Yeah, and at this point I'm absolutely fine with banishing republicans into the wilderness until they learn to stop being cunts. They can't be trusted with power anymore, fucking ruin their shit, set it on fire, kick their asses out the door and chain it shut.
Hence the hashtag I posted earlier, #GOPDelendaEst. It's gaining in popularity as people lock onto the sentiment that there should be no bipartisanship for the sake of civility, no olive branches, no "divided government works best". No "maybe now they'll do some soul-searching and reform their party" -- they've had several opportunities for that over the last 30 years and every time they chose to double-down on the crazy. The GOP needs to be destroyed and the earth it once stood on salted. Let the Trumpers continue on with a rump social conservative/nativist party, and let the sane fiscal conservatives build a new party.
Anywho, it's time for the weekly number-crunching.
I'm having something of a crisis on which numbers to use.
RCP is distinctly more negative on Biden this week, but they're also omitting a lot of polls from their average, while including R-friendly contrarian pollsters like Trafalgar Group, which skews the average (up to 3 points in some cases). They also give equal weight to all polls, regardless of age or past accuracy of the pollster. This makes it easy to game them (which is honestly what I think Trafalgar Group is being paid to do at this point).
538 includes a lot more polls, and weights them on a number of factors, including polling method, sample size, margin of error, age, and pollster rating. They also have poll data for ALL states, not just the close ones. But those states are almost entirely based on SurveyMonkey, which is only marginally more accurate than a REAL monkey. They're accounting for that in their aggregate by dropping Biden's support about 5 points from what the SM polls show, but still...
My concern is that by using their numbers, I'm essentially incorporating noted warlock Nate Silver's special sauce without fully understanding it.
So I've been tracking both sets of numbers since last week, and instead of listing raw data I'll just post the data where it's relevant.
On the whole, we're seeing what I would expect to see at this point, and probably the remaining weeks of the election -- a reversion to the mean in most swing states, and a further strengthening in many of the bastion states. In layman's terms, close races are getting closer, races that aren't even close are getting further apart.
That said, even with RCP's decidely unfriendly numbers, they still show Biden with a 357-198 win, as they have Georgia, Florida, Iowa and North Carolina all in the D camp. Of the true swing states, only Ohio is shown going to Trump, and that one is R+0.5, which is well within MOE. Of slightly more concern is the weakening of Dem advantage in a number of battleground states.
Florida down from 3.7 to 1.4 (although this is heavily skewed by one of those Trafalgar polls)
Michigan down from 7 to 6.8 (another Trafalgar poll)
Minnesota down from 9 to 6.6
Nevada down from 6 to 5.2
Ohio down from D+0.6 to R+0.5
Pennsylvania down from 7.1 to 4.4 (yet another Trafalgar poll)
Wisconsin down from 6.1 to 6
If we look at the 538 numbers (and their change from last week) for the same states:
FL -- down from 4.5 to 3.9
MI - down from 8.1 to 7.9
MN - down from 9.3 to 9.1
NV - down from 6.9 to 6.3
OH - down from D0.7 to R0.4
PA - down from 7.2 to 6.7
WI - down from 7.7 to 7.3
So the poll aggregates agree on the general tightening of these races (and the Ohio flip), and in some cases agree on the magnitude of change (MI, OH) but in other cases (FL, MN, NV, PA) disagree by up to a full 2 percent on the magnitude.
On the plus side for the Democrats, both aggregates show North Carolina and Georgia moving in their direction, with RCP flipping Georgia from R to D while 538 already had it in the D column.
One of the more interesting bits with the 538 numbers is that Nebraska suddenly becomes a potential battleground, with Trump at +6.3, down from +11 a week ago. Whether this has anything to do with Trump's weekend pissing match with Nebraska senator Ben Sasse is hard to say at this point. Kansas is also at +6.5, enough to raise an eyebrow at, but inconsequential unless it continues to narrow and quickly.
Since we were just talking about it, Texas remains largely unchanged. RCP has it at R4.4, unchanged from last week. 538 has it at R1.3, a drop of 0.2 from last week. Given Sen. John Cornyn's recent remarks about the GOP's relation with Trump being like "a woman who marries a guy thinking she can change him", it'll be interesting to see what impact that has on both the statewide Presidential race *and* Cornyn's own race. But given that over 4 million Texans have already voted, Cornyn may have made the shrewd calculus that enough of his supporters have voted that any Trump-driven outrage would have minimal effect. OTOH, recent polls have shown MJ Hegar as close as one point away from Cornyn, so he may be doing this to try and steal a few conservative Dem votes for himself. Given that his opponent is a decorated Air Force major, I have strong doubts that'll work.