Nate Cohn

Nate Cohn @Nate_Cohn

@RachelBitecofer @cityofprogress @TravisKniffin @Redistrict loll this is charlatan level work your putting out here, and you're trying to paper it over by attacking other people like you're on a reality tv show — PolitiTweet.org

Nate Cohn @Nate_Cohn

@RachelBitecofer @cityofprogress @TravisKniffin @Redistrict loll this is charlatan level work your putting out here, and you're trying to paper it over by attacking other people you're on a reality tv show — PolitiTweet.org

Nate Cohn @Nate_Cohn

@cityofprogress @TravisKniffin @RachelBitecofer @Redistrict she classified 42 seats *would flip*, and then said there were another 20 tossups and 20 lean rep seats or whatever. she now says that meant she predicted dems to gain 42, which is most certainly not what it means — PolitiTweet.org

Nate Cohn @Nate_Cohn

@TravisKniffin @RachelBitecofer @Redistrict i don't do forecasts, but she predicted dems to flip 55-60 seats, so it wasn't particularly accurate — PolitiTweet.org

Nate Cohn @Nate_Cohn

@RachelBitecofer @Redistrict ...it's far more important to know when you shouldn't have one than to have one — PolitiTweet.org

Nate Cohn @Nate_Cohn

@RBidenism @ElectionMapsCo yeah that’s not weighted by education — PolitiTweet.org

Nate Cohn @Nate_Cohn

Monmouth gives Biden a narrow lead among LVs in Arizona. https://t.co/ULOjh2sqBD — PolitiTweet.org

Nate Cohn @Nate_Cohn

We're not going to have that issue this year. A lot of new, high quality firms have entered the state polling fray, and others have upped their game since '16. I'm not saying that guarantees an accurate average, but it certainly improves the odds — PolitiTweet.org

Nate Cohn @Nate_Cohn

Or put differently, if there had been 3 MULaw polls at thend--I'm not sure they all would have been that good for Clinton given the national trend and their other two polls. If there were more high quality polls there, I really do think we might have known something was up — PolitiTweet.org

Nate Cohn @Nate_Cohn

In the end, MULaw was Clinton+6. And that was the last and only word on it. But what people might forget is that MULaw had previously shown Clinton up just 4 and 2 in their prior two surveys. After the fact, it makes me wonder whether the final poll was a little hot for Clinton — PolitiTweet.org

Nate Cohn @Nate_Cohn

Speaking personally, my interpretation of the end of the race really hinged on the final MULaw poll, which you can kind of see from this 11/2 article. There just weren't really many other high quality polls in the Midwest from that point onward. https://t.co/0diohu0woA — PolitiTweet.org

Nate Cohn @Nate_Cohn

One thing people forget about Wisconsin's huge polling error in 2016 is that there was basically no high-quality polling in town. It was basically MULaw and MULaw alone. As a result, so much of our interpretation of the race hinged on their final poll. — PolitiTweet.org

Nate Cohn @Nate_Cohn

If there's any single competitive district you could look at on Election Night to know whether Texas went blue, this might be the pick https://t.co/hF14G2HcsE — PolitiTweet.org

Nate Cohn @Nate_Cohn

@CampaignDiaries Well if everyone was Biden+5 that might be a problem. But it's interesting if the poll results don't appear to be normally distributed around Biden+5, with say a cluster there and outliers either way. Here it's kinda just spread evenly from 0 to 10 — PolitiTweet.org

Nate Cohn @Nate_Cohn

One state where we don't have much consensus among pollsters is AZ, where we have everything ranging from Biden+9/10 from three firms to Biden+3 in YouGov and +1 in AAPR/Benensen/GSG. KFF/Cook landed right in the middle today, at 5. We'll see where Monmouth lands in a few hours. — PolitiTweet.org

Nate Cohn @Nate_Cohn

As an aside, the MULaw and NYT/Siena polls in WI were both R+2 by party identification, and Biden still led by 4/5 points among LVs, and more among LVs. It's just not as easy to judge these polls by party ID or past vote or whatever as you might think! https://t.co/Ks2bLgToaF — PolitiTweet.org

Nate Cohn @Nate_Cohn

Yeah, doesn't seem like it's weighted by education. In NC, there's a truly mammoth white education gap--so that's far too much to overlook. — PolitiTweet.org

Nate Cohn @Nate_Cohn

Suffolk University has Biden+3 in NC and Cunningham+4 in the Senate. In the past, this poll hasn't weighted by education--and that's a huge caveat in NC. OTOH, they've occasionally had some of Trump's best national numbers. https://t.co/HtIceOuDNp — PolitiTweet.org

Nate Cohn @Nate_Cohn

As someone not reaching these voters, I certainly appreciate someone go through the effort. That's particularly true when they have similar results, as it may help add a little more confidence to the other data https://t.co/TEm2Mh72wm — PolitiTweet.org

Nate Cohn @Nate_Cohn

The mail-to-web element is to try and compensate for one of the risks in many state polls, including our own: not everyone on the voter file has a telephone on file. So they reach those voters by mail and ask them to take the poll online. — PolitiTweet.org

Nate Cohn @Nate_Cohn

New KFF/Cook surveys: Arizona: Biden+5 Florida: Biden+1 North Carolina: Biden+2 And a nice thing about these polls is that they're methodologically unique, with a novel live phone+mail-to-web methodology for full coverage RBS https://t.co/TraM7I8uCk — PolitiTweet.org

Nate Cohn @Nate_Cohn

Another day of strong state polling for Biden, including some eye-popping margins in MN and ME. I think we know the MN one is outlierish, and we'll have a second opinion on ME from NYT/Siena in a few days https://t.co/juAMFFB8i8 — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @MonmouthPoll: COMING TOMORROW: Arizona #GeneralElection for Prez & Senate. Our March poll had @JoeBiden over @realDonaldTrump by 3 pts.… — PolitiTweet.org

Nate Cohn @Nate_Cohn

If the next wave of polls, ideally with some national confirmation for stronger trendlines, shows something more like Biden+8-9, I think we could easily point to the last 48 hrs as when the moment the convention bounce seemed to fade out. But we're not there yet imo — PolitiTweet.org

Nate Cohn @Nate_Cohn

This is a serious question, though the national data--of admittedly quite uncertain quality--does run in the other direction https://t.co/MTTCQKnT1D — PolitiTweet.org

Nate Cohn @Nate_Cohn

@B_M_Finnigan 💅 — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @MattGrossmann: For last 2 elections, polls underestimated Republican strength in Red midwestern states & overestimated it in CA. If th… — PolitiTweet.org

Nate Cohn @Nate_Cohn

Forget about Quinnipiac, for a sec. Just logically: shouldn't this state be ground zero for Biden, given what the national polls tell us about his strengths? https://t.co/11gUQxGmZy — PolitiTweet.org

Nate Cohn @Nate_Cohn

I do think the case for a 'snapback' in Maine is pretty straightforward. It is basically the oldest, whitest state in the country. It is a state Obama won by *15* in 2012. With Biden doing way better among older and white voters, why shouldn't this be a state where he surges? — PolitiTweet.org

Nate Cohn @Nate_Cohn

Obviously the firm has had a bit of a left lean this year, but you'd might need to take 10 points off of these numbers before you could call them decent for he GOP — PolitiTweet.org