User talk:Plrk/On the creation of articles

So as long as a spammer or vandal inserted a red-link into an article somewhere, or put it on requested pages, then they could go ahead and create that article? Do you have any evidence to back up your statement that worthy new articles all have redlinks elsewhere? Maybe a recent analysis of patrolled articles against incoming internal links? Avruch^T 21:45, 14 September 2008 (UTC)[reply]

Not evidence that would stand it's ground in a court or in a research publication, but personal experience (from new pages patrolling) and common sense. Not all new worthy articles have redlinks to them, but I'm pretty sure most do, and I'm even more sure that most speedily deleted new articles do not have any. As always, there are exceptions. Plrk (talk) 21:49, 14 September 2008 (UTC)[reply]

Also, forcing an editor to insert a redlink in another article before they can create say, a vanity article about a friend, would deter many - I believe. Plrk (talk) 21:50, 14 September 2008 (UTC)[reply]

I've come across vanity spammers who absolutely insist they have red links to their vanity pages, long after they've been deleted and their sites blacklisted. MER-C 04:55, 15 September 2008 (UTC)[reply]

I've come across them too. That'shardly the point, though. Plrk (talk) 05:26, 15 September 2008 (UTC)[reply]

Comments

Studies have indicated that a majority of Wikipedia's content is added by "anons"
- Can you tell us where you found this?
  Yes, I read it somewhere (in Wired or the Wikipedia Signpost, probably). I'm guessing that the study I was thinking of was this one, reported about in Ars Technica, Scientific American. Note that the study makes a difference between "overall content" and "quality content": while logged-in users add the most content overall, anonymous users add the most quality content. In the quote above, I was referring to this "quality content", as the "overall content" includes template tagging, discussion, reformatting, etc, which I would not really call "content". Plrk (talk) 22:22, 14 September 2008 (UTC)[reply]
if something is notable enough to warrant an article, it is probably already has an article, and if not, it probably has one or more redlinks pointing to it.
- I find this reasoning questionable. Redlinkage does not imply notability.
  Redlinkage implies more notability than no redlinkage, for sure. Note "implies", not "proves". Plrk (talk) 22:22, 14 September 2008 (UTC)[reply]
you should only be able to create an article if there is one or more articles in the main namespace linking to it
- First, I hope this is only supposed to apply to unregistered editors! If this was your intent, you should change the wording. I would hate to see such a restriction to registered, especially autoconfirmed, editors.
  I did not, but I later added a variant that would allow autoconfirmed editors to always create articles. Plrk (talk) 22:22, 14 September 2008 (UTC)[reply]
- As Avruch mentions above, an IP could easily insert the necessary red links into articles and create as many articles they want. This does make it a little more difficult/annoying for potential vandals, though. Perhaps a better way to limit would be to say a certain IP address can only create x articles per day, say three.
  The more difficult/annoying part is what I'm thinking of. Instead of one step to create a dumb-ass article, there's two. (Or instead of two steps, counting the registration procedure, there's still two.) An added benefit would be that the people watching the article (there's often someone) that had a random redlink inserted, would hopefully check out this new strange link add tag it for speedy deletion if unwanted. Plrk (talk) 22:22, 14 September 2008 (UTC)[reply]

Finally, given the backlog that persists at Special:NewPages, is the benefit of good IP-created articles worth the extra burden that will face us with vandal-created articles? — Twas Now ( talk • contribs • e-mail ) 22:07, 14 September 2008 (UTC)[reply]

If it was not for the backlog, I would not make this suggestion. The primary benefit of my proposal lies not in good IP-created articles, but in the reduction of nonsense, spam, and vanity articles created. Plrk (talk) 22:22, 14 September 2008 (UTC)[reply]

I don't see how giving more people the ability to create articles reduces the amount of crap articles. If more people are given the right to create articles, the amount of crap in the NewPages backlog will inevitably increase, giving patrollers more to sort through. Additionally, the amount of good articles will increase, which means the patrollers will have even more to sort through. I make these assertions under the assumption that you will not be able to convince the community to remove article creation from new editors (was that indeed your suggestion at question 3, bullet 1?). — Twas Now ( talk • contribs • e-mail ) 23:13, 14 September 2008 (UTC)[reply]

While I do suggest that more people should be able to create articles, I am really suggesting that more people should be able to create less articles. Most of the crap that is flagged for deletion on sight on New Pages Patrol has no incoming links. That is the core of my proposal - moving the restriction from users to articles, not removing the restriction. If newly registered users can create articles without restrictions instantly after registration, my proposal would - as you say - have no effect whatsoever, except possibly increasing the amount of worthless articles. Plrk (talk) 23:19, 14 September 2008 (UTC)[reply]

Moved from the Village Pump

Although I symphatise with your idea, I think your suggestion would only make things worse (rather than better). (I am overstating, not to make a mockery of your commendable idea, but to illustrate my point). Your current idea would support the creation of yet more Pokemon character articles; while it would block (or at least slow down) articles on topics in non-western pre-historical cultures, which are less likely to be red-linked. Thus it would only increase the focus on current US-youth-culture-contemporary-topics of English wikipedia. So no, I don't think this would be a good way forward (althoug I agree with the problem you have flagged up, article creation over GA/FA article level has become a pest). Arnoutf (talk) 21:32, 14 September 2008 (UTC)[reply]

I can't see how it could be worse than it is already is. If there are links to Pokémon characters around, they should immediately be redirected to a list of Pokémon characters. The same goes for the rest of the US-youth-culture-contemporary-topics - if it is not notable, it should be unlinked. While the growth of articles on topics in areas that are unfavored due to our systematic bias may be hindered in a few cases, I think that article topics grow from the general to the specified: a general article regarding, say, the Inca civilization, link to other more specialized Incan articles, which when they are created will link to even more specialized Incan articles. And as said, if you were not able to create an article, an instruction would be shown that explained that you could introduce a link in relevant, existing articles. Plrk (talk) 21:38, 14 September 2008 (UTC)[reply]

In my opinion we need fewer, not more Pokemon articles (notwithstanding redlinks around). We should support, not limit (by adding effort of linking) expanding into yet unexplored areas such as pre-historian civilizations which are less known than the Inca.

I recognise it is pretty bad as it is at the moment, I think some limit on article that can never become FA level should be imposed. I am afraid, however, that your proposal would put more stress on including new articles that might become FA (but are not in Wikipedia because of the bias) than on articles that will never become FA, and are only borderline interesting (ok not only Pokemon, but also communities of 3 people established 5 years ago).

So yes, I underwrite your concern, but no I don't think your suggestion is an improvement (Mind you, I have no ideas myself, so I appreciate your effort and thinking on this topic, I just don't think it will work). Arnoutf (talk) 21:49, 14 September 2008 (UTC)[reply]

OK, we agree that we are in disagreement then. Plrk (talk) 21:52, 14 September 2008 (UTC)[reply]

Fair enough, let's wait for some other comments. Arnoutf (talk) 21:56, 14 September 2008 (UTC)[reply]

With your thoughts in consideration, I have added a more liberal variant that I think still would be a significant improvement. What do you say? Plrk (talk) 21:57, 14 September 2008 (UTC)[reply]

The condition "only if there is a red link" will simply encourage POV-pushers, advertisers, etc. to disfigure existing article with semi-relevant red links. Who knows which articles will become FA? Last year I found myself defending the notability of an article against a deletionist. This year the article made FA. -- Philcha (talk) 22:11, 14 September 2008 (UTC)[reply]

(I de-indented your post. I hope you don't mind.) POV-pushers, advertisers, etc already disfigure existing articles with semi-relevant red links. I doubt this would make it much worse. Considering your FA-article, I'm pretty sure it is not an orphan? (Was it an orphan at the time of it's creation, even?) Plrk (talk) 22:27, 14 September 2008 (UTC)[reply]

I can't remember whether the article that reached FA was an orphan when someone slapped a "notability" tag on it. I know I went round a few related articles and inserted links before taking on the deletionist :-)

I don't think orphan status should be made the basis of a rule. Any article on a completely new topic starts as an orphan, including WP's first ever article. -- Philcha (talk) 22:46, 14 September 2008 (UTC)[reply]

You see, even in your own experience the presence of redlinks is an argument to keep an article! As you point out, applying this concept to a new-born wiki would be ridiculous; but newly created articles now are nowhere near WP's first ever article. We have over 2,5 million articles - are there really any completely new topics? I'd say no. The creation of Wikipedia:Walled gardens has been discouraged a very long time. Plrk (talk) 22:56, 14 September 2008 (UTC)[reply]

While I like the idea of giving back to the lower caste of users the ability to create articles again, I don't think this is really the idea to go about it. This would inhibit a lot of articles related to current events (the article on Nomophobia comes to mind; to create it, one would first have to find a list of psychiatric conditions or something, then add a redlink to it, then create it the article, rather than focus on the article itself. Realistically, I think it would pretty difficult to enforce; one could simply make redlinks in their userspace or something, then create the article; at that point, keeping the article or deleting it would depend on the quality, relevance and notability of the article's subject anyway, which makes this a rather pointless idea to begin with. Celarnor ^{Talk to me} 22:41, 14 September 2008 (UTC)[reply]

I'll continue with your usage of Nomophobia as an example. It has plenty of links to it - instead of writing the article first and adding the links later, the process would simply be reversed (in the cases it isn't already). Considering enforcability, my intent is of course to have this written into the software. I also specificially suggest that redlinks in the main namespace are required (in the phobia case and many others, links from the template namespace count as main namespace links as templates are included in main namespace articles (at least according to Special:Whatlinkshere)). Plrk (talk) 22:56, 14 September 2008 (UTC)[reply]

I have some sympathy with these proposals, but I note that the current block on anons creating articles seems to have gone some way to stop the worst kind of nonsense being created. Were I to create, for example, the Glenrothes by-election, 2008 article under your proposals, would I have to find a number of links first to justify creating it? I don't understand how this would work in practise, other than . making the project less user-friendly doktorb _words^deeds 03:53, 15 September 2008 (UTC)[reply]

the Glenrothes by-election, 2008 probably had a link in List of United Kingdom by-elections before creation? You would not "have to find" a number of links - the idea is that this would be implemented on a software level. If there were links, you'd be able to create the article, if not, you would not. Plrk (talk) 05:26, 15 September 2008 (UTC)[reply]

I tend to agree with Arnoutf. If you look at it simply:

If every article always contains 4 redlinks and Subjects A and B are equivalent in real-world importance:
- Subject A contains 5 articles:
  - There are 20 more articles that could be immediately created in Subject A.
    - Those 20 are created, now there are 80 articles that could be created.
- Subject B has 2 articles:
  - There are only 8 articles that could be created for Subject B.
    - Those 8 are created, now there are 32 that could be created

If there is no point where Wikipedia becomes saturated with a subject to the extent that the number of redlinks per article drops, Subject B will never catch up to Subject A. In reality, the number of redlinks per article probably does decrease as the number of articles in a subject increases (assuming polices like WP:N, WP:NOT, etc. remain static) but something like this is still going to be aiding the increase in the coverage gap between A and B. I don't quite see what the point of this is. To switch to an economics perspective, why should we introduce arbitrary imperfections into the "article market?" The demand is still there to create the articles (and would be higher by allowing anon page creation), but by restricting what people can create, you're cutting off the "supply" at an arbitrary point. Mr.Z-man 16:23, 17 September 2008 (UTC)[reply]

There are, and has been, points where Wikipedia has become saturated with subjects to the extent that the number of redlinks per article has dropped - often to zero. I'd go as far as saying that most of Wikipedia's articles have no redlinks, although I have no evidence for this.

To switch to an economics perspective, putting restrictions in a market is often useful: in example, you have to have a capital of 100,000 SEK to go public with a company (in Sweden that is) (akin to having a number of redlinks to "go public" with an article). Another example: you have to be a licensed practitioner to get a job as say, a surgeon (again, at least in Sweden) (on Wikipedia, the "license" to "become a surgeon" would be having a redlink to become an article). There is a great demand to create articles, just like there is great demand to become a surgeon - and just like we have restrictions on surgery practice, we should have restrictions on article creation. This might lead to fewer surgeons/articles, but a better quality overall on the surgeons/articles. Plrk (talk) 21:00, 17 September 2008 (UTC)[reply]

Catch-22...?

Plrk, thanks for your reply, above, with specific relation to the example I gave. But your reply suggests to me hints of the Catch-22. If I understand your suggestion, for me to create the Glentrothes by-election article, let us say, there would have to be a link at List of UK by-elections. But for there to be a link at the latter, I would have to create the former. But I cannot create the former without the latter. Are you not setting up a real block for "as live" creations? I will concede if I have this situation wrong, but I cannot stop worrying that your suggestion will stop Wikipedia from advancing rather than encouraging. doktorb _words^deeds 21:06, 17 September 2008 (UTC)[reply]

As my proposal apparently did not gain any traction whatsoever, you have nothing to worry about. The only thing that could make this proposal go through now would be a directive from Jimbo.

Anyways, to your question. What the proposal would set up is a block against Walled gardens - that is, you would not be able to create a set of articles that only linked to each other and not any other articles. My proposal would, apart from the intended primary effect of reducing crap articles, force all new articles to be linked to the rest of the encyclopedia through at least one link. There would be no more orphans (except where articles are created and links to them removed afterwards, obviously). In your example, you could set up a link to "List of UK by-elections" from say, the article on UK by-elections, which could be linked from the article on UK politics, which could be linked from United Kingdom. Obviously this would be very cumbersome for a very small wiki; but when discussing and thinking about my proposal, keep in mind that the english Wikipedia boasts over 2,5 million articles and that my proposal is intended for the english Wikipedia only. As such, the list of UK by-elections already exists, and there is really no problem. What does not exist, however, is a list of students at X-ville high school, which would prevent articles about random classmates from being created. Plrk (talk) 21:54, 17 September 2008 (UTC)[reply]

I am grateful for your response, many thanks. doktorb _words^deeds 16:44, 18 September 2008 (UTC)[reply]

But you do not agree with my proposal? Plrk (talk) 19:39, 18 September 2008 (UTC)[reply]

I do have some problems with it. But there are very few editors who would spell out responses in such considered ways. I do share some of your concerns but am not sure it is the right way to go round it doktorb _words^deeds 19:44, 18 September 2008 (UTC)[reply]

Walled gardens aren't bad because they are easily identifiable by bots. Once identified, they may be evaluated for notability and either improved or AFD'd. Your proposal will make them invisible: if there must be a link from outside, the gardener will insert it somewhere ... I do it all the time :)) NVO (talk) 06:14, 14 April 2009 (UTC)[reply]

Comments on "The statistics"

I agree that there is a correlation between articles that are pre-wikilinked and notability. Still, I have reservations about this proposal. First, did you check to see if those incoming wikilinks were there before or after the article was created? Second, some of those patrolled articles wouldn't have been created under the revised system (the 5 of 46 that were not pre-wikilinked). That means we might not have these apparently notable articles if they needed to be wikilinked first.

Third, is there any way to check who created those A7-deleted articles? I would bet the average age of the accounts that created the A7-deleted articles is younger than the accounts that created the patrolled articles. Assuming this is the case, this only shows that the more experienced editors (the older accounts) are:

more knowledgeable about what is worthy of inclusion (and what is not)
more knowledgeable about how to increase exposure to newly created articles (e.g. wikilinks to-and-from related articles)

In other words, this is a case of experience with Wikipedia policies and guidelines. Introducing an intermediate step will be annoying and discouraging (and, for inexperienced editors, confusing), and will probably drive away some good editors (both new and old). We should first be informing inexperienced editors with deleted articles about how they can recreate the article and have it kept (e.g. by asserting the subject's notability, providing reliable independent sources, …).

I'm not trying to crap all over this idea just to get you down; I agree with the principle behind it—to prevent walled gardens and act as a filter for bad articles. I am just looking at this from different angles so that we don't add further restrictions on editing. — Twas Now ( talk • contribs • e-mail ) 18:58, 21 September 2008 (UTC)[reply]

I'll adress your points one by one:

One, no, I did not - that would require a lot more work than I was ready to do at the time. However, as my sample consisted of very recently created articles, I believe most of the in-links were there before article creation.

Two, indeed. The question is: is the risk of losing those 11% worth the chance of getting rid of 80% of speedied creations? I'd say yes. The other major restriction on article creation, that only registered accounts can create articles, probably makes us lose very many notable articles too (but it does get rid of a lot of crap).

Three, indeed "worthy" articles are probably created by more experienced contributors. That's why there's a variant on my proposal - let only autoconfirmed accounts create freely. This would force people to make at least ten legitimate edits and wait four days before they could create an article out of the blue, but anyone could still create a linked-to article. (And, by the way - as I am not an admin, I can not check the age of the article creators' accounts.)

Plrk (talk) 20:12, 21 September 2008 (UTC)[reply]