2018 Congressional Generic Ballot and House Polls Megathread - the original

[mencken]

Oh no, you and CBS News are concern-trolling hacks like LimoLiberal for even suggesting the House is anything but Likely Dem at this point.

[mencken]

Oh no, you and CBS News are concern-trolling hacks like LimoLiberal for even suggesting the House is anything but Likely Dem at this point.

Wrong. Democrats have around an 80% chance of taking the House at this point in time.

Poe's law everyone...

[mencken]

Yeah, Dems almost win an R+13 seat in AZ with a some lady and pick up R+11 PA-18 all in the last three months, but sure, the House is merely a tossup.

But those things happened months ago, you see. It's much smarter to rely on a Reuters poll that has whiplashed from R+7 to D+8 in a few weeks for literally no reason.

Polling averages have a history of being used to predict elections. Averaging all the special elections over the two years before the midterms lacks a comparable track record of out-of-sample performance.

[mencken]

Crickets from all the concern trolls and conservative pundits now, lol.

The new evidence does suggest movement toward the Dems. However, I thought GCB didn't matter, since all we have to do is average special elections and that's better than taking surveys?

[mencken]

Crickets from all the concern trolls and conservative pundits now, lol.

The new evidence does suggest movement toward the Dems. However, I thought GCB didn't matter, since all we have to do is average special elections and that's better than taking surveys?

Combine the two IMO

Yeah, special elections seem like reasonable indicators when you have a lot of them, and combined with the GCB + other more indirect factors like fundraising, recruitment, party that controls the WH, they are useful for indicating potential.

-

To mencken's post - Republicans were leaning heavily on the GCB to back their anti-Dem wave claims, so it stands to reason that once the GCB swings against them again, their narrative falls apart.

For us Democrats who have argued a wave is coming based on numerous factors, such as special elections, the GCB, and the stuff I mentioned above, this is just one more reason for us to stand by our predictions.

My issue with using the specials average is they are highly inaccurate if the environment changes significantly near the end. If the environment is consistent, though, they are usually a slightly better (but still flawed) indicator of the midterm results than the GCB.

And how do you know if the environment changes?