Wikipedia talk:Good article reassessment

This page, a part of the Good article talk page collection, is archived by MiszaBot II. If your discussion was mistakenly archived feel free to go retrieve it. Current Archive location: Wikipedia talk:Good article reassessment/archive3

Archives
2006 2007: January - May • June - Present

A new way of doing things aka How we can fix everything

So alot of people have been complaining about really old GARs and the weird limbo some articles get in because improvements have been made to the article half way through. This seems like a valid concern to me, and seems to muddle up any kind of efficiency we can offer here. So, here are my ideas, completely copied from other people and posted here as if I thought of them myself:

• Have the ability to "reset" a GAR once certain criteria have been met (length of reassessment, lack of consensus, etc).
• Reassessments are only of oldid's of articles. That way the discussion is over one incarnation of the article, not spread out amongst any number of subsequent edits. If improvements are made to the article (as they hopefully should be), the GAR can be reset whenever the article is believed to be up to GA criteria and reassessed based on the new id.

The latter suggestions in particular, seems like it would speed things up. I imagine these things would rack up a consensus much faster if the article under review is a single version - any subsequent discussion of the article would be focused on interpretation of criteria and/or its application to the article, not an ongoing to-do list. In conjunction, I believe the GAR process would be vastly improved. Drewcifer 05:14, 6 October 2007 (UTC)

• Or something. This idea has a relatively spring-fresh odor.
• I believe something might be done to deal with articles at regular, preset intervals every week or ten days rather than giving each article a week on its own without comments etc.
• I mean, swing through and clean house on every article that has been around since before the previous housecleaning. If a discussion is ongoing & many changes have been made, and so it needs to be "reset" then do so... Articles fall between the cracks because they stand alone (and they are not the 101st Airborne Division, for those who like Band of Brothers). Let's herd those little doggies! (... and by "let's" I mean "why don't you guys"; I am still nose-deep in RL) --Ling.Nut 06:42, 6 October 2007 (UTC)

I feel that setting the reassessment to be of only one oldid would just inflame drama on GA/R. Lack of consensus could always be archived as lack of consensus. -Malkinann 09:33, 6 October 2007 (UTC)

I absolutely agree with assessing only one oldid - this is not an article improvement drive, if somebody wants to improve the article to the GA status, please ask the reviewers who did GA/R for guidance and once everything is said and done, resubmit as GAN. I feel it is too time-consuming to participate in GA/Rs with moving targets, there is no need to please everybody, this is a yes/no decision process, not kindergarden. PrinceGloria 09:33, 9 October 2007 (UTC)

I think it's a good idea, but I think it should only be done if the requirement to send an article to GAR is that a notice must first be placed on the talk page with a time of one week for improvements to be made. If the article has not been sufficiently improved at that time, the article is nomed at GAR with the current oldid. Bust out recommendations and archive after either so many recommendations or a certain amount of time, say one week, whichever comes first. Like PG said, GAR isn't an improvement drive. If it's minor issues, we'll fix them like we always do, keep, archive. Otherwise, move it along. Lara❤Love 13:04, 9 October 2007 (UTC)

Why so? If the article does not meet WIAGA, it is improperly listed as one and should be promptly removed. Actually, the removal should take place without involving GA/R, the GA process is about every user being able to list OR delist an article. Only in controversial cases should GA/R be invovled, and then they can assess whether a contested listing or delisting was appropriate, which by itself refers to an oldid situation, not the later changes. If the article was, say, delisted apropriately BUT improved afterwards, the conclusion form GA/R MIGHT be that the editors are encouraged to nominate the newer version for GA again. PrinceGloria 13:14, 9 October 2007 (UTC)

I really really like Lara's idea above. Requiring that specific concerns with an an article be raised on an article's talk page first would dramatically reduce the problem of reassessing an ongoing work in progress. Besides, discussions of how to improve an article is what a Talk page is for in the first place. I imagine requiring this step would keep alot of articles from being raised at GA/R in the first place. So, here'e my idea of the basic timeline of a GA/R:

1. An editor expresses his/her concerns with an article and it's qualification for GA status on the article's talk page. If, after X number of days the article has not been improved or has not shown signs of improvement, the editor may either
   1. delist the article themselves in the most clear-cut cases or
   2. nominate the article for GA/R in any potentially controversial circumstances, using an oldid
2. The article is reviewed based on a single oldid version, and if a consensus of X votes is reached close and archive the GA/R as Delist/Keep/Etc.
3. If substantial improvements have been made to the article during the GA/R, before a consensus is reached, and to the point of addressing all undisputed concerns raised, editors may reset the GA/R by archiving the old review and starting a new one with a link to the previous GA/R and a new oldid.
4. Repeat until a consensus is reached.

Drewcifer 19:57, 9 October 2007 (UTC)

Doesn't that turn GA/R into even more of a Peer Review process than a process designed to maintain the integrity of GA status? Also, improving articles as the result of a GA/R is fine, even preferred, but what difference would spending even more time waiting for someone to fix problems make when many articles already sit on the page for weeks, even over a month? Furthermore, that idea about waiting to file a GA/R or delist returns us back to effectively deleting speedy delists (Again) which is critical to quickly removing blatant non-GA's, and I imagine is super-critical to the current sweep in progress. (I see a whole crowd of speedied articles on the log now directly due to the sweep) Should people doing sweeps be expected to just keep adding their removal candidates to a huge personal list of articles, waiting until the day finally comes when they can delist almost all of them? (I seriously doubt many articles have active editors ready to maintain article integrity all of the time) Finally, this idea of only reviewing a certain oldid will result in very embarassing situations where an article in the middle of review suddenly contains some absolutly terrible violation of the criteria. (As is likely with certain extremely controversial topics, and possible with any article) Also, to address something you said earlier, there's nothing stopping people from resetting a GA/R right now, I just did a few days ago in what I consider to be entirely reasonable circumstances concerning a church. If we want to stop GA/R's from being so increadibly long-lasting, lets formalize the suggestion of myself and Geometry Guy below about a hard-cap on the length for GA/R's. GA/R should be reassessing the status of an article first and foremost, not trying whatever it can to improve them, that's what the (currently languishing) GA collaboration should be for. Homestarmy 21:21, 9 October 2007 (UTC)

I would argue that such a process removes any hint of a peer review: what GA/R should really be is "does this article match the GA critieria? Yes or No?" Instead these things typically devolve into an ongoing to-do list, which ends up being very difficult and annoying to sort out. All of that should be reserved for the Talk page, not here. As for the first step in my above timeline, I suggested that more as a formal way of saying something that is already being done pretty informally: the decision to delist on the spot or nominate at GA/R. My point is that if an article DOES substantiate a GA/R, then the issues should ideally be brought up on the Talk Page initially, rather than going straight to GA/R, to give the article's custodians a change to improve the article before it is reassessed. This already happens quite alot. I did not mean to undercut so-called "speedy" delists or sweeps, since those don't really apply to GA/R.

I'm not sure if I understand your point about the embarrasing situation. I presume that you mean if an article gets a certain number of "Keep" or "Relist" votes based on a particular oldid, and something new pops up in a new version that might blatantly violate criteria, that this might render the old votes based on an oldid incorrect or invalid? But the same scenerio could happen now, so my propposed changes don't change that fact. And besides, that might be a case where a GAR should be reset, as something has substantially changed. Besides, how often does that particular situation come up versus the reverse where someone votes delist, the article is improved, and the voter doesn't change their vote. The solution to both situations is the same: reset the article based on the new version (the new oldid).

My suggestion for restarting GA/Rs is based on what you and G-guy have been doing, so kudos. I'd just like to see it become a useful tool in all of our GA/R toolsets, rather than something that happens once and a while at some editors discretion.

And, finally, I think a hard-cap time limit is a useful solution, but not perfect. Since a GA/R has gone on for X days doesn't mean that the review is invalid in some way. We should find a way to speed the process up, rather than impose a time limit. I believe my suggestions would do just that: simplify things into a very quick "yes" or "no" vote. Drewcifer 21:51, 9 October 2007 (UTC)

Adding on more pauses to the process makes things seem more like a PR to me because its as if the system is trying everything it can to give other editors encouragement and time to raise an article to GA status, and then only delist it as an absolute last resort after several weeks of mostly pointless waiting. Yes, people on GA/R's do tend to give lists of problems and propose that the problems be solved, but giving specific concerns about an article makes a decision look stronger. This can be especially important in reviews where one group is determined for an article's status to be changed one way or another in a way that isn't really correct. Sure, sometimes people come in and fix the problems people list, and then note the problems as compleated like a PR. But if people want to fix problems that editors have identified with an article in the hopes of retaining GA status, that's their perogative. Your suggestions sounds as if an editor must bring up problems on an article's talk page before doing anything, GA/R or delisting immedietly. The point of delisting immedietly is so that it can be done immedietly, not after a week, for egregrious violations of the GA criteria. Also, there's no rule that says you have to use GA/R immedietly when you think an article shouldn't be a GA, anyone could just drop a comment about how they think an article doesn't meet the GA criteria and offer suggestions for improvement.

Nextly, you're right about what the situation i'm referring to would entail. And yes, it does happen right now. That's why I assume we're talking about ways to reform GA/R, since that seems to be the principle problem in many old GA/R's. But the GA system doesn't promote a certain oldid of an article to GA status, it promotes just an article in general, and a Wikipedia article isn't supposed to be static. It also would GA status mean less and less the longer away the initial GA review was, since the current article (And the version that the world will see) will not be representative of what it was like when it was given GA status. Of course, not doing anything does mean GA/R's have to be reset or cancelled eventually, as I did with that one church article. The "Bringin on the Heartbreak" GA/R at the bottom of the page might be a candidate for resetting, since the nom changed his/her assessment which is dated back to August, but other editors have not followed suit, and Jayron and VanTucky still say it doesn't meet the criteria. So yes, this situation does seem to be happening more and more often.

When I restarted one GA/R and had to reccomend another be restarted, it wasn't something that I felt reflected strongly on the GA/R process, since it essentially meant that the review as it stood was a total failure. I think it would be best if all GA/R's would work the first time, and that means something needs to be done to prevent them from going on so long that article content changes too drastically for older comments to be accurate. (Unless, of course, the article actually is unstable, and fails the GA criteria) Having a time-cap would be similar to the AfD process, though of course, our cap should probably be much longer if extant at all since we have lower participation than AfD. At the very least, it should eventually get down to a point where, in any given GA/R, there isn't enough time for most comments to get out of date, and if they do get out of date, the article's stability is probably either questionable, or a very poor article is dramatically improved to GA status due to a GA/R. Homestarmy 16:39, 10 October 2007 (UTC)

I agree with most of your points. But I think you're getting caught up on alot of the wrong parts of what I'm proposing. I too think that we should find some way to speed things up, that is the whole point of my suggestions. Enforcing a time limit on a GAR accomplishes this, but it imposes a restriction on the GAR which completely ignores the actual status of the article. I could imagine many situations were valid points are brought up in a GAR, but the time limit is reached and the GAR is closed despite these valid points - simply because there was not enough activity in the GAR.

Instead, I am proposing something which, I would argue, would speed things up even more, while still being based completely on the article: reviewing an oldid (which basically boils things down to a "yes" or "no" vote by removing the potential of an ongoing to do list, a moving target if you will) while encouraging possible resets of the GAR if all valid concerns have been adressed (leading to another simple "yes" or "no" vote). Basically, my goal is to simplify the way people participate in a GAR. Simplifying the process will in turn speed it up. As separate processes (oldid and resetting) neither suggestion would improve things that much if at all, but together I believe they would help considerably.

As for the Talk page notification, this is not something I'm dead set on, nor do I think it should be a requirement. It is just a common courtesy to the article's custodians: give them a chance to fix things before putting its GA status into serious jeopardy. Perhaps this could be something that is recommended but not enforced, and perhaps the time-limit things (X number of days then nominate for GAR) as well - neither are that important to me. The main benefit of this additional process would be to keep some article's from reaching GAR in the first place, since the article's custodians would hopefully address any concerns before nomination is even necessary.Drewcifer 01:11, 11 October 2007 (UTC)

So yea, about that new speedy archive criterion...

Why a week instead of three days? And when was it changed anyway? What was wrong with three days? What was wrong with around 80 percent? If this happened the week I was in Michigan, i'd sure like to see a link to this discussion, but as far as I can tell, this was changed with basically no real point. The entire purpose of being able to archive a GA/R speedily is so that it will be, you know, speedy, especially for blatantly obvious cases. A blatantly obvious case will not require an entire week to figure out. Homestarmy 17:10, 7 October 2007 (UTC)

I agree with comments made by PrinceGloria and others that we shouldn't give precise figures in terms of numbers of days or percentages, but rely on the discretion of the archivist. The more clear cut the reviews, the faster the archiving can be. I'd be happy to shorten the archiving guidelines along these lines. Geometry guy 18:26, 7 October 2007 (UTC)

But if its left vauge like that, it only takes one archivist gone awry to create a mess by archiving disputes like a day after their filed, long before editors who have a useful opinion can see it. Homestarmy 19:19, 7 October 2007 (UTC)

I am not aware of this ever happening. There is no need to invent solutions without a problem. In my view, the most important line in the current archiving guidelines is "Closing a discussion requires taking responsibility, determining what the consensus of the reviewers is, and taking action where necessary." GAR archivists all seem to take this responsibility quite seriously. Geometry guy 19:26, 7 October 2007 (UTC)

What is blatantly obvious to one reviewer will not always be blatantly obvious to another, or more importantly, blatantly obvious to likely contributors to GA/R's. That's why I think there needs to be some kind of limit, just not a week long one. Unlike the majority of proposed regulations to GA that were searching for an extremely unlikely problem to solve, the failure of human subjectivity to please everyone is not so unlikely. Plus, what you propose hasn't even been implemented, so of course there aren't problems yet. Homestarmy 19:31, 7 October 2007 (UTC)

Okay. In my view, no GAR should be archived sooner than 7 days after it starts because the process should include reviewers who only check the page once a week. However, what really should be changed is the "do not archive until there have been 7 days since the last comment". This should depend on the consensus. If there is no dissent, archiving can happen the next day. If there is a minority of dissent, then it is best to wait a couple of days before archiving. If the discussion is highly contested, then a week might really be needed. This is what I mean by "The more clear cut the reviews, the faster the archiving can be." Do you agree with this principle? Geometry guy 19:40, 7 October 2007 (UTC)

That sounds fair, and might help with the higher activity on GA/R often resulting in reviews lasting too long. But then, what happens with reviews where there is not nearly a unanimous consensus, but there is clearly a majority for a certain option? Such a review could last forever if people continue discussing it, since it couldn't be closed as just no consensus if its around 75 percent to do something. That sort of thing is why I think there needs to be a set-in-stone hard cap on the length of time for GA/R's, because most of the time, there's no point in waiting around for more changes or more opinions after a certain long period of time. Homestarmy 20:09, 7 October 2007 (UTC)

Given that I support a minimum time of 1 week for a GAR, it seems reasonable to me that there should also be a maximum time, say 5 weeks, after which a no consensus archive is acceptable without waiting for further comments. Is this the kind of thing you had in mind? Is there general support for such a proposal? Geometry guy 20:21, 7 October 2007 (UTC)

Well, I was thinking 4 weeks when I proposed it in a section above, but you didn't seem to like it since it would of required adding another rule. Homestarmy 20:23, 7 October 2007 (UTC)

I'm not dogmatic. I think a combination of making the guidelines more flexible and adding a time limit might work quite well. Although it involves adding a rule, other rules would be shortened. Geometry guy 20:28, 7 October 2007 (UTC)

Then, if there's no disagreement about these ideas in principle, it seems to me we only need to work out what the exact numbers should be. Homestarmy 20:33, 7 October 2007 (UTC)

As I have said in previous discussions, I am against enumerating everything, but agree that a maximum time limit for GARs might be helpful. So there is only one number to fix from my point of view. I suggested 5 weeks: just over a month and easy to count. Does that suit? Geometry guy 20:45, 7 October 2007 (UTC)

I thought a month was 6 weeks long, but even though I think 5 is kind of high, I guess its fair to start with to see how it goes. Homestarmy 20:47, 7 October 2007 (UTC)

February is 4 weeks (28 days), other months are four and a half (30-31 days). But yes, lets see how 5 weeks works in practise. Geometry guy 20:52, 7 October 2007 (UTC)

Disputed review over Marth (Fire Emblem)

Hello there. I and User: Hbdragon88 have discussed this issue briefly at each others' talk pages. Basically, the uncertainty is over the "broad coverage" GA criteria which was cited as a reason to fail Marth (Fire Emblem). The article is missing a Conception and creation section as that information isn't really attainable. I feel that this omission doesn't stop the article from being broad, it just stops the article from being comprehensive—a difference between GA and FA assessment. Really, I'm asking whether this is a justifiable reason to fail GA and whether this omission really stops the article from being broad in coverage. On another note, this user thinks every video gaming article must have a development section to reach GA status. So, I'd appreciate it if somebody would clear this up for us, Thanks. Ashnard _Talk ^Contribs 15:10, 10 October 2007 (UTC)

First, I would consider opening up a full request for this on the main page, instead of bringing it to talk (which is more discussion on how articles are reassessed).

But specifically, I see a lack of broadness as defined by notability. There is very little information about the character that is not directly related to his appearance in games and the events within those games. It's not 100% in-universe, but there is a significant lack of out-of-universe information that would be absolutely needed just to meet WP:N, much less GA or FA. And I do understand that you are probably right that you'll have problems finding the creation information ( at least from English sources); unfortunately, this makes it very hard to justify why this article should stand alone.

My suggestion: follow the example of Characters of Final Fantasy VIII and merge all the Fire Emblem characters to a single article. You are likely going to have an easier time to find creation and other out-of-universe information for these characters than for one single character from the series, which lacks a strong English presence and thus is going to be lacking a large number of English references. That list itself may then be worthy of the GA. --MASEM 15:17, 10 October 2007 (UTC)

See List of characters in Fire Emblem: Path of Radiance. It failed FL; there isn't any out-of-universe info that I could find. Amazingly, FE DS has just been announced!!! It's a remake of the first Fire Emblem and will feature Marth as the main character. I know it's still in-universe info, but could this in any way aid the cause? Doh! It fails stability now anyway. Thanks for the comment. Ashnard _Talk ^Contribs 15:24, 10 October 2007 (UTC)

I'd like to list a number of examples that I discussed with Ashnard. Super Princess Peach's GA status was revoked in part due to a lack of development section; I nominated this for GAR once I realized that other GA video game articles listed at WP:CVG/GA did have dev sections. The reviewer for Big Rigs cited the need for explaining why they made such a bad game, which I could not provide. Character articles Aerith Gainsborough and Princess Peach and Sephiroth (Final Fantasy) (which has an empty "concept and creation" section, heh) are all former GAs that have no deve sections. hbdragon88 22:35, 10 October 2007 (UTC)

Yeah, that factor was unanswered. Is it absolutely essential for an article to have a development section? One thing I'd like to add—without offence to Hbdragon88—is that Big Rigs also had other factors that led the reviewer to fail it, not just the lack of development. Ashnard _Talk ^Contribs 15:36, 11 October 2007 (UTC)

Note how this time I didn't say it was the only factor; I think I said before that it was the only reason, but it's not. The important part is that the "broad in its coverage" was cited as "needs improvement" because of the lack of a dev section, which was a contributing factor to the failure of theG GA nom. hbdragon88 23:00, 12 October 2007 (UTC)

A number of articles just appeared in CAT:GA

A number of articles just appeared in CAT:GA without a listing on WP:GA. Doesn't look like these were indepedently reviewed. Gimmetrow 02:53, 14 October 2007 (UTC)

• Abdel Basset Ali al-Megrahi
• Alternative theories of the bombing of Pan Am Flight 103
• Chester Crocker
• Civil Cooperation Bureau
• Craig Murray
• Frederik Willem de Klerk
• Investigation into the bombing of Pan Am Flight 103
• Pan Am Flight 103
• Pan Am Flight 103 bombing trial

See Phase4's recent contribs, Special:Contributions/Phase4. He's been listing articles unilaterally. T Rex | talk 02:58, 14 October 2007 (UTC)

Alright, I'm off to ask him to stop doing this. Should we put these articles on GA hold or just bold delist? Many articles have problems that can't be fixed quickly. OhanaUnited^{Talk page} 03:24, 14 October 2007 (UTC)

Delisted them all. Lara❤Love 03:56, 14 October 2007 (UTC)

Now a bunch of articles have been delisted without removal from WP:GA: Edmund Burke, Iglesia ni Cristo, Organic food, Pasta, Zagreb, Chicago, Dallas, Texas, Edmonton.

New articles in CAT:GA not listed on WP:GA: Casino Royale (1967 film), Castlevania: Dawn of Sorrow, Hurricane Nina (1957), Joey Santiago, Temple of the Dog Gimmetrow 21:51, 26 October 2007 (UTC)

GAR tag

What happened to the tag for GAR articles that went on the talk page? I don't find it in the instructions. Sumoeagle179 20:22, 28 October 2007 (UTC)

I'm not sure if there ever was a template, thought it sounds like a familiar template...usage would be very helpful for automating the updating of ArticleHistory events. Homestarmy 20:40, 28 October 2007 (UTC)

There was something that said something like "It's GA, but a user has expressed a concern..." as I recall. — Rlevse • Talk • 18:55, 29 October 2007 (UTC)

Now that you mention it, that sounds like a great idea. If that template exists let's start using it, and if not let's make it happen! Drewcifer 19:53, 29 October 2007 (UTC)

GA citation debate

I believe I am one of the most productive WP:GA authors. At last count I had 38 current GA credits plus two that have been promoted to WP:FA. For some time, I have constantly had WP:GACs in the queue for review. In the month of October at one time I had 9 GACs, as shown here, in the in the queue. In October, I had at least 7 GA promotions: including Rush Street (Chicago), Ricky Powers, Washington Park, Chicago (neighborhood), Chicago Marathon, Harold Washington Cultural Center, Haystacks (Monet), Prairie Avenue, and Rainbow/PUSH.

Last month I posted Gilbert Perreault at WP:GAR (see Wikipedia:Good_article_reassessment/Archive_31#Gilbert_Perreault) because I felt it was being destroyed by another editor who was removing citations. The GAR was headed to a consensus to keep the article with the citations replaced. Then, User:Geometry guy closed the discussion as inappropriate for GAR and suggested I take it to WP:RFC.

You may recall that I had first taken the article to the talk pages of both WP:HOCKEY and WP:WPBIO without reply and then requested help at WP:PR after much back and forth editing and arguing.

Much to my surprise the RFC is headed toward a consensus to allow the removal of my citations and essentially authorizing people to remove citations at will if they are in the mood as has been pointed out in the debate. However, it is my opinion that I am being baited in the debate especially by User:Djasso who is making it appear that I am doing some unscrupulous. My most serious concern is that the way the debate is going, it seems to authorize citation removal from WP:GA beyond what I believe is appropriate and I feel I have a good sense of what is appropriate for a good article. Furthermore, the way debate is going it appears I will be handcuffed to sit by. I am very fearful that the persons I am debating with intend to tear the citations out of the GAs I have contributed and sort of need some backup so it does not happen. I had been thinking my work at finding citations was valued, but I am not so sure.

I am hoping that others here are concerned about protecting good articles will help reverse the debate back in line with the direction it was headed while at GAR. Please see Talk:Gilbert_Perreault#RFC.--TonyTheTiger (t/c/bio/tcfkaWCDbwincowtchatlotpsoplrttaDCLaM) 08:11, 5 November 2007 (UTC)