A random survey of 50 pages

edit

Inspired by TPK's random survey, I hit up 50 random pages and noted down some statistics.

Page content

edit
  • Places - 13 (26%) (including 5 Rambot articles)
  • People - 7 (14%)
  • Concepts - 5 (10%)
  • Albums - 3 (6%)
  • Companies - 2 (4%)
  • Books - 1 (2%)
  • Copyright Violations - 1 (2%)
  • Disambiguation Pages - 1 (2%)
  • Fictional Characters - 1 (2%)
  • Lists - 1 (2%)
  • Software Packages - 1 (2%)
  • TV Episodes - 1 (2%)
  • Species - 1 (2%)
  • Other - 12 (24%)

Statistics

edit
  • Number of pages with Categories - 24 (48%)
  • Number of pages with "See Also" links - 9 (18%)
  • Number of pages with external links - 12 (24%)
  • Number of pages with infoboxes or nav boxes - 7 (14%)
  • Number of pages with images or diagrams - 8 (16%)
  • Number of pages with reasonable wikification - 49 (98%)
  • Number of pages I thought to be stubs - 19 (38%)
  • Number of pages with "stub" tag - 14 (74% of the stubs)
  • Number of pages with interwiki links to other languages - 4 (8%)
  • Number of pages which I felt I had to make an edit to before I requested the next random page (e.g. untagged stubs, blatant spelling or wikification errors) - 14 (28%)
  • Number of pages which I listed on Wikipedia:Cleanup because they needed more attention than I could give without interrupting my survey - 2 (4%)

Other notes

edit

I recognised one page as one I'd edited a few days ago, and one page as one I'd read recently.

One page (Gush Dan) had undetected vandalism on it - it had been vandalised 24 days previously.

The 50 articles

edit

A Pagan Place, Come to Daddy, Saturday Night Fever (soundtrack), Lord of Chaos, ABC Family, Patria, Automatic message exchange, Center of mass, Growth management, Hypocenter, Resource starvation, Lhasa Apso, Glen, Lilah Morgan, Socio-technical systems, Bobby Hill (baseball player), Charles Pelham Villiers, Donald Trumbull, George E. Clymer, Henry Cobham, Joel Silver, John C. Turmel, Danebury, Glorioso Islands, Gush Dan, Jekabpils, Kasai, Hyogo, Komarno, Misato, Tokushima, New Lynn, New Zealand, Asherton, Texas, Byron, Illinois, Marion, Juneau County, Wisconsin, Waltham Township, Minnesota, West Springfield, Massachusetts, GNU Aspell, Accipitriformes, Lonely Among Us (TNG episode), Abwe, Contemporary culture of South Korea, Culture of Cambodia, Dean Forest Railway, Demographics of the Central African Republic, Noric language, Puerto Rico at the 2004 Summer Olympics, RACA Fourth Decade 1933-1942, Ramsgate tugboats, Singapore Airlines Flight 6, UTF-2000, Woolwich Ferry

Stormie 05:48, Sep 16, 2004 (UTC)

A random (weighted by popularity) survey of 100 pages

edit

Inspired by Matt Crypto's random articles weighted by popularity, I hit up 100 more pages, this time selected by his script.

Page content

edit
  • Books - 2%
  • Companies/organizations - 5%
  • Concepts - 13%
  • Disambiguation pages - 4%
  • Fictional characters - 2%
  • Historical events/periods - 10%
  • Lists - 1%
  • Memes - 1%
  • People - 16%
  • Places - 8% (including 1 Rambot article)
  • Species - 5%
  • TV shows - 2%
  • Video games - 1%
  • Wikipedia namespace pages - 1%
  • Other - 29%

Statistics

edit
  • Number of pages with Categories - 74%
  • Number of pages with "See Also" links - 37%
  • Number of pages with external links - 53%
  • Number of pages with infoboxes or nav boxes - 22%
  • Number of pages with images or diagrams - 32%
  • Number of pages with reasonable wikification - 100%
  • Number of pages I thought to be stubs - 32%
  • Number of pages with "stub" tag - 22% (69% of the stubs)
  • Number of pages with interwiki links to other languages - 57%
  • Number of pages which I felt I had to make an edit to before I requested the next random page (e.g. untagged stubs, blatant spelling or wikification errors) - 35%
  • Number of pages which I listed on Wikipedia:Cleanup because they needed more attention than I could give without interrupting my survey - 0
  • Number of pages which I listed on Wikipedia:Votes for deletion - 1%

I would say that this sampling of articles was superior in every way to my original sampling. Seems pretty clear that the articles with lots of reads are also getting lots of edits, and thus tend to be in much better shape.

The 100 articles

edit