Discrepancy in UK stats

[nickybisht]

Confirmed cases for 26th dec in UK is 11891292 and 27 Dec, 21 is 12209991 on JHU so the daily new cases is 318699. Whereas stats are different in WHO(https://www.who.int/countries/gbr/ ) Worldometer(https://www.worldometers.info/coronavirus/country/uk/ ), govt site(https://coronavirus.data.gov.uk/details/cases?areaType=overview&areaName=United%20Kingdom ) Worldometer for 27 Dec: 98 253 WHO for 27 Dec: 0 Confirmed cases on 26 and 27 dec :12111480 Govt site: 98515

Please confirm if github(https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_confirmed_global.csv ) is showing correct data or not. If it is incorrect, Please update the data.

[ghost]

Also USA for USA North America United States 2021-12-27 52793407 512,553 (new cases over 500k)

while https://www.worldometers.info/coronavirus/country/us has only something over 300k.

[CSSEGISandData]

Thank you for the notification @nickybisht. It appears that the UK has backfilled the data for the days where they were not reporting over the Christmas holiday. We'll work to get this integrated shortly.

@pepe78 thank you for your comment. The spike in cases on the 27th is due to the holiday backlog when a majority of states did not report new data (only 10 states published data on December 25th). Worldometers historical data is not verifiable and should not be compared to our dataset. For an example, Worldometers is reporting 265,346 cases reported on the 25th of December. A random sample of their state level data shows that they are reporting new cases for Indiana, Kentucky, and Mississippi on the 25th; all of these had previously indicated that they would not be publishing data on that day. For a more explicit example, Mississippi stated they would not be updating their data on December 23 through 26. Worldometers state specific page is reporting that Mississippi reported the exact same number of new cases (1,411) four days in a row when in reality they were not reporting. The data in our repository is organized by date of report. It is unclear what metric Worldometers is publishing.

[ghost]

Thanks for update. Looking at https://www.who.int/countries/usa/, it seems first time reported over 500k is on 28th. For 27th it is ~158k.

[CSSEGISandData]

The date discrepancy with the WHO COVID-19 Dashboard has been noted before and is explainable by a) the update frequency of the WHO Dashboard and b) the difference in time zones between Switzerland and Maryland. In both cases, the differences result in the WHO dashboard assigning identical cases with a one day delay compared to our repository.

A) The WHO COVID-19 dashboard updates three times per day while we are collecting data hourly. Our data represents any data published by 11:50 PM eastern time on the date to which it is assigned; if data is reported and verified after the last update for the WHO COVID-19 Dashboard, it is assigned to the next day.

B) The WHO COVID-19 dashboard is based in a time zone of Central European Time (GMT+1) while our repository is based in a time zone of Eastern Time (GMT-5). Any data published within the six hour period where there is a date mismatch between the two dashboard (I.e. any time between 6:00 PM and 11:59 PM Eastern Time) will be assigned to disparate dates for the two dashboard. This leads to permanent discrepancies in date assigning for some US sources with late afternoon or evening update times (e.g., Washington state updates around 4:00 PM Pacific Time daily).

These details are available on the WHO COVID-19 Dashboard notes section. Specifically, the 'Time stamps and updates' section will be most useful.

[CSSEGISandData]

Data for December 25-28 patched in #5148. We'll leave this issue open currently as the UK is similarly going to backfill data in the coming days.