Nate Silver

Nate Silver @NateSilver538

These take a bit of work to calculate, because there are some adjustments for uncontested districts and states with unusual voting systems. But we use these in our Congressional forecasts; they provide a nice complement to partisanship scores based on presidential voting. — PolitiTweet.org

Nate Silver @NateSilver538

Here is 538's State Legislature Partisan Lean Index (SLPLI) for 2020, which is based on voting for the lower chamber of a state's legislature over the past four state legislative election cycles (generally 2012-2018). https://t.co/x7WJjOvjpJ — PolitiTweet.org

Nate Silver @NateSilver538

@conorsen @mattyglesias I'd bet on there being some pretty rapid improvements in suburban restaurant quality over the next several years, including in categories that are traditionally associated with central cities, although that's somewhat independent from COVID. — PolitiTweet.org

Nate Silver @NateSilver538

@mattyglesias @conorsen I'm agreeing with you (and slightly disagreeing with Conor)! I'm just emphasizing that people's senses may be a little mis-calibrated right now because suburban amenities (especially *space*, outdoors and otherwise) mostly hold up OK under COVID whereas urban ones don't. — PolitiTweet.org

Nate Silver @NateSilver538

@conorsen @mattyglesias I think there's something a bit more subtle going on though which is the amenities that differentiate from big central cities like NYC from the suburbs are exactly the amenities you can't get (or are costly/risky) under COVID. — PolitiTweet.org

Nate Silver @NateSilver538

@BenjySarlin For sure, college students are gonna be socially promiscuous. Especially in "back-to-school" week. So it's a big ask. But at least it's an ask! A more realistic ask than "stay 6 feet away from everyone all semester!" — PolitiTweet.org

Nate Silver @NateSilver538

I wonder if a message of "if you're going to hang out, hang out in smaller groups, outdoors or in well-ventilated spaces, and not in crowds or large parties" would be more effective than just telling students to "socially distance" & hoping for the best. https://t.co/RcaY886F9z — PolitiTweet.org

Nate Silver @NateSilver538

People cite Trump as a counterexample. But he arguably proves the point. The GOP lost the House popular vote in the midterms by 9 points. And now they're on track (obviously polls can change) to lose both the House and presidential popular votes by 8-9 points again. — PolitiTweet.org

Nate Silver @NateSilver538

* Actions that excite your own party's base often also rile up the other party's base. So even if the net impact on swing voters was neutral (which it often *isn't*) these actions won't necessarily result in a net increase in turnout. — PolitiTweet.org

Nate Silver @NateSilver538

Yeah, this. Also important to keep a couple of related points in mind: * Switching a vote from D to R or R to D has twice the impact of turning someone out who wouldn't otherwise. i.e. instead of going from 0 to +1, you're going from -1 to +1 for a net gain of 2. — PolitiTweet.org

Nate Silver @NateSilver538

RT @Herring_NBA: My latest for @FiveThirtyEight: Carmelo Anthony's On-Court Cussing Has TV Networks Playing Defense https://t.co/KtbzyWyPrl — PolitiTweet.org

Nate Silver @NateSilver538

But on balance, there is likely to be more scrutiny on ensuring mail votes are counted than there might have been otherwise, from e.g. Congress, the courts, the media, post office employees, etc. And—importantly—from voters (especially Dem. voters) who may now be more vigilant. — PolitiTweet.org

Nate Silver @NateSilver538

Does that mean, with today's news, there's no reason for concern? I wouldn't say that. I think there's too little concern paid to vote suppression in general. And mail ballots can have high-ish rejection rates, even if everyone is operating in good faith. https://t.co/AoqUGYpFnU — PolitiTweet.org

Nate Silver @NateSilver538

The USPS is one of those things where if Trump were a more cunning politician, he might have maintained plausible deniability and gotten away with causing real problems for Democrats around the margin. Instead, he gave away the plot last week! https://t.co/M3uSlUPG24 — PolitiTweet.org

Nate Silver @NateSilver538

RT @daveweigel: The key to convention take-writing is combining stuff like "Do conventions matter anymore?" with "could extending AOC's spe… — PolitiTweet.org

Nate Silver @NateSilver538

Game theory optimal strategy for dealing with trolls (although 100% ignore is probably the better approach on this platform these days) — PolitiTweet.org

Nate Silver @NateSilver538

Even if e.g. NYC has 25% seropositivity and that's enough to dampen spread given mask-wearing, etc., incoming Columbia students are probably closer to the US-wide average (~10%) and they're mostly associating with one another so nothing stopping rapid spread within that group. — PolitiTweet.org

Nate Silver @NateSilver538

I know this is an obvious point, but a college that's bringing in students from all around the country is gonna produce different dynamics as far as COVID spread goes than one that's mostly drawing from the local area. — PolitiTweet.org

Nate Silver @NateSilver538

RT @ClareMalone: A new pod, post-first night of odd televised convention viewing https://t.co/bl7ZQySBQm — PolitiTweet.org

Nate Silver @NateSilver538

RT @laurabronner: If you haven't yet, check out this incredible resource detailing how to vote in every state by @baseballot and @juruwolfe… — PolitiTweet.org

Nate Silver @NateSilver538

Can you feel the excitement? And/or are stuck at home with nothing better to do? It's time for the: 5 3 8 🥳 C O N V E N T I O N 🥳 L ! V E B L O G https://t.co/JW2jM6576O — PolitiTweet.org

Nate Silver @NateSilver538

Overall, this is a period where you don't necessarily learn a lot from the polls. Now and then, a convention will be a real inflection point in the race, but more commonly, whichever party gains ground tends to give much of it back. — PolitiTweet.org

Nate Silver @NateSilver538

In addition, our model applies a *convention bounce adjustment* since the party that just held its convention sometimes sees artificially high numbers. However, it's a very modest adjustment this year as compared to the effect it has had in the past. https://t.co/nVtlpQscbB https://t.co/aA3AJEybdS — PolitiTweet.org

Nate Silver @NateSilver538

How so? Our forecast will hedge against movement in our polling average around the conventions by giving some weight to that pre-convention snapshot. So if, say, Biden gains 3% in our polling average, our forecast won't give him full credit for it until he maintains it for a bit. — PolitiTweet.org

Nate Silver @NateSilver538

So, unless there are more polls tonight, it looks like our final pre-convention snapshot of the race will be Biden +8.2. But note that there may be some divergence between our *forecast* and our *polling average* now that the conventions are starting. https://t.co/cy51vc5isJ — PolitiTweet.org

Nate Silver @NateSilver538

It seems like we need a few more gradations in talking about the USPS. A lot of the rhetoric on here falls into categories 1 or 5, while the reporting would point toward 2, 3 or 4. https://t.co/DKch1BYtMI — PolitiTweet.org

Nate Silver @NateSilver538

NYC is not nearly as locked down as people assume (and the suburbs are even less so), but it's also a long way from normal and there's fairly high mask-wearing. So ~25% seroprevalence may be enough to keep R<=1 under those conditions but perhaps not under "business as usual". — PolitiTweet.org

Nate Silver @NateSilver538

Media coverage around herd immunity has gotten more nuanced. But one nuance that's a bit lacking in coverage IMO is an understanding that a COMBINATION of moderately high seroprevalence + moderate control measures may be enough to achieve containment when either alone might not. — PolitiTweet.org

Nate Silver @NateSilver538

FWIW, I intuitively expected Biden to have a higher probability than the ~70% he came out with. One reason to build models is it forces you to be more rigorous rather than to "trust your gut" (I don't think "guts" are very good at election forecasting given small samples, etc.) https://t.co/XgGeVqhIyq — PolitiTweet.org

Nate Silver @NateSilver538

And... have some further reflections up on our forecast model and why I think other models are overconfident. This started out as a tweetstorm and we decided to turn it into an article instead, so it has a very "bloggy" feel! https://t.co/C1M0OsfzGz — PolitiTweet.org