Nate Silver

Nate Silver @NateSilver538

Objectively, it's not a slow news day, but people sure are worried about some largely irrelevant shit this morning. — PolitiTweet.org

Nate Silver @NateSilver538

3. Biden has gotten some strong state poll results over the past 48 hours, but our model still thinks the state polls are more consistent with say a 9 or 9.5 point Biden national lead rather than 10-11. — PolitiTweet.org

Nate Silver @NateSilver538

Slightly contradictory thoughts: 1. We're 15 days out from the 1st debate, which is getting to be outside the period where you could attribute Biden's gains to a short-lived debate bounce. 2. Still, a 10-11 point lead is a LOT & may be hard to sustain. https://t.co/cy51vc5isJ https://t.co/qtHxSM3DkT — PolitiTweet.org

Nate Silver @NateSilver538

RT @emilymbadger: Three months after the Postal Service began experiencing sudden delays this summer, first-class mail is still slow. An u… — PolitiTweet.org

Nate Silver @NateSilver538

RT @FiveThirtyEight: How conservative is Amy Coney Barrett? https://t.co/fcjpewXkys — PolitiTweet.org

Nate Silver @NateSilver538

Gallup has a net +53 on whether voters are more enthusiastic or less enthusiastic than usual to vote, which I believe is the largest spread ever recorded on that question. This is a pretty reliable predictor of turnout. https://t.co/oWwvzD3U3F https://t.co/Vle6rNvSty — PolitiTweet.org

Nate Silver @NateSilver538

In general, you would think this sort of pollster faces less reputational risk than a US-based firm—if their results are way off, they can just say "welp, the US is hard"—but the question is whether that frees them up to avoid herding or enables them to publish wacky stuff. — PolitiTweet.org

Nate Silver @NateSilver538

One thing I have completely no idea what to make of is the various British, Canadian and Latin American pollsters that have started to poll the US presidential race. A couple of them have quite strong results for Biden, although not all do. https://t.co/Ashqhf3WCe — PolitiTweet.org

Nate Silver @NateSilver538

@SenhorRaposa I think the big news organizations think about media narratives, e.g. MN and WI are relevant after "law and order", and they tend to think similarly to one another so you get some duplication. — PolitiTweet.org

Nate Silver @NateSilver538

@SenhorRaposa It's weird how we go through these little streaks. There was a three-day period with like 62,000 high-quality polls of Minnesota and now it's gone back to being one of the more underpolled states. — PolitiTweet.org

Nate Silver @NateSilver538

In some ways, though, higher turnout should make polling easier. In a hypothetical country where everyone voted, you wouldn't need to worry about a likely voter model. And indeed, polls show very little difference between registered and likely voters this year. — PolitiTweet.org

Nate Silver @NateSilver538

Our model assumes that turnout is a bit harder to predict because of the rise of mail voting, and that leads to some additional potential for polling error. — PolitiTweet.org

Nate Silver @NateSilver538

And with Biden's Arizona polling being a bit lackluster lately, NC has really moved into a three-way tie with FL and AZ for what is Biden's first line of defense if one of the Midwestern battlegrounds falls through. — PolitiTweet.org

Nate Silver @NateSilver538

North Carolina has been in toss-up range all year but has moved to Lean Biden after a series of polls showing Biden leads. https://t.co/YToi4OY2RV https://t.co/iav7THVLOc — PolitiTweet.org

Nate Silver @NateSilver538

@joshtpm There's still a chance that Trump can draw things closer with the 2nd debate. But if we go into Election Day with the current numbers and Trump won, it would either mean that polling was VERY broken (which our model does try to account for) or the election was (which it doesn't). — PolitiTweet.org

Nate Silver @NateSilver538

@joshtpm I think it's a sensible forecast. Overcoming an 7-8 point deficit in the tipping point states in 3 weeks is very hard though not impossible. And although there are some additional reasons for uncertainty this year, our model (tries to) account for (some) of those. — PolitiTweet.org

Nate Silver @NateSilver538

Something like long lines in GA might be typical on the first day of early voting. But that doesn't mean it couldn't have an impact, although in this case there is still plenty of time to vote. — PolitiTweet.org

Nate Silver @NateSilver538

Something like long lines at early voting locations in GA might be typical on the first day of early voting might be typical. But that doesn't mean it couldn't have an impact, although in this case there is still plenty of time to vote. — PolitiTweet.org

Nate Silver @NateSilver538

I find there's more noise than signal sometimes (sorry) in election administration stories when covered on Twitter. One thing that might help is if people think of a two dimensional matrix. Is something typical or atypical? And it is benign or impactful? — PolitiTweet.org

Nate Silver @NateSilver538

The conventional wisdom seems to be something roughly like "Biden's way ahead in polls and is indeed is a pretty clear favorite and a landslide is a distinct possibility, but stranger things have happened than Trump winning somehow" which I think is a fairly good read. — PolitiTweet.org

Nate Silver @NateSilver538

I can't figure out where the FiveThirtyEight forecast is relative to the conventional wisdom this year. It's probably fairly compatible, for a change. — PolitiTweet.org

Nate Silver @NateSilver538

RT @perrybaconjr: Democrats have an outside chance of winning Senate seats in traditionally-red Montana, Kansas and Alaska. We took a close… — PolitiTweet.org

Nate Silver @NateSilver538

In general, polling in Senate races tends to be pretty stable, and our forecasts reflect that. Not like a presidential race where voters are constantly consuming new information (although the presidential race has been fairly stable this year, too). — PolitiTweet.org

Nate Silver @NateSilver538

Our North Carolina Senate forecast has really been quite stable, with Cunningham having about a 2 in 3 chance of victory. Note that our model actually *does* take into account his scandal in our priors for the race, it's just outweighed by the polling. https://t.co/l14zlxUwrE https://t.co/xd2VTfN9Tt — PolitiTweet.org

Nate Silver @NateSilver538

As you can see in the article above, while there were changes in both directions, on balance the pre-pandemic trend in 2018-2019 was toward making it easier for people to vote. And post-pandemic, all swing states but TX now have no-excuse absentee voting. — PolitiTweet.org

Nate Silver @NateSilver538

It's not clear to me that there are extra GOP efforts this year to make it harder for people to vote, so much as that this has *always* been happening and it's getting a lot of (overdue IMO) attention this year. https://t.co/x3pDlxpQka — PolitiTweet.org

Nate Silver @NateSilver538

On the one hand, I'm still not convinced that state polls are consistent with the 10-11 point Biden lead shown by national polls. On the other hand, this is still a very robust position for Biden in state polls. — PolitiTweet.org

Nate Silver @NateSilver538

NH: Biden +10.9 MN: Biden +8.9 MI: Biden +8.0 WI: Biden +7.7 NE-2: Biden +7.5 PA: Biden +7.3 <-- TIPPING POINT NV: Biden: +6.9 FL: Biden +4.6 AZ: Biden +3.8 NC: Biden +3.4 ME-2: Biden +1.3 IA: Biden +1.1 GA: Biden +0.6 OH: Biden +0.2 TX: Trump +1.5 https://t.co/wKbW02AGO5 — PolitiTweet.org

Nate Silver @NateSilver538

RT @geoffreyvs: We took a poll of pollsters (unscientifically) to see what changes they’ve made since 2016 and what they’re worried about i… — PolitiTweet.org

Nate Silver @NateSilver538

Maybe there'd be a needle to thread for a skilled politician. I'm not sure high support for lockdowns in polls squares with people's behavior/revealed preferences. But attacking Dr. Fauci, or saying you're immune and want to kiss everyone, isn't it. https://t.co/JGVREvQXqT — PolitiTweet.org