Template talk:COVID-19 pandemic data/Germany medical cases chart

Format
Time should usually be on the x-axis. Is this the normal format? Walter Görlitz (talk) 19:35, 11 March 2020 (UTC)
 * If showing the total and the % change is intended, this is the only format that looks acceptable as time goes on. If just the bars are kept, then the chart could be easily compressed and putting the time on the x-axis would be no problem. Alexiscoutinho (talk) 00:35, 12 March 2020 (UTC)

WikiProject COVID-19
I've created WikiProject COVID-19 as a temporary or permanent WikiProject and invite editors to use this space for discussing ways to improve coverage of the ongoing 2019–20 coronavirus pandemic. Please bring your ideas to the project/talk page. Stay safe, -- Another Believer ( Talk ) 18:13, 15 March 2020 (UTC)

fishy data.
currently 14655 is given as RKI sourced number of infected for Germany

RKI website currently show 10999 infected accounted for in Germany ( link in data ).

I've seen the same overstatement a couple of times now. Allways fixed in spurts and bouts over a couple of days until they "stay"

Any commentary? ZwergAlw (talk) 18:04, 19 March 2020 (UTC)
 * This problem derives from the fact that the official RKI numbers are lagging behind the official counts of local cases by local authorities by about a day and a half. The newspaper Berliner Morgenpost regularly screens the official numbers of local authorities and adds them up to give an up-to-date number of the officially registered cases in Germany, whereas RKI waits for the local authorities to report their numbers on an electronic reporting sstem, which they don't do immediately.
 * I don't care much about whether we use the less up-to-date official RKI count or the more up-to-date unofficial sum of official local counts in this graph, but the important thing is to maintain internal consistency and not to mix the two different counts. Marcos (talk) 19:02, 19 March 2020 (UTC)
 * This is all and well but if the table says source RKI it must be RKI data and not some other unreferenced source.
 * If the intention is to show more up to date data ( questionable if they really were more up to date: they should reflect into the RKI numbers with a delay. they don't though afaics.) the source statement must reflect that. ZwergAlw (talk) 19:13, 19 March 2020 (UTC)

Recovery figures
What is the source for the recovery figures? The RKI source given doesn't list any recovery figures. —Mahāgaja · talk 20:26, 21 March 2020 (UTC)
 * The number of 2809 recoveries has been announced at today's press conference of RKI (https://www.youtube.com/watch?v=QUxBikdLrHo&t=8m4s, in german language); for further reliable german sources that mention this number, see e.g. https://www.tagesschau.de/inland/coronavirus-deutschland-rki-103.html, https://www.bz-berlin.de/deutschland/schon-mehr-als-2800-nach-corona-infektion-wieder-gesund, https://www.spiegel.de/wissenschaft/medizin/coronavirus-robert-koch-institut-fordert-konsequente-umsetzung-der-schutzmassnahmen-a-762be8f7-28c1-4996-86f9-2ab316f4ce51 or https://www.sueddeutsche.de/panorama/coronavirus-news-deutschland-1.4828033. This has also been mentioned in some (possibly unreliable) english language sources: http://www.china.org.cn/world/Off_the_Wire/2020-03/23/content_75850047.htm, https://www.aa.com.tr/en/europe/germanys-coronavirus-death-toll-rises-to-111-/1776126. The textual source at https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus/Fallzahlen.html however does not seem to mention recoveries. The sharp increase in the number of recoveries is due to the fact that RKI has usually not reported on these numbers, until today. Best, Schluppo (talk) 14:44, 23 March 2020 (UTC)

Dates of RKI figures need adjustment
RKI data is collected until 00:00 and then released in the morning. This actually means, the data officially released e.g. March 20, 00:00 covers all related information sent to RKI until end of March 19. So to associate the figures to the end date of the time range it is actually covering, we should distract one day from all entries. This might not look very nice, but that's what RKI data actually says. -- Käptn Weltall (talk) 07:47, 26 March 2020 (UTC)

Excessive data on the chart
The change in deaths should either be absolute or relative as having both bloats the chart (this is not a spreadsheet) and breaks the lines on mobile. I'll wait until a decision is made to convert the chart to the newest standard ('data' parameter). Alexiscoutinho (talk) 02:16, 27 March 2020 (UTC)
 * My vote goes for active cases (=total cases - deaths - recoveries) with percentage and deaths with the difference. Reasoning: active cases are the basis for possible deaths in the near future as well as for new infections as well as the work load of the health care system. Also active cases is what the orange bar actually depicts. And for deaths the absolute difference is a good measure against the active cases to have an estimate of the effectivity of the health care system. -- Käptn Weltall (talk) 09:29, 28 March 2020 (UTC)
 * I'd suggest we do the respective changes from tomorrow morning on (March 29, 2020), if no concurring proposals come up inbetween. -- Käptn Weltall (talk) 09:38, 28 March 2020 (UTC)

RKI is no more a trusted source
RKI is definitely slow to update data and/or it provides incorrect data. Even the John Hopkins University started to ignore the RKI data and it takes the data from other sources, as you can see from the well known map. I propose to change the data with the dataset of the John Hopking dataset, since RKI is no more a trusted source, as also pointed out by the three German newspapers that started to count on their own. --151.20.234.78 (talk) 22:01, 30 March 2020 (UTC)
 * I wouldn't judge RKI as unreliable, but as - sometimes much - delayed. And Johns Hopkins just changed their recoveries figures to the RKI estimates, while total cases have been taken from state, county or even local health administration pages since long. After all this is the same data, that finally finds its way to RKI, too. RKI still seems to be very much behind in automation. If delays get worse, I'd also vote for a switch, but IMHO we should keep with RKI for the next days. -- Käptn Weltall (talk) 08:21, 31 March 2020 (UTC)
 * I agree with Käptn Weltall and wouldn't judge RKI as unreliable. More importantly, the entire graph is based on RKI numbers and rectifying this in now will be hard to justify - i.e. are there any sources 'reliable' sources having a consistent timeline? if the source is updated, i think also all previous numbers need be updated and changed according to this alternative source. Lukas6283 (talk) 08:48, 1 April 2020 (UTC)
 * Johns Hopkins has the timeline. Click on Germany in the left table, and in the box on the bottom right you'll have the timeline graph with respective figures on mouse-over (hover). I'd like to stick with official RKI, as mentioned. But even government run news like "tagesschau" lately cite Johns Hopkins figures. So if RKI keeps on publishing figures 1000s of cases behind, we can switch. Let's see... -- Käptn Weltall (talk) 13:31, 1 April 2020 (UTC)

Daily snapshots or day of report
We currently collect the cumulated figures from RKI on a daily basis as published. But actually RKI (on their "dashboard", bottom right, tab "Fälle kumuliert") provides a history of cumulated cases with respect to the date they were actually reported to the health authorities. This is certainly a more accurate view of the development as at least no submission delays push in. Also, as I posted before, these figures are associated to the actual day of incoming reports, not to the date of the next morning when RKI publishes them, so for today, April 2, the latest figure is for April 1. But If we want to take over this data it would mean we have to adjust, or at least check, the whole history evry day. Other pro's and con's? Request for comment... -- Käptn Weltall (talk) 06:13, 2 April 2020 (UTC)
 * Hi, on a related question, the legend to the table marks the orange part as "active cases" but the the numerical data it is based on are just cumulative cases (adding the newly infected people but the deaths and recoveries are not subtracted). Or am I wrong? If not, it would perhaps be better to fix that first bcs the graphic part of the template is now quit misleading. Thanks for a response, WikiHannibal (talk) 09:19, 2 April 2020 (UTC)
 * The template calculates the active casses from overall cases, recoveries and deaths automatically. The respective figure is shown on mouse over. So if you put your mousepointer over today's orange bar, you'll get 53950 active cases, which is 73522-18700-872 calculated from the data we provided in the template for April 2. -- Käptn Weltall (talk) 09:48, 2 April 2020 (UTC)
 * Thanks. ;-) WikiHannibal (talk) 11:17, 2 April 2020 (UTC)

Slightly broken layout
The recently added note breaks the layout (number of case and plus from the percentage get on top of each other). My attempt at shifting the information to restore the layout was met with a revert. Can someone propose a better fix? David Prévot (talk) 19:17, 9 April 2020 (UTC)
 * If someone finds us a better workaround, ok. I tried to give an explanation when reverting: a recoveries related note is misleading inside cases figures. And you cannot place a note directly behind the date or bar. At least not without changing the underlying cases template. Maybe we place all notes between bar an figures to avoid a messy appearance? I found notes placement inside the parenthesis also a bit awkward. The clean solution would be to add an optional notes column at the end to the underlying template. Anyone volunteering for that task?? -- Käptn Weltall (talk) 21:13, 9 April 2020 (UTC)

Timeline is wrong by 1 day
I think the timeline is wrong by one day. Each day the Robert Koch Institute publishes an update, but these only include figures for the previous day. For example I'm writing this at 11:30 UT (12:30 BST) on 10th April, and the RKI site says that there have been 2373 deaths, but RKI says "Case numbers in Germany as of April 10, 2020, 00:00 a.m." In other words, 2373 is the cumulative death toll up to and including 9th April, not 10th April.

This would be consistent with the approach adopted on, for example, the page about the pandemic in the UK; each day's official update only includes data from the previous day.

I'd rather get this sanity-checked by other contributors before making an edit. Do people agree with my interpretation?

Stewart Robertson (talk) 11:31, 10 April 2020 (UTC)
 * I tried to address this already, see above. Moreover even RKI on their "dashboard" associate the latest cumulative cases count to the day before, see bottom right, tab "Fälle kumuliert" with mouse-over on the rightmost data point. -- Käptn Weltall (talk) 15:55, 11 April 2020 (UTC)

Percentage or new infections?
Why did the template change the daily changes from stating the percentage growth to the new infection number? I see other templates for other countries (e.g. US, France, Belgium) use the percentage. --Mahmudmasri (talk) 22:48, 22 November 2020 (UTC)

I agree that the percentage is more relevant and is used on all other countries I have checked (US, France, Denmark, Japan, Sweden, ...). — Preceding unsigned comment added by B340bf (talk • contribs) 08:09, 25 November 2020 (UTC)


 * Relevance is certainly not generated by what other wikipedias do. It's what's most commonly used in news or other publications. And there the wide majority refers to absolute changes, for the newly infected as well as for changes in the number of deceased. -- Kohraa Mondel (talk) 06:29, 26 November 2020 (UTC)


 * How to compare the rate of infections with absolute numbers? E.g. Germany with USA, 2020-11-24 to 2020-11-25.
 * New cases absolute:  GER=18.633 USA=175.631 (whaaaaaat!?)
 * New cases percentage: GER=2.0%  USA=1.4%    (ok for me)
 * I think it is much better to use percent view in this template to have comparability. ;) --It-franky (talk) 16:02, 26 November 2020 (UTC)


 * If you need comparability to other countries, the relation of new cases to cumulated cases won't help. There you go with new infections per inhabitants (usually per 7 days per 100.000). The percentage previously used in the chart can be a good measure of how quickly the disease spreads, that's why it was used in the beginning of the first outbreak in many publications. But once most of the cumulated cases are recovered, and thus not contagious anymore, this stops making any sense. You simply compare to a number that's no factor at all for the current epidemic development. And so the wide majority of publications stopped using that figure with the beginning of the second waves at the latest. Since months you'll see that at wikipedia almost exclusively, maybe because not so many people dig into the depth of templates. I can see your point and there are some possible changes, that would make sense, but percentage does not. And for the other options you'd need another or an augmented basic template. -- Kohraa Mondel (talk) 19:29, 26 November 2020 (UTC)
 * Thank you, and I concur. Percentage is only seemingly useful, but does not actually allow a comparison with anything, except maybe the last couple of days in the same region. Absolute numbers are much better. 2003:EC:AF05:AF00:B8BD:D717:6C96:E63E (talk) 07:14, 27 November 2020 (UTC)
 * Agree!!! Even added a population line into the main articles' infobox (there are „units“, e.g., with 1 million and 0.5 million of humans, yet the numbers of infections/deaths are almost the same; makes one wonder, why?) (basing on everyday “counting” of 36 countries' numbers, currently; 'cause, template can be programmed to count per cents of us, humans (e.g. X of "7,800,000,000 people equals to Y ) &#32;— Pietadè (talk) 21:01, 26 November 2020 (UTC)
 * In other words: per cents of parents, children, brothers, sisters, et cetera... &#32;— Pietadè (talk) 09:15, 27 November 2020 (UTC)

I have a question to those who keep changing the graph back to percentages. Please go here: https://en.wikipedia.org/wiki/COVID-19_pandemic_by_country_and_territory and then click on any country- or all the countries. You will see that the article on the pandemic in Germany is a singularity: it is the only one that uses absolute numbers instead of percentages. Now, the meaning of the percentage metric can be discussed. However, wikipedia should present data in these articles in a uniform way. In other words: if you think that absolute numbers are more appropriate (to which I disagree), then edit _every_ article this way. Here comes the curious point: a brief look at the GERMAN wikipedia shows that there, absolute numbers are used throughout (in a very different plot template). This fuels the suspicion that it is some Germaniacs that are selectively editing the article on the pandemic in Germany... guy and gals, this is NOT the German wikipedia. Stay at home. — Preceding unsigned comment added by 86.52.82.109 (talk) 00:44, 29 November 2020 (UTC)
 * I changed the percentages to absolute with lots of reasoning here as well as on the change comments, as you can easily see in the article history. And I did never any changes on the german version or took part in discussions there. So much for your suspected german conspiracy theory. The change was done for good reason, so maybe you just take the time to read this little section here and make your point. -- Kohraa Mondel (talk) 06:12, 29 November 2020 (UTC)

Of course I read this "discussion" with "lots of reasoning". However, for now I'm not even discussing the reasoning. You still haven't answered why the Germany article is the only one that's using absolute numbers. If your reasoning is correct, why are you not changing the articles about the pandemic in other countries? — Preceding unsigned comment added by 86.52.82.109 (talk) 10:40, 29 November 2020 (UTC) May I add that your profile clearly states that German is your mother tongue. I have to assume that you are predominantly consuming German media- and there is a clear preference for absolute numbers. But as I already stated: this is the international wikipedia. There are two options: either you make your change to all articles and see what the reaction of the general wikipedia user base is. Or you stop changing the present article to conventions you are accustomed to. I'm not interested in an edit war- please state what your course of action will be. — Preceding unsigned comment added by 86.52.82.109 (talk) 23:27, 29 November 2020 (UTC)

I strongly agree that it is very odd (to say the least) that this page for Germany suddenly shows the absolute and number and not the change in percent. It is at complete odds with all the other country tables I have seen. Moreover, pandemic has a lot to do with growth rate. — Preceding unsigned comment added by B340bf (talk • contribs) 09:00, 30 November 2020 (UTC)
 * I frequently go through press headlines and articles in googles news for some other countries. My impression there is all the same: Percentage of new cases compared to cumulated is completely out of use. Even absolute numbers of cumulated cases you'll find less often than daily new (absolute). So maybe you actually try with taking part in reasoning instead of posting unfounded accusations ad hominem. And you're welcome to do changes on other pages if you're annoyed by missing uniformity in the daily changes metrics, even though comparability - as discussed above - is no real issue. -- Kohraa Mondel (talk) 09:59, 30 November 2020 (UTC)
 * Agreed. Growth rate relative to total cases since start of the pandemic was in use during the first wave, but is only misleading now. Growth relative to yesterdays values would be somewhat useful, better yet averaged over 7 days. 2003:EC:AF05:AF00:292C:5626:8865:5A79 (talk) 07:56, 1 December 2020 (UTC)

I will not make changes to the high number of other country-specific Covid-19 articles, since I do not think these are a "contribution". If you think these changes are justified (and I do not agree), go ahead and make them consistently. If you will not make them, then I will revert the article on Germany back to the current wikipedia standard, percentages. May I note that if an individual wants to change the current wikipedia standard, then the discussion section of one specific article might not be the right place to decide this, since you are not reaching the relevant authors. — Preceding unsigned comment added by 86.52.82.109 (talk) 10:39, 30 November 2020 (UTC)
 * Yes, this is not the right place to discuss metrics of roundabout 200 articles. But it's you who does it. Ok, I can see your point (which is also mine): What talk/page to discuss, then? Talking about standards: I'd consider the module Medical cases chart as a standard, and this module explicitly offers the absolute difference metrics as well as percentage. So I do not see the necessity of all articles using the module to use the exact same metrics. I actually do think that for most covid charts the absolute difference would be the much better choice (see aboove), but it's this article where I took part in maintenance and so I did changes I thought appropriate here. -- Kohraa Mondel (talk) 13:29, 30 November 2020 (UTC)

In other words: your changes are and will remain Germany-specific. In this case I will revert (and will keep reverting) the plot to the current wikipedia standard. By the way: a roundabout has nothing to do with all of this. — Preceding unsigned comment added by 86.52.82.109 (talk) 22:10, 1 December 2020 (UTC)
 * FYI: There is no such standard, see Project COVID-19, Current consensus and COVID-19 pandemic, Current consensus. If you think, all country articles should use percentage, place a respective section there, but pls stop your edit war here. Actually I already opened a section on that topic there myself, so of cause you're also invited to take part there (Project COVID-19, Medical cases charts - change type) instead of opening a concurrent section. -- Kohraa Mondel (talk) 14:21, 2 December 2020 (UTC)

I saw that edit summary there inviting me to vote, so therefore I'd vote for the percentage for consistency.

✅ to percentage --Mahmudmasri (talk) 03:27, 7 December 2020 (UTC)


 * If you had read the paragraph right before yours, you might have noticed that there is an ongoing project-level discussion. Please read and discuss there, and also vote there whenever that discussion calls for final opinion. As some here were of the opinion, that all countries charts should use the same option, there is the place to discuss, not here anymore. -- Kohraa Mondel (talk) 13:44, 7 December 2020 (UTC)

Request for comment - cases change metrics
Please take part in discussion here: Project COVID-19, Medical cases charts - change type — Preceding unsigned comment added by Kohraa Mondel (talk • contribs) 22:44, 5 December 2020 (UTC)

Use 7-days incidence changetype?
I made 7-days incidence by date of publication available. For germany medical cases it could look like this. In my opinion the best solution, compared to absolute and percentage. It would also be much better for comparability between countries, as it neither takes historic cases into account, nor neglects the population count. Also such incidence is widely used in countries' health administrations and other publications. -- Kohraa Mondel (talk) 17:53, 31 October 2021 (UTC)


 * Ok, no one seems to care too much. As in previous discussions this seemed to be the best compromise, rather than absolute numbers or percentage, I switch now, where this option is available from the module. -- Kohraa Mondel (talk) 19:01, 7 November 2021 (UTC)
 * That is total nonsense. Nobody cares about incidence over a week. The absolute numbers are the point, so everybody can see the actual development and compare it to others dates. — Preceding unsigned comment added by 87.185.45.235 (talk) 19:26, 7 November 2021 (UTC)
 * I can't see the point of comparing figures to other states, and comparing to previous dates, daily absolute numbers are seriously biased by the day of the week. All this was discussed before, and we only chose absolute numbers as superior to percentage lacking other alternatives provided by the module. See discussion in chapter "Percentage or new infections?". -- Kohraa Mondel (talk) 05:34, 8 November 2021 (UTC)

Sticking numbers
Hello. The numbers have reached over tens of millions and therefore they appear very sticky to the bracketed numbers.


 * The very sticky German chart right now: [//ibb.co/CmYbcGh]
 * With a monospaced typeface appear on top of each other: [//ibb.co/t85fhJt]
 * The American chart as a comparison. Notice the breathing room between numbers: [//ibb.co/LgJF2ZV]
 * With a monospaced typeface, still well spaced: [//ibb.co/5MCbLpC]

I tried using the phrase  as indicated in the parent template, but that failed to make a difference. --Mahmudmasri (talk) 10:57, 11 February 2022 (UTC)
 * «|numwidth=xxww» did the job..., see template for explanations (there is a table below — ‘Optimal numwidth ranges’) ☆☆☆— PietadèTalk 17:18, 11 February 2022 (UTC)

Inconsistency between tabulated data and graphed data
The graphical display shows deaths tapering off to near zero at the end of April, while the tabulated results continue to show up to 300 deaths per day. Which is correct? — Preceding unsigned comment added by 108.52.207.45 (talk) 11:11, 27 April 2022 (UTC)


 * It is now near the end of July, and the inconsistency between the graphical display and the tabulated results has not been addressed. The graph indicates no day with more than 100 deaths since early May, while the tabulated data continue to show at least one day a week with deaths higher than 100. 98.115.160.165 (talk) 04:50, 24 July 2022 (UTC)