User:Joe Roe/Seven tips for new page patrolling

This is in an opioniated guide to new page patrolling with a focus on reviewing new articles efficiently and sustainably. It doesn't attempt to reiterate existing policies and guidelines or to cover the process of reviewing articles in-depth. Instead, I offer seven general tips that have improved my patrolling and that I think might help you too:

1. Specialise in reviewing certain types of pages
2. Be pragmatic and prioritise the most important tasks, bearing in mind that some articles need more scrutiny than others
3. Work in batches of similar articles
4. Don't worry about notability, NPP is not the notability police
5. Don't draftify, it just puts off a final decision
6. Don't try to do everything, NPP is just a first pass
7. Do something else from time to time

Why listen to me? I've never been a "superstar" patroller—someone who ploughs through hundreds of articles a week—but I have been doing it relatively consistently since 2016 and my work at NPP was one of the things that convinced my fellow Wikipedians to trust me with the admin tools. I've taken breaks to do other things (see tip #7), but I haven't burnt out. So I hope I have some insight into how to be a decent patroller.

1: Specialise

If you want to keep engaged in new page patrol for the long term, pick a set of pages you enjoy patrolling and stick to those. Sitting in front of the new page firehose and trying to handle every single article that comes at you is a surefire way to burn out – so if you only pay attention to one of these tips, let it be this one.

What you specialise in of course depends on your own skills and preferences. Some ideas:

• Articles on a topic that interests you: you can use the NPP Browser, SDZeroBot's sorted feed or WikiProject article alerts to find these. That way you will actually enjoy reading them, and you can use your knowledge of the topic to spot problems or cleanup opportunities that others might miss.
• Very new articles ("speedy deleting"): articles that are a couple of hours old are the most likely to have serious content problems and be eligible for speedy deletion. The idea here is that you focus on quickly identifying these most obvious targets and leave anything more nuanced to subsequent reviewers. It's like shooting fish in a barrel.
• Very old articles ("working the back of the queue"): almost a review process in itself.
• Probably-bad articles: use the filters available in the feed to look for articles lacking citations or one of the 'issue' flags. Requires a good eye for separating flawed-but-salvageable cases from deletion candidates.
• Probably-fine articles: filter articles with predicted B class or above, or use the automated report. This might feel like 'easy mode', but that's nothing to feel bad about – every article removed from the queue counts!
• Copyvios: detecting copyright violations can be really tricky and is in many ways its own type of patrolling. If you enjoy this kind of detective work, you can filter the new pages feed to only show articles automatically flagged as being potential copyvios.
• Redirects: see Wikipedia:New pages patrol/Redirects

You'll probably want to pick more than one area and periodically switch them up to keep things interesting. For example, I've settled into a pattern of reviewing articles on history and science (because I find them interesting to read), articles that were previously deleted (because I like investigating article histories) and articles by prolific creators (because this complements my work at WP:PERM/A). Nowadays I avoid areas with a high rate of deletion, because I don't find that rewarding, but when I was more active in anti-COI work I used to seek out pages that were G11 candidates. The important thing is that if you don't get anything out of reviewing certain types of pages, don't bother.

2: Be pragmatic

When you review an article, try to develop a sense of what really needs to be done, and what can wait. The primary aim of new page patrol is to identify articles that have serious content problems that need to be addressed right away or are obviously unsuitable for inclusion. Everything else is just a "nice to do" and the sad reality is that, since NPP has been perennially backlogged for as long as it has existed, we don't really have time to do them. Above all, remember that you're not responsible for the articles you review—the creator is—and you're not obligated to improve it to a certain standard. You just have to make sure that it's not irredeemable.

It's also important to recognise that not all articles require the same amount of scrutiny. Articles on some topics (biographies of living people, organisations and companies, etc.) are more likely to be problematic, and/or have problems that pose a pressing threat to the integrity of the encyclopaedia. These articles are where NPPers should be spending most of their time. In contrast, we can afford to give low-risk, low-readership articles on niche topics the benefit of the doubt. If you don't think you have the time to give an article the time it deserves, then skip it. NPP is a team effort and it's more important to keep the momentum going than to try and handle every article in front of you (see tip #3).

3: Work in batches

Batch working is a well-known productivity hack that exploits the fact that humans are good at pattern recognition but bad at task switching. It's what makes a production line more efficient than an artisan's workshop. Applied to NPP, this basically means that most efficient way to review is to identify batches of similar articles and work through them one after the other, rather than for example reviewing pages as they come into the queue.

The page curation tools offer several ways to identify and patrol batches:

• Articles by the same author (use the filter tool and/or database report)
• Articles with a particular issue (use the filter tool)
• Articles on the same topic (use the bot-sorted list)

4: Don't worry about notability

The most effective way to speed up reviewing is not to worry too much about notability. NPP is not the notability police. Its job is to filter out articles that are blatantly unencyclopaedic. Articles on subjects that might not have enough significant coverage behind them but are otherwise inoffensive are not blatantly unencyclopaedic. A detailed source assessment for notability is by far the most time-consuming thing you can do when reviewing an article. It also requires a lot of prior knowledge (of the relevant guidelines, the topic, and where to find sources) and is easy to screw up (because you don't have that knowledge, or access to key sources, or you just missed something). Unless there's a compounding problem, it's usually sufficient to tag articles of questionable notability with {{notability}} and move on.

Obviously, this does not mean that you should ignore articles on blatantly non-notable topics or A7/A9 candidates. How deeply to assess notability should also be approached pragmatically, based on the nature of the topic. If you're reviewing a biography of a businessman and the references look weak, it's probably worth taking a closer look at notability, because doing so will also highlight any WP:BLP or WP:NOTPROMO issues – and these are more important policies than WP:N. But not being sure whether there are enough sources to support a stub on a species of beetle is not a pressing threat to Wikipedia.

5: Don't draftify

See also: User:A2soup/Don't use draftspace and User:Paul_012/Drafts are broken

Repeat after me:

I must not draftify.
Draftification is the mind-killer.
Draftification is the little-death that brings total obliteration.

Moving an article to the draft namespace ('draftifying' it) often appeals to new page patrollers, especially less experienced ones, because it looks like a frictionless way to get difficult articles out of the queue. It is most often used when the reviewer isn't prepared to mark an article as reviewed, but also doesn't want to go through one of the deletion processes. Unlike CSD it can be applied to any article, and unlike deletion through AfD it can be carried out by an individual reviewer without affirmative consensus. It gives the reviewer an easy way out of a difficult decision and the creator a softer outcome than immediate deletion – a win-win, right?

Unfortunately, this is an illusion. Draftifying looks like a decision, but it is just putting off a decision. Consider the possible outcomes after an article is moved to draft:

1. It is (re)submitted to AfC and:
   1. Accepted, entering the new page queue again
   2. Declined, sending us back to either #1 or #2
   3. Rejected, probably leading to #2
2. It is abandoned and deleted six months later under WP:CSD#G13

In the first scenario, the article will be reviewed at least once by an AfC reviewer, and if accepted, once again by an NPP reviewer. So it is only temporarily removed from the NPP backlog and the overall burden on volunteer reviewers generated by this one article has been at least tripled by the draftification. In the second scenario, an article that had potential (or it wouldn't have been draftified) has been lost.

Both scenarios are obviously bad for the project as a whole. Even more perniciously, they're bad for your development as an individual reviewer, because it encourages you to switch your brain off (see the melodramatic Dune misquote above). When patrollers reject an article with vague statements about it being "borderline" or "not ready for mainspace", what they're really saying is that they don't know what to do with it. So they use draftspace to put off deciding. But there's always a decision that could have be made: a topic is either suitable or not, an article is either salvageable or not. There are no real grey areas and the most important skill patrollers need to develop is deciding, reliably and efficiently.

But wait, isn't it much kinder to article creators to give them a chance to work on their article in draftspace? Again, consider the possible scenarios. If the topic of the article is unsuitable, then it will be deleted, sooner or later, so encouraging the creator to work further on it is simply a waste of everyone's time. Few people will thank you for nominating their article for deletion, but if that's where things are headed, most would agree that they'd rather get it over with. If the topic is suitable but the article itself needs work, then all the evidence we have indicates that this is much more likely to happen in mainspace, where other editors will jump in and help. The vast majority of contributors experience their article being moved to draft as a rejection, and they are less likely to continue contributing because of it.

It's entirely possible to patrol without ever using the draftify tool – indeed this is what everybody did for the first decade of NPP's existence. If you're considering draftifying an article then, first and foremost, pause and ask yourself what's stopping you from marking it as reviewed. If you think the issues with the article could be fixed on draftspace, why can't they be fixed in mainspace? Marking an article as reviewed doesn't mean that you're giving it a glowing seal of approval, just that you don't think it needs to be immediately removed (see tip #6). If the article is unsuitable, and it doesn't meet a speedy deletion criterion, then simply send it to AfD and explain why.

6: Don't try to do everything

New page patrol is a first triage of articles. Behind it stand a whole row of more specialised patrols, maintenance queues, and individual editors. You can and should defer any non-essential and/or time-consuming tasks to these processes.

The main way that NPPers can defer an issue to another maintenance process is with cleanup tags. You might feel that adding an article to a maintenance backlog that's already 100,000+ pages is a little futile. But remember: Wikipedia is a megaproject. It wasn't designed to be done in a year, or a decade, or your lifetime – we'll be working on this for generations. Maybe your great-grandchild will be the one that finally gets around to adding references to that article you tagged.

There's also nothing wrong with deferring a decision on an article to another reviewer. If you don't know the topic very well, if you're not yet confident in your understanding of the relevent policy, or if you're just not sure what do (e.g. if you're reaching for the draftify button!), then just skip it and move on. There are X other reviewers waiting in line behind you, one of them is bound to know what to do with it!

7: Do something else

See also: Wikipedia:You are not irreplaceable

Few people can maintain focus on a single task over the long term. Luckily, when you get tired of NPP, there's always something else to do on Wikipedia (or off!). If you start feeling like you have to review patrols (or the backlog will explode, other people will do it wrong, or something else bad will happen), that's a good sign that you need to take a break, whether that's from NPP or Wikipedia in general. The project will tick along without you and will still be here when you get back.

This is also a good idea because NPP is, in many ways, an unusual subset of the project. Most obviously, it focuses on scrutinising content contributed by others, rather than creating it. For reasons that are less clear, it also tends to be more hierarchical and rules-based than the rest of the project. None of these things are bad in themselves, but when new page patrollers don't do anything else they risk getting cut off from the rest of the community, which never bodes well. Keeping a hand in other areas of the project—whether that's article writing, dispute resolution, or other maintenance tasks—helps you maintain a sense of perspective. You also increase the points of contact between the NPP project and the wider community, helping to keep up us in sync on issues of project-wide relevence like notability and expectations of advanced rights holders.