Sorry. This was supposed to be a response to one of @BamaATL 's posts.
I've stated this in the presidential race thread, but it bears remembering here too: the low response rates of current polling make the margin of error HUGE. Anticipate a "polling error" * this year.
Problem is, we won't know direction of error, but the direction is likely correlated amongst states with similar demographics. The magnitude of the error is likely to vary considerably amongst states.
Battleground states OUTSIDE of polling error (538):
AZ --> D + 7, Gallego > 50%
WI --> D + 5, Baldwin > 50%
NV --> D + 9, Rosen 49%
PA --> D + 5, Casey 48%
MI --> D + 4, Slotkin 48%
MT --> R + 6, Sheehy 49%
Battleground states WITHIN polling error:
TX --> R + 3, 47 - 44
NE --> R + 3, 43 - 40
OH --> D + 3, 47 - 44
FL --> R + 4, 46 - 42
Dems need to hold Ohio, which is a pretty good bet unless Trump over-performs and has a coattail effect. If this happens, we're screwed anyway.
Sadly Tester's probably toast. He's a great dude, but probably the last Dem from MT for many years. I'd trade MT for FL or TX. I'd prefer FL because I think Mucarsal-Powell is more progressive than Allred. NE seems like a hard hold also, but if that's what we get, we gladly take it to get to 50 - 50.
* It isn't really a polling "error." Polls are predictive devices. Predictions have assumed error. The potential for error is increased with greater deviation from ideal samples. Samples so small that they can't reflect the voting population have greater potential for error.