RFAs by #/participants (estimate)

All subpages of Wikipedia:Requests for adminship, along with two estimates of the number of participants:
* the number of different registered users to edit them
* the number of different user pages/user talk pages linked to.  (If both a given user page and its associated user talk page are linked to, it's only counted once.)

Both of these are overestimates - for example, the first counts users who only rolled back vandalism or corrected lint errors, the second counts users who were mentioned but never actually edited, and both count users who asked a question but never formally !voted.  (Not to mention the user being nominated!)

My first impulse was to count a given user as a participant only if they met *both* requirements, but this turned out to wildly underestimate counts due to renames (especially in older RFAs) and still erroneously counted users who asked questions without !voting.

[[Wikipedia:Requests for adminship by year]] is much more accurate, though it only goes back to 2006.

For [[WP:RAQ#RfA by number of participants?]] circa 26 September 2022.

SQL

SELECT page_title,
       COUNT(DISTINCT actor_user) AS 'non-anon editors',
       (SELECT COUNT(DISTINCT pl_title)
        FROM pagelinks
        WHERE pl_from = page_id
          AND pl_namespace IN (2, 3)) AS 'user(talk)page links',
       CAST(MIN(rev_timestamp) AS DATETIME) AS 'creation'
FROM page
JOIN revision ON rev_page = page_id
JOIN actor_revision ON actor_id = rev_actor
WHERE page_namespace = 4
  AND page_title LIKE 'Requests_for_adminship/%'
GROUP BY page_title;

SQL

Checking query status...