Wikipedia talk:Requirements for good articles/Archive 1

Active discussions

Starting GA processEdit

Hello all, to start the GA process, I will boldly state what I think:

  • GA's should be easier to make than VGAs
  • The GA criteria should be a subset of the VGA criteria. We pick a few of the VGA criteria, and say they are the GA criteria
  • The VGA criteria are here.

In my opinion:

  • We can drop criteria 2 (Comprehensiveness), 9 (Article does not need improvement); we should modify length (item 3, I propose 2,5k / 1 screen), References (item 10; references still required, external links not necessary).
  • Do we want to drop the redlink criterion (item 7) for good articles? - Speaking graphically; Mali does not meet the current VGA criteria. Would you think that is what a good article should look like?
  • We need to talk about voting requirements (min number of votes/min support) for good articles

The rest should not change really.

So, happy discussing away. --Eptalon (talk) 13:30, 29 January 2008 (UTC)

I feel that the main things that should be different from the VGA criteria are:
  1. article doesn't have to be comprehensive
  2. article has a shorter length requirement
  3. a few redlinks are okay (but all important words still need to be linked)
  4. we don't need the part about "For articles that have references or external links on the English Wikipedia, there must be at least one in the Simple English article as well" in the reference criteria
  5. voting requirements are not as high
Those are my general thoughts. As a note, I have changed the criteria page to match the VGA criteria page a little better (part about making large changes and demotion process) and also to be more up-to-date as to what we're thinking. I'm also going to go ahead and mark the points that are under discussion because I think that helps people. · Tygrrr... 15:17, 29 January 2008 (UTC)
If you had the choice between an additional red link and an unlinked important word, which would you prefer? - What do you think of the Mali article? --Eptalon (talk) 15:22, 29 January 2008 (UTC)
Redlinks. All important words should be linked. People should never leave an important word unlinked just to avoid having a redlink. Plus, if they want it to eventually be a VGA, they're going to need to have those words linked anyways. And redlinks encourage new pages to be created. · Tygrrr... 15:29, 29 January 2008 (UTC)

Point 2: length requirementEdit

Eptalon has suggested 2.5 kb. I would like to suggest 3 kb. What do others think? · Tygrrr... 15:40, 29 January 2008 (UTC)

3 kb --Cethegus (talk) 16:30, 29 January 2008 (UTC)
3 kb from me too. Razorflame 16:35, 29 January 2008 (UTC)
something that is easy to measure. I am fine with any number; i just needed to start somewhere. Original idea was: About half the size of a VGA (5k). So yes, we can say 3k too. --Eptalon (talk) 16:41, 29 January 2008 (UTC)
3kb --Gwib -(talk)- 18:38, 29 January 2008 (UTC)
3kb, I suppose. Even though that is really quite small, in my opinion. --Thamusemeantfan (talk) 03:02, 30 January 2008 (UTC)
I prefer 4kb. - Huji reply 10:13, 2 February 2008 (UTC)
4kb probably is better. I mean, look at Sound Tribe Sector Nine. That's almost exactly 3kb, and it has two sentences of prose. --Thamusemeantfan (talk) 06:38, 4 February 2008 (UTC)
The requirement reads "A minimum is 2.5 kilobytes (about 1 screen), not including infoboxes, images, references, other websites, interwiki, and categories" (italics added for emphasis). Your example of Sound Tribe Sector Nine is only 674 bytes when you don't include that stuff. It would need be approximately 5 times as long to meet a requirement of 3kb. Just so you know. · Tygrrr... 14:54, 4 February 2008 (UTC)
Either 3 or 4 is good with me. Durova (talk) 01:58, 6 February 2008 (UTC)
I'd say 3kb as a minimum is good, but I definitely would not be opposed to 4kb. нмŵוτнτ 20:06, 6 February 2008 (UTC)

(unindenting) Do we want a GA to be shorter than a VGA? - Would a VGA not be defined more through the coverage of the subject, the external references (etc) than its mere length? - If pure length does not matter, we might as well leave it at 5k (like for VGAs). What do you think? --Eptalon (talk) 13:04, 4 February 2008 (UTC)

I actually think that 5kb would be good. Razorflame 14:45, 7 February 2008 (UTC)
I think that the length requirement should be slightly lower than the requirement for VGAs. Therefore, I would be fine with either 3kb or 4kb. Maybe we could do 3.5 as a compromise for those that would be fine with either? · Tygrrr... 14:51, 7 February 2008 (UTC)
That seems like the best idea yet. нмŵוτнτ 17:54, 7 February 2008 (UTC)
3.5...I like that the best. --Thamusemeantfan (talk) 05:29, 8 February 2008 (UTC)
3.5 is good with me. Razorflame 14:02, 8 February 2008 (UTC)

Update: Consensus is forming to have the minimum length at 3.5 kb. There is still some time to discuss. (If we choose 3.5kb, about how many screens is that? I'd like to be able to update the wording from "about 1 screen" to be accurate.) · Tygrrr... 14:57, 8 February 2008 (UTC)

3.5 is good with me as well. --Cethegus (talk) 14:59, 8 February 2008 (UTC)

(update: 1.6k = 1 screen; 3.2k = 2 screens ; 4.8k = 3 screens)--Eptalon (talk) 15:06, 8 February 2008 (UTC)

This makes for a change. I support 3.2k instead of 3.5k as 3.2k is a uniform 2 screens. Razorflame 15:51, 8 February 2008 (UTC)
5k is about 3 screens. Therefore one screen is 1 (2/3) k; I would leave it at 3.5k; but say about 3 screens. What did we actually base that screen on, by the way? --Eptalon (talk) 15:57, 8 February 2008 (UTC)
Never mind my last comment. Leave it at 3.5kbs Razorflame 16:00, 8 February 2008 (UTC)
We could always leave out the "about # screeens" part too. It really doesn't serve that much purpose, I guess. As far as what we based the screen on, I don't know. And don't you mean that 3.5kb is about 2 screens, not 3, in your last comment? Just want to make sure I'm understanding you correctly. · Tygrrr... 16:53, 8 February 2008 (UTC)
1.6 k per screen; 3.2k (3.5) is therefore two screens; 4.8k (5) 3 screens; looks confusing, lets take it out (also of the VGA thing). --Eptalon (talk) 19:40, 8 February 2008 (UTC)
Agreed. Razorflame 19:42, 8 February 2008 (UTC)

Point 6: red linksEdit

I like the current wording "All important terms should be linked and there should be very few red links left. Red links point to articles that do not exist yet. Usually the important word or phrase is only linked the first time it occurs." I feel that important words need to be linked but that it's okay to have some red links. Probably no more than a dozen, or so. I don't know if we should add a specific number or not, though. · Tygrrr... 15:44, 29 January 2008 (UTC)

If you look at Mali there is a difference between red links in the text and red links in lists of places. Therefore I would suggest: "there should be very few red links left (not more than five - lists might justify an exception)". --Cethegus (talk) 16:36, 29 January 2008 (UTC)
I believe that good articles should have a few red links, otherwise, they would be considered very good articles. Good articles are articles that are good, but need improvement, so there is no reason for there to be no red links. No more than 8 links on a page. Razorflame 16:37, 29 January 2008 (UTC)
What about giving a percentage, like no more than 20% red links (with the possible exception of listings), as outlined? --Eptalon (talk) 16:39, 29 January 2008 (UTC)
15%. Razorflame 16:43, 29 January 2008 (UTC)
Then we'd have to count all the links and then calculate how many are red. This could be quite the daunting task on longer articles, and certainly not easy even on the short articles. I think it would be best to either give a number or leave it as is (without a specific limit). · Tygrrr... 16:45, 29 January 2008 (UTC)
Using percentage can cause problems. The easiest way to bypass it would be to just link everything in sight. Links are good, but every other word, every time it appears on the page, thats overkill but it would decrease the percentage of redlinks.-- Creol(talk) 00:46, 1 February 2008 (UTC)
If that's the case, then my original suggestion of 8 redlinks is what I suggest. Razorflame 16:46, 29 January 2008 (UTC)
I agree with 8. --Thamusemeantfan (talk) 23:50, 30 January 2008 (UTC)

(unindenting) What about leaving this as vague as it is now; very few redlinks means less than half, but 20 redlinks in a big article is different from 5 redlinks in a small one? - Alternative wording All important terms should be linked and there must not be many red links left. (Highlight added to show what I changed) --Eptalon (talk) 11:09, 1 February 2008 (UTC)

I agree with the vague wording; having it be too specific is bad for a bigger article. --Thamusemeantfan (talk) 16:07, 1 February 2008 (UTC)
Vague wording sounds good. I also agree with the vague wording. Razorflame 16:20, 1 February 2008 (UTC)
I agree vague is good. That lets people think about each article instead of just using math. Durova (talk) 01:53, 6 February 2008 (UTC)
Yeah, that wording makes it more lenient for different articles, but it still has the same impact. нмŵוτнτ 17:56, 7 February 2008 (UTC)

Update: unless there are any last minute objections, the current wording will remain: "All important terms should be linked and there must not be many red links left. Red links point to articles that do not exist yet. Usually the important word or phrase is only linked the first time it occurs." There is still some time to oppose. · Tygrrr... 14:46, 8 February 2008 (UTC)

Point 8: templates showing improvement neededEdit

Point 8 currently reads: "There must be no templates pointing to the fact that the article needs improvement. These templates include {{complex}}, {{cleanup}}, {{stub}}, {{unreferenced}} and {{wikify}}. The article also should not need them." Eptalon has suggested that we don't need this criteria. I disagree. I think it would be shameful to have an article on display as a GA that literally has a sign on it saying "This article has something blatantly wrong with it." We shouldn't have GAs that are stubs, unreferenced, complex or unwikified. I think this is one of the more basic requirements an article needs to fulfill in order to be a GA and should definitely be kept. · Tygrrr... 15:49, 29 January 2008 (UTC)

This needs to stay. We can't have any of these templates on them. Razorflame 16:40, 29 January 2008 (UTC)
I agree. --Cethegus (talk) 16:42, 29 January 2008 (UTC)
Ditto --Gwib -(talk)- 18:40, 29 January 2008 (UTC)
Also agreed! --Thamusemeantfan (talk) 23:51, 30 January 2008 (UTC)
I agree also. Durova (talk) 02:00, 6 February 2008 (UTC)
Yeah, I think this should be about the most basic thing for a GA. If it needs those kind of improvements... it's not a good article. нмŵוτнτ 17:57, 7 February 2008 (UTC)

Update: unless there are last minute objections, this will remain as a criteria and will keep its wording as-is. There is still some time to oppose. · Tygrrr... 14:47, 8 February 2008 (UTC)

Point 9: referencesEdit

Point 9 currently reads: "Content that is from books, journal articles or other publications needs to be referenced. This can either be done with <ref>..</ref><references/> tags, or as a list of publications. The article does not have an {{unreferenced}} tag and does not need one." We could probably drop the last sentence since, as Eptalon pointed out during the VGA discussion, it is redundant to point 8. · Tygrrr... 15:53, 29 January 2008 (UTC)

Agreed, we should drop the last sentence, as it is redundant with point 8. Otherwise, this needs to be kept intact. Razorflame 16:41, 29 January 2008 (UTC)
I agree. --Cethegus (talk) 16:44, 29 January 2008 (UTC)
Sounds good. --Thamusemeantfan (talk) 23:52, 30 January 2008 (UTC)
Yes I agree. Durova (talk) 02:00, 6 February 2008 (UTC)
Yes, remove the last sentence, but the rest sounds perfect. нмŵוτнτ 17:58, 7 February 2008 (UTC)

Update: unless there are last minute objections, the last sentence will be dropped and the criteria will read "Content that is from books, journal articles or other publications needs to be referenced. This can either be done with <ref>..</ref><references/> tags, or as a list of publications." There is still some time to oppose. · Tygrrr... 14:49, 8 February 2008 (UTC)

How to make an article good (voting)Edit

The main consensus that was forming on the on the VGA criteria talk page was 4 to 5 named voters, 70% support needed for promotion. Other suggestions? · Tygrrr... 15:56, 29 January 2008 (UTC)

I think that the percentage should be the same as the percentage needed for a successful RfA. 65% support and 4 named voters would be my suggestion. Razorflame 16:42, 29 January 2008 (UTC)
I'm in favour of a minimum of 5 named voters. --Cethegus (talk) 16:45, 29 January 2008 (UTC)
Alright. 65% support, 5 named voters. Razorflame 16:47, 29 January 2008 (UTC)
As stated in the VGA process; 5 voters; but 70% support (4 out of 5 (80%), 5 of 6(83%, 4 out of 6 is 67%), 5 out of 7 (71%, 4 of 7 is 57%), 6 of 8 (75%, 5 of 8 is 63%), 7 of 9 (77%, 6 of 9 is 67%), 7 of 10 (70%) etc... --Eptalon (talk) 09:32, 30 January 2008 (UTC)
I think we should have five votes, at least 60% support. --Thamusemeantfan (talk) 23:53, 30 January 2008 (UTC)
Eptalon, I think that 70% is a bit too high. I really think that we should make it 65% support. Razorflame 20:35, 31 January 2008 (UTC)
I think that 65 sounds good. --Thamusemeantfan (talk) 00:37, 1 February 2008 (UTC)

(unindenting) Lets go for 65 then; 5 or 6 votes: 1 opposing vote; 7 or 8 votes: 2 opposing votes; 9 or 10 votes: 3 opposing (etc) for a vote to pass. --Eptalon (talk) 00:54, 5 February 2008 (UTC)

If an article is truly good, having a high amount of support will be easy to attain. I feel that if 2 of 6 voters (i.e. 33%) don't think the article is good, it probably shouldn't be promoted. I don't think 65% is high enough if we have 6 voters. I think 70% is better because I think if we only have 5 or 6 voters, 2 oppose votes should stop it from being promoted. Therefore, I would like 70% and a minimum of 5 voters. · Tygrrr... 15:06, 5 February 2008 (UTC)
Supposing an article is really good, getting votes should not be much of a problem; therefore getting 70% support should not be either; or as to take the figures from above: 5 or 6 votes: max. 1 oppose, 7 or 8 votes: max 2 oppose; 9 or 10 votes: max 3 oppose for an article to still pass. --Eptalon (talk) 10:12, 7 February 2008 (UTC)
Yes, that's what I'm saying. Are you saying that you again are supporting having the requirement being set at 70%? · Tygrrr... 14:42, 7 February 2008 (UTC)
I do not support 70%. 65% is better. 70% is too high. Razorflame 14:44, 7 February 2008 (UTC)
Razorflame, you do realize that with a 65% support requirement that if only 4 of 6 voters think it is good it will still be promoted, right? Do you really think that's a good idea? · Tygrrr... 14:52, 7 February 2008 (UTC)
4/6 meaning 66.6%, right? I guess we can go with 70%, but I still think that it's a bit high. Razorflame 16:51, 7 February 2008 (UTC)
4/6=2/3=66%; 5/6= 83% --Eptalon (talk) 16:53, 7 February 2008 (UTC)
Alright. I'll go along with 70%, even though I think it's a bit high. Razorflame 16:56, 7 February 2008 (UTC)
I'm leaning for 70% (at least 5 voters). I mean, as mentioned earlier, if the article is actually good, it shouldn't be hard to get people to pass it. If a decent amount of people don't agree, it probably shouldn't be a GA. нмŵוτнτ 18:02, 7 February 2008 (UTC)
I'd go for 70% on this, just to make sure we're getting a really good look at the article. Archer7 - talk 13:43, 8 February 2008 (UTC)
5 named voters and 70% seems good for me. --Cethegus (talk) 15:03, 8 February 2008 (UTC)

Update: The consensus that is being formed is to have a minimum of 5 named voters and 70% support for an article to be promoted. There is still some time to discuss. · Tygrrr... 14:52, 8 February 2008 (UTC)

General commentsEdit

For comments not related to a specific section.

It seems we are reaching agreement much more quickly for GA criteria than for VGA criteria. Should we set a date that an agreement needs to be made? --Thamusemeantfan (talk) 07:31, 2 February 2008 (UTC)

There's already a deadline set: this Friday the 8th. · Tygrrr... 19:24, 2 February 2008 (UTC)


I think one of the things we need to make clear also is that by supporting a certain article nomination, the supporter not only says that this article should become a (very) good one, but that the article also meets the requirements to do so; opposing a nomination can therefore also mean that the article is not yet ready, as it does not meet some requirements. This applies to VGA's as well, of course.--Eptalon (talk) 10:17, 7 February 2008 (UTC)

I would have to agree here. People who support or oppose articles should say why they vote the way that they do. Razorflame 14:43, 7 February 2008 (UTC)
They don't need to say why they vote the way they do; they just need to be aware that if they often support articles that do ont meet the respecive requirements, they will probably lose their credibility. --Eptalon (talk) 15:42, 7 February 2008 (UTC)
Do you have a specific wording proposal, Eptalon? I'm tentatively supportive of the idea but I'd like to see an exact wording before I commit to liking it :-) · Tygrrr... 15:48, 7 February 2008 (UTC)

(unindenting); I don't know, perhaps the following will do Users can support or oppose requests that certain articles be promoted to their new status (of good article, or very good article); If they sign their name in support of a certain article, they also say that the article meets the criteria it needs to get promoted to its new status. Opposing is similar: In their opinion, the article does not (yet) meet the criteria to get promoted.; That text is supposed to go on the proposals page; at the moment it is neither short nor simple, but I hope it gets the idea across (Article promotion is not a matter of personal opinion, but rather of having checked that the article meets the resp. criteria; an editor signing in support confirms this article meets the criteria). There are now two versions of the idea; good luck simplifying it (as to timing: this is not part of the guideline...) --Eptalon (talk) 16:48, 7 February 2008 (UTC)

I like the wording, but it does need to be simplified. How about this for a simplified version: If you add your name to support, you support the article becoming good. If you add your name to oppose, then you think that it doesn't meet the requirements. People found to be abusing this system can have their rights to vote taken away. or something along those lines. You like this suggestion? Razorflame 16:53, 7 February 2008 (UTC)
Well, I'm all for changing the intro to both the Proposed good and very good article pages. It's bothered me for a while that they're one sentence then a break, one sentence then a break, one sentence then a break, one sentence then a break. They could definitely use some improvement. · Tygrrr... 17:09, 7 February 2008 (UTC)
Does anyone even like the suggestion that I put into the mix? Razorflame 17:11, 7 February 2008 (UTC)
Razorflame: One of the problems your proposal has is that (to my liking) it does not become clear that I put my name under the support votes for foo, I also say that that article meets the requirements, not so much that It would be a good idea if foo became a (V)GA. The criteria are objective; this should not become I do not want that foo is promoted, because I think foo should not be a VGA. Think about it; if a seemingly controversial topic, like for example sexual intercourse was nominated there, how many people would oppose, just because they think it should not be made a (V)GA, not because it would (currently) probably not meet the standards? --Eptalon (talk) 09:56, 8 February 2008 (UTC)
I would have to agree there. People opposing just because they don't want an article to make the {V}GA standard would be not fair, especially if it were an article on a controversial topic. We could add in a sentence about making all your voting neutral unbiased votes, but I'm not sure that this would solve this problem. Razorflame 14:06, 8 February 2008 (UTC)

(unindenting) Perhaps the word vote is wrong; it implies that it is your opinion that counts here; should it rather not be something like checks (6 people have checked that it actually meets the requirements, 2 disagree, so the thing does not get promoted)? --Eptalon (talk) 14:24, 8 February 2008 (UTC)--Eptalon (talk) 15:02, 8 February 2008 (UTC)

I would like to remind you of Amandajm's words "it makes me cross to see an article about a subject that requires quite a good knowledge of the broader field to understand it, being put up as a potential "VGA" when it has been lifted out of Wkipedia with very little simplification of its complex and sometimes confusing language." I think the decision if an article is good enough in this respect cannot be replaced by a mechanical check if all points are met. Therefore it makes sense to vote by several persons and not to check. For checking could be done by a single person. --Cethegus (talk) 15:21, 8 February 2008 (UTC)
Look at Item 9 (article does not need improvement); clearly such an article could be tagged by one of the proposed tags ({{complex}} comes to mind). It would therefore not meet the requirements; The "vote" is not about some editor thinking that this would or would not make a (V)GA; but rather, the editor being able to point out that the article should or should not be promoted ...because... (plus some reference to the resp. guideline). The argument of me opposing or supporting Condom to become a (V)GA, because of my personal views on contraception is not a valid one. --Eptalon (talk) 15:45, 8 February 2008 (UTC)
We discussed the quorum because checking by one person seems not to be sufficient. But as long as there is no doubt that the 70% must be met one could call it checking if you prefer. --Cethegus (talk) 16:27, 8 February 2008 (UTC)
I am fully aware of the amount of approval needed (discussed above); all I want here is to break the association an average user might have, between an RfA (or similar vote), and a vote for a (very) good article, thats why I proposed a more "neutral" term (as outlined above). Now whether we call that checks or strawberries, or whatever does not matter that much. All we need to make clear is that it should be (support or oppose) because of things in the respective guideline. --Eptalon (talk) 19:57, 8 February 2008 (UTC)

Good articles and very good articlesEdit

As it turns out, all good articles will also be very good articles; I therefore propose to have the Cat:VGA as a subcat of cat.GA; that way all GAs that become VGAs have their GA tag replaced by a VGA tag; otherwise we would have both tags.--Eptalon (talk) 15:37, 8 February 2008 (UTC)

Wouldn't it be the other way around? GA is subcat of VGA? Razorflame 15:50, 8 February 2008 (UTC)
No. All Very Good Article are good, not all Good Articles are Very Good.--Bärliner 15:52, 8 February 2008 (UTC)
Ah. My fault. It's early in the morning here, and my brain still isn't working to full capacity. I agree with Eptalon's point. Razorflame 15:54, 8 February 2008 (UTC)
Sorry; ofc Barliner is correct. Some day there will hopefully be Good articles that do not meet the VGA requirements. --Eptalon (talk) 16:01, 8 February 2008 (UTC)
Return to the project page "Requirements for good articles/Archive 1".