Wikipedia talk:The future of NPP and AfC/Archive 6

From WikiProjectMed
Jump to navigation Jump to search
Archive 1 ← Archive 4 Archive 5 Archive 6 Archive 7

NPR user right and the AFCHS

Since this has been mentioned multiple times elsewhere, what do we think about taking a Huggle-esque approach and requiring the NPR right to access the AFCHS: still keeping the individual projects and their respective cultures and focuses separate for now, but a step towards combining labour forces that could help both projects and also have a uniform standard of review that could lead to not needing review of drafts moved into mainspace by reviewers. Not having this as a formal proposal now, but simply as a testing of the waters. TonyBallioni (talk) 19:53, 6 November 2017 (UTC)

  • Support I’d also suggest granting NPR right to all the current AfCH approved users. I doubt there would be many and we can move toward eliminating the NPR on all AfC approved submissions. Cut the double workload. Legacypac (talk) 20:03, 6 November 2017 (UTC)
  • Neutral leaning oppose I'm not "officially" on NPR, and I don't necessarily want to be NPR, and if I weren't an admin I wouldn't want to be forced to join NPR simply to review drafts. The last time I ran the numbers it was 52:73 There is an even split of regular:NPR reviewers (with 23 20 admins mixed in). It's undoubtedly changed since then, but likely not by much. While I totally agree that NPR and AFC are two connected, they're more like the left hand/right hand than two eyeballs: each can do a task without being intrinsically tied to the other. Primefac (talk) 20:07, 6 November 2017 (UTC)updated the stats. Primefac (talk) 20:26, 6 November 2017 (UTC)
    • I suppose what I think the main benefit of this would be is that it'd allow us to get rid of the full protected check page, and ideally would get more people to do both AfC and NPP. As I said at Kudpung's talk (which I think was copied above), I really don't want a merger of AfC and NPR in terms of project space pages and workflow: I don't enjoy reviewing AfC submissions and there are some quirks with the project's internal culture that suite it but aren't really where I think I would thrive.
      I do think having it simply as a check like we do for rollback and Huggle is beneficial, though. It'd also make it easier to check if someone moving something out of draft space that was AfC was an AfC reviewer, because user rights are, in my opinion, much easier to check than a checklist. It'd also allow us to get rid of the double workload currently existing with AfC and NPP, which I think is good. TonyBallioni (talk) 20:19, 6 November 2017 (UTC)
      • Oh, I certainly agree that shifting something into the Article space (via AFCH) should mark it as patrolled, because that seems like double-duty, but my general assumption is that "anyone not using AFCH to move drafts" needs to be double-checked. But as I said, I'm largely neutral on the issue because I never really had to worry about NPR things and so I don't really have an opinion. Primefac (talk) 20:26, 6 November 2017 (UTC)
        • Yes, thanks for your feedback. Its appreciated since I think you know how highly I think of the work you do personally at AfC. A bit of addressing both you and A Den Jentyl Ettien Avel Dysklyver's concern is that this really wouldn't have any impact on workflow or anything else in terms of merging the projects structure and wouldn't lump anything together: it'd just be adapting the Huggle model and hopefully also allow us to cut out some work both on the admin side and on the reviewer side. We typically do have a somewhat higher standard for granting the NPR permission because of the Google indexing concerns, so if anything this would hopefully help cut down on the work that you and DGG do cleaning up after reviewers. TonyBallioni (talk) 20:38, 6 November 2017 (UTC)
  • Neutral weak oppose I would generally be supportive of making AfC a userright (draftpatroller or something) and in doing so removing the double review cycle with NPR. However, AfC is completely different to NPR, the process and workflows are completely different, and we should avoid lumping things together or comparing this to huggle. Huggle is a script to help rollback vandalism. AFCH is not a script to help patrol pages, it is a completely different process with a different skillset needed. However if this is a known factor and taken into account when granting the rights and maintaining the project, this is fine.Dysklyver 20:24, 6 November 2017 (UTC)
  • Neutral I am not a member of AFC and I am not sure AFC is my cup of tea. Maybe someday it will be. At the same time, I appreciate very much the work the AFC reviewers do. And being able to send some new and old articles to draft space is a helpful option. OK really, being able to send to draft space is a Godsend because I don't have to add to the AfD workload when applicable. So after reading this thread, if this will cut down the workload for NPR and help AFC reviewers then this makes sense. That is, as long as AFC reviewers don't have a problem with this. ----Steve Quinn (talk) 22:02, 6 November 2017 (UTC)
  • Weak oppose - I simply don't see the benefits of this. Since being an AfC reviewer does require an admin to put you on a page (at least if you want to be productive in any way). Because of this, I don't think that we need this extra step. Plus, it would likely make some non-NPR AfC reviewers leave AfC, because they don't want to go through the process of getting NPR. Thus, I don't think that it would be beneficial. That said, I would probably be supportive of having NPR members auto-added to AfC (with a message on their talk page telling them that they can get involved, should probably get the helper script, and that AfC standards are higher than NPP standards). This is because it would not push away people from doing NPP or AfC, but would encourage people who were already vetted to get involved in the latter. RileyBugz会話投稿記録 22:26, 6 November 2017 (UTC)
    • RileyBugz, the point is that there would be no extra step: the checklist would stop existing. That is the huge benefit here, and it would allow us to bypass NPP for AfC approved drafts. This would hugely simplify the process. If we were to go with this, we could also do a grandfathering like we did last year with the NPR right. Sorry, I'm not trying to bludgeon people, I'm just trying to clear up discrepancies in understandings of what is being proposed :) . TonyBallioni (talk) 22:35, 6 November 2017 (UTC)
Hmmm... I would be ok to weakly support this if it was combined with a grandfather clause and the auto-review of accepted drafts. RileyBugz会話投稿記録 22:38, 6 November 2017 (UTC)
  • Agreed there seems to be confusion. If someone is trusted enough to be approved for AfC what possible barrier is there to letting them be an NPR? If our AfC approver standard is too low for NPR, raise the standard! I don’t care if they actually want to do NPR, but with an appropriate transition we can have all AfC reviewers set up on NPR. Then we can have an AfC accept cover the NPR accept too. Legacypac (talk) 22:59, 6 November 2017 (UTC)
Legacypac, I certainly agree with that in priciple. see my notes lower down. 1.2.130.90 (talk) 02:09, 7 November 2017 (UTC)(Kudpung - not logged in because I'm on a public network from my car for a few minutes. It geolocates to Bangkok anyway, 550 miles from here).
I feel like the main barrier might be knowledge of CSD. You don't really need to CSD all that well to do AfC while for NPP you absolutely need to know it well. So some people who don't know CSD (perhaps like me, I know CSD generally but not that in depth, haven't studied too much) might be barred from AfC. Galobtter (talk) 07:04, 7 November 2017 (UTC)
User:Galobtter get Twinkle if you don’t have it and follow the explanations. The main CSDs are pretty self explanatory. I’ve got faith in you :) Legacypac (talk) 07:24, 7 November 2017 (UTC)
I do have twinkle, and do occasionally tag things to be CSDed - just obvious ones like author request deletion or in wrong namespace. I haven't really read up on the others and don't have any real desire to learn it (right now atleast, it doesn't really come up that much). And others in AfC may know even less. Galobtter (talk) 07:36, 7 November 2017 (UTC)

Just a few points

because I'm not sure if everyone understands what being proposed.

  • Merging NPP % AfC: This would be nothing more than creating an additional search criterion in the Curation tool dropdown. Nobody would be forcing NPRers and AfCers to work on both types of new page in the list, but if they want to, they could certainly help out.
  • There is little chance of getting yet another a policy based user group any time soon. The path to this is monstrous, remember I went through it just a year ago to get NPR agreed and up and running. Much depends on the willingness to get the WMF devs to spend the time on it.
  • PERM: Those who are not admins are possibly not aware of the extent of scrutiny we admins apply to applications for this NPR user group. There are even bots that help us. This level of scrutiny is not exerted on the kind of application that I and Primefac carry out for AfC.
  • AfC vs. NPR: These are not entirely different at all A Den Jentyl Ettien Avel Dysklyver. The reason that a merger has been discussed over the past 3 years is specifically because they are basically similar, and could certainly use the same interface. Thee differnce is that NPR is a front line triage, while AfC is a clinic.
  • Forcing approved AfC drafts through NPR is not as silly as it sounds. We need checks and balances, and while for the most part AfC reviewers do a good job, their actual knowledge of policy is not required to be as thorough, and they do not need to be so quick to recognise obvious signs of inappropriate articles. Compare with the right any confirmed users still have to watch the new page list and even mark the pages for deletion with no experience whatsoever. Odd as it sounds, this is what the community wanted, and of course, it creates double work for the accredited NPRers and admins.
  • The NPR flag, being a more important process than AfC, would automatically cover AfC. Indeed, when we set up NPR, we sent an newsletter to all AfC reviewers inviting them to apply.Kudpung กุดผึ้ง (talk) 01:26, 7 November 2017 (UTC)
  • ===Continued survey===
  • Comment--Excellent explanation. The path forward is simply change the AfCH criteria by adding a single line “NPR approved” and bring any active AfCH user who are not NPR through PERM. If someone can’t pass NPR PERM what good are they at AFC anyway? How hard is that? Legacypac (talk) 03:28, 7 November 2017 (UTC)
  • Neutral, leaning oppose--Almost entirely echo RileyBugz except the auto-adding of NPR reviewer(s).There are fundamental differences in the work-flow of both the groups (On one hand at NPR, you must have good CSD/AfD knowledge etc. and an ability to apply them instantly and judiciously but these are hardly exploited at AfC.On the other hand at AFC, you need a good co-operative cum content-building attitude which is hardly exploited at NPR)NPR folks are actually more like the frontline defenders.Echo Kudpung's specific point in the regard:) If you want to advertise AfC to like-capable or like-minded folks, chart out a banner at PERM/NPR asking applicant(s) to approach AFC and/or vice-versa.If this proposal is for easy tracking/checking of whether moves from draft space to mainspace etc. are by an AFC reviewer, make the AFCHS an user-right in it's own accord.All that being said, we can mass-message the non-AFC-NPR-Reviewer(s), informing them about AFC and the active AfC reviewer(s), who are not at NPR to try applying at NPR.Winged Blades of GodricOn leave 17:07, 7 November 2017 (UTC)
If NPP really wanted more reviewers what would be more beneficial than advertising - and really people on AfC should know of NPP already - is if their instructions wasn't a wall of text, and more like AfC's instructions with its collapsible/openable boxes. That'd atleast get me to read the instructions, which I haven't done mostly because of that. I mean I already know most of the details like in detecting copyvios, so I'd like to be able to see an reasonable overview like in AfC's instructions. Galobtter (talk) 16:12, 10 November 2017 (UTC)
  • Oppose changes that raise the barrier for new reviewers. NPR is difficult to obtain. We have enough backlog as it is. ~Kvng (talk) 15:17, 10 November 2017 (UTC)
  • Oppose What godric says is similar to what I've said above. Agree with kvng too. Galobtter (talk) 15:36, 10 November 2017 (UTC)
  • Question what is the reason that we want to keep a separate checklist for use of AFCH? I haven't seen anyone actually explain why that is preferable, and I am honestly curious. It seems to have no benefit to the project, and is preventing things that everyone wants like not having AfC drafts have to go through NPP. Every time any reform of the process is suggested, aligning the qualifications is always the first step people mention before future reforms. then every time we have the conversation about aligning the qualifications, we get a swell of opposition that doesn't ever really suggest alternatives or why the current system is better.
    I'm fine with consensus not going my way, and rarely get frustrated when that happens. What gets me very frustrated is when no real reason is given for why the current system is better, and no alternatives are suggested when there is rough agreement that we want reforms in this area. TonyBallioni (talk) 15:55, 10 November 2017 (UTC)
I think about ~15 drafts out of 250 at most get accepted a day from AfC. NPP reviews something like 500? drafts a day..so the benefit is very little. But AfC is doing pretty bad with the backlog so any reduction in the number of reviews would be pretty bad like kvng said. The skillsets overlap somewhat but there are some big differences in CSD knowledge required, and what Godric said. So if that led to losing some reviewers..wouldn't like that. I can see the benefit, and if there wasn't really (if it was shown) much reduction in actuality with the number of reviewers I'd be fine with it. Galobtter (talk) 16:04, 10 November 2017 (UTC)
This would not reduce the number of people able to review: it would automatically increase it as there are more people with the NPR flag than are active AfC reviewers. Also, I find the objection "the standard to get NPR is too high" to be a very weak one. Yes, the processes are different but the end goal is the same: publishing content that is freely available to anyone and can be indexed by a search engine. If the AfC standard is too low for adding reviewers, then raising it would be a positive thing for the encyclopedia, especially given Legacypac's (very valid) observations that most AfC reviewers are too reluctant to publish an article to main space and at the same time are too reluctant to simply say "this draft has no place in the encyclopedia, and never will", so they feed the AfC backlog by denying and allowing for resubmission countless times for a draft that will never be published. Having a common standard here would likely address both of those issues by making the qualifications (and thus knowledge of policy) the same, and I honestly see no negatives. TonyBallioni (talk) 16:16, 10 November 2017 (UTC)
You can make it unidirectional - everyone with NPR can use AfC, which I would support. That would increase the number. AfC reviewers are too reluctant to publish an article; too reluctant to simply say "this draft has no place in the encyclopedia, and never will" I'm pretty sure many or most of these people do have patroller flag. It's more like an attitude at AfC than unqualified users being unqualified and thus reluctant. We're actually having a discussion on it at WT:AFCR and everyone who was cautious in declining also has patroller flag. Also I have no idea how NPP works but for clarification: can people with patroller flag approve drafts they approve via AfC? Galobtter (talk) 16:24, 10 November 2017 (UTC)
I'd be fine with that for now as a compromise, and be willing to revisit the discussion in a few months after we see how that works out. I see zero benefit now having two de facto user rights based on an argument from differing personality types being attracted to the two projects (which is what the majority of the opposes are every time we have this discussion). My frustration is that this is being painted as somehow limiting AfC: if anything it would increase the number of AfC reviewers and make vetting easier because Primefac wouldn't be the only one doing it.— Preceding unsigned comment added by TonyBallioni (talk • contribs)
I support the unidirectional proposal giving all NPP reviewers permission to use AFCH. ~Kvng (talk) 17:25, 10 November 2017 (UTC)
It is sort of limiting AfC in that half the current reviewers - of course probably more like 10% of the actual reviews - don't have it, and that could lead to loss of reviewers. I wouldn't be totally opposed to it though. I really have no idea how strict the NPP approval is vs AFC etc so I can't make a real informed judgement. Galobtter (talk) 17:27, 10 November 2017 (UTC)
Going by the experience i.e. time spent requirements, access to AfC is easier than to NPR.Winged Blades Godric 04:17, 16 November 2017 (UTC)
  • I suppose it all really boils down to: we want a single new pages system or we want a new pages system with AFC tacked on. We can have fun working out the details of how a single system would work if we decide that AFC shouldn’t just be 'tacked on'.
Right now we have one new pages system (NPR), and AFC, which is in effect just a helpdesk, its not 'approving' anything (from a NPR viewpoint), it is just helping people move pages into mainspace with a reasonable chance of passing NPR. The idea as I see it is to allow AFC reviewers to pass an article as if they were at NPR, to avoid the double review of such pages. This could be done without merging them, but then we would have two parallel new pages systems, which is probably a bad idea. Dysklyver 16:32, 10 November 2017 (UTC)
I think AfC remaining a helpdesk does kinda make sense. I mean people with COI have to use it, and AfC helps make it not promotional, and currently "eh" articles get sent to AfC to help them pass NPR. Like I said above, the savings for having AfC articles automatically approved is almost negligible, but it'd be quite a bit of trouble to implement. Galobtter (talk) 16:47, 10 November 2017 (UTC)
I have no desire to change AfC from a help desk: I certainly don't want to be dealing with that, and I think people who want to should have a space devoted to it.
What I'm saying is this: a full protected checklist to use something is the exact same thing in practice as a user right. There is no real difference. We already have a user right that requires admins to check for the exact same things. No reason has been given why this move would be a bad thing (other than it being two different personality types, which I don't think is relevant to a technical discussion). Any discussion to create a draftreviewer flag would be shot down in an instant with the objection "just use NPR as the flag". We might as well do that now. TonyBallioni (talk) 16:58, 10 November 2017 (UTC)
Sorry, but the main opposing argument is quite relevant to the discussion.It's not as technical as you speak of and is much more of an attempt at quasi-merger of certain aspects of NPP with AfC than converting a checklist to a right-dependent-tool.Your example of Rollback and Huggle are too intrinsically linked than the case at our hand. And, count me as one who will support your last proposal.Winged Blades Godric 04:17, 16 November 2017 (UTC)
  • Comment: NPP is not 'hard to get'. Like RfA and any other user right, it ts presumed that any one applying for it would already bring sufficient knowledge and experience with them. The numerical thresholds we set (RfA doesn't have one) are really only to prevent the absolute least serious of applications and time wasters - and we have found that it works. At PERM, admins do not automatically accord rights to every candidate who has reached the level demanded by the numerical barrier. This is what the hat collectors don't realise. It goes without saying that there are a lot of hat collectors in the special rights groups. That's why even with ACTRIAL the backlog is on the rise again. The half dozen regular reviewers have taken a well earned respite, and the 444 others aren't, and weren't, doing very much beyond putting an 'I am a New Pge Reviewer' badge on their user page. Or, as it turns out occasionally, using their NPR status to patrol their own 'paid for' articles. Kudpung กุดผึ้ง (talk) 23:17, 10 November 2017 (UTC)
Kudz - you already know my patterns, and it may be that I spread my time too thin doing both AfC & NPP trying to find sources, and taking the time to evaluate potential, copy edit/expand when dealing with unsourced or poorly sourced stubs. I've since learned that G11 is a hit and miss proposition (an admin said that if only one sentence is not promotional it fails G11) so I'm taking them straight to AfD - another time sink. I do believe that after an NPR passes an AfCH, that same NPR should be able to mark that article reviewed the moment it's moved into mainspace, which I think is what the iVoting is about above regarding our curation tools and combining that part of the work. Atsme📞📧 23:45, 10 November 2017 (UTC)
No one is criticising the dedicated all-round editors. There is a tool that shows us who has asked for the NPR tools and barely used them, and viewed in conjunction with their edit history, it can be quite revealing. Kudpung กุดผึ้ง (talk) 06:55, 11 November 2017 (UTC)
  • The arguement that NPR requires CSD/AfD knowledge that is not required for AfC work is problematic. The core criteria at AfC is "will this page survive Deletion via CSD/AfD". If an editor does not know enough about deletion process to assess suitability they can't be much help to a new editor at AfC. Legacypac (talk) 20:24, 11 December 2017 (UTC)
I agree with Legacypac. I'd rather deal with backlogs at AfC than backlogs at NPP. Ok, Butch - cut 'em off at the pass! and Butch replied, I got this, Sundance. (artistic license) Atsme📞📧 01:38, 12 December 2017 (UTC)

NPR Backlog

I expect everyone has received the December NPR newsletter. Eleven years ago Jimbo Wales suggested Wikipedia contributors should shift their focus from the number of articles and instead work on improving their quality.

By preventing through ACTRIAL the possibility for first-time users to create articles directly in mainspace, we have successfully reduced the workload at the New Pages Feed by about 80%. But that does not mean we can retire. For some unexplained reason the backlog has not reduced one iota, unreviewed articles - mainly junk - is being released to Google after remaining unpatrolled for 90 days . And we are now faced with a new challenge: detecting an increase in artspam and other paid editing by users who do their 10 edits and patiently wait 4 days.

In a volunteer collaborative environment, participants can make their own minds up as to where they want to work and the extent of the work they do. One of the biggest challenges to effective New Page Reviewing is incentive and reward, but it’s not like Today’s Article for Improvement or Collaboration of the Week or the WikiCup.Those are all informal projects. NPP is an essential, but dreary core function that simply has to be done by someone, and all the time 24/7. Kudpung กุดผึ้ง (talk) 23:38, 12 December 2017 (UTC)

Apologies, but I'm taking a short break from NPP, at least until after Christmas. Hope that's ok. Atsme📞📧 01:45, 13 December 2017 (UTC)

AFC discussion summaries and moving forward

Please add your voice to this discussion on improving AFC. https://en.wikipedia.org/wiki/Wikipedia:Village_pump_(idea_lab)#AFC_discussion_summaries_and_moving_forward Egaoblai (talk) 08:38, 18 December 2017 (UTC)

RfC at Village Pump (proposals) that concerns this project

I have posted an RfC that may impact this project and/or be relevant to the users who frequent this page. You can find it at WP:VPPR#RfC: Three Strikes Rule for AfC submissions and reviews. Your comments are welcome. — Insertcleverphrasehere (or here) 09:27, 21 December 2017 (UTC)

Collecting stats of AfC vs NPP for a random dataset

I was wondering if any users that are both an experienced New Page Patroller as well as an experienced AfC reviewer could join me in a statistics gathering mission. Legacypac?

I created a list containing every 10th article in the AfC pending submissions list: User:Insertcleverphrasehere/AFC stats

The idea is to (hypothetically) review each draft as if it was an AfC submission as well as if it was a NPP submission (per the reviewing instructions of each group), and record the results of each (hypothetical) review. Additionally, a full WP:BEFORE check should be done, with a search for references and comparison to subject specific notability criteria. As an additional data point to use for the three strikes dataset, the number of previous reviews should be recorded.

Note that for the articles I have reviewed so far, after preforming a source check I often added sources and then otherwise reviewed the draft differently to what I had recorded in the table as what I thought the likely outcome would be (I don't want to waste the effort of findinga nd reviewing sources, and then adding them changes the outcome). However, if you change your mind after the source search, don't change the table. The table is meant to represent the likely outcome if following the instructions to the letter with no extra effort expended (i.e. no source search for AfC), no refs added, etc. Not necessarily what you actually do with the draft in the end (for all the stats care, after doing the 'hypothetical' reviews and searches, you could just leave the draft unchanged if you don't want to review it -- though that seems like a wasted effort).

Instructions can be found on the list page, as well as a list of options for each. Please sort by title and then pick a letter and start reviewing alphabetically from that point, so that you are not inadvertently cherry picking articles based on their titles (the total list was chosen at random, so reviewing the whole list there will be no bias, but if we want to analyse the results as we go, we need to not cherry pick).

If you choose to help out, Thanks! — Insertcleverphrasehere (or here) 14:37, 14 March 2018 (UTC)

  • Drewmutt is currently doing a significant amount of work patrolling accepted AfC submissions and sending ones that should receive broader input to AfD. Primefac might also have something to add as he maintains all of the AfC stats. TonyBallioni (talk) 14:40, 14 March 2018 (UTC)
There is often a difference between what I do and what the typical AfC reviewer will do. I'm ok with letting mainspace users fix up pages on notable non-CSDable pages instead of forcing the newbie editor to get everything perfect. I believe one of the best ways to get more notable topics and reduce the AfC backlog is to accept anything that is not CSDable and should pass AfD. The worst that happens is no one gets to the topic for a long time. Legacypac (talk) 21:23, 14 March 2018 (UTC)
Yes, I think your method should be the model for most AfC reviewers, but I do think that getting more input from others is also good. TonyBallioni (talk) 21:25, 14 March 2018 (UTC)

Challenging an assumption in ACTRIAL analysis

Here Wikipedia:Autoconfirmed_article_creation_trial/Post-trial_Research_Report#Less_low-quality_content_in_article_space WMF staff talks about increased deletions and says "This increase in deletions is not commensurate with the increase in draft creations, meaning that we see a lot of created drafts that appear to not warrant deletion."

This is not correct - it ignores timing. In Mainspace a delete worthy page will most likely get caught in NPP amd deleted within 90 days (even with the backlogs). In Draft page will not typically face deletion until at least 180 days as a G13 or much longer if it goes into the AfC review, waits for reviews 3 or more times, gets some bot or AWB or disambig edits etc. You can somewhat see this by looking at dates created on this list User:MusikBot/StaleDrafts/Report. In other words, the increased unencyclopedic content created in Draft instead of mainspace during ACTRIAL is mostly still sitting with no move yet to delete it.

It's not that the Drafts do "not warrent deletion" it is that they are not old enough to delete G13, no one is even looking at the non-AfC submitted Drafts yet, and even if we wanted to review those pages Gx CSD criteria is way narrower than Ax CSD criteria is. There is no PROD for Drafts. Further some editors insist that we can't look at Notability in MfD (see WP:NMFD) so sending lots of garbage there will not be accepted. Even editors who are happy to debate notability at MfD prefer to let G13 catch the junk later because there is so much junk, trying to debate it all is hopeless.

I 100% agree that a shift of new creations to Draft space has happened, but that is a wonderful thing. Draft space is no index and it has a clock where an Admin human will look at every Draft page in 6 months or so unless someone first deletes or promotes it. Legacypac (talk) 22:46, 14 March 2018 (UTC)

ACTRIAL

Now that it appears that ACTRIAL has been a resounding success, with the concerns expressed by some about the slight increase in submissions to AfC, it may be time to consider consolidating the two systems. They are as similar as they are different but the skillsets may not be so varied. If AfC reviewer candidates were subject to the same scrutiny as applications for the NPR right, there might be an argument now for merging the two operations or at least elevating AfC from WikiProject to an official process with its own user right. In any case, joint acollabortion is now called for. Kudpung กุดผึ้ง (talk) 14:19, 12 March 2018 (UTC)

Agreed Legacypac (talk) 14:47, 12 March 2018 (UTC)
What would a merged NPP+AFC look like, practically speaking? I suppose the Page Curation tools can be used to review pages in the draft namespace, but how would we handle requests for drafts to be moved to mainspace? – Joe (talk) 15:25, 12 March 2018 (UTC)
  • Unifying the user rights is one step. I actually prefer to let other editors add categories and wikiprojects after I accept a Draft. I don't know those as well as deletion and notability issues. So I'm not necessarily in favor of autoaccepting draft. Legacypac (talk) 15:52, 12 March 2018 (UTC)
Unifying the requirements of reviewing (standards) should be our first step. At the moment, AfC reviewers are not required to do WP:BEFORE before they decline a draft (or at least this is not made explicit at all in the reviewing instructions). This means that many notable topics are declined due to not having sources in the article, regardless of whether sources exist or not. Two things that I would like to see before merging NPP and AfC: 1) Changing the criteria for declining to require that reviewers perform a search for sources before declining a submission, as well as 2) removing the prohibition on nominating drafts for deletion for lack of notability (overturning WP:NMFD and allowing reviewers to immediately MfD a draft if it fails a notability source check). These two changes go hand in hand, work well with the three strikes rule proposed above, and together would result in essentially treating drafts as if they were articles (at least for notability reasons), making a unification trivial and essentially a formality. — Insertcleverphrasehere (or here) 08:33, 14 March 2018 (UTC)
Having reviewed the RfC that established WP:NMFD, I think we would have to be clear that the userspace would still be exempt, and perhaps only have it apply in the case of repeated submission to AfC (three declines). Even then, we might not be able to get a consensus for this, as many users seem to think that the draftspace should be used as a holding pen for stuff that might be notable in the future. Tough one. — Insertcleverphrasehere (or here) 09:02, 14 March 2018 (UTC)
Don't we need any minimum standards before an article is accepted? Under your clause (1), it looks to me as if article creation could be merged with Requested articles: the "creator" has only to name the topic and, if "notable", it becomes someone else's job to find all the references and, you know, write the article. There needs to be a "decline" process that keeps probably notable drafts in purgatory until they reach some minimum standard. Tagging doesn't cut it: tags just stay up for ever: Noyster (talk), 10:42, 14 March 2018 (UTC)

I've reviewed thousands of pages up for G13. A very small percentage that get there are on notable topics, and most of those are already covered in mainspace (a copy of mainspace, or already copy pasted to mainspace). Maybe a BEFORE search is not required of AfC reviewers but in most cases the page is either obviously notable or not notable. I see many comments suggesting reviewers are doing notability checks on the pages between these clear cases.

@User:Noyster we do have draft purgatory, its called Category G13 postpone and/or removing G13 tags and/or commenting or editimg on the page to delay deletion. The 6 month review forced by G13 very occasionally turns up good, usually unsubmitted pages that can be promoted to mainspace. I think some new users figure they wrote a Draft and that puts in Wikipedia. Legacypac (talk) 10:56, 14 March 2018 (UTC)

(edit conflict)@Noyster. Minimum standards would be CSD criteria and the BLPPROD condition (which would become main reasons for declining a notable topic at AfC). Anything else that is found to be notable in a search should simply be sent to mainspace and slapped with a {{sources exist}} template. The fact is that new editors often have no idea how to format a reference or how to find quality references even if they exist. It isn't acceptable to "keep probably notable drafts in purgatory until they reach some minimum standard" if they are notable topics, as this results in a lack of the kind of collaboration that happens in the mainspace that has a chance to improve the article. It then is abandoned by the new editor who does not have the skills to improve it further and the article is eventually deleted via G13 (unless Legacypac saves it). If the article has a bunch of bad content, cut it down to a two line stub. Seriously, we can't keep up this fiction that AfC can continue to act like it is a gatekeeper of quality for drafts that will eventually improve. The recent report by the WMF found that the proportion of drafts that are eventually published to main space is only 1.2%, which is essentially nothing. We aren't improving notable topic drafts by putting them in purgatory, we are essentially deleting them all. The proportion of drafts that are on notable topics is currently unknown, but I have a research project aimed at answering this question by reviewing and doing a notability check on a random sample of AfC submissions. If the proportion of notable topic drafts is much larger than 1.2%, then this indicates that we need to change the way that AfC works, if that proportion is very small (close to 1.2%) then no changes are needed to AfC, but we might want to question the purpose of bothering to keep AfC around at all if only 1.2% of drafts make it to main space. — Insertcleverphrasehere (or here) 11:40, 14 March 2018 (UTC)

Thank you. I haven't seen this {{sources exist}} tag before; is it ready to be added to our list of cleanup tags? More broadly with so many WikiProjects defunct, I query how much interest there still is (apart from among conflict-of-interest types) in writing articles on top of someone else's stubs, even in mainspace, when it doesn't even add to one's "Aren't I good, I've created nnn articles" total. The original creator does presumably have some motivation and is best placed to know where their information can be sourced, and skills in editing and referencing are not that difficult to acquire once they know they need to do that to ensure that their creation goes live: Noyster (talk), 12:21, 14 March 2018 (UTC)
Noyster and Insertcleverphrasehere make good points. Myself, I'm willing to help develop a weak article if I've been introduced to the newbie in a setting such as an editathon. For example, see the Women's Classical Committee, which I helped last year, or the article Tall Jawa which was created by someone in Greece that WereSpielChequers asked me to tutor. The key is to establish a good mentoring relationship by making an introduction. Picking up a random stub cold isn't going to work so well. That's more like the QPQ review process that you get at DYK which is more of a chore and so tends to be either perfunctory or becomes adversarial. Social factors are important and so we need to look at what makes projects like the Teahouse and Women in Red successful. Part of the problem must be the emphasis on secrecy and anonymity which seems to make editors too paranoid and suspicious. We need more openness and role models like Megalibrarygirl, who was recently profiled in the Library Journal. Andrew D. (talk) 12:45, 14 March 2018 (UTC)
  • I would love to require the NPR flag as a requirement for AfC, but I do not think that will ever happen as there are simply too many cultural differences between the two projects at this point, and any time it has been proposed there has been a pretty strong backlash from AfC regulars. I appreciate the work both projects do, but I think moving toward merging is not likely at this time. What I do think would be good, however, is establishing a more clear criteria for what AfC should look for, and encourage them to promote more articles on subjects that are more likely to be notable (for example, any dead biography that has some sourcing beyond a local obit has a decent shot at AfD), and be more prone to decline for reasons other than notability (promotion, for example, is an automatic decline criteria that some reviewers will ignore.) TonyBallioni (talk) 14:48, 14 March 2018 (UTC)
(edit conflict)I agree with TonyBallioni. The processes used for evaluating AfC articles is different from NPP. Also, I often find AfC articles about notable topics declined because the references aren't in the articles or they aren't formatted correctly. I don't think this a good reason to decline. I personally add references to AfC articles like these and then approve them to move out of AfC. However, not everyone wants to or should have do this, so the "Sources available" tag sounds like a good compromise. The idea about AfC is that the article should be able to stand up to an AfD and if there are reliable sources that aren't in the article, then yes, it would still stand up to an AfD. Thank you, Andrew Davidson for making me aware of this discussion and also for your kind words. Off topic, I will say, that from my experience in being open on the internet as myself is that I get less harassment than other women I know who are anonymous. I think being open says to people "I'm not afraid." It also makes me more accountable for what I say and how I treat others online, which we should all practice, anonymous or not. Thanks again, Andrew! :) Megalibrarygirl (talk) 18:01, 14 March 2018 (UTC)
  • I agree that specific criteria should be followed. I'm also of the mind that the burden of making an article acceptable for mainspace should not rest with NPP, especially if it's one that should never have been approved by AfC. We have plenty of articles in mainspace now that need work...some of which have ended-up in the NPP queue where it's not unusual to discover they were created by editors who should have known better. Another concern about not having specific criteria for AfC reviewers to follow is that, while we do our best to AGF, we may be leaving the door open for COI editors to game the system; i.e., get paid for creating stubs that NPP reviewers have to invest time in to make encyclopedic. I can understand passing articles with a few flaws if they have historic, academic or geographic importance - those are shoe-ins - but even then, if the article creator is educated enough to submit such an article, they can cite the sources, fix copyvios and the like (and/or be sent for tutoring). I don't mind mentoring editors who want to contribute and help build the project but I don't have the time to expand/fix/create encyclopedic articles from assembly-line stubs coming out of AfC. I also don't see how 500+ editors will be able to sustain it. Burn-out will happen far quicker, too. Atsme📞📧 17:59, 14 March 2018 (UTC)
  • not so sure about ACTRIAL. Now that we have no unreviewed pages indexed, and with experienced NPP/R-ers; I am not sure if we need "only auto-confirmed users can create articles" requirement. I mean, we already have revoked that right from anons. If a user creates an account, then they have to wait to get 4/10 thingy. At a given moment, around 5% articles are semi-protected; most of the templates are now semi-protected too. I think we are getting away from "an encyclopaedia that anybody can edit". When it was implemented, ACTRIAL did slow down the increase in backlog, but the huge drop in backlog was not because of ACTRIAL, it was because of the increased number of reviewers. I am beginning to think that under current circumstances, we dont need ACTRIAL, but more reviewers. —usernamekiran(talk) 20:19, 14 March 2018 (UTC)
​ The reduction wasn't just due to more reviewers. In fact, the WMFs findings indicated that we actually had less active reviewers during ACTRIAL than before (though they only analysed the first two months, before I started inviting users). While the users I invited and who joined on their own obviously helped, the period with the greatest reduction in the backlog was due to the end of year holidays (presumably reviewers with more time available), as well as the backlog drive, which involved the top reviewers working harder, as well as slightly higher numbers of reviewers. However, note that without the reduction in incoming articles afforded by ACTRIAL, none of this may have been possible. By that I mean that the total number of articles that would have been created by new editors, which were blocked by ACTRIAL is many times the amount that we reduced the backlog by. Around 300 articles per day on average were prevented, based on pre-ACTRIAL creation stats of non-autopatrolled editors; that is a whopping 54,000 article reviews that were avoided over the 6 months!(yes, seriously) Now, many of these that were avoided were amongst the easiest to review, as most of these new editor submissions were clear delete candidates (80% used to be deleted). However, this significant reduction in overall workload cannot be underestimated as the major contributing factor in the reduction of the NPP backlog. — Insertcleverphrasehere (or here) 20:49, 14 March 2018 (UTC)
The median number of deleted articles went down 227 per day. That is a massive time saving both for admins and for patrollers. Also, as I said to Morton on the research talk: yes, there are plenty of confounding variables here, and we cannot say for sure what impact they had on the backlog from the current analysis. Also, a bit earlier MER-C showed me the first article he found that got deleted after ACTRIAL ended. I'm not sure if he could dig it up and userfy as an example, but it is telling to compare the stuff that had simply not been being created at all that is now going live in mainspace. TonyBallioni (talk) 21:07, 14 March 2018 (UTC)
I'll quote the first sentence (of Special:Undelete/Shrikantarts', which was the very first article created by a non-autoconfirmed user after the trial) here: "Hello, Myself __(person's name)__ Art has always been an inspiration to me. That inspiration credits goes to my Brother and my mom, who stood up always when I started falling apart." And another "article" I deleted an hour ago was, in its entirety, "our boy tyrone is a good boy". Enough said. MER-C 21:49, 14 March 2018 (UTC)
hmmm. all valid points. but i still think if we get more reviewers, we might be able to review all the articles before indexing deadline. What happens next in the trial? I mean, is it like two sample sets? X months with ACTRIAL, X months without, X months with, and again X months without; and then compare the stats? —usernamekiran(talk) 21:53, 14 March 2018 (UTC)
We have an RfC to decide if we want to make it permanent and not a trial. TonyBallioni (talk) 21:55, 14 March 2018 (UTC)
Hopefully not, Kiran. We've already had years without for comparison. I'm supporting permanent. Atsme📞📧 00:11, 15 March 2018 (UTC)
New users are already being blocked for creating blatant spam. It looks as if turning off ACTRIAL has opened a floodgate. Kudpung กุดผึ้ง (talk) 16:37, 15 March 2018 (UTC)

The original proposal for the ACTRIAL, and the RfC dates way back in time. After becoming a reviewer, I have been trying to find out the history, but honestly speaking; understanding the circumstances of that period is not easy (especially given the fact that I was not active back then). The migration from NPR to NPP, the "ready-made" queue/backlog are another things to be considered. Points presented by Tony, and MER-C are insightful. In short, my observations so far are:

  1. ACTRIAL will take enwiki a fraction away from "encyclopaedia that anyone can edit. (the only con)
  2. It will stop blatant spam, and idiotic article.
  3. While getting/waiting to become A-CON, a user might get familiar with notability topics.
    • The short term vandals will not create stupid articles as they will have to wait a few days, and make few edits to be eligible for article creation.
    • This doesnt give any leverage over sleepers/socks, and LTAs; but creating articles is not hobby of the most of these categories.
  4. A good-faith novice user might get friendly with wiki-coding, using <ref>reference</ref> and other templates, and some basic policies.
    • In short, a lot of time of NPP/R folks will be saved from gnoming the articles by such editors.
  • In conclusion, ACTRIAL is good. I am back to supporting it. —usernamekiran(talk) 18:15, 15 March 2018 (UTC)

Three strike rule (aka 4th review is a discussion)

Insertcleverphrasehere's proposal for a three strike rule at AfC was archived without being formally closed, but my quick reading is that there is a decent amount of support. Do we have enough of a consensus to move forward with it? What would be the next steps? – Joe (talk) 16:29, 17 January 2018 (UTC)

Joe Roe probably post at WT:AFC since it concerns that mostly; have to figure out exactly how to implement it.. Galobtter (pingó mió) 17:13, 17 January 2018 (UTC)
I posted it here because this work group is interested in reforming AfC and NPP. Experience has shown that it's generally more effective to work out the details in focused discussions with a small group before opening it for wider community comment. – Joe (talk) 18:26, 17 January 2018 (UTC)
Galobtter, It could potentially be implemented as part of a change to AFCH, so that the script just automatically nominated it on the third decline. That way we wouldn't necessarily have to have a bot. As for the consensus, can we not just un-archive it and request a close? — Insertcleverphrasehere (or here) 18:37, 17 January 2018 (UTC)
A change to AFCH..seems like the best way; can request a close at WP:ANRFC Galobtter (pingó mió) 04:48, 18 January 2018 (UTC)
  • The VP is never an ideal venue for a full RfC. It's fine for testing the water though. A full RfC should be on a dedicated page related to the topic, such as, for example: Wikipedia:Articles for creation/RfC 3 strikes rule. As Joe Roe correctly points out, it is : ...effective to work out the details in focused discussions with a small group before opening it for wider community comment. That's one of the reasons why we have this project here. A good RfC has a proposal statement and supporting details carefully worked out by people who have good communication skills and are practiced creative writers. Examples of when this is not so are when RfCs tail off with out interest because they become forks of all sorts of alternative ideas. So yes, this is the place to discuss it without a lot of background noise. Kudpung กุดผึ้ง (talk) 19:20, 17 January 2018 (UTC)
I love this idea. It will cut the backlog by stopping the 4th to 15th declines, get rid of junk quicker and give notable topics a better chance of getting to mainspace. Legacypac (talk) 14:46, 12 March 2018 (UTC)
+1 - ✅ Atsme📞📧 03:54, 14 March 2018 (UTC)
  • @Kudpung. Well, I apologise for jumping the gun with it, I have a lot to learn with regards to putting together RfC proposals. It generally went well though, despite the discussion fragmenting somewhat, with a lot more support than opposition. In any case I have learned that it should be renamed to something like "Automatic notability discussion following 3 declined AfC reviews". Worded as it was as the "three strike rule" resulted in many people opposing without realizing that this proposal was meant to help good faith draft writers that were hard done by reviewers as much as it was intended to address tendentious repeat submitters. "Three strike rule" gives the impression that the proposal was simply meant to punish AfC resubmitters, but the rule would also have a positive impact on dealing with repeated declines on spurious reasoning. What should the next step be for discussing/re-proposing/implementing something like this? — Insertcleverphrasehere (or here) 04:16, 14 March 2018 (UTC)
I'm thinking that proposing it as an AfC guideline (per Slakr's closing comments) is a good idea (rather than as a bot), while also discussing with Enterprisey whether a semi-automated option to "list for discussion at MfD" could be added to the AFCH script (i.e. a button for 'list at MfD' that pops up the twinkle XfD script with a pre-loaded message, as well as a pop-up telling reviewers that they should list it for MfD notability discussion if declining for the third time). I'll also look into collecting a lot more data on exactly how many 3+ declined drafts there are in the AfC backlog as well, so that we can estimate the likely impact on MfD. — Insertcleverphrasehere (or here) 04:38, 14 March 2018 (UTC)
I also see it as giving the new editors a fair chance to get the article passed. It would also help solve a problem reviewers face - no one wants to approve a page 3 or 4 other reviewers declined - going against their opinions. With an MfD keep we can confidently move a page to mainspace without fearing some busybody will say we are moving junk in. One objection will be WP:NMFD which came about because some busybodies ran an RfC to stop cleanup efforts. We'll need to get past the crazy idea that the same editors that assess notability on AfC drafts and in AfD magically loose their Notability foo when posting in MfD. Legacypac (talk) 06:13, 14 March 2018 (UTC)

Does anyone know how to put together a sortable report of all Pending AfC drafts (2400 odd now). If we could see Draft name, date last submitted (that is sorted into categories now), and # of declines (0 to ??) that would help tremendously in assessing amd piloting this idea. Legacypac (talk) 07:06, 14 March 2018 (UTC)

Insertcleverphrase's 1/10 list will give us a good estimate. Legacypac (talk) 15:52, 14 March 2018 (UTC)
  • Update at User:Insertcleverphrasehere/AFC_stats we have randomly reviewed 72 AfC pending pages so far and 6 of those had already had 3 or more reviews. A small sample but if that 8.3% holds across the 2400 draft backlog around 200 pending pages would qualify for a The fourth review is a discussion guideline. Many of those 200 qualify for G11. Some should be mainspaced. We would not see all of the remaining pages at MfD all at once, so that seems very reasonable to handle. Legacypac (talk) 14:51, 16 March 2018 (UTC)

See User:JJMC89 bot/report/AfC decline counts for the decline counts for Category:Declined AfC submissions.​​ — JJMC89 (T·C) 02:28, 18 March 2018 (UTC)

thank-you so much. That is some very good data. Legacypac (talk) 02:45, 18 March 2018 (UTC)

Proposed AFC/MFD Guideline

When finding an AfC submission that has been declined 3 times already the reviewer should either Accept the page, seek Deletion via an appropriate CSD (usually WP:G11) or Refer the Draft to MfD for a discussion of the Draft's suitability for inclusion in mainspace or deletion. In general no AfC Draft should be declined a 4th time.

If this gets support in this working group we will take it to AfC and then MfD pages for a vote. Wording improvements are welcome.

  • Support as drafter. Legacypac (talk) 15:01, 16 March 2018 (UTC)
  • Oppose. Wrong direction of solution-finding to an agreed problem. Similar to my comment at Wikipedia:Miscellany_for_deletion/Draft:Global_Database#Draft:Global_Database, there are multiple systematic problems, and this proposed solution is part wallpapering and part punishing the author for his under-training. Before proceeding, I’d like to see all new registrants auto welcomed. Newcomers need the useful links that template includes. Ideally, IPs will be told to register to create even a draft, but until then, IPs should be auto-welcomed on their first edit.
I also note the lack of even pseudo-human communication with draft authors. They are not treated like newcomers in mainspace. I think this cultural divide is bad.
And oppose because MfD is not a suitable forum for AfC issues, where the page does not fail anything at WP:NOT or WP:UPNOT. See scope creep. This proposal will break MfD long before it fixes the problem. —SmokeyJoe (talk) 05:34, 18 March 2018 (UTC)
I've seen bot welcomes on some pages. I don't know the history of that or why it stopped. Autowelcoming is outside the scope of NPP/AFC.
I don't know how newcomers to mainspace are treated differently then those that start in draft space. It's all one project and from what I've seen checking thousands of edit logs newcomers have no problem going between editing drafts and mainspace.
Giving new pages a chance instead of stuck in AfC where each decline may increase the likelihood of another decline is not helpful to the newcomer. We are not going to break MfD - nothing has yet. If there are more noms including interesting pages potentially worth keeping we will see more participation based on what I've seen. Legacypac (talk) 05:45, 18 March 2018 (UTC)
The problem of not autowelcoming is hurting AfC using newcomers. They, many, are making pages ineptly. We need to find the place to make a noise about it.
For mainspace constructively editing newcomers, they are only spoken to when something needs saying. A talk page is used. They are addressed by username, and politely, in free text and without templating. —SmokeyJoe (talk) 06:04, 18 March 2018 (UTC)
  • Wait I think that we are close to proposing serious changes to the way that AfC aproaches declining with regards to notability. I think that the three strike rule might not be needed after that, or if it is needed, it will make a lot more sense. I suggest we approach other AfC reforms first, and come back to revisit this idea later. I've thought pretty hard on this with regards to the new revelations abut AfC that have come about due to the WMF's research as well as our own, and I think that the 3 strike rule is only trying to address a symptom of a larger issue of how AfC approaches notability. — Insertcleverphrasehere (or here) 09:50, 18 March 2018 (UTC)
  • Wait per Insertcleverphrasehere. Other developments that have implications for this idea are under development. I'm also not really comfortable with a simplistic three strikes rule as such, some drafts may legitimately need more help than just three reviews. Roger (Dodger67) (talk) 11:25, 18 March 2018 (UTC)

We will need some sort of escape valve that ends the submit-reject cycle. I've started at the top of User:JJMC89_bot/report/AfC_decline_counts to CSD or MfD the most rejected pages. Some are hanging around more than 6 months due to bot edits that keep them off the User:MusikBot/StaleDrafts/Report but they can still be sent G13. Some have been in review-reject since 2016! Eliminating these highly rejected timesink pages will help clear the backlog. Legacypac (talk) 14:43, 18 March 2018 (UTC)

Taking away the big blue "Submit" button from rejected hopeless drafts would be a good start. --SmokeyJoe (talk) 03:48, 19 March 2018 (UTC)

RfC on permanent implementation of ACTRIAL

There is currently a request for comment at Wikipedia:Autoconfirmed article creation trial/Request for comment on permanent implementation about whether or not autoconfirmed status should be required to create an article in the main space. This is a follow up to the recently ended autoconfirmed article creation trial (ACTRIAL). All are invited to participate. TonyBallioni (talk) 13:40, 19 March 2018 (UTC)

Retrieved from "https://mdwiki.org/wiki/Wikipedia_talk:The_future_of_NPP_and_AfC/Archive_6"