Wikipedia:Top 25 Report/About

=About the Top 25 Report=

The Top 25 Report is a curated weekly report of the 25 most popular articles on Wikipedia. Our archives cover January 2013 to the present. A Top 10 version of the Report usually also appears in the Wikipedia Signpost as the "Traffic Report".

This report is based on data derived from Toolforge. It is a curated report in order to provide commentary on the reasons why articles appear to be popular (which necessarily may include opinion; the Report is not in article space to which WP:NPOV applies), and to exclude articles for which high viewcounts appear to be overly influenced by non-human views. We also exclude the Main Page and non-article pages (such as script calls). We furthermore exclude any red links; prior to 2016-AUG we produced WP:TOPRED, a weekly list of the most requested red links. Unfortunately a change in data sources makes it impossible to generate such a list moving forward.

Also compiled is the Top 50 Report, an annual list of the 50(-ish) most-viewed articles each year. There are also a couple years' worth of Top 50 annual reports tracking the daily most-viewed articles.

Records

 * Number one articles - A list of articles that have topped the Top 25 since 2013.
 * Records - Most viewed articles (for one week) since 2013, most appearances, and more.

Non-human views
Regarding non-human views, typically (since October 2014) this is determined when the percentage of mobile views of an article reported by the WP:5000 is either very close to 0% or 100%. Popular articles normally have mobile views in the range of 20% to 70%. Extreme outliers draw our attention. Articles may also be excluded using data available from |Dog WMF's Pageviews analysis tool, which can reveal unusual spikes in views. For example, normally an article on the death of a well-known person will result in a large spike in views over 1-2 days, followed by a tail. An article which suddenly spikes from 1,000 views per day to 40,000 views per day, and stays around 40,000 views per day for a long period, is not normal. We also can consult with the WMF analytics teams regarding suspect entries. However, we are not perfect in detecting non-human or "gamed" views, and attempt to explain debatable entries when they arise.

History of the Report
The WP:5000 was created in October 2012 by West.andrew.g (original update). It originally reported on a 10-day cycle, and switched to 7-day week periods (Sunday 0:00 UTC -Saturday ) in January 2013, with weekly report data starting as of January 6, 2013. With that change, the Top 25 Report was also started. The WP:5000 and WP:TOP25 were first highlighted in a Signpost article in February 2013 which received some attention from the press. The Signpost first ran a Top 10 Traffic Report in April 2013; it turned into a weekly feature in June 2013.

See Category:Wikipedia Top 25 Report for a chronological list of all Top 25 reports.

Originally the WP:5000 and WP:TOP25 did not include mobile or Wikipedia Zero viewcounts, as that data was not available to users. This likely undercounted total viewcounts by at least one-third. As of October 5, 2014, the Top 25 Report includes mobile and Wikipedia Zero data. WMF changes also mandated a change in data source in late August 2016. That change resulted in (a) traffic from some automated spiders/robots being filtered out, and (b) more accurate reporting of Wikipedia Zero viewcounts. As of late August 2016, the Top25 report integrates this new data. If attempting to compare the relative popularity of articles between time periods, these discrepancies should be taken into account.

In January 2020, the WP:5000 bot that generated automated lists of the most viewed articles crashed. This has since forced the writers to compile the data manually from Topviews at toolforge, with brief periods helped by other tools.

Volunteer for the Report
Serendipodous and Milowent were the primary editors of the Report from 2013 to 2016. We have more recently had additional volunteers start to prepare reports, and welcome additional help. We also rely on editors who proofread the reports upon posting, and comment on the Report's talk page regarding the reasons for article popularity, and other issues impacting our work. The longer that this report is continued, the more useful it may become in tracing trends, so to keep this going into the future we realistically need more people. We welcome additional volunteers, so please introduce yourself.

In late 2015 we created a Template:Top25 which can be placed on the talk page of an article which has reached the top of the chart, giving notice to editors of the popularity of a subject. Community reception to the bulk placement of such notices has garnered a few complaints, so is not currently part of our operating procedure, though we have still used from time to time.