Wikipedia talk:Autoconfirmed article creation trial/Archive 1

Archive 1

Archive 2

Archive 3

→

Archive 5

Page reopened

I've archived the original page and its talk in their entirety as it existed before today at Wikipedia:Autoconfirmed article creation trial/2011 archive, and have unarchived the current page and removed any threaded discussion. If I have removed more than people thought should be removed, please feel free to add it back. TonyBallioni (talk) 05:25, 9 July 2017 (UTC)

Tasking suggestions

Kudpung's list

Some years ago, I sorted out a huge meandering project with 400 participants that had got stuck and was going nowhere, by splitting it up into separate tasks for development. It worked, and today we have one of our most important deletion systems.

Since the original ACTRIAL concept and the issues leading up to it, the envelope profile of the fire hose of new articles has changed significantly over the last 6 years. What we are getting now is a lot of artspam, non English articles (generally COPIVIO), and a great deal of stuff from non English language regions (probably most of it). This means that we have to consider the readabilty of interface messages and the Wizard for non native English users, perhaps even using snippets of Hindi, Malayalam, Arabic, and Russian on the first splash page.

What a lot of people don't realise, is that ACTRIAL is a lot more than simply preventing new users from creating pages in mainspace. At the time of its conception in 2011, it involved three separate elements, and these are my suggestions how I think the developmental tasking of ACTRIAL could be allocated:

ACTRIAL

• Code: Granular permissions in the MediaWiki control panel (if there is one, I’ve never done an instal of Media Wiki) - WMF
• Interface message texts
• Interface templates code (php, js?) for built-in logic (if there is any) - WMF
• Newsletter to admins
• Newsletter to New Page patrollers

Article Creation Flow/Landing page

• Interface message texts
• Interface message graphics
• Code - WMF

Article Wizard

This really needs distilling. Its walls of text are enough to put anyone off. I started to rebuild it offline years ago and fortunately, it had no programmed built in logic and was just made up of css, links, and templates for transclusion. Ideally, it should be a genuine conditional process using .js so users can’t override it once they are in it. Anyone who can write js can have a go at it once the visuals and texts are done, and the logic paths explained.

To keep it centralised, the above three elements could be the basis for separate sub (workshop) pages here. I don't think we would want it to be on one of the confusing WikiMedia sub-department or research pages. Kudpung กุดผึ้ง (talk) 10:38, 9 July 2017 (UTC)

I'd be interested in lending a hand if needed. I'm not that familiar with the back-office interfaces, but could act as a guinea pig of sorts, for any proposed GUI changes, etc. K.e.coffman (talk) 23:23, 9 July 2017 (UTC)

TonyBallioni's list

In addition to Kudpung's suggestions above, I see there as being two additional sets of questions that we need to work through as a part of implementation.

The off-wiki quesitons:

• What guidance should be given to first time editors who participate in off-wiki events? Should there be a semi-official NPP/AFC help document to show people the best ways to get involved on Wikipedia and how to get appropriate articles through?
• What guidance should be given to the coordinators of these events? Considering that this has been the most consistent objection to ACTRIAL, I think that a newsleter/guidance/FYI document should be sent out and made public.
• Is there anyway we can create a process to speed up granting the confirmed flag to these participants without debundling the right to confirm?
• The ultimate question, which I think would best be addressed in the post-ACTRIAL RfC but is worth thinking through now: should we debundle confirming accounts to account creators? I support this, but I'm sure there are objections.

The AfC questions

• What additional resources, if any, will AfC need once ACTRIAL is implemented?
• Is there a way that the tutorials and prompts can be worded to decrease the likelihood of the complete crap going to AfC?
• Should the AFC project list be deprecated and subsumed into the NPR user right?

Pinging Primefac and DGG on the AfC questions to see if they have any thoughts.

Obviously, some of these questions are things that will need to be thought through as the trial is ongoing as part of the analysis, but even those will require metrics and questions formulated beforehand. Others, such and providing guidance documents, we really should start the basic drafts of now as we are leading up to D-Day. TonyBallioni (talk) 23:37, 9 July 2017 (UTC)

• For the editathons and similar events, I am already working on a guideline. The key element is that someone experienced needs to take the responsibility of checking that topics are suitable--not just suitable, but safe, because someone's first experience here should be positive, and not involve them with topics that are going to be disputed. I and the other WM-NYC people have years of experience at this; we've been learning how to do this ever since we ran the first editathon--at least the first in the US-- eight or nine years ago. In terms of practicality, I will personally engage to do this or find somebody I trust for any event if nobody else is available.
  • But we also need a version ofthis that suggests safe topics that someone working on their own can follow.
• Someone associated with the event should be an admin, in person if possible; but at least remotely. And I would permit unbundling this yo account creators. (we already realize that an event needs someone with account creator permission--again, at least remotely. DGG ( talk ) 06:18, 10 July 2017 (UTC)

Questions/observations

Interesting stuff. Some responses from ignorant me:

1. Is there any reason why there's no draft message for people trying to create mainspace content [yet]? It's the most likely thing they'll do.
2. Why are the proposed drafts so long? Can't we cut to the chase? (It's unlikely what they're looking for has been deleted, for example)?
3. For anons, could we make the messaging firmly about encouraging them to take up accounts, how easy, free, quick it is?
4. Why did WMF sit on this for six years and why the change of heart now?

I know, easy to chirp from the sidelines. --Dweller (talk) Become old fashioned! 11:40, 18 July 2017 (UTC)

Hi Dweller, I'll try to answer what I can

1. ACTRIAL is only aimed at mainspace content. I'm assuming the messages here discusses project space because something in the coding linked to the title the message is displayed when using things such as Template:BASEPAGENAME. I could be wrong, but its my guess.
2. I was about to say that we actually get a fair amount of CSD recreations, but ideally this wouldn't be the case because the pages won't have been created to begin with. I agree that keeping things short and sweet. T170354 is the task that will create the new landing page for new users trying to create content, so this is something that we should keep in mind for that and work on before the rollout. Your input would be helpful on it :)
3. I think this is a good idea and is probably something that could even be improved in the current live message for anons.
4. Vetoed in 2011 for from what I can tell were philosophical reasons by the foundation. As recent conversations on-wiki built up about flipping the switch ourselves without WMF support (something that is feasible through the blacklist), the Foundation released Wikipedia:New pages patrol/Analysis and proposal, which went over like a lead ballon. As a part of the conversations following the release of that report, MusikAnimal compiled this dataset looking at the deletion rates of pages created by new users (~80% get deleted or redirected within 90 days). The reaction to their suggested solutions and MusikAnimal's numbers made the Foundation reconsider this as a research experiment. (If I got anything wrong in this account, I'm sure someone will correct me).

I hope this was helpful. TonyBallioni (talk) 13:29, 18 July 2017 (UTC)

Thanks. I'd be interested in helping craft the messages to newbies and anons. I'm guessing Phabricator isn't the venue for copyediting. If/when there's a location, please drop me a line with it! --Dweller (talk) Become old fashioned! 13:38, 18 July 2017 (UTC)

Why should we trust the WMF when we have been lied to before?

Back in 2011 we reached a consensus for a two-month trial of Pending changes with a fixed ending date. Those who supported the trial trusted those who were to run the trial to do what they said they would do and end it on that date. The ending date was ignored. In response to this, there was a new discussion and a new consensus for a "we really mean it this time" drop-dead date. This consensus was also completely ignored. Then a major RfC was closed with a clear consensus for two week deadline for removal from all articles -- which was then ignored. Consensus was ignored and that trust was betrayed.

Then we saw an admin get blocked for following the clear consensus and removing pending changes from articles. If that doesn't discourage admins from following consensus, I don't know what will. As far as I can tell this block had no repercussions.

Consider the following comments by other editors from back in 2011 who also felt that this harmed Wikipedia:

Comments by other editors

>"Many editors consider the refusal to end the trial on the date promised as a breach of trust" >"It's impossible to assume good faith when a past assurance continues not to be honored (i.e. begs the question: How can we trust you to honor the results of discussions if you're not honoring the results of a previous one?)" >"The main reason the trial got in in the first place was due to users supporting with the understanding that it would be removed, and only because it would be removed, saying they wanted to try it out. If we don't keep the promise of turning features off after the trial, this factor will be lost for future trial proposals" >"The poll that produced the original consensus to turn the feature on was for a trial with a specified end date. In the absence of any consensus to make the feature permanent or start another trial the feature should be removed from articles. Failing to do this has damaged the credibility of any future software trial proposals." >"I disagree that this is just removal solely for the sake of making a point. This is making good on the original agreement that the trial would end, by the end of 2010 for the last agreement. Anything beyond that wasn't approved, it's that simple. In the absence of any community agreement to do anything else this is the default option and the one we must follow. The only way around that is to totally ignore the original agreement, which totally goes against the whole concept of consensus." >"This is necessary to deal with negative feelings about being lied to. I must say I have trouble with those myself and feel a strong irrational urge to oppose to everything related to pending changes." >"The only consensus was for a fixed-term trial, with a clear expectation that if no further consensus arose then we would revert to the status quo ante. We need to deliver on that promise, to retain credibility for future trials in other areas." >"If 'trial' comes to mean 'turned on indefinitely', no-one else will get consensus to trial other new ideas in future." >"WP:IAR doesn't excuse conscious deceit. The extension of this trial beyond two months was a betrayal, not 'ignoring rules to make a better encyclopedia'... The decision to go back on one's word isn't something that is done by accident or mistake." >"I increasingly feel this debate has become about something much more important than pending changes. It's become about good faith. A sizable portion of the editor base clearly feels that without a clear consensus to continue the pending changes trial that the original commitment to end the pending changes trial after two months should have been upheld. ... Wikipedia is already hurting in recruiting and retaining editors, and cannot afford to reach a point where change and compromise has become impossible because of distrust."

The above quotes clearly show that harm was done. The WMF cannot unring that bell. Here it is, seven years later, and we still lack a firm and clear published policy that promising to try something for a limited amount of time and then breaking that promise will never again be tolerated on Wikipedia That is the bare minimum required to start to regain the editor's trust. Years later we still have no apology, no written policy, and no assurance that consensus will not be ignored again.

I like the idea of an autoconfirmed article creation trial, but I have little confidence that the WMF will actually allow the trial to take place and zero confidence that if it does that the trial will end when they say it will. Once burned, twice shy. --Guy Macon (talk) 05:29, 19 July 2017 (UTC)

• The trial now has the public support of Jimmy Wales, which I doubt he would have made without checking with someone at the WMF. I think that is a sign that this is going to happen with WMF support. It also has the public support of Doc James, in his capacity as a volunteer editor, but I would hope he would also represent the interests of this trial with the WMF as well. If the trial doesn't have WMF support in implementing, there is clear community support for flipping the switch and we can do it on our own. Additionally, an office action opposing would it likely be a PR disaster on par with KnowledgeEngine or Super Protect. Combine that with my belief that people typically mean what they say, I don't believe the WMF is trying to screw us over here. As I have said on multiple occasions since Danny Horn's post announcing that they were willing to try this: working with the resources of the WMF is much better than working against them. This is a community driven initiative, but their support helps us a lot on multiple fronts.
  The dates still needs to be worked out. The consensus at the 2011 RfC and recently at NPPAFC was six months. The WMF doesn't want that, but I think the argument that we can shut it down if the consensus is that things are starting to burn down around us is strong. If it gets shut down early without community consensus, we will still have the data to look at and can hold an RfC to flip the switch locally after the trial, which is the current plan anyway. Working with the WMF at this time costs us nothing and gains us quite a lot. While I do get the frustration from all the mistakes in the past, at this point they've agreed to at least 80% of what we've asked for, and we are talking about the remaining 20%. I think its time we give them a chance to show that they are listening to the community. TonyBallioni (talk) 06:03, 19 July 2017 (UTC)

(edit conflict) The trial will take place, and following the conditions set by the the RfCs that governed it. Those RfCs have not lost their validity any more that all the other new policies, guidelines, and practices that have been agreed upon by consensus have in any way expired. I rekindled these discussions (or tried to) when I was in Esino last year when my proposed talk had been usurped by the WMF, and I was given a brush off when I tried to disucss it in the street with the WMF. Since then, I have been relentless in pushing for this solution. I think we now hold a trump card (see the red comment at the top of my talk page) - accredited new Page Reviewers are giving up on their task in the face of the sheer deluge of the unmitigated rubbish and spam that arrives every day in quantities they just can't manage. Of the the rest of those who can still tag new pages (but not makrk them as patrolled), far too many of them lack the required knowldge or competency and they are biting bona fidae new users and letting unwanted content slip through. Backed by the stats that have recently been published by their own staff, the WMF realises that it's now or never if they want to save the reputation of the encyclopedia for quality and show the paid spammers that we are not LinkedIn or the Yellow Pages. They are also engaging the services of a professional statistitian to evaluate the resukts. If the WMF screws up this time, there will be a community mutiny and I won't be able to stop it. Kudpung กุดผึ้ง (talk) 06:12, 19 July 2017 (UTC)

Sounds like a perfectly reasonable trial. We are not a new start-up anymore and I think everyone now realizes this. We need to shift our focus from quantity to quality and adapt to the new reality in which we have a lot of people trying to bend Wikipedia's good name to their or their companies financial benefit. While "anyone can edit" at least some of Wikipedia and at least initially over a third of WP's top 5,000 medical articles have some form of protection.[1]

User:Guy Macon I was under the impression that PC had sufficient community support to continue? I do not apply it to medical pages as I find it too slow and clunky but that is just me. I did not think the WMF was pushing it as a solution? Doc James (talk · contribs · email) 07:30, 19 July 2017 (UTC)

That was later. We even have a template to make it easy to follow the history:

There never was a strong consensus against PC Changes itself. What got everyone upset was the broken promise. Wikipedia:Pending changes/Trial and Wikipedia talk:Flagged protection and patrolled revisions/Poll showed clear support for a two-month trial. Wikipedia talk:Flagged protection and patrolled revisions/Poll#what was and was not approved shows the first signs of people feeling that they were deceived.

Then we had Wikipedia:Pending changes/Request for Comment February 2011/Archive 3 which said "Pending changes was originally intended to be instituted on a "trial basis" and reviewed later. The trial period has been over for some time. Admins are advised in several places to use it with caution and not do anything drastic, but to date there is still no clear policy on its continued use. This RfC is intended to settle the matter of whether PC should remain activated or be disabled" Note that this was five months after the "two-month trial" started in September 2009

On 21 May 2011 Administrator Kww was blocked for implementing the clear consensus of the above RfCs: [2]

In Wikipedia:Pending changes/Request for Comment 2012 we finally got consensus to implement PC changes for longer that the mythical "two-month trial", and by all accounts PC protection has been a success.

The issue here isn't whether PC protection is good. The issue is that we were lied to when we were asked to !vote on a two-month trial. Which brings up the obvious question; we were lied to then, so how do we know that we aren't being lied to this time? --Guy Macon (talk) 09:28, 19 July 2017 (UTC)

The block of Kww was carried out by a community member. It was thankfully quickly overturned.

Is this not more the community not following through on what the community initially agreed to?

Or was the expectation that rather than the community simply not using the tool until further consensus had developed that the WMF was to remove it?

Doc James (talk · contribs · email) 09:39, 19 July 2017 (UTC)

Guy Macon is right to bring this up, but Doc James is correct that there were serious failings in the community as well. There wasn't strong consensus for or against PC; the community was divided. Over time, opponents of PC were worn down and dropped out of the fight, and consensus emerged to implement it. But at the time of the trial period's end, as I recall, there was still a widely held expectation that the agreed-upon terms of the trial would be honored. Unfortunately, there was no shortage of Wikipedians who were only too happy to support short-circuiting those terms and leaving PC in the admin toolkit before any final decision had been rendered. Unlike the WMF and developers' refusal to respect community consensus on implementing the autoconfirmed trial, the PC trial debacle was especially toxic because a significant portion of the community itself wasn't respecting consensus. Thankfully, that's in the distant past now. Rivertorch^FIRE_WATER 16:43, 19 July 2017 (UTC)

All good points. To answer the OP's question: we don't have a lot of other options. We could implement ACTRIAL ourselves, and we will have to if the WMF drags their feet any more than they already have in the past six years; but for now, they are offering to help. One way to judge their intent is to see if they try to put too many tasks on a critical path when those tasks could obviously be worked on in parallel. This has already been raised as a concern.- MrX 17:47, 19 July 2017 (UTC)

What's happening right now on ACTRIAL is that we're contracting with a researcher to help set up the trial -- User:Nettrom, who runs Suggest Bot and has been doing research on Wikipedia content production for several years. He's currently building a Research page on Meta: Research:Autoconfirmed article creation trial. He's building a first list of research questions, which will then be discussed by all of us here, so that everyone has a say in establishing what we're looking for. The list will be guided by the original list of questions here, as well as MusikAnimal's stats, and Morten will also include some questions that wouldn't have been technically feasible back in 2011, but are potential data sources that we can use now.

Kaldari and I have been working on setting up the necessary tech to make the trial work. MaxSem from the Community Tech team is working on a stripped-down version of the ArticleCreationWorkflow extension that will redirect non-AC users from non-existent pages to a special page that we'll create. (You can follow that ticket at phab:T170354.) The content and functions on that page will be determined by everyone here, including the researcher, who will probably have some suggestions based on what would give us useful information. We were also talking yesterday about how to make this trial period work for editathons; that conversation is above.

I think over the last two years, the Community Tech team has established a good record of doing what we say we're going to do, but I understand that doesn't erase the anger and frustration that people feel about WMF's bad faith actions in the past. -- DannyH (WMF) (talk) 19:13, 19 July 2017 (UTC)

DannyH (WMF), WADR, I think that is wishful thinking. There is possibky no doubt that you have done what you said you would do, but what you have also done is relagated a top level critical issue to a wish list on non-urgent gadgets and convenience features, your colleagues from the WMF have mailed me 'muzzling' orders, my talk in Esino was thrown out - wasting the WMF's money on the scholarship to send me there, and effectively making it impossible for me to go to Montreal next month to sit round a table with you guys for another chance to talk about this sensibly and maturely. Kudpung กุดผึ้ง (talk) 03:23, 21 July 2017 (UTC)

discussion on meta

Some questions have been raised at | meta concerning the effects of actrial on the retention on new editors that participants in this talk page might be interested in. The specific question is Currently, among non-autoconfirmed users who create articles, what percentage will still be active editors after one and two months? I think that omits an important qualitative aspect of measuring the contributions of non-autoconfirmed users. I care not just that we retain new editors, but also that they make constructive contributions. Mduvekot (talk) 01:29, 24 July 2017 (UTC)

Pinging JMatazzoni (WMF) to let him know of this comment here. TonyBallioni (talk) 16:52, 24 July 2017 (UTC)

The way our hypotheses are currently phrased, the vast majority of those who stick around are constructive members of the community. Do we have reasons to expect that not being the case? In other words, if someone's around a month or two after registration are they likely to still be making non-constructive edits? I also followed up the conversation on meta and mentioned that we might want to look into the productivity of new editors (i.e. the value that they add to the encyclopedia), meaning that I'm also interested in understanding the quality of their contributions. Regards, Nettrom (talk) 18:06, 24 July 2017 (UTC)

I think that Aaron has talked about an "edit survival" measure, which looks at how much text in an edit is still visible on the page after x amount of time. But I can't remember where that was, and I can't find it now, so I may have imagined it. :) But I think that's one way that you can measure the "quality" of an edit -- whether it lasts or not. At the most basic level, we could look at whether the edit gets reverted or not. That's a low bar for quality, but it's something. -- DannyH (WMF) (talk) 18:22, 24 July 2017 (UTC)

Yeah, "persistent word revision" (see Content persistence) is a typical approach to this. Looking at revert proportions could be a decent proxy given that we're interested in "constructive edits", and arguably edits that get reverted are not constructive (although the reasons for why they get reverted can differ greatly, of course). Cheers, Nettrom (talk) 19:41, 24 July 2017 (UTC)

Research progress

Hi, just want to give an update on the research process. Our researcher, Morten, is working on the research questions for the trial: see the Research page. It's largely based on the questions created in 2011, with some additions.

There are three basic areas that we're interested in:

• the impact on new accounts -- how many non-autoconfirmed users make edits, how many reach confirmed status
• the impact on the quality assurance process -- the backlog of New Page Patrol and Articles for Creation, the survival rate of new articles, reasons for deletion
• the impact on content quality -- more good pages and good edits, fewer bad pages and bad edits

This is still a draft -- it's not complete, and Morten is going to incorporate some feedback we got from other Foundation researchers this week. Still, it's interesting to look at and think about; please feel free to take a look and make suggestions or ask questions on the talk page.

Meanwhile, there's ongoing work on the extension that's going to redirect non-autoconfirmed users to the landing page -- if you want to check the progress, you can look at phab:T170354. There's also the discussion above about what to put on that landing page. I think that's everything going on right now, but let me know if I forgot something. -- DannyH (WMF) (talk) 20:58, 21 July 2017 (UTC)

Also, I would like to mention that Morten (aka Nettrom) is currently working on this project as a community volunteer. We are still in the process of securing funding to hire him as a contractor, which will hopefully be accomplished soon. Kaldari (talk) 17:54, 23 July 2017 (UTC)

We've got an update on the contract situation that Kaldari mentioned above. :) Morten's been working as a volunteer for a couple weeks to get started; he's going to be on contract starting on Monday, July 31st. So we've got dedicated research support for ACTRIAL, which is good news. -- DannyH (WMF) (talk) 21:39, 26 July 2017 (UTC)

Timeline

@DannyH (WMF) and Kaldari: do we have an update on what the timeline for launch is here? Also, any update on updating the graphics of the landing page? I can see if there are some editors who are willing to help with making the landing page look current, but it would be great if we could get WMF communications support on this. TonyBallioni (talk) 02:12, 3 August 2017 (UTC)

See the research discussion Theyv'e shelved the landing page project again, and this is just a draft of what they think they might like to do. It would be helpful if people like JMatazzoni (WMF) would engage with the community. They even think that 2 weeks will be enough for the trial. DannyH (WMF) and Kaldari are doing their best but seem to have their hands tied - people higher up are new and don't seem to be aware of what is at stake. It seems as if Halfaker is the only one actually doing anything at the moment. They'll all be off to Montreal in a few days and that will delay things for another couple of weeks. We still need to keep our options warm for a local roll out of ACTRIAL.Kudpung กุดผึ้ง (talk) 04:56, 3 August 2017 (UTC)

@Kudpung and TonyBallioni: We're still working on a scaled back version of the landing page project. See for example, https://gerrit.wikimedia.org/r/#/c/365197/ and https://gerrit.wikimedia.org/r/#/c/369592/. We're also currently working on finalizing the research questions and would love feedback at Research:ACTRIAL. (So far it's been surprisingly quiet.) I think Nettrom will be sharing his thoughts about trial duration within the next few days and then we can discuss that further during Wikimania. We're currently shooting for beginning the trial some time in September. Have you guys thought any more about potential improvements to the Article Wizard? It would be great to get that in place before the trial starts. Kaldari (talk) 07:44, 3 August 2017 (UTC)

I see you both just added comments at Research:ACTRIAL. Thanks! Kaldari (talk) 07:56, 3 August 2017 (UTC)

Hi Ryan, thanks for your prompt reply. From what we have understood, the WMF is not too keen on the landing page idea, although I started already reducing the walls of text that go with the Wizard, I guess I still have about 15 - 20 hours work on them. I am not sufficiently privileged to go to Montreal. Kudpung กุดผึ้ง (talk) 08:54, 3 August 2017 (UTC)

That's OK, we can keep discussing things here. Kaldari (talk) 15:28, 3 August 2017 (UTC)

TonyBallioni, the designer that I talked to said he would take a look, but that's all so far. I just pinged him again. Everyone's a little scattered just before Wikimania, so I'm not sure if I'll get anything useful from him before then. Community Tech doesn't have a designer, so it's kind of "beg-and-borrow" for us, unfortunately. If there are people in the community that could help, that would be great.

Kudpung, there's no problems with moving ahead on ACTRIAL; everyone at WMF is very supportive of us working on this, up to the highest levels. It doesn't stretch to a designer, unfortunately, but they're paying for Morten to work on this project full-time. We did a presentation for staff yesterday on our progress with ACTRIAL, and people were really interested. As Ryan said, the landing page is currently being built. The reason why the current landing page says "draft" is because we want to get more ideas and feedback from you (and other people in NPP). :) I saw your detailed comments on the Research talk page, that'll help a lot.

I'm sorry that you can't make it to Wikimania, I hope we can meet some people who are involved in New Page Patrol there. Would you like to set up some time to talk? -- DannyH (WMF) (talk) 17:45, 3 August 2017 (UTC)

Yes of course. Mail me when. Kudpung กุดผึ้ง (talk) 15:14, 4 August 2017 (UTC)

Is 2011 community consensus still operational 6 years later?

Off topic discussion

I have trouble accepting that a community consensus established 6 years ago should hold without any discussion today. Shouldn't we at least make sure that currently active members of the community support the 2011 consensus? --The Cunctator (talk) This was discussed at both WP:VPP and User talk:Jimbo Wales recently. There is broad consensus that the 2011 consensus to go forward with a limited trial period is still the wishes of the community and that we do not need to reassess it for a limited trial. A followup RfC will be held after the trial period to determine whether or not the changes should be made permenant. TonyBallioni (talk) 13:50, 1 August 2017 (UTC) Where are the specific links to that discussion? It would be very helpful to have this explanation on the project page. I'm not sure why you reverted my clarifying edits but I trust you had good reasons. Obviously it wasn't clear to me, despite your dismissal of my confusion. I'll trust that you'll do the edits to make the project page clearer since it's evident that you believe I shouldn't touch it. Thanks. --The Cunctator (talk) 17:18, 1 August 2017 (UTC) The Cunctator: the Jimbo-talk discussion was the main one: it was crossposted at VPP as well. This diff is Jimmy Wales' response to the inquiry and discussion on his talk page. It is linked from the introduction in WP:ACTRIAL. The ongoing discussion was in the thread above. You can find further discussion at WT:NPPAFC, WT:NPR, and their respective archives. In regards to the revert and the ping in the above section: I reverted your edits because I felt that they actually made the discussion harder to follow and less clear. I'm very open to tweaks for clarity, but I think that your version actually made it significantly more difficult for anyone who doesn't want to get into the weeds to follow. I also tried to take into account your some of your concerns by removing the specific dates and making minor tweaks. TonyBallioni (talk) 17:33, 1 August 2017 (UTC) The principle that Wikipedia has always operated on (to the best of my knowledge) is that an established consensus is accepted as remaining in force until such time as a fresh discussion establishes a new consensus. All of our policies and guidelines are established through consensus, and surely you're not going to suggest that we have to re-establish those consensuses every five years – or whatever time period – simply because the existing consensus is that many years old? I don't believe we have ever operated a principle of "consensus decay". --RexxS (talk) 18:01, 1 August 2017 (UTC) There's a difference between an active policy which has been operating consensus for 6 years, and whose active use over that time has added to its consensus. Every time someone operates at Wikipedia in alignment with agreed-upon policy it's like another vote in favor of continuing that policy. That's an active consensus. This was a discussion that had a vote 6 years ago, then was overruled by the WMF (with no explanation) that then everyone let die; it's a moribund consensus and not an active one. No one has done anything with this for 6 years, so it's a perfectly valid question about re-establishing a new discussion. Wikipedia editors have been using consensus policies and guidelines for the past 6 years even if they weren't around for their establishment, and that use is a form of consensus. There's a LARGE part of our community who has no awareness of the earlier discussion, since it's not been an active policy, but a moribund one. We should not dismiss requests for a new discussion for that reason, because THIS is a very different situation than what you are talking about. --Jayron32 15:47, 3 August 2017 (UTC) This shooting inside the tent isn't helping. Wikipedians have struggled for years with the pipeline of nonsense that lands here every minute and now a couple admins think it's wise to re-litigate the case. If you want to have an academic conversation about how consensus is judged to endure, do it somewhere else. This project is to restrict the flood of new articles. If you disagree with that goal, then just say so. You're welcome to voice your opinion. There was consensus years ago and only WMF's intransigence has delayed us. The problem has only gotten worse, so modern complaints about the worthiness of consensus ring hollow. Chris Troutman (talk) 16:40, 3 August 2017 (UTC) @Jayron32: You're responding to a question that hasn't been asked. Nobody is suggesting that a fresh debate to examine the consensus established six years ago can't happen. The objection is to these suggestions that it must happen because the consensus has somehow "expired". I reject your concept of "active consensus" as something that has no basis in policy or practice on Wikipedia. Consensus is consensus, no matter its age, until it is changed to something different. Your model of a consensus decaying if it is not actively in use raises far too many issues about appropriate timescales and subjective judgements, for no gain whatsoever. --RexxS (talk) 12:25, 5 August 2017 (UTC) A better question is "Is the logic of six years ago still sound?". One thing that is changing with time is institutional inertia. It has taken a lot of stirring to revitalise this, there has been no shortage of discussion, notwithstanding a number of people who failed to notice. To be sure, this should be added to WP:CENT. Has it not? Has it not been advertised there due to it being a long simmering discussion, and CENT tends to focus on storms? --SmokeyJoe (talk) 12:53, 3 August 2017 (UTC) CENT is exclusively for listing debates. The debate on ACTRIAL is over. See the post from RexxS above. He didn't ask a question. There has been no simmering discussion. The Foundation has taken the relay. The trial goes ahead, and this page is for them to keep us up to date with how it will be coded for us and how they will extract the data for the stats for us. The next debate will be about the results of the trial and whether or not to implement on a permanent basis.Kudpung กุดผึ้ง (talk) 13:30, 3 August 2017 (UTC) Right now we are in a holding pattern in regards to waiting for the Foundation to to the technical work here. The research questions and comments on the landing page are the things that the community could most comment on. I just posted a link to the meta discussion and here at VPP. This page has been linked there before, but been mostly ignored by the larger community. I think its a symptom of the community enjoying the discussion of what should be done, and then not as any people wanting to be a part in the implementing of the solution (which is fine and understandable). TonyBallioni (talk) 13:53, 3 August 2017 (UTC)

The summary text

It is extraordinarily confusing to have the 2011 trial language be what is being used for the 2017 proposed trial description up top. There needs to be more clarification of the summary text up top. --The Cunctator (talk) 13:32, 1 August 2017 (UTC)

1. I've reverted your edits because I think the summary in the lede made sense. I have made some tweaks as well. The 2011 trial is what is being implemented. This is not a new idea and documenting the history of it for people who are new to the conversation is important.

TonyBallioni (talk) 14:00, 1 August 2017 (UTC)

TonyBallioni Generally a good rule of thumb for whether an explanation by person A is clear to person B is whether person B believes it to be clear, not whether person A believes it to be so. I honestly did not find the summary clear - you telling me that it is clear to you doesn't change my experience. --The Cunctator (talk) 17:24, 1 August 2017 (UTC)

WADR, I think The Cunctator may wish to get up to speed - there was no trial in 2011. Kudpung กุดผึ้ง (talk) 04:05, 3 August 2017 (UTC)

Kudpung: Yes, I do wish to get up to speed. Thank you for hitting the nail on the head. I understand that there wasn't a 2011 trial. The summary language on this page are confusing and unclear, which was prompted that original question. My attempts to edit the summary description to make that more clear were reverted and I was told that the summary language was clear. I'm reasonably intelligent - if it's difficult for me to get up to speed, it's a safe bet it's difficult for others to do so as well. The summary is now better, but it's still confusing and I believe makes it difficult for people to get up to speed, which I'm sure you agree is a desirable outcome. --The Cunctator (talk) 15:04, 4 August 2017 (UTC)

The Cunctator, I reverted again because your edits make the history here less clear. I am all for mopping up any language that is confusing, but I don't thin the changes you made do that: what in particular do you find confusing so we can tweak the language? TonyBallioni (talk) 15:23, 4 August 2017 (UTC)

TonyBallioni, the summary text is a poorly organized admixture of unnecessary detail, missing context, and defensive special pleading. --The Cunctator (talk) 15:34, 8 August 2017 (UTC)

It remains largely the same as it was in 2011, because this is the same project. It has been updated to include information from the recent discussions, including links to the relevant discussion pages, which is appropriate so that people who are unaware of the lengthy discussions that have been ongoing here can find them. As I have said, I'm more than open to discussing specifics about how to make the summary more clear, but I quite strongly feel that the edits you made make the summary much less comprehensible and useful to someone who is unfamiliar with the project. If the goal is to provide clear context, removing the context to the body that most people won't read does not serve that goal. TonyBallioni (talk) 15:40, 8 August 2017 (UTC)

What goes where?

What content should go on the project page here and what should go on the Research page on Meta? --The Cunctator (talk) 15:19, 4 August 2017 (UTC)

The meta page is for discussion concerning the statistical analysis. This page is for discussion on how to roll out the project on en.Wiki and deals with things regarding that such as technical issues, design of the landing page, informing the community, etc. TonyBallioni (talk) 15:27, 4 August 2017 (UTC)

(edit conflict) :The Cunctator, I can understand if you feel it a bit confusing here. You've rather jumped into the deep end first. This page is actually about organising the technicalities of what is going to take place. To learn more, you might wish to read through the threads at WT:NPPAFC and then try to figure out what's going on on the page Danny linked you to. In the meantime, you may wish to check out the new-ish system for patrolling new pages at Special:NewPagesFeed, but you'll need to read the tutorial at Wikipedia:Page Curation, then spend an hour patrolling some pages. You're an admin, so you automatically have the New Page Reviewer right. Read also the detailed instructions at WP:NPP and if there's anything you don't understand, blame me - I wrote most of it. Kudpung กุดผึ้ง (talk) 15:28, 4 August 2017 (UTC)

Kudpung Thanks - it would be great for a link to WT:NPPAFC be added to the summary, since I can't edit the summary. --The Cunctator (talk) 16:20, 8 August 2017 (UTC)

The Cunctator done. Feel free to move it to a location you want in the summary. I thought the nutshell would be best, but don't have a strong opinion on it. Thanks for pointing this out. TonyBallioni (talk) 16:31, 8 August 2017 (UTC)

Stats discussion

Discussion about the questions and suggestions of new ones and what should be updated since the 2011 version should go here. TonyBallioni (talk) 05:16, 9 July 2017 (UTC)

• It all looks good and makes sense to me. We don't want to restrict non-autoconfirmed users from editing, and thereby creating, their own talk pages as the {{helpme}} method will be inaccessible so help from the MediaWiki development team is clearly necessary. The question now is how it will actually be implemented. Once we have gained consensus on the specifics, will we submit this to the developers? If so, through what channel and what are the practicalities of the submission? DrStrauss talk 12:53, 9 July 2017 (UTC)

DrStrauss, that conversation has been/is taking place at WT:NPPAFC. The Wikimedia Foundation has agreed to assist us in implementing ACTRIAL. There are three options for doing so, the easiest of which being doing it from the blacklist, which we can do without assistance from the WMF if needed. The other two options are updating the software to automatically redirect to the article wizard (which I oppose), or to update the 2011 code. Updating the 2011 code would probably be the best long term, but would also take the most time to roll outTonyBallioni (talk) 14:42, 9 July 2017 (UTC)

• @Kudpung and TonyBallioni:Another interesting metric would be to probably consider the number of AFC acceptances(month-wise) in the time-span of the ACTRIAL with the number of articles by non-confirmed users that currently enters WP on a monthly-basis.That will give a rough idea about how many of the supposedly good creations get lost in the process.Winged Blades ^Godric 14:27, 9 July 2017 (UTC)
  • Godric, good point. I think that might add some conflating issues because AfC tends to also be chronically backlogged (just like we are), and I don't expect their acceptances will go up unless they get more people (which is an argument for combining NPP and AfC).
    • @TonyBallioni::--Yeah that's a pretty good point.I don't think we will be able the bring down the AFC backlog to near-zero values before the initiation of ACTRIAL.Winged Blades ^Godric 16:38, 9 July 2017 (UTC)

• For the tasks involved, please see the talk page.

For the stats required, I'm updating a list I made in 2011, just give me a chance to post it. It's not prescriptive and anyone can change it or add to it, such as the important suggestion by Winged Blades of Godric. Kudpung กุดผึ้ง (talk) 21:44, 9 July 2017 (UTC)

• More metrics: Does the trial result in a in/decrease in the amount of junk pages being created in other namespaces (userspace, "articles" in the Wikipedia namespace, spam drafts, etc)? Is there a net decrease in the amount of crap?

I'm also interested in whether users try to game the system, especially using typical sockpuppet techniques. MER-C 04:31, 16 July 2017 (UTC)

@MER-C:--How do we define junk?While identifying junk is easy in mainspace(supposedly CSDable etc.), the evaluation procedure is difficult to implement in user/draft space.Winged Blades ^Godric 11:30, 16 July 2017 (UTC)

This is fairly straight forward: at a minimum it should include stuff that gets deleted, vanity pages, social media profiles, CVs, abandoned drafts of hopelessly non-notable subjects, stuff that fails WP:NOTWEBHOST, tendentiously resubmitted drafts. The giveaway is the zero mainspace edit count in hindsight. MER-C 12:20, 16 July 2017 (UTC)

The "effects on articles" questions focus on the article survivability, which doesn't tell the whole story. Many new articles that should be deleted are slapped with a maintenance tag and remain untouched for years. I suggest adding the following research question:

• Does the number of new articles added to maintenance categories, such as Unclear notability, CAT:COI, and Promotional, change?
• The prediction is an across the board decrease in both absolute and relative numbers.

Rentier (talk) 11:55, 20 July 2017 (UTC)

Newsletters

Announcements were sent as part of the August 2017 NPR newsletter and September 2017 Admin Newsletter

NPR announcement draft Hello Autoconfirmed article creation trial, thank you for your efforts reviewing new pages! This is to alert you of an upcoming change to our content creation system for new users. In June 2017 the Wikimedia Foundation agreed to implement the consensus of a 2011 discussion to restrict creation of new articles in the article main space to accounts that are autoconfirmed. This change is on a trial basis only and will run from 14 September 2017 to 14 March 2018. Afterwards, there will be a period where new users can create content in the article main space again. The Wikimedia Foundation will then provide us with data and statistics that will be provided to the community in a follow-up RfC. What this means for you: There will probably be fewer pages created in the new pages feed every day to review. This will free up reviewers to look at other pages and work on the current backlog of XX,XXX. Articles for Creation may experience an increase in articles during this time. If you are interested in working with new users to develop content, consider volunteering there as well. Your input in the conversation on how this trial is going will be needed both during and after the trial. You can join the conversation at WP:ACTRIAL or WP:NPPAFC. If you wish to opt-out of future mailings, go here. Admin announcement draft Hello Autoconfirmed article creation trial, thank you for your service to the English Wikipedia as an admin! This is to alert you of an upcoming change to our content creation system for new users. In June 2017 the Wikimedia Foundation agreed to implement the consensus of a 2011 discussion to restrict creation of new articles in the article main space to accounts that are autoconfirmed. This change is on a trial basis only and will run from 14 September 2017 to 14 March 2018. Afterwards, there will be a period where new users can create content in the article main space again. The Wikimedia Foundation will then provide us with data and statistics that will be provided to the community in a follow-up RfC. What this means for you: There will probably be fewer pages created in the new pages feed every day to review. This will free up reviewers to look at other pages and work on the current backlog of XX,XXX. Speedy deletion nominations are expected to decline. More admins will be needed to review WP:PERM to grant confirmed status to editors who are taking place in editathons. Editathon coordinators have been asked to place a notice at WP:AN in advance of an editathon to seek admin support in coordinating these efforts. Articles for Creation may experience an increase in articles during this time. If you are interested in working with new users to develop content, consider volunteering there as well. Your input in the conversation on how this trial is going will be needed both during and after the trial. You can join the conversation at WP:ACTRIAL or WP:NPPAFC. If you wish to opt-out of future mailings, go here. Newsletter discussion @Kudpung, DGG, Majora, Robert McClenon, MER-C, Winged Blades of Godric, and Scottywong: See the drafts above of the announcement newsletter mentioned on the talk page list. Hopefully the wording here and the commitment to this as a trial with a followup RfC that will be sent to 1600 users will help with Majora's concerns that it might be made permanent without an RfC and that the trial would not stop. Feedback on other wording, copyeditting, etc is appreciated as well. TonyBallioni (talk) 23:22, 16 July 2017 (UTC) That looks fine to me. Although you may want to also post a version of this to the talk pages of the Teahouse and the help desk. As I'm sure they are going to be hit with a lot of questions. I've already notified a few people who help out in the IRC help channel that this is coming as well. So notifying them of a date as well would also be extremely helpful. --Majora (talk) 23:39, 16 July 2017 (UTC) The date isn't set yet. As soon as it is we'll be sure to let anyone you think would be interested know. I appreciate your informing people, and I really would like for more people to come here to be a part of the conversation on how to set it up so it address concerns: those of us who think this is the best way forward do take those concerns seriously, and I think if you look at all the places that this has been discussed you'll see that. I'll also ping Cameron11598 since I know he commented (and I think he's involved in -help). I'll post this at the Jimbo talk as well, but I promise that I have no intent of sneaking this in as a permanent change without an RfC and I know Kudpung and Scottywong have also made that exceptionally clear to some people who want to make it permanent immediately. Hopefully if we make that clear on as many places as possible people will realize that this really is a trial. TonyBallioni (talk) 23:50, 16 July 2017 (UTC) Hey TonyBallioni thanks for looping me in. I'm sure we will get a ton of questions about this in -help on IRC. Would it be possible to get a copy of this notice that we can make into a short link to place in -helpers so our IRC volunteers know whats going on? Thanks! --Cameron11598 (Talk) 23:55, 16 July 2017 (UTC) Cameron, I wouldn't want this to be sent out yet: currently we are in the stages of working out with the WMF what exactly the trial will look like. Right now we are waiting for the coding to happen and for the WMF to get funding for a data analyst. Part of the reason that the exact date is unknown is because the preferred method for this will also involve a simplified landing page for new users that will help them find what they want to do. That involves updating old code to match with new code. You can read more at WP:NPPAFC and WT:NPR and their archives, and also join in the conversation there and here. Those are very long, but it contains a lot of the background discussion. A much quicker read that will also put the desire for this in context is User:MusikAnimal (WMF)/NPP analysis. I'm also available to answer any questions on my talk page or anywhere else you might want to ask them :) TonyBallioni (talk) 00:05, 17 July 2017 (UTC) Ideally I'd like some time to get those helpers on IRC familiar with what kind of errors/edit notices new editors will be seeing when creating pages so we aren't blind sided. So long as we have a bit of notice before the metaphorical switch is flipped that would be helpful :D. --Cameron11598 (Talk) 00:16, 17 July 2017 (UTC) Please send them our way. My talk and Kudpung's have become zones for community comment and questions on this of late, so you can send people there if they don't want to join any of the formal conversations at the places I linked. One of the tough things on this is that there is hours worth of reading material on background. I don't expect everyone to read all of it, but a lot of the answers to the questions people have can be found in those discussions dating back at least a year (and really six, but the last year has been very active.) TonyBallioni (talk) 00:39, 17 July 2017 (UTC) This has all been brought about by Majoras improptu and inappropriate posting of messages to the VP and Jimbo Wales' talk page. We are nowhere near ready for disseminating this information yet, and won't be until we are ready to revise and decide upon the GUI messages and if the WMF is planning to implement the 'AC' part of the trial or whether we will implement it with a local script. Broadcasting it on IRC is premature. Kudpung กุดผึ้ง (talk) 04:02, 17 July 2017 (UTC) Agreed that we should not be broadcasting specifics yet because they are not known. The above were drafts mainly in response to the fear that this would become permanent and to show that we would be sending out abundantly clear messaging as a trial. Its at VPP and Jimbo-talk now, so while I don't think it was secretive at all before, it certainly isn't now and people are able to join in any of the conversations just as they were before. Cameron, I think Kudpung's point is that because we are still very early in the planning stages of implementation and don't even know the date yet, we shouldn't have people worked up when the exact details aren't known. We're both very willing to answer any questions anyone has (look at our talk pages), and if you or anyone else has any questions, feel free to ask. Since you are also involved with helping new users on -help, your input in these conversations is very useful. TonyBallioni (talk) 04:28, 17 July 2017 (UTC) And please note everyone, that the above proposed newletters are only drafts at this stage. Tony has posted them in the interests of transparency so that people know what is going on, but they are in no way the definitive texts and as the preparation evolves, things can change. This trial is going to happen, but we don't know when yet so we don't want people getting worked up and creating the same kind of confusion that happened a few years ago when the BLPPROD was in the making. There is a whole team at the WMF working on this now for us and with us and we want to get it right. Kudpung กุดผึ้ง (talk) 05:12, 19 July 2017 (UTC) Yes. These were drafts created after the Jimbo-talk discussion as examples of how we could explain the specifics to concerned groups. As with virtually every edit I make, I have no attachment to them if they can be improved or circumstances change. TonyBallioni (talk) 05:26, 19 July 2017 (UTC)