There’s some weird stuff here.
Biden’s doing 4.1% better in Ohio but 1.5% worse in PA?
2.8% better in Wisconsin but 1% worse in Michigan?
Michigan and PA are correlated, as I’d expect. He’s doing worse in both - but he should be doing worse in Ohio too, then. And Wisconsin is whiter than PA and Michigan, but her he’s doing better there.
That polling is unsettling to me. On the other hand, Biden is doing 2.3% better in GA and 2.7% better in FL. That makes sense for them to be correlated, but I think they’re fools gold due to cheating and riggage.
Biden should open up an offensive in NC I think. Iowa is tempting, but it’s got a GOP governor and Secretary of State. One of the problems Biden has is no margin for error in a close election. If the only states in play from 2016 are PA, MI, WI, AZ, Biden must win PA and any two of the other three.
This is probably why his campaign has ignored Hispanic voters. Electorally speaking, Arizona doesn’t take Pennsylvania out of must-win territory. North Carolina does, though.
North Carolina: 69% white, 21% black, other groups <4%
Pennsylvania: 82% white, 11% black, other groups < 4%
Wisconsin: 86% white, 6% black, other groups < 3%
Michigan: 79% white, 14% black, rest < 2.5%
If the racial divides are magnified, counting on PA as a must win is even riskier… North Carolina could make sense as an insurance policy if the election shifts even more along racial lines. Trump could bring it more into play for Biden given the demographics there.
If you can put Trump in a spot where every step he takes toward aggrieved white voters in PA/WI decreases his odds in NC by increasing black turnout, you put him in a bad spot.
I have a feeling this is also why Trump is trying to get NV into play. Biden is strategically ignoring Hispanics because Arizona doesn’t have enough electoral votes to come into play at the tipping point in most scenarios. But if ignoring them lets Trump snag Nevada, that opens more pathways back up.