Nate Silver

Nate Silver @NateSilver538

Mondays have been *very* slow reporting days lately so keep that in mind. The numbers will likely get worse again over the next few days. Still, fewest cases since 3/25 and fewest deaths since 3/28, and the lowest positive test rate yet. So, some positive news. — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking: Newly reported deaths Today: 469 Yesterday: 656 One week ago (5/25): 523 Newly reported cases T: 16K Y: 23K 5/25: 19K Newly reported tests T: 404K Y: 406K 5/25: 441K Positive test rate T: 4% Y: 6% 5/25: 4% — PolitiTweet.org

Nate Silver @NateSilver538

p.s. I don't necessarily have a broader point here but the variation is interesting. There are a fair number of confounders to consider (e.g. weather, lockdown rules relaxing). Also, while protests may encourage movement, curfews and fear of disorderly activity may discourage it. — PolitiTweet.org

Nate Silver @NateSilver538

A few more: Denver +29% Miami +29% Richmond +12% New Orleans +7% Houston +6% Austin +6% Nashville +4% Detroit +2% Philadelphia -0% Salt Lake City -3% — PolitiTweet.org

Nate Silver @NateSilver538

Data from here and reflects an average of driving, walking and transit. https://t.co/32iSFYgSGd — PolitiTweet.org

Nate Silver @NateSilver538

Change in mobility in select US cities, yesterday vs. Sunday 5/24: Boston +52% Washington D.C. +25% St. Louis +21% Dallas +11% Minneapolis +10% New York City +7% Louisville +2% Atlanta +1% Seattle -2% Los Angeles -3% Milwaukee -3% San Francisco -3% Chicago -3% Cleveland -4% — PolitiTweet.org

Nate Silver @NateSilver538

Hard to think of a better way to facilitate complacency and undermine trust in experts than to constantly emphasize worst-case outcomes and treat them as though they're the base case. Usually the worst case doesn't come to pass. But the actual outcomes can still be very bad. — PolitiTweet.org

Nate Silver @NateSilver538

In fact, the number of deaths through the end of the month almost exactly matched what the consensus of expert models expected on May 4. But those models got little attention in the media. https://t.co/EcO0msAlUk https://t.co/jlaRKirVJx — PolitiTweet.org

Nate Silver @NateSilver538

What actually happened? We've averaged 942 deaths and 21K cases per day over the past week (via @COVID19Tracking). So way off from those supposed projections. But here's the thing: this isn't a surprise. The actual numbers have been very close to the true expert consensus. — PolitiTweet.org

Nate Silver @NateSilver538

4 weeks ago there was tons of media coverage of "leaked White House projections" that supposedly showed deaths rising to 3K/day and cases to 200K/day by today. As I said, these projections never made sense and shouldn't have been treated so credulously. https://t.co/pivgOjiJfg — PolitiTweet.org

Nate Silver @NateSilver538

RT @samswey: I wrote about the data on police violence, what’s changed since 2014, and how the data can help us identify solutions to reduc… — PolitiTweet.org

Nate Silver @NateSilver538

RT @Tonyhkchow: Nationwide, the number of people killed by police has been relatively constant since 2013 even though the police are killin… — PolitiTweet.org

Nate Silver @NateSilver538

Again, I don't think we can be super confident about the effect of the protests/unrest on the election campaign. But be wary of explanations that suggest Trump is some clever political strategist, as opposed to someone who isn't up to the job, is acting on impulse, etc. — PolitiTweet.org

Nate Silver @NateSilver538

RT @jtlevy: I don't think people should have very confident views about the political effects of any particular protest tactic, and certain… — PolitiTweet.org

Nate Silver @NateSilver538

There's poor performance in the midst a crisis (think of, say, the mayor of a certain large city on the Eastern Seaboard) but when you're not even able or willing to go the motions of providing traditional leadership that seems like another level entirely. https://t.co/LcGuybIZGN https://t.co/bJwolDWKU3 — PolitiTweet.org

Nate Silver @NateSilver538

Still, cases are also up a bit week-over-week, though there were a LOT of tests today so the positive test rate is unchanged. Overall, we've alternated between reports that show a downturn (good news) and a plateau (less good). This is more on the plateau/less good side. — PolitiTweet.org

Nate Silver @NateSilver538

So, this is the first day in several weeks when deaths are up from the same day a week earlier (656 today vs. 654 last Sunday). But, it's not a big difference and last week's number may have been especially low because of the holiday weekend. — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking: Newly reported deaths Today: 656 Yesterday: 989 One week ago (5/24): 654 Newly reported cases T: 24K Y: 23K 5/24: 21K Newly reported tests T: 441K Y: 396K 5/24: 379K Positive test rate: T: 5% Y: 6% 5/24: 5% — PolitiTweet.org

Nate Silver @NateSilver538

*prediction — PolitiTweet.org

Nate Silver @NateSilver538

Val Demings, who was formerly the Orlando police chief before joining Congress, has moved into 3rd place (12%) in predation markets as Biden's most likely VP pick, behind Elizabeth Warren (13%) and Kamala Harris (39%). https://t.co/Bv81t1GNsG — PolitiTweet.org

Nate Silver @NateSilver538

RT @BeschlossDC: “Polls are phony,” “rigged,” with “fake results”—George Wallace, Albany, NY, October 9, 1968: https://t.co/SfArDImDbv — PolitiTweet.org

Nate Silver @NateSilver538

So that provides some perspective on how various stakeholders (the media, governments, etc.) & how people with different political preferences may interpret/narrate/"spin" COVID-19 data. It is not the only perspective on these issues, but I believe it is relevant. Take care, all. — PolitiTweet.org

Nate Silver @NateSilver538

Not going to get into any arguments today. But I have a lot of hands-on experience as a both a journalist and a practitioner on questions of: 1a) how the media covers complex data in politically fraught environments; 1b) how elites argue about it; 1c) how voters perceive it. — PolitiTweet.org

Nate Silver @NateSilver538

Note particularly that last poll, which shows 56% of people (including a plurality of whites) saying race relations have gotten worse since Trump became president vs. just 10% who say they've gotten better... and that was before George Floyd. — PolitiTweet.org

Nate Silver @NateSilver538

I have no idea how the politics of a "law and order" presidential campaign would play out amidst everything else going on in 2020, but it's worth noting that Trump receives VERY poor marks from the public for his handling of race relations. https://t.co/CTQenOhrRZ https://t.co/g5wl5XHGZ0 — PolitiTweet.org

Nate Silver @NateSilver538

The good news, I guess, is that predictions of one-off events that would trigger spikes don't have a great track record. That is, the predicted spike sometimes never comes. But those are often fairly small, one-off events, whereas these protests may have wider scale. — PolitiTweet.org

Nate Silver @NateSilver538

In 2-3 weeks, you'll probably have conservatives arguing the data shows a spike of COVID-19 cases in Minneapolis (and/or other cities where there were significant protests) and liberals arguing that it doesn't. The reverse of what happened in Georgia, Florida, etc. — PolitiTweet.org

Nate Silver @NateSilver538

But, cases are up a tick from last Saturday, though the positive test rate is similar. Stay safe, everyone. — PolitiTweet.org

Nate Silver @NateSilver538

So, it's good to see the number of deaths below 1,000 (albeit barely). Saturdays are *not* particularly slow days for reporting since they mostly reflect what states were doing on Fridays (i.e. a weekday). This is the first Saturday with under 1,000 deaths reported since 3/28. — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking Newly reported deaths Today: 989 Yesterday: 1,186 One week ago (5/23): 1,091 Newly reported cases T: 23K Y: 24K 5/23: 22K Newly reported tests T: 396K Y: 492K 5/23: 366K Positive test rate T: 6% Y: 5% 5/23: 6% — PolitiTweet.org