Wikipedia:WikiProject Rankings

From Wikipedia, the free encyclopedia

✘ This proposal has failed to attain consensus within the Wikipedia community. A failed proposal is one for which a consensus to accept is not present after a reasonable amount of time, and seems unlikely to form, regardless of continuing discussion.
This WikiProject is believed to be inactive. If you are not currently a member of the project, please consider joining it to help.
Feel free to remove this tag if activity resumes or if this tag was placed in error.

After clear opposition, Old project: Wikipedia:WikiProject Rankings/old, dumped, a new project is to be drafted. --Cool Cat My Talk 03:23, 16 Apr 2005 (UTC)

Project was moved yet again because people cannot be civil. --Cool Cat My Talk 22:49, 21 Apr 2005 (UTC)

Project implementation is unknown yet, it is in the 'think about it' stage.

  • Please use discussion for comments, sarcasm included.

Contents

[edit] What do we want to achieve?

  • Determining the quality of editors.
  • Helping Arbcom members decide by gathering how good/bad a user was.
    • In most cases users just put up dirty laundry of other party. --Cool Cat My Talk 04:50, 16 Apr 2005 (UTC)
    • I think ArbCom already does a pretty good job of this. "Good" editors who have lapses in judgement are often treated more leniently (or issued more circumscribed sentences) than editors who do nothing but troll or vandalize. If any editors are concerned about any ArbCom case, they are welcome to comment and add evidence of mitigating circumstances. Further, arbitrations involve an extremely small fraction of the Wikipedia population so this might be a very substantial amount of work for very little benefit. Finally, I don't think ArbCom should (or would) be willing to rely on a single "ranking" score to evaluate the trustworthiness of an editor. --TenOfAllTrades | Talk 14:39, 16 Apr 2005 (UTC)
      • This suppose to rank people on request. It ment to be an objective insigt to arb com and be full automated perhaps. --Cool Cat My Talk 09:49, 21 Apr 2005 (UTC)

[edit] What should it be based on?

  • I see rankings as about quality editors and quality edits. If you are seriously going to go after rankings you'll need a database dump of all titles in ns0. If you are going to base it on edit count at all, to be reliable, you'll have to parse out all edits for a given user, factor out edits that are not in the article name space. Factor out/penalize for the edits that are mostly repeated pasting of the same thing all over (advert pasting). Penalize for block and bans (especially repeated ones). I'm of two minds about the namespace thing, good editors do talk. Similarly about minor edits, like spelling and grammar edits, however if you wanted to factor out minor edits, look at the Levenshtein distance algorithm for ideas. I know this sounds like information overload, but for something this sensitive, Kate's tools aren't reliable enough. (point of fact try to query dbroadwell and get an edit count) This will require significant development, space, ram, and runtime. In a perfect world considering sockpuppets, edit wars and personal attacks would be good but I don't see those as implementable. -- Dbroadwell ∞ 04:33, 16 Apr 2005 (UTC)

[edit] Implementable

  • Number of edits of a user
  • Factor out the 'minor edits'
  • Block histories, which users were blocked for what reason for how long? (not sure if I have to be an admin to do so, should be open to public.) --Cool Cat My Talk 04:36, 16 Apr 2005 (UTC))

[edit] Hypothesized

  • Factor number of reverted edits
  • Revert wars?
  • Factor out extreme cases of advert pasting and lump into subjective ranking
  • Does user have sockpuppet history?

[edit] Who should be ranked? Who should be allowed to request it?

  • Anyone
  • After a vote to save resources
  • Administrators only
  • Arbitrators and Mediators
  • Arbcom members


[edit] Is any such project a good idea?

[edit] Pros

  • Provides recognition for contributors
  • Provides an incentive to contribute
  • it's nifty and cool

[edit] Cons

  • Wikipedia is not a contest
    • Encourages competition rather than cooperation
  • Damages egalatarian ideals, promotes elitism
    • People may try and 'pull rank'
  • System may be "gamed" or abused
    • It will be hard to provide a system to promote 'lots of good stuff' instead of just 'lots of edits'
  • Having a low rank may discourage newcomers
    • Confronting people with high rank may discourage newcomers
  • is 'determining the quality of editors' inherently POV?
  • may become a 'personal attack magnet' for highly-ranked editors
  • Arbcom has expressed no interest in a system 'gathering how good/bad a user was'
  • may be confusing
  • overtones of Big Brother - "we are watching you"; the system is open and transparent, but users shouldn't feel like they are being centrally monitored and catalogued themselves.
    • to what extent would the Arbcom and the Ranking Group gradually merge?