Jump to content

Wikipedia talk:New pages patrol/Reviewers/Archive 11

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 5←Archive 9Archive 10Archive 11Archive 12Archive 13→Archive 15

Instant Indexing, again

I have another example of the accidental indexing phenomenon. This phenomenon, which I have to assume is a misfeature and not a feature, is likely to be taken advantage of by spammers to get their pages out on Google. The page Www.rkriponkhan.com was created with no content. It was promptly tagged for WP:A3. User:Bbb23 then pulled the A3 tag with the note not to apply the A1 or A3 tags for 10 or 15 minutes. The page was then listed in the New Page Patrol list with a green check mark. That is, it was listed as having been reviewed. I am sure that untagging administrator Bbb23 did not mean to be reviewing the page, only tagging it as not yet ready to be deleted as an empty page. As I have been realizing, New Page creation appears to be a system that is designed primarily to speed up the indexing of pages by Google, with a few checks on hasty indexing by Google, but the checks are all afterthoughts that do not work very well. Quick indexing of pages by Google works very well and very fast, and the various checks on it sometimes manage to work for a little while, but have cracks that crud falls through either by accident or on purpose. In this specific case, can the tagging be fixed so that just removing a speedy tag doesn't constitute an approval of the article? Robert McClenon (talk) 16:29, 24 June 2017 (UTC)

Does anybody understand what I am trying to explain? I get the feeling that I am explaining into a vacuum. Robert McClenon (talk) 16:29, 24 June 2017 (UTC)

Actually, it was I who marked the page reviewed after you marked it for A3 deletion. The log shows that Jake Brockman marked it as unreviewed. I don't understand how an article can be marked as unreviewed before its been marked as reviewed, but that another question. I'm not especially concerned about article like this being indexed by Google in the short term, unless of course they slip through the NPP queue entirely.- MrX 16:46, 24 June 2017 (UTC)
The page had previously been marked as reviewed by Bbb23. I think that the fact that it is now marked as unreviewed without being reviewed illustrates that the marking of pages as reviewed is buggy or confused. However, it seems that removing a deletion tag from a page causes it to sort of be marked as reviewed, but not permanently, or something. Will someone please look into the details of what causes a page to be marked as reviewed and unreviewed? Please?! The problems with review tagging can be manipulated by spammers. Robert McClenon (talk) 22:03, 24 June 2017 (UTC)
You User:MrX are not particularly concerned about an article being indexed by Google in the short term. The problem is that the oddities of indexing can be used by spammers, and I think that I know a way that a spammer can deliberately pull a page through the NPP queue, but I don't want to say what it is. Robert McClenon (talk) 22:03, 24 June 2017 (UTC)
It would certainly be preferable if spam articles were not indexed. I'm now questioning whether I should even be marking CSDed, AfDed, and PRODded articles as reviewed. I thought I was doing a good thing, but maybe a better practice is to leave them unreviewed. Does anyone have any guidance on this?- MrX 23:24, 24 June 2017 (UTC)
AFAIK the __NO INDEX__ is embedded in the CSD, PROD, and AfD templates rather than on the article. This means that when such a tag is removed, the page is open again for indexing. Google is very quick with its resident Wikipedia bot to index new pages. Sometimes it's only a matter of seconds and that's exactly the work around that SEO spammers will exploit. Kaldari is the goto engineer on these issues. Kudpung กุดผึ้ง (talk) 03:50, 25 June 2017 (UTC)
I just (empirically) confirmed that when articles by new users are created they have <meta name="robots" content="noindex,nofollow"/> in the head. tagging for CSD, reviewing, unreviewing, retagging, and re-reviewing does not have any effect on the noindex status. Is it possible the resident Google bot ignores the noindex directive?- MrX 15:15, 25 June 2017 (UTC)
It would be helpful if you could provide specific examples. A page can have a template that sets NOINDEX and still not have the robots meta tag. For example; List of housing cooperatives in Canada is unreviewed and at AfD, so you'd think that it should not be indexed, but it IS indexed, and the reason (per Wikipedia:Controlling_search_engine_indexing is that it is old enough that The magic word and the template do not work on them. Mduvekot (talk) 17:09, 25 June 2017 (UTC)
DAV Public School,Sawang is the specific example that I was referring to. List of housing cooperatives in Canada does not respect __NOINDEX__ or {{NOINDEX}} because it is more than 90 days old. See WP:NOINDEX#Indexing of articles ("mainspace").- MrX 17:19, 25 June 2017 (UTC)

The Problematic Sequence

I had been hesitant to go into great detail as to what the exact problematic sequence of steps is, but I think that the spammers know the exact sequence as well as we do, so that there WP:BEANS doesn't apply and there is no benefit in avoiding saying what is going on. The unintended review occurs in the following situation:

  • Editor A creates Article X about Company Y or Person Y. Article X is not indexed, and, having not been reviewed, is not indexable. For the sake of explanation, let us assume that Article X is marginal as to notability and marginal as to promotional content. That is, it isn't obviously reasonable, and it isn't obviously crud, but it is cruddy.
  • Editor B tags Article X for speedy deletion as A7 or G11 or both. What happens behind the scenes is that the article is now marked as reviewed, but that review status is overridden by the deletion template.
  • Editor C removes the deletion template from Article X. What happens now is no longer behind the scenes. The article is reviewed, and is promptly indexed by Google.

If Editor D marks Article X as unreviewed before Google indexes it, the article isn't indexed. Likewise, if Editor C marks the article as unreviewed at the same time as they pull the deletion template, the article won't be indexed, but that is dependent on Editor C being a knowledgeable NPP reviewer who understands what is going on and wants to leave the indexing decision to yet another reviewer. However. otherwise, the tagging and untagging for speedy deletion has the effect of indexing the article. Editor C (or Editor A, since spammers don't care about the rule against an article-author untagging it) is, in effect, indexing the article. It doesn't take a real reviewer. That is the problematic sequence. Comments? In the cases we have seen, the article does eventually get deleted. But in the cases we haven't seen, the article gets indexed, and no one notices except the spammer. Robert McClenon (talk) 17:01, 25 June 2017 (UTC)

That's not consistent with my experience in the example I gave in the section above. Have you verified that it actual works that way?- MrX 17:25, 25 June 2017 (UTC)
That is what I have observed. How have your observations been different? Robert McClenon (talk) 04:14, 26 June 2017 (UTC)
@Robert McClenon: For reference, the NOINDEXing feature is documented at Wikipedia:Controlling search engine indexing#Indexing of articles ("mainspace") in case there is still any confusion about how it works. The first problem here is that a reviewer improperly tagged an article as A3 immediately after it was created. The second problem is that when the admin removed the A3 tag, they didn't also change it back to being "unreviewed". I don't see any evidence that spammers are abusing this process, and if they were, it would easily be caught by people monitoring this feed of editors removing speedy deletion tags. Another possible solution would be to create an edit filter that prevents editors from adding A1 or A3 tags until 10 minutes after the article was created. If that sounds like a good idea, let me know. Kaldari (talk) 06:23, 26 June 2017 (UTC)
User:Kaldari - No. First, you are putting the burden on the reviewers. While it is generally a reasonable rule that an article should not be tagged too quickly, if there really is a hard-and-fast rule that it should not be tagged quickly, stop showing it in the New Page Feed. Second, the real problem isn't the addition of the tag in the first 10 minutes; it is the removal of the tag. If you really want to help, provide a filter that prevents non-autoconfirmed editors from removing the tag. Third, I don't think that editors or admins know that the mere removal of a tag renders the page as "reviewed". I don't think that it should, but, besides, I think that should be publicized. In general, it sounds as though you are trying to find ways to say that the problem is one for the reviewers and admins, not one that maybe involves work by the developers or the WMF or someone. (After all, the WMF is busy coming up with things that the community of editors don't want or need, but I shouldn't say that.) Robert McClenon (talk) 14:25, 26 June 2017 (UTC)
Until reading this thread I have been marking as reviewed an article that is in the NP feed that has been prodded by another user if I agree with the reason. My logic is that if I had tagged it myself using the page curation it would automatically be marked as reviewed, the trouble is that anyone can remove a Prod without justifying it (with the exception of unsourced BLP). So removing a PROD from an uncontroversially poor article is clearly permitted and this article will be referenced if I or another reviewer is not quick enough with the unreviewing. If we leave the PRODDED CSD and AFD articles as unreviewed then they will pop up every time and NP reviewers will waste time checking if the tag is justified or not. I personally feel that a script that automatically unreviews it when the tag is removed is a better way to go. Domdeparis (talk) 14:51, 26 June 2017 (UTC)
@Robert McClenon: I have been discussing such a filter at Wikipedia:Edit filter noticeboard#Special:AbuseFilter.2F29, but there doesn't seem to be consensus for it there. Please join that discussion if you feel such a filter would be useful. Kaldari (talk) 17:28, 26 June 2017 (UTC)

Backlog

We're below 20,000 for the first time in a while (at the time of writing, at least). Well done, everyone! Boleyn (talk) 20:33, 25 June 2017 (UTC)

Not to burst any bubbles, but we dipped below 20,000 yesterday. This morning, I saw that it had climbed above 20,000 again. It seems we're barely treading water here.
When are we starting ACTRIAL and when will WMF help us with the features we've requested?- MrX 20:43, 25 June 2017 (UTC)
'Barely treading water' is the right description and anything less than a reduction to its pre June 2016 levels is nothing to rejoice about - even 5,000 would be too much. Reading between the lines, MrX (which I know of course is the wrong thing to do), it seems (to me at least) that the group of Foundation developers that is showing concern for these issues has no intention of actually physically addressing them by developing the required software. Although as Evad37 has demonstrated, these features need only hours rather than months to code up, DannyH (WMF) constantly pleads the classic 'no time, no personnel, no money' mantra, probably knowing full well that good natured volunteers like Evad37 will do the work for free that the WMF is supposedly being paid for out of the donations generated by our work. The move to draft feature was first registered on Phab 18 months ago and there has been not one single attempt to address it. The problem is that Wikipedia has already become such a hotchpotch of useful user developed .js that browsers can't cope with them. The vast majority of them should be MediaWiki extensions or API.
The only way to get anything done at this stage, AFAICS, is to escalate through the echelons of the totally vague hierarchy of the WMF until someone in a position of responsibility , senior to Danny Horn, takes note, and preferably Maher herself. At the moment, the WMF smacks of Yes, Minister - which even Margaret Thatcher remarked was closer to reality than the sitcom audience realised.
A small group of editors is working to decide how and when ACTRIAL will be rolled out. It's mainly a question of deciding which of the proposed scripts to use, and being sure that now that Scottywong has retired, we know how to obtain and process the stats that are needed to monitor the trial. Even the 'profis' at the WMF are having difficulty knowing how to use the database. Kudpung กุดผึ้ง (talk) 03:25, 26 June 2017 (UTC)
@Boleyn and MrX: where can one see the backlog? —usernamekiran(talk) 03:34, 26 June 2017 (UTC)

Usernamekiran Special:NewPagesFeed, right-hand corner above the newest/oldest options.

Personally I do rejoice to see it going steadily, though very, very slowly, down. Yes, there's still a big issue and other solutions need looking at as most people agree, but I don't think ther's anything I can personally do about that. However, acknowledging the achievements of people working hard on this project is important too, and editors' hard work has helped. I find the doom and gloom on this page affects my morale and ability to keep working on it. I hoped pointing out asmall positive might be encouraging. Boleyn (talk) 05:38, 26 June 2017 (UTC)

@Boleyn: My comment was not intended to diminish the accomplishment, or to not acknowledge the 4080 reviews that you have done in the past three months which is a huge contribution to having reduce the backlog from it's high of over 22,000. But I am concerned that the trend will reverse if even a few of us stop performing this Sisyphean task.- MrX 16:03, 26 June 2017 (UTC)
Boleyn, thank you for your post here and providing much needed cheer. I agree the doom and gloom can be over the top sometimes, and your efforts for this project really do deserve recognition more than a barnstar can provide. All I can offer is my thanks. I agree that there are concerns with it going back up again, but the current dip is a positive and shows what this project can do when we have people focused on it. TonyBallioni (talk) 16:09, 26 June 2017 (UTC)
I agree with all of you guys. @MrX and TonyBallioni: Kudpung is correct as always. There are like 10% editors who are actively doing this task. And then there are few editors from that 10%, who have no idea whats going on this talkpage, or with the project. They dont share their opinions, or their observations (if they have any).
I always felt more editors at this task will be beneficial. Setting up high goals for oneself, or for others isnt always wise. But if we had around 1000 reviewers, and they did only 5 reviews per day on an average, that would be ~5000 articles per day!
@Boleyn: thanks for the update, it was nice to know that our efforts are not going useless. :-) But as MrX noted, this is a never ending task, inversely which seems to increase with each day. Under the current circumstances, all we need is more reviewers. And thanks for the information about finding the backlog number. I never use that. NewPagesFeed. —usernamekiran(talk) 18:03, 26 June 2017 (UTC)

There are advertisement on enwiki like "Wikipedia loves uniforms" submit photos of Indian law, and enforcement forces, and win prizes. There is one more ad currently going on about photos of heritage sites. Why cant there be a "notice" similar to these ads regarding NPP/R? Describing wikipedia needs reviewers? pinging @Kudpung, Kaldari, MusikAnimal, and DannyH (WMF): —usernamekiran(talk) 18:09, 26 June 2017 (UTC)

@Usernamekiran: I'm not sure what advertisement you're referring to. Could you elaborate? Kaldari (talk) 18:18, 26 June 2017 (UTC)
(edit conflict) @Kaldari:, I think Usernamekiran is referring to things like 'Wiki Loves Monuments,' 'Wiki loves the Dominican Republic,' and 'WP:Wiki Loves Pride,' which I believe are always advertised on e.g. watchlist pages (and possibly elsewhere, I tend to block the bloody things ASAP). A possible reason- and I'm only painting pictures- might be that these things are global 'outreach' projects, whereas WP:NPR is purely internal, and might even be in breach of WP:CANVAS (I hope I'm not speaking out of turn, but I think I'm right in saying that even Kudpung's occasional mass mailings wound some people up some of the time!). Just MHO of course. — O Fortuna semper crescis, aut decrescis 18:34, 26 June 2017 (UTC)
@Kaldari: I took that screenshot after you posted the reply. —usernamekiran(talk) 18:30, 26 June 2017 (UTC)
FIM, I don't think it'd be a CANVASS violation (there's no proposal). I also just think it wouldn't make sense for a watchlist or site banner: we don't advertise Huggle or CopyPatrol in that way, partially because they are maintenance tasks that require experience. I think the same would apply here. TonyBallioni (talk) 18:37, 26 June 2017 (UTC)
FWIW, it's possible to create CentralNotice banners (the kind that are used for fundraising) that are only displayed to users with certain rights. For example, you could only show it to users that are admins or page reviewers. I'm not sure if this would be considered too "spammy" for CentralNotice though. You could always propose it at meta:CentralNotice/Request and see what people think. Kaldari (talk) 19:12, 26 June 2017 (UTC)
What about only to extended confirmed users who are browsing recent changes (instead of on their watchlist) and who also don't have NPP? TimothyJosephWood 19:16, 26 June 2017 (UTC)

ACTRIAL update

I mentioned this above, but the WMF has agreed to ACTRIAL in principle. You are invited to join the discussion on WT:NPPAFC. TonyBallioni (talk) 20:48, 27 June 2017 (UTC)

Notice has been requested!

The request for the notice/banner/advertisement can be found here: link to meta
—usernamekiran(talk) 19:03, 27 June 2017 (UTC)

Mass pinging. @Kudpung, MusikAnimal, Kaldari, Boleyn, MrX, Fortuna Imperatrix Mundi, I dream of horses, TonyBallioni, Timothyjosephwood, Scope creep, BigHaz, SweetCanadianMullet, Ymblanter, Jupitus Smart, Jupitus Smart, Atsme, KGirlTrucker81, and Chrissymad:
I apologise for such an unnecessary mass ping. But I thought the editors who had been putting a lot of time n energy to this project should know about this. —usernamekiran(talk) 19:19, 27 June 2017 (UTC)
I... wouldn't necessarily say that the primary purpose is to ward away unwanted articles. Maybe that's about half of it, but the other half is incorporating good new articles coherently into the project (MoS, links, tags, projects, stub sorting, categories, sources, etc...). TimothyJosephWood 19:25, 27 June 2017 (UTC)

Thanks, Usernamekiran, I think that's a positive step. I know I had no idea you could nominate yourself as a patroller, or even really that pages were patrolled systematically, until I was given the right. There are lots of people who would be very good reviewers who could be recruited. Boleyn (talk) 19:53, 27 June 2017 (UTC)

It will be interesting to see what develops - great idea, Kiran. Atsme📞📧 19:56, 27 June 2017 (UTC)
  • I have no real opinion either way on whether this is a good idea or not. I'm only commenting to say that if this goes live you might want to point the ad towards WP:NPP, since that is where the tutorials are. I also just BOLDly updated that page to point people at PERM. TonyBallioni (talk) 20:03, 27 June 2017 (UTC)
(edit conflict)@Boleyn: Same here. I had absolutely no idea about it either until I came across a user who had the right. After that, I stalked this talkpage for two weeks, and then requested the right.
@Timothyjosephwood: yes, you are right. But I didnt know how to put it there. It would be great if you guys could help with wording/design. This is another reason for the mass ping. :-)
And thanks for the support guys.
—usernamekiran(talk) 20:06, 27 June 2017 (UTC)
I changed the landing page as per your suggestion Tony. —usernamekiran(talk) 20:13, 27 June 2017 (UTC)
(edit conflict × 2) The ad should probably target editors who have been patrolling without the right. Otherwise, just a thought. KGirlTrucker81 huh? what I've been doing 20:17, 27 June 2017 (UTC)
@KGirlTrucker81: I thought about that too. But then I realised, the current filters would involve these users automatically. And if we use only that filter, then we might lose some other prospective users. —usernamekiran(talk) 20:59, 27 June 2017 (UTC)
  • It can't hurt to try to get more volunteers, but the banner text probably needs to be worded in a more positive way. (Also, you didn't link to a banner in the third box). I suggest wording along these lines:
    Experienced editors: Help preserve Wikipedia's integrity by patrolling new articles. Become a "New Pages Reviewer"!
- MrX 21:04, 27 June 2017 (UTC)
  • it is very impressive X! I will update it. Yes, i didnt know how to select a banner. I thought, a good banner can be created after the request is accepted. —usernamekiran(talk) 23:59, 27 June 2017 (UTC)

New Editor Question (again)

The statement was made a week ago, by an administrator and functionary, "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." What I would like to know, and I think some other editors would like to know, is a two-part question about the new editors. First, was that statement meant to refer to all new editors, or specifically to new editors whose first major contribution is a new article? If it is meant to refer to all new editors, then maybe it is outside the scope of this page, but, since it refers to rejection, it appears to apply to those who contribute an article (or occasionally group of articles). Second, is the statement that they come to contribute to the encyclopedia based on empirical knowledge, or is it a statement of belief? I am aware that it is an article of belief that new editors come to contribute to the encyclopedia. I do not have a whole lot of evidence with regard to those who throw an article over the fence, but perhaps I have a lack of faith, or have not seen the right metrics, or am looking in the wrong way or wrong place.

If it is empirically true that they come to contribute to the encyclopedia, and we are losing a large potential resource, then it is important to know that this is empirically true. First, in that case, we need a volunteer corps of meeters and greeters to supplement New Page Patrol. (Just dumping on New Page Patrol for not being sufficiently welcoming to editors who contribute crud will only burn them out, although dumping on the existing volunteers is very much the Wikipedia way.) Second, if they really do come to contribute to the encyclopedia, maybe WMF can actually figure out some way to engage them to further increase our meaningless metrics. Robert McClenon (talk) 16:02, 27 June 2017 (UTC)

  • I think this would be important as a part of our research when ACTRIAL is implemented and we can work with the WMF to figure out what percentage of new users become active regularly during that time period vs now. That data could be useful to help us see in what areas we could improve welcoming new users and if direct page creation of inadequate pages prevents potential good users from participating. For those who don't follow the work group the WMF has agreed to ACTRIAL in principle. The question is if an RfC is needed now, and what questions we should be asking in the trial period. TonyBallioni (talk) 16:19, 27 June 2017 (UTC)
Well, since I haven't gotten an answer to my question, I am inclined to assume that the statement is a essentially a statement of belief, and is not meant to be questioned factually. In any case, if ACTRIAL goes into trial (experiment), then we will see a decrease in new articles, and presumably a decrease in the number of new articles that are speedy-deleted. Robert McClenon (talk) 23:30, 29 June 2017 (UTC)
If it is empirically true, as opposed to a statement of belief, then we do need to retain and utilize them, which means that we need volunteer meeters and greeters to mentor them. If it is an article of belief, it may be enough to repeat it. Robert McClenon (talk) 23:30, 29 June 2017 (UTC)

Editors with most unreviewed pages

I have been reviewing new articles for a few days (thanks Kudpung for granting me the right) and made an observation.

A few months ago DrStrauss asked about a way to sort users by number of unreviewed pages. I created a script that generates such list, see User:Rentier/NPP. Listed below are aggregate numbers.

Unreviewed Articles Editors Articles Backlog fraction
>50 10 650 3%
>20 56 1996 10%
>10 207 4108 21%
>5 529 6450 32%
>2 1434 9759 49%
- 10228 19851 100%
Generated: 2017-06-29 15:20:58 UTC

One fifth of the backlog can be removed by reviewing articles created by just 207 (2%) most prolific editors. Since articles created by the same user tend to be of similar quality, I think that reviewing is made much more efficient by grouping the reviews by user rather than just going through the list chronologically.

Rentier (talk) 16:14, 29 June 2017 (UTC)

@Rentier: thanks for the script, I'll check it out!
@Kudpung: would granting some of these editors autopatrolled be an idea?
Thanks, DrStrauss talk 16:17, 29 June 2017 (UTC)
Thank you! I had a hackish script that did something similar and found that way of reviewing very effective when my interests and expertise overlapped with the editor. Once you have the user name, you can use the Were Created by filter in the Page Curation tool. The only problem I see is that the lack of diversity in reviewers might give the editor the impression that you're reviewing them in stead of a "random" article. Not all editors appreciate that, especially if you keep bringing up the same issues. In other words; this is only an effective way of reviewing if there are few or no problems with the articles. Mduvekot (talk) 16:26, 29 June 2017 (UTC)
From first glance, a lot of the articles created by the top creators are mass creations of similar subjects that, well, how to put this diplomatically, leave a lot to be desired. One entry lists like List of universities in Sint Eustatius and rawly formatted statistics like 1979–80 FC Barcelona season aren't really ideal for granting autopatrolled to. Pinging Primefac since he's helped with cleanup of similar things before. TonyBallioni (talk) 16:31, 29 June 2017 (UTC)
As far as I can tell (I'me at #15( allmost all the top entries in the list are candidates for categorical deletion. For example, I would seriously consider the deletion of every article that sockpuppet User:Envale has created. Granting autopatrolled is out of the question, as far as I'm concerned. Mduvekot (talk) 17:47, 29 June 2017 (UTC)
To clarify, I was only talking about the ones worthy of autopatrolled. DrStrauss talk 18:26, 29 June 2017 (UTC)
@Mduvekot: and re the Envale pages, how could we go about a mass deletion? DrStrauss talk 18:28, 29 June 2017 (UTC)
Hi DrStrauss, there is a process for nominating multiple articles for deletion. Mduvekot (talk) 19:20, 29 June 2017 (UTC)
@Mduvekot: I've done it. DrStrauss talk 20:12, 29 June 2017 (UTC)
and several editors workingwith academic articles, myself included, have asked for a speedy keep. We do not automatically delete articles by sockpupetts if reliable editors contribute substantially to them, and several of us have offered to do so. We need time for that. This was in my opinion a reckless use of group nomination, done without consideration of WP:BEFORE. Several of the people are unquestionably notable under WP:PROF, such as holders of distinguished chairs, and I think It will not be difficult to show the notability of at least half the academics and a good many other articles--one is about a major prize from the leading professional association in the field. I agree it is frustrating when poor editing makes work for others, but that's part of what comes from working on a open editing project. DGG ( talk ) 00:42, 30 June 2017 (UTC)
Oh my... those FC Barcelona lists are hideous. Burn them with fire, please. If no matches, then just redirect to the "seasons" article. If users are creating shite like that, we might have to take the SvG approach and just nuke 'em all and start over (though obviously that would take an AN discussion). Primefac (talk) 23:21, 29 June 2017 (UTC)
Primefac, those are the seasons articles from what I can tell... TonyBallioni (talk) 23:35, 29 June 2017 (UTC)
Sorry, I meant redirecting the terrible "XXXX season" articles to List of FC Barcelona seasons, which has just as much (if not more) useful information about the individual seasons. Primefac (talk) 23:57, 29 June 2017 (UTC)
@Mduvekot: Yes, that's something to keep in mind. On the other hand, sometimes it may be better for the creator to receive a single thoughful comment rather than several one-liners from a bunch of different people. Rentier (talk) 03:05, 1 July 2017 (UTC)
It's also a good idea to review these quickly, since students are typically engaged with Wikipedia for a short time. Thanks Rentier! – Joe (talk) 12:04, 1 July 2017 (UTC)
@Rentier: Whats up with these so many articles about football clubs, and players? I know nothing about football/soccer/rugby so I have not reviewed any article yet.
I just reviewed few articles of films/short films by same creator. Harddly 3-4 lines, and no sources at all. One of the articles is already ProD'ed. I added notability tag on others. I will ProD it after 2-3 days if they dont improve. I didnt CSD or ProD them only because it is claimed that the films were produced by state government. —usernamekiran(talk) 12:18, 1 July 2017 (UTC)

Stats

Hi,
Is there some way to find how many articles/pages were reviewed in June? Thanks a lot. —usernamekiran(talk) 12:20, 1 July 2017 (UTC)

If I did this query correctly, it is 20,080.- MrX 12:46, 1 July 2017 (UTC)
Thanks X. Kindly feel free to make changes at User:Usernamekiran/NPR Stats, or to suggest changes. Best, —usernamekiran(talk) 14:00, 1 July 2017 (UTC)

Improper A-PAT?

I recently came across an article of a non-notable Bollywood song. The creator is A-PAT, and more than 90% of articles created by him are not notable. Most of the sources used are dependant of subject. I wanted a second opinion of an experienced user. Would somebody please look at it? X seems to be online. —usernamekiran(talk) 14:06, 1 July 2017 (UTC)

I sampled about 25% of their new articles. Most were acceptable, and only one relied entirely on unreliable sources.- MrX 14:13, 1 July 2017 (UTC)
The articles are in encyclopaedic tone, and they have sources. But the sources are not "independent". For the articles of songs, the sources are music sites, music review sites, celebrity gossip sites and similar. They are not "independent" of subject. The ones that are independent, make passing reference to the subject (source talks about the film, passing reference to the song from that film). By this logic, I can create an encyclopaedic article for every person that has a profile on IMDB. —usernamekiran(talk) 14:33, 1 July 2017 (UTC)
(edit conflict) I think a more general issue you're going to see here is that people simply aren't familiar with South Asia enough to comment on whether something is a notable band or song. Hell, I'm not familiar enough with the North American music scene to comment on that (my rule would be 'if they aren't played on a mainstream radio station, they don't get an article' but that would never get consensus and I also have no idea what is played on some mainstream stations.) Kiran, if you are familiar enough with them to make an argument for it at AfD, I think you should, but I also don't think the crowd here is likely to be that familiar with the topic area. TonyBallioni (talk) 14:37, 1 July 2017 (UTC)
The sources are not very independent, that's true, but you will find other editors who use a much stricter definition of independent. Since I can't distinguish between these particular sources being promotion websites disguised as news sources, or actual news sources, I used the stricter definition of independent.- MrX 14:52, 1 July 2017 (UTC)
(edit conflict) Yes, I should wait for 3-4 days though, before taking the article to AfD. Hopefully, editors from India will participate in that discussion. May I know where are you from Tony? —usernamekiran(talk) 14:55, 1 July 2017 (UTC)

An article indexed on google before reviewed

I selected the name, clicked "search on google", then moved the article to Angira Dhar from Angira dhar, I skimmed through the sources (then closed the tabs); edited the article, and then marked it as reviewed. When I opened the google tab, I saw the first result was her wikipedia article. ie: The article was already indexed by google before it was reviewed. It was created on 13 June 2017, and reviewed just a few minutes ago from now. —usernamekiran(talk) 20:46, 28 June 2017 (UTC)

Unfortunately, I am not surprised. I think that some actions that do not formally mark an article as reviewed cause it to be indexable. The article was categorized on 21 June. Although it was not marked as reviewed then, is it possible that categorization has the unintended backdoor effect of reviewing the article? Robert McClenon (talk) 03:15, 29 June 2017 (UTC)
Here is one more article that has not been reviewed yet, but indexed on google: Alice Torriani. Pinging @Kudpung, MusikAnimal, Kaldari, Danny H (WMF), and Robert McClenon: —usernamekiran(talk) 01:22, 3 July 2017 (UTC)
She has been out there for months without being formally reviewed. In any case, we have evidence that articles get indexed without being reviewed. Robert McClenon (talk) 02:44, 3 July 2017 (UTC)
@Robert McClenon: Sorry, I forgot the 90 days thing. —usernamekiran(talk) 05:01, 3 July 2017 (UTC)

Question about The Wondrous Tale of Alroy

Hello, folks. I was under the impression that, unless the creator had the "autopatrolled" right, any articles newly moved from AfC into Main space would not be indexed until reviewed by someone here. But the above-named article, moved into Main space just a few hours ago, is already showing up in a Bing search. And I see no evidence that it has been patrolled. Am I misunderstanding something about the process? Any comments you have will be greatly appreciated. NewYorkActuary (talk) 21:12, 5 July 2017 (UTC)

Does User:TheSandDoctor have the autopatrol privilege? Robert McClenon (talk) 01:04, 7 July 2017 (UTC)
Thanks for responding. No, SandDoctor doesn't have "autopatrolled", but does have the "new page reviewer" right. Might the software be treating those two rights as the same? NewYorkActuary (talk) 01:11, 7 July 2017 (UTC)
I would assume that the software is treating something as something. We already know that the software does not behave as we are told it behaves. Robert McClenon (talk) 01:29, 7 July 2017 (UTC)
That's ... ominous. I'll keep my eye on other acceptances over the next few days, to see if that is indeed what is happening. Thanks again for the response. NewYorkActuary (talk) 01:35, 7 July 2017 (UTC)
I understood it to be that, since I have the "new page reviewer"/patroller right, that it simply patrolled it for me since I can patrol any non-patrolled page that I did not create. Which, to my knowledge, that particular draft (and any others I approve) fall into the category. @Robert McClenon: @NewYorkActuary:. --TheSandDoctor (talk) 02:20, 7 July 2017 (UTC)
Yes, that makes sense. The confusion is that the "patrolling" doesn't show up in the logs for the article. NewYorkActuary (talk) 02:28, 7 July 2017 (UTC)
@NewYorkActuary: Oh, I see. That makes sense. That is indeed odd. I wonder why it doesn't show in the logs. --TheSandDoctor (talk) 06:34, 7 July 2017 (UTC)

Unreviewed articles by keyword

I made another list, this time grouping the unreviewed articles by keyword. It's a big (ca. 1MB so be careful if you are on a mobile) and ugly list, more of a proof of concept than anything else: User:Rentier/NPP/Unreviewed_articles_by_keyword. Enjoy Rentier (talk) 15:46, 1 July 2017 (UTC)

Excellent work Rentier! I think a list like that could be very useful to page patrollers so that they can focus their efforts where it would do the most good. Is this something that we could have bot generate a few times a day and perhaps make it a subpage of WP:NPP?- MrX 16:00, 1 July 2017 (UTC)
Yes, definitely. I requested access to the Tool Labs — it seems to be the optimal environment in which to automate the script. Rentier (talk) 16:44, 1 July 2017 (UTC)

I created a web interface for browsing the backlog - you can filter by any keyword or username and sort the results: http://176.58.102.28/NPP/public/ If it proves useful, I will move it to the Tools Server later. MrX, do you think that an in-wiki list is still worth doing? I can add an "export to wikitable" option quite easily, but the web app is more convenient, at least for me. Rentier (talk) 00:11, 4 July 2017 (UTC)

I commented as requested. As to the keyword list, great idea, but there are some errors. For example, you'd expect Fencing to appear under Construction, but instead it's under Sport. EEng 03:17, 4 July 2017 (UTC) That was a joke, in case that's not apparent.
  • Rentier, I do think having a list on enwiki is a good idea, if it doesn't require excessive effort. There will likely be reviewers who are uncomfortable going to an unknown website to access such a list. I also agree with Kudpung that filtering should be built into the page feed interface. I envisage a drop down box with a large list of categories and/or possibly a text search functionality. Something like the "New filters for edit review" beta gadget might be useful also.- MrX 12:41, 4 July 2017 (UTC)
  • I added an ability to browse the full list by keyword to the web tool, along with an ability to export to wiki markup. I will configure it to update automatically twice a day. Perhaps someone else can make a bot to update the list on enwiki, if there is such need. Simply fetch the content from this url. I'm open to any ideas with regards to reorganising the category-keyword scheme, layout of the wiki-list etc. It would be nice to have the filtering built-in into the page feed, however I don't expect the tool I made to be directly useful in this regard, except perhaps as a justification of the need to have such feature. Rentier (talk) 15:31, 4 July 2017 (UTC)

Rentier, this is so useful, thank you. It loads very slowly on my computer, so I've kept another tab opened and copied and pasted the titles into the searchbox that way. If this could be incorporated into Speacial:NewPagesFeed, and a newsletter go out about it, it might encourage some editors to come back. I've certainly been reviewing more, as I've been able to focus on disambiguation, which I know well, and searching for rivers, where I know the notability criteria well and so I'm able to patrol efficiently. Boleyn (talk) 09:05, 6 July 2017 (UTC)

Boleyn, thanks for the feedback. That's exactly the kind of improvement in the efficiency of reviewing that I was looking for. I agree that incorporating the search into the NewPagesFeed would be best as it would enable the widest participation. In the meanwhile, you can use the web tool I created for a better and faster experience (and you can search by any keyword), though I understand if you are not comfortable going to an external site. Rentier (talk) 16:31, 6 July 2017 (UTC)

Rentier, I've been using the web tool and it's made my reviewing much easier and I've stayed more motivated and more efficient. Thank you so much for taking the time and trouble to create this. Boleyn (talk) 20:08, 7 July 2017 (UTC)

One more essay of personal opinion

This one is first from me though: User:Usernamekiran/Strategies for NPR.
I also created a template to invite worthy editors to become new page reviewers. Worthy has been defined in the essay. —usernamekiran(talk) 12:26, 5 July 2017 (UTC)

Well done, Usernamekiran. Boleyn (talk) 20:11, 7 July 2017 (UTC)
@Boleyn: lol, thanks for the compliment, and reviewing the page too. —usernamekiran(talk) 20:29, 7 July 2017 (UTC)

Steve Roach album entries

Hi all, I've been looking at some of the top-backlogged creators, one of whom has made a large number of album entries for a prolific musician named Steve Roach (musician). The album entries I've checked so far all have zero or one secondary source (and if they have one, it's AllMusic). I reached out to the creator of the those in the backlog to see if they could be improved, but the editor concedes there are no further sources and so the entries don't meet NALBUMS; they made the entries because of a personal view of the worthiness of the artist. As a compromise I was going to suggest this editor make a "Discography of Steve Roach" page if they'd like to preserve track listings...but upon closer examination, I'm not sure the Steve Roach page itself (and thus a related discography) would survive AfD: right now it's very thinly sourced for GNG, and so far I haven't found a single album that has as many as two reviews for NARTIST, although I have not searched exhaustively. It wouldn't make any kind of sense to go through the work of creating a discography page and then redirecting dozens of entries to it if this is just not a topic we have the sources to cover in an encyclopedic manner, and we need to delete the whole. Might someone have a chance to look over the Steve Roach page with an eye toward this issue? Would be grateful for input on how to proceed. Thanks so much. Innisfree987 (talk) 18:24, 2 July 2017 (UTC)

Innisfree987, in my opinion the sources absolutely do not contribute to notability. Many creators believe that a plethora of fleeting mentions, listings, interviews, primary sources, and even other Wikipedia pages, are enough that we won't check them out. The subject has certainly produced an impressive number of works, but as far as Wikipedia is concerned , I think per WP:MUSICBIO it's a reasonable candidate for PROD (would probably be dismissed), or AfD. Kudpung กุดผึ้ง (talk) 01:44, 3 July 2017 (UTC)
Thank you for taking the time to look it over Kudpung; I share your impressions. I think I'll go ahead and send to AfD; my guess is that that may be unpopular with the fair number of editors who've worked on the page but at least it'll get the notability question settled (well, we can hope), and then we can deal with the many album entries that are indisputably not independently notable. Innisfree987 (talk) 04:58, 3 July 2017 (UTC)

I've been looking at these too, Innisfree987, and have been redirecting to artist. New editors often think albums are inherently worthy of inclusion, unfortunately! Thanks for your hard work on this, Boleyn (talk) 20:07, 7 July 2017 (UTC)

Ah, many thanks Boleyn. It looks like the AfD will likely close as keep, so indeed redirecting seems like the way to go. Much appreciated--there are so many! Hopefully the editor will understand that entries are meant to be based not on the existence of the album, but on secondary source coverage of it or other external recognition like charting... Innisfree987 (talk) 01:21, 8 July 2017 (UTC)

Duplicate Articles

Occasionally, in the queue of new articles, I will see two articles with almost the same title, and the same or almost the same content. The most recent example is Little Caesar (singer) and Little Ceaser (singer). In these cases, the articles are either the same, or one is slightly longer than the other. My assumption is that in these cases the editor doesn't know how redirects work, and wants to make their information accessible under an alternate title, and so has created a duplicate article. In these cases I have converted one of the two articles, typically the one with the variant spelling, into a redirect. (It is commonly, and correctly, said that redirects are cheap. I would add that storage is cheap, and the problem with duplicate articles is not that they use storage, but that the two articles will get out of sync.) Has anyone else seen this? Do other reviewers agree that if the reviewer notices that there are duplicate articles, one of them should be a redirect (probably because the creator doesn't know about redirects). Robert McClenon (talk) 00:58, 7 July 2017 (UTC)

If you have duplicate articles, then all but the oldest/best/use-your-best-judgment should be redirects. That's true whether it's one user who doesn't know how to use Wikipedia, someone who does a copy/paste page move (assuming the end target is a valid title), or someone who doesn't realize an article already exists (usually with some form of disambiguation). Nothing wrong with it, and as you say, redirects are cheap! Primefac (talk) 01:52, 7 July 2017 (UTC)
Redirects may be cheap but they lend a legitimacy to the typo. Many wiki-clones will happily pick up articles and present them with their typo title and correct content. Redirects from typos aren't necessary, a search on both Wikipedia and Google will happily point you to the right destination. What the redirects do succeed in doing however is propagating the typo and adding Wikpedia's imprimatur to it.
This has been a bugbear for me since Sparrows Herne Turnpike Road / Sparrows Hearne turnpike in 2009. Hearne was a far less common typo before that redirect was created. Cabayi (talk) 12:18, 8 July 2017 (UTC)
I seem to recall seeing another super-old redirect being nominated for deletion and kept. The actual policies regarding typos etc in redirects is probably more of WP:RFD's domain, but my point was more that we don't need two duplicate articles, and whichever is better should be the actual title with the actual content (I'm not really fussed if the second page is redirected, A10'd, or R3'd). Primefac (talk) 13:15, 8 July 2017 (UTC)

Sometimes, on New Page Patrol, I encounter an article that simply looks and feels like it was published in a journal or in a trade publication. It just doesn't look like it was written for Wikipedia, and looks like it is a polished piece of work, but was polished for a purpose other than Wikipedia. Normally I try to search for phrases in it using Google. Sometimes this works, and that is G12, but usually not. Earwig's copyvio detector is another source. If I don't find the original, my usual action is to tag it with {{copy-paste}}. Is there another reasonable action? Robert McClenon (talk) 00:29, 13 July 2017 (UTC)

In one case today, I tagged a page with copy-paste, and it was removed by the author with a comment that it was copied by the copyright author into a new page with a different title. Now that, to me, looks like an acknowledgment precisely that it is copyvio, because Wikipedia can't accept copyrighted material even from the copyright author unless there is a proper release of copyleft. Is that correct? What do I do now, other than re-apply the tag? Robert McClenon (talk) 00:29, 13 July 2017 (UTC)

My normal course is run through Earwig first, then Google partial sentences and phrases that are unique. That will normally turn up something. I've never encountered someone admitting a copyvio without naming a source like you describe. Despite what people think, the authors permission to include in Wikipedia alone is not enough. We need proof they irrevocably have licensed it under a compatible license, which means allowing commercial use. What you described seems like a clear case of a copyright violation since we wouldn't have proof of a compatible license. If you had a suspected source, it would make sense to send it to a copyright investigation. Not sure what is ideal since we don't know the source. Diannaa might have thoughts. TonyBallioni (talk) 01:15, 13 July 2017 (UTC)
Huh. If I've got the right page (now at AfD?), that's really an odd one. As an admin, can you Is there an admin who might be able to look and see whether the draft page the user claims to have pasted it from ever existed? If not then clearly something fishy's going on. That said, (Nevermind, found it.) I do wonder where the line is for material we don't know to have been published elsewhere, with an author we'd have to verify for proper licensing. It's plausible to me this is a repurposing of a rejected journal article, in which case, is that any different from me drafting an entry in a Word doc and then pasting it over? Isn't my pasting it in effectively the compatible license? Absent indication the WP editor and the material's author are not the same, we do tend to take it on faith that the person adding the material is the person who wrote it/owns it... Innisfree987 (talk) 01:45, 13 July 2017 (UTC)
Robert McClenon, you're not an admin?? How've you gotten away with that this long?! Innisfree987 (talk) 01:52, 13 July 2017 (UTC)
User:Innisfree987 - That is it. If the author didn't publish it in some other medium, then there isn't a copyright issue. The deletion debate may turn on the matter of style, that it isn't in an encyclopedic tone. Maybe I should use that argument more often. Robert McClenon (talk) 02:22, 13 July 2017 (UTC)
Not an admin. In past years, there were editors at RFA who had a strong preference that new admins had to be "content creators", because some editors thought that content creators were being mistreated by admins. I'm ready to be an admin now if someone wants to nominate me. Robert McClenon (talk) 02:22, 13 July 2017 (UTC)

Reminder

Reminder: We are doing a push on July 15. If you are experienced in patrolling, we want you. Thanks! RileyBugz会話投稿記録 17:28, 13 July 2017 (UTC)

To editor RileyBugz: I was sort of inactive in past few days, probably will be for few more days (sleep issues); so I don't know what a push is. Would you please elaborate? Thanks a lot. —usernamekiran(talk) 18:50, 13 July 2017 (UTC)
It is basically where we try and do as many quality reviews as we can do in a healthy manner. RileyBugz会話投稿記録 18:57, 13 July 2017 (UTC)
  • I wasn't sure where this ended off at based on the previous conversations. If we are in fact making it a formal thing, I do think we should send a newsletter out alerting people. I also have DGG's concerns that this type of thing will push through hasty taggings, etc, but a newsletter to qualified reviewers with the right could help minimize that risk. There are other things that should probably be included (user scripts that weren't in the tutorial, decrease in backlog, etc.). Pinging Kudpung for general thoughts here. I have no problem working on the newsletter and sending it out tonight if need be. TonyBallioni (talk) 00:42, 14 July 2017 (UTC)
I don't really have an opinion - I'm supposed to be retired from micromanaging NPP since February although the elected successor(s) have not taken up their 'duties'. From AfD experience I find that such drives are a net negative and that's why I never organised one for new articles. They encourage fast, superficial work. The additional problem is that with maintenance areas being a magnet for younger users, it will encourage total newbies, and other inexperienced users to patrol new pages and the result is that it just makes more work for qualified reviewers and admins. This is unfortunately exactly what the community faction led by one admin wanted, based on the premise: '...everyone can edit' . What they failed to mention (obviously) was that there are plenty of areas where precisely not every one can edit and IMO, NPP should be one of them. Personally, I'm more concerned with devoting my time to ensuring ACTRIAL is properly prepared and gets off to a timely start. Kudpung กุดผึ้ง (talk) 00:57, 14 July 2017 (UTC)

Is there an IRC room or Slack channel or anything that will be going on Saturday? Power~enwiki (talk) 03:55, 14 July 2017 (UTC)

Check out the very lonely #wikimedia-npp connect channel :) — MusikAnimal talk 04:00, 14 July 2017 (UTC)
And IMO, it's best staying lonely. Kudpung กุดผึ้ง (talk) 09:45, 14 July 2017 (UTC)
There's not enough volume to keep it populated normally, but if there are new people with questions I'll be there most of the day today. Power~enwiki (talk) 20:00, 15 July 2017 (UTC)
Any word on how the big push went? Chris Troutman (talk) 02:47, 17 July 2017 (UTC)
Chris troutman, it looks like the backlog is down ~1,000. I wasn't involved really: Saturday was for my content project then the Jimbo-drama hit. RileyBugz might have more thoughts. I also haven't looked at the quality of the reviews. TonyBallioni (talk) 02:53, 17 July 2017 (UTC)
Chris, and Tony In last 24 hours I reviewed something like 20-24 articles. Which is less than my usual average. But as of 03:08, 17 July 2017 (UTC) the backlog is 17,581. This is the biggest drop I have ever seen over the period of 2-3 days. So I would say the push went well. Not sure about the quality though. —usernamekiran(talk) 03:08, 17 July 2017 (UTC)

Do you think a mass message should be sent to all flag holders thanking them, and boasting about stats Tony?
We can organise another similar event in mid or end of August. This thank you note, and boasting will serve as motivation for everybody. —usernamekiran(talk) 04:13, 17 July 2017 (UTC)

I'm personally of the less is more philosophy when it comes to mass messaging. I also tried to avoid wording it as anything "official" in the newsletter: just a commitment that some people were making. I (or anyone else) can put it as a blurb in the next newsletter under the backlog section. If there is a consensus to update people now, another mass message sender or admin could do it. TonyBallioni (talk) 04:32, 17 July 2017 (UTC)
I think the push went pretty well overall. I didn't check other people's reviews, but what I saw in the backlog was stuff that would take a bit to clean up, which I think shows that 1. we mainly focused on numbers 2. the quality was pretty good. I do think that we did pretty well, though, and I think that we probably did more bad-article-cleanup than we usually do, at least based on my experience. RileyBugz会話投稿記録 17:18, 17 July 2017 (UTC)
Hi Tony. I understand your point. Do you think in the next newsletter/message to NPR folks, links to certain pages should be included? You know, the pages where the unreviewed articles can be seen in categories. I think this will increase activity of many flagholders.
RileyBugz put my thoughts in words. —usernamekiran(talk) 20:36, 17 July 2017 (UTC)
Also, something that I want is some kind of line graph that shows the backlog numbers over time. Maybe put it on the NPP dashboard? RileyBugz会話投稿記録 20:47, 17 July 2017 (UTC)
User:MusikAnimal (WMF)/NPP analysis is supposed to be self-updating (90 days back, though). I'm not sure if it currently is, but MusikAnimal might be able to shed some light on it and tweak to include a total backlog graph in addition to the current ones (I don't think its a super-huge need now if it is a lot of work or he is too busy.)
I didn't include the lists from Rentier because I was under the impression that they were a prototype and had not updated. If they are actively updated, I think it would make sense to let people know about them in a future newsletter. TonyBallioni (talk) 21:19, 17 July 2017 (UTC)
yes. Actually less the interval time, the better. And real time update would be the best. —usernamekiran(talk) 21:34, 17 July 2017 (UTC)
The external tool http://176.58.102.28/NPP/public/ updates itself every six hours. The on-wiki lists are not self-updating at the moment, but I tend to update them semi-manually every few days. Rentier (talk) 21:36, 17 July 2017 (UTC)
I proposed and asked for feedback about a bot that would update those charts at User:MusikAnimal (WMF)/NPP analysis, but no one said anything! If you are interested I can make this happen :) However it's important to note that script records data going by the time of article creation. From my research >95% (estimate) of articles that don't survive are deleted within 90 days, perhaps even less, so indeed I only want to report data 90 days from the present, if that makes sense. Let me know what you think — MusikAnimal talk 22:41, 17 July 2017 (UTC)

MusikAnimal, for some reason I thought I had read it as it was already updating. I think it would be helpful. I have a few other questions, but I'll ask it on the talk page to avoid cluttering this space. TonyBallioni (talk) 00:08, 18 July 2017 (UTC)

I think these graphs are excellent and may even cover many of the datasets that we require to be made to monitor ACTRIAL.Kudpung กุดผึ้ง (talk) 00:17, 18 July 2017 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:New_pages_patrol/Reviewers/Archive_11&oldid=1078449092"