Nate Silver

Nate Silver @NateSilver538

In fact, fundamentally I don't know there was even agreement about whether we were pursing mitigation (flatten/slow the curve to avoid overloading ICUs but don't ultimately prevent the spread) or suppression (significantly lower the overall number of cases until a cure/vaccine). — PolitiTweet.org

Nate Silver @NateSilver538

Important to note: This is not what everyone calling for social distancing assumed. There were a wide variety of stances, all of which were served by the same *near-term* goal. Also, their report was maybe unduly pessimistic about alternatives to distancing, e.g. contact tracing. — PolitiTweet.org

Nate Silver @NateSilver538

People say "nobody actually called for intensive social distancing for 18 months until a vaccine is found" but that's exactly what the famous Imperial College report did call for. (Though it allowed for periodic breaks in social distancing.) https://t.co/0lITMpvxQH https://t.co/Q5BC83vGV3 — PolitiTweet.org

Nate Silver @NateSilver538

@ClaraJeffery It's not clear whether that would snuff it mostly out. Or it might snuff it *mostly* out, but far from *completely* out (with a large number of essential workers unable to socially distance) and it would start to increase again on July 2. It's genuinely a really hard problem. — PolitiTweet.org

Nate Silver @NateSilver538

I think a lot of governors would tolerate that. It actually seems marginally optimistic, if anything. There's a lot of talk about opening up in places where death counts are currently higher on a per capita basis. That's still an additional 50K-70K deaths in the "plateau" phase. — PolitiTweet.org

Nate Silver @NateSilver538

The US and most other countries will likely tolerate some baseline level of deaths to open their countries back up. How high? I don't know. But say the US gets down to 1/5 of the current levels and hold there for 6 months until effective treatments lower the IFR, so 300-400/day… — PolitiTweet.org

Nate Silver @NateSilver538

The thing about some of these death count projections is that they all assume we eventually get down to ~0, and that is very unlikely to happen (until therapies/vaccines/herd immunity). — PolitiTweet.org

Nate Silver @NateSilver538

RT @ClareMalone: New pod: we talk COVID, Trump's slipping numbers, and yes, a little more about Biden's basement campaign. https://t.co/hB… — PolitiTweet.org

Nate Silver @NateSilver538

Outdoor time is also a potential confounder for a lot of things people have tried to study, such as weather, density and wealth. California is pretty outdoors-oriented despite being dense and wealthy, for instance, and it has had a surprisingly low rate of cases. — PolitiTweet.org

Nate Silver @NateSilver538

Has anybody looked at the relationship between COVID cases and the amount of time people spend outdoors in different states/countries? — PolitiTweet.org

Nate Silver @NateSilver538

Meanwhile, after several days in a row with 200K-300K tests, we reverted back to 136K. It's the fewest new positives since March 30, but that partly reflects the relatively low volume of tests; the positive rate (16%) was pretty good but not great. — PolitiTweet.org

Nate Silver @NateSilver538

We've had two days in a row where death counts are closer to 1K than 2K. That's good, but Sundays and Mondays are typically low days for reporting. Tomorrow will likely be worse. Still, these numbers are down from previous Sundays/Mondays. — PolitiTweet.org

Nate Silver @NateSilver538

US daily numbers via @COVID19Tracking. Weird day. Newly reported deaths: Today: 1,163 Yesterday: 1,095 One week ago (4/20): 1,689 Newly reported cases: T: 22K Y: 27K 4/20: 25K Newly reported tests: T: 136K Y: 208K 4/20: 149K Share of positive tests: T: 16% Y: 13% 4/20: 17% — PolitiTweet.org

Nate Silver @NateSilver538

One theory is that countries follow the tone set by their leaders. Most places *currently* have popular leaders who got a COVID-19 approval-ratings bounce. In the US, however, Trump is unpopular and distrusted on COVID. So people oppose re-opening because he seems to be for it. — PolitiTweet.org

Nate Silver @NateSilver538

I also find it interesting that American liberals are generally wary of reopening but (given that American liberals are generally fond of Europe) Europeans are more equivocal and Sweden is the most "YOLO, herd immunity!" place in the world. — PolitiTweet.org

Nate Silver @NateSilver538

Certainly *some* correlation between wealth and desire to reopen, with poorer countries wanting to re-open sooner, but it's a noisy one, e.g. Mexico is against re-opening Germany is relatively for it. — PolitiTweet.org

Nate Silver @NateSilver538

I have no idea what the pattern is here. It's not particularly correlated with how bad the outbreak was, either currently, or at its peak (if the country already peaked; some haven't). But it's interesting that the US is in the low-to-middle-end of the pack. — PolitiTweet.org

Nate Silver @NateSilver538

On a typical day in NYC, something like 10% of newly-reported *confirmed* deaths actually occurred 2 or more weeks ago. I doubt NYC is alone in this regard. — PolitiTweet.org

Nate Silver @NateSilver538

At appears that on average, it takes about 6 days for a death to occur before it shows up in the "confirmed" column. That's an average, though; there's a long tail of deaths they're counting that actually occurred weeks ago. Probable deaths can often catch these sooner. — PolitiTweet.org

Nate Silver @NateSilver538

The further back you go, the more likely a death is to be confirmed instead of probable. Around 80% of (confirmed + probable) deaths that occurred in March were confirmed, for instance. But only ~50% of deaths to occur in the past 5 days are confirmed. — PolitiTweet.org

Nate Silver @NateSilver538

This is interesting data from New York City, which has been i) tracking probable in addition to confirmed deaths and ii) tracking deaths on the date they think the death actually occurred, not when it shows up in their records. https://t.co/O1f8OmAlrI — PolitiTweet.org

Nate Silver @NateSilver538

Over the past 10 days or so, the rate of decline in deaths in New York state has been consistent with an R of around ~0.75. That is decently steep. — PolitiTweet.org

Nate Silver @NateSilver538

Usual caveats apply about how these studies could be overestimates, *or* underestimates. https://t.co/UUhObb5aDW — PolitiTweet.org

Nate Silver @NateSilver538

NYC now up from to 25% of people having antibodies, up from 21% in earlier testing. Not clear if the increase is because of the larger sample sizes or because a bit more time has passed, allowing more to develop antibodies. — PolitiTweet.org

Nate Silver @NateSilver538

RT @baseballot: Wondering why every state isn't just switching to vote-by-mail? Turns out it's super hard! https://t.co/M6iTvE6bQe — PolitiTweet.org

Nate Silver @NateSilver538

RT @galendruke: How predictive are presidential election polls 200 or 100 or 5 days out from Election Day? @geoffreyvs has some great visu… — PolitiTweet.org

Nate Silver @NateSilver538

My guess is 3 > 2 and 2 > 1, especially given the narrow range of interventions available in the short run. (If every country could "do South Korea", it might be different, but they can't on short notice.) That could lead to some wrong conclusions about what works, what doesn't. — PolitiTweet.org

Nate Silver @NateSilver538

The buckets are something like: 1) Policy interventions 2) Immutable location-based characteristics, e.g. geography, density, age structure of the population, "culture", weather 3) Luck. How much undetected community spread was there before March? Perhaps strains matter also. — PolitiTweet.org

Nate Silver @NateSilver538

A big challenge in determining the effectiveness of policy interventions (i.e. social distancing) is that there are really several buckets that determine how bad the outbreak is in any given area and it's not clear policy interventions are the most important one in the short run. — PolitiTweet.org

Nate Silver @NateSilver538

Gates is quite optimistic about treatments and vaccines and quite against opening up right now. Conversely, you'll read takes that are like "we may never get a vaccine... so we might as well open back up and learn how to live it." — PolitiTweet.org