Jump to content

Wikipedia talk:WikiProject Highways/Assessment/A-Class Review/Archive 1

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1

Basis for the initial setup

Most of the setup is from the process over at Wikipedia:WikiProject Mathematics. Some other items have been added from various other WikiProject A-Class reviews. Comments are welcome regarding ways to improve the system. --TMF Let's Go Mets - Stats 07:38, 5 November 2007 (UTC)

Minimum level of support?

Should we impose a minimum number of support votes to get something passed? This would ensure that there is a sufficient number of people that review the articles. How does 4 supports counting the nominator? --Holderca1 talk 00:14, 28 February 2008 (UTC)

Sounds good. --Rschen7754 (T C) 22:50, 28 February 2008 (UTC)
Works for me as well. --Imzadi1979 (talk) 02:36, 29 February 2008 (UTC)
Sounds good to me, but I'd change it to four net support. --TMF Let's Go Mets - Stats 02:27, 29 February 2008 (UTC)

Over 7 days?

What do we do about discussions that have been open for over 7 days with no votes? I see quite a few of these on the page now, some open since the middle of February, with no votes, much less the 4 net support votes, only comments. Should they be closed? —Scott5114 [EXACT CHANGE ONLY] 17:38, 18 March 2008 (UTC)

I would say if no one has addressed the comments, then close. If the nominator is actively working on improving the article it should remain open, they shouldn't be penalized because of lack of reviewers. --Holderca1 talk 18:01, 18 March 2008 (UTC)

M-28

I think M-28 (Michigan highway) was closed a bit prematurely. From what I can see, O has an objection to the article that hasn't been addressed. While it has the sufficient number of support votes (6), I don't think I would close it with open oppose votes. --Holderca1 talk 15:55, 21 March 2008 (UTC)

I'll wager a guess though that's it's the same objection he's raised at the NY 174 ACR that Polaron objected to being a valid objection (a skewed view of English grammar.) The only other thing he objected to was citations in the lead, which per WP:LEAD, M-28 follows. In fact given a comment made about this at the NY 174 ACR, it's been mentioned at WT:USRD to formulated a more concrete project guideline about citations in the lead, where consensus now seems to be leaning away from redundant citations. Given that those were his last objections, I'll leave it to others to decide if closing it was valid. We already had O mention if it was proper for someone voting in an ACR close the review, and decided that where possible, we should get a third party to close it. In the event that all active editors participate, we'll discuss that situation then. Imzadi1979 (talk) 15:41, 22 March 2008 (UTC)

Procedure wording issues

The procedure, as currently worded, seems to omit the fact that this review process is also intended to be used to nominate an A-Class article for demotion to the sub-A classes (either GA or B depending if it went through GA). As it reads right now (particularly the "Participating in discussions" section), it covers only the promotion half of this process. I'll look into rewording it in the days to come if no one else does by that time. --TMF Let's Go Mets - Stats 14:55, 1 April 2008 (UTC)

I agree, for the most part they should be pretty similar although the nominator should list what is wrong with the article so someone can fix it. The voting procedure would be reversed as a support vote would indicate that you support demotion (which should require something that can be fixed to sway you the other way). We should probably require a person that is nominating an article for demotion to contact the person that originally nominated the article and any other major contributors to the article. --Holderca1 talk 15:29, 1 April 2008 (UTC)
I'd only modify that to suggest we use the terms keep or demote as appropriate instead of support and oppose. Imzadi1979 (talk) 04:15, 13 April 2008 (UTC)

Absentee commentors

Since this process is billed as aiming to "provide a FAC-style venue without the FAC-style atmosphere", I think we should do something about a problem with the FAC atmosphere that I sadly see slipping into ACR. People leaving a laundry list of comments (sometimes even with an oppose) and then wandering off and not checking the nomination again. It is extremely frustrating for a nominator to fix all of a person's concerns, only for it to be to no avail in winning the person over to the support column.

I would propose something along the lines of "If someone asserts all concerns raised by a reviewer have been addressed, and the reviewer does not reply within 5 days (or at the close of the nomination, whichever is first), then the comment counts as a support". This helps motivate commentors in opposition to a proposal to watch over the nomination. Also, if someone says "Fixed", and it hasn't been fixed, all that needs to be done by the original reviewer is to pipe up and say "No, it hasn't," and outline why, so those concerns can be fixed, and eventually the commentor will be satisfied. —Scott5114 [EXACT CHANGE ONLY] 22:05, 5 April 2008 (UTC)

Support — I can support this proposal. It sounds very reasonable. Imzadi1979 (talk) 03:34, 13 April 2008 (UTC)

Comment: I've already run into this as GA reviewer. Leaving a message on my talk page helps me, and I'm sure other reviewers, a lot, especially when I'm doing 5 reviews in parallel. (After the GA backlog is down to a reasonable level, I'll consider doing A-level reviews.) —Rob (talk) 04:17, 13 April 2008 (UTC)
Strong oppose - I specifically put my comments behind a comments rather than a oppose for a reason. I don't catch every problem and I typically want to see what other reviewers have to say and before I will give my support. If this is how it is going to be handled, than I will just stop reviewing articles. --Holderca1 talk 16:34, 13 April 2008 (UTC)
That's all well and good, but that's what the peer review process we have (which is seriously neglected, by the way) is for. While ACR is can certainly be used to highlight issues, that should primarily be done at PR. The main point of ACR is so we can decide promote/not promote. Really, we should be able to decide these things in 7 days, but everyone's commenting and nobody's voting, so it's bogging the process down to a painful level. M-35's review has been open since the end of February simply because nobody is voting either way on it. And besides, if the commenter isn't going to support you even if you fix their concerns, what's the point in fixing them? I mean, yeah, it makes the article better, but... —Scott5114 [EXACT CHANGE ONLY] 19:10, 13 April 2008 (UTC)
The issue is that we don't have enough reviewers, I haven't had time to make full reviews of the articles that are currently being reviewed, I spent about a minute on each review. Like I said, if someone is going to come along and change my vote (or lack thereof) without my consent, then I am going to refrain from submitting comments at all, because I don't want it to look like I support an article I haven't even fully reviewed. --Holderca1 talk 10:59, 14 April 2008 (UTC)
Ok then. I would comment on some, but that would mean I couldn't close any. And the last time that I commented on one and closed it, someone got all bent out of shape with me, so I'm not doing that again. —Scott5114 [EXACT CHANGE ONLY] 02:44, 15 April 2008 (UTC)
Oh, and if nobody's voting on anything in this process, I don't see any reason for it to exist. If nobody votes and everyone just comments, then we can't decide pass/not pass on A-Class, and this just turned into a sexed-up peer review. So if that trend continues, I think we should just deprecate ACR in favor of HWY/PR. —Scott5114 [EXACT CHANGE ONLY] 02:47, 15 April 2008 (UTC)
Deprecate this process in favor of one that works even worse...good call. Anyway, the reason I leave comments first is that I only look for blatant flaws to begin with for the sake of time. I haven't had time lately to give anything a full review, and to be honest I still don't. The essence is this: there are only about a half-dozen people that participate in this process, so anytime there's more than one article nominated, there's going to automatically be a backlog since I doubt this process is the highest priority item of most of said participants. – TMF 03:58, 15 April 2008 (UTC)
There is nothing wrong with closing reviews out that you have commented on. If I recall correctly, the one you closed out that you are referring to had open oppose votes that hadn't been addressed. I even think the nominator of the article responded to that oppose vote by saying that he already had enough support votes, so he wasn't going to worry about that oppose. If an article has nothing but the required number of support votes, there isn't a problem with closing them out. I have closed out several that I have commented on and no one has had an issue with it. Also, I don't see the big rush here, there isn't that big of a backlog, only four articles presently. If you don't have the patience to deal with this process, then take the article straight to FA. --Holderca1 talk 10:45, 15 April 2008 (UTC)
  • Strong oppose per discussion above. That, and frankly it's just wrong to change someone's vote. If someone forgot about their ACR comments, nudge them a little bit on their talk page or on IRC to get them to take a second look at the article. If they've suddenly disappeared off Wikipedia, well, that kinda sucks, but I think 99% of the time, people don't start off opposing, so at least it doesn't get stuck as a permanent negative vote. (Plus, God willing, people won't suddenly disappear, so it's a moot point. :-P) Basically what I'm saying is that you have to suck it up and give the process time to work. Give the reviewers a chance to give the article a second look to catch anything they may have missed, and more importantly, confirm that their concerns were addressed to their liking.

    On the tangential issue, I agree that this process has become somewhat redundant to peer review, but I don't think that's necessarily a bad thing. By making it peer-review style, I think it just helps more articles get to A-class by showing people exactly what they need to do, and encouraging it to get done. If we were to deprecate one (not that we should), it should be WP:HWY/PR, not this one. -- Kéiryn talk 15:24, 16 April 2008 (UTC)

The issue with deprecating HWY PR in favor of this is that it leaves non-USRD articles out in the cold, so to speak. However, non-USRD articles never got any kind of meaningful response there anyway so it'd probably be no different in practice. :P
The fact that this has become somewhat of a peer review (people oppose/comment, nominator fixes concerns, people support) does contribute to the backlog somewhat, since an article that is not up to snuff and not fixed during the process will be pushed through (and failed) much more quickly than those that undergo reviews and edits. But I see this as more of an unavoidable consequence than something that can be or needs to be fixed. Really, the only solutions that will both unquestionably work and make perfect sense are to 1) get more reviewers and 2) be attentive of the review: to the commenter/opposer, if your concerns have been fixed, strike them and support if desired. The first item is out of anyone's control; the second is well within. – TMF 16:36, 16 April 2008 (UTC)

Vote counting

I think there's a fundamental flaw in the way we count votes, namely counting the nomination as a support vote. Granted I haven't seen that many articles go through this process, but from what I've seen, it always goes the same way. Someone nominates an article, and suggests it get promoted to A-class, even though it's really not ready yet. Then people suggest a bunch of changes, the article gets fixed, and those people vote support. Basically what I'm saying is that I don't really think a support vote before the article gets fixed should count the same as a support vote after the article gets fixed, if you catch my drift. From what I've seen, other procedures on Wikipedia work the same way, i.e. the original nominator can't vote in a good article review.

Plus, it seems we're only counting the nomination if they explicitly stated that they want to promote the article, which essentially means we're punishing editors for their honesty/humbleness when they give no explicit opinion.

It doesn't matter much to me if we reduce the number of net supports required to 3 if we discount the nominator, or keep it at 4 – I'm just trying to point out that it seems really odd to count the nominator as a support vote. (God I hope any of this made sense.) -- Kéiryn talk 15:24, 16 April 2008 (UTC)

I think I agree with this sentiment; I never really understood why the nomination counts as a vote, since that's not how either GA or FA (the two processes that sandwich this one on the scale) does it. To be honest, I think the only Wikipedia process that counts the nomination in a tally is RFA—and that's only because the nominator is allowed to vote. – TMF 16:15, 16 April 2008 (UTC)
FL counts the nominator, but in all honesty it doesn't really matter, we could change it to 3 votes not counting the nominator. I also don't understand why "net" was added. Is that suppose to mean support votes counting the nominator? --Holderca1 talk 17:05, 16 April 2008 (UTC)
Net means for every oppose vote the article gains, it needs an additional support vote along with the minimum required number of support votes. See wikt:net as an adjective. – TMF 17:10, 16 April 2008 (UTC)
So if an article had 10 support votes, and 6 oppose votes, it would pass? Not sure I completely agree with that. If the oppose votes have valid concerns that can be corrected, they should be addressed prior to being closed out. I also believe that if a voter votes oppose and all of their issues have been resolved, but they don't come back to review the fixes and strike their vote, then it would fall upon the person closing it to decide whether to discount that oppose vote, but not changing it to support as the above proposal suggests. --Holderca1 talk 17:22, 16 April 2008 (UTC)
From the ACR page: "Please remember that the process is not entirely a vote, and articles with majority support can still fail if they don't fully meet the criteria." If it has six oppose votes, 1) it probably doesn't meet the criteria anyway and 2) someone spammed the ACR in a high-traffic location since I don't foresee any ACR getting six votes period anytime soon (and certainly not 16 as your example above has). The net is intended to prevent articles that have three or four supports and one legitimate oppose from passing without discussion of the opposer's issues. – TMF 17:26, 16 April 2008 (UTC)
The 16 votes were just a hypothetical situation. I just don't want to see a nomination get closed out if it has 5 support votes on day 6, then on day 7, someone with an issue with it and votes oppose is just ignored without addressing their issues. --Holderca1 talk 18:39, 16 April 2008 (UTC)
That situation could occur even if supports are counted without factoring in opposes. Anyway, I was under the impression that four net support votes was the minimum needed for a successful nomination and was in no way a threshold for automatic closure, especially if there are open concerns. Going back to your hypothetical example above, 10-6 isn't consensus - I certainly wouldn't promote the article with that tally (plus as I said before, an article with that bad of a support percentage probably wouldn't be that good anyway). – TMF 20:08, 16 April 2008 (UTC)
The reason I bring it up is because of this comment [1], with the nominator basically ignoring a commentor's concerns because the nomination would meet the time requirement and the net level of support. --Holderca1 talk 20:27, 16 April 2008 (UTC)
Hm...yeah, that's really not the attitude that nominators should have since the 7-day period and four net support things are the minimum requirements for an ACR. Now, if the nominator thinks that the opposer's concerns aren't valid, that's one thing, but to dismiss them just because "it currently enjoys the necessary level of support to be passed" is sort of both a slap in the face of the reviewer and not really what this process is about. – TMF 22:20, 16 April 2008 (UTC)

So... were we pondering actually changing this? -- Kéiryn talk 16:47, 28 April 2008 (UTC)

IMO the change should be made. – TMF 00:49, 6 May 2008 (UTC)
So do we continue to require 4 net supports, or only 3 now that we're discounting the nominator? -- Kéiryn talk 01:23, 6 May 2008 (UTC)
I say 4 net excluding the nominator. The more eyes that are required to look over the article the better it is in the long run. – TMF 01:49, 6 May 2008 (UTC)
Agreed as the nominator support is usually a throwaway support. --Rschen7754 (T C) 01:53, 6 May 2008 (UTC)
Okay, so basically we are now requiring one more support vote? Before it was three not counting the nominator, now its four? If so, are the articles currently nominated subject to this change, or new noms from this point forward? --Holderca1 talk 02:10, 6 May 2008 (UTC)
Yes, it would essentially require one more support vote in most cases. IMHO, common sense would dictate that it would apply to any new noms, not any of the articles currently nominated. -- Kéiryn talk 02:33, 6 May 2008 (UTC)

I-15 in AZ

If someone wouldn't mind closing out the I-15 in AZ review, it would be appreciated. I would do it myself, but to avoid the appearance of a COI, I will refrain. It meats the required support and time to pass. I have already updated the articles talk page to make it easier, so all that is needed is to close out the discussion. Thanks. --Holderca1 talk 15:20, 1 May 2008 (UTC)

U.S. Route 491

U.S. Route 491 has been here since June! We need more voters from the project to come here. ~~ ĈőмρǖтέŗĠύʎ890100 (tĔώ) 15:59, 3 August 2008 (UTC)

Well, once my minor concerns are met, it will have five net support votes and will pass, plus there's been at least one review that's taken two months to get done. But I agree that more participation is always good. -- Kéiryn (talk) 17:27, 3 August 2008 (UTC)

Participating editors closing reviews

Ok, I think we need to cement something here. It's been discussed before, but to me it's now time to finalize the concept that if you are participating in a review, you shouldn't be closing it. That would even apply to withdraw a nomination. Once a nomination is made, it's the "property" of the community of editors, not the nominator, IMHO. There are plenty of non-USRD editors out there that would could draw on to close reviews. I've even abstained from a review so I could close it, only to see a participant in the review close it before I got to it.

Recently I gave a review of an article. Some might even call it a tough review. I wanted to ensure that once the issues I mentioned were addressed that this article would be, in my opinion, worthy of the A-Class designation. Nothing in the article was impossible to update, nor would it have required any major overhauls. I actually felt cheated that the article's nominator closed the review before addressing the issues. Now once those updates are made, and many of them have, a new ACR will need to be started.

FACs are closed by a designated set of people, so I think the idea of neutral parties being asked to close our ACRs is valid. What does everyone else think? Imzadi1979 (talk) 01:52, 4 August 2008 (UTC)

I agree that we should have one (or two) people set to monitor the A-Class review section. --Admrb♉ltz (tclog) 02:01, 4 August 2008 (UTC)
I would be happy to volunteer to refrain from reviewing so that I could neutrally oversee the process. -- Kéiryn (talk) 02:13, 4 August 2008 (UTC)
Then maybe we should nominate and elect an "ACR director" of sorts, make a position somewhat official then? Imzadi1979 (talk) 02:17, 4 August 2008 (UTC)
Agreed. I would also like to volunteer for this as well. --Admrb♉ltz (tclog) 02:19, 4 August 2008 (UTC)
Just as a side note, the FAC deputy director, SandyGeorgia, does participate in FACs, but only as a means to clarify questions and replies, in order to determine consensus. I'd only suggest we had one person regularly doing this role though since the pool of quality reviewers is so small, I wouldn't want to decrease it that much more. That being said, if the director had an article to nominate, then I'd suggest someone else be nominated to perform the task for that review only, and not a designated deputy. Imzadi1979 (talk) 02:24, 4 August 2008 (UTC)
I don't mind not reviewing, as Kéiryn (talk · contribs) provides much better feedback than I usually do. --Admrb♉ltz (tclog) 02:30, 4 August 2008 (UTC)
An idea could be to rotate whoever reviews teh discussion for consensus as well. I would be fine with having an ACR director, however. --Rschen7754 (T C) 03:26, 4 August 2008 (UTC)
I think ideally there would be 2 or 3 "directors" – looks like we already have two volunteers – so that if I see something that needs attention that no one else caught, which IMHO happened with US 491, I could participate in the discussion and let the other director(s) know, "Hey, would you mind closing this one for me?" Plus, I know I haven't been all that active recently, but eventually I'll get back into article work, too, and I'd like to be able to nominate my own articles for ACR too. -- Kéiryn (talk) 11:45, 4 August 2008 (UTC)

I am willing to help. ~~ ĈőмρǖтέŗĠύʎ890100 (tĔώ) 17:35, 6 August 2008 (UTC)

That's fine, but keep in mind that the main point here was that the person closing shouldn't be someone who participated in the discussion. Both of the ones you just closed were ones you supported. -- Kéiryn (talk) 18:26, 6 August 2008 (UTC)
Okay. I will stop and thanks for reminding me. ~~ ĈőмρǖтέŗĠύʎ890100 (tĔώ) 21:57, 6 August 2008 (UTC)

Discussion at WT:USRD

See Wikipedia talk:WikiProject U.S. Roads#ACR and GAN changes. --Rschen7754 (T C) 21:15, 16 November 2008 (UTC)

Proposal to decrease supports needed

I propose that the required net support should be decreased to 3 votes. A-class should be a class between GA and FA, so articles should pass through this process more quickly. Also, many of the supports as of now are by the same people. We are not a large project by size, so having 4 votes is a little ridiculous and many articles end up waiting for long periods of time awaiting review. — PCB 01:52, 11 April 2012 (UTC)

I think decreasing the net support to 3 votes is a good idea based on the amount of active editors available. This should speed up the ACR process. Dough4872 02:02, 11 April 2012 (UTC)
I am neutral on decreasing the net support to three votes. There are good arguments on both sides. In favor, we have only about 10 people in the project who are comfortable performing an A-Class Review. As it stands, a significant percentage of the project—excluding the nominator, who is also likely to be a potential reviewer at ACR—needs to get their hands dirty for an article to be promoted. In opposition, having that extra review improves the chances that something that will be chosen as a reason to oppose at FAC will be caught before the article is sent to FAC.
Because it is relevant, I disagree with PCB's contention that A-class should be a class between GA and FA, implying a significant quality difference in quality between A-Class and FA. Articles that successfully complete the ACR process should be as close as possible to being FA-class in all but having the bronze star. While I am generally in favor of getting articles through ACR as quickly as possible, quality is a much more important result of the process than expedience. I think someone mentioned how our A-Class articles usually do not stick around long because they get promoted to FA. I see that as a very good thing. That means our ACR process is working to properly prepare articles for FAC. I am not arguing against improving the process, but I have yet to be convinced that reducing the number of required reviewers will improve the process.  V 17:37, 11 April 2012 (UTC)
I am in agreement with VC here. Quality should always trump quantity. This isn't GAN where you can flood the queue with scenery and construction costs. –Fredddie 21:36, 11 April 2012 (UTC)
I'm on the fence on this one. When I first started bringing articles to ACR (when it was USRD-only), the standard was at one point a net four supports, including the nominator, or essentially what PCB is proposing (net three not counting the nominator is the same as net four with the nom). The CRWP edition of ACR had a net three rule for promotion, but in the merger, accepted net four under the promise that there would be a greater reviewing pool. Wikipedia:Version 1.0 Editorial Team/Assessment/A-Class criteria by the way only requires two editors to agree to promote an article to A-Class; we're the ones levying additional conditions.
I tend to agree that there is a difference between what A-Class means and what FA-Class means. In some respects, we should be encouraging articles to be promoted to A-Class regardless if the nominator has any plans to take the article to FAC. No matter what we would like, there is a difference, however minor. A-Class will never be a fast-track to the FA star, but the track record does demonstrate that a good review here saves trouble there. That said, quality trumps quantity, but we need to work to cultivate this process so that items aren't back logged. If the rigor of the reviews remains high, I have no trouble returning to a net three supports, but if those future reviews are rubber stamps, I can't. I think before we formally change the support requirements, we might want to clarify a set of A-Class criteria first. Imzadi 1979  22:02, 11 April 2012 (UTC)

Late to this discussion, but I agree that we need to toss the net support rule once and for all. Yes I know we technically "did", but we're still using it internally. I agree on the A-class criteria proposal Imzadi has proposed, not by choice, but that's because we're rubber-stamping stuff that really shouldn't be and not demanding egregious information such as WHERE THE HELL DOES ROAD PASS THIS FARM AT THIS LOCATION?. Mitch32(There is a destiny that makes us... family.) 13:56, 23 April 2012 (UTC)

KMLs on A-Class articles

Closures based on opposes

Closures based on inactivity

Demotions of ACRs and reviews of FAs

Source reviews at ACR

We could be saving some time with future FACs if we'd incorporate a source review at ACR in the spotcheck. A spotcheck reviews fidelity to the sources used to write the article, but a source review would verify that the sources used are both high-quality and reliable (in the Wikipedia meanings of the terms). A source review would also check for consistent formatting of the citations in the article. This is another easy thing that could be done at the ACR stage so it could be referenced at the FAC stage to save time and effort. Thoughts? Imzadi 1979  23:24, 21 July 2013 (UTC)

  • I support adding a source review to the spotcheck. TCN7JM 23:52, 21 July 2013 (UTC)
  • Support - Makes it easier for the article at FAC by doing it here. Dough4872 00:04, 22 July 2013 (UTC)
  • Oppose for a few reasons. Already, the requirements for a passing ACR are growing more complex and difficult to understand, and this will not help in that regard - and while the three reviews do come in a lot quicker than the four reviews, finding someone to do the spotcheck has become increasingly difficult already, without this proposal. Secondly, I'm concerned about this being used to support the belief that our FACs are being passed by inbred reviewing. I think that for something not so black and white as checking if sources are reliable, and that they are properly formatted, that having an outside reviewer for this benefits credibility; though behind the scenes I do check reliability at ACR, having this done from an outside reviewer gives us credibility. Finally, I'm just not seeing the need here - this isn't a spotcheck that is quite annoying and difficult to find someone to handle. If we're concerned about the time it takes to have our FACs pass, then we need to start reviewing other FACs, which the delegates nearly mandate - in fact, MILHIST strongly recommends that nominators review 2 FACs for every FAC submitted, and I think this is a practice that we should be considering as well. --Rschen7754 02:31, 22 July 2013 (UTC)
  • Oppose I have to agree with Rschen here. I would not be opposed to the FAC reviewing suggestion either, even if it's just a gentlemen's agreement. –Fredddie 02:36, 22 July 2013 (UTC)
  • Oppose per Rschen7754. Also, while we like to consider ACR a dry run for FA. They are different and we need to differentiate the two. A-Class does not equal "WP:HWY internal FA class". I also agree that we should strongly recommend (but not mandate) the 2 for 1 review guideline for an FA nominator. -- Nbound (talk) 02:41, 22 July 2013 (UTC)
  • Switch to oppose - Rschen7754 is correct. Although Nbound should note that most, if not all HWY ACR reviewers review based on the FA criteria. TCN7JM 02:45, 22 July 2013 (UTC)
While that is true, they are in general, a little more forgiving (one would hope) -- Nbound (talk) 03:01, 22 July 2013 (UTC)

Demoting articles

It seems a little odd to me that articles that are up for demotion will be closed as keep if noone can be bothered to review and/or maintain/improve them. I think this should be changed, but I am aware that others may want checks and balances to make sure that a change isnt abused by editors which dont have the best interests of the project at heart. (I think this is unlikely but the discussion is open to compromise). Similarly some may wonder what happens if articles miss out by one or two votes, and so on?


Thoughts? -- Nbound (talk) 03:15, 13 September 2013 (UTC)

This isn't that odd to me. The default position behind FAR(C) is to attempt to keep articles at FA status. There needs to be a consensus to move from the review (FAR) to removal (FARC) stages, so for ACR to mirror the requirement to have a consensus to demote only makes sense to me. Imzadi 1979  03:18, 13 September 2013 (UTC)
(edit conflict)Its more likely that a stale article will maintain a no longer deserved status here, as noone might be bothered to give it a second look. Rschen7754 has put Balt-Wash Pkwy up for discussion and noone else has looked at it. I trust him to have made a legitimate complaint in regards to its quality, but unless there are three demotion votes it will remain A-class for who knows how long, slowly but surely experiencing the quality rot that affects all long-standing articles even more than it has already. If anything this is more likely to spur on more input into the articles, and truly have them at the required standard, as people would be more likely to vote save an article. At the moment doing nothing is the easiest option and there is no consequences for it. -- Nbound (talk) 03:31, 13 September 2013 (UTC)
I'd have to think about this one a bit. This issue has become a concern with the increased number of demotion candidates, but I am not entirely sure that adopting a blanket demote policy won't get us into trouble further down the road. --Rschen7754 03:20, 13 September 2013 (UTC)
What issues are you worried about in particular? I cant think of anything too major, on the off chance an article was wrongly demoted it can always be repromoted back to via the standard method (or skip A altogether with a run for FA). The worst case as far as I can see is that we go back to the old rules if whatever is thought of here doesn't work out. -- Nbound (talk) 03:31, 13 September 2013 (UTC)
Well, for one, the standards between A and FA are slightly different, and I would hate to see an article removed from A-Class just because the author was not planning on taking it to FA. --Rschen7754 03:40, 13 September 2013 (UTC)
Sure, but as long as that editor or others are around to maintain the article at the A standard, there is no issue. Besides, generally an article isnt even suggested for demotion unless there is a clear failure in some regard. Its very unlikely that an existing A-class article is going to be demoted for something minor. -- Nbound (talk) 03:53, 13 September 2013 (UTC)

Something else to consider is that over time our A-class reviews have gotten better and better (compare: 2008 with 2013). Older articles that were given A-class earlier on, may not be up to the current A-class standard; and may never have been. -- Nbound (talk) 04:07, 13 September 2013 (UTC)

I agree that our reviews have gotten better (we no mandate image reviews and spot checks that weren't required before), but I don't know that the default should be to demote an article nominated for demotion. Like I said, FAR(C) defaults to keeping articles unless there is a consensus to delist. Our aim should be to do whatever is needed to retain the A-Class assessment on our articles. We should only revisit that stance if apathy sets in; I don't think we're at that state though, even if momentum as slowed at ACR. Imzadi 1979  04:36, 13 September 2013 (UTC)
We could demote by default only if noone other than the nom gives input. That rules out absolute apathy. -- Nbound (talk) 04:47, 13 September 2013 (UTC)
I can't support that either. If it takes 3 reviewers to promote, it should take multiple editors to demote. WP:WIAACA specifies that absent a formal ACR by a wikiproject, it takes two editors in agreement with a nominator to promote, and why should it be easier to remove the assessment than it was to give it? If an article truly sucks bad enough, it wouldn't be that hard to get multiple people to agree. If the process breaks down that much, it's best to shutter ACR in general and resort to the WP 1.0 defaults for promoting to, and demoting from, A-Class. Imzadi 1979  04:54, 13 September 2013 (UTC)

ACR for lists

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.


As many of you know, the US Roads project is working on sending some lists to FLC. We've never formalized how this was to be handled at ACR, so perhaps we should explicitly allow this, though reviewers should be examining the articles based on the featured list criteria. Because lists are easier to review, the reviews should be shorter. A spotcheck would not be required, as that is not very helpful for lists, and is not required at FLC.

Following a successful ACR nomination, the A-List class can be given to the article. If a national project does not want that classification, they can choose to opt out.

Thoughts? --Rschen7754 05:54, 21 March 2014 (UTC)

  • Sounds good to me. In general, our lists will be collections of data from the listed articles, and that data is rarely part of the spotchecked information in those articles, so I say skipping that step will be ok. That said, any reviewer at any time may make any reasonable request to check or challenge information in a nominated article, even if that's not part of a formal spotcheck. Imzadi 1979  05:59, 21 March 2014 (UTC)
  • I can only see good things coming from extending ACR to lists. –Fredddie 06:39, 21 March 2014 (UTC)
  • I like the concept, but I'm hesitant only because there hasn't been much hammered down in the way of standards for lists. I don't want to see the standards of the first promoted list (at ACR and/or FLC) becoming the defacto standards without due discourse. - Floydian τ ¢ 06:48, 21 March 2014 (UTC)
    I kinda have the same reservations. Maybe we could wait until 10 articles have gone through the process and through FLC before we start assigning the A Class level to lists. –Fredddie 06:53, 21 March 2014 (UTC)
    Well, even back in the early days of ACR (2007), we didn't really have a FA standard either. And though that did lead to some errant A-class ratings in the early years, I think that all of our current A-Class articles meet the standard. --Rschen7754 06:56, 21 March 2014 (UTC)
  • Seems reasonable so long as achieving A-class does not become a project-induced mandatory prerequisite to nominating a roads list for FLC. That being said, an ACR could double as the suggested peer review process prior to nominating a FLC. On an aside, though rather inactive on roads in WP as of late, I've been quite active and successful at FLCs for community lists, and have begun returning the favour by reviewing other FLCs about topics in my areas of interest. I look forward to participating in road-related FLC reviews as the nominations come in. Hwy43 (talk) 07:54, 21 March 2014 (UTC)
    It wouldn't be required; rather, highly recommended, just as sending any road article to ACR before FAC is recommended now. The main purpose of ACR is generally to get the opinions of other editors on whether or not it's ready for FAC, and it'd most likely be the same idea here. TCN7JM 12:41, 21 March 2014 (UTC)
  • It might still be a good idea to do a spotcheck, just for the prose sections, which should be at the same standard as A-class articles - accurately reflecting what the sources say without copying or paraphrasing too closely. It should be easier to do than an article spotcheck as there would be less content and fewer refs. - Evad37 [talk] 09:45, 21 March 2014 (UTC)
  • I support this proposal. TCN7JM 12:41, 21 March 2014 (UTC)
  • I generally support the proposal, however, I disagree with the idea of ditching the spot check. We should be conducting some sort of check to make sure the tables aren't filled with garbage data and that the author hasn't misinterpreted the source (which can be easy to do with some states' route logs). —Scott5114 [EXACT CHANGE ONLY] 22:16, 21 March 2014 (UTC)
  • I agree that ACRs for lists should look against the FL criteria much as we use the FA criteria for articles at ACR. I also think we should do image and source spotchecks at list ACR to look for possible issues. Dough4872 00:26, 22 March 2014 (UTC)
  • I support spot checks during the A-Class List Review process. I also support holding off on awarding the letter right away. We should wait until there have been 10 reviews or the end of the year, whichever comes first, before awarding the letter. In the meantime, any list that passes the review can be given Provisional A-Class List status. This would be a notice on the talk page that says this list qualifies as an A-Class List but will be awarded the letter at a later date. If any significant changes to the list review criteria occur after a list receives provisional A-Class status, that list would need to be brought up to the new standards before it is awarded the official letter.  V 12:26, 22 March 2014 (UTC)

Revised proposal

Okay, so based on the feedback:

  • This is something we want to do.
  • We should do a spotcheck to make sure that the sources for the tables are interpreted correctly, but this would be more based on facts than sources.
  • The first 10 successful reviews will only give provisional/temporary A-List status. At the end of the 10th review, we will do a general review of any lists remaining that have not passed FLC, and determine if the status should be kept.

Thoughts? --Rschen7754 02:03, 30 March 2014 (UTC)

I think we can go through with this. Dough4872 02:08, 30 March 2014 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

KMLs for AL-Class lists

I'm assuming we will use the AL-Class used by WP:MILHIST when the time comes. However, I'm curious how we will handle KMLs in relation to lists. It is possible to make a single KML that has multiple road segments in place. What concerns me is that a KML for List of Interstate Highways in Michigan might be feasible, but List of state trunklines in Michigan has so many more entries that a KML might not be a good idea. As it stands, KMLs are required, but that requirement was put in place two years ago about articles, not lists. Imzadi 1979  23:01, 9 April 2014 (UTC)

It is possible to make a single KML... But file size is limited, which could be a problem for some lists. I tried to combine the seven KML files listed at Highway 1 (Australia), and the resulting file was 2055kB, too big for Wikipedia to save. - Evad37 [talk] 23:30, 9 April 2014 (UTC)
This was an oversight, and I don't think we should require KML for lists. --Rschen7754 23:33, 9 April 2014 (UTC)
I agree, list articles do not need KMLs as they can be found in the individual route articles. Dough4872 00:10, 10 April 2014 (UTC)
Nice to have, shouldn't be required. –Fredddie 00:15, 10 April 2014 (UTC)

AL-Class

Just an update, but AL-Class has been implemented in the banners for HWY and all of the subprojects. The assessment categories need to be created at some point however, but otherwise the banners are ready to handle it. Imzadi 1979  06:54, 12 April 2014 (UTC)

Spotchecking

What is our procedure for spotchecking? There seems to be a bit of disagreement over this, ranging from "just 10 sources" to "25% with a maximum of 20". Lately, the spotchecks have been holding reviews up, because there's very few people willing to do spotchecks lately since they're so time-consuming. What do we want to do to solve this? --Rschen7754 04:14, 3 July 2014 (UTC)

I like doing a certain percentage with a maximum set number but I feel 20 might be too high. I say we combine the two ideas and do "25% with a maximum of 10." The maximum of 10 will be used for articles with 40 or more sources while those with less will follow the 25% rule. Also, spotchecks may be hard to do when a lot of sources are either print sources that only the nominator has easy access to or paywalled articles. If a user opts to do a spotcheck, they should be able to have enough sources that they can easily access without having to burden the nominator to provide a copy or access to the sources. Dough4872 04:29, 3 July 2014 (UTC)
Well, any nominator should be willing to e-mail sources or upload them into a Dropbox-type account (or similar) so that the reviewer can access a copy for the spotcheck. In theory this will keep the nominator honest and prevent fabrication of offline sources. (Several FAs were delisted a while back when it was discovered that the nominator was just making up sources.) Imzadi 1979  04:42, 3 July 2014 (UTC)
WT:Reliable sources/Noticeboard/Archive 4#Legolas2186 possibly falsifying references is what I was thinking about earlier; I have no reason to believe such a thing would ever happen in our projects, but it is possible. The Grace Sherwood affair is why FAC implemented the spotcheck requirement, and that situation with fabricated sources only cemented the need to check periodically. Imzadi 1979  06:05, 3 July 2014 (UTC)
For FAC spotchecks, how many sources are they checking nowadays? --Rschen7754 09:16, 10 July 2014 (UTC)
What would people say to the following: we aim for 25% or 10 sources, whichever is lower. However, the reviewer has the right to check more at their discretion (first nomination by a nominator, bad track record of nominator...), and the nominator has the right to ask for a more thorough check if they think that one will be needed at FAC. Thoughts? --Rschen7754 06:01, 14 July 2014 (UTC)
I think that is a good idea. Dough4872 00:12, 15 July 2014 (UTC)
I like the gist of what you're saying, but I think it's wordy. How about "...at the very minimum, we aim for 25% or 10 sources, whichever is lower. At the nominator's discretion, a more thorough check may be requested." –Fredddie 12:56, 16 July 2014 (UTC)

Is this of any use to anyone? Nowadays Echo pings seem to work out well enough... --Rschen7754 00:31, 18 July 2014 (UTC)

As I mentioned on IRC, I found it difficult to use. Logically speaking, some of the parameters weren't where I thought they should be. I would be OK with decommissioning. –Fredddie 00:33, 18 July 2014 (UTC)
Mind as well get rid of it, I don't use it when I review articles as I can easily ping the user or they probably have it on their watchlist. Dough4872 03:34, 18 July 2014 (UTC)

Discussion on A-class at Village pump (policy)

You are invited to join the discussion at Wikipedia:Village pump (policy)#A-Class_article_assessment. – discussion on whether to abolish, modify, or segregate A-class out of the WP1.0 quality scale. Thanks. Evad37 [talk] 04:30, 24 July 2015 (UTC)


There have been some proposed changes to the A-Class review process at the discussion above. --Rschen7754 20:50, 5 November 2017 (UTC)