Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
WHO: Novel Coronavirus (Covid-19) Situation Dashboard (arcgis.com)
30 points by pintxo on March 12, 2020 | hide | past | favorite | 35 comments


Not to start a political conversation, but why are Taiwan, Hong Kong and Macau not listed as separate countries? All 3 have their own measurements and cases, surely that shouldn't be included with data from China...

Not only that, it gives a false perception of what's going on in those areas. Macau only has 10 confirmed cases, with everyone recovered. Hong Kong and Taiwan both have less than 200 cases, definitely a much different situation than in Mainland China.

EDIT: I see they have a separate option to show by province / autonomous region, specifically for China.


It's just sad that when we point out the obvious, the pro-dictatorship camp just dismisses it by saying "don't politicize science". But this data is in fact the result of politicized science.

By bootlicking China, WHO is just expensing her own credibility. Many no longer follow the authority of WHO. For example, many countries started restricting travel from China[0] to give themselves more time to prepare when WHO opposed it[1]. Experts had long questioned WHO's decision to refuse to declare it a pandemic[2], before WHO's late announcement yesterday.

There are some efforts to stop the ripple of this China political power play, particularly by Taiwanese and Hong Kongers[3]. We should all be more alert.

[0] https://fortune.com/2020/02/06/countries-china-travel-restri...

[1] page 5 of https://www.who.int/docs/default-source/coronaviruse/transcr...

[2] https://time.com/5798797/un-who-coronavirus-outbreak-classif...

[3] https://github.com/CSSEGISandData/COVID-19/pull/418


Almost every other data source lists Macau, Taiwan and Hong Kong as separate countries, and for good reason.

China, Hong Kong, Macau and Taiwan all publish their own individual statistics, why would the WHO aggregate them into one.

> https://github.com/CSSEGISandData/COVID-19/pull/418 I hate to disagree, but they should be listed as separate countries.


yeah it's especially idiotic when a freaking cruise ship gets its own entry but an independent country-island doesn't.


And bootlicking certainly puts you at hazard of obtaining the coronavírus.

Sorry, I know jokes are frowned upon here.


Doesn't seem unreasonable; the "W" in WHO stands for World. They aren't here to report on city states and islands. Even if those city states and islands are well run politically independent entities.


Well, they listed a cruise ship. Singapore, Bahrain, San Marino(!), occupied Palestinian territories, Liechtenstein, etc are listed too and they re all tiny islands or city-states.

The only reason WHO is not listing HK and TW separately is to avoid displeasing China.

They should be listed separately as in the fight against the virus they act as independent entities using their own resources and local policies. These actions have no relation to the political claims made by China over these territories. China is making everything into a political issue, so


They made choices for reasons. Borders aren't a universal truth, countries disagree on these matters and the WHO has to work with that because they are a global entity.

It is obvious why it happened, but it is nevertheless perfectly reasonable. If anything the Europeans are the problem; resisting being aggregated into sensible units. But nobody begrudges them that.

> They should be listed separately as in the fight against the virus they act as independent entities using their own resources and local policies.

Don't the Scots have their own health system? Maybe Scotland should be listed separately from England?

The situation reports list Occupied Palestinian territory separately; they probably aren't going it alone. Maybe the WHO should have merged them?

There is no end to the nitpicking that is possible. There isn't anything particularly important in the aggregations they are using.


While they're at it they can fold UK's numbers into France and New Zealand's cases into Australia. Cyprus's numbers can be divided equally and assigned to Turkey and Greece.


Doesn't that make sense? Realistically having separate numbers for each European country is excessive. Anyone who cares what is happening in somewhere like, eg, Latvia can go look up the Latvian public health page.

Latvia is smaller than most Chinese provinces. It isn't a useful amount of data at the global scale. The UK is smaller than the major Chinese provinces.

It doesn't make any sense to give the UK as much paper space as India, for example. Taiwan, HK and Macau are not where the injustice is here. India is nearly 20% of all people.


Olympics is about world too.


Can we discuss the impact of visualization parameters on perception?

On first look it seems Europe is way worse than China regarding case numbers. But actually checking the scale for the bubbles shows that they are categorical and the largest bucket is ">12,000 cases" thereby putting China (80k) and Italy (12k) into the same visual bucket.

Why would you do this?


Two possibilities:

- the goal isn't to convey information, it's to appear to convey information so they can defend themselves later on

- we tend to hire artists who are concerned with appearance into roles that require visual engineers

The former seems to be something that the UN struggles with. For instance, they can't indicate how many people have coronavirus in Taiwan because they need to consider the feelings of their member states. And they can call one recent coronavirus "Middle East respiratory syndrome", whereas another has to be called "SARS-COV-2" to avoid racist implications of a geographic name.

The latter is a widespread problem. For instance, once upon a time, software design gave consideration to making it easy to approach and use, especially by people who might be using computers for the first time. Nowadays, it's primarily about making sure the software looks fashionable.

My trust for international organisations has gone down the loo thanks to this outbreak. I am sad.


This article has some decent thoughts on the quality of data-viz, which has been pretty shocking [1]. ArcGis seems to have pretty bad scaling generally

https://www.futurehealth.live/blog/2020/2/10/coronavirus-dat...



Good question. At least the bubbles aren't red...


This version shows cases by state in the US:

https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.h...


I'd recommend everyone use this one over the parent articles.

1. This one is supposedly maintained by John Hopkins. I'd expect at least some level of quality control.

2. The parent link seems to be a copy of JHU's dashboard.

3. The parent link does not break out smaller regions (mainly the US).


I have found the worldometer page easier to use than the others : https://www.worldometers.info/coronavirus/


What I don't like about that page is the way they hide the active cases graph, and when you show it it's tiny. I would like a graph that shows the active cases as its most important figure. Total cases is interesting from an immunity standpoint, but active cases is important from a "will i catch it" standpoint.


I'd prefer a graph that shows exactly how many tests are done by each country per day. All other extrapolation hinges on that. If the data is not available in a country (e.g. US), then everything for that country should come with a warning that says "we have no clue what is happening because we don't test"



My dashboard aggregate data from Worldometer and some people prefer a simple view: https://coronavirus.gg

I want to add deaths per country as well and maybe other ideas you guys have that would not clutter the view too much...


Amusing technical flaw: If you select the dot representing US cases, then click the plus symbol, it zooms straight to a field in Kansas, and makes it appear as though all cases are in that field.



I heard an interesting comment on the Recovered figure, this doesn't mean the patient has recovered and can go back to their life. I heard it just means they don't have a fever and have passed two tests for not having the virus. They might still be fighting for their life on a ventilator.

Is this true?


At least for China, the requirement is no fever for 10 days and two negative swabs. Only then they can get discharged. I'm not a doctor but I'd be surprised if you still need oxygen by then. You probably won't run a marathon either.


They won't be fighting for their life, but they might have long-term damage that requires continuous care.


It’s highly improbable. Medical recovering I assume it also includes being discharged from the hospital.


If you're interested just in the numbers with the option to show the development over time with logarithmic axes, i found this site quite interesting:

https://coronavirus.arik.io


They're showing 987 for USA. Surely, that's a bit outdated?


Johns Hopkins Hospital version seems more up to date

https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.h...


Same for other countries - this seems to be lagging.


Yes, it's been lagging for days. Worldometers is up to date.


This is old data, Iceland is already up to 90 cases




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: