Wikipedia talk:New pages patrol/Reviewers/Archive 10

This is an archive of past discussions on Wikipedia:New pages patrol. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 5

←

Archive 8

→

Comments on Reviewing and Institutional Cultural Disconnect

It occurs to me that there are two basically unrelated aspects of institutional culture that both contribute to the review backlog, and, more importantly, toward the predictable (and almost planned) decline of quality at the expense of quantity. The first is the WMF, which imposes a culture from above. I have no personal knowledge of the culture of the WMF, but I can see that they like metrics, have a staff, have a budget, and (as User:Guy Macon points out), have an expanding budget and expanding staff. Controlling the expansion of the English Wikipedia must run contrary to what I see as a growth-oriented outlook by the WMF Board. However, if it isn’t controlled, we will soon enough have fifteen million articles, of which eight million are spam and three million are other kinds of crud. That is the first aspect of institutional culture that is contrary to reviewing. A focus on numerical expansion doesn’t place a high value on controlling the expansion. If we, the reviewers, are trying to control the growth, we are fighting against a culture from above.

The second and more subtle factor that works against improving reviewing is a cultural value among editors. That is the extremely high value that is attached to the need to be welcoming to new editors, as illustrated by the rule of Do not bite the newcomers. While that is officially a behavioral guideline, less formal than a policy (although more formal than an essay), I think that it has become a dogma, an almost religious tenet, in the culture of many editors. While as written it is a good guideline, it is my eccentric belief that as implemented it does more harm than good for two reasons. First, it is one of the first Wikipedia rules that many combative new editors, eager to use Wikipedia to right great wrongs, find, and they use it as a cudgel against more experienced editors who try to caution them. (In my perverse opinion, if you have been editing Wikipedia long enough to quote BITE, you are no longer a complete newbie, and you have been editing Wikipedia long enough that you should let a third editor decide whether you are still entitled to kid-glove handling.) One effect of the extremely high value placed on the need to be welcoming to new editors is that editors such as New Page reviewers, who have to be unfriendly to some new articles and some new editors, are sometimes rebuked or degraded.

Just a few comments that as reviewers we seem to be working against both a culture from above and a culture from within. Robert McClenon (talk) 03:23, 20 June 2017 (UTC)

^{On the road to creating: The Urban Encyclopedia.^{Atsme📞📧} 03:27, 20 June 2017 (UTC)}

First paragraph: I believe every word of that.

Second paragraph, I largely agree, but some tentative comments: Be gentle with newcomers is important, and it has the right level of current respect, except that an important caveat is needed: The newcomer must have genuine interest in contributing to the project. I don't have a firm idea on how to do this right, but newcomers should be treated more gently for wiki markup mistakes, and less gently for spamming. I still like the old idea, pulled from ACTRIAL rationales, that newcomers with new topic ideas should be encouraged to add mentions of the topic in existing articles, and discouraged from starting a new orphan page (draft or otherwise) until that is done. On that thought, DraftSpace may be doing more harm than good due to mainspace->draftspace wikilinks being forbidden. --SmokeyJoe (talk) 03:34, 20 June 2017 (UTC)

Regarding the expanding budget and expanding staff, see my essay at WP:CANCER. --Guy Macon (talk) 04:21, 20 June 2017 (UTC)

All I can say to that Robert McClenon is:

First paragraph: I never heard anything so succinct and more accurate in all my time as an editor and admin.

Second paragraph: Ditto.

Kudpung กุดผึ้ง (talk) 05:20, 20 June 2017 (UTC)

can't say I agree with either part.

1.If the WMF is of the opinion that the growth of the enWP should be encouraged, I'df agree with them. Even for articles, while we do have saturation coverage of current works in a few areas of hobbyist interest, but everywhere else we are woefully incomplete, even for the chronological, geographic and language areas our current editors find easiest to work in. More important, we will always need new editors, to if only to maintain and improve the current articles. Very few of even the most active editors stay for more than five or six years, and WP is a longer term project than that. It would be very good to have a quantitative measure for article quality, but in its absence, what can we measure but article number and editor count? There's no cultural difference between the WMF and the editors in this. And while the bureaucracy may want an expanding budget, although this may affect various aspects of the culture and relationship, I don't see how it affects or even requires the necessity for new editors.

2. Most new editors come to contribute to the encyclopedia. It is essential to retain them, but at least 80% of those who have their first article rejected quite understandably never return. We may still need to reject their first efforts, but this an be done with a rational targeting explanation in ordinary language, explaining the wikiterminology and encouraging them on how to continue. It can not be done with templates. No matter how well we word them, who in the world actually like an obvious form message sent on the internet? Fpr the ones who come to promote their views, they can be directed for how to explain them, and many of them will learn to once they are told what is expected. For those who come to advertise, they can be directed on how to add articles on their fields of knowledge--some of them will, but at least the others can be sent away understanding politely why this is not the appropriate place. Some do come to make trouble; some come to do undeclared paid entering--even within this groups, still some can be reclaimed. Most can't of course, and here if they are dealt with properly, they can be at least persuaded to leave without making further trouble. A form message doesn't even do that, any more than does the wording on a parking ticket. DGG ( talk ) 06:11, 20 June 2017 (UTC)

DGG ( talk ) 06:11, 20 June 2017 (UTC)

I have a question. User:DGG states that most new editors come to contribute to the encyclopedia. I would like to know whether this statement is being made specifically about new editors who contribute an article as their first contribution, and whether this statement is being made as an empirical and factual statement (as opposed to being stated as an axiom or an article of religious faith). The answer does make a difference, because if new editors who contribute an article that is declined did come to contribute to the encyclopedia, then we have both a need for special action to retain them, such as a group of greeters and meeters, and an opportunity for significant improvement in the English Wikipedia, by refocusing the efforts of the large number of new editors to other important tasks such as the improvement of existing articles. Is that a statement of empirical fact, or simply a statement of required belief? Robert McClenon (talk) 03:14, 23 June 2017 (UTC)

DGG, Kudpung with regards to the quality vs quantity argument, I wholeheartedly support the "move to draft" action after tagging, and reaching out to the article creators to fix the problems. Getting articles deleted is a root canal. A good number of them are being created for promotional purposes by editors who may be employees or have other COIs. A few examples:

Geneshift - game makers are using WP as a launchpad for their new games, no RS, the results of the AfD were keep (the same is happening with films still in the can, long before distribution);
Marc A. Zimmerman - someone close to the subject wrote a promo piece about this academic citing nothing but work authored by the BLP himself - I recently added two sources but the entire article needs to be rewritten, and should not have been published;
Maks SF - multiple relistings in AfD

Resources are low in all departments. We need less bureaucracy and more active editors. Oh, and we also have the irritation of drive-by taggers who are too busy tagging articles with orphan and category tags instead of fixing the issues. It simply doesn't make sense to allow new articles into article mainspace when they have blatant promotional and/or sourcing issues. They stand a better chance of being fixed in a timely manner by their creators while in draft space, and maybe we could also create a task force to work with those editors and encourage them to grow with us. ^{Atsme📞📧} 13:06, 20 June 2017 (UTC)

DGG "Most new editors come to contribute to the encyclopedia." Can you point to evidence that that is actually the case, or are you perhaps expressing a hope. In my observation, the vast majority of the 80% of article that are deleted are from editors who are trying to promote something (a company, a person, a favorite record, and amateur sports club). Based on some quick sampling, my guess is that fewer than 10% of non-autoconfirmed users make any edits non directly connected with the original subject that they edited after two months. It should be easy enough to look at a larger sample to see what the actual stats for this are, but I think I'm in the ballpark when I say that a huge majority of new users come here to promote something. Sometimes their promotion overlaps our goal of creating an encyclopedia, but about 80% of the time, it doesn't.- Mr X 13:47, 20 June 2017 (UTC)

I didn't say most of the articles, I said most of the editors. I think by now at least half of the new editors know not to start articles immediately, and won't be harmed. A lot of what AC stops are sockpuppets too unsophisticated to get autoconfirmed first--all that adopting AC will do is teach them to edit trivially a few articles. I think it would be very wrong in any case to discourage even 10% of the new editors. We need them. The rest of us are at some point going to either go on to other things or die in harness. DGG ( talk ) 18:07, 20 June 2017 (UTC)

I know you said most new editors. Assuming we talking about new editors who create new articles, your statements simply don't reflect reality as far as I can tell. It's incredibly frustrating hearing these grand generalizations about Wikipedia's mission while the rest of us here are in the trenches trying to save it from being overrun by spammers and vandals. If we're going to make any progress on this, we need to work from a set of facts and not a bunch of unfounded assumptions and platitudes that only distract from this effort.- Mr X 20:36, 20 June 2017 (UTC)

First paragraph: Pretty good, although it seems, as DGG touched on, that the view of the WMF is like Wikipedia editors, but Wikipedia editors only want growth in certain areas (editors), while the WMF wants growth everywhere.

Second paragraph: I do think that your core idea, that people take WP:BITE to religious levels, is correct. But, as, again, DGG said above, we do need to keep BITE in mind, as we need more editors. So, it seems that the WMF is paying too much attention to some things and not enough to other things, including the core problem, the fact that we need more good editors. So, really, we need to get the WMF to focus a good deal of their energy on getting new, good editors. And we should probably focus on that too. My conclusion, then, is that we need to keep BITE in mind and try and do more explaining and less automated messages. RileyBugz ^会話 _投稿記録 20:52, 20 June 2017 (UTC)

And so the stalemate continues...the promotion & unsourced articles continue while we wait for the Utopian community that can clean everything up with the sweep of a magic broom...the spam continues...the vandalism continues while we remain in denial...the beauracracy grows....the project inherits more garbage because there aren't enough editors to go around and keep things reviewed...veteran editors continue to feel overwhelmed and retire...the integrity of our articles continue to diminish because we can't keep up...the drama continues on the notice boards because the PAGs are wishy-washy...and in the end, where does that leave us? Do a review of admin TPs...take a look at what ends up at COI, AfD, ANI...study the articles at NPR...nuff said. ^{Atsme📞📧} 00:30, 21 June 2017 (UTC)

Just wanted to add - when savvy admins and editors go the extra mile and take the necessary actions regarding new articles that are clearly not ready for mainspace, such as move them into draft space...editor morale soars because the feelings their fighting a losing battle disappears!!! The tags are in place on the article which helps the article creator become aware of what's required, and they can collaborate with others, ask questions, and eventually create an article that's worthy of the encyclopedia. We need more actions like this one if we hope to maintain the quality and integrity of WP: Draft:Marc_A._Zimmerman ^{Atsme📞📧} 02:50, 21 June 2017 (UTC)

@Atsme:Just wanted to add that I ain't an admin!Thanks for the elevation though!

!Winged Blades^Godric 04:10, 21 June 2017 (UTC)

My apologies, Winged Blades of Godric, I clarified my statement, and THANK YOU!! We need more "movers & shakers" like you. ^{Atsme📞📧} 13:27, 21 June 2017 (UTC)

@Atsme: Just wanted to clarify: BITE does not mean, at least to me, not doing shit and just slapping tags on articles. It means cleaning up articles (that aren't obvious attempts at promotion) and telling the creator what they should do if they want to add more info to the article. It means removing unsourced stuff from BLPs with prejudice, but then telling the people who added it that all they need to do is add some sources. I do know, although, that my statement about convincing the WMF was incredibly idealistic. So sorry about that. But, continue moving the borderline articles to draftspace and deleting unsourced statements in BLPs. RileyBugz ^会話 _投稿記録 14:16, 22 June 2017 (UTC)

RileyBugz

On the same page and I wear "kid gloves" when communicating with new editors. I was one myself, seemlngly not that long ago. ^{Atsme📞📧} 20:17, 22 June 2017 (UTC)

"Most new editors come to contribute to the encyclopedia.". Actually, very few new users who create an articke as their first attempt to edit Wikipedia come to contribute to the encyclopedia. Anyone who patrols from the 'Were created by new editors' filter, clearly identifies without any difficulty that at least 80% of all the new article are created without the slightest consideration at all for what could even be broadly construed as 'Wikipedic'. These are 'pages' crated by spammers, autobiographers, vandals and trolls, COPYVIOS, hoaxes, gibberish,and foreign laguage pages (most of which also turn out to be completely worthless). The creators only have one goal: to get themselves or their org on Wikipedia or to leave vandalism on it. Those 'users' will never become useful members of the editing community, as clearly evidenced by the very low number of contested CSD and disputed PROD. Kudpung กุดผึ้ง (talk) 18:12, 21 June 2017 (UTC)

I would like to provide my own follow-up comments on two topics that have been mentioned in this discussion, expansion of the English Wikipedia, and new editors in the English Wikipedia. Robert McClenon (talk) 23:27, 21 June 2017 (UTC)

Expansion of Wikipedia

I think that my concerns about unlimited expansion of the English Wikipedia may have been misunderstood. I said that if a focus on growth continues, we will eventually have fifteen million articles, of which eight million are spam and three million are other kinds of crud. (That does mean four million articles that should be kept. Since we currently have somewhat more than five million articles, I did mean that approximately one million of the existing articles are crud that should be deleted. Four million is still a lot of articles, and the English Wikipedia is an impressive electronic work.) I think that some editors may have thought that I meant that expansion of the English Wikipedia is not needed or is undesirable. I did not mean that. Expansion of the English Wikipedia should be encouraged, in particular in areas where coverage is known to be incomplete. Areas where coverage is known to be incomplete include those sciences where the scope of knowledge is both broad and expanding, including biology, astronomy, and chemistry. Every species that has been formally described is worthy of its own article, even if only a stub. Every star that is in a star catalog and every galaxy that is in a catalog is worthy of its own article, even if only a stub. Every distinct chemical substance is worthy of its own article, even if only a stub. Also, there are areas or sub-areas in which it has been determined that our coverage is incomplete or needs to be improved (women writers, medieval art). However, continued expansion of the English Wikipedia as such should no longer be a primary objective, and an increase in the number of articles should not in itself be reported as a measure of continued well-being. (An increase in the number of spam articles, far from being a sign of health, is, to modify Guy Macon’s metaphor, a cancer.)

I would further submit that any need for continued expansion of the English Wikipedia in areas where knowledge is expanding or where its own coverage is inadequate should be viewed entirely separately from any need for new editors. We should not think that we need new editors because we need to continue the expansion of the English Wikipedia. Expansion of the English Wikipedia in areas where it needs expansion, such as invertebrate species, is being done by editors who specialize in the areas of our need. In general, we should not look to new editors to improve our coverage of areas where our coverage is lacking. (There may be special cases where a WikiProject determines that we need to try to recruit new editors with specialized interests. Such a policy, if in order, is unrelated to the handling of over-the-transom editors.) Robert McClenon (talk) 23:27, 21 June 2017 (UTC)

If it hadn't become clear by WP:ACTRIAL that the Wikimedia Foundation is oly interested in boasting about raw page creation growth irrespective of quality, content, or deletions, it should now be blatantly obvious that this is the leitmotif of the essay by DannyH (WMF) and the phraseology of his rejection of the many user comments both there, here, and at WT:NPPAFC, and his admitted lack of comprehension of what ACTRIAL was all about. Before the Foundation embarks on another course of its own reputation self-destruction, they need to read, and properly read what the community is writing rather than just leave a new comment at the bottom of every new thread. Kudpung กุดผึ้ง (talk) 01:14, 22 June 2017 (UTC)

New Editors

One respected functionary says that most new editors come to contribute to the encyclopedia. Others disagree. We agree that when their contributions are deleted by NPP (or declined by AFC), most of them leave and do not return. The statement that new editors come to contribute to the encyclopedia can be interpreted in either of two ways, as an empirical statement, based on some assessment of their contributions, or as an axiom, a principle that must be taken to be true and is not subject to proof. I am aware that it is an axiom that new editors come to contribute to the encyclopedia, and, as a result, the fact that they are not welcomed and are lost is a loss to the encyclopedia and a problem. But MrX asks: Can you point to evidence that that is actually the case, or are you perhaps expressing a hope.? Good question, and the answer does matter. If we do have a large number of new editors who come in order to improve the encyclopedia, and these new editors are not welcomed appropriately, then we have not only a shortage of reviewers, but a twin shortage both of reviewers, and of a corps of greeters and meeters. (We cannot expect the same volunteer editors both to protect Wikipedia from crud and to extend a special profuse welcome to new editors.)

My experience, first in AFC, and more recently in AFC and NPP, is that new editors of new articles fall into three overlapping classes: those who want to contribute to the encyclopedia; those who are clueless; and those who have self-serving agendas. If it is true that the majority of new editors fall into the first class, then it is true that we have a serious new editor problem in that we treat new editors badly, as if they are either clueless or self-serving or both. I would like to see an analysis of new editors. I would be more interested in an empirical assessment than in a quasi-religious statement of belief. If we are indeed losing new generations of editors because we do not welcome them properly, then we need an additional volunteer corps of greeters and meeters, not to nag our current volunteer patrol editors to do double duty.

Also, regardless of the merits or the numbers of the new editors whom we are losing, we should not count on new editors to facilitate a rounding out of Wikipedia or the addition of knowledge in areas that we know need improvement. We should not be assuming that new editors will fill in areas of knowledge and interest, such as invertebrates, medieval art, women writers. Regardless of the rate at which we should be expanding the English Wikipedia, any filling-in of incomplete areas must be done by conscious effort or active recruitment, not by idle hope that new editors will satisfy our needs.

Also, does “most new editors” mean most new editors whose first edits are to create a new article, or just most new editors? Many new editors introduce themselves to Wikipedia in some way other than by creating an article, such as by discussing at a talk page, or by copy-editing, or even by playing in a sandbox. If most new editors who submit an article really come to contribute to the encyclopedia, then we really do have an available resource that is being lost and is not being engaged. But do most editors whose first edits are a new article really come to contribute to the encyclopedia? The answer does matter. I am aware that it is often stated as an axiom or belief, but is it empirically true with regard to editors whose first edits are a new article? Robert McClenon (talk) 23:27, 21 June 2017 (UTC)

Again an excellent assessment by Robert McClenon which probbly the WMF will simply scoff at again. I will just emphasise however, that in the context of these recent discussions, the term 'new users' is used to mean new uses who create a new page in mainspace as their first effort editing the Wikipedia - that needs to be clear.

"new editors of new articles fall into three overlapping classes: those who want to contribute to the encyclopedia; those who are clueless; and those who have self-serving agendas." Precisely, and that and its associated comments pre-empt the comment I was already drafting to post on one of the other talk pages in this round of discussions, in which I was going to ask for statistics that show:

Number of articles created by new users,
edit count of those users, (which of course will be <10).
Number kept,
Number deleted by 'each CSD criterion',
number of actually deleted CSDs that were refunded within, say 7 days
umber of PRODS,
number of successful dePRODs, (i.e. that were not sent to AfD.
number of new articles by new users sent to Afd by patrollers
number of new articles by new users sent to Afd by patrollers and were kept
number of new articles by new users sent to Afd by patrollers and were deleted and/or redirected.

The problem will be in getting those stats, because the WMF has now clearly demonstrated by asking the help of a long retired volunteer, that they don't actually know how to consult their own data bases.

Kudpung กุดผึ้ง (talk)

We agree that when their contributions are deleted by NPP (or declined by AFC), most of them leave and do not return. : I would just add that they have usually gone before they even knew that their articles have been tagged, PRODed, or CSDd. Kudpung กุดผึ้ง (talk) 01:04, 23 June 2017 (UTC)

Yes. Yuck. That explains why the vast majority of the PRODs that I apply on NPP expire in seven days, and the tagged articles expire with the PROD tags. Preventing the one-time drive-by "authors" from throwing articles over the fence would make things simpler. Robert McClenon (talk) 06:37, 23 June 2017 (UTC)

From what I have observed the majority of new users that create articles and then disappear create articles that are either for companies or people very often with serious notability problems. Would it not be better to limit the creation of articles to registered users that have a certain number of edits in the main article space? Domdeparis (talk) 08:38, 23 June 2017 (UTC)

User:Domdeparis - Maybe. That would require a new privilege, something more than autoconfirmed, but not quite the same as extended confirmed, that editors could achieve, for instance, by copy-editing. Is that sort of what you are saying? Robert McClenon (talk) 02:56, 24 June 2017 (UTC)

Kudpung, DGG, Robert McClenon - I liken the current process of "anybody can edit", as it often applies to article creation, to laboring as an unpaid Busboy. I'm of the mind that unless an editor has autopatrolled rights, all new articles should begin and stay in draft space until they've been reviewed and passed. Another option would be for a BOT to automatically headline new creations in mainspace with the label DRAFT ARTICLE, unless the editor has autopatrolled rights, and include instructions in edit view so the creator will know how to get the label removed. I believe either the former or latter, or both, will encourage article creators to work at getting the label removed, and that's when the training intervention begins to nurture and keep new editors. Either way, such a label advises our readers so they'll know what articles are most trustworthy. ^{Atsme📞📧} 14:51, 23 June 2017 (UTC)

Atsme, an intesting suggestion, and it would certainly help the genuine good faith artice creators, but it wouldn't stop the creation of blatant spam and other junk. For that, ACTRIAL is still the only realistic solution. Kudpung กุดผึ้ง (talk) 18:22, 23 June 2017 (UTC)

Autopatrolled rights have only been given to 4k editors and the guidelines are that an editor has to have 25 valid articles already. As it is the draft review is completely snowed under and articles have been in there for months now. I submitted an article myself and after nearly forty days waiting I decided to ask advice of some other editors modified the article and then published it in the main article space. This suggestion would create an impossible backlog and most definitely scare off new editors and even discourage experienced editors from creating articles. This would in my opinion just displace the problem from one group of editors, new page reviewers, to the AFC reviewers. And probably what will happen is that the number of page creations will radically drop because those that aren't autopatrolled will lose patience and abandon the creation of articles. This will mean that only 4000 people will be creating article on WP and we would lose the encyclopedic nature. Domdeparis (talk) 16:56, 23 June 2017 (UTC)

Would it not be better to limit the creation of articles to registered users that have a certain number of edits in the main article space? Domdeparis, if you take a moment to catch up on the talk above on this page and its recent archives, and at WP:NPPAFC, WP:KNPP, Wikipedia talk:New pages patrol/Analysis and proposal. You'll then see that WP:ACTRIAL corresponds with your suggestions and addresses your concerns for AfC. Kudpung กุดผึ้ง (talk) 18:08, 23 June 2017 (UTC)

Kudpung, as you probably know by now, I wholeheartedly agree with WP:ACTRIAL and will support it but in the interim, perhaps the DRAFT ARTICLE label (even on stubs) would be easier to get implemented until that happens, or maybe not

. What's happening now reminds me a bit of WP:ProjectAccuracy (now deleted) wherein I attempted to get a quality ratings system implemented that would "certify & semi-protect" articles that passed a high-quality editorial board review (using academia and experts) which would make such labeled articles more appealing to all levels of education as acceptable resources for term & research papers. I submitted it to WMF and....well....

. Gotta add that trying to change the status quo is a bitch and requires the help of $\S/$ . ^{Atsme📞📧} 18:48, 23 June 2017 (UTC)

Atsme, the problems facing us have been steadily getting worse over the last 10 years since the 'watershed' year of 2007 when development and contributions to Wikipedia reached their apogee. I have been in the vanguard of reform to NPP and its solutions for the last 6 years and I personally think that the time now is to implement an urgent solution, such as giving ACTRIAL a chance (as its name implies) rather than prolonging the agony by considering alternatives. Such suggestions could certainly be examined if the trial were to prove that the WMF is right and the Community is wrong, but history has shown that almost without exception, the Community is always right and the WMF is always wrong. In fact the WMF has a clear history of making costly blunders and the recent staff reshuffle isn't going to change things - it;s still the same people. Kudpung กุดผึ้ง (talk) 19:13, 23 June 2017 (UTC)

@Atsme: Well... I think that it would certainly be good to have some sort of system where the creation of articles by non-autoconfirmed users is somehow tagged. It would be nice to have a banner (for about 7 days or so) on articles created by non-autoconfirmed users, as that would make it clear that the article is not to be "trusted". I am coming around now to the temporary ACTRIAL position, though (and as a trial, as Kudpung has made clear), but I certainly hope that we can find another solution. But, I do think that prior to ACTRIAL it should be known what percentage of non-autoconfirmed users whose articles are not delete become autoconfirmed, and how they get those 10 edits (if there are stats on this, please direct me to them). RileyBugz ^会話 _投稿記録 19:34, 23 June 2017 (UTC)

^[FBDB] Perhaps a peaceful protest like what we've seen at Berkeley will garner the attention of WMF?

^{Atsme📞📧} 14:42, June 23, 2017‎

RileyBugz, according to this, we have ~1.15 million users who have made 10 or more edits ever. Out of a total of ~31.2 million users who have ever registered that means ~3.5% reach autoconfirmed status. Not an answer to your exact question, but maybe it provides some insight. According to the same numbers, we have been gaining between 5-7k new autoconfirmed users a month over the past year. TonyBallioni (talk) 20:15, 23 June 2017 (UTC)

Yuck. That's a sad but interesting metric. Most users who create accounts never make the minimal 10 edits. ?!?! Does the WMF have anything to say about that? I assume that they will argue against ACTRIAL, because they will take some interpretation that is consistent with their focus on quantity and is contrary to the community's ideas about quality. Robert McClenon (talk) 03:44, 24 June 2017 (UTC)

Yeah, I'm going to be talking on Monday with some people at the WMF who want to help us analyze the situation, and get the stats that will help everyone to understand the full picture, and make decisions. I agree that the WMF hasn't been as helpful as we should have been for a long, long time, and that's something that my team is working on changing. Atsme, if it helps -- you've gotten the attention of my team, at least.

I think it's important that we gather as much info as we can about new users and new pages. Some of the questions that have been asked here and on the other recent discussions are very complex, with multiple if/then branches, like the one that RileyBugz is asking a few comments up.

I'm thinking about the next step on our side. Right now, it seems like the most helpful thing that we can do is come up with a list of stats questions that we think will help provide a full picture. My attempt this week to ask if there's already a set of success metrics for ACTRIAL didn't get a helpful response, so I'm thinking that my team should put a list together, and then run it past the folks here to see what we've missed. I'll be compiling the suggestions posted here and in the other recent discussions, but if somebody feels like pointing out the ones you think are most important, that would be great. DannyH (WMF) (talk) 04:16, 24 June 2017 (UTC)

DannyH (WMF), to be clear, while Kudpung might not have a particularly sunny view on your efforts here, he did provide you with a link to the questions that were originally planned on being evaluated before ACTRIAL was vetoed in 2011. I also answered your questions about how this actually fits in better with the WMF's vision for healthy communities than the deletion templates that greet 80% of articles created by new users. While I get why you might be frustrated with the reaction to your efforts here, and what has been the categorization of your proposal as sweeping the problem under the rug, people did address the questions and points you raised. TonyBallioni (talk) 04:33, 24 June 2017 (UTC)

TonyBallioni: Sorry, I should have responded more specifically about that list of evaluation questions. The question that I'm asking about a way to evaluate success for ACTRIAL is different from that list; I'll explain why.

This isn't meant to be a slight on the work that was done, just an observation: The list of "Questions we could ask" on the ACTRIAL page is making predictions for what we might expect to see -- number of users registering might go down, proportion of accounts making 100 edits in the first month might go up. It's a really good list of indicators that we could look at. But it doesn't actually say "this is what we'd call a success: more good pages (defined as XXX), fewer spam pages (defined as YYY)".

If we're going to do ACTRIAL as an experiment -- six months on, one month off to evaluate -- then we need some idea of what "success" would look like. Right now, it sounds like the most important indicator is the number of pages in the NPP backlog. Obviously, we know that ACTRIAL would make the number of new pages go down, and the number of pages in the backlog would also go down. If that's the only meaningful measure of success, then we already know that ACTRIAL worked, so it's not actually an experiment, it's just a change in the way that Wikipedia works. So my question is: are there other measures of success or failure for ACTRIAL? That list on the ACTRIAL page doesn't actually answer my question. DannyH (WMF) (talk) 05:03, 24 June 2017 (UTC)

The list of questions provides a basis for metrics that can be used and assessed holistically rather than an exact rubric for success or failure where X% decrease coupled in new pages created with Y% decrease in CSD and only Z% decrease in users reaching autoconfirmed. Success can be data driven and evaluated holistically, especially when the things we are evauluating have no direct monetary value or impact on an organization's bottom line. What we are all ultimately looking for is a way to improve the handling of new pages that aids the encyclopedia. If we look a the trial asking those questions and determine it harmed the encyclopedia as a whole, then the switch should be unflipped. If we look at those questions and decide that on the whole it helped the encyclopedia, then it should be made permenant. Other questions can be added, but ultimately the only question that matters is if considered as a whole, it had a positive impact. TonyBallioni (talk) 05:20, 24 June 2017 (UTC)

TonyBallioni: Yeah, that's a good point. Thanks for expressing it like that; that totally makes sense. Okay, when I talk to folks on Monday, we'll look at that list and see what we can establish baseline stats for. I'll let folks on these pages know what we come up with. Thanks very much. DannyH (WMF) (talk) 05:23, 24 June 2017 (UTC)

TonyBallioni, it will all come out in the wash when we've let ACTRIAL run for its 6 months. One of the problems is possibly partly of my own doing: We now have these discussions spread over four pages including now the talk page of my highly critical essay at WP:KNPP. One thing however that gives me great satisfaction, without feeling smug, is that I have caused more to be discussed on this topic in the last couple of months than was written on it over the last 5 years. That's progress. It's a shame however that DannyH (WMF) will still not read up on the history of ACTRIAL, and follow the current discussions in detail and has decided that he will now only continue a dialog with people who represent his own views. That's not progress. Kudpung กุดผึ้ง (talk) 04:54, 24 June 2017 (UTC)

Yes, I can take some of the blame there too since I restarted the conversation at NPPAFC, which I thought would be a good place to centralize (and still do, to be honest, but you can't really control where people talk.) I do think its a good thing that the conversation at the WMF essay has died down: its a lot more difficult to talk about future plans when you are reacting to something that was generally not well received, and I think shifting the location of the conversation has also shifted it to be broader. TonyBallioni (talk) 05:03, 24 June 2017 (UTC)

DannyH (WMF)...if you get a chance, please take a look at Wikipedia:WikiProject_Rihanna. I'm of the mind that it's not representative of what Jimbo intended when he declared WP aims to offer "the sum of all human knowledge". A recent NPR of a Rihanna marketing spin-off Westbury Road Ent. demonstrates to what extent promoters/publicists/fans will go to get an article in WP. It also demonstrates why ACTRIAL is greatly needed. I doubt the Westbury Road Ent. is an isolated incident. The article was rejected 3 or 4 times including my moving it to draft space which was quickly reverted by the article creator who appears to have 2 or 3 usernames. Consider the promotional/marketing effort surrounding Rihanna, and the time sink it was for at least 4 or 5 editors to get it removed. No wonder there's such a NPP backlog and we're losing good editors due to infinite discouragement. Perhaps we should exert as much effort into keeping our competent editors as we do recruiting new ones. As we've learned, most publicists/promoters/fans/COI editors don't stop with just the one WP article - they create spin-offs as evidenced by the Rihanna project - it's like WP:WikiProject Med for a single BLP. WP is filled with song promotions/marketing, game promotions/marketing, promotions for new actors/actresses/singers/films and so on. Do we really need a standalone article about every hit song ever recorded by a notable singer? How about films that were produced but not yet distributed, or every sports participant who failed to win an Olympic medal, or hockey players who competed in at least one international hockey game, and so on? It seems to me that there really needs to be some serious tightening of WP:GNG to remove the ambiguities and questionable requirements. Our encyclopedia is being inundated by promotional marketing, fan clubs and political advocacies, not to mention all the poorly written articles that are submitted each week. I truly believe ACTRIAL will bring some much needed weed eaters and motorized pruning tools to the project which will unequivocally help protect the integrity of WP by improving the quality of the articles that are published. It will certainly help boost editor morale to know we're not fighting a losing battle. ^{Atsme📞📧} 16:02, 24 June 2017 (UTC)

@Atsme: I don't think that upping GNG would solve anything. If it was taken to AfD because of promotion, then there are two outcomes. One, the article would get deleted because of its promotional nature, or two, it would be cleaned up and, if it passed GNG, kept. If we made GNG stronger, on the other hand, then we would just get more articles deleted that should really not be. And, I don't know how upping GNG would help article creators fight systematic bais. It is hard for me to find sources for the Japanese politicians that I'm creating articles for, and having GNG be made tighter would just make them easy targets for deletion (just to note, I am still able to find sources, so they do satisfy GNG currently, but my point is legitimate articles like those could be at risk for deletion by tightening GNG). The reason, it seems, that there are still some promotional articles getting through is because of a lack of participation at AfD. This is getting better now, it seems, but GNG would not be the solution. It wouldn't help improve participation at AfD, or at least very much. RileyBugz ^会話 _投稿記録 16:18, 24 June 2017 (UTC)

Atsme, I completely agree about the Westbury Road Ent. article, and the need to keep out obvious marketing spam. I disagree with you about WikiProject Rihanna, which looks like good-faith people who are interested in a subject going to a lot of trouble to make sure that the articles in their subject area are as high-quality as they can be. In my view, that's just like any other subject, big or small. I guess the difference for you is that you don't think Rihanna is an interesting or worthy subject, maybe because she's a pop singer, and you're not interested in pop music? (I could be wrong about that, sorry if I'm misstating your view.) But that WikiProject has produced eight Featured articles, and 50+ Good articles. I don't see how that could be seen as having a negative impact on the encyclopedia's quality.

As a similar example, I was recently researching various aspects of the Beatles' career for a writing project about the late 1960s, and Wikipedia's high-quality, detailed coverage of the Beatles was unbelievably helpful. I read the articles about every album, most of the songs, all of the spin-off projects, as well as the articles about the Beatles' breakup and their visit to India. The history of the Beatles is very important for people who are trying to understand pop music, rock music, 60s counterculture, the sociological development of the concept of "teenagers", media in the 60s, and so on -- legitimate areas of academic interest. I don't know if Rihanna will someday be seen as having the same kind of cultural impact -- it's hard to top the Beatles -- but that's the kind of scholarship that a comprehensive, well-written history of Rihanna's career could support.

So conflating the obviously promotional Westbury Road Ent article with what appears to me to be the serious, good-faith work of WikiProject Rihanna -- that's the thing that concerns me about the direction of the NPP over the last couple years. Having a limited group of people deciding which good-faith topic areas are notable risks bias creeping in, and closing off legitimate areas of academic interest. DannyH (WMF) (talk) 19:32, 24 June 2017 (UTC)

@DannyH (WMF): Please AGF. I don't, but you should. Our patrollers aren't persecuting editors because of an WP:IDONTLIKEIT mentality. You should believe that if we think commercial interests and partisans are the problem, it's because we see it all too often and it's true. Treating us like jingoists is only going to further alienate us. Chris Troutman (talk) 19:38, 24 June 2017 (UTC)

@Chris troutman:: Okay, fair enough. I don't know the situation around the Rihanna pages, and you do. -- DannyH (WMF) (talk) 20:09, 24 June 2017 (UTC)

@RileyBugz and Atsme:, re: both of your comments on GNG: its important to remember that GNG alone does not satisfy our notability guideline, because it is only half of the two-pronged guideline. It must satisfy GNG and not fail WP:NOT. A pass of GNG is not a pass of WP:N if it fails NOT. DGG has explained his views on this in the past, which I will try to paraphrase accurately by saying: GNG and WP:N are ways for us to approximate if something fails WP:INDISCRIMINATE. Ultimately NOT is the policy that guides us as to whether something should be excluded from the encyclopedia. A promotional fluff piece can and should be excluded under it, especially if it is just a borderline pass of GNG. TonyBallioni (talk) 20:00, 24 June 2017 (UTC)

Yes, that's an accurate statement of my view. And the problem of whether something fails NOT INDISCRIMINATE and NOT ADVOCACY will always be a matter for judgment, with many equally valid position. For about half of the article at AFd, either a keep or delete could be rationally justified. (Even, as I hope but is not likely, we agreed on firm cutoffs in the various fields for INDISCRIMINATE, I do not think that there ever will be a sharp cutoff for what we want to consider impermissible advocacy.) And as an additional factor, we all know that reasonably well written articles which are proportionate to the apparent importance and where the sources are not absurdly unreliable, will often get passed at AfD, in contrast to a barely literate one filled with exuberant detail, and with mostly on spam references.Article quality does matter. DGG ( talk ) 20:33, 24 June 2017 (UTC)

(edit conflict) RileyBugz, I understand exactly what you're saying as I've participated in AfDs regarding academics in non-English speaking/relatively undeveloped countries. It's not so much about media coverage and fame as it is the overall impact they've made in their respective fields - DGG really helped bring that to light for me. There are always going to be some exceptions to WP:GNG as I've long since learned, particularly when it involves women academics in countries where women have been oppressed - but those exceptions simply need to be discussed as they materialize because there simply isn't any way we can include all the caveats and exceptions to the rule. I think the stats will confirm that the largest number of problems arise from business entities, professionals, and advocacies as I alluded to above. Much of it is pure ole marketing 101 made to appear encyclopedic...but is it, really? Quite a few have gotten by us for a number of different reasons, and I would think a connection to "fame & fortune" is high on the list as demonstrated by the recent Westbury Road ordeal. ^[FBDB] WP is literally being forked senseless and spun like a top,^{terminology I borrowed from dating blahgs}, and I wouldn't be surprised if the implementation isn't being handled by payrolled publicists and marketers, not to discount overzealous fans and politicial advocates. Without belaboring the pros and cons of GNG, I say kudos to editors like Kudpung who have worked hard to draw attention to these issues, and whose experience speaks volumes to what the project and its editors are facing today. The fruits of their labor is in front of us, ripe for the picking. I can't begin to tell you how relieved and inspired I was when I read some of the comments by DannyH (WMF). There really is light at the end of the tunnel.

^{Atsme📞📧} 20:25, 24 June 2017 (UTC)

List of things we need to know

This is, sorted by most-important to least important, the things we need to know to judge what effect ACTRIAL could have.

How many new editors get to autoconfirmed in part by creating non-deleted articles, and, on average, how many non-deleted articles do these users (the ones who got to autoconfirmed in part by creating non-deleted articles) create while non-autoconfirmed?
How many edits do new editors have, on average, after getting their article declined at AfC (only include those that have edited within the past 2 days before getting their articles declined) compared to how many edits they, on average, have after getting their articles deleted. For this latter one, please include the stats for each CSD criterion, because we don't really want the stats of those creating attack pages to be the same as those who are just unaware and creating non-notable articles.
How many new editors stay until at least 2 days before having their AfC submission declined or ok'd (whatever it is at AfC). This is important to include with the statistic above, as otherwise we will not know whether or not the people who get declined at AfC and stay are just the people who really want to help Wikipedia, or our average new editor.
What percentage of the backlog do articles created by non-autoconfirmed editors make up and what percentage of article creations do non-autoconfirmed article creations make up. This is near the end because there are stats on this, but we just need it to be more updated.
How many new articles are created by non-autoconfirmed editors that get tagged for speedy deletion and deleted and how many go through PROD and AfD? (Going through PROD and AfD is much more work)

These are the things that we need to know, in my opinion. RileyBugz ^会話 _投稿記録 15:36, 24 June 2017 (UTC)

RileyBugz, this is really helpful for us. Thank you for putting this list together; I'll include these when I talk with folks this week about how to get more stats. I'll let you know what we figure out. -- DannyH (WMF) (talk) 19:36, 24 June 2017 (UTC)

DannyH (WMF), I would just like to strongly emphasise again once and for all, that ACTRIAL is a trial. There are no notions of 'success' or 'failure' attached to it. It's an experiment to prove or disprove the theories posited by a) the volunteers (more specifically the patrollers who speak from empirical experience), and b) the WMF who don't understand these practical things while clinging onto what might well be an outdated ideology. Kudpung กุดผึ้ง (talk) 21:03, 24 June 2017 (UTC)

It's not primarily a problem of insufficient data

Any method at all with dealing with new users and their articles will be trying to minimize both the proportion of unsatisfactory articles that get accepted, and the proportion of satisfactory ones that get rejected. These can not be both minimized at the same time--any system that tries to prevent any possible bad article, will always have the error or rejecting more good ones, and vice-versa. We need to find an optimal point , and then figure out what measures we can take that will get there. But there is no really objective criterion for what is an optimal point Most of us have our preferences; I know my preferred point has changed: I was 8 years ago trying to minimize any possible rejection of a fixable article much more so than I am today. The optimum also varies among different types of articles, and among different types of bad, and no two of us will see this just the same.

I know to some extent what my own preferences are, and I can argue for them, but I do not realistically expect anyone else here to have the same goals as mine. (for example, It does not in the least bother me keeping articles on every possible geographic feature, but I am very concerned that we not lower the standard for eminence in any profession--both professions I care about like science, and those I do not care about, like sports. I am much less interested in the standards for minor differences in notability in either direction, as compared with keeping out lapses towards promotionalism. These are just samples--there's a whole list). Every one of us wi;l have analogous differences, even if not explicit. If we all had the same goal, we could rationally collect data and design the system towards attaining that goal. But we don't, and we cannot optimize towards different goals at the same time. I can --and I do-- give extensive argument why my preferred goal is the best for WP, but I would be surprised if anyone else sees it the same way. In arguing for or against a particular method, I naturally judge according to how far it will match my preferences in different aspects and in different fields. And regardless of goal, there is no magic trick that will permit us to simultaneously optimize accepting the good and rejecting the bad.

There are in fact other criteria than the results. We also want a system we can deal with using the available people. We want a system which will encourage other to join. We most of us want a system that involves a minimum of complicated procedures. We most of us want a system that avoids exaggerating the tendency to dispute minor issues. I at least want consistency, so we can tell people what will be accepted and what not--and I am willing to sacrifice many other preferences to get this--most others here seem to think this less important.,And so on. But we can not do all of these at the same time. (As an example, there will be fewer errors if we have multiple levels of review--but this will also increase the workload on the people dealing with this).

We will never settle these differences. There is for me little practical advantage in having detailed data that different of us will interpret in opposite directions. More important, we can not simultaneous collect detailed data, and also move quickly towards solving out problems. We will always need to proceed on incomplete data, and there is no optimum place to stop collecting data and decide. I think we certainly should continue gathering data, but that's partly because I'm interested in this data for understanding our system, even if it will be of little immediate practical importance. In a different direction, I am so much involved in doing the practical day-to-day work on these problems that I want almost anything which will attract more people to help immediately, than get what I think is an optimal balance of inclusion. Therefore, I tend to support suggestions like ACTRIAL even though I think it will have the wrong balance, because it will decrease the immediate workload. DGG ( talk ) 00:41, 25 June 2017 (UTC)

To encapsulate what you so elaborately described DGG, it's what I referred to above as "unsustainable development" which requires some level of modification to accommodate the rapid growth rate and avoid potential implosion. ^{Atsme📞📧} 01:35, 25 June 2017 (UTC)

I do not see it that way. The problem of reviewing the new articles was here when I joined in 2006. It was essentially the same problem, but the reason we didn't see it as prominent was that things disappeared from the feed in 30 days, and we couldn't track them. A he result was even worse than now. Deletion was even more erratic--many admins were simply deleting single-handed, the standards at AfD were even more inconsistent, and there was even less agreement on what the interpretation of the notability guidelines were. We had little agreement on NOTPLOT and on articles about trivia. We didn't have BLPPROD, and there was still incomplete development of BLP standards and ONEEVENT. There was much less awareness of the problems of coi and promotionalism. The result was a heritage of CCOI promotional articles that we are very far from dealing with. Most of our longstanding articles on non-profits were coi then, and remain so now. We did have fewer problems with barely notable business enterprises, as most of them didn't know about us yet, and the SEO industry was much more primitive. Over the years, there have been many improvements, and much more internal awareness of our processes. Most of the truly erratic people here then have left for one reason or another, and we This has been compensating by the increasing difficulty in getting enthusiastic new recruit, and by the increasing flood of articles, along with general public recognition of our international reach. We're bigger, and in my opinion considerably better. We are doing everything we used to do better than we did in earlier years, and could probably continue growing at our present level indefinitely without anything imploding.

But in another sense, you are right: there is an unmet need: What we have not done is to develop standards appropriate for the role in which people see us. We ourselves never thought we would be in any sense a reliable authority, but people now think we are. It is irresponsible not to try to match the expectations--but the difficult in being a reliable crowd sourced encyclopedia is enormously greater than in just being a crowd-sourced encyclopedia.

Since you feel I've been too elaborate, I'll try a shorter summary: we do not agree of what we ought to be trying to achieve, and we are consequently trying to achieve inconsistent things. We are trying to optimize something which has no optimum.

And even shorter: my own priority is recruiting & encouraging new editors who are or will become competent. This is the only way to maintain and improve standards. DGG ( talk ) 05:33, 25 June 2017 (UTC)

I understand - did not intend for my use of "elaborately" to mean it was too much - and trust you will understand why I see things much differently regarding the "unmet need". I'm of the mind that if we don't take action to resolve the looming issues plaguing editors and this project, and continue with a focus on recruitment rather than retention, the result will be throwing the baby out with the bathwater as evidenced by current "editor retention" issues. We're neglecting the resources we currently have by not focusing on the problems that plague them which includes a growing amount of work with no light at the end of the tunnel. If the goal is editor turnover, then we're currently headed in the right direction. I think we need to focus more on what's happening in that dark tunnel, and finding a way to escape the darkness of the unknown by incorporating achievable goals, like ACTRIAL. If we aren't able to maintain what we have because of burnout and low morale, how can we expect to attract new recruits? Fewer prospects are willing to buy a house that has a gaping hole in the roof...and that speaks volumes when you stop to consider the damage a gaping hole can cause. I'm of the mind that it's of the utmost importance to fix the problems that are plaguing editors now - focus on editor retention not editor turnover - which in turn will fix new editor recruitment issues. To do otherwise perpetuates starting from square one, only instead of dealing with 2 million articles, we're dealing with the reality of 4x+ that amount, a shrinking editor base, and a growing amount of garbage, promotion, and spam. ^{Atsme📞📧} 20:54, 25 June 2017 (UTC)

Draftify/Userfy

I don't think I'm alone in forgetting to mark pages as patrolled before moving them to draft or user space - where the [Mark this page as patrolled] option no longer exists. Personally I'd find a reminder useful in the text at MediaWiki:Movepagetext along the lines of:

Note to admins and reviewers: Mark the page as patrolled before moving it out of mainspace.

Any thoughts? Cabayi (talk) 19:18, 24 June 2017 (UTC)

Shouldn't we leave them unpatrolled, so they'll go back in the queue if/when they're moved back to mainspace? Or am I misunderstanding? – Joe (talk) 19:48, 24 June 2017 (UTC)

Why would we mark them as patrolled? As Joe said above, when they go back to the mainspace, they won't necessarily be better. RileyBugz ^会話 _投稿記録 20:37, 24 June 2017 (UTC)

I honestly don't know if the review is persistent through the move to draft space and back, but I would expect that it is, meaning we should definitely not be reviewing prior. TimothyJosephWood 21:16, 24 June 2017 (UTC)

I agree that they should probably remain in an unpatrolled state.- Mr X 23:26, 24 June 2017 (UTC)

Thanks to all who offered an opinion. That leads us to the easiest fix... I stop worrying about unpatrolled & unpatrollable User pages on Special:NewPages. Done. Cabayi (talk) 17:08, 26 June 2017 (UTC)