Wikipedia talk:Good articles/Archive 6

Where is all of this going?

Reading the TFD vote on the {{goodarticle}} badge made me question the entire purpose of this project. It seems useful for preparing a CD release and giving users a pat on the back for writing a medium-quality article, but I no longer think that it serves as a real intermediary to the featured article process. In fact, from the description on the project page, it seems like it is intended for many articles which supposedly can never become good articles. I have no problem with the template spamming and self-promotion which many users allege against this project, but can someone explain to me why this process is preferable to simply just opening a peer review before submitting an article to be featured. It seems like the fundamental difference between the two is that (and this has been my experience for the few articles that I did not promote) any suggestion for improvement is taken much more hostilely on this project because users percieve that they are being denied some kind of award, whereas such suggestions are usually met with thanks during a peer review because users are glad to get feedback there. This is of course not universal, but looking at the contributions list of any GA regular makes it appear that most articles are rated in between 15-30 seconds. This is great, if our objective is to locate articles with a table of contents, references section, and aescetically pleasing appearance while using the "page down" key repeatedly, but it doesn't seem like it is conducive to much feedback.

One suggestion seems to be to add more process. That would certainly address the objection that some good articles are crap, but its not clear why we should be spending even more time on process for this. If the objective is just to give users a pat on the back, then perhaps a barnstar or a nice talkpage comment would be better. I won't be promoting any more articles for a while, but I will be involved in discussions on where this project is going. savidan^{(talk) (e@)} 00:04, 27 June 2006 (UTC)

Well it's never served as an intermediary between non-FA and FA, it's not big enough yet :/. People who pass articles are as I understand generally expected to try and offer a little advice on articles advocating changes and the like, though not all nominators do this, I think most do however. Homestarmy 02:39, 27 June 2006 (UTC)

Out of interest, I took a look at the last article I could find that Homestarmy promoted, Crohn's disease, and he took at least 7 minutes to review it and possibly anywhere up to 10 minutes. He is not alone, I also find I spend a good 5 to 10 minutes reviewing articles, sometimes making minor alterations. Furthermore I rarely find people react in a negative manner to the good article feedback (at worst, they tend to just ignore it). The truth is this process helps to identify good articles and helps identify flaws in articles that are not as good. Beyond the feedback for articles that don't make the cut, identifying good articles might be helpful if Wikipedia was ever releases a published version however it can also be helpful to casual readers who want to find the good articles in their subject area. There are many articles, such as highly technical articles that tend to go astray in the featured article process. There are few (and probably no) featured articles that are at the same mathematical level as the good articles on homotopy groups of spheres or the Riemann hypothesis. Articles like Fort de Chartres might become bloated if they went through the featured article process. And finally then there are articles where the author's work is good but he is just not up to going through the painful featured article process. The good articles project can help in all these situations. So don't think of it as a step in the featured article process (which is not anything special - it was really just a process started one day by a Wikipedian), but as part of the identification of good articles. It might also interest you to know that for the 1190 good articles there are around 220 delisted good articles (18%) and at least 370 failed good articles (31%). So far from being a rubber stamp the good articles process is making a good attempt at identifying good articles. Plus with the recently added list it is now much harder for mediocre articles to slip through the good article process. Cedars 06:32, 27 June 2006 (UTC)

If we focus our attention to where it is supposed to be, which is the reader, well they need to know if an article has been given a standardized review and if so, they need to know it just to make the article meet the WP's policy. I normally assess articles in about 5-25 minutes which is a considerable amount of time figuring that I read it thoroughly at least 3 times and give a point-by-point assessment of what is missing to obtain GA status or if it reaches it, then I give more advices into what to work on to achieve a better quality/status. People will continue to see GA as a copy of PR which it is not because we rate the articles and we help the articles meet WP's policy & guidelines. It is not a first step to FA as of now because the community doesn't like that there will be more steps in order to achieve FA which I understand completely though if it doesn't meet FA it doesn't mean that it was good enough for GA and by keeping that in mind, the editors may have more to handle to meet FA than GA to begin with but after they meet the GA standard, the GA process will help them reach the FA status. Lincher 12:11, 27 June 2006 (UTC)

I understand all of these objectives and I don't want to be too hard on the GA process, but it seems like there is some inherent tension between them. It seems like a substantial portion of the GA movement is an offshoot of the portion of users who have some beef with the featured article process. I'm not comfortable with accepting (although there there is much truth to it) that certain articles can never become featured or, worse, that the featured article process would make some articles worse. If thats the case then we need to improve that process rather than attempt to circumvent it with another distinction with different (lower?) standards. Even if reviewers are spending 10 minutes + per article, how thorough can we really claim this review process is? That's what I'm worried about the most, is that some Good Articles are perhaps better than Featured Articles, but some may contain some glaring errors that just haven't been found yet. savidan^{(talk) (e@)} 19:26, 30 June 2006 (UTC)

Delisting change

I propose to change the original text:

Wikipedia:Good articles is an unbureaucratic system to arrive at a quick consensus set of good articles: everyone can nominate good articles, and everyone has veto power. However, sometimes editors disagree whether an article reaches the good article criteria. This page is for dealing with such disputes.

to:

Wikipedia:Good articles is an unbureaucratic system to arrive at a quick consensus set of good articles: everyone can nominate good articles, and everyone has veto power. When editors disagree whether an article reaches the good article criteria, discussions to find consensus on the status are listed here, and consensus reached at this page cannot be unilateraly vetoed, but requires new discussion at this page.

-- Kim van der Linde ^{at venus} 15:17, 27 June 2006 (UTC)

• Sounds fair enough to me. Homestarmy 15:26, 27 June 2006 (UTC)

• Sounds good. What we've been seeing too much of is POV warriors unilaterally delisting articles on all subjects that are even slightly controversial. --Cyde↔Weys 15:49, 27 June 2006 (UTC)

• Oppose. This proposal would invalidate the veto power entirely by allowing editors in favour of an article to simply relist it after ignoring anything discussed on disputes, as has been tried on two separate occasions in the case of the JP Cartoons. It's yet another attempt to facilitate the imposition of a tyranny of the majority. — JEREMY 15:54, 27 June 2006 (UTC)

No, it would allow editors in favor of the article to relist it after gathering consensus, people can't just ignore the dispute, if consensus isn't reached than any relisting would count as trying to sneak around the process. The majority is sometimes right. Besides, this project isn't all that anyway in terms of importance to the Wiki, what's the big deal? Homestarmy 15:58, 27 June 2006 (UTC)

If you call consensus tyranny of the majority, than the standard policy of wikipedia is tyranny of the majority. As far as I can tell, consensus has still a higher priority than objections of an individual unless that violates the key policies of wikipedia. -- Kim van der Linde ^{at venus} 16:44, 4 July 2006 (UTC)

• 100% support, Wikipedia is all about consensus, let GA be in accord with that. Netscott 15:56, 27 June 2006 (UTC)

• Oppose. Neutral at best If there is a slight reason why some user would use his veto power to remove it from being GA then the article should comply to this reason or he will not be allowed GA status. Lincher 15:57, 27 June 2006 (UTC)

I like this because it allows for disputed delistings to be addressed by a wider group. Only suggestion is that the new discussion has a time period before it can be relisted for delisting. Gnangarra 16:01, 27 June 2006 (UTC)

Timeframes sound reasonable. When I think of things like AfD and CfD seven days comes to mind. Netscott 16:07, 27 June 2006 (UTC)

Lincher, please reconsider. Vetos as you explained them would surely cripple the project from handling any politically-sensitive topic. If a concern is reasonable it will be supported by the consensus as is often the case for featured article candidates where one user votes in support of another user's suggestions. Cedars 01:21, 28 June 2006 (UTC)

The proposed change does not make it impossible to delist an article, to the contrary, that mechanism still remains. But it does make it impossible to unilateraly delist an article after extended discussion and consensus building. -- Kim van der Linde ^{at venus} 16:44, 4 July 2006 (UTC)

• Oppose. Instead of overruling the concerns by voting, the problems should be addressed. This would indeed improve the articles, instead of lowering the Good Article standards. Raphael1 23:52, 27 June 2006 (UTC)

• Support. This was how I expected the disputes forum would work in the first place. Notice even the first text does not suggest a single user has veto power when it comes to the disputes page. That notion is a fallacy. Please if you support this project, support this modest proposal so we can put these events behind us. Cedars 01:15, 28 June 2006 (UTC)

Though I assume that even if when the disputes end and the article retains its GA status, it would be changed to reflect all the changes asked by all the people that have commented in the discussions or else, the process is a bit counter-veto and also counter-GA in the fact that when we want like NPOV for example, if an editor disagrees then it means the article is still POV and even if a consensus or a 100-1 vote is done, there is still POV in the article thus it shouldn't become a GA. Lincher 01:44, 28 June 2006 (UTC)

And if the article is not POV, and all instances cited to the contrary have been dismissed by the 100, what do we do then? Articles certainly can't be fixed in situations where the dispute is baseless. Homestarmy 01:49, 28 June 2006 (UTC)

If you have the time Lincher, you might like to read this article which describes some of the dangers of objectivity. It touches on two points: Firstly it restates the widely acknowledged notion in academia that there is no such thing as objectivity (as the article states "there is no objective history"). Secondly, it points how the NPOV policy leads to weak writing or what it terms "waffling". The truth is articles should not aspire to a unanimously accepted neutral point of view - that's more than likely impossible - they should aspire to a standard of neutrality that is accepted by the majority. This is a good article process, 1% dissent should not block an article from acheiving good article status - that wouldn't even stop an article from acheiving featured status. Cedars 09:28, 28 June 2006 (UTC)

A 100-1 vote doesn't necessarily mean there is POV in an article. It can mean the article is neutral but the one editor wants it to reflect his POV in a massive way. Reality check: blatant systematic POV pushers do indeed exist. Weregerbil 12:55, 28 June 2006 (UTC)

I am going to contact the TheGrappler, Worldtraveller and Walkerma, to get their feedback on this issue. These are the individuals who started the dispute process as can be seen here. Cedars 09:35, 28 June 2006 (UTC)

• Thank you for the invitation to comment. I think it's worth mentioning here that at WP:FAC, any "actionable" oppose vote (one that specifies an improvement that needs to be made to bring the article up to WP:WIAFA) actually does provide a veto. However, in that situation, the "featured articles director" has the interprative power of whether it really is an actionable point or whether this is just stick-in-the-mud opposition, or somebody who has got a content dispute against the consensus version of the article. So the one vote veto on FA is a little bit of a myth, but on the other hand it's true that consensus on FAC has got to be very strong, stronger even than at WP:RFA and a lot stronger than what typically gets closed as a "consensus" vote at WP:AFD. I don't believe that "super-consensus" is a good model for the good articles system, for the simple reason that WP:GAC is about checking whether an article meets certain standards whereas WP:FAC is largely about improving an article until it is the best it could possibly be (so obviously has to keep going until nobody can think of a way of making it any better). The good thing about FAC is that there is a featured articles director who can decide whether to count an oppose vote as just a piece of blatant systemic POV pushing; it would be harder to have an analagous situation here. "Everyone has a veto, except blatant POV pushers" is not workable as a system. Additionally there may be instances where people oppose on the grounds "well, this article isn't ready, but it needs to be brought up to standard by doing X, Y and Z" which would be fine as actionable opposition at FAC, but for a good article there may be consensus among other reviewers that it is still good enough to pass WP:WIAGA. For that reason - and because it is simply normal Wikipedian good practice - I think consensus building is better than vetoing. "Everyone has a veto" wasn't something I was happy with but it did seem to be a quick way of avoiding bureaucracy. The problem is that it can be blatantly abused. Consensus shouldn't be brought about by vote counting alone, though. I think for a consensus system to work better than the veto system, comments need be judged on their content (how well they relate the article concerned to the WIAGA criteria) not just be counted as a vote. Remember that POV pushers can also vote "support" to an article that suits their view! TheGrappler 09:44, 29 June 2006 (UTC)

• Thanks for asking my opinion. In fact I had little to do with setting up the disputes page (which is clearly very useful, thanks Grappler!), but maybe I can add some comments anyway. I presume this change came about because of this lively discussion on the Danish cartoons.

1. There does seem to be a disconnect between delisting and relisting procedures that can lead to a cycle of relist-delist. I think Kim has identified the problem - the issue of the single person veto vs. the consensus. I support the idea of inserting a phrase into the main description at Wikipedia:Good articles/Disputes to indicate that one person cannot ignore the consensus and delist or relist again. I'd make a minor change to the last part of the rewording, to break the long sentence in two and to correct the typo in unilaterally: When editors disagree whether an article reaches the good article criteria, discussions to find consensus on the status are listed here. Consensus reached at this page cannot be unilaterally vetoed, but requires new discussion at this page.
2. TheGrappler nicely points out that consensus can be a judgement call, and I believe that a level-headed moderator is needed to resolve such things with "The judge's decision is final". In some groups there is an official or de facto leader, but at the disputes page there is no one person able to do that. Do we need to vote on a moderator for those times when consensus is not obvious? Walkerma 01:12, 1 July 2006 (UTC)

• • • I don't like the idea of having elected moderator's - I fear that can end in all kinds of disputes and recriminations, which we could do with avoiding. I think the best thing to do is to let the dust settle on a debate, then a more experienced reviewer to declare the matter closed by taking it off the review page and putting in a note to the appropriate talk page. Is there a "disputed GA" template in use at the moment so people can find these debates while they are ongoing? If not there probably should be and the reviewer can change that too, either to "delisted" or the standard good article template if it is kept. I think it's worth pointing out that the vast majority of disputes seem to run pretty smoothly, and on the whole it's probably a good thing that they avoid becoming linked with content disputes (they should merely be "criteria disputes": does the article fit the criteria? Having said that, the demarcation line is not always so clear). What seems to be problematic is a minority of cases where we lack the Holy Grail of Wikipedia, the Non-Voting Consensus. I'm not sure that moderators are the solution, but I can see where the idea comes from. The cause of the problem does seem to be the lack of linkage between the dispute and relist processes, leading to the cycling problem. There is an alternative solution which is to require consensus to list an article (which is what is done on de: but not so far on en:); perhaps for longer articles and for previously disputed ones this is the flip side solution. But it involves the dreaded bureaucracy again. Rather than go for something so radical and with such an obvious downside, just to solve a minority problem, I'd rather go with the suggestion above. TheGrappler 00:41, 3 July 2006 (UTC)

Delisting proposal two

Reading the statements above the changes have put emphasis on gaining consensus during the disput process. Please read Wikipedia:Consensus Maybe Delisting of current GA articles should go through a seperate process that requires a detail response stating the exact criteria that the article doesn't comply with.

Currently the dispute give the following procedure for delisting GA articles

---

If you believe an article should be delisted

If you find an article listed as good that does not actually satisfy the good article criteria, then you can delist it:

1. Check the good article criteria to see which criteria it fails to meet.
2. If the problem is easy to resolve, it might be better to be bold and fix it yourself.
3. If you can't fix it, remove the {{GA}} tag on the article's talk page and put in {{subst:DelistedGAbecause|reason}} (at the bottom of the page) or {{DelistedGA}} (at the top). Do not use {{FailedGA}}.
4. Remember to explain what the problem is and what needs to be improved to meet the criteria.
5. Remove the article from the list at Wikipedia:Good articles.
6. If you find an article that you suspect should be delisted, but aren't certain, then you can ask other editors to review the situation by adding the article to the list below.

---

maybe it should be something like this

If you find an article listed as good that does not actually satisfy the good article criteria, then nominate it here:

1. If you find an article that should be delisted, then ask other editors to review the situation by adding the article to the list below.
   1. Check the good article criteria to see which criteria it fails to meet.
   2. If the problem is easy to resolve, it might be better to be bold and fix it yourself.
   3. Remember to explain what the problem is and what needs to be improved to meet the criteria, on here and the talk page with a link back to this page
2. With consensus if the article cant be fix, remove the {{GA}} tag on the article's talk page and put in {{subst:DelistedGAbecause|reason}} (at the bottom of the page) or {{DelistedGA}} (at the top). Do not use {{FailedGA}}.
   1. Article editors should be given 7 days after consensus to fix the problem.
3. Remove the article from the list at Wikipedia:Good articles.

This discussion should be allow to take place without support/oppose votes. That a straw poll should be conducted seperately to the discussion. Gnangarra

I like this proposal much better to the current one. But I think you shoudl exclude the "7 days to fix the problem" part. If they can fix the problem, they should relist it for nomination again.--Konstable 03:16, 29 June 2006 (UTC)

I realize the problem on politically sensitive topics, for instance, but GA is fairly easy to get compared to FAC, immediate delisting shouldn't be a problem. Otherwise, it might discourage people from delisting the articles that really need to be delisted. Maybe an idea might be to allow immediate delists but then make it come here if there is a dispute and then have that dispute binding - sort of like a PROD? RN 03:30, 29 June 2006 (UTC)

I agree with RN. Good articles need to be easy to delist to ensure the quality of good articles is mantained. The good articles project shouldn't be about [edit: making] editors feel good (you can use barnstars for that), it should be about identifying good articles. Cedars 04:37, 29 June 2006 (UTC)

i've struck out the seven days, the repeated immediate delistng by one editor is the cause of this proposal. Thats why I suggested the discussion approach first, that way the reasons have got to be defined against the criteria, malicious delisting can then be quickly identified. Gnangarra 08:16, 29 June 2006 (UTC)

While true that repeated delisting (especially after it violated consensus) got us here mostly in the first place, the right to have at least one immediete delist seems to me a conveinent one. Considering that the GA system is rapidly growing beyond the size of the FA systems article list, I think we'll probably need a much easier and quicker way to remove articles than the FA system does, and even then, we won't be as important, that is to say, if somebody does start doing silly delists, it won't hurt things as much as, say, immedietly delisting the FA on the main page. After the first delist, then the conversation can start, but the result of that conversation I think needs to be binding somehow. Whether its something like 80-20 consensus or some other rule, im not exactly sure, but it seems to me that the immediete delisting thing is fine and perhaps even helpful, it just can be abused without a stronger disputes system. Homestarmy 01:24, 1 July 2006 (UTC)

One problem is that content is dynamic so a designation of good or not should not be eternal (or at least eternally without review), nor should consensus be taken as forever binding. Also, it's true there is a degree of subjectivitity in the criteria. While I fully support the idea that delistings should specify which criteria they are based on, there is still the problem of the POV-pusher who will claim "factual accuracy" or "NPOV" or even "offensiveness" problems. It's finding a way of breaking out of the list-delist cycle which is the problem. I'm becoming quite partial to the idea of having a time limit (e.g. 7 days to fix) rather than immediate delisting, with the possibility of contesting at GA/D. While some delistings are simply immediately clear, it's no tragedy if they take 7 days to work off the system (some FAs or ex-FAs are shockingly poor, but get given several weeks or even several months to secure their status; since GA is a lower hallmark than FA there's presumably less reason to be trigger-happy). This does have the advantage of giving editors more of an opportunity to improve an article to be sure it's reaching the criteria, reduce heat and acrimony (which immediate delisting certain has a track record of generating) and substantially address the problem of arbitrary delistings. It also seems to deal with the structural problems underlying the promote/delist cycle. TheGrappler 01:00, 3 July 2006 (UTC)

Im a bit concerned that without the ability to immedietly delist, the disputes page may be overwhelmed; already, many articles don't get commentary from many people. Homestarmy 04:27, 3 July 2006 (UTC)

That's why I put my suggestion like I did. I can't see that it would cause any increase in the number of disputed delistings: in fact I suggest it might do the contrary. The basic idea is: (1) reviewer notices a "good" article doesn't meet certain criteria, and leaves a note saying that there are 7 days to fix those particular problems; (2) an editor can try to fix those problems within the time limit or can take it to the disputes page; (3) if not taken to the disputes page and no improvement is made before the time limit expires, the original reviewer can revoke the GA status. A lot of these would likely go through uncontested. Cases where the reviewer trigger-happily delists but the problem is quickly fixed (I recall a series of disputes about orders of discographies that might be a good example of this) sometimes end up unnecessarily on the disputes page when there is no longer anything that is disputed; this proposal would avert that. I suppose some stuff would go through the disputes page rather than delist followed by virtually immediate renom (which might take 3 weeks to get looked at) but it is the "delist then immediate renom" which is causing the problems at the moment with list/delist cycles. That's the place to break the chain. TheGrappler 04:45, 3 July 2006 (UTC)

Where to put Autostereogram?

I'm at a loss of where to catagorize Autostereogram. Its sort of computer graphics, sort of art, sort of biology and perception... I'm just not sure where it should go. Right now its in 'Art', which doesn't feel right, but felt better than anything else. Anyone have any great ideas as to where it should go, or a new sub-category that should be created for it to fill? Phidauex 15:51, 29 June 2006 (UTC)

Record month for GAs; proportion of good articles continues to rise

I've just updated the good article stats page and it turns out that July was an excellent month, with an increase of 146 good articles. This is the biggest rise since the 152 in February, and is the highest recorded under the current, more stringent, nomination system (the all-time monthly record is 316 in December 2005, and is unlikely to be broken for some time).

Moreover, the proportion of all articles that have "good article" status is continuing to rise, despite the rapid growth of Wikipedia, as shown in this graph. The GA system is currently outpacing overall article growth, unlike featured articles, which are continuing a long-term decline as a proportion of all articles despite their near-record 38 more articles in June. Hopefully this is just a sign that the standards required for WP:FAC have been getting higher. TheGrappler 02:09, 3 July 2006 (UTC)