. I learned of it playing around with some scenarios of areas needed to cancel out another area, and this seems pretty solid:

-Add together the totals and margins of the two strongest R congressional districts. Currently that's 6 and 7. Take the R margin of victory in vote totals from these together. That's Value X.
-Add together the totals and margins of Minneapolis and St. Paul. Just the cities themselves, not districts 4 and 5. Take the DFL vote margin of victory from this. That's Value Y.

If Y > X, then the DFL wins.
If X > Y, then the Republican wins.

And of course it's been a very long time since X > Y.

How does this hold up in, say, the 2006 and 2010 gubernatorials?

[I spent the winter writing songs about getting better]

Sort of. I learned of it playing around with some scenarios of areas needed to cancel out another area, and this seems pretty solid:

-Add together the totals and margins of the two strongest R congressional districts. Currently that's 6 and 7. Take the R margin of victory in vote totals from these together. That's Value X.
-Add together the totals and margins of Minneapolis and St. Paul. Just the cities themselves, not districts 4 and 5. Take the DFL vote margin of victory from this. That's Value Y.

If Y > X, then the DFL wins.
If X > Y, then the Republican wins.

And of course it's been a very long time since X > Y.

How does this hold up in, say, the 2006 and 2010 gubernatorials?

It may not, but I think it's pretty accurate now.

The thing is this: The other Republican seats are the 1st and 8th. Not as Republican and have significant DFL enclaves unless the 6th and 7th. But if they're running a bit more Republican, and the DFL has to cancel them out with votes from the suburbs, considering that the suburbs are probably less D in this scenario and plenty of the suburban territory is already pretty Republican...you can see the path.

But even for Keith Ellison 2022, he got a wider vote margin out of Minneapolis/St. Paul than the deficit he had in the two most conservative districts, as close as it was.