Nate Silver

Nate Silver @NateSilver538

Also, empirically, the best way to forecast the popular vote is to forecast the result in each state and combine them, rather than to use national polls. We don't have that calculation yet but will soon. Might yield a different impression of the popular vote vs. EC gap. — PolitiTweet.org

Nate Silver @NateSilver538

I wouldn't totally sleep on Nevada and New Hampshire as factors, by the way. Yeah, Nevada's getting bluer, but it was a very good state for Trump in the (2016) primary and a poor one for Biden. And New Hampshire is weird and nearly went for Trump in 2016. — PolitiTweet.org

Nate Silver @NateSilver538

Rather small gap at the moment between tipping-point state polls and national polls: National: Biden +7.7 Florida: Biden +7.5 New Hampshire: Biden +7.1 Michigan: Biden +7.0 Pennsylvania: Biden +6.7 Wisconsin: Biden +6.6 Nevada: Biden +6.5 https://t.co/nWUo9X5F4y https://t.co/D0IccKEzCL — PolitiTweet.org

Nate Silver @NateSilver538

@mattyglesias It's possible they believe that. But, of course, parties almost always come up with electoral excuses/rationales when they want to achieve policy or ideological goals and given the current state of the pandemic that is among the thinner fig-leaves. — PolitiTweet.org

Nate Silver @NateSilver538

Given that it's an election year, I don't understand why Republicans wouldn't just want to stimulate the living 'eff out of the economy. One of the clearer examples of where ideological dispositions outweigh electoral incentives. — PolitiTweet.org

Nate Silver @NateSilver538

@joshtpm Estimates of the herd immunity threshold have a wide range, from as low as 20% to as high as 60-70%. Personally, I think the lower values are too optimistic. But 20% (likely more like 25% seroprevalence) in NYC *could* matter a lot. — PolitiTweet.org

Nate Silver @NateSilver538

@jbarro I don't know. Low-ish density + can do everything outdoors in the summer? I'll grant you it's all pretty complicated. But if you're building a mental model, this is one of the important factors. We KNOW it matters to some degree (i.e. in an SIR model) and it COULD matter more. — PolitiTweet.org

Nate Silver @NateSilver538

How big a factor is hard to say (see article below). And having had a bad past outbreak may make people more cautious, in addition to providing immunity to some percentage of the population. But, you cannot just ignore it when comparing regions. https://t.co/2YOsC0idji — PolitiTweet.org

Nate Silver @NateSilver538

A continued pet peeve of mine: given that there is at-least-short-term immunity for the large majority of cases, you have to account for the high rate of past infections in hard-hit areas (say the Northeast) in assessing why they're now doing better with the virus. It's a factor. — PolitiTweet.org

Nate Silver @NateSilver538

OANN is currently the only organization that we've had to apply this policy to, but please be aware that we'll also apply it to other sponsors that selectively release data to the benefit of one candidate or party, which clearly qualifies as partisan behavior. — PolitiTweet.org

Nate Silver @NateSilver538

Polls sponsored by media organizations are generally *not* considered partisan, even if that organization has a clear viewpoint. But we make an exception if they cherry-pick or selectively release data, as OANN has been doing. https://t.co/JaiAGybfAU — PolitiTweet.org

Nate Silver @NateSilver538

So as an FYI, we fixed that tonight, which slightly changed the averages in some states. Also, here is our newly-updated polls policy, which explains what is considered a partisan poll. https://t.co/37N8bDqsnB — PolitiTweet.org

Nate Silver @NateSilver538

Namely, we use a stricter house effects adjustment; the model assumes that these polls are biased until proven otherwise. And it weights them less. However, for boring technical reasons, the model wasn't catching all the partisan polls and adjusting them like it was supposed to. — PolitiTweet.org

Nate Silver @NateSilver538

As an FYI, there have been a fairly high number of polls with partisan sponsors in the presidential race this year. (Usually these are more common in races for Congress.) We DO include these polls in our averages, which is a change from the past, but with an adjustment. — PolitiTweet.org

Nate Silver @NateSilver538

Well, this was a pretty miserable day, and seeing weekday death tolls >1,000 may be the new normal for a bit. There are real signs of improvement in AZ and of plateaus in FL and TX. But between deaths lagging cases and other states getting worse, it's not a pretty picture. — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking: Newly reported deaths Today: 1,126 Yesterday: 1,029 One week ago (7/15): 858 Newly reported cases T: 70K Y: 63K 7/15: 65K Newly reported tests T: 796K Y: 749K 7/15: 762K Positive test rate T: 8.8% Y: 8.4% 7/15: 8.5% — PolitiTweet.org

Nate Silver @NateSilver538

There are some reports of re-infections now. They could be false positives. They could be people with longstanding infections that came and went in waves. Or they could be real but rare (since immune defenses run on a spectrum, etc.) But they do not seem to be common. — PolitiTweet.org

Nate Silver @NateSilver538

One thing I'd add is that if you take the "immunity disappears after 3 months!" headlines completely at face value, you'd have expected to see tens or even 100s of thousands of re-infections by now given that there were already millions of people infected 3 months ago. — PolitiTweet.org

Nate Silver @NateSilver538

Couple of good articles out today on immunity and the "can you get COVID-19 twice?" question. Much better and more nuanced (and *mostly* fairly reassuring, though not completely so!) than some of the other coverage I've seen. https://t.co/RPNcgeKGgj https://t.co/bIlRYH6Al8 — PolitiTweet.org

Nate Silver @NateSilver538

I know that's pretty basic stuff, but there's a lot of hard work that goes into distinguishing the 90 percent chances from the 95s, and the 95s from the 98s, and the 98s from the 99.5s. And you have to think carefully about the structure of your data & model for these edge cases. — PolitiTweet.org

Nate Silver @NateSilver538

For instance, a forecast saying there's an 85% chance of an event and a forecast saying there's 98% chance may sound similarly confident. But in the former case, the event will fail to occur around 1 in 6 times, while it's 1 in 50 times in the latter case. Big difference. — PolitiTweet.org

Nate Silver @NateSilver538

Today's the day when I finalize all the parameters in our model related to uncertainty and it always gets me in a mood. One thing it's important to bear in mind is probabilities get scrunched up once you get closer to 100% in ways that can conceal important practical differences. — PolitiTweet.org

Nate Silver @NateSilver538

@IChotiner I mean, the Biden coalition is supposed to include some Obama-Trump voters (from the "white working class" and other groups) who have soured on the president. So that's basically mean-reversion, on an individual level. — PolitiTweet.org

Nate Silver @NateSilver538

@SopanDeb Yeah, basically. I mean, Texas is clearly becoming more purple. But, if Biden wins by 9 nationally and wins Texas by 1, odds are that it's still a Republican-leaning state in 2024/2028/etc. — PolitiTweet.org

Nate Silver @NateSilver538

Another implication of mean-reversion is that Trump's Electoral College edge could be smaller this year, since state leans could look like some hybrid of 2016 and 2012 (when Obama performed strongly in tipping-point states). And I'd argue that you see that in the data, too. — PolitiTweet.org

Nate Silver @NateSilver538

Yes, Biden is sometimes leading in polls of those states (especially Arizona). But he's way ahead nationally so that's not a huge surprise. Right now our averages have him +2.0 in Ohio but +2.3 in Arizona, so pretty similar. All of this is consistent with mean-reversion. — PolitiTweet.org

Nate Silver @NateSilver538

And I'd argue that you're seeing this in the data this year. Biden is doing "surprisingly" well in the Midwestern states like OH, MI and WI that drifted Republican from 2012 to 2016. But his numbers have been pretty tepid in AZ, GA, TX, states where Dems were gaining ground. — PolitiTweet.org

Nate Silver @NateSilver538

One thing that I think people miss is that the partisan lean of a state is slightly mean reverting. Say a state was an R+20 state eight years ago, and an R+10 state four years ago. You might expect it to be a swing state now, right? No. On average it'll be like an R+13. — PolitiTweet.org

Nate Silver @NateSilver538

Obviously, any poll showing the Democrat up in Texas is a pretty good result for the Democrat, but if Quinnipiac has Biden up 1 in Texas when he's up 8-9 nationally (and more than that in Q-Pac's last national poll), it's nowhere near the tipping point. https://t.co/TMncRtcsWi — PolitiTweet.org

Nate Silver @NateSilver538

Per @youyanggu, we could be looking at as few as 50,000 new infections per day (which would likely translate into something like 8K-10K detected cases given that many cases go undiagnosed) or as many as 550,000 infections (around 100K detected cases). — PolitiTweet.org