Wikipedia talk:Autoconfirmed article creation trial/2011 archive

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

More helpful header wanted

As I see it, a group here and a group on mediawiki.org are both talking about solving the same set of issues, related to article creation and what we know about the editors trying to create the article (including their autoconf status). I'm not sure why this gap exists (except that everyone prefers their own favorite wiki for brainstorming :) but you could make the header here much more informative.

It would be good to unify the discussions about how to move forward. And I don't see a reason to 'shut down' discussion of practical policy and message-crafting matters, as you began to here. There are ways that part of what was agreed upon here on en:wp could be started without strictly changing userrights. For starters, if you show a different set of MediaWiki messages to autoconf editors and you can dramatically change where they go (on viewing a redlink, or completing a search for a missing topic) and what they think constitutes "creating a new article".

And if we have a way to do real-time surveys of editors after they've contributed, or as they are leaving, we could get feedback on the different reactions to various methods in a week, not six months. Which isn't a bad direction to head in, if we want to do much more data-driven testing as this proposal suggests. – SJ + 09:48, 17 September 2011 (UTC)[reply]

Thanks for the feedback. It would not be appropriate to rename this page - it was already a major, well published cental RfC.

Hi Kudpung, I did not suggest a rename -- only a more informative header template. It could for instance point readers to all places where there is currently active discussion taking place. – SJ +

Unless you took time to follow or catch up on the history of this issue, you will not have understood that it was either the WMF and/or a developer who has unilaterally decided not to accept the clear consensus that was reached for Autoconfirmed article creation, and its trial. It would be worth investing the three or so hours to do so - you will then be in a good position to decide objectively which side of the argument you stand and where you will lend your support..

Kudpung, by the way -- your work on NPP is great and helpful part of the current page creation process on this wiki. So thanks for that - obviously your views on how to improve matters bear consideration. One interesting aspect of the discussion so far is that a number of people at key points in the process have assumed that those they were speaking to hadn't done their homework (partaking in NPP, reading discussions to date, doing background statistics, developing a nuanced set of ideas surrounding the RFC to support participation by newbies). I see a lot of people who have done a great deal of work, who aren't recognizing one another's efforts.

I did follow the mw, en:wp, and bugzilla discussions - my comments reflect my understanding of the current situation, which is slightly different from yours. I see no unilateral decision, just two different multilateral groups, having slightly misaligned discussions around the same topic in two different places. And while I may be missing something obvious, I do not see argument, just people talking past one another. The awkward part of having discussions by two disjoint groups in separate places is precisely that it is not possible to crystallize differences into a productive argument. – SJ + 12:15, 18 September 2011 (UTC)[reply]

In the meantime, the WMF have closed further dialogue - at least on the open discussion at Bugzilla - and started their own different proposals for solutions loosely based on the original proposals, but apparently neither based on fact nor supported by statistics.

Discussion on bugs remains open indefinitely -- independent of the "open"/"resolved" status of the bug. The disadvantage of bugzilla discussions is that they are less easy for most editors to contribute to than a wiki discussion, hence the request to move to wiki pages. – SJ + 12:15, 18 September 2011 (UTC)[reply]

These are the relevant discussions:

Kudpung กุดผึ้ง (talk) 18:45, 17 September 2011 (UTC)[reply]

This is Wikipedia. If you believe the header needs to be reworded, then reword it. We don't own this page. —SW— ^comment 15:38, 18 September 2011 (UTC)[reply]

Soft security concepts

A soft security method might work here -- trying to implement what is envisioned in this trial, but after the fact rather than before, via scripts or bots.

There's no need to tell any new user "you can't create an article" -- but the default place that the article goes on creation could be different. – SJ + 09:09, 17 September 2011 (UTC)[reply]

The developers appear to be under instructions from a member of the WMF to revert such changes too. --Kudpung กุดผึ้ง (talk) 18:47, 17 September 2011 (UTC)[reply]

Do you have a link to support this idea? I don't know why anyone would revert what admins here choose to do to mediawiki-namespace text to improve the on-wiki process. – SJ + 12:15, 18 September 2011 (UTC)[reply]

Anyone who believes the "improvement" is not an improvement would do that. WhatamIdoing (talk) 16:33, 18 September 2011 (UTC)[reply]

mw:Article creation workflow may be of interest. →Σ τ c. 20:04, 17 September 2011 (UTC)[reply]

Thanks Σ, that's what we've been talking about here all along ;)

@SJ & WhatamIdoing: I would personally not risk any wheel warring with the salaried staff over UI code changes that I could do through my admin access to simulate a similar system based on scripts and bots . The clear message from Erik Moeller and Brandon, appears to be that the WMF can (and will) override anything they don't like on this issue, without needing to support their reasons with facts or proper research - I always assumed that the individual Wikipedias were able to govern themselves by consensus on such policy issues, but maybe I have been wrong. However, the reason why there are two discussions in different places on this is because the WMF appears to want any solution to our NPP problem to be one of their own making, through discussion as part of their mw:Article creation workflow project. If I thought it would help, I would participate in it, but through their action and comments at Bugzilla I have lost confidence in their GF and decision making. Broad based WikiMedia discussions of that kind take months, if not years - if ever at all - to find their focus and crystalise into workable propositions that can be implemented. Either that, or the WMF will simply decide among themselves what the solutions are to be and implement them by fiat. I nevertheless still firmly believe that Snottywong and Blade with their autoconfirmed rule, with its three ways to aid new mature creators fast-track serious new articles, have come up with the most logical and pragmatic solution, and it's the only one I can support. Kudpung กุดผึ้ง (talk) 00:29, 19 September 2011 (UTC)[reply]

mw:Article creation workflow appears to be focused on making the article creation process easier and more successful for new editors. It does not seem to touch on the problems that ACTRIAL is designed to deal with; I'm not sure why people keep linking to it in discussions about ACTRIAL. ErikHaugen (talk | contribs) 17:58, 19 September 2011 (UTC)[reply]

WMF's response is essentially that they believe restricting article creation will have too many negative side effects which will cause new user retention to plummet. We tried to explain to them numerous times that a brief trial would prove or disprove their theory (and it is a theory, backed up by no hard evidence). Their response was that in their experience, even brief trials of controversial ideas like this resulted in negative effects that lasted well beyond the end of the trial. Take that as you will.

The reason they are linking ACTRIAL to the Article creation workflow idea is because this is their response to our trial. They don't want to implement the trial as proposed, but they recognize the problem and this is their proposed solution. Most (if not all) of us disagree that the proposed changes to the editing interface will resolve any of the problems that ACTRIAL was intended to resolve. Nevertheless, WMF has devoted resources to the problem, and mw:Article creation workflow is where they are being discussed. You might want to express your opinion there if you want the WMF developers and staff to be aware of it.

In any case, they have made it clear that the trial isn't happening, and any attempt to implement the trial on our own (i.e. using a method that doesn't require developer access) will be reverted. If you're interested in trying to steer where they go with their own solution, comment at the mw page. I, for one, am burnt out on the topic and don't wish to discuss it much for the time being. —SW— ^speak 20:02, 19 September 2011 (UTC)[reply]

I can certainly understand why you feel that way. Thanks; ErikHaugen (talk | contribs) 20:51, 19 September 2011 (UTC)[reply]

Six months is probably not within the WMF's definition of "a brief trial". It certainly isn't within mine.

As for whether the WMF should oppose this—well, I don't own Wikipedia, and I've never suffered from the illusion that I did. The WMF's job is not to do the bidding of "the community", but to take steps that, in their best judgment, will reach their charitable goal of sharing educational information with the world. I don't have to agree with them, but I do recognize that it's their job to prefer their best judgment to my best judgment. The WMF is not actually accountable to me, or to any other individual volunteer. Consequently, I am not going to demand that they explain, justify, or otherwise try to account for their decisions to me. WhatamIdoing (talk) 01:35, 20 September 2011 (UTC)[reply]

I've said more about the process from my perspective here, here and here, and I'll leave it at that, except to say again that I'm sorry that we've not been able to be more responsive and engaged on these issues sooner. :-( But we are now: we're trying very hard to help as serious and engaged partners.

The design drafts that we've posted are intended to look at the various problem areas that are targeted by WP:ACTRIAL from different perspectives: how do we increase the efficiency and quality of new page patrolling; how can we motivate users to more effectively self-sort into "bold" vs. "careful" users (those who want to create a page right away, and deal with the consequences, and those who want to learn to do it correctly); how can we more effectively communicate key policy concepts; how can we reduce biteyness of the experience.

Our overarching hypothesis is that it is possible to maintain a high (and possibly even increasing) standard of openness while also maintaining high standards of quality (without burning out patrollers) and achieving an experience that's at least appropriately friendly (i.e. where folks who are truly engaged receive a high degree of human touch, while those who are dumping poor quality content receive appropriate, polite notices). Moreover, we think that the reason that we haven't achieved this yet is in part due to lack of focused support for the community from WMF -- many people in the community have been saying these things for a long time.

We think that's a hypothesis very much worth exploring, and one which could have very important implications for the long-term health and development of Wikimedia projects.

While the first page we've created focused on the article creation workflow, I'd also like to point to Brandon's new design specifically related to improving the new page patrolling interface, which can be found at mw:New Page Patrol Zoom Interface. And yes, we're going to post more data, and we're hoping you'll add relevant data points as well.

(The reason, BTW, we've created these pages on MediaWiki.org is that these are software specifications that likely will have implications for other Wikimedia wikis, and potentially non-Wikimedia wikis, as well, and we're aiming to bring more people from outside en.wp in the conversation. We realize that there's inherent awkwardness in moving around between wikis -- sorry for the inconvenience. I'll try to keep an eye on this page as well.) --Eloquence * 09:20, 20 September 2011 (UTC)[reply]

The tooling can only help so much; to me, most of the work of patrolling is thought, reading, research, etc. The lack of tooling doesn't slow me down a significant amount. So, while the Zoom interface might be helpful and appreciated, I doubt it will do anything to address the concerns that ACTRIAL would address. I think other existing tools have already taken care of the low-hanging fruit here. To really take another significant bite out of NPP workload, I don't know of anything short of ACTRIAL that is going to work. ErikHaugen (talk | contribs) 23:35, 20 September 2011 (UTC)[reply]

Actually there are a whole bunch of ways in which we can make the process work better. Some of them as uncontentious and major as showing the [mark as patrolled] tab to experienced editors who come across new pages other than via special Newpages. As one of the substantial minority who opposed the trial I've avoided taking part in discussions as to how we can successfully trial something that I was deeply opposed. But if we now have commitment from the Foundation that resources are available to improve things then I think we should take advantage of that - who knows we may wind up with something that pretty much everyone thinks is an improvement. Ϣere SpielChequers 09:50, 22 September 2011 (UTC)[reply]

The WMF team that has taken over this NPP issue most definitely does not appear to be looking at the problem areas targeted by WP:ACTRIAL. What they are doing is looking at making new page creation a more enjoyable experience for potential serious contributors of new articles. What they are trying to do for NPP is to create a tool that will only be as good as the workers in whose hands thery put it. WP:ACTRIAL addresses a serious problem of the total ineficiency of a system that generally attracts only the most inexperienced of all Wikipedia users, because it gives them a power that requires no user right, no experience whatsoever, and no demonstration of maturity and knowledge of policy. The authors of WP:ACTRIAL have provided substantial research and data that corroborates these facts, while the WMF has provided no proof whatsoever for their reasons for refusing even a trial. Furthermore, any vague evidence-based arguments were centred around a completely erroneous flyer published by the WMF for the Haifa Wikimania that suggested that 30% of en.Wiki's best contributors began their Wiki careers as vandals.

Where is, or was there any proof of 'burn out' of NPPers? That notion was also invented at a WikiMedia discussion. The only people who suffered 'burn out' were those who were desperately trying to stay on top of the disasterous patrolling by the patrollers, and provide the research and stats for WP:ACTRIAL. Those people have now largely given up hope that their work will be recognised, and have lost a great deal of confidence in the competence and sincerity of the WMF. The vast majority of NPPers is a base of transient individuals who don't give a hoot for WP:NPP.

It`s a bit late in the day to labour for more months or even years under new suggestions that pretty much everyone thinks is an improvement - if user retention is a priority, what is needed fast are some concrete suggestions - such as WP:ACTRIAL with its three fast-track avenues for serious creators, and to stem the tide of totally and uncontroversially unwanted new pages, while drastically enhancing the new user experience for new article creators. New users are being turned away by the 100s every day due to both the walls of text and rules they are confronted with, and the bitey messages they get when they are trying to create a decent article in good faith. Kudpung กุดผึ้ง (talk) 12:45, 22 September 2011 (UTC)[reply]

WSC, if that was directed at me, I am certainly not saying I don't like Zoom or that I don't want anyone to work on it or "take advantage of that". I'm looking forward to using Zoom. I am simply making the point that it is naive to think that Zoom addresses any of the concerns behind ACTRIAL. I think Eloquence is wrong that Zoom is the kind of "focused support for the community from WMF" that will bring us closer to the described NPP nirvana. That is all. I'm looking forward to using Zoom, but I hope something is also done to address the concerns behind ACTRIAL, since those, I think, are pretty serious. ErikHaugen (talk | contribs) 17:07, 22 September 2011 (UTC)[reply]

This is essentially my perspective as well. —SW— ^babble 18:55, 22 September 2011 (UTC)[reply]

Kudpung, since you brought it up and it was mentioned on the bug, I just want to make a side note and say that the 30% stat isn't in the final summer research findings on Meta or in the PDF of the flyer you're talking about because it was simply a stupid miscommunication between the researcher who did the study and us when we were all writing the flyer on a tight deadline. Their number was 33% of the current top 100 by edit count were reverted in their first edit session, not reverted for vandalism. It was a confusion in terms, that's all, and I'm sorry it got published. Steven Walling (WMF) • talk 19:44, 22 September 2011 (UTC)[reply]

Do a lot of NPPers actually use the NPP interface itself, or intermediary software products (like NPWatcher of old, or Huggle, or whatever is the current standard)? Nathan^T 23:21, 23 September 2011 (UTC)[reply]

I use Twinkle, which is fine for me. Every so often I'll do it without tools, to keep it fresh in my mind (I manually patrolled pages for months before using Twinkle). The Blade of the Northern Lights (話して下さい) 05:52, 27 September 2011 (UTC)[reply]

Two tools I know of are Kissle and NPWatcher, though they both seem to require some form of pre-approval to use and I'm not sure how actively they're used. My guess would be that the long tail of patrollers are using the Special page, perhaps combined with Twinkle. Steven Walling (WMF) • talk 17:39, 27 September 2011 (UTC)[reply]

I've thought about trying Kissle, but I move faster on Twinkle than I've seen people go with Kissle. I don't know how much of that is my reading speed (well within the top quartile of the 99th percentile) that lets me go fast versus the tools being used, so I might give it a shot to see what happens, though installing code is not one of my strengths. The Blade of the Northern Lights (話して下さい) 02:44, 28 September 2011 (UTC)[reply]

I naturally use Twinkle - Since it was recently completely overhauled, Twinkle is a superb interface, is quick to do what it does, is cross-platform compatible, needs no user permission, and takes up no screen space. It also gives me an opportunity to reflect upon, and modify some of the bitey and/or TL;DR uw that have been written for it.. I remain fully convinced that any other forms of automation for NPP are unnecessary, and even dangerous in inexperienced hands (so is Twinkle actually). Kudpung กุดผึ้ง (talk) 12:52, 28 September 2011 (UTC)[reply]

Wow...

I can't believe how this got halted at the eleventh hour. Missed the tussle during the move back to uni. Thank you, editors who put so much effort into moving the idea forwards (I shan't list names, because I'll inevitably forget some, but you know who you are); it's a real shame that hasn't paid off. Crossing my fingers that the WMF will reconsider this at some point in the future before the turning point is passed. Cheers guys. Brammers (talk/c) 18:24, 29 September 2011 (UTC)[reply]

Not running ACTRIAL will deny us of the valuable feedback that it would have provided, and it's true that the WMF has had it's fair share of disagreements with several Wiki communities of late. However, I do believe the WMF has now taken all this on board and although ACTRIAL as an immediate experiment may not happen any time soon, there is positive resonance from them and some serious challenges to the problems of New Page Patrolling (which triggered everything in the beginning) are going to be met head on, and most importantly, now with the close collaboration of some of us 'ornary lot' . --Kudpung กุดผึ้ง (talk) 05:54, 30 September 2011 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.