Wikipedia:Wikipedia Signpost/2018-04-26/Special report
ACTRIAL results adopted by landslide
Kudpung has been a Wikipedia contributor since 2006 and and an administrator since 2011. His focus is on policy changes concerning deletions/notability, RfA, and the improvement of the new page patrolling and AfC processes. The views expressed in this article are his alone and do not reflect any official opinions of this publication.
Confirming the massive 2011 consensus to put an end to the inappropriate pages that comprise up to 80% of a day's intake of new articles, the new ACTRIAL debate was closed on 18 April this year after running for exactly 30 days almost to the hour, with a clear majority to permanently implement a new rule: in future, the creation of articles in mainspace is to be limited to users with confirmed accounts. Champagne corks were heard popping from as far away as New Zealand.
The RfC registered 88.8% consensus for the most important change in editing policy since the Foundation limited article creation to registered users in December 2005. Faced with an imminent roll out of ACTRIAL by community administrators in 2017 using a locally developed filter, the Wikimedia Foundation offered support this time round in the form of professional statistical research. The results and those of the ensuing debate illustrate that while the WMF brusquely rejected the trial in 2011 (The Signpost, 26 September 2011), under the organic evolution of Wikipedia they are able to respond to urgency. The volte-face was welcomed by the community.
- How they voted
Among the users opposing the introduction of the new rule, the arguments were centred mainly around an expected increase in the workload for reviewers at Articles for creation (AfC), and claims made by several Wikipedian-in-Residence (WiR) employees and other Outreach and editathon volunteers, that the move would inhibit the creation of new articles by session participants. Wikipedians-in-Residence, who primarily maintain a link between their institutions and the Wikimedia editing community by organising, for example, training events and editathons, are generally paid for their work either by the institution or by a Wikimedia-related organization. Many editathons are also organised by the outreach teams of Wikimedia chapters, and Wikipedia projects such as Women in red. Solutions suggested by those countering the arguments included the need for these organisations to be more aware of developments throughout Wikimedia and related projects, providing more support from administrators (Sysops), or even the creation of special user rights for the organisers.
I am deeply concerned that AfC is simply not fit for assessing new articles. I have reviewed a tiny fraction of the biographical articles requested for speedy deletion recently from this project, and have found a worrying error rate, far in excess of the usual NPP errors. In a very small sample (not had time to look at more, for firefighting all the problems uncovered so far) potentially viable articles are being rejected for relatively trivial stylistic reasons, for not having inline citations, for being written in the wrong tone or for suspected COI.
– Espresso Addict
The difference between the admin workload and quality of new articles with and without the autoconfirmed requirement is like night and day. It also makes paid for spam stand out sufficiently enough that it can be detected with simple behavioral heuristics. (...) Anti-abuse tools in MediaWiki are sufficiently derelict that a subject matter filter can only currently be implemented by the edit filter. (It is near impossible to get any useful software development out of the WMF in any reasonable amount of time.)
Finally, I had the privilege of deleting the first page created by a non-autoconfirmed user after ACTRIAL ended, and yes, its creator made no effort to understand what an encyclopedia is before he edited one.
– MER-C
Opposing on the premise of much increased AfC submissions, admin Espresso Addict who has herself deleted over 3,500 pages, expressed her concerns in the discussion section that reforms for AfC appear to be overdue:
| “ | I am deeply concerned that AfC is simply not fit for assessing new articles. I have reviewed a tiny fraction of the biographical articles requested for speedy deletion recently from this project, and have found a worrying error rate, far in excess of the usual NPP errors. In a very small sample (not had time to look at more, for firefighting all the problems uncovered so far) potentially viable articles are being rejected for relatively trivial stylistic reasons, for not having inline citations, for being written in the wrong tone or for suspected COI. In mainspace these problems could be fixed by any of the hundreds of thousands of active editors, alerted via the template-warning system or via categories, and relevant Wikiprojects are notified by talk page tagging & the various newbots. In draftspace, only the creator has the opportunity to fix them, without any assistance from the reviewers, and largely doesn't have the knowledge. | ” |
To which Insertcleverphrasehere replied "New users ending up at AfC is entirely a result of the fact that AfC exists, and therefore becomes the only option for a new user who wants to publish now. ACTRIAL was never about sending new users to AfC instead of NPP, but rather about stopping them from creating new articles in the main space altogether (for a time). (...) AfC reforms can come later, and I for one am committed to making sure that happens, but ACTRIAL is needed now to prevent an overwhelming deluge of terrible submissions from overwhelming New Page Patrol."
Jim Henderson (supporting) exposes in the discussion section his experience as an editathon facilitator with an arresting description of those who attend his many sessions in New York:
| “ | Interesting that the majority of OPPOSE and most of DISCUSSION are about editathons. As it happens I coach at such events more than once most months and they are a useful institution. However, they don't generate many new persistent editors. They mostly attract attendance by promising new biographies, and newbies arrive expecting to make articles about their friends. Friends who, through no fault of their own, are unfortunately still alive and able to benefit. We tell them how difficult this is. (...) Having arrived with the wrong expectation, they may ask for new a direction. We may say they should find and fix Wikipedia's plentiful old, bad articles. (...) If they succeed in finding a few refs and write a new biography in userspace, that's when they run into the 4-day 10-edit barrier. We usually have enough coaches to take care of these directly by a mainspace rename. So, the major problem is the concentration on new BLP, and the Autoconfirmed barrier is a minor one. That barrier could be further shrunk by giving Account Creators the right to boost newbies for one day or four, and maybe someone can suggest other methods, but it shouldn't be a reason for allowing the rawest newbies to create articles on their own. | ” |
Where such debates often serve to highlight related problems, commenting early (vote #16), admin MER-C highlights the difficulties in communicating the needs for assistance from the Foundation: "It is near impossible to get any useful software development out of the WMF in any reasonable amount of time."
- AC-TRIAL goes AC-PERM
Summing up, closer Primefac concludes by saying "Other concerns that were brought up were that this change moves us further away from 'The Wiki Way' (where anyone can create and edit a page immediately after joining), it gives a barrier to those wanting to immediately use the Content Translation tool, and for 'philosophical reasons'."
The RfC resolution was passed by 207 editors supporting, against only 26 in opposition. ACTRIAL is scheduled to go ACPERMANENT (or ACREQ) on 3 May, or earlier, by WMF core software developers. A debate is taking place to discuss a suggestion made by TonyBallioni for solutions requested by the event coordinators.
Moving forward – AfC meets NPP
In another venue, workshopping some future suggestions for the two systems, according to Kudpung the Wikiproject Articles for creation (AfC), and the core function of Page Curation (aka NPP) share the same goals:
- Ensuring that inappropriate new articles are not published in main space.
- Helping the creators of articles with potential to better understand how they can efficiently prepare their articles before publishing in mainspace and thus avoid deletion.
These two systems, if they were to function at an optimal level, would appear to complement each other well by addressing not only the needs of the creators of new pages, but also those of the reviewers.
Clarifying that they are as similar as they are different, Insertcleverphrasehere explains that "while the the two systems are almost the same, they operate at different stages of content creation and apply slightly different criteria for retention or rejection: while AfC is proactive, the reviewer doesn't have to prove anything, the submitter must demonstrate notability/suitability, while NPP is reactive, the the reviewer has to prove non-notability/unsuitability". He goes on to imply that "NPP is the 'necessary evil' (if we want to have quality control) and AfC is the 'necessary kindness' (if we want to be able to provide a path for very new editors to create articles)".
- New developments in the aftermath of ACPERM
As the systems of reviewing new pages and for processing drafts are inextricably correlated, talks are on-going at The future of NPP and AfC and at AfC Process Improvement with the participation of Foundation envoy Marshall Miller who is looking into the possibility of supporting improvements to the AfC process.
Other recent talks on improvement to AfC include suggestions that the New Page Reviewer user right (NPR) should become the default qualification for processing submitted Drafts. In order to demonstrate their understanding of the challenges of better reviewing, a large number of active AfC reviewers have already applied at Requests for permissions for access to the NPR user group. Administrators have been able to accord many applicants that key to the Curation tool set.
Other suggestions include the AfC process sharing the New Pages Feed and the Page Curation tool, and encouraging the WMF to address the list of required improvements to the new page patrolling software.
- Outlook
While regular AfC reviewers are expecting some possible help from the WMF on the design and wording of their templates, the New Page Reviewer faction remains optimistic that the deciders of Foundation engineering projects will accord top priority to upgrading the Page Curation software developed by the WMF 6 years ago.
!!STOP PRESS!!
Per the RfC, the page creation limitation enacted during ACTRIAL is now permanently implemented as of 26 April 2018 22:37 (UTC)
Discuss this story