Wikipedia:WikiProject COVID-19/Case Count Task Force

Scope
Maintain COVID-19 confirmed cases, deaths and recovery counts:
 * 1) Referenced with official government health agency counts as provided directly or by secondary reliable sources (WP:RS).
 * 2) As up-to-date as either official sources or reliable sources allow.
 * 3) Consistent between templates and pages (e.g. COVID-19 pandemic data and 2019–20 coronavirus pandemic in mainland China.

Locations

 * This is a guide to reliable sources for latest data about confirmed COVID-19 cases, deaths and recoveries per territory.
 * Add a note to a territory's entry if other epidemiological COVID-19 data parameters are available (such as daily numbers of lab tests).
 * If you are willing to update counts for a territory at least once a day, please, add your username to the adopters column.
 * If you are looking for a country to adopt, you may want to look for countries where your language is spoken and/or have a timezone similar to yours.
 * See for other countries.

Common errors
Aggregate sources like Worldometer, the Johns Hopkins University or 1point3acres offer timely updates based on other sources. However, there are some common errors in these sources that should be considered:

Territory double-counting
The JHU CSSE counts cases in overseas France twice. They add them to France's total, but they are already included in the official figures for France. See discussions: 1, 2.

Case double-counting
JHU CSSE and Worldometer used to add cases in nursing homes to the total number of cases in France. However, the official total for France already includes these cases. This results in more than 30,000 cases double-counted. See discussion: 1, 2. See acknowledgement by the JHU. As of mid April, this problem is fixed both in JHU CSSE and Worldometer.

Suspected cases
Aggregate sources may include presumptive or suspected cases in the total of confirmed cases. Worldometer and JHU CSSE consistently do this for countries like France, Canada or New Zealand. Note that JHU CSSE inclusion of presumptive cases is not an error, as it conforms to their own definition for "Cases" and its documentation.

Erroneous metrics for recoveries
1point3acres may include incorrect metrics in recoveries. For example, for Northern Mariana Islands they included all people released from quarantine as recoveries (now fixed). However, most people in quarantine were not confirmed cases in the first place. See discussion: 1.

Fallback sources
Some countries have some kind of daily (or even hourly) online reporting by state agencies (e.g. France), state-run news agencies (e.g. Iran) or a reputable independent media outlet. These sources are preferred and we track them at the section. However, when these are not available or when they lag too much, we need to look for other sources. Here are some tips to find them:
 * Worldometer often cites their sources in the Latest Updates section. You may find reliable sources there.
 * John Hopkins University is also a reputable source with frequent updates for countries and territories across the globe. There is also a 2019-nCoV data repository on GitHub.
 * Reuters has a high reputation as a reliable source and often publishes reports on new COVID-19 cases. You may use a search engine to find Reuters reports searching for site:reuters.com covid-19 (Google) and setting up filtering for the past 24 hours.
 * Xinhua News Agency publishes reports on new COVID-19 in many countries that are usually underreported. These reports might be found on xinhuanet.com or china.org.cn. Note that the date of the report might be previous to the date indicated in the dateline. When this happens, the actual date may be found in the first paragraph of the text.