Wikipedia talk:Featured article candidates/archive72


Post-mortem required?

I saw, but did not contribute to, Wikipedia:Featured article candidates/Black Friday (1910)/archive1, which has now been closed. It's a canonical example of a FAC which went off the rails, occupied a huge amount of volunteer time, created bad feelings, and didn't lead to a commensurate or collegial review process. Is there any appetite to have a conversation about why it went wrong? --John (talk) 15:36, 2 July 2018 (UTC)

  • wp:consensus is pathetically easy to game, doubly so with weak, laissez faire FAC coords. Anyone, but especially longstanding editors and doubly especially one in particular, can put a wall of text Oppose up regardless of merit, just dig in their heels and decline to yield no matter how overwhelming the evidence is they are wrong, then wait for FAC coords to reliably fold and declare No Consensus. FAC coords must evaluate each and every Oppose reason, accept or reject or call draw on each and every Oppose reason, else wall of text Opposes always game and always win. Here the coords will swear they overlook Opposes but not wall of text ones. Also, if each argument isn't explicitly called win/lose/draw/not wiafa and then the FAC is closed, the Opposers can recycle and reargue the completely unevaluated Oppose reasons next time, regardless how weak or nonextant their merit. That creates a MASSIVE disincentive for returning to FAC. Please refer to Bengal famine fac for most objectively obvious destruction of Oppose reasons in recent memory, plus "too many footnotes" for not wiafa, doubly true since the other tag team opposed added large amounts of footnote text... Same tactic employed in this fac, reliable outcome. Wp: consensus is eminently gameable, and has been gamed.  Axylus.arisbe (talk) 23:38, 2 July 2018 (UTC)
The underlying problem is the practice of the sandbox-to-FAC rewrite. An editor rewrites an article from scratch in a sandbox, adds it to mainspace in one edit, then takes it quickly to peer review—often the next day—where it's reviewed by supportive editors. From there it might go straight to FAC, where the same people support. Once the article is promoted, it becomes very difficult to change it.
This means that the text has bypassed the usual collaborative processes. The page watchers may not realize the rewrite is happening until it's added in one edit, at which point it's hard to object. Editors may be told they should no longer use the talk page, because someone has opened a peer review. No one wants to revert when an experienced editor adds thousands of words, but if an article you care about goes too quickly to FAC after a rewrite, it means dropping everything to accommodate the nominator's timetable: sending off for books, reading, gathering thoughts, writing up concerns, etc. It's a lot of work. We saw this cause repeated problems at Bengal famine of 1943 (a 14,000-word rewrite) and now again at Black Friday (1910).
As I wrote at the FAC, working in sandboxes can create an emotional mismatch between the rewriter and page watchers. There's none of the usual gradual transfer of "stewardship" to the editor who's earning it by putting in the most work. The page watchers see someone arrive out of the blue, inserte a rewrite in one edit, and on that basis insist on complete control. But from the rewriter's perspective, he's been working on the content for weeks and has indeed earned stewardship.
I was therefore thinking of proposing somewhere that sandbox rewrites should be given time to settle in mainspace (say, six months) before being nominated for peer review or FAC. SarahSV (talk) 02:04, 3 July 2018 (UTC)

As someone who had no involvement with this article or nomination previously (I literally just read it today), a few "outside observer" comments:

  • We can't simultaneously expect reviewers to provide actionable details of their concerns, then complain that they produced too many detailed complaints as part of an oppose. Personally, I would rather have a lot of specific feedback (even if I think some of it is wrong) than vague objections or death by inattention.
  • It's hardly original but still relevant to remind everyone to focus on content rather than on contributors. I think this is especially true at FA. If you don't like how a someone has handled edits to the article, take it to their User talk; what's relevant is whether the end result meets the criteria. If you don't like the way someone addresses you, do your best to ignore it and re-focus on the substance: is their concern a real (FAC-relevant) problem, and if so, has it been addressed. This advice is applicable to all parties and is severable -- it's helpful to follow it even if (often especially if) the other person doesn't.
  • I'm skeptical of layering in more specific process requirements, such as mandating wait time between steps on the road to FA. What mattered most was that seemingly knowledgeable editors had substantive reservations about the article's quality, not how much time or how many edits it took to reach that unsatifactory (to at least some participants) state. I'm not saying that there aren't helpful suggestions to be made about how to approach major article changes, but WP:CREEP is also good advice.

To those initial thoughts I would also add that I hope the discussion here can focus on how best to make FA discussions constructive and not get too much into re-litigating the specific issues in the nomination. --RL0919 (talk) 03:23, 3 July 2018 (UTC)

@RL0919: I agree about WP:CREEP, and with a lot of articles it won't matter whether it's rewritten in a sandbox or in the usual way. But rewrites with a political dimension are likely to be contentious, so I think we do need guidance. I also agree that what should matter at FAC is whether editors with experience in the area have reservations, but in practice a brief oppose might be overlooked if supports are arriving from experienced editors, so anyone opposing has to add detailed commentary. That's a lot of work to spring on a page watcher who didn't realize the article was being rewritten elsewhere. SarahSV (talk) 05:15, 3 July 2018 (UTC)
To me this article sounds like a case of a contentious topic where research was a little flunked. Now I am speaking from the perspective of someone who has written articles in sandbox form and on relatively uncontentious topics before, so. Jo-Jo Eumerus (talk, contributions) 06:43, 3 July 2018 (UTC)
  • I've only just looked at the FAC, and haven't read the article, but while I have sympathy with both sides, I think the bigger threat to FAC is ignoring detailed opposes rather than putting off nominators. We don't have enough reviewers, and the number of those really examining the content rather than the prose & referencing etc is much lower than it used to be. At the same time the level of pushback against such reviewers is sometimes much higher than it was. I have some sympathy with the complaints about a quick movement from sandbox to FAC, although I've done similar FAs myself - but built up in articlespace, which may be better. Johnbod (talk) 13:44, 3 July 2018 (UTC)
  • Comments from a passing editor here, I'm curious about this conversation and have been following it for a bit because I myself am working on expanding an article (Siamosaurus) in sandbox space, (link here[1]). Once it is finished I was planning on editing it into the mainspace and sending it to GA review, then likely peer review, a copy-edit at the guild, and finally FAC. But it seems to me there's some discord at the moment on whether or not this is a good idea. Are we talking about expansions or rewrites? Or is this discussion due to the more political aspect of the article being discussed? Just wondering if I'll get into any trouble later on at FAC for proceeding with this plan, thanks. ▼PσlєοGєєкƧɊƲΔƦΣƉ▼ 16:47, 3 July 2018 (UTC)
  • Chiming in as someone who has barely skimmed the surface of FAC for the last decade or so, and with effectively zero interest in the topic of the contentious FAC's article. Having read through the nom, my clear impression of the salient point, from a FAC perspective, is that the issues should have been hashed out on the article's talk page before nomination. As it was, the nomination failed the stability criterion and should have been quick-failed on that account. There was a lot of various forms and levels of objectional behaviour in there (and I'm deliberately not getting into details), but I saw no signs that anyone was acting in bad faith or being deliberately disruptive. Having bad-faith actors involved is grounds for toughing it out and keeping a nom open in the face of heated exchanges; but good-faith actors who are at worst misguided (and more likely simply not interacting optimally for situational reasons) is a good reason to archive until the differences are hashed out elsewhere.
    From a FAC perspective, letting noms like that linger on and fester is far more harmful to FAC (and, I would even go so far as to claim, the project as a whole), than quickfailing. It may sting for nominators, and there need to be some safeguards against bad-faith actors, but having this sort of acrimonious exchange play out as a part of the FAC process is infinitely worse in the long run.
    And to be perfectly clear, the practical effect of this position is that this situation was caused by the sandbox-to-FAC approach taken here. Entirely irrespective of your opinion on the merits or form of objections raised, by not eliciting them and addressing them before nomination, the nomination was (and should be) doomed to fail in the way it did. Still irrespective of form or merit, the process up to nom clearly left those who objected feeling they had the choice between objecting the way they did or tacitly support a nom with which they had fundamental reservations. By going the sandbox-to-FAC route the nominator gambled that there were no such lurking issues, and in this case the gamble blew up in their face. PaleoGeekSquared above faces the exact same gamble with their sandbox-expansion: the only material difference is the odds of the gamble. On Siamosaurus I'd take those odds. On Black Friday (1910) I would not. --Xover (talk) 18:42, 3 July 2018 (UTC)
  • To be fair, many articles sit in article space for years without attracting any serious analysis or changes, but under the FAC spotlight things can be different. In this case, there was a peer review in April, though it was closed after 3 days and 1 review (by Tim Riley) with "Unfortunately I am closing the PR, as I don't think I am in a position to be able to work on this article further. I'm sorry that you spent your time reviewing, much as I am sorry I spent my time writing it!". What might have happened if it had run longer, who knows - the PR response is highly variable. Johnbod (talk) 00:30, 4 July 2018 (UTC)
  • @PaleoGeekSquared: I might add that most articles can be worked on in mainspace. I would not bother with a sandbox unless an article is really high traffic and your edits are getting lost. Dinosaur articles are so quiet you may as well just do it in mainspace. Cas Liber (talk · contribs) 03:33, 4 July 2018 (UTC)
@SlimVirgin:, for a case of cooling off period for sandboxes, there would need to be some sort of record of other FACs going in a similar way...? Do we have others - i.e. I understand the reasoning behind the idea (of time in mainspace) but would be prudent to see if this was borne out in other FACs. @John: this is a good discussion to have. Cas Liber (talk · contribs) 03:33, 4 July 2018 (UTC)
Hi Cas, if you mean do other people leave time after a sandbox rewrite before FAC, I don't know. I've only noticed the ones where there's little or no time in mainspace. SarahSV (talk) 03:50, 4 July 2018 (UTC)
No no, I meant a list of other FACs (of articles that were expanded in sandboxes and put into mainspace and then nominated) that either failed or else had long and drawn out FACs? Cas Liber (talk · contribs) 04:21, 4 July 2018 (UTC)
I know of one sandbox rewrite that caused the same problem (the Bengal article, several FACs); one other in which nothing was said but it had similar issues; and a third that was close to being nominated but I objected. SarahSV (talk) 04:26, 4 July 2018 (UTC)
@SlimVirgin: agree this is evidence of a problem. My thinking that the first step would be some general advice along the lines that although a sandbox can be a good place for developing content, it runs the risk of not being reviewed properly as it is not in mainspace....is it worth adding something to Wikipedia:Drafts or even the FAC nomination page? Cas Liber (talk · contribs)
I have almost nothing to do with FAC but watch this space as a guide to best practice. To me, sandboxes are for testing and really should not be required at all be experienced contributors of the type likely to be nominating articles. Rewriting an article in a sandbox and then dumping it as a replacement for an existing article seems like the exact opposite of collaborative editing and it should be discouraged. I really cannot envisage an exception to this: if an article is subject to heavy traffic, such as would be the case with some political topics, then that makes collaboration more important, not less. And if such traffic means that the thing is never stable then so be it. - Sitush (talk) 03:42, 5 July 2018 (UTC)
  • Xover raises a point about 1e. It seemed to me that the Black Friday nomination did fail that, and there was even a series of reverts during the FAC. But if the coordinators don't archive when someone points that out, reviewers have to spend time writing up detailed opposes to address the other criteria. Can Ian and Andy offer advice about what reviewers should do in that situation? SarahSV (talk) 03:59, 4 July 2018 (UTC)
    Not Ian or Andy, but I am looking at the page history and I only see three reverts on one day between many more edits. I don't think that is nearly enough to justify archiving under point 1e. Jo-Jo Eumerus (talk, contributions) 06:41, 4 July 2018 (UTC)
    @Jo-Jo Eumerus: Hmm. Consider the full implications of that reasoning. Reverts are the visible artefact of disagreements over content. In themselves they are mere technical logkeeping: the important aspect of them is their cause (the disagreements). In this particular case there were very strong disagreements evident, it's just that these disagreements were made manifest through comments and discussions in the FAC and the article's talk page (and, it seems, also spilled out over various editors' talk pages). In terms of the concerns 1e tries to address, is the difference in expression really relevant?
    And further, the disagreements got their expression through FAC rather than mainspace reverts because those who objected were both 1) very experienced editors well familiar with our mores and modes of collaboration, and 2) acting in good faith. Had either of those two factors not been present, the exact same situation would likely have found its expression through means visible as reverts in the article's history. To read 1e as literally as this proposes creates strong incentives for editors to disregard our policies and practices, and rewards those who act in bad faith. Would we really prefer that those who objected here had edit-warred, in mainspace, to the very edge of 3RR rather than raise their objections in the FAC?
    In short, I think this level of disagreement in a FAC needs to be treated as a 1e fail and sent back to article talk to hash it out. --Xover (talk) 07:25, 4 July 2018 (UTC)
    Well, the criterium says it is not subject to ongoing edit wars and its content does not change significantly from day to day, except in response to the featured article process. to me this sounds like it requires an actual conflict on the page, not simply disagreement in a discussion. Jo-Jo Eumerus (talk, contributions) 07:38, 4 July 2018 (UTC)
    That's a good point. In the past it has been, and I believe it still is, the practice for the coords to archive noms where the nom itself gets out of hand, which I would assert was the case here. Whether that's anchored on 1e (i.e. formally a failure to meet the criteria), or simply as a procedural close, I hadn't really reflected on. To the degree it is the latter, and that the reliance on procedure is why it was not sent back to the article's talk sooner, I would argue that this speaks to a need to expand the scope of 1e to cover such cases. It would be disastrous in the long term to have a practice that "rewards" bad faith disruptive behaviour and "punishes" good faith constructive behaviour. --Xover (talk) 08:27, 4 July 2018 (UTC)
    I'm not convinced a change is needed. It's part of the nominator's job to assure peace on the article talk page, and deal with any disputes that may arise there. As a practical matter, it makes sense to scout out the territory, see who the largest contibutors are, check the article history, read the talk page, before beginning work on a renovation. I don't think FAC need greatly concern itself with how the nominator does it, because it's at their own risk.--Wehwalt (talk) 09:37, 4 July 2018 (UTC)
    Would keeping what you think of as the status quo result in further botched FACs like this one though? Speaking selfishly, but I imagine others may feel similarly, I wanted to contribute to this discussion, but at no point could I figure out a way that I could constructively do so. The right result was achieved (no consensus to promote) only through the eventual withdrawal of the nomination after over a month, and 170 k of discussion from 24 editors. Is there a way this could have been expedited? Or even better, avoided? --John (talk) 11:20, 4 July 2018 (UTC)
    I would like to see such things avoided, enthusiasm lost is never a good thing. But I'm not sure the answer is adding to already long FAC instructions.--Wehwalt (talk) 22:36, 4 July 2018 (UTC)
    John is right, it could certainly have been expedited; the nominator could have withdrawn the nomination at this point, more than two weeks earlier than they ultimately chose to do so. Already at that point on 24 June, User:Victoriaearle had made clear 13 days earlier, that illness would cause difficulty with her further contribution to the FAC. MPS1992 (talk) 23:00, 4 July 2018 (UTC)

John, to address your question about how to avoid a situation like this, (a) it would be avoided if nominees were required to leave sandbox rewrites in mainspace for six months before a nomination; and (b) it would be avoided or closed down quickly if coordinators and other reviewers would recognize that 1e applies. The problem was that reviewers carried on supporting as though nothing was happening, and the coordinators barely commented. It's an overused term these days, but it felt like gaslighting. Wehwalt, consensus could be sought to add a six-month sandbox "settling-in period" to WP:STEWARDSHIP.

Xover and Jo-Jo Eumerus, re: 1e, the reverting was taking place across several articles during the FAC, related to similar issues (one was about Black Friday content), and DS notices were being added. I won't give details per the request above, but they were discussed during the FAC, and the coordinators were pinged. SarahSV (talk) 00:21, 5 July 2018 (UTC)

That does not sound like it'd fall under it is not subject to ongoing edit wars and its content does not change significantly from day to day, except in response to the featured article process. regardless if the issue is on other pages. Jo-Jo Eumerus (talk, contributions) 07:12, 5 July 2018 (UTC)

Why not require the article to be a minimum GA-class, which means the edits are more likely to be in main space for longer, and it will have completed some form of formal review. Kees08 (Talk) 03:40, 5 July 2018 (UTC)

Because the GAN process is heavily backlogged. Jo-Jo Eumerus (talk, contributions) 07:12, 5 July 2018 (UTC)
Tks for the suggestion Kees. This has come up before but I think the difference between GAN and FAC is a bit too large to make the former a particularly useful pre-req for the latter. We're talking a single-reviewer process vs. a community one, not to mention the difference in standards. I also don't think the fact that an article has passed GAN necessarily precludes a major rewrite between that and FAC. I do think that articles should have the widest viewing possible before FAC, which is why I've twice in the past floated the idea of making PR (or A-Class Review if available/appropriate) a pre-req for FAC, but on neither occasion did that gain much traction. I don't recall a bedding-down requirement for sandboxed rewrites being suggested before, but if that did have consensus to become part of the FAC instructions then of course it might have resulted in an earlier procedural closure of the Black Friday FAC. Cheers, Ian Rose (talk) 09:00, 5 July 2018 (UTC)

While I agree that there should be an interval between the posting in mainspace of a sandbox rewrite and the start of any review process, a statutory six months would be absurdly long in the great majority of cases. I am not going to play the blame game with regard to the Black Friday fiasco, but I believe that similar situations would be avoided, not by further legislation but by application of a few basic common-sense courtesies. For example:

  • Anyone proposing to do a rewrite or expansion should give early prior notice of their intention on the article's talkpage, so that page watchers can be aware of what's afoot. Similar notice might be given on project pages where these are still active.
  • If the rewrite is in a sandbox, the transfer to main space is better done in instalments rather than in a single block so that the shape of the new article becomes evident over a period of time rather than in a single, possibly indigestible lump.
  • When the rewrite is complete, the author should invite comments on the talkpage, while the revised text simmers. If nothing substantial is forthcoming after say a month or so, I think the author should be entitled to seek a formal review if they so wish
  • A peer review (or other process) should not be seen as some final polishing pre-FAC stage, but as an important element in communal article-building. Cheerful encomiums ("Brilliant work, Brian!") are gratifying but are not particularly useful. PR has gone out of fashion – no more than a handful of the current FACs went through it – but properly used, it can save a multitude of problems at FAC. A lot of the adverse discussion at the BF FAC would have been far less toxic if it had taken place in the calmer atmosphere of a peer review.
  • Assume good faith at every stage in the proceedings. Brianboulton (talk) 10:18, 5 July 2018 (UTC)
I agree with all Brian's points. Mike Christie (talk - contribs - library) 11:13, 5 July 2018 (UTC)
Bunch of additional comments that are in part self serving as I plan to send some more articles (I've already done Tutupaca to FAC):
  • I do sometimes expand or rewrite articles in userspace sandboxes, typically by writing a first draft in a sandbox, merging it into the existing article and expanding and completing the draft text in articlespace. This hasn't created problems so far, probably because I write on uncontroversial topics in low traffic areas where few other people edit, which isn't the case for the article here discussed.
  • I've refrained from using formal peer review because it's stuck to me as being fairly inactive, and if memory serves inactivity has been part of the reason why earlier proposals to require PR before FAC were rejected. I wonder if requiring any pre-FAC review process may make it unduly difficult for editors working in low traffic areas.
Jo-Jo Eumerus (talk, contributions) 11:35, 5 July 2018 (UTC)
  • Clearly there's a case for a note, however phrased, warning nominators not to take the article's watchers completely by surprise. But I worry that the FAC instructions are already overlong and underread.--Wehwalt (talk) 12:26, 5 July 2018 (UTC)
  • The "natural" level of views for this article seems to be about 80 per day, much higher than most FAC candidates, but still fairly low. Many (most) wiki-projects are moribund. Most FACs only attract real attention at FAC (if then!). I don't think any of this should be mandated, though all the points are good. In particular, most articles, including this one, can and should be built up in articlespace. I always do this, and very rarely does it cause problems, even on far more popular pages. At the same time, you almost never get comments on the talk page, even if you ask for them. Johnbod (talk) 12:37, 5 July 2018 (UTC)
  • The perceived moribundity of PR is indeed one of the factors that leads to premature FAC nomination – editors believe they won't get attention elsewhere. As to whether most article can and should be built up in mainspace, I'm not so sure. Perhaps, JB, you are a quicker writer than me; it takes me weeks or months to create an article, and if I were to do this in mainspace the article would look a sorry and unreadable mess for ages. I prefer to do my spadework in a sandbox and then transfer in reasonably coherent blocks. But each to his own – it doesn't matter provided that the process is transparent. Brianboulton (talk) 13:21, 5 July 2018 (UTC)
  • It certainly takes weeks, but what you have in the meantime is generally a rather unbalanced article, or one with refs missing etc, and WP readers are well used to both of those! A really messy section can be hidden until it's ready. Johnbod (talk) 13:54, 5 July 2018 (UTC)
  • I'm an American woman about whom more than one English woman has said "Oh, she's very direct, isn't she!" So here goes: yesterday after trying to steadfastly sit on my hands to prevent myself from saying anything, I lapsed for a moment and posted this, then I changed my mind. But, I can't live with myself for not speaking up. You all can go around in circles discussing PR and the like, but the bottom line is that it comes down to dominance issues: dominating how and where content it created and controlling who may or may not edit; controlling the reviewing environment by handpicking time & place and participants, and disparaging those who show up uninvited; and, apparently, back channeling about "crazy women". Is this really how Wikipedia wants to be perceived in post #MeToo 2018? I'm deeply embarrassed that I flamed out and slapped an retired tag on my page, but I shouldn't be. I should be asking myself what drove me to that point and whether this the best use of my limited volunteer time. Victoriaearle (tk) 13:53, 5 July 2018 (UTC)
  • Victoria, the above getting-it-off-my chest post might make you feel better, but at this point it's not helpful. Nor is it wise to dismiss people with "You all can go around in circles discussing PR and the like..." No doubt you feel angry, but so do others, and not everybody accepts the narrative as you have summarised it. So be it. People here are trying to find a way whereby something similar can be avoided in future; we should concentrate on that, rather than keeping the temperature up. Brianboulton (talk) 14:35, 5 July 2018 (UTC)
  • But that may be one of the problems; men talking down to women on a feminist article, telling them to "play nice", "assume good faith", and dictating what we should be concentrating on. By censoring Victoria, you may be preventing us from reaching the solution we need, and yourself from understanding why some female editors have taken umbrage at the well-meaning paternalism of some male editors here. Victoriaearle, please unstrike the comment. Your input is valuable. --John (talk) 18:05, 5 July 2018 (UTC)

Break 1

Regarding sandboxes - my question the is that if a topic is controversial or high-traffic, then maybe that is precisely the reason that it should not be written in sandbox but in article mainspace? I guess the more I think about it the more I think sandbox editing is the antithesis of live collaborative editing...(happy to be proven wrong)? Cas Liber (talk · contribs) 20:39, 5 July 2018 (UTC)

I agree that the use of sandbox editing on high-traffic or controversial topics can be problematic. Certainly the "all-in-one-transfer" approach should be avoided in such cases. However, mainspace editing by editors with perhaps widely differing perspectives can also cause problems. The sandbox can be a useful tool, particularly for editors like me who work slowly, often with significant gaps – I'd be embarrassed to see some of my half-formed deliberations exposed prematurely to all eyes. There's no reason why sandbox editing has to be secretive; interested editors can be given access to the sandbox, something which I've offered in the past, thereby maintaining the collaborative approach. Brianboulton (talk) 21:54, 5 July 2018 (UTC)
Ah, sorry @Casliber:, didn't see your comment. As for my reply, I have to say I'm a lot like @Brianboulton: in this aspect, like he describes above I'm currently just slapping information from the relevant literature onto the Siamosaurus draft anyway I can, even if the wording sounds overly simplistic or somewhat ridiculous. Only later on do I actually start to piece everything together into a coherent article with more proper writing. So I think it's fine if I just continue editing it in the sandbox as it is. ▼PσlєοGєєкƧɊƲΔƦΣƉ▼ 22:04, 5 July 2018 (UTC)
Brian, re: "mainspace editing by editors with perhaps widely differing perspectives can also cause problems", yes, and that's called Wikipedia! That's its weakness and its strength. If a rewriter's approach is going to cause a problem, it will be caught early on and not after he's written 10,000 words. (Note: I'm not talking here about drafting parts of an article in a sandbox. I'm talking about top-to-bottom rewrites that are often added to mainspace in one edit.)
Victoria's post wasn't about feelings ("feeling angry", etc), but about facts. The sandbox–PR–FAC approach, where sandbox rewrites are taken almost immediately to peer review, is about control. PR is part of the problem, so it ought not to be offered as a solution. The suggestion that rewrites of contentious topics be allowed to live in mainspace for six months before formal review would make sure the new text had at least some community exposure. If no one reverts or comments on talk, you can assume there are no strong objections and proceed to the next step. SarahSV (talk) 22:55, 5 July 2018 (UTC)
Sarah. I largely agree with you concerning top-to-bottom rewrites. I think your view and mine about peer review are different. Yours appears to be based on what I think is a misuse of that process, a tendency towards back-slapping, rubber-stamping and exclusion. That does happen. But a peer review needn't be like that; for one thing, it can bring in a wider range of editors beyond the group deeply involved in the topic. I've experienced (and contributed to) some excellent peer reviews in the past and would like to see this process rehabilitated and used properly. On your six-month hiatus suggestion, that might be fitting for a small number of highly contentious articles, but in most cases the necessary incubation could be a good deal shorter, if the processes that I've outlined above are regularly observed. Brianboulton (talk) 23:22, 5 July 2018 (UTC)
Brian, the reviews at PR are not robust; they're mostly FAC-type reviews where people suggest that the third sentence be slightly reworded, or that there are too many blockquotes or not enough non-breaking spaces. The same people then support at FAC. Peer review should be a last step before FAC, not a first step after a rewrite. SarahSV (talk) 23:48, 5 July 2018 (UTC)
Sarah, I am not defending the kind of PR you're objecting to. I want the process to be refashioned into something far more more robust. Because there has to be some filter between article completion and FAC; except for MilHist, what else have we? GA will not do; the element of community involvement is lacking and the standard of review there is, to day the least, variable. PR, however, could provide a forum for communal article-building; in the old days I remember articles sometimes having two PRs before FAC nomination, one for the main work and one for final polishing. Brianboulton (talk) 10:17, 6 July 2018 (UTC)
I wouldn't disagree that in an ideal world PR would work like that. But there's the PR we want, and the PR we actually have. Until the PR we have is closer to the PR we want, having FAC rely on it is only adding a suboptimal process to an imperfect one. That doesn't mean we shouldn't try to improve PR too, of course, but it does mean that any improvement to FAC now will have to use a different tool. --Xover (talk) 12:30, 6 July 2018 (UTC)

I often create new articles in sandboxes. You're more or less forced to; a partially written new article may get pounced on by the NPP and moved to the draftspace, and you don't want that to happen. I haven't done it with an existing article though; I still edit these in the mainspace. It doesn't enhance the readers' experience though. And new or revised article I immediately send to DYK or GA (and then DYK). The whole idea is to get as many eyes on the article as possible. For example Operation Black Buck got 18,777 page views in 24 hours at DYK - roughly as many as it normally gets in a year. Asking for articles to be left in the mainspace for a certain amount of time is dubious because so many articles get so few page views. If I intend to take it to FAC, I then go to PR or ACR. ACR is fine for MilHist, but for anything else all we have is PR, which is indeed like FAC suffers from a dearth of reviewers these days. But I think taking to PR should be encouraged, not discouraged. When it comes to FAC, I find that it is no longer the gauntlet it once was, due to fewer reviewers. If confronted by an overwhelming amount of comment, I would withdraw the article; but there is a daunting penalty for this: a two-week break before you can nominate another article. And most articles now spend about two months at FAC, so there is a calculation at play here. Hawkeye7 (discuss) 23:43, 5 July 2018 (UTC)

Submitted for consideration, the discussion here is focused on the wrong aspect of that FAC. The problem is not sandbox rewrites per se, nor is a new "rule" regarding same needed. A well placed, well considered, actionable oppose should always cause the nomination to be archived, regardless of support (I once archived a FAC that had 26 fan supports) and the coordinators should always be alert to which reviewers/nominators customarily support each others' work. Nominations need not remain open because they get some support; a well-reasoned oppose can negate all customary support. No need for a new rule; just a return to more active archiving when there are well positioned opposes. SandyGeorgia (Talk) 23:47, 5 July 2018 (UTC)

Slightly differ on A well placed, well considered, actionable oppose should always cause the nomination to be archived given the concern raised above about FAC taking too long. It'd may be reasonable to differentiate between the oppose(s) which can be resolved within the FAC process and these which can only/should only be handled outside of FAC, with the latter ones (such as the FAC we are discussing here) being archived. Jo-Jo Eumerus (talk, contributions) 07:40, 6 July 2018 (UTC)
I don't necessarily see a contradiction there though. There are some opposes which are easy to fix, some which are hard, and some which in the circumstances are impossible to fix. In the case we are discussing, I'd have closed and archived around here, not because the problems Victoria raised are insuperable, but it was obvious even then (on 15 June) that there would be no compromise between the nominator and the reviewers. Comments like this would have made this obvious. Instead we had a further two weeks of back and forth in which little positive was achieved. This was also interesting; a bland support vote in the middle of the serious dispute about sourcing and content, which did not [seem to] take any account of Victoria's or SV's objections. --John (talk) 10:17, 6 July 2018 (UTC)
I think we should make the good faith assumption that a support, cast in the middle of such a discussion, did take those objections into account; Brian can say otherwise if he wishes. Naturally it's useful when a supporter comments on existing opposes, but it's not required. Mike Christie (talk - contribs - library) 11:09, 6 July 2018 (UTC)
I personally have no difficulty in assuming good faith on everybody's part here. Indeed, it's nice when we can all assume the assumption of good faith in one another. When we have to start reminding each other to assume good faith, it's often a sign that normal working has broken down. In a situation like this, where there are strong feelings on several different fronts, it's very important that we can disagree respectfully, but we need to be able to disagree. Telling people to "play nice" or to calm down can often be counterproductive in real life and here. Being able to engage with others' good-faith and reasoned dissent is vital to solving a problem like this, which I think is a wee bit of a systemic one. Simply talking past someone (or several people) who have taken the trouble to write a well-argued oppose can be very discouraging. --John (talk) 13:42, 6 July 2018 (UTC)
Just to be clear: mine was not a "bland support" offered regardless of the state of the discussion. At that point (20 June) there were no opposes; Victoria had said on 18 June that she wasn't going to oppose but would "walk away"; Sarah hadn't contributed to the discussion for nearly two weeks. Nor was mine the final support registered, incidentally OK, shortly thereafter things went ballistic, but that was not the situation on 20 June. This was a good faith support, on the basis of how I saw things at that time, after due deliberation, not an attempt to override existing opposes. Brianboulton (talk) 14:05, 6 July 2018 (UTC)
I've struck the adjective as, on reflection, it does not do justice to your motivation in making the support. I still think, in light of Victoria's comment below, that it would have been better to engage with the oppose than what you did, but I don't want to ascribe carelessness to you. --John (talk) 17:45, 6 July 2018 (UTC)
Thanks John. As somebody who has been reviewed several times in serious, minute detail by Brian over the years, "bland support" is just not in his make up. He was supporting under different criteria is all. St. Caurgula (talk) 15:15, 7 July 2018 (UTC)

Break 2

  • Just chiming in here to say that I've appreciated reading feedback from the community about how coords could have better managed this particular nomination. I've read and considered lots of comments about the manner in which the article was nominated and the manner in which the coords moderated it. Community feelings seem to ebb and flow in terms of how rapidly coordinators should archive troubled nominations or how stridently coordinators should deal with behavioral issues. We serve at the pleasure of the FAC community so we'll definitely take this feedback on board. --Laser brain (talk) 12:24, 6 July 2018 (UTC)
  • Brian, Mike, Andy - I have zero wish to fight with anyone. Seriously. Yes, I didn't lodge an oppose because I believed the article wasn't so far off the mark that with some work it might have been brought to a point where I could support. Weeks went by, nothing happened, none of my suggestions were taken seriously, so I moved to oppose, as is my right, based on sourcing, structure and prose. Nothing went ballistic, the nominator seems to believe I suggested men can't write about women but to suggest this to the woman who writes about male artists and modernist writers is, frankly, absurd. What I wished to see, and would still like to see, is the context added to the article that's found in someone like Jorgensen-Earp's work. What happened, and Brian is right in that I said I'd walk away, is that in the face of consistent & daily requests for specific feedback, and, (I hesitate to show weakness on Wikipedia because it's fatal) a health issue that comes & goes, there were days I could't keep up. I didn't ask for help, either on wiki or off, but I did flame out, which isn't optimal. And clearly, as my first remark indicates, I flamed out in part because when suggestion after suggestion is basically considered rubbish, to me, it seemed, and still seems, that I have little credibility. I get that. I can live it. But I don't want the coords upset, nor do I wish to be arguing with everyone. Bottom line is that the article isn't quite there in my opinion. I would like to believe that after however many rounds at this rodeo that I have some sense of the requirements. Victoriaearle (tk) 14:24, 6 July 2018 (UTC)
  • I think it was Cas who said that..gosh I'm on a cellphone and looking back is difficult... But said that in controversial articles live not sandbox editing is key.Please lets all recall reality here. Serious editing on a controversial article is often literally practically impossible. It's common to see people worn down by drip dip argue argue. I spent... I think initially 1 year in sandbox working on Bengal famine. You can look at the sandbox history and see hundreds and hundreds and even thousands of massive changes. I don't care what sv and nick claim as oppose reasons, anyone who is familiar with wp can look at the birdseye view big picture and conclude at the very least 2 things: first the article went from garbage to something fac worthy, perhaps minus a few points that appeared in SVs view to blame victims... Second, that massive amount of change would have been all-caps underline impossible if I had tried to do it in mainspace. My cellphone correct feature sometimes adds odd words if you see something strange. Axylus.arisbe (talk) 05:07, 7 July 2018 (UTC)

Just one small point, not related to any of the above – has the nomination been properly closed? On my screen it appears to be still open on the article talkpage. I've tried purging, to no effect. Brianboulton (talk) 09:38, 7 July 2018 (UTC)

The bot cannot close the nomination because the closing template was added by ShroCat and not one of the coordinators. (https://en.wikipedia.org/w/index.php?title=Wikipedia:Featured_article_candidates/Black_Friday_(1910)/archive1&diff=prev&oldid=848334271) It will have to be closed manually, or closed again by the coords. Graham Beards (talk) 10:46, 7 July 2018 (UTC)

One thing I think is being missed here in the discussions about PR, is that yes it not is a properly functioning system, nevertheless it does have one area of shining light, which the pre-fac system that Brian (with Tim) almost single handedly built up and have maintained over the last six or seven years. The two things about this is that it generally leads to FAC noms that hit this page being fully screened and prepared, and, wait for this; it is perhaps the most congenial and collegial part of wiki I have ever participated in, both as a reviewer and nominator.[2]. Yes we had one major break down here, for various and compounding reasons, but am encouraged that we have a lot of very thoughtful and nuanced stewardship in evidence in this post event analysis. In terms of an ideal working model, coupled with the final FAC process here, what Brian has built up and lead by example here is to me is the best working model I have come across. St. Caurgula (talk) 15:03, 7 July 2018 (UTC)

Most generously put, if I may say so, although I must disclaim any special credit. In a post above, Xover distinguishes between "the PR we want, and the PR we actually have", but I think that Caurgula shows that at least in some instances we do have the PR we want, and that it's worth building on, rather than being written off as "part of the problem". Brianboulton (talk) 15:43, 7 July 2018 (UTC)
Obviously agree. Overall I think pragmatism has to win out in an imperfect world, nor should we legislate for blue moon events in terms of discretion on opposes. That said, I note the irony that we were crying out for lack of opposes only a few weeks ago. My view on full-scale sandbox to mainspace transfers is that it is patently not working, and ends up being counterproductive and very hurtful to the nominator, who now has to deal with comments on a full article, rather than the usual incremental feedback main space editing attracts (on contentious rather than tumbleweed articles). Plus if the topic area is not within your usual editing areas, credibility has to be earned; and that takes time and comes from evolved demonstrated performance and the building of relationships with topic incumbents.
Alas, because of the nature of the internet "I am an expert" wont cut it. The frustration of being in that situation as nominator is self inflicted, but perhaps understandable after the fact. Individual sections worked out in sandbox and then transferred over, ie piece by piece, is to be encouraged. St. Caurgula (talk) 16:18, 7 July 2018 (UTC)

@Brianboulton: Just to be clear, I would not say that PR is part of the problem, nor that it should in any way be written off. My assertion above was simply that it is, right now, insufficiently perfect for FAC to be able to solve its own problems by outsourcing them to PR. As a general recommendation I would still suggest everyone take articles to PR before nominating at FAC (I certainly will if any of my hobby projects ever get sufficiently "finished" to merit it); I just don't think FAC can require that right now. No criticism was intended; and if it came across that way I apologize!

Incidentally, the "part of the problem" phrasing makes me suspect that you are thinking of SarahSV's argument above. If that's the case, I understood her point as being that the rush from sandbox to PR suffers from the same problems as a rush to FAC. I.e. not that PR in itself is a problem, but rather that the approach they find problematic happens to involve PR, and it would still be problematic entirely irrespective of how well or poorly PR functions. (Please correct me if I misunderstood your point Sarah.) --Xover (talk) 16:28, 7 July 2018 (UTC)

While I periodically bang this drum, I'd suggest that a much wider roll-out of A-class reviews could really help with situations like this. From my experiences with the Military History Wikiproject's A-class review process, they provide a good way for new or re-written articles to be assessed by people with expertise concerning the subject they cover before the article goes to FAC. Nick-D (talk) 00:13, 8 July 2018 (UTC)

One of the reasons MILHIST-A works is the tireless efforts of the temporarily delegated leadership. Another is the NPOV sentiments that militaria and military history encourages regarding the subject matter. A third is the significant concentration of an active editorial base, which has deliberately supported itself and its quality. HIST-A is moribund as of 2011. I would it were not so, given my background is social history. But I'd rather have the pay-off of helping an interested, polite, active community excel. :/ One, two, many MILHIST grade projects. Fifelfoo (talk) 03:53, 8 July 2018 (UTC)
I like the MILHIST-A criteria. Per WP:MILMOS#SOURCES, editors are expected to "identify every source considered to be authoritative or significant; these sources should, if possible, be directly consulted when writing the article." WP:FACR doesn't mention "authoritative"; it requires "high-quality reliable sources". Editors don't always consider whether a source is an authoritative one for the text it supports. SarahSV (talk) 06:29, 8 July 2018 (UTC)
How is it possible for a source to be "high quality" but not authoratitive? I would have thought that authority was part of the definition of high quality. Brianboulton (talk) 10:52, 8 July 2018 (UTC)
Consider the French Revolution and Napoleonic Wars. Hundreds of documents are published each year in the scholarly press. They aren't necessarily authoritative or significant texts. Also, MILHIST-A requires that the seminal and field defining sources be cited if possible. 1c only requires "a thorough and representative survey of the relevant literature." The difference is under MILHIST-A you must consult the seminal texts of the field, whereas under FACR you only have to representatively depict the literature. Compare citing "Seminal conclusion" against Bloggs, Jane (1955) Seminal, versus citing it against Smith, John (2018) "Review Article," in Journal. Fifelfoo (talk) 13:19, 8 July 2018 (UTC)
The current Arbcom case suggests that what MILHIST says is not necessarully what MILHIST does. —SerialNumber54129 paranoia /cheap sh*t room 13:30, 8 July 2018 (UTC)
I'm sure it varies, but having just reviewed the Yeomanry Cavalry FAC, which had both a MILHIST-A & PR, I was rather underwhelmed by the thoroughness of the former when I looked. The PR (earlier, and just Hawkeye7) was better, but FAC dug deeper than either. Johnbod (talk) 14:08, 8 July 2018 (UTC)
Brian, you and I discussed an example in a suffrage article that used Paul Foot's book The Vote to support that women's suffrage had been considered by the Levellers, which is an unusual claim. The book is a reliable source, and you could argue that it's a high-quality source. But it's not an authoritative source for that point about the Levellers. You wrote that you wouldn't use it that way in an article about the Levellers, but in an article about women's suffrage, it was fine. I can't see what difference that makes.
Similarly, regarding the Black Friday claims that Churchill instructed the police not to arrest the suffragettes for obstruction, authoritative sources include Churchill's and other statements at the time, letters, etc, and the Churchill secondary literature. Those sources don't seem to have been explored much. It looks as though WP:MILMOS#SOURCES would have required that those sources be used.
The problem with some FAC source reviews is that reviewers look at the works cited, see a list of high-quality publishers, and declare the sources acceptable. They often don't consider how the sources have been used, whether they're up to date, and whether important sources are missing or underused. SarahSV (talk) 14:11, 8 July 2018 (UTC)
Thanks for that, it inspired me to review better and it'll improve my reviewing. My personal checklist for conducting FAC reviews just hit 20 instructions! Fifelfoo (talk) 14:48, 8 July 2018 (UTC)
Yes, but across how many disciplines could you truly make the judgements which Sarah is suggesting should be made? Ideally, source reviews should be done by reviewers with subject expertise, but the reality is that they rarely are. There aren't enough experts prepared to review, so the work is done by those like me (and you, and others) who do their best but cannot be expected to have a grasp of the range of sources for every topic. In fairness, it isn't that often that we reach the impasse that arose over sources with the Black Friday article. Brianboulton (talk) 15:39, 8 July 2018 (UTC)
I don't think this is a question of expertise. You don't need to be an expert on the Levellers to question the use of Paul Foot, or an expert on Churchill to suggest that the Churchill literature be mined rather than the suffragette literature. It's a question of checking for the most authoritative source for each point, not only high-quality sources. An expert will see faster what is missing, but a non-expert can do Google searches and ask questions. SarahSV (talk) 15:55, 8 July 2018 (UTC)
I can't accept that reasoning. Adding to the workload of the present sources reviewers just isn't practical. The answer is to persuade more of those with specific knowledge to get stuck in. Brianboulton (talk) 16:36, 8 July 2018 (UTC)
This is about writers too, not only reviewers. We don't encourage people to think about the best source for each point in their text. SarahSV (talk) 23:39, 8 July 2018 (UTC)
Yes, writers should obviously look for the best sources, but my point is about reviewing and reviewers, who cannot reasonably be expected to assess the quality/authority of each and every source in areas where they have no expertise. Only those with deep knowledge of the subject and its literature can do that properly. They ought to be doing far more source reviews than they do. Brianboulton (talk) 10:18, 9 July 2018 (UTC)
Brian, do you mind moving User:Brianboulton/Guidance on source reviewing at FAC to project space, or is it okay if I go ahead and do it? Suggested titles: Wikipedia:FAC source reviews, Wikipedia:Reviewing sources at FAC, Wikipedia:Source reviewing at FAC, or if you prefer the longer title, Wikipedia:Guidance on source reviewing at FAC. SarahSV (talk) 00:15, 10 July 2018 (UTC)
By all means, move it, and if you feel it's necessary to add or amend the text, feel free. It wasn't written as a personal manifesto, but as general guidance, so I think the longer title is probably the most appropriate. And I suggest a permanent link on the FAC talkpage, as at the moment few can find it – I even have trouble locating it myself. Brianboulton (talk) 09:33, 10 July 2018 (UTC)
Brian, that's done. See Wikipedia:Guidance on source reviewing at FAC. Shortcut: WP:FACSOURCE. SarahSV (talk) 22:36, 10 July 2018 (UTC)
Great, thank you, Sarah. How can we get a permanent link? Brianboulton (talk) 22:55, 10 July 2018 (UTC)
Brian, I've added a section about it at the top of this page. I left it unsigned, which means the bot won't archive it. SarahSV (talk) 23:20, 10 July 2018 (UTC)
Last time I took an area of my "external" expertise (Australian Honourslevel) to WP:FAR I got kicked in the teeth so hard it made me want to leave wiki. Nine years later the article is just as bad and the most recent talk page section is precisely on that weight/content issue. The English language scholar on the missing weight area's output still isn't cited. Fifelfoo (talk) 16:24, 8 July 2018 (UTC)
I'd be interested to look at it, unless it's a sore point. SarahSV (talk) 16:39, 8 July 2018 (UTC)
  • As this is winding down, I'd just like to nail one point often made above with a detailed example. A watcher of Black Friday (1910) would have seen:
14:59, 18 April 2018‎ SchroCat (talk | contribs)‎ . . (+45,812)‎ . . (a ttweak or two...) (undo | thank) [total now 50,894 bytes]

- this was on their lists for seven minutes before:

15:06, 18 April 2018‎ SchroCat (talk | contribs)‎ . . (+90)‎ . . (→‎top) (undo | thank)

- and later loads of other small edits. At no point was there anything at all about the expansion on the talk page. Once moved, work continued, and the article would have been frequently at the top of watchlists, but only the very sharp-eyed would realize it was now 10 times larger than previously. At Burning of Parliament, there was at least 15 hours between:

15:22, 28 April 2015‎ SchroCat (talk | contribs)‎ . . (+42,460)‎ . . (undo | thank) - [total now 52,320 bytes] and
08:21, 29 April 2015‎ SchroCat (talk | contribs)‎ . . (+19)‎ . . (→‎Background) (undo | thank)

-again, nothing on talk. Johnbod (talk) 18:01, 9 July 2018 (UTC)

My impression from the discussion is that there is a lot of agreement that copying whole-article changes from a sandbox in a single edit is not the best practice, especially if there is no mention of it on Talk, and an editor who soon follows that with placement on PR and/or FAC is asking for trouble. I don't think there should be a special rule around it, and I definitely don't think using sandboxes in general is bad (I use them a fair amount myself), but the implementation of sandbox changes should be handled carefully. --RL0919 (talk) 18:50, 9 July 2018 (UTC)
Johnbod, the point about watchlists is a good one. RL0919, if we don't introduce some guidance, it's likely to happen again. SarahSV (talk) 00:22, 10 July 2018 (UTC)
Agreed: somewhere in the above thread before the first break (see 5 July 10.18) I posted four possible points which, suitably polished and formalised, could perhaps form the basis of such guidance – although getting the precise wording right might be tricky. The essence of what I suggested was:
  • Anyone proposing to do a rewrite or expansion should give early prior notice of their intention on the article's talkpage, so that page watchers can be aware of what's afoot. Similar notice might be given on project pages where these are still active.
  • If the rewrite is done in a sandbox, the transfer to main space is better done in instalments rather than in a single block so that the shape of the new article becomes evident over a period of time rather than in a single, possibly indigestible lump.
  • When the rewrite is complete, the author should invite comments on the talkpage, while the revised text simmers. If nothing substantial is forthcoming after say a month or so, the author should be entitled to seek a formal review if they so wish
  • A peer review (or other process) should not necessarily be seen merely as some final polishing pre-FAC stage; it can be an important element in communal article-building. Properly used, PR can save a multitude of problems at FAC.

Brianboulton (talk) 09:41, 10 July 2018 (UTC)

Brian, interestingly, your points are quite similar to the guidance we give educators and students through Wiki Education. From their tutorials:

Not bad advice for anyone engaging in sandbox article development. --Laser brain (talk) 11:34, 10 July 2018 (UTC)

I agree that this is good advice for handling article sandboxes generally. But I don't think it should be added to the specific guidance given at WP:FAC. It's excessively specific detail about an issue that only seems to come up occasionally. I would expect to see this kind of advice on a page that gives general guidance about sandbox editing. However, I didn't readily find any existing pages focused on guidance for using sandboxes for article changes (as opposed to new article development). Maybe there is a good place I didn't notice, or alternatively, several potential project-space names for such a page (Wikipedia:Sandboxes, Wikipedia:Editing in sandboxes, Wikipedia:Using sandboxes for article changes, etc.) are redlinks. Regarding updated guidance for FAC more generally, there are several existing essays giving extended advice about preparing articles for FAC, several of which are linked at the bottom of WP:FACR, but all the ones I know about are quite dated -- the most recent one I found had its last significant update in 2012, and many of them give advice along the lines of "inline citations are now expected", which is true but hardly news to anyone who has done significant editing in the last five years. Is anyone up for creating an updated FA advice essay? --RL0919 (talk) 13:40, 10 July 2018 (UTC)
No, I wasn't suggesting adding any of this to the FAC documentation, which is probably already bloated to the point of comedy. I dislike instruction creep and scope creep in general and would rather leave most things to common sense or defer to existing guidelines. I do agree that the existing FAC documentation could use a thorough review and update, and I would be happy to participate. --Laser brain (talk) 15:32, 10 July 2018 (UTC)
On the matter of sandboxes, I've started a draft essay at User:RL0919/Using sandboxes and would welcome input and revisions. --RL0919 (talk) 20:00, 10 July 2018 (UTC)

Close paraphrasing

Having been out of the loop for a number of years, what is the current conception of "close paraphrasing" and its appropriate and inappropriate uses? Fifelfoo (talk) 04:57, 24 July 2018 (UTC)

Citations in lead

https://en.wikipedia.org/wiki/Wikipedia_talk:Featured_article_candidates/archive70#Curious

Discussion at WP:MOS. SandyGeorgia (Talk) 15:28, 24 July 2018 (UTC)

Billy Martin

For some reason this is showing on the nom page as having one support. It actually has three. I don't know what's causing it.--Wehwalt (talk) 19:39, 1 August 2018 (UTC)

I make it four. You may get a better summary of the state of each FAC from this table which I tend to update daily.Brianboulton (talk) 22:19, 1 August 2018 (UTC)

Tumbleweeds a-blowin' thru FARC...

as usual, FARC is really quiet. It'd be great if even one or two folks could offer an opinion on an article up for FA review....Cas Liber (talk · contribs) 00:53, 10 August 2018 (UTC)

FAC reviewing statistics for July

Here are the FAC reviewing statistics for FACs ending in July.

# reviews Type of review
Reviewer Image Source Prose Total
Nikkimaria 21 2 23
Brianboulton 12 3 15
Fifelfoo 1 9 10
Peacemaker67 7 7
Wehwalt 7 7
Casliber 5 5
Nick-D 5 5
Factotem 2 3 5
Jimfbleak 5 5
Serial Number 54129 4 4
Tim riley 1 3 4
Squeamish Ossifrage 1 3 4
Aoba47 1 3 4
Dank 4 4
FunkMonk 4 4
SchroCat 4 4
Gerda Arendt 3 3
Jens Lallensack 3 3
J Milburn 3 3
SlimVirgin 1 2 3
Seraphim System 2 1 3
Ceoil 3 3
Hawkeye7 3 3
Johnbod 3 3
Jo-Jo Eumerus 2 1 3
GreenMeansGo 2 2
Chetsford 2 2
TheJoebro64 2 2
Carabinieri 2 2
JennyOz 2 2
John 2 2
Laser brain 1 1 2
KJP1 2 2
A. Parrot 2 2
HJ Mitchell 1 1
Lusotitan 1 1
<none> 1 1
Ian Rose 1 1
Richard Nevell 1 1
Tintor2 1 1
MarshalN20 1 1
Victoriaearle 1 1
L293D 1 1
Dudley Miles 1 1
Ssilvers 1 1
Outriggr 1 1
Argento Surfer 1 1
Mgiganteus 1 1
Giants2008 1 1
Vami IV 1 1
Iazyges 1 1
Parsecboy 1 1
Rjensen 1 1
Midnightblueowl 1 1
Sturmvogel 66 1 1
Mike Christie 1 1
Ltwin 1 1
SNUGGUMS 1 1
Zingarese 1 1
AustralianRupert 1 1
Ceranthor 1 1
MONGO 1 1
Juliancolton 1 1
AmericanAir88 1 1
Kaiser matias 1 1
RileyBugz 1 1
Hanamanteo 1 1
Dunkleosteus77 1 1
Mattximus 1 1
Aa77zz 1 1
Hijiri88 1 1
Freikorp 1 1
Koncorde 1 1
Krimuk2.0 1 1
Alanscottwalker 1 1
Damian Vo 1 1
Axylus.arisbe 1 1
PericlesofAthens 1 1
Paparazzzi 1 1
Famous Hobo 1 1
Grand Total 26 23 152 201

There were 105 declarations of support, 5 opposes, and three struck opposes; no opposes were converted to supports. The reviewer "<none>" in the table attests to a FAC that had no reviews at all. There were 88 reviews that did not declare support or oppose; 48 of these were image or source reviews. One media review is included in the image review statistics. The raw data is available for copying into Excel for those interested: there is a list of FACs, and a list of reviews. Mike Christie (talk - contribs - library) 14:33, 11 August 2018 (UTC)

Fix loops

I've written an essay in my user space about what I've called "fix loops": the situation where a reviewer has to return to a review repeatedly. The essay concludes that "It should be acceptable behaviour at FAC to oppose on prose and not be expected to give more than one or two examples, and not be expected to return and reread the article or provide a second round of examples". I'd be glad to hear comments about that conclusion, or anything else in the essay.

I've seen multiple discussions over the years on the underlying issue, and I think most FAC habitués will recognize the arguments. The essay was prompted partly by comments from Popcornduff, an excellent writer who was dismayed by the fix loops they ran into when reviewing at FAC. Yesterday I opposed at the FAC for Saving Light, and the nominator responded with a suggested fix that doesn't address what I was concerned about. Rather than get into a fix loop, I wrote the essay. Mike Christie (talk - contribs - library) 13:57, 19 August 2018 (UTC)

  • @Mike Christie: I think that this essay is very important. I have definitely reacted poorly to oppose votes in the past, and I am continually trying to be better with receiving constructive criticism. Do you think it would be helpful in these situations to make it very clear in an oppose vote that the reviewer will not return to the nomination for further comment as the issues extend beyond those expected for an FAC? That way the reviewer's intentions are made clear to the nominee. I hope that makes sense, and apologies if this is already covered in the essay. Aoba47 (talk) 03:41, 20 August 2018 (UTC)
    I think that's the key question -- is it acceptable to do that? It's a question of the reviewing norms here at FAC. Mike Christie (talk - contribs - library) 13:59, 20 August 2018 (UTC)
  • Thanks for the kind words, Mike. I think your essay makes sense - but then I would say that.
If it were acceptable for editors to look at articles, decide the prose is not close to good enough, oppose with a clear rationale and then fuck off, I'd be happy to do that. I just get the impression that sort of behaviour isn't smiled upon. Popcornduff (talk) 08:17, 20 August 2018 (UTC)
Mike, thanks for writing this. I think these "fix loops" are part of the reason people are reluctant to review. There's a sense that those opposing are expected to help rewrite the article. It should be enough that the prose isn't good enough and is unlikely to become good enough during the FAC. This should be accompanied by archiving after, at most, four weeks. Keeping FACs open for longer contributes to the idea that FAC is where reviewers routinely help with rewrites. SarahSV (talk) 03:02, 23 August 2018 (UTC)
Frankly, I'm of mixed minds about this as I've been stuck in the loops myself more than once because I haven't been able figure out the things in my prose that the reviewer was objecting to, although I've been usually been able to receive help from other editors to fix things that I couldn't recognize. I'd be pissed if a reviewer walked away after I fixed the specifics, but nothing else, but I could deal with that. I've also been that reviewer who's disgusted enough to throw up my hands and want to leave so I don't have to point out all the problems with a couple more paragraphs. So, to sum it up, I'd be OK with doing that with the same proviso that Sarah mentioned.--Sturmvogel 66 (talk) 03:27, 23 August 2018 (UTC)
Frankly I don't think there's much more frustrating as a writer to have someone give comments and then never revisit them. Even a simple "I'm not enumerating all the issues but it still has these problems" is a better result. What the hell is the writer supposed to do with feedback that they attempt to action but is left standing in perpetuity, because run the FAC again later and pointedly ignore the comments? It's not just an FAC issue but reviewing articles comes with a certain set of responsibilities, I would think. If it were really easy to do, then we'd have everyone doing it, wouldn't we? Der Wohltemperierte Fuchs(talk) 22:32, 23 August 2018 (UTC)
The solution is that the editor of the archived FAC needs to find outside help before resubmitting.--Sturmvogel 66 (talk) 22:57, 23 August 2018 (UTC)
I don't think this approach would be used often. For articles with a lot of minor issues, but which are clearly close to FA-quality, I'd be fine with listing scores of small points to fix. That's not a loop; that's a long list of minor issues. Perhaps "suggest withdrawal" is another way to approach this, but that doesn't address the issue of what to do if the nominator responds with "I copyedited it" or "I had GOCE go through it" and asks you to take another look. Mike Christie (talk - contribs - library) 09:53, 24 August 2018 (UTC)

Speaking as a reviewer, I suppose the way I avoid unusually long fix loops is that when I review I generally prefer to just dive in and copyedit, and save my commentary for points -- whether they be prose, referencing, or whatever -- that I simply can't work out during that ce. This works for me partly because I quite enjoy editing other people's work and I seem to be able to do it in such a manner that it doesn't get the writer offside. Of course there are times when I start copyediting -- or simply reading -- and feel it's a losing battle and then I will stop and oppose and/or recommend withdrawal. On that point, and speaking as a coord now, a recommendation for withdrawal early in the nomination almost always results in pretty speedy archiving on the grounds that the nom is unprepared and would take a long time to get right during FAC. I sympathise with David's frustration from the nominator's perspective -- and as Aoba suggests, the least we should expect is that everyone shows each other a bit of respect for their efforts -- but we have a long-standing imbalance between the number of articles wanting review and the number of reviewers willing to look at them, which means that reviewers are bound to reduce their time on some articles to brief -- and perhaps blunt -- commentary. Cheers, Ian Rose (talk) 14:23, 24 August 2018 (UTC)

Splitting FAC into three

This is not really a serious suggestion, just a trial balloon to see if the idea has any traction. What if we were to split FAC into three separate review processes -- images, sources, and prose/MoS? (We could allocate the FACR amongst the three.) I'd suggest Ealdgyth as the coordinator for the source review process, and Nikkimaria for the image review process. The current FAC pages would be for MoS/prose articles that had passed those image and source reviews; any article that had not passed them would be removed by the coordinators. Reviewers here would be expected to comment on any changes to sources or images, and could in fact raise new source/image questions if they think they've found something, but it would not be a requirement. If an article has substantial changes to sources/images reviewers could recommend a return to the other review processes. I would think comprehensiveness would have to be addressed here, but it would also be considered in the source review. An article would only be featured if it passed all three reviews.

This would address Ealdgyth's concern; the only focus of the source review process would be on ensuring top-quality sources and comprehensiveness. It might move the backlog to those processes, but that might in turn encourage other reviewers who are scared of reviewing at FAC because they're not great writers. Image reviewing and source reviewing take some experience but they are absolutely learnable, more so than prose reviewing. We might get a cadre of specialized reviewers who focus on those processes. And if not, and those processes are backlogged, it would be very clear to the nominators how to fix it -- go review sources, or images. That would address Brian's earlier comment about the lack of interest in source reviewing.

Frankly, we already have this process -- there are three independent review processes going on at WP:FAC; they just all have the same coordinators and the same nom page. Splitting and serializing them might reduce the backlog and make it clearer where there's a shortage of labour. Mike Christie (talk - contribs - library) 01:23, 27 August 2018 (UTC)

Mike, it's an interesting idea to split off and prioritize source reviews (but not images; they can be dealt with later). But there's no point spending time looking at sources if the prose is poor, and no point fixing the prose if the sourcing is poor. These issues arrive as a package (prose, organization, whether the sources are high quality, whether the sources have been used appropriately, and whether each source is the most appropriate for each point, which is rarely examined). Ideally, all reviewers would pitch in with each issue in some way. SarahSV (talk) 01:44, 27 August 2018 (UTC)
Mike, that's a very imaginative suggestion and, unlike Sir Humphrey Appleby, I mean it as a compliment. I do tend to agree with Sarah that we have a chicken-and-egg situation with sourcing and prose, though -- if anyone can come up with a way round that we'd probably be on a winner... Cheers, Ian Rose (talk)
I think Ealdgyth's list above includes quite a few things that could be addressed without any reference to prose. Sarah, yes, the issue of which source is best for each point would have to be looked at in the prose review, but that's by no means the majority of Ealdgyth's list. It's true that there's no point fixing prose if the sourcing is poor, but the reverse is only partly true. A clean up of sources, with no attention paid to prose, would address a lot of problems.
I wonder if perhaps it would also make it easier for reviewers to oppose. The narrower scope might cut down on lengthy commentary. Mike Christie (talk - contribs - library) 02:26, 27 August 2018 (UTC)
On your most-recent point, Mike, yes, it might make it easier. But it would inevitably make the process longer and more cumbersome—especially for coordinators. The current integrated critical process is far superior, despite the current shortcomings. I'd sooner clarify a few things about opposing (in the instructions, or not), which might change the culture:
  • It is considered normal to oppose a nomination pending the satisfaction of objections to one or more criteria.
  • If a reviewer opposes, they are expected to watchlist the page and revisit the page to determine whether the objections have been satisfied.
  • Nominators are encouraged to notify an opposing reviewer when they believe they have satisfied the objections; but there is no requirement to do so. Tony (talk) 02:42, 27 August 2018 (UTC)
Tony, I agree with the three bullet points you give, but exhortations to reviewers to be active in opposing have been tried before, and have not had much beneficial effect. I don't think separating source reviewing would make the process more cumbersome -- in fact, if an article's sourcing is a problem, the FAC for that article is likely to lead to a lot of wasted effort, since much will be written about prose issues before the source review is done, but this is wasted commentary if the sourcing has to be redone. If a FAC went through a sourcing review first, no time would be spent on prose until we had some confidence that the sourcing was appropriate. It might not even take longer, if we have an efficient source review process, and it would eliminate FACS twisting in the wind while they wait for a source review.
It might be possible to set up something like this on a voluntary or pilot basis, to see if it would be effective. We could put together a list of criteria, ask a couple of serial nominators if they wouldn't mind putting one of their FACs through the process before nominating at FAC, and see how it goes. If nobody else comments here I'll try to put together a more detailed suggestion for how it could work. Mike Christie (talk - contribs - library) 11:02, 29 August 2018 (UTC)
Mike, if this worked, sure. But where would the source reviewers come from? Tony (talk) 11:56, 29 August 2018 (UTC)
We already have a source review process embedded inside FAC -- there is always a separate source review for each (successful) FAC. So my first answer would be "the same place they come from now". However, a specialized source review process might attract some additional reviewers who are nervous about contributing at FAC, which has a bit of an "experts only" reputation. It's also true that most nominators at FAC never do source reviews; most review prose instead. Separating the processes might encourage some current nominators to review both there and here, which would even out the burden. Mike Christie (talk - contribs - library) 12:14, 29 August 2018 (UTC)
Mike, as you know, I trailed the problem of source reviewing on this page a few months back, when I was pretty well the only person doing source reviews. I was asking for a wider involvement, and to that end I produced a "how to do it" guidance essay which has since been improved by Sarah and others. But it's clear from that page's viewing history that hardly anyone looks at it, and with one or two exceptions there has been little inclination among FAC regulars to take up source reviewing on any regular basis. Thus only about a quarter of current FACs have source reviews (as against about 80% with image reviews). I would favour the idea of making the source review the first stage in a tripartite FAC process, but only if there is a clear commitment from more people to actually engage with it; otherwise, we could have FAC articles waiting for weeks or months before they can progress to the prose review. How about appointing a dedicated panel of, say, three, from our more experienced regulars, who for a period of time, say six months, would make source reviewing their main priority? After that period they could choose to serve longer, or could be replaced by someone else with equivalent expertise. This would of course involve an element of self-sacrifice, but for the greater good. Brianboulton (talk) 16:02, 29 August 2018 (UTC)
Source reviews are certainly disproportionately done by a few reviewers. I'm tempted to speculate why that is, but perhaps the reasons don't matter. Yes, if we can find people willing to commit for a while, it could help, but I don't think it would result in any change in behaviour on the part of most nominators. If we make passing a source review a prerequisite to a FAC nomination, the backlog at that process will look exactly as it does here, and I think that will provide much more motivation to review sources. Here, a nominator can say "my FAC has three supports and has passed image review; I just have to wait for a source review". There, a nominator would see absolutely nothing happening and it would be much clearer that this is a process that needs everyone to chip in.
The two ideas aren't incompatible, though. If we had a separate source review process, we could recruit volunteers to review regularly, at least to start with, to ensure the process gets started on a sound basis. Mike Christie (talk - contribs - library) 16:43, 29 August 2018 (UTC)