31 ms·
Problems with Japan's Covid-19 reports
- cmehdy 7y agoWhile things like masks and social distancing are more likely to be already in place in Japanese culture, there definitely seems to be a bit of a surprise regarding actual numbers vs. means put in place to prevent the spread. The reports seem to reflect the opacity of communication of Japan's situation about the Covid-19. It's good that the author put out actionable suggestions at the end, but I'm wondering if the opacity is a result of internal issues or intended opacity. Makes me glad to see Canada's stats[1] being so clear in comparison. [1] https://www.canada.ca/en/public-health/services/diseases/2019-novel-coronavirus-infection.html https://www.canada.ca/en/public-health/services/diseases/201...
- jesuschroist 7y agoThe strategy here is clearly fighting Covid-19 with 'Security through obscurity'.
- btilly 7y agoI am extremely suspicious that Japan's numbers look good because they have not been testing enough. Given that honesty in testing consistently precedes the eventual slowdown in deaths by several weeks, I strongly suspect that Japan is maliciously incompetent and a lot of dead Japanese will be the result.
- standardUser 7y ago" I strongly suspect that Japan is maliciously incompetent and a lot of dead Japanese will be the result." When? Japan has had this coronavirus a lot longer than most places. If we expected the transmission rate to explode, we would have expected it to happen already, and we would be seeing the consequences.
- btilly 7y agoAre you sure we aren't seeing consequences? https://www.vox.com/covid-19-coronavirus-explainers/2020/3/28/21196382/japan-coronavirus-cases-covid-19-deaths-quarantine https://www.vox.com/covid-19-coronavirus-explainers/2020/3/2... shows that I am hardly alone in my suspicion that the numbers look good because reality is not being tracked.
- hilbertseries 7y agoIt looks Tokyo is starting to have an out break https://www.chicagotribune.com/coronavirus/ct-nw-tokyo-olympics-coronavirus-20200330-fwlj7xw7rfb6jfkjh5iafplzsq-story.html https://www.chicagotribune.com/coronavirus/ct-nw-tokyo-olymp...
- btilly 7y agoThe most interesting point in that article is that Japanese citizens and politicians are suspicious of the timing. All numbers looked good until the Olympics were called off. Then Japan started reporting climbing cases and talking about emergency. I certainly hope that the politicians in Parliament who asked questions got satisfactory answers. I doubt that they did, but there should be time to ask those same questions later when there is a clearer picture. I just hope it isn't as bad as I fear it will be.
- unishark 7y ago> All numbers looked good until the Olympics were called off. Then Japan started reporting climbing cases and talking about emergency. Typical causation versus correlation problem. Exponentially increasing pandemic outside Japan can also be the cause for everything: the olympics cancellation, the increasing govt concern and testing, and the sudden surge of cases within Japan from travelers. No conspiracies needed.
- btilly 7y agoFirst of all, lack of testing and government denials that there is a problem here been a hallmark of the response worldwide. It is hardly a conspiracy theory to say that it happened in Japan as well. Secondly for a couple of weeks now I've been running across articles of the form, "Japan should have had a problem by now, why are they an exception?" Which is more likely? That they were an exception until now, or that the Olympics provided extra motivation to stonewall? It is clearly not proof. But extra motivation to fail to try to detect their own problem combined with reason to doubt their version of reality (in the same way that happened over and over again elsewhere) makes my estimate of the conditional probability very high that their government was simply a bit more deliberately blind than elsewhere. And, as with everywhere else, the cost will be dead citizens. Only the further you let the problem go and the older your population, the worse it is. Japan has the highest concentration of truly old in the world. And may have let it go longer than anywhere else. The result? The final body toll won't be pretty.
- l_davis 7y agoJapan's numbers are probably low because of limited testing. Still somewhat reasonable to compare against countries that also have limited testing. Regarding the presentation of the data, I've seem enough poorly formatted presentations over the years to say that is unique to Japan.
- rwmj 7y agoSome anecdata here: My friends run a mid-sized hair salon, about 8 workstations, on the west side of Tokyo [edit: Kichijoji] and are still operating. They see no reason to close down, haven't been told to close down by anyone, and although their business is down, it's still brisk. Meanwhile the shops are full of both people and stock (except for loo roll). Some young people in Tokyo have been returning by shinkansen to their parents' homes in the countryside (which is obviously a dreadfully bad idea). I persuaded my friends not to do that, but they were thinking about it.
- asiachick 7y agoTokyo was mostly closed this weekend. IIRC by government request. TV shows showed all of Ginza closed. I walked down to Shibuya on Saturday, had one errand to run, it was mostly empty. Every chain store I passed was closed with a notice in the window. All the stores on the 4th floor of Mark City were closed. The basement food stores were open but the number of people were ~10th normal. I agree though that until recently things were pretty open. Maybe they were open again today. I walked by a coffee shop and people were in it. The local Target like store was busy but main street in front of it was dead compared to normal.
- rwmj 7y agoTheir salon is in Kichijoji which is not really central, as in well west along the Chuo line. I updated the original comment above. I'm interested if you know more about this absurd-sounding Aso/wagyu beef tokens story? It's not reported at all in the western media.
- efreak 7y ago> I'm interested if you know more about this absurd-sounding Aso/wagyu beef tokens story? It's not reported at all in the western media. This sounds strange enough for me to want to know more, and I tried looking it up, but only found weird porn when searching without quotes, and weird blockchain stuff from China when searching with quotes. Do you have any links I can run through a translator?
- phoenixstrike 7y agoPopular opinion in the US regarding the integrity of various countries' coronavirus reporting is hilarious. Whatever China says must ALWAYS be a lie, on the other hand, the US numbers are not "suppressed," it's just incompetence. Meanwhile, nobody is batting an eye at Japan's ridiculously and obviously doctored numbers. Hey guys, maybe China was just having literally the same organizational problems the US is having now leading to underreporting of numbers, back in Jan-Feb when every commenter on reddit and HN were saying China was covering it up. Meanwhile people just assume Japan's numbers must be honest because... why? Anyone who has the slightest awareness of how Japan works knows that this is par for the course. And no, it's not just the government brainwashing the people. The people are complacent in this and defend their government by rejecting criticism and continuing to vote them in office.
- LorenPechtel 7y agoI think this is a case of someone simply scanning various reports into the website--doing it the easiest way possible, not the right way.
- noobermin 7y agoSeriously, why all the skepticism regarding this? I understand it but it's strange to me. Hanlon's razor seems like an easier explanation, why does there need to be a nefarious plot under every bed.
- guardiangod 7y agoKnowing Japan, I bet some poor freshly graduated but elite sod in the health ministry was, out of the blue, told by his boss "here's a bunch of reports from different prefectures. Make a report out of them." So he goes off and try to make sense of the data, which is why you see those magic and unexplained equations in the report. The kicker is the most complex computer thing he ever did in his life was that Word presentation in university. HTML? Isn't that something you write by hand?
- reaperducer 7y agoHTML? Isn't that something you write by hand? In a good number of very large (Fortune 100) companies, yes. Even in large tech companies. Just as mechanical lathes and nail guns didn't put carpenters out of business, there are plenty of web sites that benefit from the attention of human hands, not just trusting the latest fad framework to concatenate several thousand dependencies pulled from random GitHub repositories.
- downerending 7y agoHaving looked at a lot of machine-generated HTML, "by hand" is still looking pretty good.
- astrobe_ 7y agoIncidentally, I always though that in animes it is often a big deal for a character to catch "a cold" because plot; unless their cold are actually more aggressive than western colds (or flues)? I wonder more and more which one it is.
- krackers 7y ago>to catch "a cold" Not to mention that mere exposure to inclement weather is enough to result in its onset, germ theory be damned
- kenhwang 7y agoMy mom is responsible for updating all of Los Angeles' numbers. It's literally "someone sent an email with updated new numbers, put them on the government website/in the database". There's no "making sense of the data" phase, just copy, paste, and hope whoever delivered those numbers did it correctly. She described the situation "all hands on deck, extra shifts, and losing staff to sickness daily"; the last thing they care about is some number on a screen.
- reustle 7y agoThe team behind https://covid19japan.com https://covid19japan.com fights with this nonsense every day. We're open ears to improvements, and are playing with ways to possibly automate bits. Our GitHub: https://github.com/reustle/covid19japan/ https://github.com/reustle/covid19japan/
- robinduckett 7y agoI wish there was something like this for the UK!
- bshimmin 7y agoThis isn't quite as pretty but is fairly good: https://www.arcgis.com/apps/opsdashboard/index.html#/f94c3c90da5b4e9f9a0b19484dd4bb14 https://www.arcgis.com/apps/opsdashboard/index.html#/f94c3c9...
- Angostura 7y agoMobile version: https://www.arcgis.com/apps/opsdashboard/index.html#/ae5dda8f86814ae99dde905d2a9070ae https://www.arcgis.com/apps/opsdashboard/index.html#/ae5dda8...
- xrd 7y agoThis is fantastic, terrific way to visualize this information. It's interesting that Hokkaido is red for infections but the growth rate is really slow. Are the distant prefectures less likely to follow guidelines from Tokyo and reporting more honestly?
- panny 7y ago>It's interesting that Hokkaido is red for infections but the growth rate is really slow. Hokkaido is about the size of South Carolina and approximate population of 6 million people. ~2.5M of which live in Sapporo. Most of the island is more rural with many tourist destinations. It's a popular destination for Chinese tourists in winter. >Are the distant prefectures less likely to follow guidelines from Tokyo Tokyo was seen as mostly following Hokkaido governor's lead early on. Hokkaido was the first to close schools, etc. >and reporting more honestly? Could HN dwellers stop implying Japan is handling this dishonestly? That is extremely rude coming from the nation whose CDC was angry to accept 14 infected citizens from the only ally nation that would accept their US Diamond Princess cruise ship and provide care for passengers.
- dataatemytime 7y agoHas anyone applied Benford's law to the raw data? https://en.wikipedia.org/wiki/Benford's_law https://en.wikipedia.org/wiki/Benford's_law
- kevinventullo 7y agoI think this could make an interesting post!
- rurban 7y agoLooks mostly right but he is wrong about the Diamond Princess. Those people all went off home, which is not Japan. Mostly western states. So they don't count to the Japanese numbers, only did in the very beginning. The rest looks indeed cooked, on purpose. Pasting images instead of tables is the usual trick.
- longtom 7y agoHe is also wrong about there having been no follow-up. Here it is, though only about a smaller fraction of the passengers. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3551331 https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3551331
- yufeng66 7y agoNever attribute to malice that which can be adequately explained by stupidity.
- Loughla 7y agoI believe that is an accurate statement, only when applied to the actions of one individual. When speaking of the actions of corporations, governments, or groups of people working in tandem in any fashion, malice needs to be your first assumption. Stupidity can be dealt with after you fix the problem, but malice will not only make the problem worse, but will also make dealing with the problem more difficult.
- objektif 7y agoOh man, china would like to have a word with you about that statement.
- dang 7y agoNo nationalistic flamebait on HN, please. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
- DangitBobby 7y agoI don't think this sentiment is so wise as its popularity would lead you to believe. I have no reason to believe that most bad things that people do happen entirely by accident. Further, I suspect that the first line of defense for anyone who's been caught doing something naughty is to play dumb, and I see no reason to make it easier to get away with murder, so to speak.
- lainga 7y agoThe reason why I used a two day old report on March 29 is simple, there was no full report yesterday nor today. I'm guessing the Tokyo lockdown means that nobody can get to work, and PCs are too expensive to buy for home use so the people at the Ministry of Health can enjoy a nice weekend with hoarded pasta. How's corporate Japan as a whole doing with remote work?
- svara 7y agoAt the end of the day, hiding exponential growth of this is impossible. That's what really bugs me about those who love to point fingers at how this or that country is counting things wrong (often revealing their biases in the process). Maybe every country has a random constant factor between actual cases and reported cases. But it doesn't matter - if it goes exponential, we'll know.
- gwern 7y agoYes, obviously once it goes exponential, exponentials are hard to hide. The point is to stop it before it goes exponential. That's why people were discussing, say, Iran or NYC testing problems and lack of social distancing, rather than saying "well, at the end of the day, hiding exponential growth is impossible, so who cares"...
- hujun 7y agonow days hacker news is often filled non-tech related discussion, specially anything relate to China or politics; come on guys, this is Hacker news, not political news; if you are so passion about politics, go to some forum dedicate for those things;
- dang 7y agoThis isn't so different from how HN has always been. Many topics have political overlap. That's inevitably the case on a site that's dedicated to intellectual curiosity (https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html). This is a challenge for moderation, but the approach we take has been stable for a long time. I've written about it here: https://news.ycombinator.com/item?id=21607844 https://news.ycombinator.com/item?id=21607844, and there's lots more explanation at https://hn.algolia.com/?dateRange=all&page=0&prefix=true&query=by%3Adang%20political%20overlap&sort=byDate&type=comment https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que... for anyone who wants it. Concerns about HN getting more/too political go back almost as far as the site itself: https://news.ycombinator.com/item?id=17014869 https://news.ycombinator.com/item?id=17014869. There's no systemic breach, just topical fluctuation. How we handle major ongoing stories (like this crisis) is mostly by downweighting routine follow-ups, so that just the stories with significant new information get primary coverage: https://hn.algolia.com/?dateRange=all&page=0&prefix=false&query=by%3Adang%20major%20ongoing%20significant%20new&sort=byDate&type=comment https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu.... That's the intention, at least. Of course there are always exceptions for submissions that have intellectually curious angles, like the detective aspect in the OP here.
- ronsor 7y agoStatistics is a science
- exolymph 7y agoToday you learn: https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
- nemetroid 7y agoMy hypothesis is that the "with no symptoms" and "with symptoms" items actually do include deaths. > Remember to not include death, because this is a quasi-hierarchy. The table tempts you to, but it's excluded. (The quasi-nature is because it is under the PCR tested positive umbrella on level 2.) So what about integrating level 3? > That's simple. You don't. Because no matter how hard you try, the numbers won't add up. (e.g. Give it a try - you'll end up with 129 != 131 and 1147 != 1191. We'll need one of these numbers later.) Level 4 adds up though, so the plot thickens. Note, however, that (131 - 129) + (1191 - 1147) = 46, i.e. the number of deaths. So if you adjust the 131 and 1191 figures to match their subdivision, everything would sum up correctly, including the top level (pcr_tested_positive), which would then include deaths. This would imply that 2 people without symptoms died, and 44 people with symptoms did. Taking a guess, "deaths" used to be two separate columns: one under "with no symptoms" (2 deaths), and one under "with symptoms" (44 deaths). It was then decided that "deaths" should be its own top-level item, and someone decided that this was the best way to achieve that goal. I can't find a way to make sense of the daily increases though, so this might be entirely wrong.
- icoxfog417 7y agoThe official format is terrible but you can get CSV data from here https://github.com/kaz-ogiwara/covid19/blob/master/README.en.md https://github.com/kaz-ogiwara/covid19/blob/master/README.en....
- noobermin 7y agoI know, data science (was it called that?)...didn't exist until tableau
- zadkey 7y agoI've been trying to access this website for the last 2 days, it is either down or overwhelmed.