Research:Portuguese Wikipedia trends and behavior/Social Structure

Snapshot of positions

Portuguese Wikipedia has similar positions as some of the other Wikipedias, and in fact many of the positions were adopted following the lead of the other Wikipedias. The breakdown of the five primary positions can be seen below.

	Burocratas	Verificadores	Conselho de Arbitragem	Administradores	Eliminadores
Year started	2004	2007	2009	2004	2010
Primary responsibilities	Granting the status of administrator, bureaucrat, bot and eliminator Removing the status of bot and eliminator Renaming users	Checking user IP informations and login string	Conflicts between two or more users Serious disregard of the policies and common practices of the community Behaviors and situations which could affect the normal operation of the Lusophone Wikipedia	Protecting and unprotecting pages and edit protected pages Deleting and undeleting pages or revisions Blocking and unblocking users and IPs Editing the edit filter Granting and removing the flags of rollbacker, autopatrolled, confirmed editor and IP block exempt	Deleting and undeleting pages or revisions
Selection process	Vote	Vote	Vote	Vote	Analysis of requirements by everyone (no need of right to vote) and final decision by bureaucrats.
Removal process	1 year term, removed if desysopped or Arbcom decision	1 year term, removed if desysopped or Arbcom decision	1 year term or vote (2/3 of support to remove)	vote (1/3 of support to remove), Arbcom decision, sockpuppetry, legal theats, inactivity (less than 50 logs in 6 months)	At discretion of bureaucrats, inactivity (less than 25 logs in 6 months)
# of target participants	5 + 3 alternates	3 + 2 alternates	5 + 2 alternates	no superior bound	no superior bound
Notes		Elections temporarily suspended	Relatively inactive	First policies brought in from EN:WP	Status first created for ptWiki, now also in use on hiWiki

Administrator workload

The number of administratores on wiki.pt is low when compared to other big language projects. The number of successful elections has declined over time. Between 2005 and 2007, ~30 new admins were elected each year. In 2008 the number decreased to 18, in 2009 only 13 admins were elected, and in 2010 this number decreased once again to 9. As of October 2011, 9 new admins have been elected. The number of request for block on administrators noticeboard has increased over time.

Workload areas

The following are areas that would indicate over-work required by admin.

Blocking	Reversions	Page protections
		TBD :)
Total number of block requests for PT-WP has increased over time	Admins share the reversion workload with Hugglers, but they still do over 50% of all reverting although they are less than 10% of total reverters.	TBD :)

Comparison with other language projects

Note: These are from the last 6 months only, and compare only current sysops. Source: Toolserver

	ptwiki	frwiki	eswiki	itwiki
# active users
# edits per day
# admins	33	196	139	102
# admins >50	33	104	66	80
# blocks	3921	7478	9140	11059
# protections	3475	18969	2940	7437
# Meta-Wiki edit	751	632	202	369
# deletions	65392	78439	76372	65719
# imports	0	6	4	3
# right changes	217	58	174	141
# total actions	73666	105582	88787	84728

Number of administrative actions in the last semester as of October 18, 2011. The role admins >50 indicates the number of admins with at least 50 actions in the last 6 months.

Conclusions

Certainly the number of administrators on pt.wiki is lower than other language projects, even when we count only those administrators who fullfil the pt.wiki requirement of activity.
With such a low number of administrators, every administrator would have to be on line at least 5 hours per week in order there to be at least one online admin. each moment. This calculation consider that all administrators are online at disjoint periods of the week, so it is expected there to be periods of the day uncovered by the work of admins.

Next steps

Analyze the number of edits by administrators, mainly those made on wikipedia namespaces
Analyse the number of administrative actions over time, mainly those which are more likely to be controversial like blocks to well established users.
Analyze total block requests per admin over time
Look at trends in protection requests and permissions