Nate Silver

Nate Silver @NateSilver538

p.s. This would suggest the decline in social distancing from where it was just after lockdown to where it is now will raise R by something like 0.1. That's not *that* much. But since we were barely (if at all) reducing cases before, it could be enough to matter. — PolitiTweet.org

Nate Silver @NateSilver538

FWIW, there may also be some effects from partial herd immunity in edge cases, e.g. cases are falling faster than the model thinks they "should" be falling in NY. This effect is likely to be negligible for most of the country, though. — PolitiTweet.org

Nate Silver @NateSilver538

R's are estimated using the technique described here and attempt to account for the number of tests. They reflect a median of the previous 4 weeks, which should capture the post-lockdown period fairly well. https://t.co/7CxfmiXPWZ — PolitiTweet.org

Nate Silver @NateSilver538

But controlling for mobility, they're also falling more in rural states than urban ones. Montana isn't having many COVID-19 cases even though mobility is high there, for instance. https://t.co/2Xw5KWgyc7 — PolitiTweet.org

Nate Silver @NateSilver538

This is perhaps interesting. You can do a pretty good job of predicting recent R in each state based on Apple's mobility data and urbanization. No surprise, but cases are falling more where people are moving around less. https://t.co/LGkYr6sxFi — PolitiTweet.org

Nate Silver @NateSilver538

Because there is now a fairly explicit policy debate about what to do next—more so than a few weeks ago—that's likely to spill over into a lack of consensus about how to interpret the data. And with COVID data, there are few givens and most things are subject to interpretation. — PolitiTweet.org

Nate Silver @NateSilver538

I am fairly pessimistic about the near-to-medium term also, because social distancing is fraying and we don't have a Plan B. At the same time, there are a lot of unknowns, including exactly how big the effects of fraying social distancing are, which we won't know for a few weeks. — PolitiTweet.org

Nate Silver @NateSilver538

Let me put this carefully: there is often a disconnect between what I'd call the "mood of the internet on COVID-19" vs. the data/evidence on a given day. Today the mood was pessimistic but the numbers were comparatively good. Can think of other times when opposite has been true. — PolitiTweet.org

Nate Silver @NateSilver538

First day since April 1 with <1000 deaths. And the 9% positive test rate is tied for the lowest so far. — PolitiTweet.org

Nate Silver @NateSilver538

So... Mondays are often slow days for reporting, especially for death tolls. Many states don't just do as much counting over the weekend. Tomorrow's numbers will likely be considerably higher. With that said, these are pretty encouraging numbers even accounting for that. — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking: Newly reported deaths: Today: 938 Yesterday: 1,158 One week ago (4/27): 1,163 Newly reported cases: T: 21K Y: 26K 4/27: 22K Newly reported tests: T: 232K Y: 239K 4/27: 190K Positive test rate: T: 9% Y: 11% 4/27: 11% — PolitiTweet.org

Nate Silver @NateSilver538

The NYT is doing great work on COVID-19. But especially given how much confusion already exists about projections and models, it was irresponsible of them to release the slide deck and put such bold, somewhat misleading headlines on it without providing any of that context. — PolitiTweet.org

Nate Silver @NateSilver538

So the Washington Post adds a lot more context, though I think their headline is misleading. The 200k cases/day number was from an incomplete draft version of model and apparently one of many scenarios, not some sort of official White House estimate. https://t.co/jfS0m2aDN6 https://t.co/u0AGMMYVcq — PolitiTweet.org

Nate Silver @NateSilver538

RT @svscarpino: Mean number of days between someone getting infected with #COVIDー19 and their *infector* showing symptoms is negative 0.19… — PolitiTweet.org

Nate Silver @NateSilver538

That is better than where the US has been recently (~1,800 per day) but not a whole heck of a lot better. Canada's numbers also seem to be getting slightly worse over time. https://t.co/ZtcplJfYgb https://t.co/4kVOV82l7y — PolitiTweet.org

Nate Silver @NateSilver538

Also, Canada's data is starting to look less positive. Here are their daily death tolls for the past 7 days *scaled up to the US population* (i.e. multiplying by ~8.7 since the US has 8.7 times more people than Canada). 1,397 1,013 1,528 1,807 1,641 1,196 1,327 — PolitiTweet.org

Nate Silver @NateSilver538

Yeah, could also be: someone wants to scare Trump out of touting re-opening, so they're releasing some sort of worst-case scenario or conjectural scenario and it's being portrayed as the base case. https://t.co/dO8uFcdxX5 — PolitiTweet.org

Nate Silver @NateSilver538

@SeanTrende I don't know, but I think the lack of context in NYT's reporting on this is regrettable. There are some things in that document that don't make much sense on the surface, and maybe they have good explanations, but we don't have many explanations at all. — PolitiTweet.org

Nate Silver @NateSilver538

@JayCostTWS Yeah I guess if tests increase tenfold by the end of the month, something like this could make sense. I doubt that's very likely, though! — PolitiTweet.org

Nate Silver @NateSilver538

Without more context and explanation, are we sure that the White House isn't just leaking these to make the actual, merely-awful numbers look better by comparison? — PolitiTweet.org

Nate Silver @NateSilver538

Indeed, this also doesn't make much sense. https://t.co/RzV2Hufgh2 — PolitiTweet.org

Nate Silver @NateSilver538

Projecting ~200K new detected cases each day **by the end of the month** from a current baseline of ~30K is also a LOT, even with quite pessimistic assumptions. I would encourage some caution with taking these at face value. — PolitiTweet.org

Nate Silver @NateSilver538

Why are *reported* deaths in this slide so different than projected deaths? Was this a model originally developed in March or something? It is very strange. https://t.co/Pb4XdigjgT https://t.co/xJxNftpj7Y — PolitiTweet.org

Nate Silver @NateSilver538

RT @onyxfish: A reminder that we have a continuously updating tracker of *six* COVID-19 models, for each state and the nation. https://t.co… — PolitiTweet.org

Nate Silver @NateSilver538

@ForecasterEnten Seems like you're ignoring the millions of Americans who had their votes thrown out because they tried to vote for both Elizabeth Warren and Amy Klobuchar. — PolitiTweet.org

Nate Silver @NateSilver538

RT @aedwardslevy: This goes, like, quadruple if you yourself have a wide social media footprint or any status as a public figure -- if you'… — PolitiTweet.org

Nate Silver @NateSilver538

RT @aedwardslevy: Friendly reminder that anyone can write down numbers on Twitter and claim they're a poll (or an epidemiological stat, or… — PolitiTweet.org

Nate Silver @NateSilver538

@joshtpm There is a lot of support in polls but i) it seems plausible that the month of May will be an inflection point ii) actions speak louder than words/attitudes iii) if a significant minority of the population (say 1/3) disregards social distancing, that still creates issues — PolitiTweet.org

Nate Silver @NateSilver538

@minakimes The other day I sat through several minutes of rambling on a video about a stuck laptop key, only to realize the entire video was just a guy rambling about his stuck 'H' key and there was no actual advice on how to fix it. — PolitiTweet.org

Nate Silver @NateSilver538

Appeals along the lines of —Here's why we need to *continue* the lockdown (even though it's longer than we hoped for originally). —Here's what it will gain us —Here's a plan for when it will end …may be a more persuasive approach than presuming the lockdown to be the default. — PolitiTweet.org