Wikipedia:Wikipedia Signpost/2015-01-14/WikiProject report



This anniversary issue, the WikiProject report is returning to WikiProject Articles for creation for one of our largest interviews ever. Last looked at in 2011, AfC is the method used by unregistered or new users to create articles, and provides an effective filtering system to remove all unsuitable or unsourced submissions to save them needing to be found and deleted later. Four years on from the last interview, we openly invited all active members of Articles for creation to come and have their say in this interview. So, the resulting report below is quite long, but there is a general message coming across: they need your help.

'''One of the perennial problems with Articles for Creation has been the large backlog of unreviewed submissions, currently at around 2,500 ("severe"). How long does the average submission wait before being reviewed, and how many participants actively review submissions to keep on top of the backlog?'''
 * AmaryllisGardener: I don't know exactly, but if I had to guess I'd say a few weeks, it depends on the subject too. BLPs and companies take the longest, I would say, things like animals and plants, stuff like that, don't usually take that long, because the notability of the subject is easy to verify.
 * Aggie80: It depends on the complexity of the article. In too many cases, the submitter hasn't paid any attention to the rules of Wikipedia, those are easy to see and eliminate. A good article that is properly done takes longer to validate. A couple weeks is pretty accurate.
 * Hasteur: It really depends on the effort put into the article. Simple ones that would be quickly CSDed if they had been created in mainspace are easily dealt with. Ones that are on the edge of being acceptable for mainspace take more time as the reviewer's reputation is on the line if it gets put up for WP:AfD.
 * Bellerophon: That really depends on the individual submission. Some are easily determined to be not-notable or otherwise unsuitable, and some reviewers prefer to try and filter these types of submission and get them out the pending reviews queue. Others require a more in depth examination and reviews may have to do a significant amount of clean up work themselves to make sense of a submission, or make it policy compliant. The length of time submissions wait around for review is directly proportional to the number of submissions awaiting review. Some submissions can sit around for months, but on average I would say one to two weeks based on the current backlogs. Reviewers are encouraged to review the oldest submission first.
 * Ritchie333: It depends on the context. A submission that would normally be speedy deleted if in mainspace can be handled within 24 hours. I tend to review from the back of the queue, which tends to be 1–2 months. Sometimes, and indeed with increasing regularity, the time lag is so long, that by the time I review the submission, another editor has created it anyway. Infamy-address-1.gif's Infamy Speech, with many corrections as is required in most AfC submissions]]
 * Libby norman: It really depends on the article – some are obvious fails/passes but more complex or lengthy articles may 'languish' because (in my personal opinion) everyone looks at them and gulps. I've seen submissions that are a month or more old and haven't been touched. Active reviewers – couldn't hazard a guess as this varies – do their best, but it's tempting to skip an article, especially if it's in a specialist field or would require an hour or more to check properly.
 * Onel5969: As the other editors have rightly said, it depends on the article, although I would disagree that it depends on the length. Some very lengthy submissions are well cited, and very easy to pass. Others I have seen have virtually no citations, and are easily declined due to lack of meeting the Wikipedia criteria. I can't speak for other editors, but I know that articles can sit due to a lack of familiarity with the subject matter, or lack of expertise (e.g. articles on technical subjects like physics or medicine). These can take an inordinate amount of time simply familiarising oneself with the subject matter. Personally, I focus on the "very old submissions", and try to approve/decline 10–15 a day. Every few days I'll spend a bit more time on the list, and try to "pick off the low-hanging fruit", by going through a few dozen articles and checking for obvious reasons for declining, like no references, or WP:COPYVIO issues.
 * Sionk: Yes, it varies a lot. I personally preferred to tackle the older, more difficult submissions while others prefer the 'quick wins'. Both strategies are equally valuable, I'm sure! Will be interesting to know whether someone keeps stats.
 * SecretName101: Currently, far longer than it should. It does depend on the article though. Many are addressed in a reasonable time period, while some take longer before they are reviewed. A number appear to fall through the cracks (so to speak) and seemingly are overlooked, remaining unreviewed for an inexcusabley long period time. It has proved difficult for Wikipedia's limited number of volunteers to keep up with the demand for reviews.
 * Theonesean: On the main page of the WikiProject, it says, "Following huge backlogs on the page, Kwsn and others started WikiProject Articles for creation on 2 June 2007.” So, obviously, the backlog is integral to our mission. We have a fairly solid core community made up of experienced Wikipedians whose main or major goal is to review articles. However, it’s been quite some time since anyone was “on top of” the backlog. On the submission box on every article up for review, it says that the reviewing process "may take several weeks, to over a month,” and, indeed, we have a large amount of articles that are older than a month and are still awaiting review.
 * Joe Decker: "Average time" is going to be a misleading statistic, let me phrase it another way. A lot of severely problematic article drafts, perhaps accounting for half the incoming stream of drafts, are handled within a day. Two-thirds of the remainder seem to peter out over a period of a couple weeks. Of the remainder, they often take *months*, and some of those have excellent potential. A good deal of the work is probably done by a dozen or two dozen participants.
 * Dodger67: Drafts on complex or "niche" topics often take the longest as very few active regular reviewers feel competent to review such submissions. Requests for assistance from relevant Wikiprojects do sometimes help but often they end up at "dead" projects. I have found that it's usually the "scientific" projects (such as Medicine, Mathematics, Physics, etc) that respond best to such requests, the Military History and Aviation projects are also responsive to requests for "expert" opinions. We need a more systematic and consistent way to draw subject-specific projects into reviewing drafts and thus avoiding or at least reducing the delays for such submissions. Multilingual reviewers are also in short supply to verify non-English references. A set of "automagical" tools to filter out the bulk of obvious rubbish drafts that constitutes the vast majority of new submissions are desperately needed. They could probably be adapted from the existing vandalism detection bots as there isn't really a difference between detecting "Johnny is a poopypants" type of vandalism in mainspace articles and finding similar content in draft submissions.
 * KvnG: Promising submissions are now taking about 1 month to be reviewed. Obviously flawed or incomplete submissions may get reviewed faster by reviewers who concentrate on such things. We have over 500 reviewers listed but have not recently gone through and determined how many are actually active.

'''How do you think the backlog could be significantly reduced? In the past, backlog drives have been held to push the numbers down – are these successful or do they present too many problems, such as sloppy and rushed reviewing?'''
 * AmaryllisGardener: I think that backlog drives are pretty good, and that we should have more of them. I think the best thing would be to try to get more experienced editors involved in reviewing AfC submissions.
 * Aggie80: Some gates 'up front' to get through before submission. Blank and nonsense submissions could be filtered out by macros or bots. Force anyone with fewer than a few dozen actions to go through the wizard and not allow them to bypass it.
 * Becky Sayles: I have not participated in backlog drives, but this does not seem like a reasonable solution. I agree with Aggie about upfront gates before submission. It would help to have some automatic way to ensure new editors have taken minimal steps towards an acceptable draft. Right now it's too easy to create garbage and submit it for review. Backlog drives by their nature encourage sloppy/rushed reviews.
 * Hasteur: Frankly we need more volunteers to review drafts. I believe that we get around 200 drafts in for review every day. And as referenced below we're on a trend of about 50 more drafts in than out so we're consistently losing ground. Add to this are the CSD:G13 restores where it takes nothing more than a few clicks to get a abandoned draft back and put back into our review queue without any substantial effort to improve beyond the previous decline (or that the draft was never put in for review). As a veteran of multiple backlog drives we suffer a significant problem of gamification where you are rewarded one point for every draft that you review (either as an accept or decline). If you review hordes of submissions you can bury the quality check reviewers (who look at the reviews to determine if the review was correct) and win one of the top 3 prizes as long as your reviews are correct for the most part. There is a significant incentive to do sloppy reviews to get the points and move on during the backlog drive,, Ergo this is why I've opposed doing backlog drives under the current regime as it does not solve the underlying exploting problem.
 * Drmies: More bodies. We could always send Jimbo Wales around to ask to come back—it was the zeal by people like him that made this thing work in the first place.
 * Kudpung: I'm with and  on this one. Reward-driven backlog drives are just asking for sloppy reviewing. What we need, just as we desperately do at WP:NPP, is a greater number of more competent reviewers. I see recently that some users are literally hovering over their edit count to get their name on the AfC reviewer list, and like NPP and many other maintenance places, it's the wrong reason as a newbie, to join Wikipedia.
 * Bellerophon: The current backlog drives are somewhat successful in reducing the backlogs and experienced reviewers do carry out checks on the quality of reviewing, but this is a short-term fix. What's really needed is a significant increase in the number of active reviewers, and we need reviewers with the right kind of experience on Wikipedia. Editors with experience at Articles for Deletion would be good candidates, as a solid grasp of the notability guidelines is pretty essential. The entire community needs to take some responsibility for Articles for Creation, along with the New Pages Feed, it is one of the only methods Wikipedia has for quality assuring new articles and helping new editors.
 * Ritchie333: Like others, I am dead against backlog drives as we used to do them as they encourage low-quality reviews that cause more harm than good (see ). I have proposed the idea of an "AFC Cup" which would reward passed reviews, with increasing points for reviewing older submissions, but it would require a considerable amount of maintenance to run. To be honest, I think we need to think of something completely different, and one solution is to allow more co-operation between NPP and draft space. Instead of NPPers CSDing or AfDing articles, they could be moved to draft space where the creator can work on them. Conversely, if a submission sits unreviewed for say 90 days, it is deemed "no consensus to reject" and automatically moved to mainspace. That sounds radical, but it's no different to a new article that remains unpatrolled for 30 days and drops off the end of the queue (which admittedly is rare). AfC as we know it now would cease to exist, and become an auxiliary function of NPP. I think that's quite possibly the answer.
 * Libby norman: I think it would help if the backlog were more clearly divided up into unreviewed and reviewed and failed, then at least people could target the list more logically and wouldn't be so disheartened every time they looked at the backlog. Currently it feels like the more you do the larger the pile grows and that's because, unfortunately, there are some boomerang articles that are resubmitted repeatedly – either the editor doesn't understand the comments made or doesn't know how to make the required improvements. These repeat fails also clog up the system. In an ideal world, there would be further intervention to assist the editor to make the article work, although I'm pretty certain all regular reviewers have come across the 'hopeless case'. I think backlog drives are quite useful, although they present potential problems if people focus too much on the prize. They are just one tool though and I think more could be done to incentivise reviewers to tackle the boomerangs that might be salvageable but need more targeted intervention.
 * Onel5969: Not in favour of backlog drives, for all of the reasons stated above. Aggie's point is a good one, especially if we could have a bot that looks for COPYVIO or no references. I also like Libby's suggestion of dividing the list into new submissions, versus resubmitted versions of failed submissions. Sometimes it's difficult, since so many folks who submit articles don't put effort into their articles, they don't make an effort to bring it up to even the most rudimentary wiki standards. I think the break-out of re-submitted articles would help significantly. I find that those who submit articles, and honestly want to do it correctly, contact me and try to work to get the article in shape for the mainspace. When they do, I make every effort to work with them (not doing it myself, except for perhaps making an edit as an example), to help them get it ready. The only time this doesn't work is if the subject is simply not notable (like a beloved local HS football coach). But, as the other editors have pointed out, we need more bodies.
 * Sionk: I think reviewers often raise the bar too high, though I can understand why. The purpose of AfC should be to weed out the clear and blatant failures. We should simply ask ourselves "If I saw this article in mainspace would I immediately speedy delete it or nominate for AfD?" If the answer is "No" we should accept it and leave to others to clean-up/scrutinise more deeply etc. Unfortunately I didn't get any satisfaction from letting articles through that were borderline, or probably gaming the system. But to clear things more quickly we should not do such a thorough job (just a minimal one), I'm afraid.
 * Theonesean: Staying on top of the backlog is an extremely relative term, and truly the only thing that could reduce our backlog is more experienced and knowledgable reviewers devoting more of their time to reviewing. Backlog drives seems like a good idea, but they are a keen double-edged sword; many newer reviewers who have great potential end up valuing quantity over quality, and pumping out as many reviews as possible. This, obviously, makes waves that other reviews have to deal with.
 * The WikiProject desperately needs new people to help review, but backlog drives can take eager, inexperienced reviewers and misguide their energies into gaining points rather than experience.
 * Joe Decker: I no longer believe that the problem can be entirely solved by begging for new reviewers. I think that there either needs to be systemic change to better explain what the minimum requirements are for accepting an article before an article is submitted, or there needs to be automatic acceptance after a short period of time. The backlog pushes away new contributors in a way that is harmful to the encyclopaedia.
 * Anne Delong: WP:AFC is a vital part of Wikipedia's mandate to see that "anyone can edit". The Wikipedia community should all pitch in to make sure that the drafts are reviewed promptly. In order for the process to work properly, most drafts should be reviewed within a day, and those needing expert knowledge should be referred to a WikiProject within that time. This isn't happening for a couple of reasons:
 * Most AfC reviewers are serious Wikipedians who value the opinion of their peers. Even though the mandate of AfC is to accept any submission that has a better than even chance to pass an AfD, accepting a lot of submissions which are then deleted looks bad, so reviewers are overly conservative. Some of the reviewers would like to become admins in the future, and editors who are overly liberal at AfC are criticised for it at RfA. This leads to a lot of declined submissions and resubmits which clog up the process.
 * Because of the backlog, a lot of editors complain that AfC is discouraging new editors and slowing the growth of the encyclopaedia. This leads to other editors not wanting to be part of it for fear of being criticised, creating a vicious circle, since lack of participation is what is causing the backlog.


 * Solution: Backlog drives are at best a temporary solution. The obvious, but difficult to achieve solution, is to change the culture of criticism so that more editors will participate in reducing the backlog. A more achievable solution may be a way to get marginal submissions out of AfC without a reviewer approving them. For example, after a submission has been checked for the most severe problems (blank, copyvio, article exists, etc.), then when a draft is declined, the reviewer would have the option of including an extra notice on the draft creator's talk page which would say something like "Please read the notice on your draft to see why it was declined. You can continue to make improvements and resubmit (recommended), or you can move your submission into the encyclopedia at any time by (explanation of move or requested move process here). Be aware that if your new article does not meet Wikipedia's standards for inclusion, it may be deleted, so you may want to keep a copy of your text."
 * KvnG: We currently don't have a core group of reviewers volunteering to run a backlog drive. That may be due to fatigue, problems with previous drives or whatever. I think this is a secondary problem. The primary problem is that, to address the backlog, we need more AfC reviewers. To make the project sustainable, we should be concentrating on recruiting more reviewers. I'm not even sure this can be successful in the context of net exodus of contributors to Wikipedia as a whole. We don't want to burn out editors by running back-to-back backlog drives. We also don't want to burn out editors by asking them to contribute to numerous critical projects on Wikipedia. I believe our best available means of addressing the backlog is to get comfortable with having a backlog. A lot of reviewer time is spent rejecting unacceptable submissions. If authors have to wait 30 days for their submission to be reviewed, they may put more effort into meeting Wikipedia inclusion standards. Of course there is also the risk that promising submissions and the new editors behind them will get discouraged by the wait and will abandon their submissions and perhaps Wikipedia altogether. We try to address the latter by pointing authors to the teahouse as part of the AfC review process. We also have a few reviewers concentrating on evaluating abandoned submissions before they're automatically deleted (6 months after last edit).

'''How many submissions does the project receive in an average day, and has this figure changed since the last interview? Are most submissions accepted or declined? What are some ways an unregistered user can improve an article's chance of being accepted?'''
 * AmaryllisGardener: I don't how many exactly, but I would guess around 150. That's a lot of work for the reviewers like me. Most submissions are declined, the new users don't know how to write content that follows policy, the most common problem in drafts is notability and referencing, as simple as that. Some advice I would give to the users submitting the drafts would be, take things slow, and don't submit it until you're ready, and to get a better understandings of the policies, especially policies on citing sources and notability. Some new users take it the wrong way when you tell them "X isn't notable", thinking it's a personal thing.
 * Aggie80: A couple hundred submissions a day. About 25% are probably re-submissions after something has been reviewed. About half don't come close to meeting the basic requirements and another 25% are being submitted for advertising or self promotion. That leaves about 50 with a legitimate subject and some attempt to meet the requirements. (These are seat of the pants estimates and not based on actual numbers.)
 * Becky Sayles: The numbers vary quite a bit, but according to Category:AfC pending submissions by age The number still waiting by day are: 1d 37, 2d 19, 3d 34, 4d 58, 6d 12, 7d 14, 8d 2, 9d 7, 10d 33, 11d 61, 12d 60, 13d 28, 14d 24, 15d 65, 16d 72. These seem to reflect a drop because of the holidays. But remembering from before December, as Amaryllis said above, 150 seems about right for an average. Most submissions are declined, many each day are resubmitted. Unregistered users can improve their chances by taking advantage of all the resources available for new people and drafting. It's all there for them, too many simply don't want to read instructions. With the number of autobiographies and promotional drafts submitted, it would help to make it clearer requirements for notability and references.
 * Hasteur: I would like to see see some sort of line in the "create a new page" that says something to the effect of "Many drafts that are submitted early in an editor's tenure at Wikipedia are not compliant with policy. Please familiarize yourself with what a good article looks like so you have a better chance at getting your article accepted". I know it's pie in the sky, but it would be nice to remove many of the "Johnny Doe is the best-est bass player in the whole world and has the most kickass garage band in the universe." type of submissions.
 * Bellerophon: The number has fluctuated over the years, I would say on average AfC receives between 100 and 200 new drafts per day; it has been known for it be much higher. When you correlate this with the relatively small number of active reviewers and the amount of additional work most submissions need, it's easy to see why the backlog builds up so fast. The majority of submissions are declined on first review, because AfC is almost exclusively used by new editors who are not really familiar with what Wikipedia is all about. We see a great many submissions that are simply not suitable for an encyclopaedia. However, many submissions are improved after initially being declined and go on to become acceptable articles. We do have a problem with submissions that are repeatedly declined, and we probably need to work on how we communicate problems to new editors and how we deal with those who disrupt the process. This has been a work in progress for several years.
 * 78.26: Most submissions are declined, and it isn't even close. This is because writing an encyclopaedic article isn't an easy thing to accomplish, and AfC of course is a submission portal for new wikipedians. Even articles on historical figures or geographic locations may take several submissions before it is accepted, so that the article is verifiable and readable. A very high percentage of the submissions are designed for product visibility, and even if the subject is notable, weeding out promotional language is usually a prolonged process. To improve the chances of being accepted, if the submitter would take the time to understand WP:Notability and WP:REFB we would have a much better "accept" rate. However, these are not easy policies to understand, and I'm very glad that users whose articles are declined are sent to the teahouse.
 * Onel5969: I'll address the last point, since the others have been very well answered above. While the resources are there for new editors, many don't know about them, nor even where or how to look for them. Some things are easy fixes, and you can point a new editor to the right guide, e.g. WP:BIO, WP:CITE, WP:CIT or MOS:LAYOUT. I find those four help in the vast majority of circumstances. I don't know how to fix the problem, but when a new editor wants to create an article, they should be required to look at and acknowledge that they've read certain guidelines before creating an article.
 * Theonesean: Most submissions are not ready for mainspace, and some take two, three, or four iterations before they are ready. Obviously, most are denied the first time. However, most users can take some simple steps, like making sure the subject of the article is indeed notable, using inline citations to support claims, and making sure that the writing is clear and concise. Reading the Wikipedia guidelines for notability and generally being familiar with how Wikipedia works will work wonders in making your article mainspace-ready.
 * Joe Decker: 200–300, almost all are eventually declined. Most of them will be failed because they lack two or more reliable sources independent of the subject providing in-depth coverage of the subject. Because nobody ever told the contributors of that requirement, nor explained what we mean by those terms. A fair number of the rest are cut and paste copyvios, or spam.

Do many accepted submissions go on to further improvement after being approved, such as appearing on DYK, being expanded, or even become recognised content?
 * AmaryllisGardener: Absolutely. Some submissions start out at C-class. And as you can see here, two articles that were once drafts are now FAs, 28 are GAs, and 765 are B-class. I don't know about DYK though.
 * Aggie80: You bet! People are encouraged to tighten up and add to an article that has been accepted. Often they will shepherd it, coming back to review and monitor it, as well as update as new information is received.
 * Hasteur: One of DYK's suggestions is to look through recently promoted AFC submissions for potential DYK nominations. One of my best acceptances was F. O. Oertel which I also nominated for DYK (which subsequently passed).
 * Ritchie333: If an AfC submission piques my interest, I will do what I can to improve it, sometimes to DYK (eg: The Minories, Colchester). I passed 2013 meat adulteration scandal through AfC and has seen great interest from a wide variety of editors. One AfC submission, (Haim (band)) is sitting at WP:GAN as I write this. Very occasionally an AfC submission gets AfDed, and I will improve the article to ensure its survival (eg: Articles for deletion/Direct Ferries). Experiences like this, though, are very much in the minority.
 * Libby norman: There are many articles I'm pleased I put through, such as North End, Detroit, Scarlet Road, All Saints' Cathedral, Cairo and Design School Kolding. I tend not to ask myself whether this is a potential star, but whether it's an article that adds to Wikipedia/meets criteria to be on Wikipedia. Having also checked out the articles that arrive via other routes, I would say that standards on articles that arrive on mainspace through AfC are quite often higher because reviewing editors take pains to put the basics in place. And provided articles are noteworthy and reasonably well constructed and sourced, they always have the potential to go further.
 * Theonesean: The process of article review is a collaboration between reviewers and authors. If an author works to make their article ready for mainspace, it’s much more likely that they will continue to tweak and expand their article once it’s on Wikipedia at large. So far, we have a long list of featured articles, good articles, and otherwise recognised content.
 * Joe Decker: Rarely, but I'm sure it's happened. Improvement does happen often, though. I certainly try and "level up" anything I accept at least in terms of easy improvements, and tag it for things that can summon appropriate gnomes (e.g., categories), but yes, articles do get better.
 * Anne Delong: AfC graduates should have a better chance to become respectable members of the article community than new articles created directly in mainspace, because they have already had many of their weaknesses fixed up before approval, been selected as not copyvios or spam, and end up on the watchlists of a number of experienced editors, who may continue to improve then or at least prevent COI editors from degrading them.
 * KvnG: Primary AfC acceptance criteria is that accepted submissions are not likely to be deleted. Many reviewers here use a much higher bar than WP:LIKELY so I believe most AfC articles find a permanent place in the encyclopaedia.

'''Can you remember the best submission you've ever reviewed? Was it well-sourced, well-illustrated, skilfully and engagingly written?'''
 * AmaryllisGardener: Flavobacterium psychrophilum was the best draft I ever reviewed. It's well-sourced and well-written overall (for a draft), it was started by an IP editor.
 * Cwmhiraeth: Mites of livestock had been worked on for months and was C class by the time it was accepted. I successfully nominated it for DYK and the creator has continued to work on it since then.
 * Ritchie333: There are AfC submissions I've enjoyed working on (see those above), but as a rule of thumb, anyone who knows how to write a good article out of the box doesn't need AfC. Every submission that's been promoted above and beyond the basic sub or start-class acceptance has required extensive copyediting and finding more or better sources. I'm not sure I have ever assessed a submission at even C-class, unless I've worked on it myself.
 * 78.26: The best submission I reviewed happened to be on a subject matter I am familiar with, and was written by a new editor who obviously understood the subject and already cared about wikipedia's goals. The submitter and I have since struck up a friendship, making this indeed a most rewarding experience. I have assessed a couple of articles at B-class, but I've a feeling these were not by a new editor. I suspect WP:CLEANSTART in these instances.
 * Sionk: My first ever submissions to DYK were ones I came across at AfC. My favourite was the fascinating Great Polish Map of Scotland!
 * Anne Delong: The large number of basic, informative articles about topics for which people are searching contribute at least as much, and probably more to Wikipedia's usefulness and popularity than the smaller number of excellent, scholarly articles. For example, Afrocubism, a small article about a record album, gets a steady 400–500 hits per month, and there are many, many other articles like this. I guess my answer is that I don't remember which of the thousands of articles I've reviewed was the best, and I don't think it's important.

How can a new member help today?
 * AmaryllisGardener: A new member can help today by reviewing the submissions, of course. They should take it slow, and if they're not sure, skip it.
 * Becky Sayles: Aside from reviewing, the best thing a new member could do is to recruit more reviewers.
 * Hasteur: Any editor who has experience with editing and handling multiple mainspace articles is welcome to come assist. We ask that new reviewers follow the reviewing ruberic that has evolved as best practices for handling AFC submissions. We don't throw the entire book of CSD or Mainspace policy at the advocate for the draft, but encourage the advocate to continue improving the draft to the point that it has at least a 50% chance of passing a AfD nomination.
 * Bellerophon: New members are very welcome indeed. Aside from accurately reviewing articles, editors who are great at problem solving or have technical skills, such as programming and coding could help in continuing to improve and maintain the various templates and scripts we use in the reviewing process. AfC also covers Redirects and Categories for creations and Files for upload, so any editors experienced in those areas are particularly welcome.
 * Ritchie333: Review the "very old submissions", particularly ones that you think should pass. Identify cases at CSD and AfD where "move to draft" is an acceptable option. Promote the idea that a submission can be moved from draft to mainspace, and even back to draft.
 * Libby norman: Adopt an article that you think deserves to be on Wikipedia but needs a helping hand and get it through to mainspace. It is a very satisfying feeling.
 * 78.26: New wikipedia members shouldn't be reviewing, but they can make improvements to the pending articles, as no one owns AfC drafts any more than a mainspace article. If you are a new AfC member, try to find articles in an area you have some knowledge and expertise in. It will make your reviewing much easier if you know what is and is not a reliable source for the given field.
 * Theonesean: I’d urge any new member to hang around the talk page of the WikiProject and familiarise themselves with the community before reviewing. Perhaps they could see how other, more experienced reviewers review certain articles, before trying one on their own. Also, there is no shame in asking for help or skipping articles that are too complicated or difficult to review.
 * All in all, the most important thing is to keep working at it. One only becomes a better reviewer by reviewing articles.
 * Joe Decker: Start at the shallow end of the pool, the quick fails are good practice. Over time, notice which explanations help people "get" what is needed (the general notability guideline is very difficult to explain to completely new editors), and consider writing some of your own boilerplate to explain those issues in your own words, you'll find yourself repeating yourself a lot. Keep on assuming good faith.
 * Dodger67: New reviewers (not new editors) can get their feet wet by doing "quick fail" reviews until they become more familiar with the standards and move on to more complex reviews. Don't be afraid of asking for help – some of our most competent and experienced reviewers regularly ask for help and second opinions. Don't bring your ego with you into AFC, it will get bruised quite regularly.
 * Anne Delong: A new AfC member could help out by looking over the submissions and leaving messages about them at appropriate WikiProjects. Some of the submissions need subject experts, and others could benefit from editors who know which sources in a certain area are reliable.
 * KvnG: The most recent submissions are probably the easiest for a new members to handle. Until you gain some experience reviewing, you're encouraged to just skip anything that you're not sure about or seek help from more experienced reviewers on our reviewer help talk page.
 * LaMona: I see a lot of new members trying to start by creating a new article, which is much harder than editing an existing one. If we could pair up new members with folks creating articles then they could learn to edit together in the Draft space. Maybe we could point new editors to the drafts?

Anything else you'd like to add?
 * NE Ent: AFC is yet another example of someone coming up with an idea for other people to do work based on the fantasy there's an infinite pool of ready volunteers. (See also – let's stick a tag on top of this article!) Should be abolished as soon as possible.
 * Ritchie333: One additional problem not covered is reviewing an AfC submission seems to carry a risk of public humiliation if you get it wrong (eg: Articles for deletion/Ahead Learning Systems, Articles for deletion/Clemmie Moodie, Wikipedia talk:Articles for creation/Creative UK – all three I passed through AfC although in each case I felt notability was very borderline). This definitely leads me to procrastinate over reviews and only taking them on if I'm prepare to do a "GA Lite" review and have extreme confidence it will withstand an AfD.
 * Libby norman: AfC is a valuable testing ground for new editors and articles – and both need nurturing. There are some very hardworking reviewers out there that make this happen.
 * Onel5969: I think you have to overcome a fear of "getting it wrong", as Ritchie pointed out. I've had 2 or 3 articles I've accepted which have later been marked for deletion as non-notable. My feeling is that AfC should be a nurturing tool, and that we should approve borderline cases, and then let the community as a whole make a decision it's not notable. I'd much rather that, than declining an article and then have that editor decide that Wikipedia isn't worth it. After rejoining AfC, I spend more time on it, between reviewing submissions, responding to questions from submitting editors, and helping get articles in shape for publication, than I do on any other areas. I think the folks that are involved in AfC need to remember (as most of them do), that this is the portal for new editors coming into the Wikipedia community.
 * 78.26: There are two areas of AfC which are the least fun for me. The most common is trying to review an article with "walls of sources" trying desperately to establish notability, usually by inheritance. It takes a long time to read through these, only to find that the subject isn't mentioned at all in the source. The second (and fortunately not all that common) is to be "taken-to-task" for approving or rejecting an article. Reviewers are often dealing with marginal cases which take judgement calls, and even experienced editors may disagree. I think in general reviewers welcome collaborative discussion about our acceptance procedures, but I don't know anyone who feels they become a better editor/reviewer because they received a "what the hell were you thinking you idiot, obviously another reason AfC should be shut down, nincompoops."
 * Sionk: I've taken to leaving AfC for periods to recharge my batteries (it can be frustrating and tiring). When I return I often find the templates and procedures have changed, sometimes beyond recognition. Some sort of stability would be beneficial, if I have to relearn everything each time I return it may be more trouble than it's worth!
 * Theonesean: Thank you for doing this interview, and if anyone has any questions, please feel free to drop them on my talk page. I’ll be more than willing to help.
 * Joe Decker: I do a lot of work at AfC, and I think there are a number of extremely dedicated, hard-working people doing good work there. That having been said, I wonder sometimes if we would be better off without AfC entirely, whether funnelling all the weight of that effort, and the articles that would end up at AfC, into NPP instead, would be of a greater net benefit. I know this is not a popular view, and I'm actually quite conflicted about it myself, but I do wonder.
 * Josve05a: I would like to add a note (shoutout) about the Wikipedia help channel on IRC, . Many users who has gotten their drafts declined come to that channel for guidance, what can be fixed and so on. During daytime (GMT) we have a few helpers that can help out, but during nightime, between 3 and 7 am (GMT), I see quite a few helpees coming search for help on why their draft has been declined, only to leave after ten minutes, because nobody was available to help out. Given that many helppees are trying to write about themself or a company they work for, many of their submissions do not get accepted, even after our help, but still.
 * Even though this is not a real rule or guideline, it's one that I tend to follow when I accept draft: If a draft stands a 50/50 + chance of surviving a potential AfD, I like to accept that draft.
 * Dodger67: Devising a way to "force" draft writers to RTFM before they are allowed to write the first word of their draft could help.
 * There should also be penalties for submitters who intentionally or through incompetence waste our time by repeatedly resubmitting blatant rubbish or make no attempt to actually fix the issues pointed out in reviews – AGF should not be an everlasting "get out of jail free card".
 * Regardless of the aforementioned we need to help some submitters understand that AfC is not an adverserial process, we do not put drafts on trial, we're here to help newbies write acceptable articles.
 * If we could automagically kill off most of the daily overburden of dross it would go a long way to help unclog the process.
 * Reviewer burnout is also a problem, I have just begun being active here again after taking several months off, though the cause of my burnout was not purely due to AfC it was a significant factor.
 * Anne Delong: I believe that potential article creators should have to answer a question before submitting an article, such as "All articles need references to reliable sources. Have you read Help:Referencing for beginners? (1) yes, and I have added references (2) no, send me there now (3)Yes, but I still need help, send me to the Teahouse (4) No, but I already knew how to add references and I have done it (5) Never mind, I'll submit later. This would cut down on the number of blank and unsourced submissions.
 * LaMona: I wish the AfC process had some of the qualities of the AfD process: a discussion and a consensus. Having 2 or more editors give the "publish" assessment would ease the "OMG am I right about this?" dilemma. I'd like to see the comments pushed to the talk page of the article, where it is easier to have a conversation with the editor, especially because many are new and need to be able to easily ask questions when things aren't clear.