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

I am currently reviewing the article Billy Strachan which has been nominated by The History Wizard of Cambridge. I hadn't really looked at the sourcing side of the article yet because I tend to do that towards the end of my reviews. The nominator has recently been the subject of this discussion on the administrators noticeboard about the neutrality of their editing. I have now been told that some of the main sources in the article are non-neutral. It would be helpful to have some advice on what to do here. Llewee (talk) 11:01, 23 August 2023 (UTC)

See also Talk:David Ivon Jones/GA1, and I have lodged source-to-text integrity and paraphrasing/copyvio concerns also at Talk:Billy Strachan. SandyGeorgia (Talk) 12:03, 23 August 2023 (UTC)

Additionally, Trevor Carter will need a GAR if the issues aren't corrected (POV as well as the same sourcing issues and puffery). SandyGeorgia (Talk) 12:05, 23 August 2023 (UTC)

And see also the FAC for Strachan where many of the sources were deemed unreliable. Hog Farm _Talk 15:55, 23 August 2023 (UTC)

And see the ANI for BLP vios now as well. SandyGeorgia (Talk) 16:01, 23 August 2023 (UTC)

I've learned to always do the source checks first. If there's something that isn't immediately apparent but can bring the review to a halt, it's almost certainly going to be found there. Thebiguglyalien (talk) 16:30, 23 August 2023 (UTC)

If only that had happened at most of WP:DCGAR, where AGF took over after a certain number of GAs. SandyGeorgia (Talk) 17:09, 23 August 2023 (UTC)

Yeah, got to say the best (in terms of learning for me) GA review I had was my first. I'm not saying that the others were bad - they were good also - it's just that having a microscope run over your sources and how you used them was a really useful experience that wasn't repeated to the same degree in the subsequent GA reviews. Hopefully that's because my use of sourcing was better in the latter articles! FOARP (talk) 08:04, 24 August 2023 (UTC)

This is something I've given a lot of thought to the last few weeks, having reviewed ~30 articles for the drive. I've both learned and taught a lot about Wikipedia best practices through the GAN and FAC processes. I've started to make a point of going into a little more detail when explaining corrections if the nominator has no previous GAs. Thebiguglyalien (talk) 16:24, 24 August 2023 (UTC)

@FOARP and Thebiguglyalien: FYI,

• January 2021.
• Wikipedia:Arbitration/Requests/Enforcement/Archive297#The_History_Wizard_of_Cambridge
• July 22, 2023, Ruchell Cinque Magee
• August 2023 ANI

SandyGeorgia (Talk) 16:28, 27 August 2023 (UTC)

As a result of a recent conversation at a FAC, I have proposed a wording change to WP:CITE. Please comment there if interested. Mike Christie (talk - contribs - library) 19:07, 27 August 2023 (UTC)

The change to CITE has been made. I have now suggested a corresponding change to the close paraphrasing essay, here. Mike Christie (talk - contribs - library) 11:50, 29 August 2023 (UTC)

I found some minor close paraphrasing in Talk:William L. Keleher/GA1 (not enough to quickfail the review on its own but there were other more serious issues that led me to quickfail it). As the nominator has 45 approved GAs and multiple ongoing nominations, further attention may be warranted. —David Eppstein (talk) 23:40, 28 August 2023 (UTC)

Could you post the comparison? It would be easier for commenting if we could see it side by side. ♠PMC♠ (talk) 00:54, 29 August 2023 (UTC)

Source [1]: Father Keleher was born January 27, 1906, in Woburn, Massachusetts. After attending Boston College High School, he graduated from the College of the Holy Cross and entered the Society of Jesus in 1926. He was ordained a priest in June 1937. Before his appointment as president he served as assistant to the Jesuit Provincial and as director of Jesuit novices.

Our article as reviewed: Keleher was born on January 27, 1906, in Woburn, Massachusetts. He studied at Boston College High School and then the College of the Holy Cross. Keleher entered the Society of Jesus in 1926. He was ordained a priest in June 1937. He then became the assistant to the Jesuit provincinal superior. On November 1, 1942, he was made the province's master of novices.

Source [2]: Father Keleher was later professor, administrator and trustee at Holy Cross College. He was also associated with the Jesuit retreat house in North Andover, Mass. Three brothers and a sister survive.

Our article as reviewed: Keleher was a professor, administrator, and trustee at the College of the Holy Cross. He also was worked at Campion Hall, the Jesuit retreat center in North Andover, Massachusetts. ... Three brothers and one sister were alive at the time of his death.

—David Eppstein (talk) 01:37, 29 August 2023 (UTC)

That doesn't look good to me. (t · c) buidhe 01:38, 29 August 2023 (UTC)

It's not the best look, but a lot of it is basic biographical fact that's hard to restate. ♠PMC♠ (talk) 01:55, 29 August 2023 (UTC)

Which is why I put it at the level of not serious enough to be the main cause of a quickfail. But it's a lot of text to be so similar. —David Eppstein (talk) 03:43, 29 August 2023 (UTC)

Close paraphrasing is the right term to use. It may be simple biographical fact but wording and sentence structure needs to be changed from the source more substantially than this. — Bilorv (talk) 22:11, 29 August 2023 (UTC)

When I first starting reviewing, I found the Wikipedia:What the Good article criteria are not essay by WhatamIdoing to be really helpful in teaching me how to review and how not to review. But now having nominated a few dozen good articles myself, I've found that many reviewers–probably more than half–make several of the "mistakes to avoid" in a given review. The most common is the one highlighted on that page in bright yellow. Besides the fact that this allows reviewers to enforce their personal preferences, this is one of the things that makes GA a heavier and more demanding process than it needs to be. I say it would be beneficial to make this essay more visible and to keep it more actively maintained. It might also be a good place to describe standard practice on copyediting during GA: how much is necessary, and when it's more efficient for the reviewer versus the nominator to do it. Thebiguglyalien (talk) 07:43, 20 August 2023 (UTC)

I don't know that GACRNOT has full agreement. (I don't know that the guideline it explicitly contradicts in parts does either. That's GACR interpretation for you.) I find Renaming GA to something like Wikipedia:Articles that, in the opinion of a single human, meet six specific criteria, which suggests they are probably better than most articles but you wouldn't necessarily want to call them 'good' because there is definitely room for improvement, especially since they're not required to comply with all of the policies and guidelines, some of which are obviously important might give editors a clearer idea of what the process is really supposed to achieve^[1] way, way lower-reading than I'm comfortable with, and that's the GACR reading it's written from. There's probably a gap for an RGA supplementary-essay that incorporates the clear-cut-common-mistakes in GACRNOT.

On the references thing: this is something I think about a lot, because of just how common the perception is that GACR asks for something "higher than not-a-bare-ref". There's a tricky balance here between not violating CITEVAR and having...understandable references. Technically GACR doesn't require that refs have dates of any kind, but "bare refs don't have dates of any kind" is the exact reason they're discouraged. I saw "does GACR require ref dates?" come up very recently, so this isn't a hypothetical, and either answer feels unsatisfying (it "technically doesn't", but if it doesn't, why do we prohibit bare refs?). Vaticidalprophet 08:03, 20 August 2023 (UTC)

What we really need is, similar to what's just been done at DYK, a reorganisation of these diverging and often-contradictory guidelines (WP:GACR, WP:GANI, WP:RGA, WP:GACRNOT). GACRNOT is out of date, RGA is looked at nowhere near as much as it should be, GANI is a structural and organisational mess, and GACR is superficially fine but is a little wonky underneath. Until someone finds the time to draft a reorganisation, we'll continue having this discussion every six months. ~~ AirshipJungleman29 (talk) 12:44, 20 August 2023 (UTC)

I was being rather vocal about this exact suggestion a few months ago. There wasn't much interest in doing the legwork, so I tried to figure it out myself. User:Thebiguglyalien/Good article reviewing guide is my progress in that area. Thebiguglyalien (talk) 15:28, 20 August 2023 (UTC)

I think it's worth noting that some reviewers (myself included) would review an article and give you lots of things to benefit the article. Not everything I mention will be to do with the GA criteria, but they should (hopefully) improve the article. Once the nominator has looked at all my points, I would then judge to see if it meets the GA criteria. For instance, reforder is not part of the criteria (and we've even had a long discussion about it not being a barrier for any reason), but I don't see a reason not to mention it if you see it (and also care about these things). If someone said "ah, I see it's wrong, but it's not part of the criteria", I would agree and not require the change. Personally, my opinion of the "no-bare refs" argument is that if it's a website, you can run one script to fix it. If it's another type, (say a deadlink, or a badly written book/newspaper ref), I'd want to be able to read it. I don't consider it to be a big deal as it's usually very fixable with minimum fuss. Lee Vilenski ^{(talk • contribs)} 14:36, 20 August 2023 (UTC)

In my experience, noms (especially inexperienced noms) assume that everything you put in the review is required for GA status. Or, more precisely, is required to for you personally to accept the article as GA. Noms often experience a review rather like being stopped by a police officer. If the officer says "Please take one baby step forward", very few of us feel like saying "Actually, I have a legal right to stay where I am."

The common problems seem to run in waves (e.g., bare URLs, extra MOS pages, citation formatting, minimum number of refs...). The overall trend, however, is that reviewers exceed the requirements. If you want to know more why we created that page, then Geometry guy could also give you some of the history. WhatamIdoing (talk) 16:43, 20 August 2023 (UTC)

Short answer (in archive): In 2007 GA was not widely well-respected as a process, so it needed to be improved with more consistency and clearer guidelines. Now it is well respected. Geometry guy 03:28, 15 September 2023 (UTC)

This is what I thought as well. There's definitely a power imbalance aspect, even if it's over something trivial. This is why I always specify in reviews if something is "just a suggestion", "a personal preference", "not part of the GA criteria", etc. Even my usual requests are often phrased in a way that puts it to the nominator whether they think it's a good idea or not; GA should be a collaboration when possible. And on the other side, when I'm the nominator, I absolutely feel an implicit pressure to go along with what the reviewer said, even if I'll grumble to myself while running whatever non-GA errand they sent me on. Thebiguglyalien (talk) 16:56, 20 August 2023 (UTC)

I ran into a lot of trouble with trying to get History of penicillin through GA. I thought the article was fine, but eventually had to abandon the article completely. Hawkeye7 (discuss) 20:37, 20 August 2023 (UTC)

If I were in your shoes, I'd have told the reviewer flat-out that I thought they should back down a little. Have you considered returning to the review? It's been a month. Mackensen (talk) 21:56, 20 August 2023 (UTC)

Wow, that reviewer has started 30 reviews during the nine months since creating their account.

Some reviewers find that the FA or PR processes are more to their taste. Wikipedia:Peer review, in particular, does not have any criteria, so you can provide feedback according to what you think is important. WhatamIdoing (talk) 04:33, 21 August 2023 (UTC)

Please do not let GAR becone anything like the opaque, unfathomable DYK process. Billsmith60 (talk) 18:51, 20 August 2023 (UTC)

There's clearly a balance to be done between making a review difficult to follow, and one that is a straight reflection of what is on the article with no means to improve it. I get that there is a power imbalance, but if something improves the article, there is rarely a need to not do it. The issues come up when something is debateable if it improves the value of the article. Lee Vilenski ^{(talk • contribs)} 11:17, 21 August 2023 (UTC)

Fair enough, but I think it's good practice to prefix non-GACR comments with "not necessary for GA" or something like that. As well as making it clear the nominators doesn't have to address it, it also helps new nominators learn what is and isn't required for GA. I would bet that some reviewers who ask for e.g. ref formatting do so because they were asked to do that in GA reviews. Mike Christie (talk - contribs - library) 11:43, 21 August 2023 (UTC)

My allusion to ref formatting is this: quality assessment criteria are not written in stone. If we have a constant phenomenon throughout basically the concept's entire history of the absolute majority of reviewers believing ref formatting of some kind is in GACR, then it is -- just de facto rather than de jure. The categories were made for man, not man for the categories, and if practice is very unlike the written GACR document, then that implies the document should be moved towards practice. (Extrapolate this across thousands of PAG disputes all across the project.)

Ref formatting specifically is a mess for this, because you very quickly run into CITEVAR, which for all its status as "probably the single most IARed major non-MOS PAG" (how often do you go out of your way to make sure you're reproducing the existing citation style when rewriting from scratch a stub no one else has cared about since 2009?) is still vitally important in a system with several major competing styles and literally thousands of minor ones. Nonetheless, there are still meaningful questions. I refer back to the ref dates issue -- if we don't require ref dates, why don't we permit bare URLs? Bare URLs aren't discouraged because Policy Sez, they're discouraged because they're hard to fix in case of linkrot -- access dates are the one missing element that actually makes this hard, because you can't plug the link into IA and know immediately what the last guy was looking at. Vaticidalprophet 13:59, 21 August 2023 (UTC)

At the top of the guideline in which CITEVAR lives, it says While you should try to write citations correctly, what matters most is that you provide enough information to identify the source. Others will improve the formatting if needed. There is, therefore, according to that guideline, no requirement that you always match the citation style yourself.

Also, it may be helpful to know that back in the day, when we talked about a consistent citation style, we meant that you shouldn't have half the article using little blue clicky numbers and the other half using the now-deprecated parenthetical citations. GACR has never required consistency beyond that, and every past attempt to apply the FA standard has been rejected.

Speaking of which, the FA standard is that they never actually reject an article over citation formatting. If that's the only remaining problem with an article at FAC, someone will just fix the citation formatting. FA rejections over sources are because the books and articles are bad, not because the citation has a comma in the wrong place. WhatamIdoing (talk) 20:08, 21 August 2023 (UTC)

No one archives FACs because the refs aren't All Standardized To Title Case...though I honestly think they would be archived if I pointed out "no, I think this bit of MOS is absurd and I'm not going to do it" and dug my heels in, which is why I don't dig my heels in and dutifully standardize all refs to title case during or prior to FAC. (I think that bit of MOS is absurd, and I think the GACR-FACR gap on source formatting leaves GA writers coming to FAC for the first time underprepared. This is not necessarily all on GAN.) There's a pretty large chunk of FAC that I think amounts to "a lot of people think this is absurd, but no one wants to fail their FAC solely because they didn't do it". I think this is a bigger matter at FAC than GAN, because the single-reviewer structure of GAN gives less of an 'onslaught' impression and more of a 'level playing field' one, and because FAC's reputation precedes it. I strongly disagree that 'FAC rejects this source' is a synonym for 'this source is bad', but that's another matter for another place.

I've read probably low double digits of the currently-29 WT:GAN archives, though obviously large chunks of that in the earlier ones are skimming or clicking to the interesting bits of the TOC. Still working through, though. Vaticidalprophet 20:41, 21 August 2023 (UTC)

There's more fun to be found on other pages. I think that this link is the earliest version of what became CITEVAR in 2011. WhatamIdoing (talk) 22:53, 21 August 2023 (UTC)

There is a commonality to reviewer comments that exceed the GA criteria: they are cosmetic changes rather than substantial content changes. It is easy to check how a source is formatted; it is harder to check it is used correctly in the article (#2: Verifiable); it is harder still to check the sources that are not cited (#3(a): Broad; #4: Neutral). It looks like you have done a very thorough job if you ask that the inline citations to refs #20 and #17 be swapped to maintain ascending order. A legitimately thorough review requires thought about what could be missing from the article entirely. — Bilorv (talk) 16:49, 28 August 2023 (UTC)

I agree. Many comments that exceed the criteria are putting appearances ahead of substance.

Maybe we should consider a change to the reviewer templates, to say things like "Did you think about what might be missing? It may help to check articles on similar subjects for ideas." (Although I say this with some trepidation, as I remember the new reviewer, looking at a Nobel Prize article, and asking that it be expanded to describe the food served at the celebratory dinner.) WhatamIdoing (talk) 18:28, 28 August 2023 (UTC)

Of course everything is bad when implemented wrongly—some reviewers will just struggle no matter what (and in my opinion not every GA nominator has to be a reviewer)—but I would be interested in ways to encourage reviewers to check broadness more substantively. Or to inform reviewers that an intelligible reference list (that won't linkrot) is good enough for GA. — Bilorv (talk) 22:13, 29 August 2023 (UTC)

I propose that we change WP:QF so that instead of closing the review immediately, the nominator is allowed a chance to respond before the review is closed as a failure. The sudden fail with no immediate recourse is what makes quickfails so unpleasant. As standard practice, I've begun leaving reviews open when I expect to quickfail so that the nominator can respond. I've found that nominators are generally more accepting of the review closing as unsuccessful if it's a discussion rather than an imposed decision. This also has the benefit of the nominator being able to address it more easily if the reviewer made a mistake (for example, failing on copyright grounds when it's a case of backwards copying). Thebiguglyalien (talk) 00:26, 28 August 2023 (UTC)

I don't think having a hard rule either way is desirable (many QFs are obvious enough to make forcibly keeping the review open bureaucratic; whether an individual interprets an insta-quickfail or a quickfail-after-receiving-a-message as worse sounds fairly prone to individual variance). However, I'd support a thumb-on-the-scale encouragement to incentivize 'slowfails' for borderline articles, both between "QFing immediately or failing slowly" and "failing slowly or passing a substandard article after a bunch of frustration". (Fun note: the now-vestigial wording on holds is literal, as in, "for much of GAN history you were outright supposed to fail after the timer was up". I've been thinking a lot lately about reform to the categories, because they're holdovers from a very different GAN era and 2Os in particular were wildly contentious and unclear even when they were established, but the fact holds have drifted over time from "the pass-fail lightning round for borderline cases" to "a glorified ping sometime between three minutes and three months before an article is promoted" may have had complex consequences.) Vaticidalprophet 00:42, 28 August 2023 (UTC)

After posting the above, I wondered if I should have stuck the word "suggestion" somewhere in there. The main point is that we should try to prevent quickfails from being unnecessarily harsh. Thebiguglyalien (talk) 01:01, 28 August 2023 (UTC)

If we insist that nominators must have a chance to respond to a quickfail, what is the point of having a distinction between a quickfail and simply a fail? ♠PMC♠ (talk) 01:03, 28 August 2023 (UTC)

I always understood a quickfail to be a fail before the review was completed. If I review all six criteria and then fail, it's just a regular fail in my mind. I think of a quickfail as "there's no point in doing a proper review because the whole article needs to change before it reaches GA". Thebiguglyalien (talk) 02:17, 28 August 2023 (UTC)

In my mind the distinction between fail or quickfail isn't about the thoroughness of a review, it rests entirely on whether or not the reviewer gives the nominator a chance to respond and/or work on it before failing. I've done reviews where I brought up lots of issues, but felt they were reasonably fixable, so I kept the review open for the nominator to respond. I've done equally thorough reviews where it was only at the end that I concluded that the article was a long way from meeting the GACR and shut it down without waiting for a response. ♠PMC♠ (talk) 03:28, 28 August 2023 (UTC)

Yes. If, as a reviewer, I thought that a quick-fail review could drag out into a month-long "but what if I change each of the individual concrete examples that you listed as an example of a pattern of badness without addressing anything else in the article" battle, with added "just tell me what it should say so that I don't have to put the effort into it myself", the way DC used to drag out negative reviews, I would never begin the process. —David Eppstein (talk) 04:52, 28 August 2023 (UTC)

+1 on this. Vaticidalprophet 04:56, 28 August 2023 (UTC)

All right, but the next time a nominator demands to know why I had the nerve to fail their subpar article without giving them a chance to "fix" it first, I can't promise I won't get just a little snippy with them. Thebiguglyalien (talk) 05:23, 28 August 2023 (UTC)

Quickfails are explicitly included in the instructions, so it's not as if doing one is wildly off-book. It's not your fault if a nominator sends up an article that's well below the GACR. Unfortunately, disappointing people is sometimes part of the job when you're doing any type of content review, and sometimes people choose to express that disappointment. I would point them to the GACR/GAI and remind them they can nominate again whenever they like - there's no time limit. ♠PMC♠ (talk) 07:28, 28 August 2023 (UTC)

At the end of the day, it is a review, conducted by a reviewer independent of the nominator. If the reviewer comes across a situation where a reasonable person would quickfail, they should not be required to delay this inevitable action because of red tape. We should avoid adding more bureaucracy to GAN without good reason, in my opinion. Trainsandotherthings (talk) 01:45, 28 August 2023 (UTC)

When you say "respond" do you mean that as "ask questions" or "fix all of the problems"? For the first, I think that's respectful and have made a point not to close the discussion when marking a quick fail as not passed. Regarding the second, I think that would entirely depend on the individuals, Rjjiii (talk) 06:45, 28 August 2023 (UTC)

Definitely not. I think we’ve more or less set an expectation at GAN that the reviewer should essentially try their absolute hardest to ensure that the nominator can get the article passed, no matter how substandard it is, as long as it doesn’t have copyvio in it. This is absurd for a volunteer process where the actual requirements are only to provide comments on articles that are reasonably close to passing all the criteria. AryKun (talk) 03:27, 31 August 2023 (UTC)

If we wanted to change the culture to make it clear that no one is owed a GA and the reviewer is expected to be strict on the criteria, I'd be fine with that too. Right now it's the disconnect that I don't like. Thebiguglyalien (talk) 03:44, 31 August 2023 (UTC)

Two variables that I think I take into account when quickfailing are how well I know the nominator's work, and which criterion is the problem. I've had no hesitation in quickfailing nominations with multiple serious copyvios or multiple unreliable sources, for example, but in the unlikely event I was to run into a nomination by someone I personally knew did good work, even then I might hold off and post the issues first before failing. That probably applies to everyone who's commented so far in this section. Another way to say that is that if, say, a nomination by PMC (to pick a name at random from above) appeared to be quickfailable, I would have enough respect for their prior work to ask about it first, knowing how annoying quickfails are. There are other markers for editing ability that I use if I don't know the editor's work from my own experience: if they have tens of thousands of edits; if they have multiple GAs and/or FAs; if some of the material in the nominated article impresses me as the work of a very good editor; perhaps a couple of other things. Those could also persuade me to delay a quick fail. I think that's probably not entirely fair, but I would justify it by saying the quickfails I do are sticking to the rules, and it's up to the reviewers judgement to not go the quickfail route if they wish. Mike Christie (talk - contribs - library) 07:53, 28 August 2023 (UTC)

Maybe, then, there should be more explanation of which failed criteria can be quickfailed right away vs giving time for the nominator to respond. I don't think, for example, that an article with 70+% copyvio (and not to a WP mirror) could be fixed anywhere near quickly enough to pass GAN. But if it's just prose or [moderate] broadness issues, then that theoretically could be fixed if the nominator is determined. SilverTiger12 (talk) 14:32, 28 August 2023 (UTC)

It's explicitly stated in the GACR that any of the six criteria can trigger a quickfail if the reviewer believes the article is a long way from meeting it. Yes, this means there is a grey area where some nominators will QF and some will hold out for improvement. There is no way to create an objective measurement for what ought to be a QF that will hold up every single time with no exceptions, so there will always be some element of subjectivity. That's a major reason that we don't put any limit on renominations. ♠PMC♠ (talk) 22:04, 28 August 2023 (UTC)

I struggle to think of when a quickfail would be justified on grounds of #6 (media), but I can think of cases where the other 5 in isolation could all be valid grounds to quickfail. It is a case-by-case process, so I'd be wary of trying to codify it. It's almost an in-built IAR for the GA review process: "if dotting the 'i's and crossing the 't's in a full review is a waste of time then don't bother". — Bilorv (talk) 22:11, 29 August 2023 (UTC)

Yeah, I almost wrote something similar about #6, but you never know - conceivably an article could misuse non-free images so blatantly that it would have to be revised from the ground up to make any sense, or could have a grotesquely overblown gallery that the nom refuses to trim. Perhaps a NOTCENSORED issue one way or another. I agree that one is the least likely for a fail, but overall the point is that we don't need to qualify which criteria can be quickfailed, because the GACR already state that any of them could result in a quickfail. ♠PMC♠ (talk) 22:22, 29 August 2023 (UTC)