An update on a pre-registered result about the coronavirus • Hacker News Board

Re: An update on a pre-registered result about the coronavirus

#111

Post by ceejayoz » Tue, Apr 21, 2020, 6:38 PM UTC

SpicyLemonZest wrote: post #92

XorNot wrote: post #86

Earlier quoted context omitted.

Quarantine measures were widely implemented in major cities during the 1918 pandemic, but at different times in response to different catastrophes and with a very different result in deaths per capita. See the bottom graph. Deaths/week vs when closures were implemented. Another more wordy source: https://www.businessinsider.com.au/history-of-how-st-louis-v... > Around the same time, the city of St. Louis was closing…

Your more wordy source does not describe general business closures and doesn't say the word "quarantine". The lessons of history are clear that we needed to do something, certainly.

"It doesn't say the word quarantine, it only describes it" is a weird hill to die on.

Re: An update on a pre-registered result about the coronavirus

#112

Post by davidw » Tue, Apr 21, 2020, 6:45 PM UTC

ohazi wrote: post #95

Most of this seemed blindingly obvious to me and my peers (in the US) at the time that Patrick tweeted his hashes. Several of the replies to his tweets correctly guessed the topic, and most came to similar conclusions. Anybody with half a brain knew what was going on. Maybe Japanese culture really is "different" and things really did need to be done this way, but I find myself agreeing with other comments here that t…

Yeah, I think it didn't take much to put two and two together, even without much familiarity with Japan:

https://twitter.com/davidnwelton/status/1241556680833298434

Re: An update on a pre-registered result about the coronavirus

#113

Post by clairity » Tue, Apr 21, 2020, 6:47 PM UTC

SpicyLemonZest wrote: post #49

cortesoft wrote: post #25

Earlier quoted context omitted.

I understand the problem with some of the claims people make about things being really bad being unfalsifiable. At the same time, the expected outcome of social distancing is for the outcome to be not as bad... so people will always be able to make this complaint. What exactly do you expect from people who are advocating social distancing?

I expect a degree of epistemic humility. "We don't know how bad it's going to get, but we know that the worst case is very bad, so let's go to extreme lengths to make sure we don't hit it" - no objections to that message. But in my local area, and from what I understand many others, people were saying in mid-March that it was already too late and there would surely be bodies in the streets. Which hasn't been that bad…

i'd love some media humility as well. the self-righteous corona-panic bias of the media i tend to consume (npr, nytimes, etc.) is unwavering and nauseating.

the last couple months of experience tells us (imperfect) physical distancing has done the heavy lifting in flattening the curve, and yet the media orthodoxy is still that stepping one unsheathed toe outside your door is just asking for death. this morning an npr reporter openly mocked a governor for even suggesting a loosening of the lockdown. it's brainless.

Re: An update on a pre-registered result about the coronavirus

#114

Post by sneak » Tue, Apr 21, 2020, 6:56 PM UTC

rmccue wrote: post #80

The clearest communication about the use of hashes rather than direct publication is laid out quite simply in patio11's original memo: > I have concerns about saying this loudly because I have an immigration status and because Japan has previously shot the messengers in times of national emergency, including in the wake of the 2011 disaster. As someone who has spent their entire adult life in Japan, who lives here wi…

At which URL did you find that text? I can't locate it on his site.

I can confirm, though, as an expat, that there is a special kind of fear/restraint involved in needing an opt-in, discretionary permission from a government to be permitted to physically return to your home (which you already paid for) and your family. It causes a lot of security-conscious people to err far on the side of caution, which, indeed, may not even be an error at all.

Immigration policies generally affect a very small minority of people in a country, so strengthening the rights around them simply isn't a priority in most societies, and risks/chilling effects like this fall out as a result.

Re: An update on a pre-registered result about the coronavirus

#115

Post by btilly » Tue, Apr 21, 2020, 6:57 PM UTC

nostromo wrote: post #60

btilly wrote: post #47

Earlier quoted context omitted.

Japan is a country where at the best of times, a lot of people die at home and are not discovered for some time. If you remove social contacts thanks to lockdown, this is going to become more common. See https://en.wikipedia.org/wiki/Kodokushi for more.

Yes, that happens sometimes, but no, there are not thousands of unnoticed dead people in their apartments in Japan. Occam’s razor in this case is that the pandemic in Japan isn’t as bad as the author projected. It’s not a conspiracy or coverup and there aren’t dead bodies hiding everywhere, the author was just wrong.

Here are the facts as we have them.

From https://www.worldometers.info/coronavirus/country/japan/ Japan admits to about 10,000 cases now, with a doubling time of about 11 days so they will admit to 20,000 cases by the end of the month.

If you read https://www.who.int/bulletin/online_first/20-254565.pdf you will find credible reasoning for 90% of COVID-19 cases not being tracked in official numbers, with a wide uncertainty. That correction would make it 200,000 cases in Japan by the end of the month. If you update that argument with the lower IFRs suggested in https://www.cebm.net/covid-19/global-covid-19-case-fatality-... then you get to a projection of 500,000 cases in Japan. (The admittedly problematic Stanford study came up with a suggestion of similar underreporting factors in both Santa Clara and Los Angeles.)

Alternately we could arrive at a similar conclusion if you assume that Japan is doing less testing than average. And there is considerable evidence that this is true.

We therefore do not have reason from the officially reported figures that he is wrong.

Now let'd deal with the dead bodies argument.

Deaths trail infections with roughly a 3 week lag. (A week to get sick, 2 weeks to die.) Let's go back to the simple model of 90% of infections are not reported and the true fatality rate is 1%. On March 31 there were 1700 official cases. On that model there were 17,000 infected of which 170 should die. Guess what? Today there are officially 263 dead bodies.

We. Don't. Even. Have. A. Shortage. Of. Known. Bodies.

The stated conclusions only look absurd to anyone who HASN'T tried to understand the dynamics of exponential growth, and the problems with the official data. He might or might not have been correct. But directionally he was clearly a lot more right than the official government position at the time.

Re: An update on a pre-registered result about the coronavirus

#116

Post by gregmac » Tue, Apr 21, 2020, 7:07 PM UTC

> A long time ago I did a bit of work on disaster alert systems. They’re not dissimilar to fire alarms. The engineered purpose of a fire alarm is not merely to let people know there is a fire. Many will have already perceived the fire. The alarm, buttressing training delivered far before the alarm rings, gives you unquestionable and immediate permission to evacuate. We know that otherwise some people, smelling smoke and feeling uneasy, would look around the room, see other people not moving, and conclude “Who am I, to disrupt everything going on by shouting ‘Fire! Fire! Fire!’?” The history of humanity has seen far too many rooms where no one shouted ‘Fire!’ early enough.

This alone is a pretty interesting point. Can apply to many other systems that are much less significant than literally "life or death."

Re: An update on a pre-registered result about the coronavirus

#117

Post by throwawayiouyt » Tue, Apr 21, 2020, 7:17 PM UTC

This thread is full of people who, never having been immigrants in societies hostile to immigration, dismiss the tone of the article as ridiculous.

Those people are simply wrong. Speaking up in places where you are not lucky to have that societal right has consequences. Some of those are legal, but some are not.

Criticizing the author for the politeness and secrecy of all things is a statement of immense privilege. Criticizing falsifiability, on the other hand... One usually expects more precise estimations for a hashed prediction.

Re: An update on a pre-registered result about the coronavirus

#118

Post by robocat » Tue, Apr 21, 2020, 7:20 PM UTC

BerislavLopac wrote: post #89

> There are few conspiracies in the world. There are many systems with complicated decision-making processes, internal data flows, and incentive structures for actors within them. The second sentence here is absolutely correct, but it directly contradicts the first one. Because a "conspiracy" is nothing more than having a group of those actors, with aligned incentives, coordinating to work together while hiding that…

I disagree. The second sentence is about uncoordinated actors making decisions, without anyone conspiring together. The emergent behaviour may look like a conspiracy, which is what he is pointing out.

Re: An update on a pre-registered result about the coronavirus

#119

Post by throwawayiouyt » Tue, Apr 21, 2020, 7:22 PM UTC

gregmac wrote: post #116

> A long time ago I did a bit of work on disaster alert systems. They’re not dissimilar to fire alarms. The engineered purpose of a fire alarm is not merely to let people know there is a fire. Many will have already perceived the fire. The alarm, buttressing training delivered far before the alarm rings, gives you unquestionable and immediate permission to evacuate. We know that otherwise some people, smelling smoke…

This novella was worth reading just for this gem. It is equally applicable in "boiling frog" scenarios as in the "fire" cases.

Re: An update on a pre-registered result about the coronavirus

#120

Post by Klonoar » Tue, Apr 21, 2020, 7:26 PM UTC

lalaland1125 wrote: post #45

This article frankly sounds like the rambling of a crazy person with all of the attempts of secrecy and defensiveness. > A distant secondary consideration, but a real one, was that choosing to publish could bring down heavy sanctions, through predictable pathways. Why is the author afraid of "sanctions" in a first world democracy with strong freedom of speech protections? A much more interesting and useful article wo…

I used to work in the Japanese journalism scene, on the tech side. My cofounder at the time was a veteran journalist.

He (and a litany of his peers) would agree that Japan is not how you would describe it, and I'm inclined to agree with them after building a product in that space.