Wikipedia talk:New pages patrol/Coordination/Archive 6

Archive 1 Archive 4 Archive 5 Archive 6 Archive 7

Talk pages

We have a lot of talk pages:

I think this is contrary to the way most WikiProjects do it, which is to redirect all talk pages to the main one unless it really needs to be separate. AfC has just two, for example: WT:AFC and WT:AFCP (which is their version of WP:PERM). Previously there was a consensus to do the same here and redirect everything to WP:NPR, but it seems to have been chipped away over the years. Still, none of the above pages are especially active except WP:NPR, and expecting new reviewers to watchlist 7 different pages to be fully involved in the project seems rather exclusionary to me.

I'd like to propose slimming this down to just two pages and the following redirects:

The reason I think the main page should be WT:NPP instead of WT:NPR is that that's where most people expect it to be, and time has proven that there isn't a need for a dedicated page to discuss the tutorial. There have been three threads on WT:NPP in the last year and they were all misplaced (and by the same user). I don't actually like the implied distinction between 'coordinators' and regular ol' reviewers that maintaining two pages entails, but I can see that merging them all into one would probably be a bit much. – Joe (talk) 07:10, 20 June 2023 (UTC)

Good thinking. I'd be onboard with this, with a couple of tweaks.
If you feel strongly about WT:NPP being the main talk page, I propose we discuss that more/separately, and if there's consensus, do it as a second step, so as not to hold up these other good changes. –Novem Linguae (talk) 11:48, 20 June 2023 (UTC)
I think it would be nice as well if WP:NPPN could redirect to the newsletter. Zippybonzo | Talk (he|him) 19:50, 20 June 2023 (UTC)
I boldly changed the redirect target of WP:NPPN just now to point to the newsletter page. Hope that helps. –Novem Linguae (talk) 00:13, 23 June 2023 (UTC)
In general, i agree with this. If no one objects in 24-hours, I'll move the content of the talk page (and archives) of the awards and PCSI pages, directly to this talk page's latest archive, since most of the discussion on those pages is not presently relevant. -MPGuy2824 (talk) 03:28, 21 June 2023 (UTC)
I'm pausing on this until NL's question below, about the best way to do this gets a good answer. -MPGuy2824 (talk) 04:03, 22 June 2023 (UTC)
I'm happy to start a separate RM on WT:NPRWT:NPP after the merges. I know most of us are used to that by now, but if you look through the WT:NPP archives it does quite regularly trip others up. – Joe (talk) 14:42, 21 June 2023 (UTC)
I went ahead and redirected WT:NPP to WT:NPPR just now. I also archived everything on the WT:NPP page so that it wouldn't get overwritten, then I merged the two page's archive boxes manually. See this diff for an example.
I'd be fine with redirecting the NPP subpage talk pages to this coordinator's talk page. But let's make sure we have a good plan for archiving everything. I'm not sure cutting and pasting it into the coordinator's archive is the most organized way to do it, but creating merged archive boxes is also a lot of work. Thoughts? –Novem Linguae (talk) 23:20, 21 June 2023 (UTC)

For talk pages, IMO for a project the fewer the better. But it could be a lot simpler to handle the deprecated ones and retain the archives. Simply note that at the page and advise persons to go to the discussion page elsewhere.North8000 (talk) 11:45, 23 June 2023 (UTC) North8000 (talk) 11:45, 23 June 2023 (UTC)

I agree, it's not worth spending too much time on archives. We could just make Wikipedia_talk:New_pages_patrol/Archives with a list of the various old talk pages, and put a link to that somewhere on WT:NPPR. – Joe (talk) 15:46, 23 June 2023 (UTC)

So the idea is to redirect every NPP-related talk page to this talk page, right? I started redirecting a couple more NPP talk pages to here, and I started trying to add their archives to the archive box above. I ended up stopping because I realized how many talk pages would potentially be redirected, and how many archives we'd have to track. Seems to me like the following talk pages could potentially be redirected to here, which is a lot:

  • Template talk:NPP dashboard ‎
  • Template talk:NPP backlog
  • Template talk:NPP redirect backlog
  • Wikipedia talk:New pages patrol/Header ‎
  • Wikipedia talk:New pages patrol/School
  • Wikipedia talk:New pages patrol/Awards
  • Wikipedia talk:New pages patrol/Newsletter
  • Wikipedia talk:New pages patrol/Newsletter/Draft
  • Wikipedia talk:Page Curation/Suggested improvements
  • Wikipedia:New pages patrol/Backlog drives
  • Wikipedia:New pages patrol/Backlog drives/November 2021
  • Wikipedia:New pages patrol/Backlog drives/July 2022
  • Wikipedia:New pages patrol/Backlog drives/October 2022

I'm feeling unsure. Pausing to gather my thoughts and hear other thoughts. cc some other coordinators since their subpages could be affected: @MPGuy2824, Buidhe, Zippybonzo, Atsme, Dr vulpes, and Illusion Flame:. Maybe we should only direct some or no pages. Maybe some discussion will help clarify if this is a good idea. –Novem Linguae (talk) 21:44, 23 June 2023 (UTC)

why not just keep the archives with the respective talk page. That way they don’t get lost. I don’t see any reason to gang them all together in one humongous archive.Atsme 💬 📧 23:11, 23 June 2023 (UTC)
I'm not sure about the Template talk pages. They would be very specific to that template and wouldn't be interesting to a wider group of coordinators. Similarly, the individual backlog drives' talk pages are very specific to that particular backlog drive. Wikipedia:New pages patrol/Backlog drives itself can be merged and redirected. For Awards and PSCI, we could first archive all the current talk page sections, and then redirect the talk pages here. Instead of merging the archives, I like Joe's idea above, of just making a page with a list of links to older archives. -MPGuy2824 (talk) 03:46, 24 June 2023 (UTC)
The specific backlog drives shouldn’t be redirected as they are too specific, and are better as they are. The newsletter is probably better to have its own talk page, as is awards because they are pretty specific, I’m unsure about the archives because it’s a lot of effort to bung them together into 1 set of archive pages when it is probably not the best solution anyway. Zippybonzo | Talk (he|him) 08:53, 24 June 2023 (UTC)
Both the newsletter and the award talk pages were created about 9 months ago. Before then we got along just fine with one talk page... – Joe (talk) 17:59, 24 June 2023 (UTC)
I don't think this has to be a big deal. Bear in mind that we are just restoring the status quo that other projects use, and which NPP did until recently. As others have said, it's easier to just leave the archives. People following links to old threads will find them easier that way, which is the main thing. And I know you probably didn't mean anything by saying "their" talk pages, but one good reason to centralise discussion is to hopefully broaden participation and avoid giving the impression that some coordination activities (the newsletter, the awards, etc.) belong to a select group. – Joe (talk) 09:37, 24 June 2023 (UTC)

I feel like this discussion is currently "no consensus". I don't feel a plan that enough people agree with is crystallizing. Let me know if anyone disagrees. Further discussion is also welcome. –Novem Linguae (talk) 16:36, 24 June 2023 (UTC)

...literally everyone has agreed? The only minor point of contention I can see is whether to redirect the newsletter and awards talk pages. – Joe (talk) 17:56, 24 June 2023 (UTC)
If someone can read consensus in this (to me) confusing conversation, and feels they can implement this without making coordinators upset and without breaking a bunch of archives, go for it. –Novem Linguae (talk) 22:53, 24 June 2023 (UTC)

  All done except the templates (per MPGuy2824) and specific backlog drives (per Zippybonzo). Thanks all. – Joe (talk) 07:33, 26 June 2023 (UTC)

Template:NPP dashboard

The new Template:NPP dashboard has been boldly swapped in to the header. I went ahead and redirected its talk page to here.

Low/medium/high/color coding

Thanks for taking a stab at this. Any chance we could get a color-coded low/medium/high field based on the total # of unreviewed articles/redirects? We can copy the colors and cutoff numbers from the old templates. This "at a glance" information is very useful for seeing if we're safe or if we need to panic/increase output/hold backlog drives. –Novem Linguae (talk) 21:09, 23 June 2023 (UTC)

This was actually one of my main motivations for developing the template. I don't think the total size of the queue is actually tells us whether we're safe or not. For example, the backlog now is a little over 4000—"moderate" on {{NPP backlog}}'s scale—but growing so fast that it will be "very high" in about six weeks unless we do something. Conversely, last October it was 8000 ("very high"), but down from nearly 12,000 in September, so obviously much less worrying. {{NPP dashboard}} tries to highlight the derivative of the backlog, which is much more informative in terms of what we should now. I think over-focusing on the total number of unreviewed articles is one reason why we've historically had very large cyclical backlogs. – Joe (talk) 07:09, 24 June 2023 (UTC)
I think the backlogs going up and down correlates most strongly (like 100%) to bus factors such as Onel5969 and John B123 quitting and resuming. How we present the info in this template doesn't seem like a factor to me, but who knows. –Novem Linguae (talk) 16:16, 24 June 2023 (UTC)
That certainly has a lot to do with it. But the interesting thing is that before Onel5969 (and I really don't mean to lessen or take for granted Onel5969's current contribution here) there were other superstar reviewers doing the lion's share of the work – these reviewers have come and gone, but the cyclical pattern has been constant. Probably a discussion for another day, though. – Joe (talk) 17:33, 24 June 2023 (UTC)
The NPP dashboard template is still lacking color coding based on the total unreviewed articles. Changing the font color of the numbers based on their size could be a nice subtle way to do this. –Novem Linguae (talk) 16:24, 24 June 2023 (UTC)
Sorry if I wasn't clear. I don't think we should do this, for the reasons I explained above. – Joe (talk) 17:09, 24 June 2023 (UTC)
@Novem Linguae: I've added colour-coded warnings about the total size of the backlog (using the same scale as {{NPP backlog}}, starting at 'moderate'), for redirects and articles. Is that an okay compromise? – Joe (talk) 07:44, 26 June 2023 (UTC)
I'd prefer to change the text color of the total number, since that won't add height to the banner. Up to you if you want to implement though. Looks like I'm in the minority on this dashboard stuff. –Novem Linguae (talk) 09:12, 26 June 2023 (UTC)
To give some context, I view Wikipedia and most websites at 150% browser zoom. So something like this NPP dashboard with 3 warnings fills up about half my screen. I'm not old yet, but my eyes are not getting any younger. Screenshot.Novem Linguae (talk) 09:19, 26 June 2023 (UTC)
It's obviously really hard to design something that works equally well for all people. Apart from the zoom level, the impact of height depends on which skin your using, your system fonts, the (physical) size of your screen, and so on. While developing this I've been looking at it on a 2560x1440 monitor, 1920x1200 laptop, and a 1080x2220 phone, and at 100% and 110% zoom (because I also find the default size font size on Wikipedia to be on the small size). Here's what I'm looking at now, for example. That said, I think designing primarily for the default settings (i.e. Vector 2022, 100% zoom) is a reasonable choice. And to be fair, but there is always the option to collapse it, and the combined height of the NPP header is still significantly less than say WT:AFC, WP:AN, or just an average article talk page. – Joe (talk) 09:49, 26 June 2023 (UTC)
Oh yes, I forgot to mention that we can't just use colour to convey information about the backlog severity, because that would screw over people with colour blindness or visual impairments. Some amount of additional text is unavoidable. – Joe (talk) 10:06, 26 June 2023 (UTC)
The border in your screenshot looks nice, but is missing on some pages such as Wikipedia talk:New pages patrol/Coordination. Might be worth investigating and fixing. –Novem Linguae (talk) 10:07, 26 June 2023 (UTC)
Yeah it's because of Template talk:Start tab#Disabling frame functionality on talk pages. Pretty annoying. – Joe (talk) 10:12, 26 June 2023 (UTC)

top and bottom padding/margin

The two boxes are very tall. Would you be willing to compact them as much as possible in the up/down dimension? Seems like they are about 3 times as tall as the old ones, which is a lot of screen real estate. –Novem Linguae (talk) 21:11, 23 June 2023 (UTC)

I've made some changes. I think it's now comparable to the old header, if you take into account the fact that I reduced the height of the tabs before making the change. In case you missed it, it's also collapsible. I'm reluctant to change the margins or padding because these were chosen to match those used in the default Vector 2022 skin, and reducing them is going to make it look awkwardly cramped in comparison to the rest of the page. – Joe (talk) 07:25, 24 June 2023 (UTC)
Thanks, looks better. Still double the height of the old one and there's room to shrink it more, but good progress. –Novem Linguae (talk) 16:27, 24 June 2023 (UTC)
To be precise, the combined height of the header before my changes was 215px, after (with the changes) it's 268px. Without the "growing backlog" warning (which disappears when it's not growing), it's 238px. Quite a bit less than standard article talk page headers, in any case. – Joe (talk) 17:21, 24 June 2023 (UTC)

table of contents is broken

See example at the top of this page. Can this be fixed please? –Novem Linguae (talk) 21:14, 23 June 2023 (UTC)

I'm afraid I'm not seeing this. Which skin? Can you provide a screenshot? – Joe (talk) 07:25, 24 June 2023 (UTC)
  Fixed. Looks like you accidentally fixed it with your recent edits. Before there was an unclosed HTML tag or something that was causing the border and background of the table of contents to merge with the NPP dashboard on pages that had a table of contents. –Novem Linguae (talk) 16:22, 24 June 2023 (UTC)

Survey: Put NPP dashboard template in header?

Which templates shall we place in Wikipedia:New pages patrol/Header, which is loaded at the top of most NPP project pages?

Current (Old):


Very high unreviewed pages backlog: 12218 articles, as of 18:00, 6 May 2024 (UTC), according to DatBot

Very high unreviewed redirects backlog: 24947 redirects, as of 18:00, 6 May 2024 (UTC), according to DatBot

Proposed (New):

NPP backlog
Articles
12218 ↓524
Oldest article
2 years old
Redirects
24947
Oldest redirect
4 months old
Article reviews
3266
Redirect reviews
3913
  •   There is a very large articles backlog
  •   There is a very large redirects backlog

Data provided by DatBot and MusikBot. Last updated 2 minutes ago.


  • I prefer the old ones, but am fine if consensus shifts the other way. –Novem Linguae (talk) 16:33, 24 June 2023 (UTC)
  • I prefer the new one as it turns 2 boxes into 1 and provides more information. Zippybonzo | Talk (he|him) 17:15, 24 June 2023 (UTC)
  • Dont have a lot of time to chat, currently using my neighbors internet while I wait for mine to return. I’m leaning towards the new as it has some more stats that I find helpful and condenses the size down. I also like the old ones, so I’d be fine with keeping those too. - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 17:15, 25 June 2023 (UTC)
  • I don't mind the new one. It does take up more height on my screen than before. I've made some changes to the proposed layout which can be compared here: Template:NPP dashboard/testcases. See if anyone likes them better. Some points:
    • The oldest article stat is not correct. It won't be correct until phab:T157048 and any related bugs are fixed.
    • I'd guess that reviewers are used to the warning colors right now. Maybe we can add the colors for now to the article backlog number, and remove them after a couple of months?
    • For consistency, we should add the redirect backlog increase/decrease as well. Unless, this was a conscious decision given the discussion in #Encouraging NPPs to focus on articles instead of redirects in the newsletter. -MPGuy2824 (talk) 03:56, 26 June 2023 (UTC)
      The redirects trend isn't there because that data isn't currently available to templates or modules. See User_talk:DatGuy#DatBot:_Page_Triage_Updater for a discussion of how we could get it. – Joe (talk) 04:13, 26 June 2023 (UTC)
      @MPGuy2824: And thanks for the tip about phab:T157048. But are you sure that's what's happening here? I had thought the same, but the oldest redirect-turned-article in the queue right now is from 2001. "Two years ago" seems to correspond roughly to the first pages created in another namespace (i.e. draftspace), then later moved to mainspace. I don't know if that's a bug exactly, or if it's tracked on phabricator. – Joe (talk) 08:09, 26 June 2023 (UTC)
      There's also T38930. I think there was one more which i'm not able to locate right now. -MPGuy2824 (talk) 08:15, 26 June 2023 (UTC)
      I was thinking that the text “NPP Backlog V T E” should be removed as it’s quite obvious what’s being shown, and it takes up unnecessary room on the page itself. Thoughts? - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 02:50, 27 June 2023 (UTC)
    I like Template:NPP dashboard/sandbox. If others like it too maybe we can switch to that one. –Novem Linguae (talk) 06:19, 26 June 2023 (UTC)
    That's only very slightly shorter (152px) than what we have now (176px), at the cost of more crowding/visual variety in the stats box and the loss of the collapse functionality. I'm not convinced that's a trade-off worth making... – Joe (talk) 07:36, 26 June 2023 (UTC)

What is a sustainable rate of reviewing?

Zippybonzo's comment hereif all 717 reviewers and 3 admins patrolled around 14 pages/redirects, poof, backlog is gone—reminded me of an old problem. Statements like that used to be a mainstay of the newsletter and backlog drives, and of course are supposed to motivate people by showing that if we all pitch in it's a not of lot of work. In reality, we know that we don't actually have 717 reviewers (or ~1600 if you include admins) – the majority of them aren't active. So with the amount of reviewers we actually have, how much reviewing would we all need to do to stay on top of the backlog? I'm sure we've worked this out before, but can't remember where now, and the calculation is probably due an update.

According to quarry: new page patrols by user we have about 250 active reviewers, taking a minimal definition of 'active' to be more than a dozen or so reviews a year. And according to quarry: daily summary of the new pages feed, an average of 539 non-autopatrolled articles and 647 non-autopatrolled redirects were created per day over the last month. So based on some back-of-the-envelope calculations, each active reviewer needs to review about 15 articles and 18 redirects a week to stop the backlog growing. Over the last year, only about 23 reviewers have achieved that rate for articles, and only about 18 for redirects. The mean reviewing rate for articles is 8.5 and for redirects 13. One reason the redirect backlog is in a better place is that a higher proportion of redirects are autopatrolled: 43% compared to just 25% for articles.

The upshot I suppose is that these are not terrible, but also not super encouraging, numbers. To manage the articles backlog sustainably—without relying on super-reviewers doing half the work—the project has four options:

  • Double the average rate of reviewing – encourage people to go for "two a day" (two articles and two redirects) instead of "one a day"
  • Double the pool of active reviewers
  • Double to triple the number of autopatrolled articles – this would half the number of articles to be manually reviewed
  • Half the number of articles created – probably not feasible unless we come up with another WP:ACPERM

– Joe (talk) 10:10, 28 June 2023 (UTC)

If we gave out redirect autopatrol more leniently, that would start to lower the redirect backlog. Zippybonzo | Talk (he|him) 10:22, 28 June 2023 (UTC)
That's not the answer. The threshold to receive the pseudoright is already low (rough criteria is basically just 100 problem free redirect creations) and it's rarely denied unless the user has created problematic redirects. Based on the old requests archive, there have been 84 approved requests in 2023 vs 82 requests between December 2019 and the end of 2022. Need more people to nominate users whose redirects don't need to be patrolled instead of just adding more people to reduce the backlog. Hey man im josh (talk) 14:23, 28 June 2023 (UTC)
I'm going to write up a bit more for this later, but I wanted to suggest that you perhaps modify the quarry query since MB is included in the "active" list. Their last edit was January 8th. Hey man im josh (talk) 14:15, 28 June 2023 (UTC)
It's just a rough estimate. There are also people who no longer have the right, and sadly at least one that has since died. In practice, the pool of active reviewers, however you define 'active', is constantly changing as some people leave and others join. – Joe (talk) 14:47, 28 June 2023 (UTC)
"What is a sustainable rate of reviewing?" depends on the person. For a while 50-100 articles a day was sustainable for me. Now that would no longer be sustainable. I think NPP needs to increase its bus factor but, controversially, I'm not sure the "many hands" theory is going to do it. So let me add two other ideas:
  • Prioritize which backlogs we can live with and which we cannot - I continue to think equating redirect and article backlogs is a mistake in a world of limited resources.
  • Double the productivity of the top 15 human reviewers (would have been over 5000 more article reviews in the last 30 days or more than what we lost from Onel)
  • Double the number of reviewers who are capable of doing more than 100 reviews in a 30 day period
I think the NPP tent should be large for a number of reasons. But in terms of keeping the queue balanced that doesn't mean it's the right approach. Also, importantly, I think we only hit "backlog" territory when there are lingering articles that have gone past the 90 day mark (in other words when things get added to Google without prior review). So by that definition while the queue might be at 5000 the backlog is only 10% of that. Best, Barkeep49 (talk) 14:57, 28 June 2023 (UTC)
I guess I skipped a bit of preamble there. By sustainable I mean that articles are reviewed at roughly the same rate as they are created – no cyclical backlogs. We had that for about six months after the queue hit zero last October, but since Onel stopped it's been going up again. I tend to use "backlog" and "queue" interchangeably but, even though I generally agree that we should not panic every time we see a big number, it's currently rising so rapidly that it's just a matter of time until we are looking at a large genuine backlog as you define it.
I agree that trying to grow the number of active reviewers isn't the best strategy, because it's probably not going to work. That said, I don't think NPP's historic strategy—wait until Onel comes back or someone else decides to start reviewing hundreds of articles a day—sounds great either. A combination of my options 1 and 3 and your option 3 seems to the most plausible, and the most fair, way forward to me. Also 100% agreed about the redirects; I worry that using backlog zero as a reason to pivot to redirects, instead of trying to keep articles at zero, will prove to be a major missed opportunity. – Joe (talk) 15:46, 28 June 2023 (UTC)
I view the math as sort of a pyramid. The super reviewers are the top level, the other highly active reviewers the next level down etc.. The lower layers are also where tomorrow's highly active reviewers will come from. Another big dichotomy is the folks that have developed the knowledge and experience enough that they have the confidence and capability to do reviews rather quickly. They are the only ones shorter term who can make a significant dent in any big backlog. Me personally, I consider myself competent to do article reviews fluently but stepped back when the backlog went to near-zero (sort of like "I'm not needed here") and haven't restarted. I don't consider myself yet competent to do redirect reviews fluently. And one more dichotomy at the newer person level is the "hump" to get over to know enough to be confident to do basic reviews, even if takes a long time on each one, and to start doing some. Finally, one thing that we should probably should focus on is the ability to respond to major changes such as a super-reviewer stepping back or a major burst in new articles/redirects. At the higher levels, this might mean trying to develop and keep active the "second from the top" active reviewers involved and actively reviewing which would mean emphasizing that tier handling more of the workload currently handled by super-reviewers. North8000 (talk) 14:50, 28 June 2023 (UTC)
That's a really good point. It's a little thing, but I think the loss of "the graph" has really hurt there. As someone perpetually hovering between your third tier and complete inactivity, seeing the graph spike is usually what jolted me into reviewing again. I hope we can develop alternative tools that can 'activate' reviewers in response to changes. – Joe (talk) 15:44, 28 June 2023 (UTC)
Thanks for posting this, Joe. I like your pragmatism here. In the short-term:
  • I think an MMS about the rising backlog could perhaps get folks like you and North8000 who only review when they feel they're needed to come back and help. Perhaps we should do this ASAP.
  • I think a backlog drive once the article backlog gets closer to 10,000 will be important to get the backlog down to a manageable level, although there is a risk of burnout, so double edged sword.
  • I think finding someone who wants to step up as recruitment coordinator could help a lot. With WP:PERM/NPP criteria higher than ever (and I notice you're trying to raise the NPP criteria more, probably not the best timing for this), we need to keep the new blood coming in. And the recruitment coordinator would be recruiting experienced users who are unlikely to be a problem.
In a true emergency (article backlog >20,000):
  • The folks who used to do NPP recruitment and designed the recruitment process have given the process a lot of steps. Very thorough background checks are performed on folks before they even receive a user talk message, and before they even step foot at WP:PERM/NPP. This is a bit laborious. In an emergency, we should gut these requirements and just design a quarry query that targets users who have been highly active this month and have >10,000 edits, and MMS them in batches, and let WP:PERM/NPP do the background checks.
  • In an emergency, we could start skipping articles by experienced editors so that we can focus on newer articles / articles of less trusted users. There'd be a ton of ways to implement this (PageTriage, bot, just tell reviewers, hand out a bunch of autopatrol). This would be an option of last resort since it would let un-screened articles through by un-screened editors.
Novem Linguae (talk) 22:20, 28 June 2023 (UTC)
Please see: Wikipedia:New pages patrol/Newsletter/Help draft for all the people that aren’t in the other thread. - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 23:34, 28 June 2023 (UTC)
Regarding your last point, we can grant autopatrolled for a time-limited period now. I keep thinking we should make more use of that. – Joe (talk) 04:13, 29 June 2023 (UTC)
Regarding recruitment, I agree that that list is... strict. Just scanning for usernames I know, the rejects include a 17-year editor with 85 FAs, one of the project's most active translators, and the leading contributor to WikiProject Editor Retention. More to the point, the 'vetting' already gets done at PERM. I don't think we should wait until an emergency to get rid of that duplication of effort. Of course you shouldn't invite obviously unqualified people, but since most people are going to say thanks but no thanks anyway, you might as well cast the net wide. Since the requirements at PERM are de facto experience-based, I think it would be better to focus on identifying batches of editors with relevant experience—people active at AfD recently, people with X number of recent DYKs/ITNs, and so on—and send out invitations en masse, without further vetting. (And by the way I don't see my question at WT:PERM as "trying to raise" the standard, more about aligning the written criteria with the reality, and it looks like the outcome will be to slightly widen them, so win-win). – Joe (talk) 04:51, 29 June 2023 (UTC)
it looks like the outcome will be to slightly widen them. I see at least 3 people in that discussion who object to any increase at all in the NPP criteria, so I think I'd like to see an RFC if that idea is to be pursued further. –Novem Linguae (talk) 06:12, 29 June 2023 (UTC)
Don't panic Novem. This isn't, in fact, my first rodeo when it comes to reading consensus or developing policies. By widen I mean "loosen", if that's not clear. – Joe (talk) 06:49, 29 June 2023 (UTC)

Due to my own fault, the main point of my post was not clear. Per the math (which says that our main unmet need is to be able to respond to variations) a main objective should be to develop a substantial second-from-the-top-tier reviewers who remain at least semi-actively reviewing. This would require emphasizing them taking over some of the work that the super-reviewers are doing. North8000 (talk) 13:07, 29 June 2023 (UTC)

Ideas to increase recruitment

Auto accept AFC non-probationary reviewers?

Just a thought, but perhaps we could consider auto accepting users that express interest who are considered "Active reviewers" at AfC? Primefac does a good job over there and those that are considered active reviewers I view as competent individuals who have a good understanding of what is and isn't notable. Hey man im josh (talk) 13:16, 29 June 2023 (UTC)
I think the bar for AFC is lower, because AFC reviews also get a second check by an NPP. So while I think AFC experience is fantastic when applying for NPP, I am not sure they'd be a perfect group to auto accept. Hope that makes sense. –Novem Linguae (talk) 14:11, 29 June 2023 (UTC)
I do believe the bar is lower, but I also think that those who have passed the probationary stage (just below the active reviewers) are more than likely qualified based on their experience in assessing notability. If we didn't want to grant the perm on a permanent basis I do still think a blanket "yeah, we'll give a trial to anybody on that list" would be appropriate. Hey man im josh (talk) 14:49, 29 June 2023 (UTC)
When I was actively doing the perms that's effectively what I did, along with a spot check of their AfC just in case there were any norm differences between AfC and NPP that felt worth noting. Best, Barkeep49 (talk) 14:54, 29 June 2023 (UTC)
Ah, I missed the part about non-probationary reviewers only. Yeah, I think full members of AFC would be a pretty safe category of folks. In fact, I think AFC probation usually lasts longer than 6 months. –Novem Linguae (talk) 19:56, 29 June 2023 (UTC)
This is an interesting idea, especially in light of the recent alignment of NPP and AfC rights in the other direction. I'd like to hear from Primefac first (are all these editors likely qualified for NPP? Would we be 'poaching' volunteer resources from AfC?), but in principle I wouldn't be opposed to batch assigning NPP to them. – Joe (talk) 21:26, 29 June 2023 (UTC)
Probably? I will be honest, I haven't really had much to do with NPR for a few years now; AFC membership isn't quite as easy as PCR, but "has a clue" is my primary metric there; it's mostly edit count and AFD stats that determine whether someone gets added as a proby or not. Primefac (talk) 09:02, 30 June 2023 (UTC)
I like the idea, especially if we're more lenient on handing out trials and not the permanent right to these users. The one thing that feels weird is how interlinked NPP and AFC would be after this, though – you apply for one, and you're basically already part of the other if you want to be. I think we'd need to clear up the status on patrolling your own AFC reviews at least before going through with this. Skarmory (talk • contribs) 06:02, 30 June 2023 (UTC)

Loosen the WP:PERM/NPP criteria too?

IMO a low bar MMS and for entering a trial period is a good idea. And implicitly "trial period" means needing to meet some additional criteria to achieve perm status which is a good idea. But the criteria for perm status should also be low.....doing some reviews, and not making big errors. Because, even at the end of the trial period they are still newbies and we need to keep people meeting that minimal criteria aboard so that they can progress further. North8000 (talk) 16:00, 29 June 2023 (UTC)

Which criteria would you loosen? – Joe (talk) 21:28, 29 June 2023 (UTC)
Well, overall I'm sort of recommending tightening because I'm saying everyone go through a trial period. On the MMS criteria, I just recommended loosening but don't have the expertise to nail it down. On getting it initially / trial period I'd recommend the gatekeeper discretion except give them the benefit of the doubt. To get perm I'd say having done 5 reviews in each of 2 successive months with no big problems noted. Sincerely, North8000 (talk) 22:19, 29 June 2023 (UTC)

Special MMS message to alert NPPs to the growing backlog?

This has been discussed in a couple places, but I'd like to give it its own section. We are drafting a special MMS at Wikipedia:New pages patrol/Newsletter/Help draft. Feel free to edit and improve the draft. I feel the draft could use more text, an image, and a title or signature saying it's from the NPP team. Feel free to borrow formatting/text from here or here

  • Is there consensus to send this out?
  • If so, when? We JUST sent the newsletter last week I think. Does it make sense to space these out? Or to wait until the article backlog gets higher?
  • Should redirects be mentioned at all, or just focus on the article backlog?

Thanks. –Novem Linguae (talk) 06:27, 29 June 2023 (UTC)

Given that we are still in "moderate" territory (according to our dashboard), i think folks might get a bit irritated. Let's at least wait until the backlog reaches 8K, which is our "very high/large" territory. There is also an admin newsletter that is going to be sent out soon. I think Kudpung added a line to one of those previously. We should consider doing this too. -MPGuy2824 (talk) 06:35, 29 June 2023 (UTC)
It's the rate of growth that's the problem. We'll be at 8K in three weeks. I think we need to get out of the habit of reacting to backlogs only when they hit an alarming headline figure – it just makes the problem worse in the long run. – Joe (talk) 06:48, 29 June 2023 (UTC)
Let's wait till 6K ("high"), at least. The urgency in the message would be commensurate with how the dashboard treats the backlog number. Also, then it would more than 14 days since the newsletter.
On redirects, I'd say don't mention them. -MPGuy2824 (talk) 07:06, 29 June 2023 (UTC)
I agree with Joe Roe in that it's the rate of increase, not the actual number that matters. We need to be more proactive, and consider that if the rate of increase is likely to lead to what is widely considered a backlog, then we should be actively reducing it. It should start with asking all reviewers/admins, including inactive ones to try to review 1 or more pages a day. We've got the people who know what to do, it's just getting them to review the pages. Zippybonzo | Talk (he|him) 12:09, 29 June 2023 (UTC)
Nice job with Wikipedia:New pages patrol/Newsletter/Help draft @Zippybonzo and @Illusion Flame. I think it looks really good. I think that graph is an awesome image for it too, as it visually conveys what is happening with the backlog and why there is some urgency. I am leaning towards sending this out immediately, because I think that sending it will get some reviewers who took their foot off the gas after the redirect backlog drive to come back (two reviewers talked about this on a talk page... that they don't currently think their help is needed but if they were told that it is, they would come back), and the sooner they restart their efforts, the better the backlog will be in the long run. If I'm reading this discussion right, multiple editors favor sending immediately, so let's go ahead and execute that. Friendly reminder, don't forget to add a signature to the draft. Thank you everyone for your thoughts in this discussion. –Novem Linguae (talk) 21:23, 30 June 2023 (UTC)
Also let's look at the graph again in two weeks and see how much this helped. If this has a major effect on the graph, we should make a mental note and send these out more often. –Novem Linguae (talk) 21:25, 30 June 2023 (UTC)
So are we approved to send? - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 21:26, 30 June 2023 (UTC)
Yes please :) –Novem Linguae (talk) 23:23, 30 June 2023 (UTC)
Okay, thanks. We will send it when Zippy comes online. - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 23:28, 30 June 2023 (UTC)
Please do some adjustments so that the image is fully within the purple div. -MPGuy2824 (talk) 02:26, 1 July 2023 (UTC)
  Fixed. I added a {{Clear}} template. –Novem Linguae (talk) 02:39, 1 July 2023 (UTC)
Sending after I fulfil some other MMS requests. Zippybonzo | Talk (he|him) 06:50, 1 July 2023 (UTC)
Sent, let's hope I don't have to fix the message. Zippybonzo | Talk (he|him) 06:58, 1 July 2023 (UTC)
An issue I've just noticed is that it's sent to a few people who can't help, as they aren't reviewers, so we might get some new reviewers soon. :) Zippybonzo | Talk (he|him) 07:00, 1 July 2023 (UTC)
But it sent correctly. Zippybonzo | Talk (he|him) 07:01, 1 July 2023 (UTC)
Yeap. You called it :) –Novem Linguae (talk) 07:22, 1 July 2023 (UTC)
That’s great! - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 12:14, 1 July 2023 (UTC)
Sadly, it appears the mass message only temporarily slowed the backlog. For a detailed graph, see here. - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 20:00, 14 July 2023 (UTC)

Move all reviewers to MPGuy2824's draftify script

At this point MPGuy2824's fork of the draftify user script is a clear upgrade compared to Evad37's original. In particular, the improved multiple-choice message templates, warnings about too new or too old pages, and addition of a #moveToDraft tag make it significantly more policy-compliant than the original, which I think is reason enough to fully deprecate Evad37's in favour of MPGuy2824's (as previously suggested).

My question is, what's the best way to achieve this? Has anyone asked Evad if he's okay with passing on the torch? Could we redirect the old script to the new? Or mass message people asking them to switch? Courtesy pings @MPGuy2824 and Evad37: – Joe (talk) 17:32, 16 June 2023 (UTC)

I don't think anyone has asked Evad directly, but he must have noticed me sniffing around his script's talk page, and you now have pinged him to this thread.
There are a couple of options to deprecate Evad's script:
  1. Add a small message to the UI of the script: "This script is no longer being maintained. Please switch to the current version: Edit your common.js file by changing User:Evad37/MoveToDraft.js to User:MPGuy2824/MoveToDraft.js".
  2. Replace Evad's UI completely with a message asking people to switch. (screenshot attached)
     
    2. Replacing the UI completely
  3. Redirecting from the script to my fork.
All of them require either Evad (or an int-admin) to make changes. I think we should definitely do #1 first. I'm torn between #2 and #3. #2 is more jarring to the user, but #3 might seem like we are bamboozling them without their consent. @Evad37, thoughts? -MPGuy2824 (talk) 02:19, 17 June 2023 (UTC)
3 seems too much, 2 seems a bit pushy, but 1 seems most reasonable. Zippybonzo | Talk (he|him) 12:42, 17 June 2023 (UTC)
We need to proceed here with respect and sensitivity. Someone should definitely reach out to Evad37 via their user talk page and see what their thoughts are, then we should discuss further. Let's make sure they are involved in this discussion. –Novem Linguae (talk) 18:46, 17 June 2023 (UTC)
I left a user talk message for Evad37 just now. I want to make sure they are included in this discussion. –Novem Linguae (talk) 11:50, 20 June 2023 (UTC)
Unless Evad37 responds, we should wait a week before posting an int-admin edit request for option 1 (adding a small UI message to his script). -MPGuy2824 (talk) 07:42, 24 June 2023 (UTC)
I did this edit request, and the change has been made to Evad's script. I'll report back with how many people switched over after 15 days. We can decide on any further action after that. -MPGuy2824 (talk) 05:13, 30 June 2023 (UTC)
Changes after 15 days (source)
Script MPGuy Evad
Date Active Total Active Total
1 July 119 144 327 717
15 July 137 164 316 705

Active just means active on wikipedia, not necessarily actively using the script. I recommend that we wait for 15 more days before doing anything else major. -MPGuy2824 (talk) 04:00, 15 July 2023 (UTC)

There are only a few holdouts that are actively using Evad's version. If they don't switch by the end of the month, I'll send them individual messages asking them to. -MPGuy2824 (talk) 04:15, 15 July 2023 (UTC)

Coordinator request

Hello there! I saw the outreach for possible NPP coordinators, and needless to say, I'm interested. As we already have good people for backlog, awards and newsletter, I'm thinking I can help in recruitment by sending invitations to fellow editors for joining NPP and AP. Furthermore, given my interest in counter-vandalism acts, I can audit editors who may be abusing their rights of both NPP and AP. I sure do hope I will contribute efficiently and will benefit from our coordination. Best. ❯❯❯ Raydann(Talk)

Pinging Novem Linguae for faster response =D ❯❯❯ Raydann(Talk) 04:53, 29 June 2023 (UTC)

Seen. Will check with some folks and get back to you shortly. Thank you very much for your interest. –Novem Linguae (talk) 06:09, 29 June 2023 (UTC)
As a thought, we could also possibly have multiple recruitment coordinators. It'd be beneficial to have multiple users looking for qualified members to add to the team.
I also appreciate that Raydann mentioned inviting users to apply for the autopatrolled right, as that in of itself is a helpful way to reduce the backlog now and in the future. Hey man im josh (talk) 13:27, 29 June 2023 (UTC)
Hey @Raydann. Thanks for your enthusiasm. Would you be willing to help out with NPP recruitment on an informal basis for awhile, and then if everything goes smooth we'd add you as an official coordinator a little bit later? Also please follow the section right below this closely, as we may be loosening the criteria we use to determine who we reach out to about joining NPP. –Novem Linguae (talk) 14:15, 29 June 2023 (UTC)
Wow, I actually thought of doing so even before requesting here. So yes, I'll do the outreach informally for now. And thanks for the heads up, I'll keep myself updated with the below discussion. ❯❯❯ Raydann(Talk) 17:47, 29 June 2023 (UTC)
Hey @Novem Linguae, do you reckon I can add my name to the Coordinators list now? Helping in Invitations and outreach (probably auditing too). ❯❯❯ Raydann(Talk) 20:48, 13 July 2023 (UTC)
@Raydann. Just a suggestion, not entirely related to adding yourself to the list, but would you be able to help out with preparing the recruitment MassMessage lists and such, and then I’ll send them upon request from you/NL. Zippybonzo | Talk (he|him) 01:52, 14 July 2023 (UTC)
Added. Thanks for helping out with recruitment. Nice job. (Un?)fortunately I think we've automated recruitment with the MMS project, so please keep an eye out for other jobs you may be interested in, or feel free to help Zippy with tasks. Sounds like he has one for you above if you're interested. –Novem Linguae (talk) 08:17, 14 July 2023 (UTC)
@Zippybonzo sure I can make a MMS list for people who I have decided to invite, but it would only delay the process. Usually I look for the user's contribution in other areas, then I screen their talk page to see any negative issues, then I see their responses to generate civility and then finally if satisfied, I leave a message on their talk page. If I continue this process but instead of directly inviting them, add them in a list, it will ultimately increase the workload for me, you and NL. Instant invite would be best for manual screening, and Mass message would be best for list generated through a query.

@Novem Linguae, thank you for adding me. Even with the MMS system, I would continue to look for prospective editors, quite possibly, active AfC reviewers who do not possess the NPP right. Also, not to mention I screen editors for the Autopatrolled right, and we currently do not have an automated process for that. ❯❯❯ Raydann(Talk) 14:05, 15 July 2023 (UTC)
@Raydann, if you are able to help with automating autopatrol (screening and setting up small batches, of around 10-20 to allow tweaking) and then I will be able to send upon request, though if you end up requesting sending from me regularly, I'm sure @Novem Linguae can probably set you up with the MMS right if needed. Zippybonzo | Talk (he|him) 15:05, 15 July 2023 (UTC)
@Raydann. Sounds good on inviting AFC reviewers for NPP and others for autopatrol. Thanks. Keep up the good work! –Novem Linguae (talk) 06:11, 16 July 2023 (UTC)
Raydann, thanks for picking up autopatrol recruitment as well. Could you also add "checking for autopatrol abuse" in your list of tasks? Removing a user right is a sensitive area, so please bring your first couple of suspected abuses here just for a second opinion. -MPGuy2824 (talk) 06:18, 16 July 2023 (UTC)
Oh, I just had an idea as well. Can we get a WP:QUERY detecting folks who have made more than X edits to AFD subpages in the last 30 days. Then we can manually cross-check those results to https://afdstats.toolforge.org/ and invite folks with an accuracy rating higher than 70%. –Novem Linguae (talk) 06:35, 16 July 2023 (UTC)
@Zippybonzo that's a good idea and the way I would prefer it. I got the autopatrol invite template deleted (on Joe's suggestion), but will have a talk with @Joe Roe to see what could be done.

@MPGuy2824 sure, I will try to find problematic new pages created by Autopatrolled user's and see if any one of them is abusing the right. Thank you.

@Novem Linguae I believe the candidates added by a query to a list should go through further manual screening. While using a query is a very efficient way to find probable editors, we should not invite them solely based on that. But a great idea nevertheless. ❯❯❯ Raydann(Talk) 20:14, 17 July 2023 (UTC)

Concerns about autopatrol auditing

@MPGuy2824 and Raydann: I would very strongly recommend against any newer editor/non-admin actively patrolling for misuse of rights – for your own sake. I've done a fair bit of it and, honestly, even if you're as careful and tactful as you can be, it's a surefire way to get dragged to the dramaboards every other week. Unless you have a very thick skin and are absolutely sure you don't want to do an RfA in the future, it's best left to others. – Joe (talk) 09:14, 19 July 2023 (UTC)

Seems like good advice. Maybe Raydann can just come back here with such users, and you/NL/other willing admins can take the issue forward, depending on the severity of the issue. -MPGuy2824 (talk) 09:29, 19 July 2023 (UTC)
@Joe Roe true enough. It is a difficult and time consuming task to audit the misuse of rights. While patrolling, I did find some borderline editors, but I dare not accuse them for misusing rights, as I would not be assuming good faith. I reckon I will currently not perform this particular task, and in the meantime, I can do other useful things. Thank you very much for the suggestion. ❯❯❯ Raydann(Talk) 09:47, 19 July 2023 (UTC)
I think that’s for the best @Raydann. If you ever find misuse, maybe you could contact an admin about it privately. This would solve both issues - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 19:22, 19 July 2023 (UTC)
If anyone ends up doing this role in the future, I envision them bringing it to the other NPP coordinator's attention on Discord or this talk page, and then there would be discussion, and others would handle contacting the person in question and other measures if needed. They would not have to be the one delivering the message. Hope that makes sense. –Novem Linguae (talk) 20:36, 19 July 2023 (UTC)

Edit for Wikipedia:New pages patrol/School

(Courtesy ping Novem Linguae) Just a note that I've boldly edited the school page to remove this link to Wikipedia:Online Ambassadors, which states that [t]his page is currently inactive and is retained for historical reference, which has been the case since 2014. Please discuss if anyone disagrees, thanks. VickKiang (talk) 02:14, 19 July 2023 (UTC)

All good. Thanks for your edit. –Novem Linguae (talk) 06:39, 19 July 2023 (UTC)
Perhaps this should be asked at the general NPP talk page? There's 370 page watchers there vs only 55 here. Hey man im josh (talk) 12:48, 19 July 2023 (UTC)
I thought as the school talk page redirects to here this place is fine, but you could also link it to the main page. VickKiang (talk) 22:06, 19 July 2023 (UTC)
Edit seems uncontroversial and not needing further scrutiny, imo. –Novem Linguae (talk) 00:21, 20 July 2023 (UTC)

Coordinator page alterations

I was wondering if the following table could be added to the coordination page instead of the current one:

Table
User Coordinated areas
Lead Coordinator
Novem Linguae Technical issues, WMF liaison, Discord
Backlog Drive Coordinators
Buidhe Backlog drives
Zippybonzo Backlog drives, newsletter coordinator
Illusion Flame Backlog drive assistant, newsletter assistant
General coordinators
MPGuy2824 PCSI clerking, technical issues
Dr Vulpes Awards
Atsme NPP school, NPP liaison with the board of trustees.
Novem Linguae Technical issues, WMF liaison, Discord
Newsletter coordinators
Zippybonzo Newsletter coordinator/sender
Novem Linguae Newsletter reviewer
Illusion Flame Newsletter assistant
Former Lead Coordinators
Kudpung Ca. 2010-2018
Insertcleverphrasehere Ca. 2017-2018
Barkeep49 Ca. 2018-2019
MB Ca. 2022-2023

It separates each person into the respective 'teams' (and makes us look bigger). Additionally I think it might be worth exploring turning NPP into a WikiProject, being that it's pretty much a WikiProject in its current state. Zippybonzo | Talk (he|him) 12:10, 18 July 2023 (UTC)

Thanks for suggesting this table. After thinking about it, I prefer the simplicity of the two bulleted lists for now.
We are already a WikiProject, in my opinion. Did you have some specific changes in mind? –Novem Linguae (talk) 13:27, 18 July 2023 (UTC)
I think we are technically a Wikiproject, but definitely not a normal one, given that there’s a user right associated with it. - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 02:12, 19 July 2023 (UTC)
The changes I have in mind are to allow non-reviewersto sign up as helping with weeding out bad articles and allowing us to patrol the good ones quicker. Zippybonzo | Talk (he|him) 12:19, 20 July 2023 (UTC)
I think that’s a great idea! - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 13:04, 20 July 2023 (UTC)
I think I’d prefer a blend of this idea and out current list. I like that our current list avoids repeating names, but I like the idea of separating us into teams. - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 02:10, 19 July 2023 (UTC)
I don't like that individuals and their roles are listed in an inconsistent fashion. For instance, you're listed as "Newsletter coordinator/sender" under "Newsletter coordinators" but you're listed as "backlog drives, newsletter coordinator" under "Backlog Drive Coordinators". The consistency of capitaliization for "coordinator" also needs to be fixed, as it's capitalized in three of five section headers and not under the description portion I guess you'd call it?
Why would we want the team to look bigger than it is? NPP is also already a WikiProject. Hey man im josh (talk) 13:17, 19 July 2023 (UTC)
I would prefer more of a consistent look, with a more merged approach to the table so we don't look too big, but I prefer the 'teams' concept. Zippybonzo | Talk (he|him) 12:16, 20 July 2023 (UTC)

Participate in discussion

Please see Wikipedia talk:New pages patrol/Backlog drives/October 2023. Your input is requested in said discussion. Thank you. - 🔥𝑰𝒍𝒍𝒖𝒔𝒊𝒐𝒏 𝑭𝒍𝒂𝒎𝒆 (𝒕𝒂𝒍𝒌)🔥 01:21, 21 July 2023 (UTC)