Nate Silver

Nate Silver @NateSilver538

RT @nataliexdean: Most common issues I see in scientific writing: - Doesn’t explain why the problem is important. - No clear takeaway messa… — PolitiTweet.org

Nate Silver @NateSilver538

RT @PathogenScribe: @nataliexdean I love this quote on the topic: "When writing about science, don’t simplify the science; simplify the wri… — PolitiTweet.org

Nate Silver @NateSilver538

This is true, deaths may increase as cases increase. But if states have different (lower) expectations for what the IFR is going forward, that's liable to affect their policy choices. https://t.co/IDnISm4Zlb — PolitiTweet.org

Nate Silver @NateSilver538

I do think the decline in deaths is weirdly under-discussed, though. And although states are engaging in some degree of hand-waving arguments for why they aren't more locked down, the death rates are likely to be pretty important in what policies they adopt in the long run. — PolitiTweet.org

Nate Silver @NateSilver538

However, cases and the positive test rate are still up week over week, reflecting rises in many Southern and Western states. Today was a tiny bit better than the past few days, but as I said Sundays can be a bit slow, so probably not worth putting too much stock in that. — PolitiTweet.org

Nate Silver @NateSilver538

Sundays are often slow, but this is the fewest deaths reported in a day since March 25, and the 7-day average is down to 592 and has continued to fall at a fairly linear clip. — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking: Newly reported deaths Today: 297 Yesterday: 630 One week ago (6/14): 358 Newly reported cases T: 27K Y: 32K 6/14: 21K Newly reported tests T: 518K Y: 584K 6/14: 485K Positive test rate T: 5.3% Y: 5.5% 6/14: 4.4% — PolitiTweet.org

Nate Silver @NateSilver538

Like you would think from looking at the headlines that the Falklands War was a considerably more important event in world history than the 2004 Indian Ocean tsunami, which is pretty obviously not the case IMO. — PolitiTweet.org

Nate Silver @NateSilver538

Been working on a project for our election model that involves looking at historic NYT headlines & one thing that's striking is how disproportionately likely stories about military conflict are to get multiple days of ALL CAPS full-width headlines as opposed to everything else. — PolitiTweet.org

Nate Silver @NateSilver538

Crowd sizes are perhaps the most meaningless metric and one good outcome from last night would be if the media paid less attention to them going forward, but probably the opposite will happen and they will become an even bigger deal. — PolitiTweet.org

Nate Silver @NateSilver538

@joe_sheehan That's my impression, although one would need to think carefully about how to define homogeneity. — PolitiTweet.org

Nate Silver @NateSilver538

@NAChristakis @youyanggu I do not believe he's age-adjusting his numbers, so that definitely would account for some of it. I do wonder though if NYC did uniquely badly in protecting older people and so it also was not a great indication of the true IFR. — PolitiTweet.org

Nate Silver @NateSilver538

In general, small countries seem to have done reasonably well with the pandemic, although with some exceptions (e.g. Belgium). And so I wonder if there are some advantages to not having the cognitive dissonance of spiking cases in some places and declining ones in others. — PolitiTweet.org

Nate Silver @NateSilver538

I'm not sure what the fix for this is, but part of the problem is trying to apply national frames/narratives to what is essentially a series of simultaneous localized epidemics that at times have taken wildly different trajectories. — PolitiTweet.org

Nate Silver @NateSilver538

Deaths continue to fall, but not clear how sustainable that is. @youyanggu, whose model has been accurate, estimates the IFR is in the midst of falling from 1.0% of symptomatic infections at the start of the pandemic to 0.5%. Good news, but it could be offset by more infections. — PolitiTweet.org

Nate Silver @NateSilver538

Most cases in a day since May 1. In better news, this was a record day for tests. But, note that the positive test rate is ticking UP, not down. So while tests are increasing, cases are increasing faster. — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking: Newly reported deaths Today: 630 Yesterday: 648 One week ago (6/13): 695 Newly reported cases T: 32K Y: 31K 6/13: 25K Newly reported tests T: 584K Y: 579K 6/13: 520K Positive test rate T: 5.5% Y: 5.4% 6/13: 4.9% — PolitiTweet.org

Nate Silver @NateSilver538

RT @kmedved: @NateSilver538 This is the cost of past reporting on spikes in cases without mentioning increased testing. It makes it much ha… — PolitiTweet.org

Nate Silver @NateSilver538

There were a lot of false alarms on Florida a month or two ago. That may have bought DeSantis some time and political cover. But this is increase is real, by all indications, and clear enough that it's increasingly outside the margin of BS'ing, so to speak. — PolitiTweet.org

Nate Silver @NateSilver538

I've long thought that states would tolerate plateaus. But in Florida and Arizona, these are actual, fairly spiky spikes. And not just caused by more testing: positive test rates are high and growing. — PolitiTweet.org

Nate Silver @NateSilver538

Yeah, people keep effing up predictions about particular events having a discernible impact on virus spread. It's not because spread isn't happening at them (it probably is) but because of the scale. It's a big country with lots of COVID, so hard to see impact of any one thing. — PolitiTweet.org

Nate Silver @NateSilver538

One faint, perhaps naive hope I have is that red-state governors will become aggressively pro-mask if they deduce that the other alternatives (re-lockdowns or lots of people getting COVID) are less palatable. — PolitiTweet.org

Nate Silver @NateSilver538

At least when NY/NJ/CT/MA/MI/LA were spiking in the spring, there was the hope/expectation that lockdowns would eventually slow things down. There's not really that to look forward to in places where cases are increasing now. — PolitiTweet.org

Nate Silver @NateSilver538

And yet, deaths are still declining week over week, with the 7-day average having fallen to 610. I'm not sure how sustainable that is. We do seem to be seeing increasing hospitalizations in the South/West, which could be a precursor to death tolls increasing there later. — PolitiTweet.org

Nate Silver @NateSilver538

That's a LOT of new cases today, folks. 32,000 is the most in a day since May 1st. There was ALSO a lot of testing. But the positive test rate has crept up from where it was a couple of weeks ago, so that doesn't explain all of it, especially in the South and Southwest. — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking: Newly reported deaths Today: 650 Yesterday: 695 One week ago (6/12): 775 Newly reported cases T: 32K Y: 27K 6/12: 23K Newly reported tests T: 579K Y: 466K 6/12: 583K Positive test rate T: 5.5% Y: 5.8% 6/12: 4.0% — PolitiTweet.org

Nate Silver @NateSilver538

RT @FiveThirtyEight: We've launched our general election polling averages. https://t.co/mHMsjTf7pw — PolitiTweet.org

Nate Silver @NateSilver538

Almost half the new cases are coming from one of four states: FL, TX, CA and AZ. Note that three of those are very populous, so that's part of it (Arizona has a TON of cases on a per-capita basis right now). But still, not a great situation in any of these places. — PolitiTweet.org

Nate Silver @NateSilver538

This is consistent with trends we've seen in recent days, but a more extreme version. Most cases since May 8. And the highest positive test rate since May 27. Not good at all. But: there's a big week-over-week decline in deaths, with the 7-day average falling to 624. — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking: Newly reported deaths Today: 695 Yesterday: 782 One week ago (6/11): 936 Newly reported cases T: 27K Y: 24K 6/11: 22K Newly reported tests T: 466K Y: 489K 6/11: 448K Positive test rate: T: 5.8% Y: 4.9% 6/11: 4.9% — PolitiTweet.org