Template talk:COVID-19 pandemic data/United States medical cases/Archive 2

Archive 1

Archive 2

Retroactive cases in California

According to this source https://globalnews.ca/news/6856758/coronavirus-california-early-deaths, 2 deaths from COVID were confirmed to have happened on Feb 6 and Feb 17 in Santa Clara County.

also, the Feb 26 deaths from Covid aren't included in the table from Seattle either 135.19.2.130 (talk) 17:42, 24 April 2020 (UTC)

 Not done: it's not clear what changes you want to be made. Please mention the specific changes in a "change X to Y" format and provide a reliable source if appropriate. Aasim 10:17, 25 April 2020 (UTC)

Retroactive Deaths in California - Thanks for updating cases table, but the Deaths table still has not been updated to reflect deaths on Feb 6 and 17 in California. The associated Wikipedia tables of cases/deaths "by state and by date" provide an extremely important historical record. The fact that the first COVID-19 deaths occurred Feb 6 and 17 in California is not reported in the Wikipedia Deaths table. Below are references:
- https://www.latimes.com/california/story/2020-04-23/morgues-hold-key-to-early-spread-of-coronavirus
- https://www.nytimes.com/2020/04/22/us/coronavirus-first-united-states-death.html?searchResultPosition=1
- https://www.cnn.com/2020/04/22/us/california-deaths-earliest-in-us/index.html
- https://fortune.com/2020/04/22/new-coronavirus-deaths-california-us-covid-19-victims-earlier/
- https://www.cdc.gov/nchs/nvss/vsrr/covid19/index.htm

Formalizing this request in "Change X to Y" format...
Please change the table "Deaths from non-repatriated COVID-19 cases in the US by state" to show that 2 deaths occurred on Feb 6 and 17 in the California deaths column:
- Add row for Feb 6 to Deaths table, and add 1 death in the California column
- Add row for Feb 17 to Deaths table, and add 1 death in the California column
Including this historical information in the Wikipedia COVID-19 Deaths table is extremely important because it shows the first COVID-19 deaths occurred on Feb 6 and 17 in California.

Thanks for following update to the cases table: change the table Cases from "Non-repatriated COVID-19 cases in the US by state" to show that 2 deaths occurred in the Deaths New/Cml columns. To be clear, please add rows for Feb 6 and Feb 17 as follows:
- Add row for Feb 6, and add 1 death in the rightmost Deaths New/Cml columns
- Add row for Feb 17,and add 1 death in the rightmost Deaths New/Cml columns
Morebits (talk) 01:33, 3 May 2020 (UTC)
Morebits (talk) 13:00, 7 May 2020 (UTC)

 Not done: From a quick overview it does look like most of the references used on this template are being sourced from public health government databases. Secondly, when you make an edit request on a semi-protect page be sure to format it in a "change X to Y" format with easy to find sources. Now, its not enough to just list random sources and make editors dig through them to see which one sources which statement. What we're looking for is: "change California death count on Feb 6 to 1, see New York Times citation. I'll reopen your edit request, make sure you re-word and re-source your request. Thank you for your patience. Don Spencer^{talk-to-me⛅} 17:34, 9 May 2020 (UTC)

@Don Spencer - thanks, below is updated request to correct issues with following Wikipedia page:
https://en.wikipedia.org/wiki/Template:COVID-19_pandemic_data/United_States_medical_cases

Summary - the Wikipedia page above contains two tables:
1) Cases by State by Date.
2) Deaths by State by Date.

For the past few weeks multiple requests have been made to retroactively correct the fact that Cases and Deaths occurred in California on Feb 6 and 17. These retroactive cases/deaths reported by public health departments have been widely reported by various sources including the CDC. Despite multiple attempts to get these tables corrected two issues keep re-occurring:

1) CASES Table - When rows are added to the Cases Table for Feb 6 and 17 for cases in the California column, as was done a few days ago, the Cases table later regresses back so these rows no longer appear (this situation occurred again on 5/8 at 10:15 pm ET). Why are these rows being deleted after people add these rows? Why do people keep saying this issue is fixed when it is not?

2) DEATHS Table - The Deaths table is not being updated with rows showing that the first deaths occurred in California on Feb 6 and 17? This issue is important because those deaths were the first certified deaths to occur in California and the US, and these occurred more than three weeks earlier than is currently incorrectly shown in the Wikipedia Cases Deaths tables.

Formalizing this request into "Change X to Y" format where X=column and Y=row:

1) Change Cases table "Non-repatriated COVID-19 cases in the US by state" as follows:
1.1) Change California Case Count to 1 on Feb 6 per references below:
1.2) Change California Case Count to 1 on Feb 17 per references below:
- https://www.nytimes.com/2020/04/22/us/coronavirus-first-united-states-death.html?searchResultPosition=1
- https://www.cnn.com/2020/04/22/us/california-deaths-earliest-in-us/index.html
- https://www.cdc.gov/nchs/nvss/vsrr/covid19/index.htm

2) Change table "Deaths from non-repatriated COVID-19 cases in the US by state" as follows:
2.1) Change California Death Count to 1 on Feb 6 per references below:
2.2) Change California Death Count to 1 on Feb 17 per references below:
- https://www.nytimes.com/2020/04/22/us/coronavirus-first-united-states-death.html?searchResultPosition=1
- https://www.cnn.com/2020/04/22/us/california-deaths-earliest-in-us/index.html
- https://www.cdc.gov/nchs/nvss/vsrr/covid19/index.htm

Correcting this historical information in the Wikipedia COVID-19 Cases and Deaths tables is important because it shows COVID-19 deaths occurred on Feb 6 and 17 in California. Thanks Wikipedia team for your persistence correcting this important information in these two Wikipedia tables.
Morebits (talk) 01:33, 3 May 2020 (UTC)
Morebits (talk) 13:00, 7 May 2020 (UTC)
Morebits (talk) 18:55, 10 May 2020 (UTC)

 Question: Is any body who is more comfortable with this template going to update this? @Poklane:@Bz8x8c: Since you two seem to be doing the lion share's of the work. I see no obvious fault with this edit request. Thanks, RandomCanadian (talk | contribs) 03:55, 17 May 2020 (UTC)

personally I don't really care anyway Poklane 13:31, 17 May 2020 (UTC)

Why is this case and death information being suppressed? Morebits (talk) 22:54, 17 May 2020 (UTC)

I don't think anyone is suppressing information. I wouldn't have any problem with adding these two rows with their respective cases, but note that the total cases for California is the sum of all the daily death quantities in the table for California. That sum matches the number of deaths currently being reported by the state. If someone adds these two cases on their respective dates, then two cases will need to be removed somewhere else in the table. If you know when these two cases were "acknowledged" by the state of California authorities, then maybe those two cases could be removed from the count for that date. However, for every state we simply record the quantity added to the total on each day in that day's cell. There is no effort to reflect the counts based on the date the death occurred, as that would be very tedious to maintain as states constantly change their records over time. Bz8x8c (talk) 15:11, 20 May 2020 (UTC)

Thanks @Bz8x8c and @RandomCanadian - @Morebits is in communication with California Department of Public Health (CDPH) regarding how/when these deaths/cases were accounted. Morebits (talk) 19:13, 20 May 2020 (UTC)

@Morebits - I found a news article from April 23 that appears to be when these early deaths were first made public as a result of autopsy findings. However, this news article notes that Governor Newsom has directed coroners to look at other deaths as far back as December. I mention this because there's more involved in this edit than adding a couple rows and changing a value or two from late April. In fact, every row between the added rows and the row when the two cases are subtracted will need the cumulative total updated as well, in both the cases and deaths table. By my calculations, that would likely involve editing the cumulative deaths value in over 100 rows. While that can be done with a bit of work, and in this case might be justified, it's worth stating that it's not trivial. I would not want to have to do it more than once, so if there's any chance California is going to end up identifying even earlier deaths, it might be a good idea to wait until the additional investigation is complete before putting in that amount of work. A note could be added to that effect to insure nobody thinks the information is being suppressed, but the editing effort involved means it should only be done one time! Bz8x8c (talk) 03:58, 21 May 2020 (UTC)

 Not done for now: please establish a consensus for this alteration before using the {{edit semi-protected}} template. This edit request is very long and is turning into a discussion. A consensus should therefore be sought before the edit is made. — Tartan357  ^(Talk) 14:15, 25 May 2020 (UTC)

Nevada and Iowa Updates

I just reverted an edit by Scholaroh to daily death counts for both Nevada and Iowa, where it appears the historical daily death quantities were edited to match the quantities reported by date on the source sites.

We do not attempt to maintain the death counts by date because states change the counts by date to reflect when the deaths actually occurred. It would be a nightmare to try to update the historical deaths by date and maintain the daily total and cumulative quantities by date. Instead, we simply add the change in the total for each date as they are reported.

I also see that both Nevada and Iowa have updated their total deaths since the quantities were retrieved for the May 27 update (yesterday). On Iowa, it seems they are updating the reported totals throughout the day instead of once a day.

Do we need to have a discussion to make sure we're applying a consistent approach for states like these two? Bz8x8c (talk) 11:57, 28 May 2020 (UTC)

Confirmed Cases in Pennsylvania

Starting April 22, 2020, the Pennsylvania DOH site began showing both Confirmed and Probable cases and deaths. Should the data be corrected? Covidtracking has the screen shot every day. — Preceding unsigned comment added by Gongminmin (talk • contribs) 17:52, 6 June 2020 (UTC)

Will do it soon. Poklane 18:28, 9 June 2020 (UTC)

Update: I've not done it yet, but unfortunately I can only go as far back as May 22nd as some days before that only show the combined confirmed and probably cases on Covidtracking. Poklane 18:53, 9 June 2020 (UTC)

Thanks! My bad, I want to say May but somehow typed April. Yes, it's starting from May 22. — Preceding unsigned comment added by Gongminmin (talk • contribs) 21:20, 9 June 2020 (UTC)

Probable Cases

I just want to clarify that Arizona has been reporting probable cases in the total case count. Today (June 6th), Arizona reported having 25167 confirmed cases, not 25451. It's not much of a difference but I figured the template should at least stay consistent if it's reporting on confirmed cases.

And on a side note, I want to let you know that Michigan is also including probable cases as well. https://www.michigan.gov/coronavirus/0,9753,7-406-98163_98173---,00.html — Preceding unsigned comment added by 68.55.55.59 (talk) 21:31, 6 June 2020 (UTC)

Also, Delaware has separated confirmed cases and total cases since June 11th. — Preceding unsigned comment added by Gongminmin (talk • contribs) 16:36, 11 June 2020 (UTC)

Starting from June 3rd, West Virginia doh's website can click the total number to see the confirmed cases and probable cases. — Preceding unsigned comment added by Gongminmin (talk • contribs) 16:43, 11 June 2020 (UTC)

Puerto Rico also has a separated confirmed cases and total cases. And the gap is huge... — Preceding unsigned comment added by Gongminmin (talk • contribs) 17:10, 11 June 2020 (UTC)

Idaho's confirmed cases can be found in the detailed view: [1]

Negative Case Numbers

Are the negative case numbers due to states updating their counts and us not knowing which day to subtract them from? Is it likely that these negative numbers will be redistributed to their appropriate dates and we'll get more accurate day tracking? The negatives are making for very ... interesting plots using this data. 38.108.53.234 (talk) 03:50, 6 May 2020 (UTC)

Yes, states will sometimes lower their death count based on revised accounting. And some of those states might never indicate which days those negative numbers apply to. For plots, I use 7-day averages exclusively, not just because of these occasional numbers but also because many states are highly cyclic and/or erratic in their reporting. Tuesday is often a big day as they catch up from the weekend. Mark Taylor (talk) 16:09, 7 May 2020 (UTC)

That is what I figured to be the case. I also use 7 day averaging since daily numbers have a lot of variation -- like you said weekends make a big difference (as do occasional non-reporting days). Usually a 7 day moving average cleans the data enough to take care of days with negative values. Even so, the negative numbers have recently led to some plots that are by definition not functions (specifically with Maine). My graphs can handle it, and if it is the best data we have then it will have to do. 38.108.53.234 (talk) 01:03, 8 May 2020 (UTC)

Does it make sense to add a 7-Day Average column at the right between the New and Cumulative columns? Thisisnotatest (talk) 08:11, 25 May 2020 (UTC)

What's the deal with PA on May 22? Totally throwing off my graphs. Any way we can smooth out the negative counts over a week or so? --SavageCabbageCowboy (talk) 14:41, 10 June 2020 (UTC)

WTF Puerto Rico on June 11?? Anyone have an explanation? --SavageCabbageCowboy (talk) 13:10, 12 June 2020 (UTC)

Need Google Data Help - Missing visual data.

When you search on google "US coronavirus map" It brings up a blue graph with daily new cases in the US. Since its inception months ago it has actually provided a number you can see when you mouse over a data point. So on the high days its usually 25,000 cases...and low days its maybe 17,000.

Well today that ability to scroll over and see a number disappeared....Oddly on the same day where the data shot up fast over that 25,000...A number which was reasonably consistent for a couple of months

At the bottom of the graph it looks like the data is from Wikipedia...But is that mouse-over data also from Wikipedia? Or is the whole graph google generated? Either way....how do we get this fixed? The graph becomes pretty useless without the scroll over to give the "real" number.

thanks in advance. — Preceding unsigned comment added by Paul10583 (talk • contribs) 20:08, 20 June 2020 (UTC) Paul10583 (talk) 20:16, 20 June 2020 (UTC)Paul10583

I’m not sure if posting here or pestering google did it....but the problem has been fixed. Thanks to any and all who helped. Paul10583 (talk) 20:06, 21 June 2020 (UTC)

Something is wrong with PR's data

it doesn't add up to the data on the PR page itself. And one glaring error is the -3,926 on June 11th. Where does this data get sourced from? By a bot, or by hand? — Preceding unsigned comment added by Douglaswyatt (talk • contribs) 19:03, 22 June 2020 (UTC)

Cumulative Deaths Short by 65

The Total New Deaths value is 65 deaths short on April 2nd (seemingly West Virginia's 65 deaths were missed).

Add 65 deaths to the Total New Deaths value for April 2nd, from 1075 to 1140, and increase all Total Cumulative Deaths values from April 2nd to present by 65 deaths. Koszuta (talk) 05:12, 19 June 2020 (UTC)

 Not done: please provide reliable sources that support the change you want to be made. Darren-M talk 22:57, 19 June 2020 (UTC)

There's no change in the raw data, the New Deaths sum is simply incorrect for April 2nd. Koszuta (talk) 01:25, 20 June 2020 (UTC)

I have confirmed that indeed the daily deaths total is incorrect for April 2nd, based on the existing daily death counts in the table for each state/territory on that date. However, the problem is that the daily death value of 65 for WV on 4/2/2020 should not even be in the table. I will remove the "65" from the WV deaths cell on 4/2/2020, which will resolve the problem. Bz8x8c (talk) 16:53, 27 June 2020 (UTC)

California additional cases

Yesterday's cases weren't updated on California's COVID-19 dashboard, but when I watch TV about the California's COVID-19 update by the governor Gavin Newsom, there were 5,898 confirmed cases yesterday.
Then the California Department of Public Health released the stats, which yesterday had 5,898 cases and an additional 3,842 cases received since February, totaling it to the July 1 row to 9,740. I don't why 3,842 cases were added. Evan0512 (talk) 04:22, 2 July 2020 (UTC)

Update: Instead of 9,780 cases, the dashboard said 9,480 cases, and then the next day (July 1), there were 4,056 cases and 73 deaths. The death toll on June 30 remains at 110. Evan0512 (talk) 18:08, 2 July 2020 (UTC)

Semi-protected edit request on 2 July 2020

I want to update the data (number of new daily cases and deaths) using the official websites from each state linked as the sources for each state's tally. Currently, there is a delay from when each state reports to when the page is updated. I will reduce that delay. JemezHound (talk) 22:27, 2 July 2020 (UTC)

 Not done: this is not the right page to request additional user rights. You may reopen this request with the specific changes to be made and someone will add them for you, or if you have an account, you can wait until you are autoconfirmed and edit the page yourself. Eggishorn (talk) (contrib) 00:48, 3 July 2020 (UTC)

Colorado updates off one day from July 4th onward

Starting on July 4th, all Colorado updates have been input incorrectly. They have been input too early and are this one day off. The Colorado numbers are updates at 4pm mountain time, 6PM eastern time — Preceding unsigned comment added by 97.118.137.5 (talk) 20:19, 7 July 2020 (UTC)

Texas death total for July 27 is wrong. Is 75, not 675, per their DOH's page.

Texas death total for July 27 is wrong. Is 75, not 675, per their DOH's page. — Preceding unsigned comment added by 67.77.82.160 (talk) 13:08, 28 July 2020 (UTC)

TXDoH counted 631 previously unreported fatalities on Jul-27. Thosbsamsgom (talk) 13:38, 28 July 2020 (UTC)

Urgently needed review of this template

I find it confusing and rather contradictory that for some states, we're including both confirmed and probable cases/deaths, but for some others we are not. I would love to go back and change the data so that the criterion is the same for all states (confirmed only, or confirmed+probable). Unfortunately, given the size of this table, I cannot edit it with VisualEditor, which makes editing the whole table a rather daunting task.

DE, MS, ID are three of the many states for which we're counting both confirmed and probable cases, whereas for other states we're counting only confirmed.

It would be awesome if someone could go back and change the data but first there must be an agreement on how we should tally it up. --Spaastm (talk) 20:01, 23 August 2020 (UTC)

I'd be happy to help. Which *cells and by what amount? Thosbsamsgom (talk) 04:40, 24 August 2020 (UTC)

AR's data is in confirmed cases before, but (wrongly) uses total cases in Sep 12. — Preceding unsigned comment added by 73.59.70.78 (talk) 17:55, 13 September 2020 (UTC)

Semi-protected edit request on 15 September 2020

93 Eaglejfm (talk) 02:37, 15 September 2020 (UTC)

 Not done: it's not clear what changes you want to be made. Please mention the specific changes in a "change X to Y" format and provide a reliable source if appropriate. JTP ^{(talk • contribs)} 06:48, 15 September 2020 (UTC)

South Carolina

There is significant discrepancy between SC DHEC reporting and the number reported in Wiki, particularly starting from September 11: Wiki reports almost 3 times as many cases, and from Wiki it makes to Google Coronavirus data. If any alternative source is used, how come it’s no referenced? Ikhulor (talk) 17:22, 19 September 2020 (UTC)

Semi-protected edit request on 25 September 2020

September 22 for Texas is showing 17,820 new cases. the Texas Dashboard shows only 3,964. The Google statistic page uses this information. so the information is now skued. cite "https://txdshs.maps.arcgis.com/apps/opsdashboard/index.html#/ed483ecd702b4298ab01e8b9cafc8b83" Southtexasattorney (talk) 03:08, 25 September 2020 (UTC)

TxDSHS frequently adds delayed cases from labs and other sources to the statewide total, but not to the daily confirmed cases for any specific county. If you visit the state's dashboard, check the panel at the top right. There will be a note explaining the older cases included in today's statewide total. Thosbsamsgom (talk) 22:22, 30 September 2020 (UTC)

Takes 10 seconds to load this template

Once I open this page, it takes ten seconds rather than 3 because the data size of this template exceed 500,000 bytes. — Preceding unsigned comment added by Evan0512 (talk • contribs) 19:50, 9 August 2020 (UTC)

Why not reply to this now? It takes 10 seconds to load due to the large page size (500KB). It happened to me very frequently. Evan0512 (talk) 17:01, 5 October 2020 (UTC)

Illinois error on Nov 6?

I think there might be an erroneous leading 1 before the daily case totals for Illinois on November 6. This page lists 18,049 cases. Most other data sources report something much closer to 8049. Am I missing something? The error adds 10,000 cases. TAGLASS716 (talk) 20:04, 7 November 2020 (UTC)

Following up, I had not yet checked the Illinois Dept. of public health website. It turns out that the number reported for IL for 11/6 does match what is posted on the IDPH website (18,049): https://www.dph.illinois.gov/covid19 There is a footnote under the daily cases graph that reads:

Beginning November 6, 2020 and going forward, IDPH will report confirmed cases and probable cases combined. Due to this change, all probable cases previously reported separately, 7,673, have been added to the one-day, November 6, 2020 total of new cases.

I should have caught this earlier. The TEMPLATE table is correct. TAGLASS716 (talk) 20:17, 7 November 2020 (UTC)

Wikipedia page delays, proposed to remove mouse overs and separate cases and deaths tables

I was looking at this page, shocked as of today. The page exceeds 500,000 bytes on November 6 and I propose to remove some mouseovers seen in the template data.

At the beginning of today 863 bytes were increased in this page. Yesterday 684 bytes were increased. It takes at least 10 seconds in almost every time I visited this page since October.

The increase is caused by the abbreviation of dates and states throughout the data both horizontally and vertically, such as {{abbr|Jan 1|January 1}} and <abbr title="x">XX</abbr>. These are examples.

At the end of the month another header was used, using the same one used as the last month's header, stating U.S. regions and states.

I'm going to propose separating cases and deaths to reduce the page size by half, so they won't experience interruptions in both data and data speed in this page.

Evan0512 (talk) 05:12, 9 November 2020 (UTC)

Either remove the abbr's or the wikilinks to the state pages for all but the most recent month. Thosbsamsgom (talk) 05:38, 9 November 2020 (UTC)

Confirmed cases in American Samoa

Three cases were confirmed in American Samoa on Nov 9. Source.

Should a new column be added for this territory? --Spaastm (talk) 20:27, 23 November 2020 (UTC)

nebraska

NE is not tallying correctly. I have it as 1,228, not 1,205 deaths2601:603:207F:C170:C5AA:E300:EF5:63C (talk) 17:59, 7 December 2020 (UTC)

I'm not sure where you're getting your Nebraska death data, but I pulled 1205 deaths from the Nebraska dashboard yesterday and the updated quantity of 1236 deaths from the dashboard today. If you're using a source that shows different values, then it's not sourced from Nebraska's official published data. Bz8x8c (talk) 05:03, 8 December 2020 (UTC)

Percentages in the Bar Graphs

I really do think percentages for how much the total confirmed cases and deaths increased featured in the bar graphs should include tenths for more consistency. Hotspots like New York and New Jersey include them but not Michigan nor California. Oh, and I do think all the graphs should track deaths as well — Preceding unsigned comment added by 68.55.55.59 (talk) 20:52, 3 April 2020 (UTC)

Incorrect Totals

The sum of total deaths on June 4 do not sum to 909. The sum equals 916. Can someone please update? — Preceding unsigned comment added by SavageCabbageCowboy (talk • contribs) 13:53, 6 June 2020 (UTC)

Will do it soon. Poklane 18:28, 9 June 2020 (UTC)

Hello. The sum of deaths on Nov 11 is incorrect. Can someone please update soon? Thank you. SavageCabbageCowboy (talk) 14:30, 17 November 2020 (UTC)

I also checked the total deaths for Nov 11 and SavageCabbage is correct. The total shown is 1462 but the data in that row sums to 1485 a difference of 23. Since there is no specific cell with 23 I don't know what went wrong. Please correct. TAGLASS716 (talk) 21:37, 28 November 2020 (UTC)

@SavageCabbageCowboy and TAGLASS716: I missed these comments about Nov 11 deaths until today, since they were posted in a much older section of the talk page. The error was due to an entry of 24 instead of 1 on Nebraska's Nov 11 daily deaths, which I found by comparing to my compiled historical spreadsheet data. I've fixed the entry error, and now the total of 1462 should be correct. I'm suspect I mis-entered the value because Nebraska is usually the last update of the day and I was perhaps getting tired! Anyway, thanks for bringing it to my attention, even though it took me a month to see it. Bz8x8c (talk) 16:41, 28 December 2020 (UTC)

Deaths row total incorrect for Jan 1.

Sometimes the row totals don't sum properly in these tables. Not sure why. The totals in my spreadsheet for Jan 1 are 2467 using the same row of data captured using cut and paste. That differs from the total in the table (2444) by 23. Most likely that's the Virgin Islands deaths (23) that are missing. Please adjust the formula in the table to capture the correct row total for deaths. TAGLASS716 (talk) 01:57, 3 January 2021 (UTC)

Thanks for bringing this to my attention. The deaths total was correct - I fixed the problem, which was an incorrect "23" in the VI deaths for January 1, 2021. It was not there until the changes done on the deaths table this morning. If you see any death rows that don't sum properly since earlier last year, it's likely a deaths value not entered correctly because the sum comes from a spreadsheet on my end that provides all of the data entered in the table. (There is no formula in the table - it's in my spreadsheet!) Bz8x8c (talk) 04:00, 3 January 2021 (UTC)

"Deaths" removed from second table headers

This edit appears to have accidentally removed "deaths" from the second table: https://en.wikipedia.org/w/index.php?title=Template:COVID-19_pandemic_data/United_States_medical_cases&diff=997857312&oldid=997831685 . In particular the table title no longer includes "deaths from" and the column headings have changed from "Deaths" to "Confirmed". 2601:1C2:1000:2:D264:FF73:E2D4:6FCE (talk) 13:17, 3 January 2021 (UTC)

Thanks for pointing out these inadvertent changes to the deaths table, which I have now reverted back to the original text. Bz8x8c (talk) 14:33, 3 January 2021 (UTC)

Dates without year will begin duplicating on Jan 21

Now that 2021 is upon us, dates cannot be assumed to fall within 2020. That is already causing trouble for formulas in a spreadsheet I maintain that uses this page as a source. It will get a bit worse on "Jan 21", when that date will appear twice. Would it be too disruptive to change to full dates like "21 Jan 2021"? MikeLaBonte (talk) 03:31, 2 January 2021 (UTC)

Thanks to @Thosbsamsgom for fixing this. — Preceding unsigned comment added by MikeLaBonte (talk • contribs) 12:24, 6 January 2021 (UTC)

Anytime.! Happy to help. Thosbsamsgom (talk) 04:38, 13 January 2021 (UTC)

When to "Heal" apparent bad data from a state?

@Evan0512, Poklane, Rider0101, Spaastm, and Thosbsamsgom:

I'm interested in any other viewpoints on a problem I encountered in data from Nebraska yesterday. Here's what I found and how I dealt with it...

Three days ago, Nebraska's dashboard showed 1703 deaths, resulting in a calculation of 11 new daily deaths.

Two days ago, Nebraska's dashboard showed 1811 deaths, resulting in a calculation of 108 new daily deaths, which seemed excessive but not completely out of the normal for various states from time to time. I thought maybe they had a dump of backlogged death certificates into their system.

Then yesterday, Nebraska's dashboard showed 1733 deaths, which was a sharp decrease from the previous day of -78 deaths. I confirmed the previous day's total deaths quantity on covidtracking.com, which also reflected the 1811 deaths. This confirmed that I didn't just misread the total deaths from Nebraska's dashboard.

Seeing this sharp decrease, I made a guess on what happened, suspecting that the previous day's value should have been 1711 and was incorrectly entered on the dashboard as 1811. So, I updated the previous day's death value on that assumption, making the daily deaths 8 instead of 108, and then added yesterday's total deaths of 1733, making the daily deaths calculation as 22 new daily deaths.

My question to everyone is does it make sense to assume an error and correct the data in Wikipedia, or should we not assume an error on the state's end and enter the data exactly as they release it? Bz8x8c (talk) 13:37, 9 January 2021 (UTC)

Use the most recent posted numbers and add an item to the footnote about the change. I suspect other states health officials have made similar changes to their previously posted data, too—which would be maddening if true. Good work spotting this discrepancy. !high five! Thosbsamsgom (talk) 15:50, 9 January 2021 (UTC)

I didn't get much response to my question, which I will take as some form of approval for my chosen action. However, I checked the data at covidtracking.com, and for the record they left the daily deaths for these two days as +108 and -78 instead of +8 and +22, which is precisely why I asked the question. Bz8x8c (talk) 03:12, 19 January 2021 (UTC)

I agree with Thosbsamsgom. Use the numbers from state health officers, even if they change what they post without explanation. Otherwise, we're taking "sides" and that's a slippery slope Fostrdv (talk) 12:27, 19 January 2021 (UTC)

Total cases for 4/4/2021 is incorrect

The total cases for 4/4/2021 is incorrect. The correct total should be 28,521,865.— Preceding unsigned comment added by 2601:240:4980:24e0:b553:aa32:fae6:c076 (talk) 20:57, 7 April 2021 (UTC)

The accumulated total cases for 4/4/2021 is shown as 28,526,077. This is incorrect. The correct total should be 28,521,865.— Preceding unsigned comment added by 2601:240:4980:24e0:4cdd:8a7a:6f97:f480 (talk) 23:27, 15 April 2021 (UTC)

The April 4 total cases is based on the sum of the state total cases as of April 4, with supporting data that was collected from all of the states on that date. If you think it's incorrect, can you explain why you believe it's wrong and explain your source for that information? Bz8x8c (talk) 02:17, 20 April 2021 (UTC)

Today I discovered that I had a typo in my data spreadsheet on the daily cases for two different states on two different dates. March 6 CA cases, which Evan0512 had fixed on March 7, and April 4 MD cases, which had not been caught by anyone yet. These entry errors affected my calculated total daily cases that were entered at the end of the daily row for these two dates but had no affect on any other totals. It turns out that the total error in these two daily case totals was the same as the difference noted in the unsigned comments above, although the total cases was actually correct. If the OP was summing the daily case totals, that would account for the claimed discrepancy. The daily cases for these two dates has now been fixed, so this issue should be resolved. Bz8x8c (talk) 00:41, 28 April 2021 (UTC)

Semi-protected edit request on 8 March 2021

The Confirmed Total cases for March 6, 2021 is incorrect. Change 27,047,306 to 27,043,306. 2603:7000:4502:FAF6:4DE9:BF3B:75AB:EBA5 (talk) 01:25, 8 March 2021 (UTC)

 Not done: please provide reliable sources that support the change you want to be made. Ididntknowausername (talk) 01:44, 8 March 2021 (UTC)

In the following talk page section, I explain an error I discovered in CA's case data entered for March 6 that resulted in the daily total cases being too low by 4,000 cases, which contributed to the issue with total cases raised in that talk page section. Looking back at this section, I realized this is the reason for the impression that the total cases was incorrect for March 6. It turns out the total cases is correct, but the daily cases was what was incorrect. Another user caught and corrected the error in CA's daily cases back on March 7, but the daily cases was still off by 4,000 until just now, as I just finished fixing it. This should resolve the original issue raised by the OP above, where they thought the total cases was off by 4,000. Bz8x8c (talk) 01:51, 28 April 2021 (UTC)

Visual Editor does not work on this page, splitting proposal

As of November 30, the visual editor doesn't work in this table. Once I click on the blue button, it loads, then it received an error message. The error message is nothing in there, which is challenging with 530K bytes.

Page size

The page size adds at least several hundred characters a day, and on the 1st day of each month a month header was added, increasing its page size to 8,800 bytes.

@Bz8x8c: I've heard that this page has a missing American Samoa case column. It takes up at least several thousand bytes more.

American Samoa does not have a reliable source releasing case data. I believe covidtracking.com relies on social media posts and local informants. Fostrdv (talk) 05:28, 2 December 2020 (UTC)

covidtracking.com currently shows no cases for American Samoa. The link above suggests the 3 cases were sailors, so they were not citizens of AS - these may not be counted as AS cases for that reason. I think we should hold off until some official source actually shows cases for AS. Bz8x8c (talk) 18:36, 2 December 2020 (UTC)

Daily sources

Daily sources had stopped updating its sources on April 1st. On March 6, these sources were more aggressive until on March 20 it decreases. It completely stopped and no new sources were updated for 8 months. It is worth about 217,112 bytes worth of daily sources references.

I agree, the daily sources should be redacted; however, a not so insignificant portion of that 217k was redundant white space and in-line formatting which has been removed. Thosbsamsgom (talk) 02:08, 2 December 2020 (UTC)

list of daily sources removed. the collapsible list template they were wrapped by was one of the reasons for the page load woes. Fostrdv (talk) 05:16, 2 December 2020 (UTC)

Deaths table

The deaths table is 122,237 bytes, and it increases every day. Separating the two tables make it load a bit easier. It takes 10 seconds to load this page, as I said a number of times.

Splitting proposal

@Bz8x8c, Rider0101, and Poklane: I propose that sections Deaths table section be split into a separate page called Template_talk:COVID-19_pandemic_data/United_States_deaths. The content of the current page seems it takes more than 530K bytes in two tables including daily sources and we need to separate two tables.Evan0512 (talk) 22:46, 1 December 2020 (UTC)

I get why you're proposing to split the two tables into two separate pages. FWIW, I quit using the Visual Editor very quickly because it was painfully slow even then when there was far less content. Now, I just edit the content within the normal "Edit" function, which I now have down to a science. I am not opposed to splitting the page in two, but I will voice my concern about the impact to my required editing effort to maintain the data on two pages instead of one page. That being said, if the decision is made to proceed with separating it into two pages, I'll just have to adjust and deal with it! I don't know anything about Daily sources, which I haven't touched since I started helping maintain the page, so I guess I have no problem with them being removed. I only use the source links at the bottom of each table to access the same source sites each day. As for a new column for American Samoa needing to be added, I wonder if there's an easier way to do that without editing every row in both tables to insert the column? Bz8x8c (talk) 23:08, 1 December 2020 (UTC)

I'm personally not entirely opposed to splitting, however since I've yet to experience any issues with the page as it is I would prefer keeping things at is. I do however would like to say that I think it would be a good idea to split the page into 2020 and 2021 pages a month from now. Poklane 00:17, 2 December 2020 (UTC)

Before splitting, remove all the redundant state and date abbr.'s, as well as redundant links to the various state and region pages first. That will significantly reduce the page loads. Thosbsamsgom (talk) 00:46, 2 December 2020 (UTC)

Splitting into a new calendar year makes sense to me - nice idea. However, I will need to trust someone else to make those changes after the December 31, 2020 updates (or before, I guess), since my wiki effort to date has been limited to editing existing pages. Since I've added the new monthly headers the last few months, I'll need to make sure I don't reintroduce any content removed by recent trimming edits. (I did remove the inline comment about Florida non-residents in my saved header template.) Bz8x8c (talk) 02:59, 2 December 2020 (UTC)

As a regular watcher of the page, I'd be for the positive cases table and the death count tables being split. It takes forever to load, anymore. Devilmanozzy (talk) 03:09, 2 December 2020 (UTC)

I haven't been able to use the Visual Editor since August or so. I used to actively contribute to this page but I've stopped since then. As mentioned above, I also managed to master editing using the normal "Edit" function but eventually it became too cumbersome. I am in favor of splitting the page in two. Spaastm (talk) 15:03, 3 December 2020 (UTC)

The problem is not the amount of data, at least that's my impression. Modest images are far larger and still load quickly. The problem appears to be how long it takes to render the data in HTML table format. Splitting the file only solves the problem temporarily. It won't be long before both of the new tables are too big again. The solution is to find a better way to represent that data so that it displays faster. [anonymous user] — Preceding unsigned comment added by 161.97.246.132 (talk) 21:21, 4 December 2020 (UTC)

• Page is painfully slow to load - and I've got high speed Internet! Someone be bold here ... probably need a cases and deaths page for each year. Nfitz (talk) 07:31, 11 December 2020 (UTC)

Using Chrome on a Chromebook, the page loads for me in ~3 seconds. How slow is "painfully slow" for you? Is it a lot longer than that? In an old browser version on an older OS I have, it takes much longer to load, so it's possible it depends on your browser/OS combo. Bz8x8c (talk) 23:22, 12 December 2020 (UTC)

7 or 8 seconds to load, using Windows 10 and Chrome, User:Bz8x8c. It's by far the worst Wikipedia page I load regularly - and yet my Internet connection is good enough to watch 4K television. And if anything, it's getting worse. Page is over 340 kB now! Nfitz (talk) 03:59, 21 May 2021 (UTC)

I suspect there is an open Template somewhere on the page (starting with {{ ) that is is not closed (a matching }} ). Thosbsamsgom (talk) 11:17, 21 May 2021 (UTC)

A text search says there's 244 of each - for what that's worth. Nfitz (talk) 15:20, 21 May 2021 (UTC)

Mine still loads on Chrome on a Chromebook in about 3 seconds. I'm not really familiar with the wiki technical aspects, as I just edit existing page content, but I'd be interested in whether wikimedia offers any diagnostic tools for analyzing page delivery issues. I don't have a clue how templates impact rendering or why you suspect an unclosed template, but I'd be glad to get input from anyone on how to diagnose or improve page delivery. Bz8x8c (talk) 20:31, 21 May 2021 (UTC)

@Evan0512, Poklane, Rider0101, Spaastm, and Thosbsamsgom: I just added the January, 1, 2021 rows to the existing page, along with the new monthly header for January 2021. As I stated in an earlier comment, I have no experience creating new article pages with necessary next and previous links. Can someone else take care of splitting this table into a new 2021 page and changing this page to specify it's the 2020 page, and add appropriate navigation between the two pages? If someone takes care of that change, I'll switch to updating the 2021 page with future edits. Until then, I'll just continue to add new data in this page. Any complaints or feedback? Bz8x8c (talk) 15:40, 1 January 2021 (UTC)

@Evan0512, Bz8x8c, Rider0101, Spaastm, and Thosbsamsgom: I think the best thing to do would be to move the current page to Template:COVID-19 pandemic data/United States medical cases/2020, and then make Template:COVID-19 pandemic data/United States medical cases/2021 the page for 2021. I'm also not too sure on how to do proper navigation, but I think simply adding "For 2021 data, see Template:COVID-19 pandemic data/United States medical cases/2021" at the top would be good enough for now.--Poklane 17:00, 1 January 2021 (UTC)

I disagree with such a hard split without more input from the editors who are using the data from this template/table as a source for other articles. Leave the tables in place for now, is my recommend. If we have no amicable solution by Jan 31/Feb 1. Then split. Thosbsamsgom (talk) 19:40, 1 January 2021 (UTC)

OK by me - I trust your judgement. I'll just continue editing this existing page until such time as someone decides to perform the split. Bz8x8c (talk) 20:23, 1 January 2021 (UTC)

Personally, I created a LibreOffice spreadsheet for my own use that's been importing this data every week or so, ever since it first became available several months ago. If it's split into separate pages by years I could adapt my spreadsheet easily enough, though of course there would be some effort. (On the positive side, if the 2020 data becomes frozen as its own page, that somewhat reduces the future archival risk of page vandalism and/or format changes. Also it would load faster, though the loading speed hasn't been a big problem for me so far.) If the years stay together, please start adding years to the dates on each line. I had to change some formulas today because LibreOffice suddenly decided that all dates without explicit years must be in 2021. Also the redundant monthly headers often complicate the calculations. But I can understand leaving those in place if it makes reading the data easier for site visitors. To all of you who have been collecting and entering this data so diligently for so long - thank you! Bouncey (talk) 01:48, 2 January 2021 (UTC)

As for whether to split the deaths from the rest: I mostly just use the combined national data. The state data would be much more useful if it included recoveries along with cases and deaths (or split recoveries into a separate page too). I don't know how common that is in the state-level sources though. I tried going directly to the source for one particular state, but found that they published in formats too complex for LibreOffice to import. Bouncey (talk) 02:04, 2 January 2021 (UTC)

@Thosbsamsgom: With respect to your changes this morning, I like that you moved the "Notes" out of the totals row and into their own row, which will ease some of the editing effort (The notes really complicate the task of updating the totals). However, some of the states had notes for just cases, notes for just deaths, and notes that applied to both sections. It appears you didn't create the same notes structure in the new notes content. Do we need to discuss that detail and decide how to approach that distinction between data sections' notes? Also, what was the reason for adding value formatting? That will require some additional effort in the value pasting part of the edit effort, as I currently just paste integer values into each cell. Bz8x8c (talk) 14:58, 2 January 2021 (UTC)

I am working on the bottom table ATM. Will post in next 30–45 minutes. Thosbsamsgom (talk) 15:07, 2 January 2021 (UTC)

Formatting the daily values helps with their readability. Otherwise, the table is a wall of incomprehensible 0s, 1s, 2s, etc. imho. Fostrdv (talk) 15:17, 2 January 2021 (UTC)

Lower table updated. And yes, agreed. Number formatting was for readability. However, that can be removed if need-be. Thosbsamsgom (talk) 15:58, 2 January 2021 (UTC)

well done with the edits.!! Fostrdv (talk) 16:59, 2 January 2021 (UTC)

@Thosbsamsgom: Thank you for your work cleaning up (and hopefully speeding up) these tables. Was it necessary to remove the state headers from the beginning of every other month? I have a massive Google Spreadsheet that pulls data from this page, and had compensated for all the monthly headers, so now everything is out of whack. (My personal preference would be just freezing the headers at the very top so they are always visible so monthly headers aren't neede, but I have no idea if that's possible on Wikipedia.) If I'm the only one scraping this data, I'm happy to adjust my spreadsheet, but I suspect I'm not the only one doing so, so I don't think we should be making changes to the positioning of data, like deleting headers and splitting the tables onto multiple pages, willy-nilly. Craig Butz (talk) 21:00, 2 January 2021 (UTC)

You can restore the monthly headers, yes? Nothing is set in stone on Wiki. I hope you understand. Fostrdv (talk) 04:24, 3 January 2021 (UTC)

@Thosbsamsgom: After entering one day's worth of data, I must say I like being able to paste a group of state totals into the table without having the notes get in the way. The only minor frustration I will voice is having the grand totals on the lower-right side appear below all the notes in the edit frame, since they appear to the right of the notes instead of the right of the state totals in the table! This means the final edit of the day requires me to scroll through all of the notes after entering state totals in order to update the grand totals. As I said, a minor frustration, but still an annoyance none the less. I'm also wondering if it's possible to stuff some hidden markers into the totals rows in order to identify the region groupings, since the notes actually served as 'markers' before to figure out where to enter totals values. Everything else was easy-peasy, nice and easy. Thanks for all your efforts! Bz8x8c (talk) 03:42, 3 January 2021 (UTC)

@Fostrdv: Thank you for taking the initiate to move the grand totals up a line last night. However, we need to think about the removal of the state labels at the bottom of the table. Right now, it's not a problem, but in two weeks I'm going to have a tough time finding the right source links for each state when the state labels are off the top of the screen. I'm trying to figure out where to reinsert them in a way that preserves the new position of the grand totals but restores my ability to locate specific states' refs once the top labels are no longer visible. Any suggestions are welcome. Bz8x8c (talk) 17:21, 3 January 2021 (UTC)

How about the last row? Fostrdv (talk) 05:52, 4 January 2021 (UTC)

Split by type and year proposal

Due to excessive length, I propose splitting Template:COVID-19 pandemic data/United States medical cases into the following Templates ordered by data and by year:

• Template:COVID-19 pandemic data/United States daily cases
• Template:COVID-19 pandemic data/United States daily deaths
• Template:COVID-19 pandemic data/United States daily cases in 2020
• Template:COVID-19 pandemic data/United States daily deaths in 2020

In all likelihood this data will continue to be complied well into 2022, and possibly into 2023. Therefore, templates may be further split by year, etc. I have boldly created two Templates for 2020 by truncating the existing tables and merging the Efn's. Many thanks extended to Evan0512 for initiating discussion to split this template in December 2020.

• Split Thosbsamsgom (talk) 04:11, 22 May 2021 (UTC)
• SPLIT!! Thank you, yes. Fostrdv (talk) 04:23, 22 May 2021 (UTC)
• split Agree. Wanderer0 (talk) 08:18, 22 May 2021 (UTC)
• Split way too long currently. Adoring nanny (talk) 18:28, 22 May 2021 (UTC)
• Split yes - has become exponentially slower in recent days. Nfitz (talk) 18:51, 22 May 2021 (UTC)
• OK with split I do think this will help with my browser resources, as repeated edits eventually force a browser close/re-open cycle! I'll just need to edit two pages instead of a single page, but I'm willing to do that for the sake of everyone else! Bz8x8c (talk) 19:52, 22 May 2021 (UTC)
• Split deaths from cases, but not years. I have an extensive analysis spreadsheet pulling data from this page. Every change to the format (including the banners proposing splitting the page) breaks the import. It seems reasonable to split cases from deaths, but splitting by time period just makes it harder to reaggregate the data. Can we please abide by the Zero one infinity rule and keep this data in one place? Craig Butz (talk) 06:43, 23 May 2021 (UTC)
• Split, as proposed Wiki is for humans. Humans have complained page load has been a problem for weeks/months. Excel spreadsheets are not humans, yet. Rdzogschen (talk) 11:02, 23 May 2021 (UTC)
• Split It should have been done at the start of the year, but whatever. Rube Dali, the DodoHorse (talk) 04:39, 23 May 2021 (UTC)
• (add your vote/comment, please.)

The fans on my computer have spun up when rendering this page. So yes, is very much needed. Great job with the bold-ly-go work you've already done. Fostrdv (talk) 04:27, 22 May 2021 (UTC)

I was content to leave the page as is, but I looked at today's update using a ten year old mac. The fans went warp speed as the page overflowed and each column came into existence from the cyber-ether. Welp.! Might as well with the split, I say. ;) Thosbsamsgom (talk) 04:30, 22 May 2021 (UTC)

Excellent work User:Thosbsamsgom! Opens cleanly and quickly now!! Nfitz (talk) 16:06, 13 June 2021 (UTC)

Much appreciated. Next step is renaming this template, which I will do in a week or so. Give other editors time to discover the template has been split, etc. Thosbsamsgom (talk) 20:22, 13 June 2021 (UTC)

State Data Publication Changes

Back on June 5th, Florida stopped publishing daily data updates, instead switching to a weekly data report every Friday. This skews the Friday totals quite a bit due to Florida's volume of COVID cases and deaths, but at least we're still getting data from Florida. In addition, a large number of states have stopped publishing weekend updates, resulting in a significant jump on Monday case and death values.

However, Nebraska stopped publishing any data at all on July 1st, removing any trace of their dashboard from the web. I sent an email to the Department of Health, asking if there was any alternative source for their ongoing case and death data, including any potential for just receiving the updates via email to me. I received a strange response from a state attorney that included this introductory paragraph, followed by several paragraphs of lawyer-speak:

Thank you for your public records request dated July 1, 2021 seeking daily COVID-19 data and information. Please be advised that the state of emergency declaration by Governor Ricketts on March 13, 2020 officially ended June 30, 2021 at 11:59 p.m. The coronavirus dashboard was also discontinued on June 30, 2021. Please see the DHHS Press Release on this at https://dhhs.ne.gov/Pages/State-COVID-19-Dashboard-Concludes-as-Emergency-Ends.aspx which includes alternative ways to obtain Nebraska data. Your request must be respectfully denied under Neb. Rev. Stat. § 71-503.01(1).

None of the alternative ways to obtain Nebraska data mentioned in his response from the linked press release seem to provide daily case and death data from the Nebraska Department of Health for our purposes. I'm blown away that my request triggered a response from a lawyer, but this obviously means we are not going to see any additional data from Nebraska. Any thoughts? Bz8x8c (talk) 15:04, 2 July 2021 (UTC)

The CDC provides a daily chart app. [2] Select Nebraska from the dropdown, and hover your mouse near the bottom right. A pop-up should show the daily cases, etc. Leave the "View (right axis)" option untouched. I do not know when this updates… Thosbsamsgom (talk) 22:27, 2 July 2021 (UTC)

Thanks! Other than the dates being off a day (dated per the "Thru" date instead of the release date), it looks correct. Gee, you think Nebraska could have answered my email with that info instead of having a lawyer respond with legal drivel? At this point, I'm thinking there's too little intelligence in Nebraska and far too many lawyers! Bz8x8c (talk) 05:01, 3 July 2021 (UTC)

I found the chart by following the CDC link on the state's notice board, so that attorney-person technically offered a publicly available solution to your query. But yes, the 'smell you later' attitude is disappointing. Even from a Nebraskan. ;) Thosbsamsgom (talk) 11:22, 3 July 2021 (UTC)

How to get to tables after split?

You’ve split the tables and removed the suggestion for the split (the only place I’ve seen for the link), now where is the path of links that reach the tables? I’m going to take a wild stab that perhaps you all have bookmarks to the pages? Oltemative (talk) 08:10, 1 September 2021 (UTC)

Someone else removed the Split section (which contained those links) that was originally added prior to the split being performed, with the justification that the split had been performed. While it made sense to remove the section asking for discussion on the split, I agree the links themselves were useful. Therefore, I have re-inserted 3 lines containing these links to provide the ability to navigate to them from this page. If anyone has issues with how I've formatted these links, please feel free to make adjustments or provide comments here. Bz8x8c (talk) 01:25, 3 September 2021 (UTC)

Much better! Thanks! Oltemative (talk) 03:17, 3 September 2021 (UTC)

You're very welcome. Thanks for your input! Bz8x8c (talk) 13:12, 3 September 2021 (UTC)

Strange Data on Tennessee Dashboard for September 29, 2021

In today's update (Sept 29, 2021), I encountered some really strange data for TN, and I felt I should explain what I found and what I ended up doing with the TN data update...

The dashboard for TN today had 2,448,804 total cases and 30,198 total deaths. However, the previous day's total cases was 1,220,784 and total deaths was 15,055. These new totals would mean more than a doubling of both the cases and deaths today. That seemed unlikely, but I decided (don't ask me why because I don't know) to see what I would get for daily cases if I used exactly 1/2 of the new total cases, or 1,224,402 cases. This new total cases value resulted in a daily cases value of 3,618 cases, which just happened to be exactly 1/2 the daily cases value of 7,236 on the TN dashboard. When I did the same with the death total, using 15,099 total deaths, the resulting daily deaths value was 44, exactly 1/2 the daily deaths value of 88 on the TN dashboard.

These two results based on 4 dashboard values is beyond any statistical chance of happening randomly this way, so I am convinced that someone updated the TN dashboard today with exactly doubled values in all 4 cells of the dashboard, so I chose to base today's update of TN data on this assumption. If someone has any complaint, I'd love to hear any counterarguments to my logic. Bz8x8c (talk) 02:24, 30 September 2021 (UTC)

I noticed that, too… TN's people may correct the value if it was an error. Thosbsamsgom (talk) 02:27, 30 September 2021 (UTC)

Thanks for the quick response. It's nice to know someone else is watching and I'm not just doing these updates for myself! Bz8x8c (talk) 02:32, 30 September 2021 (UTC)

Hahaha! You're doing great work, BTW. Pure chance I was looking at TN's dashboard wondering 'wait … that's odd' Thosbsamsgom (talk) 02:42, 30 September 2021 (UTC)

TN did indeed correct their dashboard at some point to the values I assumed yesterday. I don't see an update for today yet though. Bz8x8c (talk) 21:04, 30 September 2021 (UTC)

Tennessee Covid data

Tennessee did issue a report on Jan 12 but but the report shows it to be only through Jan 8. Shouldn't that be reflected properly in the table?--2600:1702:15A0:6700:DD60:C783:A3F8:4DDE (talk) 04:35, 13 January 2022 (UTC)

Thanks for your question. I am aware that Tennessee reflects an end date for their weekly report several days earlier than when the report is published. However, most states reflect that their data published on one date is through the end of the previous day, so it would be difficult to justify reporting the data on the "through" date of the published data for one state and not all of the rest. We have always reported new cases and deaths on the day the new data is actually published to simplify the data update process and avoid extensive editing of historical data points. The only exception to this approach is when a state or territory publishes a certain day's data late and it is clear which publishing date is associated with the data. I hope that helps explain the approach that has been taken since the beginning of this effort. Bz8x8c (talk) 00:51, 20 January 2022 (UTC)

Odd Pennsylvania Dashboard Case Data Update

The update to the Pennsylvania dashboard today (Jan 28) included a decrease in the total confirmed case count of over 250,000 cases, but checking the data on the CDC site reflected an increase in cases today, consistent with the increase in previous days. I believe the CDC data includes potential cases, where the dashboard value only includes confirmed cases, but I expect them to track each other to some degree. I searched for anything on the PA site and online to explain why the dashboard reflected such a drastic decrease in cases and could find nothing to explain it anywhere. Since the CDC data does not reflect a decrease in cases, I'm suspicious that there's an error in the PA dashboard. For that reason, I've chosen to not include today's case update for PA until I can confirm the decrease is real or an error, adding a PA footnote to this effect. If anyone has any additional insight, please feel free to share your knowledge. Bz8x8c (talk) 04:13, 29 January 2022 (UTC)

In the January 29 update, the total cases returned back to the previous level plus an amount roughly equal to two days of typical PA cases. This indicates my suspicion about the update error was correct, so I have left the January 28 daily cases blank. I adjusted the footnote accordingly. If anyone has any concerns, please feel free to share your thoughts. Bz8x8c (talk) 03:31, 30 January 2022 (UTC)

Time to split this thing again?

IMO we should split it off by years. So as of the time of this post, this page would have January 2022, only. The 2021 numbers could be on a page for 2021.

My electricity bill, my mouse's scroll wheel, and greenhouse gas emissions will all benefit for this change. Adoring nanny (talk) 14:08, 1 February 2022 (UTC)

I've been expecting Thosbsamsgom to perform the split, since he prepared links for it within the existing page just before the new year. Bz8x8c (talk) 20:45, 5 February 2022 (UTC)

Oh.! You want ME to do it.? ;) Thosbsamsgom (talk) 00:35, 6 February 2022 (UTC)

OK.!! Split. I will crush the daily temps. after you guys complete today's rows. Thosbsamsgom (talk) 01:22, 6 February 2022 (UTC)

Ha ha! Yes, I want YOU to do it!  ;-) FWIW, I have never created a new WP article. Bz8x8c (talk) 02:51, 6 February 2022 (UTC)

All done now. Thanks for your help with the split. Bz8x8c (talk) 03:49, 6 February 2022 (UTC)

Bz8x8c: Happy to help! Thosbsamsgom (talk) 19:54, 7 February 2022 (UTC)

Haha, I wanted to make both of you do it. Seriously, thank you very much. Adoring nanny (talk) 10:26, 8 February 2022 (UTC)

March 20, 2022 Daily Recovered Value

On the COVID confirmed cases template, on the March 20, 2022 entry, in the "Recovered" column, the value entered is 5,883,699 (which is a daily value). The data prior averaged around 172,561. It seems to me the March 20, 2022 entry of 5,883,699 is an errant entry. Could you review your recovered entry on this date? — Preceding unsigned comment added by 65.30.236.132 (talk) 12:24, 23 March 2022 (UTC)

Thanks for asking your question, as I was just as surprised as you. You are correct that typical daily recoveries have been much lower than this value. However, this value is calculated based on the change in total recoveries from the previous day based on the data source, and on that day the total recoveries experienced a significant increase. This was likely due to one or more states performing a catch-up on their recovery totals on March 20, 2022. Therefore, the calculated daily recoveries was significantly higher than was typical. If the total recoveries value the following day had returned to a much lower level, we might have drawn the conclusion that the previous day's total was incorrect. However, that wasn't the case, so we have to assume the source data was correct. I hope that helps explain the value you're seeing. Bz8x8c (talk) 22:42, 23 March 2022 (UTC)

Negative Deaths on March 22, 2022

There are -1809 deaths listed for March 22, 2022. Is that a typo? — Preceding unsigned comment added by 69.242.127.203 (talk) 13:09, 26 March 2022 (UTC)

Thank you for your question, but it is not a typo. The edit summary from when I entered the daily total deaths on March 22 was "CA, CO, ID, MT, NM, OR, UT, WY March 22; Add March 22 totals, reflecting large decrease in SC deaths", and I also added a note to SC on the Deaths page that documents the decrease in SC total deaths on that day. Bz8x8c (talk) 18:16, 26 March 2022 (UTC)

March 14 Decrease in Total Deaths

Please note that on March 14, 2022, Massachusetts reported a change in the methodology used to count confirmed COVID deaths. This change in methodology resulted in a decrease of over 4,000 totals deaths, which caused a decrease in the overall deaths and a negative value for the daily deaths for the country. Part of their notice is shown here...

As announced last week, today’s dashboard (3/14) reflects an updated surveillance definition used to count COVID-associated deaths, adopted following guidance from the Council of State and Territorial Epidemiologists which recommended a standardized approach to counting COVID deaths.

Also note the increasing number of states changing to only periodic updates, as opposed to daily or weekday updates. I'll continue to note each schedule change as it occurs, but I expect this trend to continue with other states in the coming weeks. Bz8x8c (talk) 01:41, 15 March 2022 (UTC)

Effect of Calculating Active Cases using Recoveries from a Different Data Source

For some time now, I've been concerned about the fact that the recoveries value, and indirectly active cases (which is calculated as total cases - total deaths - total recoveries), is based on the source worldometer. Since the cases tracked by this source are over 9 million more than the total cases pulled from states and territories data, which is mostly confirmed cases only, I knew there might come a time when this difference would cause problems for our data. This weekend, the problem came much sooner than expected when recoveries jumped by about 13.5 million in just two days. The result is that our calculated "active cases" total ends up being a negative value, which really doesn't make any sense. While I knew this problem might eventually need to be addressed, I didn't expect it to happen so soon.

I'm not sure what to do about it, since there's no chance the active cases value will ever return to a positive value. I'm open to any suggestions on how to deal with the problem. Bz8x8c (talk) 02:12, 11 April 2022 (UTC)

Been wondering where you all were getting the recovery numbers from. Yeah, looking over the worldometers.info, it says 1,125,278 for active cases on April 12th. So the state numbers are not from the same source as the recovery numbers? So either the numbers get a rude adjustment noting a source, or the state tables split from the daily totals (making two articles of data), or you leave the weird negative numbers be. The rude adjustment option takes the current numbers from April 9 or 10 and crudely uses the recovery number the previous day (8 or 9) Compares it to worldometers.info number and forces the number the same.

Example: Say April 8 was the reset date, Take the numbers "66,311,893" and (worldometers.info active cases number April 8: "1,216,824") and subtract, giving the number "65,095,069" as the recovery adjustment number. That be my suggestion on how to reset it, as any other method might not be do-able at this point. Devilmanozzy (talk) 04:08, 14 April 2022 (UTC)

Thanks for your input. Any thoughts from Thosbsamsgom or Rider0101 on this or other possible solutions? Bz8x8c (talk) 01:40, 15 April 2022 (UTC)

I would suggest removing the negative cells from the recovery column. Thosbsamsgom (talk) 01:47, 15 April 2022 (UTC)

I take it you mean just omit the negative values in the Active column cells, leaving them blank. Is that correct? Bz8x8c (talk) 02:17, 15 April 2022 (UTC)

Indeed, empty cells would be prudent given the source of the numbers. We should also include a note on the reliability of the "Recovered" patients numbers, since Worldometer (?) is itself a tertiary source. But Rider0101 should weigh-in on this. Thosbsamsgom (talk) 04:12, 15 April 2022 (UTC)

I agree we need his input too! Thanks for chiming in on the issue. Bz8x8c (talk) 04:25, 15 April 2022 (UTC)

I would not use Worldometer at all. It's red at WP:RSP. Not reliable. Adoring nanny (talk) 17:34, 2 May 2022 (UTC)

Thank you for your input. I understand your point about the data source, but it was added to the page quite some time ago by someone else. So my question back to you is what do you think we should do about both the existing historical data and future recovery and active cases data? Should we just cease adding future data, or should the historical data for recoveries and the calculated active cases be removed from the table entirely? Bz8x8c (talk) 21:22, 3 May 2022 (UTC)

If the source is unreliable, then any results in the table that rely on it should be removed. Brutal, I know. Thosbsamsgom (talk) 18:27, 4 May 2022 (UTC)

Brutal but honest. I don't disagree with your assessment, but how do we reach out to any consumers of the page before making such a brutal change, especially if the data is used elsewhere on Wikipedia? Bz8x8c (talk) 21:56, 4 May 2022 (UTC)

Template:COVID-19 pandemic data/United States medical cases chart also uses the recoveries information from Worldometer. That infomation should be removed if it is not reliable. Rpvt (talk) 01:01, 5 May 2022 (UTC)

Leave a note up top, but otherwise remove the recovery totals from this year and prior years. Thosbsamsgom (talk) 08:51, 7 May 2022 (UTC)

Two questions: Should we place a warning note at the top for some amount of time BEFORE we remove them (maybe a week)? Does anyone have a way to remove columns from a table in one action instead of deleting lines from each row in the table? Bz8x8c (talk) 23:10, 8 May 2022 (UTC)

As much as I'd like to give folks time and such, the data should be removed immediately. Thosbsamsgom (talk) 00:11, 9 May 2022 (UTC)