Nate Silver

Nate Silver @NateSilver538

There's also likely to be a fair amount of almost literal randomness/stochasticity in outcomes. If a group of tourists from Bergamo decides to visit NYC rather than Toronto in February, that could make a big difference. Also, there may be different strains of the virus. — PolitiTweet.org

Nate Silver @NateSilver538

For instance, it's easy to think of some disadvantages the US has, also. Lots of international visitors. Many people with pre-existing conditions. Higher average household size than parts of W. Europe (though low compared w/other regions). Also, NYC is its own unique beast. — PolitiTweet.org

Nate Silver @NateSilver538

If you want to move in the direction of trying to control for other factors, and thereby tease out a residual for "quality of governmental response"... then I'm OK with that, at least in theory! In practice, which variables to pick from is a hard problem. — PolitiTweet.org

Nate Silver @NateSilver538

RT @galendruke: 🚨 NEW PODCAST ALERT 🚨 Today FiveThirtyEight is launching PODCAST-19, an exploration of the science behind the coronavirus… — PolitiTweet.org

Nate Silver @NateSilver538

Two things can be true: * Biden may well have something to hide. * "If you have nothing [major] to hide, you have nothing to fear from opening up records" is not an obviously correct argument given how American politics and media work. — PolitiTweet.org

Nate Silver @NateSilver538

@kmedved @youyanggu @CT_Bergstrom Yeah, I will see if we can add it! — PolitiTweet.org

Nate Silver @NateSilver538

Please actually read the original post? Canada's outcomes started out much better than the US, but they've been con… https://t.co/J90PzetOo7 — PolitiTweet.org

Nate Silver @NateSilver538

We sure as hell aren't building our own COVID-19 model. But we built an interactive so you can track other people's COVID-19 models. https://t.co/ksZhgvgzso — PolitiTweet.org

Nate Silver @NateSilver538

@mattyglesias BTW, I don't know that I'd make the comparison with Italy and Spain, which may have been unlucky in some ways (i.e. it hit hard very early on). The UK seems salient, though, as does Canada, and Canada's outcomes are becoming similarly bad to ours. https://t.co/hWwG3TKFGV — PolitiTweet.org

Nate Silver @NateSilver538

@mattyglesias Well, clearly there's something different about outcomes in East Asian countries vs. in the North Atlantic countries. Is it knowledge from SARS? "Culture"? A different strain of the virus? Mask-wearing? A different conception of individual rights? All of the above? I don't know. — PolitiTweet.org

Nate Silver @NateSilver538

@SeanTrende @Tracinski @asymmetricinfo There are also theories that cases are likely to be more severe on average if people have multiple exposures, as is likely amidst very bad outbreaks (Lombardy, NYC). So there are several reasons the IFR might not be perfectly constant. — PolitiTweet.org

Nate Silver @NateSilver538

And all of these countries are struggling with the "reopening conversation" just as much as we are. This might suggest that authors of these plans are not thinking enough about what the actual political (and logistical, etc.) constraints are. — PolitiTweet.org

Nate Silver @NateSilver538

Among what you might consider our closest peers, the UK's outcomes have been somewhat worse than the US across a variety of metrics. Canada's have been somewhat better but are trending toward the US. Germany better (though more similar than you'd think outside NYC). France worse. — PolitiTweet.org

Nate Silver @NateSilver538

The fact that many peer countries are struggling with COVID-19 roughly as much as the US suggests there's some naivety to claims such as "the US could greatly improve the situation with [my proposed solution]… IF ONLY WE HAD BETTER LEADERSHIP!!!". — PolitiTweet.org

Nate Silver @NateSilver538

@asymmetricinfo @SeanTrende I mean, I basically don't get why there's been so much focus on the Santa Clara study. There are a LOT of other studies that are more methodologically sound and/or which occur in higher-prevalence areas. — PolitiTweet.org

Nate Silver @NateSilver538

@asymmetricinfo @SeanTrende The pitch included a lot of rationales for why one might want to get tested. It also asked for "HEALTHY" volunteers, and someone who is symptomatic is unlikely to leave the house to get tested in any event. So there are lots of potential biases in all directions. — PolitiTweet.org

Nate Silver @NateSilver538

RT @Colinoscopy: made a Rube Goldberg machine https://t.co/gWRNnmm5Ic — PolitiTweet.org

Nate Silver @NateSilver538

@kmedved Yeah. There's also something about the Trump era where (especially among cultural and/or political liberals) pessimism signifies a sort of savviness for how one navigates the news. — PolitiTweet.org

Nate Silver @NateSilver538

I'm not sure I totally understand the mindset of the people who only retweet the most dire, pessimistic (and sometimes misleading) headlines. (This is a subtweet, but of like 12 different people.) — PolitiTweet.org

Nate Silver @NateSilver538

Very good example of how testing backlogs can skew what the curves look like. — PolitiTweet.org

Nate Silver @NateSilver538

COVID-19 in Canada: —Not much sign of a slowdown; R seems to be ~=1. —Testing per capita ~same as US —Starting to approach US levels of deaths *per-capita* (about 2/3 as high over past week). —Provinces re-opening on similar timelines to US states. https://t.co/XipuBTePjR https://t.co/KFvoIDfYV8 — PolitiTweet.org

Nate Silver @NateSilver538

Having not quite been able to choose between suppression and mitigation the US may wind up halfway in between. Possibly Europe also. — PolitiTweet.org

Nate Silver @NateSilver538

This seems smart. A pleateu maybe more likely than a distinct 2nd peak. But not necessarily a much better outcome than a 2nd peak. — PolitiTweet.org

Nate Silver @NateSilver538

*Some* of that is driven by NY, where the rate of positive tests has plummeted (genuinely good news). But there are some signs of progress elsewhere. Outside NY, the positive test rate has declined to ~12% from a peak of ~17%. — PolitiTweet.org

Nate Silver @NateSilver538

On the number of new cases, the news continues to be OK-ish. Cases are steady/declining even as tests have increased. The rolling 7-day average positive test rate is now around 13%, down from a peak of ~20%. — PolitiTweet.org

Nate Silver @NateSilver538

We may never capture some deaths, other than through statistical inference, and there are likely to be a lot of methodological debates over how to make those inferences, so it won't be a perfectly precise count. — PolitiTweet.org

Nate Silver @NateSilver538

In theory, the gap between **actual deaths** and **reported deaths** will close over time. Even if everyone in the US were miraculously cured of COVID-19 today, the **reported** death count would continue to rise for weeks, maybe even for months. — PolitiTweet.org

Nate Silver @NateSilver538

We know that considerably more people have died in most states than the confirmed death counts. What "considerably" means will vary from state to state. Some states are better at counting deaths that occur in nursing homes. Some count "probable" deaths and some do not. — PolitiTweet.org

Nate Silver @NateSilver538

It's become hard to know what the trajectory is on death tolls. Some states are going back and counting deaths they missed before, e.g. at nursing homes. The # of newly-recorded deaths from day to day may not align perfectly to when people actually died. https://t.co/s1Asv30ePH — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking: Newly reported deaths: Today: 2,041 Yesterday: 2,700 One week ago (4/23): 1,877 Newly reported cases: T: 28K Y: 28K 4/23: 32K Newly reported tests: T: 207K Y: 231K 4/23: 192K Positive test rate: T: 14% Y: 12% 4/23: 16% — PolitiTweet.org