Nate Silver

Nate Silver @NateSilver538

Kind of shame that we don't seem to be getting many live-caller polls of Texas in the closing days here. — PolitiTweet.org

Nate Silver @NateSilver538

We've run our presidential forecast 38 times since yesterday morning and at no point have Biden's odds been lower than 88.8 percent or higher than 89.8 percent. Trump can win but polls that people are getting very excited about/mad at/etc. aren't really changing the picture much. — PolitiTweet.org

Nate Silver @NateSilver538

RT @galendruke: 🎧 New pod is up 🎧 We started off with some light SNL viewing... Then discussed the latest polls... And finally, clarifie… — PolitiTweet.org

Nate Silver @NateSilver538

@DKThomp @LPDonovan None of this is an especially coherent plan. — PolitiTweet.org

Nate Silver @NateSilver538

@Taniel Yeah, the states with competitive Senate races that aren't presidential swing states have gotten very ignored down the stretch run here. — PolitiTweet.org

Nate Silver @NateSilver538

In the next tier of states down, such as GA, TX, IA or NV, further polling could be more influential. However, to the extent there's polling outstanding, most of it seems to be coming from the Big 6 states. — PolitiTweet.org

Nate Silver @NateSilver538

This is why it'll be hard for further polling to move our numbers all that much in the Big 6 states (PA/WI/MI/FL/NC/AZ). One new fairly decent poll of PA will still get only 5% of the total weight there, or thereabouts. A very good poll with a large-ish sample might get 10%. — PolitiTweet.org

Nate Silver @NateSilver538

These are the states sorted by the total amount of recent polling. A very recent poll of ~600 people and slightly-above-average quality gets an influence weight of 1. So in Florida, for example, we have the equivalent of 24 recent polls. https://t.co/IpsTCrCdo4 — PolitiTweet.org

Nate Silver @NateSilver538

Tired of people framing "Trump will declare victory prematurely" as a media story. Most likely—in part because the White House has telegraphed it so much in advance—the media will be fairly well-prepared. Instead, it's a story about Trump seeking to undermine faith in democracy. — PolitiTweet.org

Nate Silver @NateSilver538

This is basically what you get if you only use high-quality polls instead of (as 538 does) using all polls but weighting high quality polls more heavily. About 1 point better for Biden on average. — PolitiTweet.org

Nate Silver @NateSilver538

@johnhollinger Or NC for sure. I think it's become a pretty underrated state. — PolitiTweet.org

Nate Silver @NateSilver538

I actually think you can make a case for GA as a plausible tipping point state if something goes wrong for Biden in PA. With OH it's much tougher because OH and PA polling errors are likely to be highly correlated. — PolitiTweet.org

Nate Silver @NateSilver538

For better or worse, a lot of Biden travel seems oriented around trying to have a clear and unambiguous win rather than maximizing the chances of 270 EV per se. — PolitiTweet.org

Nate Silver @NateSilver538

RT @MorningConsult: @NateSilver538 We will be releasing new state-level polls... stay tuned! — PolitiTweet.org

Nate Silver @NateSilver538

RT @laurabronner: Four reasons Biden has a better shot than Clinton did in 2016 -- and 2 reasons there's still uncertainty. A summary 🧵:… — PolitiTweet.org

Nate Silver @NateSilver538

I think Trump can win legitimately! (A 10% chance, per our model.) I am not sure if the Trump campaign does, though. — PolitiTweet.org

Nate Silver @NateSilver538

Among all the other problems with this, this is not something a campaign says if it thinks it has much chance of winning legitimately. — PolitiTweet.org

Nate Silver @NateSilver538

Well, it looks like from YouGov/CBS, we got modeled *forecasts* rather than new polling per se. It's a cool product but we can't use these in our own forecasts as we are averaging polls and not averaging other people's forecasts. https://t.co/MOJbO1nPkN — PolitiTweet.org

Nate Silver @NateSilver538

* Morning Consult has mostly been keeping its state polls paywalled, don't know if we'll get something public. * Maybe one more round of Ipsos state polls? Of course we'll get weird/random stuff too, a lot of which will probably herd, but this is ~the list of polls I care about. — PolitiTweet.org

Nate Silver @NateSilver538

* NBC/Marist hasn't released a PA poll this round so that may be coming. * Presumably Quinnipiac has something up its sleeve? * A handful of one-off state polls from local universities/newspapers may weigh in again, but they can be budget-constrained and timing hard to predict. — PolitiTweet.org

Nate Silver @NateSilver538

What's still out there, polling-wise? * Presumably we'll get some YouGov/CBS in 10 minutes here. * Emerson's releasing polling all day. * Monmouth in PA. * Rumors of another round of Fox News state polls, though I'm not 100% sure about those. — PolitiTweet.org

Nate Silver @NateSilver538

To the extent there's been tightening in national polls, it's been from Trump clawing away at undecideds. Biden is essentially unchanged from his peak at ~52, but Trump has improved from a nadir of 41.7 to 43.5 now. https://t.co/3bJ4MsGDJ4 — PolitiTweet.org

Nate Silver @NateSilver538

It can be a somewhat fine line, although at one end of it, there's a range of behavior that definitely isn't OK (e.g. now and then you learn later that a pollster declined to publish a number that it thought was an outlier). But in any event, the final polls can be anticlimactic. — PolitiTweet.org

Nate Silver @NateSilver538

Pollsters often take precautions in their final polls to protect vs. outliers. That can range from increasing the sample size (100% kosher) to adopting methodological changes they think will result in less variance or bring them more in line with other polls (maybe OK/maybe not). — PolitiTweet.org

Nate Silver @NateSilver538

Pretty much every poll this morning coming in close to what you'd expect from the respective polling firm in the respective state. The final polls are usually not a time when you get big surprises. https://t.co/9AeTKX5o0O — PolitiTweet.org

Nate Silver @NateSilver538

RT @jodyavirgan: This is exactly the right framing. Important to think of polling, and the work that Ariel and 538 and others do, as mappin… — PolitiTweet.org

Nate Silver @NateSilver538

What I'm getting at here is ... things look pretty locked in. Not the election outcome itself! That ain't locked in at all! But our final averages may not change much. Especially after adding these polls because they have huge sample sizes and so get a lot of weight in our model. — PolitiTweet.org

Nate Silver @NateSilver538

Oops, link to their polls. https://t.co/8mhHglfPIs — PolitiTweet.org

Nate Silver @NateSilver538

How the NYT/Siena polls affected our averages: AZ: They have Biden +6. Our average goes from +3.2 to +3.4 FL: They have Biden +3. Our average goes from +1.7 to +1.8. PA: They have Biden +6. Our average stays at +4.9. WI: They have Biden +11. Our average goes from +8.2 to +8.7. — PolitiTweet.org

Nate Silver @NateSilver538

Biden +4.9 in PA after adding that poll (was +4.7 before) and +1.7 in FL (was +2.0). https://t.co/nWUo9X5F4y https://t.co/H2rYol0NO3 — PolitiTweet.org