Wikipedia talk:Wikipedia is failing

Archives

/Archive 1

/Archive 2

/Archive 3

/Archive 4

This talk page is automatically archived by Werdnabot. Any sections older than 5 days are automatically archived to Wikipedia talk:Wikipedia is failing/Archive 4. Sections without timestamps are not archived.

What to do about it

The essay (intentionally) leaves the solutions unstated. If there is a lack of quality articles, what should be done about it? CMummert · talk 02:12, 13 February 2007 (UTC)

I hope that if I can persuade people that there are serious problems, there might be an appetite for radical change. These are my ideas.

• Unassessed articles should not be publicly visible, so that some form of assessment is compulsory for any article whose authors want it to be seen.
• Assessment should be a multi-stage process. Basic assessment would simply check for grammar, spelling, style and layout. Intermediate assessment would check for comprehensiveness and appropriate length. Advanced assessment would include rigorous checking of facts.
• Anyone would be able to edit an article during its assessment, but only trusted editors could edit articles that had passed all assessment stages.
• Articles that do not have any references a week after they've been created should be speedily deleted.
• Stubs that are still stubs six months from today should be deleted.
• Unreferenced articles that are still unreferenced six months from today should be deleted.

To me, the central cause of the problems I've outlined is the incentives in place. The fact that whatever you write is instantly visible on one of the most popular websites in the word provides an enormous incentive to write, but no incentive at all to write well. The FA process provides far too little incentive to write well. We need to change the incentives, and if articles did not become publicly available until they'd been assessed, that could turn the situation around very quickly. Worldtraveller 15:32, 13 February 2007 (UTC)

I agree with your notion that there needs to be some kind of incentive for good writing, not just writing in itself.

However, I would counter your first suggestion with the fact that most new pages are at the very least screened for nonsense, hoaxes, completely non-notable, etc by people doing new page patrol, which by the limited numbers of new articles, is possible to do for each page. A new policy of non-visibility could very well discourage new editors from joining, which over the long term is likely worse for WP than a few articles that slip through the cracks.

I would also disagree with your last point about deleting unreferenced articles in a specific time frame. Unfortunately, many unreferenced articles find there way to AFD after a while and are deleted, but lack of references is not in itself reason for deletion, so long as such references do exist.

Certainly a new assessment process for all articles (some of which occurs in Wikiprojects, Peer review, FAC, etc.) could be of benefit. Limiting the number of editors may work well once you get to the FA level, but if you move that bottleneck farther down the editing process, you risk choking off the process of improvement. Joshdboz 14:58, 14 February 2007 (UTC)

On the references point, my idea about deletion, if implemented, shouldn't really result in many deletions at all. Ideally it should just result in every single article getting at least one reference. We cannot afford to allow completely unverifiable content to just sit around indefinitely, and it would just provide a very strong incentive for articles to be made verifiable. I really don't think we need to worry about numbers of people joining, seeing as Wikipedia is among the 20 most popular websites in the world. What we need to worry about is quality writers joining, and I think non-visibility of unassessed pages would encourage much better standards from the off. Worldtraveller 16:05, 14 February 2007 (UTC)

Instead of outright deletion unsourced and otherwise suspicious articles could be hidden from non-registered users until the problems are fixed.--JyriL ^talk 13:10, 15 February 2007 (UTC)

Here is what I suggest.

• Articles should go through a stepping process in which who can edit the article changes
• Articles that don't exist yet or are essentially stubs should be open to anyone
• During this stage if the community views an article as important they can submit the article for expert review, otherwise an article remains open to all
  • It may be that an article remains this way forever due to lack of professionals willing to edit the article or because the community wants the article that way
• Once an article is in expert review it is essentially locked, no one but the person(s) in charge may edit the article. Someone or a group are given responsibility for editing/adding to the article for content, clarity, and truthiness
  • Here the talk pages become more important. People who want to submit to the article but are not in charge or in the group can do so through a special section in the talk pages
  • During this stage if the reviewer(s) become errant people can vote to have them kicked out and someone else given charge of the article
  • Stagnant articles that do not move to the finalized stage can be moved back to the open to anyone stage
• Once an article has grown large enough and the reviewer(s) are satisfied an article may be voted and moved to the final stage. Articles in this stage are featured.
  • This is when an article is given to a governing body to maintain. Whether this is a hierarchy of body -> maintainer -> community, where the community submits new information and edits to the maintainer who reviews and adds it but who answers to the governing body, or its body -> community I dunno which is better.
  • Discussion and disagreements with the article are again made within the talk pages
  • An article in this stage can be voted to be moved back to review stage.

I think all important articles should move to the final stage, however, the rest should remain open to all as community articles. Wiki in my mind is not just about competing with big named encyclopedias, it is about the community and working together to share information and opinion.

Dharh 16:12, 14 February 2007 (UTC)

I think that the section below addresses almost all of these points.204.193.75.21 19:17, 14 February 2007 (UTC)

The problem is that most articles are written by only a few people, and in some cases only one editor is responsible for most of the content. Certainly you can say that once an article reaches a certain point it should be placed under the authority of a wikiproject or some group akin to that, but that it would likely become unwieldy if thousands of articles are restricted in this way. The real problem I see with bringing many articles through assessment stages is this: Because of the number of articles that would need to be assessed and then maintained, the process would have to be decentralized. But who decides which editors have the authority for which articles? -- Joshdboz 17:36, 14 February 2007 (UTC)

I think you have identified one of the main problems: Where is the incentive to write good articles? And so, in my naiveté (for I am unfortunately not a regular contributor here) I propose a solution: put people's reputation and voice at stake for writing poorly, or rather inaccurately. As a very frequent user of Wikipedia I often don't care if an article strays from the standard formats and style guides slightly, but that the information is accurate and that the immediate source links at the bottom of the page are up to date. So each person gets one account and one account only. They can have it display their real name or not, it doesn't matter so long as they're held responsible for what they do. Vandals, well intentioned or not, can be flagged so their edits don't get used until some peer review process approves the edits. Or some set of editors can sift through a list generated by the flagged user edits. Maybe even have a list for first time editors, and one for regular commiters? So the incentive comes to exist by the nature of accuracy. If someone is not accurate or relevant in his edits, then they lose the privilege to freely edit all topics.

I don't mean to go beyond the scope of this discussion, but I am starting to see a need for centralized knowledge-building structures. In other words, a place to find Truth. I think Wikipedia is well poised for such an endeavor if we can get past the issues at hand. I look forward to doing what little I can to make it happen. J.H. Gorse 18:38, 14 February 2007 (UTC)

I think most articles should remain outside FA. Some topics on wikipedia just do not lend themselves to the same ideals of encyclopedias. Anime and Manga for instance are huge sections in wikipedia with at least a hundred articles covering different shows and books. There is no reason for these to fit into the restrictive guidelines being discussed around here.

Even if someone is given final say over a specific article or a set of articles theres no reason why peer review and user participation can't happen. I think thats what talk pages can be used for. If the person in charge flakes out or is seen to be biased they can be ejected.

Another idea I had was that a given state of an article could be flagged as factual or of a certain quality such that even if it is edited and messed up, a link to that particular state shows up at the top of the article. Or edits do not reflect on the article until they are deemed satisfactory, again showing a link at the top of the article to the draft page. Dharh 07:05, 16 February 2007 (UTC)

I agree with you, Dharh. We could follow a simplified software development cycle tags such as alpha, beta, RC, GM, current, stable, release, whatever! So long as it lends itself to the flux of Wikipedia development, why not tag it? I did a little more reading and found that someone already proposed something similar and got shot down: http://en.wikipedia.org/wiki/WP:STABLE J.H. Gorse 01:32, 20 February 2007 (UTC)

Re: Wikipedia is failing

Almost every time I search through another encyclopedia for my computer science research, I come up empty handed. Wikipedia almost always has something about what I want to know. It may not always be in the right format, or meet some silly guidelines, or even give me all the information I need to know, but usually there is at least something. Thus I can use Google to continue my research, which has even more breadth but much much lower quality.

I wouldn't be nearly as effective without having Wikipedia as a starting point. —The preceding unsigned comment was added by 70.41.124.25 (talk) 14:14, 14 February 2007 (UTC).

This is an interesting point that I wish more editors would pay attention to. WP has reasonably thorough and accurate introductory articles on many science and mathematics topics. It is becoming more and more common to hear speakers at scientific conferences say that they looked at WP to get a rough definition of some concept they are not familiar with. WP beats EB both in terms of coverage and in terms of depth. For example, group (mathematics) (a fundamental concept in undergraduate mathematics) is a single paragraph definition in EB, and EB returns no results at all for the more advanced, well-known topic groupoid.

If the rest of WP is failing, as claimed, maybe it would help to analyze why the science and math sections are doing so well? CMummert · talk 15:11, 14 February 2007 (UTC)

Umm, it's pretty obvious. The science and math sections are doing well because Wikipedia is Internet-based, and Internet use is going to be associated with science and math (and especially computer science). Ken Arromdee 17:08, 14 February 2007 (UTC)

Actually, there are two much more basic reasons:

• Science and math topics aren't so often cited at junior high school levels, so they have a much lower rate of vandalism.
• There is much less PoV-pushing, and more of what there is isolated and isolatable cranks pushing a pet theory. Fortunately, the genuinely divisive issues, like the status of intuitionism or category theory, have not yet inspired a crusading editor to Tell the World the Truth. Septentrionalis PMAnderson 14:24, 20 February 2007 (UTC)

Yeah, I agree. I find it particularily worrying that articles like this generate so much consternation. We need to stop focussing on what one person thinks the wiki is doing wrong, and focus on all the things that we all think the wiki is doing right

Sure there's garbage on the wiki, maybe even a lot of it. But so what? There's also some absolutely amazing articles, articles that have no peer anywhere else in the world, either online or in print. To ignore this fact and focus instead on the problems does everyone, not just use wikipedians, a disservice.

And let's be honest, the metric used in this article is arguably useless. GA has been backlogged forever, so using that as a measuring stick is silly. One might then argue that the GA process is flawed, but who cares? Just because someone didn't stamp an article with GA doesn't mean it isn't a GA. And looking over the list in question, I consider every one I saw GA.

So basically, another problem in the making. Maury 15:18, 14 February 2007 (UTC)

How is focussing on problems a disservice? When I was studying for my A levels and took a mock exam, I didn't focus on what I got right, but what I got wrong. There's no point sitting smug and ignoring all my faults. By acknowledging and correcting those faults, I got a better score the next time. It's vital that we critically assess Wikipedia's progress to ensure it's being made in the right way. I'm not saying that what's been done so far isn't good; I'm saying, let's make it better. --Seans Potato Business 01:36, 19 February 2007 (UTC)

I think I have to expand a bit on this, because the more I read the article the more I disagree with everything in it.

"In fact, of those 1182, only 72 are featured articles. This means that 94% of the essential topics that should have excellent articles fall short of the standard."

No it doesn't, it just means they haven't been FA. FA's are one a day. There are over a million articles on the wiki. This isn't exactly surprising. In fact, its specious.

"131 are listed as good articles, which, according to Template:Grading scheme, means that 'other encyclopedias could do a better job'."

• sigh* See the first point above. I consider it extremely unfortunate that grading implies a hierarchy with FA at the top and GA below. GA's, IMHO, are just as good as FA's, they're just on topics that aren't important enough to go up on the home page. HiPER meets the criterion, but I would never put it up for FA because the topic is too obscure.

"The rest, presumably, are B-class or start-class on the assessment scale" "This means that slightly more than 99.8% of all the articles on Wikipedia are not considered well written, verifiable or broad or comprehensive in their coverage."

It means nothing of the sort. These (and similar exampels) are grossly misleading statements. Simply put, judging the wikipedia by tagging is a very bad idea. Consider, for instance, the very first article on the VA list, Paul Cézanne. It's not tagged at all. And it's a fantastic article.

Maury 15:37, 14 February 2007 (UTC)

The very problem in the first place is there isn't enough focus on what might be going wrong. Just saying 'but some articles are great!' does the project no favours. Sure, some articles are great, but as the numbers show, they're the vanishingly small minority. If the metric used here is 'arguably useless', well, what's a better one? Sorry, but your assertion that Paul Cezanne has an excellent article is not the kind of thing we can use to judge the overall quality of Wikipedia. Similarly GA's, IMHO, are just as good as FA's doesn't help us judge accurately how successful the project is.

As for whether science is actually doing that well, look at WP:VA. According to that, 19 of the 216 most essential science articles are FAs, making 8.8% instead of the overall value of 6%. Mathematics has two out of 59 - 3.4%. Remember, the aim here is Britannica quality or better.

If, as seems to be the claim here, there are large numbers of high quality articles, where is the assessment that shows that those articles exist? WP:1.0/I shows that out of 290,000 articles assessed, 186,000 (65%) are stubs, 77,000 (27%) are 'start class', 22,000 (8%) are B-class, and 1255 (0.4%) GAs. Just over 2,000 are defined as A-class or FA - 0.7%. How can this project be seen as successful when just 0.7% of all articles that have been assessed actually count as high quality? Worldtraveller 15:45, 14 February 2007 (UTC)

It seems the point has been missed.

My point is that using FA and GA tags for statistical puposes will fail because those processes are not working as well as they arguably should be. It says nothing about the wikipedia or the articles on it. Remember, GA requires active work on the part of editors who vote on articles that they are not a part of. There's an anti-reward, and as a result it sees little attention. Most articles that I consider great to fantastic never appear on GA in the first place. Check out rope trick effect. It's a fantastic article. Now is it "less fantastic" because the author didn't put it up for GA? No, of course not. Yet that's precisely what this essay claims.

Further, the grading process is relatively new. As a result it is not being widely applied to older articles. Yet those are precisely the articles that are most stable and more likely to be GA or high-grade. And to top it off, you repeat the confusion between quality and FA. FA is not about quality. I've written hundreds of high quality articles that I will not propose for FA, because I really don't think zgrass or HiPER is of wide enough interest to be on the home page. FA implies both quality and many other things, and using it as a metric of quality alone remains a serious problem.

I'll reiterate: using FA and GA is a bad metric. Tagging only slightly less so. That I can't offer a better one doesn't change anything. Maury 17:10, 14 February 2007 (UTC)

If we want to discuss whether Wikipedia is succeeding or not, we do need some objective way of judging that. The table at WP:1.0/I seems about the best we can do, with a sample of 18% of the encyclopaedia assessed. What we see is that there are vastly more articles at the bad end of the scale than at the good end. Where are all the high quality articles that you say exist? If there were actually large numbers of great articles, but no-one could find out which ones they were, that would also be a failing.

As for rope trick effect, well, it seems rather similar to this web page. It doesn't seem to offer anything that you can't find elsewhere, and it doesn't provide any references, just external links. It's not bad but it's not amazing by any means. Worldtraveller 17:36, 14 February 2007 (UTC)

Yeah - I don't think that's a great example. But look at articles about cars - there are thousands of them, only three are FA's , three ex-FA's and less than a dozen are GA's. Are they all terrible articles? By no means - pick one at random: Audi A8 - not a terrible article. A bit light on references - but then who writes learned tomes about recent cars? Still - it wouldn't take much to make a GA out of it. Trust me, nobody will ever try. Wikipedia:WikiProject Automobiles/Assessment#Worklist is a good snapshot of car articles...look at all of those B-class articles - about 80 of them. The assessment scheme won't push any of them higher than B-class until they pass GAC - but look at them - they are all really close to GA quality. Why aren't there 80+ GA's in the automobile project area? Simple - it's too much pain to go through the process. The author of Mitsubishi i actually told me that it was too much bother to deal with PR, GAC, FAC - I had to work quite hard to pursuade him to take this great little article through the process. That's a GREAT article - 25 references, nice English, nice photos - good mix of technical stuff and interesting background reading. It should be an FA...but it's overwhelmingly likely that it won't ever be. SteveBaker 18:03, 14 February 2007 (UTC)

Whether we need an objective "measure" on the success or failure or not, the one you use is a piece of rubbish for the aforementioned reasons. For some reason that defies logic, some other people actually bought the rhetoric and then started the cliched discussion about how to improve wikipedia, including restricting access to editing, which doesn't even have anything to do with the "problem" you decided was so obvious! Sir, the measuring stick you use is useless, and pretty much the results you come to are as such completely bogus. There may or may not be problems to solve here, but this discussion is so... useless and besides the point. Zalle 18:10, 14 February 2007 (UTC)

One more point: "If there were actually large numbers of great articles, but no-one could find out which ones they were, that would also be a failing." If somebody needs to read an encyclopedia article on a particular subject, they'll read it and perhaps think then that the article is good - but no one but encyclopedia geeks want to "find out which articles are good" so that they could then read those articles. You nor anyone else need to tell the reading public at large that some particular articles are good. They'll figure it out for themselves.Zalle 18:20, 14 February 2007 (UTC)

I'm sorry Worldtraveller, but I can't accept that you're still trying to draw conclusions based on what even you seem to admit is questionable evidence.

The majority of articles on the wikipedia are not accessed. The vast majority of articles have not been, and likely will never be, offered for GA. The completely overwhelming majority will never be offered for FA. Yet you're still suggesting we should use those statistics as an "objective measure", based on nothing more than the fact that you don't have anything better. Are you serious? If you don't have anything better, you don't draw conclusions. That much should be obvious to anyone, least of all someone who appears to have some sort of technical background.

And as I suspected (and the reason I chose the particular article), you "fail" rope trick on one point due to a lack of external references. If I added one, would the article suddenly "improve"? Only if one narrowly defines "better" based on some completely arbitrary definition that ignores the content of the article and bases it instead on the format. And sadly, the tagging process is being driven almost entirely by editors that do make this fundamental mistake. This confusion over the difference between process and product is precisely what I think makes the assessment process so inherantly flawed.

And yet your flawed analysis has been Slashdoted and is being read by thousands of people. People who are going to question my hard work because of your personal musings. That's simply not fair.

Seriously, if you really believe that the wikipedia is "failing" then I can only shake my head in astonisment. The world loves the wiki. And as Zalle so eloquently pointed out, they're the only people who matter. Maury 19:26, 14 February 2007 (UTC)

No, I don't think it's questionable evidence at all. Perhaps not perfect but what evidence would be, short of independent assessment of every single article? You seem to be trying to say that there are orders of magnitude more quality articles than the number of FAs, A-class and GA articles, but you can't offer any evidence to back that up at all.

Yes, if you added a reference to Rope trick effect, it would be demonstrably better, because the information in it would become verifiable. If you re-wrote it so that it wasn't so similar to the text in the external link, it would be objectively better. Your definition of what constitutes good is not the same as the widely held view, arrived at through discussion and set out at {{grading scheme}}.

Don't get too concerned about the slashdotting. Many people are capable of reading an essay and considering for themselves whether or not they agree with it. If people are going to question the reliability of the encyclopaedia, they're also questioning my work. The world may love the wiki today, but that doesn't mean they will always love it. The world loved the Bay City Rollers once. Worldtraveller 23:11, 14 February 2007 (UTC)

I'm really sorry for the ad hominem, but at the moment I have to say you're a textbook example of why editing should be restricted. Dude, just because someone "can't offer evidence" does not mean that the point they are arguing is not valid. People have explained to you several times why there aren't any more "good" or "featured" articles, and yet you just go "la la la I can't hear you." Do you even understand why there possibly can't be more featured articles than there currently is? And the most amusing part is that you seem to think that an encyclopedia should be called good or bad depending on how many articles with completely irrelevant labels it has. Have you ever considered that Britannica has _zero_ featured articles? Or do you think that they're all "good" enough to be featured articles? If you do, I have a bridge to sell you. Oh, and indeed, no evidence short of assessing each and every article is proper evidence of the quality of an encyclopedia. That's why this is so idiotic. Thank god this doesn't really matter. Zalle 10:01, 15 February 2007 (UTC)

Zalle - I've written more featured articles than you've made edits to the encyclopaedia. Worldtraveller 18:27, 21 February 2007 (UTC)

I agree with the observations by Maury. FA statistics might have some bearing on the health of the FA initiative, but offer little in the way of assessing the health of WP as a whole, nor do I think this is the right entree into the debate. To begin with, there are two distinct judgements to form: the health of the WP content on its own merits, and the health of the WP community and political culture around that. In other threads a comparison is made between WP and other open source projects. Well, one of the fundamental differences between WP and open source software projects is that software projects can enforce mechanical regression testing, in multiple domains (coding style, correctness, performance, complexity of expression). In WP, the vast majority of quality criteria are presently enforced by human culture. There are plenty of voices here for the proposition "let's lock out the bad guys" on the premise that without so many steps backward that quality will ascend monotonically, or in some fascimile of such. But in software, it's never that simple. The well meaning contributor who contributes a performance enhancement can incidentally impact correctness in a negative way, or any other complex tradeoff of "one big step here versus many accidental or small steps backward elsewhere". An example of a POV-pusher in open source is someone who contributes a performance patch for x86 while compromising portability to every other platform, or a performance patch for one device driver that negatively impacts all other devices within the same driver family. As I see it, WP is a long way away from effective regression testing. For one thing, a software regression test doesn't throw away undesired edits: they get built into the test framework to be rejected automatically in future. It's no surprise to me that when WP deletes an unwanted topic, it keeps coming back. Finally, in open source quality tends to progress as a fractal. FreeBSD recently end-of-lifed support for their 4.11 release, yet a large vocal community still regards 4.11 as superior to all 5.x releases, and all releases thus far in the 6.x series (though holding out hope that the 6.x could potentially regain the quality crown). On the other hand, the 4.x series was never going to thrive in the aggressively multicore reality we now face, so the intensive and disruptive rework in the 5.x series was necessitated by the long term view. In the face of this reality, the goal-line-ism of FA status as a primary quality metric can only manage to create a hidebound political process, which would be a far larger failure than the issues we face now. MaxEnt 19:41, 15 February 2007 (UTC)

It seems a common viewpoint that the number of FAs doesn't say anything about the overall standard here. But what numbers do we have, besides FA and A-class totals, to even estimate how many excellent articles there are? If we simply don't know how many there are, then that's a failure as well. If you could design an ideal system to estimate the current standards of quality of all articles, what would it look like? Worldtraveller 18:27, 21 February 2007 (UTC)

I found something the other day which might help with that: User:R. fiend/How many articles does Wikipedia really have?, which uses random sampling to estimate the number of articles in several loosely defined "categories". The same metodology could be used, with only small changes, to get the answer to your question. --cesarb 21:36, 21 February 2007 (UTC)

Redirect

This article should be redirected to WP:NOT#Failing. — BRIAN0918 • 2007-02-14 15:41Z

Highscores and status rewards motivate people to excel - let's use that!

I'd like to seperate the problem (which exists, but does not bug me much) into two areas.

• Continued interference from vandals and agenda-pushers.

A lot has been done, successfully, to limit the impact of dumb drive-by-edits. The vandals and agenda-pushers that remain, being much more highly motivated, haven't left. And they won't. Get over it.

Every system is hackable and the Wiki continues to grow in influence/usefulness. Nothing but the most draconian of measures will keep those people out, and even if some sort of super-safe solution could be found, the reputation of Wikipedia would only rise, and the incentive of hack that system and use its reputation would rise with it.

So this part of the problem won't be solved and we should, accordingly, focus on the other.

• Edits from well-meaning but uninformed people.

As others have observed, some haven't read the guidelines or don't understand their importance. Others don't understand their subject well enough, and that specifically includes the inability to distinguish between facts and the things they've been told are facts, but aren't.

These people are not the problem, their lack of understanding is. From experience, I assume that in many cases lack of understanding is caused by a more fundamental problem: their lack of understanding of the fact they may lack understanding. Unknown unknowns. Psychology has a word for it: optimism bias or the tendency to see the outcomes of ones actions overly optimistically.

This problem is solvable, because it only happens when people need to make their own predictions about their competence. People just need to be reminded, prominently, that there are more competent editors as well as less competent ones, and that they have (only) a certain competence although it may not be measurable in any precise way.

The logical conclusion is some sort of editor rating system. I disagree with those on Slashdot who have said their karma system would be helpful. People are complex, and if they and others are to use a rating system for estimates of their competence in a broad range of areas, the system better be complex. I have in mind something like a list of user templates, which are mandatory parts of user's personal pages. They could look like this:

Status bar for user RandomWikipedian
A member for 3 years, rank 34,120 on the Helpfulness table.	This user has made 156 edits, averaging +72 bytes.	6 edits (4%) by this user have been reverted.	This user has never been banned.	This user has not completed the tutorial.	This user is a major contributor to 7 articles.	This user is a major contributor to 1 good article.	This user is a major contributor to 1 featured article.	This user contributed 7 images.	This user contributed 1 featured image.

Or something like that. The details would be to be figured out. I'd just like to emphasize that this doesn't just give people clear measures. It is also a rewarding system, and according to all modern learning theories theories, it is thus much more able to produce positive learning effects compared to restricting/punishing systems, especially in the long term.

This is important because almost all ideas for changes previously given focus on imposing some sort of restriction on the current, extremely free editing process. Even requiring the approval of an appointed expert before an edit goes live, is a restriction and a punishment because it will be perceived as an insult.

Restrictions emphasize a conflict of interests and thus cause all sorts of friction, mostly in the form of complaints and negative publicity. Reward systems focus on solutions, on possible avenues for cooperation. And since we want cooperation, not just absence of disruptive behaviour, the system sketched above is much more likely to make Wikipedia succeed than any sort of "protection".Denial 16:33, 14 February 2007 (UTC)

It's an interesting idea, but doesn't it already exist in one form or another. Under the above system, anyone could still edit, but it would be easier to differentiate between experienced and amateur editors. But today, as long as editors update their user page, it's pretty easy to get a good idea of who has been here a while and who is brand new. Wouldn't experience have to be linked with greater capability (like admins today) in order to actually affect the editing process? -- Joshdboz 17:43, 14 February 2007 (UTC)

I agree that is a good idea. I am not sure if the stats could be retroactive or not, since I am unaware of what, if any, statistical data is recorded on its users by Wikipedia. But, what I want to call the "Wiki Editor Card" (based on the Xbox Live Gamercard), would be a grand idea. I don't know if it should have any weight on what the user can and cannot do - but it would at least be somewhat of an incentive to do good work on here. Chad78 17:50, 14 February 2007 (UTC)

I'd like to suggest a more fine-grained approach to editorial differentiation: Editing points (EP) and Editor levels. Give each user points based on work that they've done, with deductions for various forms of "bad" behavior, and assign users to levels proportional to log(EP). Unlike karma, there will be no upper limit to the number of levels. Allow editors to lock articles so that they can only be edited by users of the same or higher levels. This would provide a hierarchy of articles with assurances that quality work is rarely vandanized. It would provide an incentive to create and use user-ids to allow ones work to be properly creditted to them, while still allowing the occasional "drive-by" edits of non-mainstream articles. Finally, it would keep all of the "fun" aspects of WP, such as revision wars and POV battles, but with a much smaller number of possible participants. 204.193.75.21 18:01, 14 February 2007 (UTC)

After a bit of thought, I'd like to suggest that a user's editorial level be equal to int(log2(1+nbr_of_edits-2*nbr_of_reversions) + sqrt(nbr_of_barnstars)). Logged-in users would have the following special abilities:

• can lock any article against edits by users with a lower level
• can permanently unlock any article against locking by a lower level user
• can award barnstars to users with lower levels

By adding barnstars into the equation, we allow quick recognition of new users. OTOH, even someone with a billion edits would only be at level 30, and winning a hundred barnstars would only push this to level 40. Hard-code Jimmy Wales at level 50, and life will be good. 204.193.75.21 18:53, 14 February 2007 (UTC)

Those are interesting ideas, but they do introduce conflicts of interest. This may be non-obvious, so please bear with me.

Any inequal distribution of power, however many advantages it has, means some have power that others do not have. Those others perceive their relative lack of power as an insult ("they don't trust me"), i.e. as an attack. This perception happens regardless of whether it "should" happen or whether people "should" accept they are entitled to less trust. It causes hostilities between the inequal groups, as I believe is already apparent between admins, registered users and anonymous IPs. Don't you, too, see most people minimize their communication with members of the two groups they don't belong to? To have more dividing lines inside the editing body, and have people start to treat each other as inequals, would predictably tempt each editor to identify with his/her group rather than all Wikipedians. This provokes decrease of cooperation. I assume that privileged groups would increasingly concentrate on articles that only they can edit, perhaps produce a respectable increase of good and featured articles, while the "lowly" others (and make no mistake, everyone recognizes the less trusted for what they are) become increasingly frustrated and disruptive. It works in states and economies, to some degree, but it doesn't work on a web site that everyone is free to leave.

Fictional quote: "Why can't I edit this article anymore? I built it, I guarded it, I extended it, now you privileged user types come and take it away from me! This place is full of (expletive)! I'm leaving!"

Systems of inequality have historically shown themselves to gradually become ever more inequal until a major crisis disrupts the whole system. In economics: a crash, in politics: a revolution, in religion: a schism, in Wikipedia: probably a fork. Although that may be the best solution for everyone involved, I don't think it is the best solution for the vast majority that is not involved: the silent masses that we produce free knowledge for. To best act in their interest, the potential for conflict inside Wikipedia must be as low as possible. I think a pure reward system that gives people status incentives but does not restrict/insult them can help a lot with the problems noted in the article without introducing more potential for conflict.

It is also not just a ego-stroking system, and not just a measure of how much to trust an editor to be right. Its most important effect would be to make people aware of their merits and capabilities, and of the limits that these have. It tells the well-meaning but clueless, who don't know they are, that they are. It also tells them the guy who reverted their edit may know what he doing - not just "because he can" using his privilege. Denial 20:14, 14 February 2007 (UTC)

I sincerely hope that status bar thing above was not meant to be serious. In particular, that bit about "average edit size". Suppose I revert a vandal who just added 1,000,000 bytes of nonsense to a page. My average change size will increase rather a lot. Now suppose I revert a vandal who blanked a very long page; a similar thing will happen. Such a metric would be useless – Qxz 07:56, 18 February 2007 (UTC)

You failed to notice that this was just an illustration. "Or something like that. The details would be to be figured out." is what I wrote. On a more basic level, highscores and status rewards motivate people to excel, so why not use that effect? It'd help, probably a lot, and there is no need to tie in new (and certainly disruptive) distributions of power. I do not claim to be qualified to design the exact shape of such a system. We could distribute:

• points in a wiki-wide ranking of contribution excellence
• medals for the creation of FAs
• honor badges for reverts of vandalism
• special titles for donations
• various statuettes for winners of competitions

Or whatever else is both agreed on and technically possible. The implementation seems rather trivial to me, although it'd increase server load. Denial 20:13, 21 February 2007 (UTC)

But we could do that no matter what. If you think it should be done, go ahead and start doing it. That wasn't your idea, though; your idea was to require people to advertise certain things on their userpage. That is not just a reward, because for whatever you put there, official requirement implies that the things listed should be used to judge an editor's value. And no system will fairly do that. For example, your "illustration", even discounting the useless things, almost completely fails to recognize that I have any value. I've been here for 5ish months (not that much), I've never been banned (which should not be viewed as exemplary conduct; being banned means you did something bad, not just that you weren't exemplary), I had never seen the tutorial before you mentioned it and I tried to figure out what you were talking about, and I have no sort of featured content at all. The only thing that could concievably be viewed as showing that I'm good is about 6000 edits, and even that is really hard to care about in the age of massive AWB. -Amarkov moo! 20:30, 21 February 2007 (UTC)

Distributed Proofreaders

I agree with many of the sentiments raised in the article. In my opinion the problem of vandalism is not stressed enough though. While on recent change patrol, I find that about 10-20% of all anonymous edits are unconstructive, if not outright vandalism. There is an anonymous edit every 2 seconds, sometimes more. You do the math, there is a lot of dedicated patrolling needed to repair such a noise level. (I only do anonymous edits, I'm sure the problem is similar with registered users.)

To stress again, the wikipedia ideal of 'encyclopedia through natural selection' is mostly failing because of vandalism pressure. If you look at the edit list of almost any important page, about 2/3 of the edits will be either vandalism of reverts. Some pages are significantly above that number.

Yet it can work. In my opinion we need to look at the Distributed Proofreaders project. For those who are unaware of the project, its a community effort to proofread the ocr output of books. Its output is contributed to Project Gutenberg. Although the potential for vandalism is there just the same, they deal with it effectively. For example they have a hierarchical layer of functions. Each new level has its entrance requirements. Some of those requirements are in terms of number of edits, number of days since registration etc. Others require peer review or taking tests on the knowledge of the guidelines. Atmittedly it gives the project sometimes a bit formal atmosphere. One other thing they have is that each book has a project leader.

In the end it gets the job done and the system is wildy succesfull: producing books at a high sustained rate, and with systematic high quality. One could wonder though wheter it is possible to lift wikipedia to such a higher level of quality control, at this stage of the work.

For me, I would like to see a new phase in Wikipedia of professionalism.

Sander123 08:35, 15 February 2007 (UTC)

Yes I agree, but I think vandalism is just the tip of the iceberg. The well-meaning but misinformed edits are far more insidious and much harder to deal with, particularly if you don't have an someone with significant knowledge of the topic constantly monitoring each article. (MichaelJLowe 11:03, 15 February 2007 (UTC))

I think we must create a some sort of review system, where a trusted expert reviews an article. When the article is found to be accurate, the reviewed article is given a "verified" status. After the article has been changed, a link to the trusted version should be available on the article in a visible place. This could make Wikipedia more reliable and give readers a change to read the actual featured/good article instead of a possibly messed up version. Reviewing must not be limited to top-quality articles, even stub articles could get a "verified" status. The point is that the article must be reliable.

What comes to anonymous vandalism, I think it is too serious issue not to be dismissed although vandal edits are usually easy to revert. It is very frustrating to revert several articles in row. Every time an anonymous editor has changed a page, it has to be checked if he is a vandal. Really, if somebody wants to edit Wikipedia, it should not be too hard to register first.--JyriL ^talk 14:10, 15 February 2007 (UTC)

Exactly. Informally we really have a system like that already. If you see a doubtful edit, you check the history of the person now as well. If that guy has thousands of quality edits you are more inclined to giving him the advantage than if he has only ten poor edits to his name. What distributed proofreaders does is make this process a bit more formal and visible. Sander123 08:07, 20 February 2007 (UTC)

Thoughts

Like many others I think there are issues, but that this essay goes too far and sometimes misses the forest for the trees. For instance, "There are about 1,300 featured articles. There are also about 1,700 good articles. However, there are currently 1,639,963 articles on Wikipedia. This means that slightly more than 99.8% of all the articles on Wikipedia are not considered well written, verifiable or broad or comprehensive in their coverage."

That conclusion seems clearly false. The cited statistics do NOT mean that 99.8% are not considered well written / et cetera. It means that 99.8% either are not considered well written OR just have not been assessed yet. Based on percentages seen in the large 'article assessment' projects I'd argue that the 'not assessed' portion is significantly greater than those 'assessed to be poor'... and a significant portion of those un-assessed articles are good.

Greater focus on pop culture is true, but not IMO a problem... because the number of 'core' encyclopedia articles of high quality is greater NOW than it was three years ago. 'Failure' would require that the overall core article quality be going down, rather than up.

Degradation of quality articles over time happens in some cases, generally when the growing user population overwhelms compromises on contentious topics which were forged by the handful of earlier editors, but it is not a universal problem and work on 'stable versions' is ongoing and presents a viable solution to this issue. 'Failure' would require that this problem be common to ALL featured articles, and that no plan for addressing it existed.

Finally, this essay focuses far too much on 'featured' and 'good' article percentages. These are new and subjective concepts which really have little to do with the value of Wikipedia to date. Personally, I think the current standards for even featured articles to be lacking the rigor of print encyclopedias... in that we do not systematically review the references to verify that they confirm the information they are given for. However, that doesn't mean Wikipedia is 'failing'... any more than Wikipedia was 'failing' back when we didn't have 'featured articles' at all. The existence of the mere concept of featured articles is proof that Wikipedia is succeeding. We've gone beyond just people showing up and writing about stuff they like to actually having formal requirements and review procedures, which have gotten consistently stricter over time, for identifying high quality content. Despite systematic reviews of old featured articles to remove those which no longer meet the heightened standards our count of featured articles continues to go UP... not down. That is indicative of success, not failure.

Wikipedia has not 'succeeded' in establishing an encyclopedia which is of equal/higher writing and reference quality as Britannica for all articles common to both yet... but it is definitely moving in the direction of success on that front rather than failure. And on the IMO equally important issue of providing useful information on topics which Britannica and the like do not, Wikipedia has certainly succeeded and continues to build on that success. --CBD 12:46, 15 February 2007 (UTC)

This is complete nonsense, at least if we go by what is happening to the article that I most understand, and have studied most, namely the Philosophy article (since Sep 2003). Perhaps that article is unique, but I think not. The article was always degrading, but the pace is accelerating. Look at it now. The current position is 2 expert editors against 3 cranks who are determined to wear us out, by concerted action over world time zones. Of the last 50 edits, all without exception were reverts of some kind, not against vandalism, but against cranky and bizarre contributions by people with some strange theory or other. I really am trying to let go and walk away, in which case the other guy will go, for sure. Then, let the deluge begin. These guys are really determined. It is a hell of a lot of work just to defend what we have left, and if you divert your attention for a second, some other part is shot to pieces. It feels like that scene in Apocalypse now where they get to the bridge, and 'Charlie' is everywhere. They really have won. Dbuckner 13:14, 15 February 2007 (UTC)

One article. On a contentious subject. Precisely the sort of thing I described being subject to degradation. NOT indicative of Wikipedia as a whole. --CBD 21:24, 15 February 2007 (UTC)

Right, this is one article with one man's struggle to convey truth. However, this is still relevant to our discussion in that we need to systematically and comprehensively remove inaccurate and misinformed posts. Automation would be a nice plus, but not necessary. The unfortunate nature of truth is that there are always more voices for fiction to confuse the matter than there are voices of non-fiction. Therefore, we must determine a process for protecting articles against the uninformed, the misinformed, and even the maligned. Where can we have this discussion? J.H. Gorse 01:47, 20 February 2007 (UTC)

I have to wonder if this article is not a prime example of an unwanted side-effect of WP:OWN. In simple terms, if no one owns an article, then no one cares if that article declines in quality. (I'm sure that the mis-named principle, the tragedy of the commons applies here.) This is not a sugestion that we repeal WP:OWN, but that perhaps tuning that rule -- say allow a WikiProject to adopt important articles & administer some form of the often-discussed "protected version" idea. However, I'll admit that even if a "protected version" tactic becomes policy, it will address only one piece of a larger problem. -- llywrch 21:26, 21 February 2007 (UTC)

Something like {{maintained}}? --cesarb 21:53, 21 February 2007 (UTC)

Utter, utter failure

Wikipedia will fail, if some editors are so blindly in love with it that they reject any attempt to point out its shortcomings. Worldtraveller 11:39, 18 February 2007 (UTC)

Imminent death of Wikipedia predicted. People have said Wikipedia would fail so many times it's got boring already. It's much more productive to discuss how to improve the places where it isn't working very well. --cesarb 12:51, 18 February 2007 (UTC)

It's much more productive to discuss how to improve - that's what I've been trying to encourage all along. Nothing will improve if people deny that there are any problems. I didn't expect that people would so blankly refuse to consider that there are significant problems. I didn't expect to find out that possibly much more of a problem than any measures of how many excellent articles there are, would be how aggressively some editors will try to cover up constructive criticism. Worldtraveller 14:19, 18 February 2007 (UTC)

Worldtraveller, you really seem to be soapboxing at this point. Yes, maybe the sky is falling, but I have yet to find a single real piece of evidence to that, just statistics interpreted creatively. Perhaps you should get a blog for this sort of thing. ^HighInBC ^{(Need help? Ask me)} 13:23, 18 February 2007 (UTC)

What, to you, would constitute evidence of failure? Worldtraveller 14:19, 18 February 2007 (UTC)

Well, I would say some sort of information that does not require assumptions or arbitrary interpretation. I don't think you can prove it is failing, because it does not seem to be. This is a half built house, not a broken one. ^HighInBC ^{(Need help? Ask me)} 14:23, 18 February 2007 (UTC)

What sort of information would that be? Perhaps we can define it, use it and then introduce it into the essay. Worldtraveller 14:28, 18 February 2007 (UTC)

You are asking me? I don't think the proof really exists, because I think you are wrong. So I am the wrong one to ask. If you can provide any sort of argument that does not rely on assumptions and arbitrary interpretation then I will certainly consider it with an open mind. ^HighInBC ^{(Need help? Ask me)} 14:38, 18 February 2007 (UTC)

The question is really, what evidence, that doesn't rely on assumptions or arbitrary interpretation, do you consider indicates that Wikipedia is succeeding? Worldtraveller 15:01, 18 February 2007 (UTC)

I don't know, tough thing to prove, that is why I would choose a name like "How Wikipedia can be improved". The fact that you chose an argument that is difficult to prove without assumptions and interpretation does not lend credit to the argument. ^HighInBC ^{(Need help? Ask me)} 15:03, 18 February 2007 (UTC)

I think you misunderstand the point of the essay. You must realise that it presents a thesis, and evidence to back that thesis. Nothing can be 'proved'. If you disagree with the thesis, you need to provide evidence to the contrary. Just saying 'I don't know, tough thing to prove', and 'You can't prove it's failing because it does not seem to be' does not actually tell us anything at all. Why do you consider Wikipedia successful? Why do you disagree with the analysis presented here? Worldtraveller 15:11, 18 February 2007 (UTC)

I am taking this page off my watchlist, so don't think I am ignoring you. I am willing to discuss further if anyone else wants to, just drop a line on my talk page. ^HighInBC ^{(Need help? Ask me)} 15:08, 18 February 2007 (UTC)

Things which might constitute actual evidence that Wikipedia were failing:

1. Number of featured articles decreasing - rather than increasing
2. Number of featured articles on 'core topics' decreasing - rather than increasing
3. Standards for 'featured article' status becoming significantly more lax - rather than much stricter
4. Number of active users decreasing - rather than increasing
5. Donations to Wikimedia decreasing - rather than increasing

In short, any sort of quantifiable negative motion would be indicative that Wikipedia was on a "failing" track. The term for showing positive motion towards a goal is, "succeeding". Most of the arguments in this essay really translate to 'Wikipedia has not succeeded YET - therefor it is failing'... which doesn't hold water. The one claim of 'negative progress' made in the essay is 'featured article degradation', but as noted above most ex-featured articles were demoted because the standards increased rather than through degradation. Even if we pretended, like the essay, that all demoted articles WERE due to degradation it still doesn't indicate 'failing'... because we've promoted more new FAs than we've demoted. The net change is still positive. That measurement still shows that Wikipedia quality is increasing. Indeed, name any project wide measurable indicator of 'quality' on Wikipedia and current statistics show that it is going up. They may not be going up 'fast enough' for some people / by some interpretations of the statistics, but that still isn't 'failing'... it's 'succeeding less quickly than some would like'. Claims that 'it will take two hundred years to succeed at this rate' inherently concede that we are moving towards success... and seem obviously overwrought to me, as increasing numbers of users will inherently lead to decreasing completion times. --CBD 16:16, 18 February 2007 (UTC)

Do you not think that by the time this project gets around to producing an encyclopaedia's worth of high quality content, by the middle of next century on current trends, that it will have long, long since been totally left behind by the competition, and will therefore have failed? Do you not think that taking 200 years to do something that could be done in 5 years is a bit of a failure? Worldtraveller 17:17, 18 February 2007 (UTC)

As noted, I think your 'middle of next century' estimate is wildly inaccurate both because it relies on unrepresentative measurements (number of featured articles - which is constrained by promotion rate rather than development rate) and because it assumes zero increase in productivity... which is inconsistent with the project history to date. However, even assuming your 200 years estimate were correct... you have not shown that our 'competitors' are succeeding more quickly than we are. Frankly, I don't see any 'competitor' progressing at anything remotely close to the rate we have. If it takes us 200 years and nobody else ever succeeds in building a reference of equivalent scope and reliability... then yes, I would call that a success. Who wouldn't? Personally, I'd estimate that we'll have articles of equal or higher quality on every single topic currently in Britannica (and tens of thousands more) within ten years at the outside. If you think there is some way that we could be succeeding 40 times as fast (by your 200 vs 5 numbers) I'd be curious to hear about it. --CBD 18:44, 18 February 2007 (UTC)

What facts and assumptions lead you to estimate that we'll match Britannica within 10 years? What leads you to conclude that FA numbers are constrained by promotion rate rather than development rate? Why haven't all the excellent non-FA articles been graded as A-class? Why are there fewer A-class articles than FAs? Can you sincerely not imagine that over a timescale far, far shorter than 200 years, the playing field is likely to change beyond all recognition? Do you not think it's wildly optimistic to assume that 'nobody else ever succeeds in building a reference of equivalent scope and reliability'? Worldtraveller 18:57, 18 February 2007 (UTC)

My ten years estimate was founded on my own analysis of our growth in scope and quality over the past four years and the assumption that this curve would continue into the future. Of course, my estimation of our rate of quality improvement seems to be vastly different from yours. That FA are constrained by the promotion process is simply a fact. There have consistently been less than four featured article discussions closed per day... further up the page I calculated that if we increased that to five per day, and they were all successful, it would still take 55 years to get to 100,000 featured articles. Currently only about 40% of candidates for FA are successful... though I'd argue that most of the failing 60% are still 'excellent' articles. You ask about why there aren't more things graded 'A class'... the grades are a relatively new concept and have not been applied to the vast majority of Wikipedia. They also are seldom updated after being initially set. You can find successful FA candidates which are still listed as the 'Start' class they were initially assessed at. As to imagining the playing field changing over the next 200 years... I am certain it will. What prevents you from imagining that Wikipedia will change WITH it? One of the things which I am counting on in that '10 years' figure is that we'll have 'stable versions' within two (hopefully sooner)... which will be a paradigm shift in quality and sustainability. --CBD 20:14, 18 February 2007 (UTC)

Can you write something about how you carried out this analysis? I'd be very interested in it.

I do not know if it is 'simply a fact' that FA is limited by the process and not the number of excellent articles. But if it is, what can we do about that? What would be a better way of assessing which articles are excellent quality? Seeing as many people (including Raul654) believe that every article should eventually become an FA, how can we change the system to make that possible? At the moment it simply isn't.

What prevents me from imagining that Wikipedia will change as well? Well for one thing some of the astonishingly conservative opinions expressed here, and the obviously strong resistance in some quarters to any suggestion that change might even be necessary.

I'm absolutely with you on stable versions. I wrote an essay at Wikipedia:Static version arguing for it. I think serious strides in quality are simply not possible without it. If we have to wait two years, though... two years is an extremely long time in the world of the web. Do you remember how quickly google demonstrated that it was superior to the existing search engines? Worldtraveller 20:47, 18 February 2007 (UTC)

The ten years figure isn't a statistical analysis. Call it an 'educated guess' about trends. Amusing anecdote: Larry Sanger predicted back in 2002 that we'd make it to 50,000 articles within seven years. It actually took about seven months... just before the big 'towns dump' and well before article creation really took off - as seen in Image:Size of English wikipedia 2001-2007.png. I believe that 'quality article count' is following a similar trajectory which has only really started to ramp up in the past year or so... community ideas of what makes a 'decent' article have gone through a radical transformation in the past two years and the 'base line' quality of actively edited articles has undergone a corresponding increase. An active article of 'average' quality in Wikipedia looked like this in 2002, this two years ago, this a year ago, and this now. The quality is clearly increasing and the rate of increase also seems to be growing. For instance, if you look at those 'vital articles'. Yes... we only have 72 'featured', but a year ago we had half that - and most of the ones which aren't 'featured' are of high enough quality that they probably could be if we cleared the slate of featured noms and concentrated only on them. Look at some of those 'non-featured' / 'non-good' articles on that list and tell me they aren't every bit as good as what you'd find in Britannica. Maybe they don't have inline references for every fact (though it is getting to the point that alot of them DO), but in most cases those facts are correct. The major exception, and only area where I do think we could currently said to be 'failing', is on 'contentious' topics. I don't see that being solved until we get stable versions or define contentious articles as a class which only a limited group can edit... possibly a practice of 'long term protection, but admins are allowed to edit at will'. Though that's not particularly palatable.

On FAs, well if we want 'every article to be a FA' then the definition and process for FA needs to change because that just isn't possible so long as the standards continually go up and the approval process remains slow. I prefer something more like, 'every article should be of equal quality as what you would see in Britannica'... which I think includes most of our current FAs, GAs, 'A class', and even some 'B class' articles. I'm not a big fan of the current 'inline reference everything' philosophy as it seems to me both an over-reaction and insufficient solution to our 'verifiability gap' with paper encyclopedias. Basically, all the facts in those have been checked and then they list a handful of references covering everything at the bottom. Since we can't 'fact check' every detail, most significantly because text is constantly being changed, we decided to require inline referencing of everything... but those still aren't all actually checked. Logically, I think the solution to this referencing gap AND the 'FA count' is again... stable versions. Once we can lock down a version for display it would make sense to have people who specifically fact check the entire article against the references and mark that as an 'officially approved' version. Then hopefully we could go back to a saner system of just a few references at the bottom for most articles, consider every 'officially approved' article to be of 'encyclopedic quality', and possibly keep 'FA' just for things which also contain the most brilliant prose, prettiest free-content pictures, et cetera.

Yes, Wikipedia needs to continue to change and evolve... but I think it is doing that. Stable version capabilities are a high priority for the foundation. Heck, the structure of the foundation has completely changed from just a year ago. There are challenges, but efforts are being taken which to me seem likely to meet them... and I still don't see any 'competitor' which has shown that they have the potential to overtake and leave us obsolete... not that it would be a tragedy if one somehow did. If someone finds a better / faster way more power to them. But it seems obvious that nobody has yet. --CBD 22:35, 18 February 2007 (UTC)

'increasing numbers of users will inherently lead to decreasing completion times' ... and if it takes a grad student six years to do one PhD thesis, then six grad students working together can write one in a year, right?

Going back to the random-article test: based on a very small sample of 250 mainspace changes, about 5% are significant additions of content: User:Opabinia regalis/Article statistics#Recent mainspace changes survey. Note that this statement makes no judgment about the quality of that content, which was generally rather poor. Someone who's not at work all day should do this during the day on a weekday, while schools are in session. Opabinia regalis 17:57, 18 February 2007 (UTC)

No, but the six grad students can write six PhD theses in the same amount of time. :] As to your '5% of changes are significant additions of content'... Wow! I had no idea it was so high. We're doing better than I thought. So, if that holds true as the user base continues to climb we might eventually get to the point where a million active users make 10 million contributions per day and 500,000 (5%) of those would add significant content. Plus the cumulative effect of the 95% of other updates. Amazing. --CBD 18:44, 18 February 2007 (UTC)

Don't neglect the caveats; no judgments about the utility of those additions. It's too late now to fish out exactly which edits I classified as major content additions, but I can tell you that two of the thirteen were formatted and referenced, and a third (by a redlinked registered user) put the reference material in html comments with a comment about the difficulty of using cite.php. And the article that got the largest change in content was about a Digimon character.

If I really wanted to keep up with the users = PhD students analogy, I'd estimate that at the end of six years, one would produce a thesis of high quality, two would produce a thesis of acceptable quality, one would be plugging along doing solid work very slowly, one would still technically be in the program but nobody's seen or heard from him in months, and the last guy is the one everybody else thought must have been related to someone on the admissions committee because he was such a screwup. That guy almost blew up the lab in his first year, failed quals three times before starting a harassment proceeding against his advisor, and finally got kicked out of the program after he was discovered falsifying his own results and contaminating the reagents of everyone else who worked in the lab. Opabinia regalis 02:40, 19 February 2007 (UTC)

This is a self-referential section break

You assume that all changes are positive. I think we all know that's not the case. The recent changes survey is very interesting and I'm very grateful for Opabinia regalis for having done that. It adds more useful data to what we can consider. But something that's completely omitted from this discussion is, what articles are people working on? We can see that Vital Articles have not got as much attention as they should have, but are people working on traditional, encyclopaedic topics generally, or is the core business of an encyclopaedia being left to wither while people write about fictional characters, individual episodes of TV series, porn stars and computer games? Worldtraveller 18:57, 18 February 2007 (UTC)

That last seems to be your root bugaboo... but strikes me as largely irrelevant. Yup, we are advancing faster on fictional characters, individual episodes of TV series, porn stars, computer games, and other elements of modern 'popular culture' than we are on 'traditional encyclopedia content'. Hardly new or surprising... and in no way a problem IMO. Some of the people who primarily contribute to 'popular culture' articles also work on the more 'traditional' stuff. Those who don't aren't 'hurting' anything... people with interest in history, botany, science, mathematics, et cetera are still free to write articles on all those subjects, and indeed are doing so. Just not as quickly as the far more numerous fans of popular culture... it's called 'popular' for a reason after all. Where's the bad? You ask if the 'core is being left to wither'... no, it isn't. Growing less slowly isn't withering. A sequoia takes centuries to reach 'full size' while dandelions can grow to full size and reproduce several times in a single year. Unless you expect the dandelions to suddenly start turning into sequoias I don't see where this is anything for surprise or complaint. --CBD 20:14, 18 February 2007 (UTC)

Unfortunately I think that a 'reference work' that concentrates increasingly heavily on pop culture will drive away the editors who want to work on serious topics. You'd have to think that a highly qualified academic might feel a little bit reluctant to contribute in an environment where one could perceive that pokemon is seen as more important than physics. Is there not a danger that Wikipedia will end up being almost entirely a pop culture reference work (if it isn't that already), instead of a general, scholarly encyclopaedia? Worldtraveller 21:11, 18 February 2007 (UTC)

I don't see why they would overlap at all. Why does an editor who spends all their time on mathematics care, or even know, that there are twenty other editors spending all their time on Pokemon? The only place such trends are noticed is in the subset of the 'community' who discuss the project as a whole... a tiny portion of the active editing force. Mathematicians have looked at our articles on mathematics and said that they are excellent. That doesn't change no matter how many Pokemon articles we have. Yes, we are a Pokemon encyclopedia. And a Stargate encyclopedia. And a porn star encyclopedia. But we are also an encyclopedia of mammals which is growing close to rivalling my Walker's Mammals of the World. We cover history, science, math, and various other 'encyclopedia' topics in greater depth than any other single collective reference and are getting to the point where we cover each of them better than any specific reference for just that topic. So no, I don't see academics suddenly leaving us in droves because the 'pop culture' section is growing faster. It HAS been growing markedly faster since 2003 and that hasn't stopped the 'core' topics from continuing right along. If anything, I think the pop culture strength of Wikipedia benefits the core topics by strengthening the site as a whole, bringing in contributors who have interests in both, and establishing the 'brand' as a known and trusted commodity. -CBD 22:46, 18 February 2007 (UTC)

One could dispute your reasoning that a vast array of popular culture and fiction articles - many of which are poorly sourced, poorly written, and and contain no 'out-of-universe' information - improves the image of Wikipedia's 'brand'. Outside perception of Wikipedia as 'the encyclopedia that slashdot wrote' means fewer new contributors to the academic topics. Also, the content issues that arise in editing popular culture articles are quite different from those that arise in editing academic articles, and a population of editors and admins who only have experienece in pop-culture will be ineffective in organizing and supporting the academic topics. (For a minor example: breaking a DNA or protein molecule is called 'cleavage'. Mindless vandalwhacker sees only the word and reverts.) Opabinia regalis 02:40, 19 February 2007 (UTC)

I can only speak for myself, who am not a scholar nor an academic, but the presence of excellent articles such as Bulbasaur and (my fave) Vulpix hasn't deterred me from contributing to Featured Articles on traditional encyclopedic topics such as Photon and Immune system. People who really care about improving Wikipedia will be charmed, rather than dismayed or daunted by such articles. And such people will always exist, I feel certain; Sir Thomas More writes in Utopia of the people of every society and generation who will work for the common good with no reward for themselves. I see the noble impulse, the yetzer hatov, at work every day among women and men of my acquaintance. We may be outnumbered someday, but we'll still exist and doggedly work to improve things. Willow 12:41, 20 February 2007 (UTC)

I can only speak for myself as well. I am an academic, and I feel utterly put off working on serious topics when I see people writing about cartoon characters and porn stars at enormous length. No-one at all seems to understand the concept of appropriate length, and no-one seems to see a problem with having 60kb on a porn star and 60kb on the most important astronomical observatory yet built. I'm no longer willing to do writing on academic subjects in a project where that's clearly not valued very much by most people. I would imagine plenty of academics feel the same. Worldtraveller 16:01, 20 February 2007 (UTC)

Worldtraveller, I really don't understand your position on this particular issue. You seem to equate appropriate length with proportionate length, and as long as articles are chosen for development in an individual and ad hoc way, their lengths will not be proportional to their subjects. This effect is magnified by the fact that articles have a natural length floor (any smaller and they'd be deleted as contextless substubs) and ceiling (eventually they'll split into subarticles). So the longest well-written individual article you'll run into is maybe 80k total, but that's not reflective of the total amount of space dedicated to the topic, because the rest of the information is in subarticles.

Well I'll jump in mid-comment to try and explain a bit more. I don't see how 'appropriate' and 'proportionate' are distinct, in this context. I don't see how having an article on a porn star that's longer than an article on a whole planet is not simply embarassing. It gives critics a very easy stick to beat us with. "any site that's got a longer entry on 'truthiness' than on Lutherans has its priorities straight" and so on. Many or most of the people who write crufty articles do so because they are fans of the subject, and they write at excessive length in a way that is exceptionally boring to most people. But fellow fans vote on FAC nominations... Worldtraveller 19:03, 21 February 2007 (UTC)

'Appropriate length' is a function of the topic and its sources, not of the length of other articles on unrelated topics. There just isn't enough space, in the range between a 3k barely-more-than-a-stub and an 80k behemoth of an article to 'proportionately' represent all subjects, if in fact it were even possible to determine where every topic falls in the importance scale. I'm actually surprised, reading Mercury, at how few subarticles are explicitly noted (compared to Venus, which has three quite bulky ones). There's just no possible way to maintain proportional representation when articles don't get improved at regular rates, and when topics get subdivided in different ways by different groups of editors. Is it 'embarrassing' that RNA interference is longer (total size) than cell nucleus? (Picking on articles I've contributed to, which you may call an embarrassment with impunity if you wish ;) You could say phi value analysis is a completely self-indulgent article on a minor little subtopic of little interest to anyone outside a very specific field, and that it shouldn't exist at all as long as protein isn't an FA - but wiki isn't hierarchical like that, and limiting our coverage for the sake of evading more snarky articles in The Onion is shooting ourselves in the foot. As for crufty FAs, well, go oppose them ;) If any article can be an FA, then Jenna Jameson can, and the article compares favorably in both quality and cruftiness level to a number of existing FAs. Opabinia regalis 03:57, 22 February 2007 (UTC)

That said - as long as the cruftiest of the pop-culture people work essentially independently of the more traditional people, I don't see much problem from an operational perspective. (Above discussion about 'brand' still applies.) However, Wikipedia does not have a culture of 'lurking' or generally of knowing one's limitations; to the extent that there is a problem, it comes when people who don't recognize their own ignorance feel compelled to unconstructively edit or criticize articles on subjects they don't understand. The fact that this pattern occurs mostly in 'pop culture' editors commenting on 'academic' subjects is more reflective of the demographics than anything else, but can still create an atmosphere in which knowledgeable people perceive that their knowledge isn't valued because they can be out-lawyered by kids who spend their time trying to be the first one to write the plot summary of the newest episode of $cartoon. Hence you get someone who's had 'citing sources' explained to him ten times in the course of elaborating on the many differences between death metal and doom metal sneezing fact tags all over an article on a scientific phenomenon, or some kid arguing that his 2005 copy of Cartoon Guide to Genetics is a better source than a 2003 Nature paper because his is published later and is a 'secondary' source, which "the reliable sources guideline says is better". I don't see how that is going to be addressed on a policy level, because it's fundamentally a matter of acculturation. Opabinia regalis 05:14, 21 February 2007 (UTC)

I'm not an academic, and I'm put off by the amount of pop culture material. There's nothing wrong with a normal amount of it, but years ago the project must have collectively let the cat out of the bag by deciding to keep articles on every album ever made, episodes of TV shows, minor cartoon characters, minor bands, minor members of minor bands, vehicles that exist in computer games, etc. These articles exist to a degree that makes me ask "why bother" with this project. Article relativism, of course, is easier than decision behavior that will make someone cry "cabal". We can see why Bugs Bunny deserves an article, and why Pet Sounds deserves an article; they've been around for +/-half a century, and we still care about them. But this is a real-time encyclopedia, and it's more trouble than it's worth to try to argue the fairly reasonable point that Aang should not have an article—and incidentally, that's probably a longer article than most world-famous visual artists have here—nor should Air Nomads (yeah, I'm picking on today's FA). I have said nothing new, expect nothing to change, and am writing for my own gratification. ;-) –Outriggr § 06:11, 21 February 2007 (UTC)

Let's play a better game! :)

Let's make a game of it! boys like a little competition, no? In one day, I made sixteen referenced, categorized articles that didn't exist before from the "D" hotlist at Wikipedia:WikiProject Missing encyclopedic articles. If you truly care about truly encyclopedic content, let's see what you can do. Anyone who bellyaches about the absence of encyclopedic articles and doesn't create sixteen equivalent articles by tomorrow has to...oh, I dunno, there must be a suitable punishment...any suggestions? ;) Willow 17:00, 21 February 2007 (UTC)

Here's my alternative competition - get as far up WP:WBFAN as you can. Bonus points if you get a formerly missing encyclopaedic article as one of your nominations. Interesting to note that it looks like before too long, I'll be top of that list, without even doing any more writing, because Emsworth's FAs are sadly a regular sight at WP:FAR. Worldtraveller 19:03, 21 February 2007 (UTC)

How many bonus points do I get for Cyclol? But your game is too slow; even the best of us can only produce a Featured Article every two weeks. My wee brain can't keep score that long. Besides, everyone has done excellent things in the past, and we definitely honor their contributions, but the game is still afoot. Come on, play along, it's more fun than complaining; the game goes like this: Dactylella, Daedala, Daima, Daonella, Darwin-Radau equation, Dasycladales, David Mermin, Decadentism, Delay equalization, Demetrius (Bible), Demetrius Cydones, Demetrius Lacon, Demetrius Triclinius, Dendroceratida, de Sitter effect, de Vaucouleur's law,...catch me if you can! ;) Willow 19:23, 21 February 2007 (UTC)

Oh, loads for cyclol. I'd be very happy if I felt motivated enough to write more articles, and it would be more fun than complaining, but sadly I don't. I sometimes wonder what Wikipedia would be like if there were a hundred or two hundred people who could write one FA every two weeks, instead of just a handful. OK, here's another small challenge - start an article, and get it to a standard where you can nominate it for FA within 24 hours. It can be done! Worldtraveller 20:00, 21 February 2007 (UTC)

Rosie the Riveter says, "We can do it!" :) Willow 22:07, 21 February 2007 (UTC) PS. Note ironic placement of excellent pop-culture ref. ;)

To that, I can respond that, as "Townie" from The Waterboy famously said, You can do eet!. –Outriggr § 03:00, 22 February 2007 (UTC) P.S. Note ironic placement of mundane pop-culture ref. :) (P.P.S. Let me look into your letter-D challenge...)

No new missing articles today, but I did write one the other day on flying ice cubes. Wait, is that cruft? ;) Opabinia regalis 04:01, 22 February 2007 (UTC)

Cool article! ;) that's easily worth ten of mine. Ten bonus points, too, if you add an explanation on how they remove the net rotational motion — do they just substract Ωxr from each particle? But wait, that would cool off the outside more than the inside? Better reply on one of our Talk pages, we're diffusing in topic-space ;) Willow 11:14, 22 February 2007 (UTC)

A different approach

Since I said above that it's much more productive to discuss how to improve the places where it isn't working very well, I decided to try to do something along these lines, and created Wikipedia:Problems with Wikipedia, where we can objectively list the problems we find with Wikipedia and brainstorm implementable solutions for them. --cesarb 21:19, 19 February 2007 (UTC)

Rethinking the process for the problematic articles

The conversation here has highlighted a couple of problems in wikipedia

1. Pathological articles like Philosophy which are subject to frequent rewritings.
2. Core/Vital/Top and High importance articles which are woefully lacking

if anything productive is to come out of this article I feel we need to rethink how we tackle both.

The problem with an article like philosophy is that there are probably ten or more equally good ways of writing the article in an NPOV way. You just need to look at the number of books on Philosophy which all have a different presentation of the topic. We could happily spend our time arguing over which is the best such presentation, but quite frankly that is a waste of everyone's time. You will also see a new version of the intro text added every month.

So what can be done about it? I would say basically be WP:BOLD in reverting changes to these pathelogical articles. Moreover, establish a culture where substantial change which have not been discussed on the talk page then it will by default be reverted. If such a process is in place then editors should be made aware that such a process is in place, say by placing a banner on the talk page. To make this work and avoid ownership issues it really requires more than one editor watching the page, it maybe that a public watchlist of these articles could be created with a team dedicated to ensuring these articles do not degenerate. Wikiprojects may also help, mathematics has the advantage of a strong wikiproject and major changes to articles are often brought up on the project talk page.

For the core and other important topics we do already have much of mechanism for improving them. WP:ACID is there for improving articles, use it. Some improvement to top importance articles was achieved last year by using the WP:MATHCOTW and submitting the start class core topics articles, these all won the vote and improved somewhat.

Finally lets not be too despondent, over the last year the number of featured articles has grown from 892 to 1240 so its roughly keeping pace with the growth of the encyclopedia as a whole. --Salix alba (talk) 18:22, 18 February 2007 (UTC)

I think one significant problem is that the more fundamental an article, the more the average passer-by feels they know something that needs adding. These piecemeal changes are often highly destructive in the long term. Sun and Venus are two articles I've worked on, both featured but both subject to huge quantities of edits which are slowly dragging them down. On the other hand, Herbig-Haro object doesn't really suffer from this, because most people haven't got a clue what they are. Unfortunately this also indicates to me that there's not so much interest in this article.

Unfortunately your final paragraph is not true. According to WP:GAS, on 1 January 2006 there were 849 FAs and 894,000 articles; on 1 January 2007 there were 1205 featured articles and 1,559,000 articles. So, FAs increased by 42%, overall size of the encyclopaedia increased by 74%. The proportion of featured articles has been in continuous decline for well over a year. Worldtraveller 20:41, 18 February 2007 (UTC)

That decrease means nothing, other than the number of articles grows faster than the number of featured articles. Either way, both are growing; which one grows faster doesn't matter. --cesarb 20:56, 18 February 2007 (UTC)

If we've got excellent and non-excellent content, as defined by FA and non-FA, then it certainly does matter if non-excellent content is increasing at a vastly higher rate than excellent content. Worldtraveller 21:15, 18 February 2007 (UTC)

No, it doesn't. If at first you have ten polished jewels and a hundred unpolished jewels, and later you have twenty polished jewels but a thousand unpolished jewels, you aren't any worse; you just didn't have time enough to polish them all! --cesarb 15:14, 19 February 2007 (UTC)

Ah, but to take another analogy, if you add wine to sewage, you get sewage, and if you add sewage to wine, you get sewage. If you add a bit more wine but a load more sewage, then you get sewage. Obviously we could trade analogies all day :) But the main problem is the rate at which high quality articles are growing. Even if we take the position that it doesn't matter at all how many bad articles we have, we should be very concerned that the rate of growth of FAs will not give us even 20,000 excellent articles for 50 years yet. Looking at WP:VA, there were 41 vital featured articles on 1 January 2006, and 71 on 1 January 2007, at which rate even the vital articles won't all be excellent until 2044. Worldtraveller 16:00, 19 February 2007 (UTC)

Reverted - again?

All my good work has now been removed by Worldtraveller in yet another revert. I don't mind criticism, but what about the corresponding view? Just because you don't hold that view, doesn't mean that it shouldn't be stated. - Ta bu shi da yu 22:40, 18 February 2007 (UTC)

Clearly, you do not agree with the title of the essay. What you have been doing is changing the very scope of it, and rewriting it so that it doesn't conclude that anything is wrong. I especially dislike the bolting on of assumptions that weren't made. Presumably your aim is to say those assumptions are false and so therefore the conclusions are false. You also changed the evidence that was being considered, entirely scrapping the 'coverage of wider topics' section. You actually say Just because you don't hold that view, doesn't mean that it shouldn't be stated, and yet you're trying desperately to prevent the originally documented view from being stated. However, good points were raised by your edits and I tried to incorporate those into what I see as a much more intellectually honest version, which states the actual assumptions made rather than what someone else has assumed later.

I think the discussion on this page has, on the whole, been very very useful, stimulating and productive. I think it muddies the waters totally to try to rewrite the whole essay, without discussing your proposed changes on talk first - how is anyone supposed to meaningfully discuss the thesis when it suddenly changes to express the opposite point of view? Can you not, just for a short while at least, point out here on the talk page what you think is wrong with this essay, rather than just turning it into a vague and ultimately inconclusive piece of nothing? And why have you not been working on WP:NOTFAIL, or WP:EVAL? These would also seem to be productive outlets. Worldtraveller 22:53, 18 February 2007 (UTC)

I don't have to work on those articles. This essay is making some comments that are clearly shortsighted and narrow minded. The assumptions at work are faulty. Anyway, I'm not going to waste my time on this stupid essay any more. Honestly, this really is a stupid piece of work. Give us some plausible criticism, not this ridiculous "there are 1.6 million poor or terrible articles and crazed admins are out of control!". - Ta bu shi da yu 07:05, 19 February 2007 (UTC)

I think it's obvious you haven't actually read it if you think it says anything at all about administrators. I'm very glad you're not wasting your time here any more. Worldtraveller 18:17, 21 February 2007 (UTC)

"A modest proposal"

I think I have a pretty clear solution for this problem:

1. Flag all articles with their rated Class and Priority (averaged if more than one project has rated them)

2. Flag any unrated article with "Unrated, but presumed a Low-priority Stub-class"

3. Outright warn readers that any article not rated F.A. should be treated with caution.

4. Readers that articles rated below G.A. should be treated with blatant suspicion

5. Have Wikipedia not return searchs results for articles rated below G.A. directly, but with an intermediate page saying basically "we have no good article on this topic, but do have one that is inferior; do you still want to see it?"

6. 1-year limit on every article, to achieve G.A. status, or be auto-deleted (can be recovered to userspace for further work)

7. 1-month warning before this happens.

8. 3-month limit on Stubs to achieve Start class.

9. 1-week warning.

10. Direct advocacy of removing unsourced information from articles; reverting such a deletion will trigger a new 1RR rule with regard to unsourced information. The uw-unsourced tags for warning users against the addition of such material would be used with the same vigor and consequences as the uw-vandalism warnings.

What else tough but quality-reinforcing can be thrown in...?

11. De-sysop every admin, without exception (but maybe a 2-month warning) that does not pass the enhanced Diablo test, and make it part of Policy at RfA.

12. Add fame and importance criteria back into the notability concept, until such time as consensus agrees we have G.A. or better articles on every important topic that should be in an encyclopedia.

How's that? I think every article but maybe 10, maybe even less, that I've spent any significant work on would get nuked under this policy, but I'd sure be working hard to improve those few gems up to better-than-Britannica F.A. standards instead of futzing with 300 stubs of dubious merit here and there! And we'd be rid of the "fancruft" that may complain consumes 90% of WP's human and electronic bandwidth.

Or then again, uh, maybe Wikipedia isn't failing. Perhaps some people just have highly personal particular expectations that aren't being met?

— SMcCandlish [talk] [contrib] ツ 23:17, 18 February 2007 (UTC)

Your first set of suggestions are amazing. But I do not agree with the second part starting from "What else tough but quality-reinforcing can be thrown in...?"

I recommend this suggestion to be seriously considered. --169.229.6.145 00:27, 19 February 2007 (UTC)

[I'd forgotten all about this post, did not watch this page, and certainly didn't expect this level of feedback about it.] Um, just in case this wasn't clear to anyone: I WAS KIDDING. I thought that the "A Modest Proposal" heading would give that away. I actually think a handful of the ideas have a grain of merit apiece, if de-extremized, but please. The proposal as a whole was meant as an ironic comment on the projectpage to which this talk page is attached. Anyone who knows me on here at all would know this was a joke from point #12 alone, give how hard I argued for months against such subjective nonsense. — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

I took your last two points as an obvious joke. But you should have been much more outlandish with the rest if you wanted them to be clearly humorous. Worldtraveller 00:52, 27 February 2007 (UTC)

I made your list numbered, hope you don't mind. I think number 5 would do amazing things for overall quality, as would number 8 (I suggested something similar to that a while back). 6 and 7 would also help a huge amount. Of course, many people would probably object to GA being used as a standard, but in principle I think those four measures would absolutely transform things in a very short time. I'd also add that new articles which cite no sources a week after their creation should be speedy deletion candidates. We have to get some things right from the very start.

For me as well, that would mean a lot of my articles getting wiped. Even though I enjoyed writing FAs particularly, and did 24 of those while I was actively editing, I made loads more unreferenced stubs. I expect the majority are still stubs, a year or two years on. The encyclopaedia would hardly be much worse off if they were temporarily removed. Worldtraveller 00:30, 19 February 2007 (UTC)

Your numbering didn't seem to stick; looks like someone reverted it, not understanding WP:REFACTOR. For the record, people should feel free to recast a talk page bullet list as a numbered list if they honestly think that will be somehow useful, and people need to get over their YOYOW possessiveness about their talk page posts; Wikipedia is not The WELL. I put the numbering back, just so these followup discussions make sense (and noting again, as above, that the entire post is an ironic joke.) — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

These are excellent ideas, although I need to mention that there are numerous B-class articles of very high quality, whose authors have not initiated the whole "nomination" thing.

Consistent with your ideas, I'm thinking of starting a WikiProject Unassessed Articles that would allow us to rank articles for importance and quality, even if they don't have a sponsoring WikiProject of their own at present. Does that seem like a good idea to the people here? Hoping for constructive feedback, Willow 00:34, 19 February 2007 (UTC)

This would be duplicative of existing efforts. The WP:1.0 project is already handling this. The WikiProjects, like WP:BIO are using their article rating system. — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

Other than our stub, there is no information on the web about Meade Island. Would you delete a valid, attributed, irreplaceable (on the Web) article simply because it has a stub tag on it? I share Worldtraveller's views on unattributed information, but I find the idea of deleting brief but fully attributed stubs absolutely deplorable. Hesperian 00:44, 19 February 2007 (UTC)

The fictional "me" that wrote this piece would say, "Yes! If no one cares enough about your silly little island and its article, it is self-demonstrably non-encyclopedic, and will never, ever have a Featured Article, and since the goal of WP is for every article to eventually achieve F.A. status, it must go." In reality I'm a quasi-inclusionist, so of course I would not actually advocate removal of a well-sourced stub, and I think hard-core immediatists need to go soak their heads. — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

Heh, following these standards Wikipedia would never have gotten off the ground at all. :] --CBD 00:48, 19 February 2007 (UTC)

That was half the point (the other half being that an arguably rational position can be defended that things have changed ergo WP's nature has to, to "keep up with reality". I leave it to the reader/editor to determine just how defensible that view is; cf. User:Worldtraveller's material below, and the very existence of this document. If it were absurd it would be ignored. But the fact that is sparks honest debate doesn't mean that it ulitmately represents the truth either.) — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

Well, Hesperian, if actually the article says all that reasonably can be said about the island, then it shouldn't be marked as a stub. Stub means 'very small article which will be expanded later'. I think when SMcCandlish said 'delete' he made clear that didn't mean delete really, but move or hide or otherwise take out of article space.

I actually meant (in a devil's advocate way) delete, though intentionally left in the caveat that any so-deleted article would be userspaced on request for further work. — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

CBD, you're quite right, but then what got Wikipedia off the ground will not necessarily correspond to what will transform it into a reliable and authoritative reference work. Worldtraveller 00:55, 19 February 2007 (UTC)

I don't think that it says all that reasonably can be said. But I do think that anyone seeking information on this obscure island would search the Web with very low expectations, and would be overjoyed by how much information Wikipedia has. Making this information inaccessible would be cutting off our nose to spite our face. Hesperian 01:00, 19 February 2007 (UTC)

Or, to put the devil hat back on, requring people who believe the island to be of encyclopedic value to really do their homework and make a fantastic article about it. — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

So what is stopping it saying all that could be said? How long is it likely to stay a stub, in the absence of an incentive to expand it? Worldtraveller 01:02, 19 February 2007 (UTC)

(i) The difficulty of sourcing information, which I believe to be out there, even if I haven't found it yet. (ii) In the absence of an "improve or delete" imperative, it will probably still be a stub in 100 years. That is an argument in favour of such an imperative, only to those who think well-written, neutral, fully attributed stubs are a bad thing. Hesperian 01:10, 19 February 2007 (UTC)

Do you see it as a problem if unfinished articles remain unfinished indefinitely? Shouldn't we have an incentive to encourage people to finish articles rather than just stub it and run? By the way, fully attributed does not apply to the vast majority of stubs, and I think perhaps the article we're talking about right now shouldn't be called a stub. Worldtraveller 18:09, 21 February 2007 (UTC)

Indeed, sounds like a "Start"-class article, though I haven't looked at it myself (and not like I'm an authority on such things; just saying I'm agreeing in principle, not on specifics. :-) — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

Actually, changes like this now would be worse than having it that way from the start. With the restrictions in place from the beginning you'd have a very slow process which would probably die of inertia, but could plod along and amount to something after a hundred years or so. Basically... Nupedia. Putting strictures like this into place NOW? Immediate insanity of unimaginable scope. A hundred thousand screaming Pokemon/South Park/Boy band/Whatever fans all going ape shit simultaneously. The site would be completely trashed and never recover. --CBD 01:06, 19 February 2007 (UTC)

Addendum to "Proposal": 13. So, when they cross the line, ban them. — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

I doubt they'd have that much impact actually. I doubt many people would get angry, especially if restrictions were only placed on new stubs with a much more generous timescale for existing ones - similar to how unsourced images were purged. Worldtraveller 18:09, 21 February 2007 (UTC)

To me (1), (2), (3) and (4) seem excellent suggestions. The rest are a bit radical though. --Aminz 02:44, 19 February 2007 (UTC)

Ya think? >;-) I'm frankly surprised that 1-4 went over well! That's kind of scary actually. — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

I imagine that implementing such an idea would devestate a large proportion of the encyclopedia. Assuming that people could stomach the idea of working to have it deleted, to work to have it deleted (like building a sandcastle where the waves are lapping). Some articles might be recreated over time and deleted again a year later. Result: more wasted effort. In reality, most people would just leave. --Seans Potato Business 02:49, 19 February 2007 (UTC)

That was the idea I was trying to get sarcastically across, in opposition to the essay this talk page belongs to. — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

I wouldn't believe for a second that most people would leave. Yes, according to WP:1.0/I, assuming no-one did any stub expansion, 64% of all articles would be under threat; but without any incentive at all to expand or improve, then most of them will never become non-stubs anyway. The project would be much better off if they were merged into broader articles. If we had a culture that viewed barely-started articles as a bad thing, and a suitable length of time before sanctions against them were enacted, huge numbers of stubs could be expanded. Worldtraveller 18:09, 21 February 2007 (UTC)

Obviously this is a terrible idea, but there might be something useful to fish out of it. Some of the worst elements of this proposal:

• Any unrated article is assumed starting out to be an unimportant stub. What? 1.0 has about 193,000 articles assessed out of 6,919,219. The others are all unimportant stubs? I just had a look at a few of my own articles - computational phylogenetics, implicit solvation, and homology modeling all fall into the unassessed category; they may not be that important, but they certainly aren't stubs. This proposal assumes that we can realistically get all (or at least the vast majority) of articles assessed. I don't know, I'd rather be writing them than writing 'B' on the talk page.
• In practice, 'B' is used to mean 'this article is in a usable state'. There is no accessible rating above B for most individual assessors. Claiming that all B-level articles are of low quality or should be treated with 'blatant suspicion' is nonsense.
• The GA process cannot cope with the massive influx of articles that would result from something like this. GA is already perpetually backlogged, and the quality of the reviews varies so enormously that it's effectively useless as a means of getting feedback on an article. The number of people just clearing the backlog, reviewing articles that they know nothing about, can only increase. If I can use my own as an example again (those are the ones I remember, you know ;), computational phylogenetics failed GA with a review that said, and I quote, "the prose, I was starting to fell asleap". (It later got a much more useful review from an expert, fortunately.)
• Since assessments are given by a single user, what's to stop me from thinking, 'hm, I wrote this pretty okay stub 2.9 months ago and nobody's edited it since, but it's pretty big for a stub, I'm just going to call it start since it would be dumb to delete it'. With 1.6 million articles, are these things going to be noticed? If they're going to be policed more strictly, where will we get the police?
• Unsourced statements aren't evil. I categorically oppose any variation of 'all unsourced material will be deleted, period' because some people have such incredible difficulty with concepts like 'common knowledge', and this causes a have-you-stopped-beating-your-wife situation, since nobody wants to be on the side of lax referencing and low standards.

A less invasive idea: introduce a simple wikiproject certification scheme. Binary decision, 'this article is usable as a reference on the topic' or 'this article is not usable as a reference on this topic'. The intuitive response to the limitations of the existing assessment system is to add more dimensions, but that would only worsen the inter-project inconsistencies. Also, the emphasis on utility eliminates the tendency in grading schemes to use them as relative rather than absolute ranks; if an article is 'start' under the current system but hits all the basic points on a subject and contains a reference to a source of more information, it could still be certified as usable. Articles that are claimed by a wikiproject but not certified as usable can be tagged to notify the reader of their deficiencies. Opabinia regalis

Exactly. WP isn't broken. It's doing a great job as-is, with the human resources available. It's a true wonder, and expecting it to be wonderful x 1000 is just silly (like Wikipedia:Wikipedia is failing). — SMcCandlish [talk] [contrib] ツ 10:16, 26 February 2007 (UTC)

What's silly is replying to data showing significant problems with a statement as bland as 'it isn't broken'. That really doesn't advance the discussion at all. If you have evidence to show exactly how Wikipedia cannot be improved and is not suffering from any significant problems in producing high quality encyclopaedic content, please share. I see nothing at all in what you've said here that demonstrates any serious problems with the analysis in the essay. I invite you to have a look at Tool, Litre, Market, Dictatorship, Mind, South America, and Herodotus and consider whether they meet the standards you'd expect of a high quality encyclopaedia. All are on WP:VA. Worldtraveller 00:52, 27 February 2007 (UTC)

Wrong assumption

That the sample of 300,000 articles assessed, with results listed at WP:1.0/I, is representative of the whole encyclopaedia.. Considering 100% of FAs are assessed and listed there, but only 20% of all articles we have are assessed, this sample is not representantive: it is biased towards higher quality article (which are more likely to attract attention and be assessed).-- _{Piotr Konieczny aka Prokonsul Piotrus | talk}  07:43, 20 February 2007 (UTC)

List of like-minded

Some good views expressed here. In these situations I always like to have a list of like-minded people that I and others can refer to. I have started one on my user page. To be like-minded you have to agree there is some sort of problem with Wikipedia, and you must believe there is a way of changing it, perhaps gradually, or a little bit, from the inside.

If you would like to put your name down, please feel free to edit my user page, or leave your name on my talk page. If you edit my page, please restrict the edit to the format that is already there, under the 'list of like-minded'.

I have put some names there already - if any of those people feel they are not like-minded, or don't like being on lists, please remove! Thanks edward (buckner) 11:31, 20 February 2007 (UTC)

Is FA a good test?

This fundamentally assumes that FA is an accurate assessment tool; it is not, it is going the same way as GA: arguments over citation by editors who frequent the place, but who understand neither the subject of the article, nor the function of citation. It should not be surprising, since many of these are the same editors. Septentrionalis PMAnderson 14:30, 20 February 2007 (UTC)

FAC is currently the only means of identifying which articles are the very best. It may not work as well as we like, but apart from FA numbers we have no other means of saying how many of our articles are of the highest quality. I haven't frequented FAC for a few months, but I do see a lot of people voting in support of articles just because they have a lot of citations. There seems to be an extremely unhealthy mania to cite crazy things, like saying Firstname Surname¹ is a popular² singer³. Worldtraveller 14:40, 20 February 2007 (UTC)

Wikipedia is a social structure and FAC is a clique within that structure. It may be doing useful work but it is far from clear that the articles it identifies are "the very best." It's criteria are more formalistic than substantive and it makes no pretense of reviewing more than a tiny fraction of articles. In its present form is not suitable as a way to evaluate Wikipedia. Wikipedia lacks a broadly applied evaluation system and it may not be possible to build an effective one internally. Perhaps the best approach is for outside experts to evaluate a sample set articles in their field. The few times this has been tried, Wikipedia did quite well. --agr 15:57, 20 February 2007 (UTC)

Yet Another Modest Proposal

Okay - I'm new here. Please don't kill me. But for the past few months I've been wandering about through both the article space and the back alleys of Wikipedia, and I recently started using some of my idle time to clean up obvious vandalism. I would not contend that I am intimately familiar with the way things work around here, though I would say that I'm passing familiar with a lot of it.

With that preamble out of the way, here's what I notice. Wikipedia's main flaw, at least so far as the problems illustrated in this essay goes, appears to be not the constant threat of anonymous vandals (I've had to give up just grabbing random IP edits to check for vandalism from the Recent Changes page because so many of them are good, constructive material), but that the project as a whole seems to have little in the way of management. Absolutely everything is controlled by the community, except for the things that aren't (and hopefully that description accurately depicted how blurry the line can get).

This is a fine way to run a democracy, if that's your thing, but for finishing a project with a goal, it doesn't work out so well. The reason why none of your vital articles are in good shape is because what experts Wikipedia attracts will likely be interested in working on things that they have expertise in. Or, to put it differently, they'll be more interested in working on the theoretical physics articles (things that are interesting to experts) than they are on the basic physics article (wouldn't YOU be bored with it if you'd been working with it so long it was burned into your soul?). Given the fact that any attempt to build consensus around basic concepts, like those that the vital articles concern themselves with, are likely to spark dispute and unpleasantness, why would anybody, least of all an expert, want to work on them?

That disincentive to work on vital articles isn't going to go away. There are always going to be more people that think they're qualified to edit the general Philosophy article than the specific article on Hyperreality (as evidence, I would say, by the fact that some recent Colbertorrism managed to persist there until I happened to notice it on a random trip through, while the rest of it was cleaned up in neat order). The problem is that these articles need to be written. The encyclopedia doesn't need greater quality and specificity in non-vital articles, if the goal is to build from the vital articles out.

Now, if this were a project at any business, all that would happen is the manager would go find some person and order him to write the article. Maybe we should brainstorm some ways to get people to work on things the encyclopedia needs before they get the opportunity to work on the things that they might really want to? Not sure how to implement it (forcing one update to an essential article before opening you up to edit anything else for the day is the first thing that comes to mind, but that seems like it might cause more harm than good), but it seems to me like we have to fill the management vacuum and start getting people to do what needs to be done. I could be wrong, though. Just thought I'd put it up as an observation from a relative newcomer.Cool moe dee 345 15:45, 20 February 2007 (UTC)

This is a volunteer project. In a job, if your manager wants you to do something that is less interesting than what you want to do, you'll still do it, because you are being paid to do it. In a volunteer project, if people are prevented from doing what they want to do, they do not just work on what other people tell them to. And if it's enforced, they say "This is stupid. I am volunteering my time, only to be told that I must do boring things. I quit." Then, we have next to no experts, so the vital articles are still not very good. And the specialist articles got worse, because they now have no editors. -Amarkov moo! 18:25, 21 February 2007 (UTC)

Volunteers are, by definition, altruistic people. I don't think there would necessarily be an exodus of editors if a modicum of editorial direction and control was applied. In addition, many of the people who might work on the vital articles are a bit put off by the fact that 75% of the 'encyclopaedia' seems to be about cartoons and pop stars. For me, writing about vital articles was fun, until I gradually came to realise that there really weren't that many other people doing so. Worldtraveller 18:40, 21 February 2007 (UTC)

Working on many of those vital articles is not fun for most people. I'm going to go edit one of them right now, and see how my change is reverted. -Amarkov moo! 18:58, 21 February 2007 (UTC)

Why would it be reverted? are you going to vandalise it? Worldtraveller 19:05, 21 February 2007 (UTC)

Okay, so it's not as bad as I thought (excluding philosophy articles). Still, the fact remains, it's not fun for some people, and this is a slap in the face; "Your contributions are not good enough unless you contribute in this way." -Amarkov moo! 19:13, 21 February 2007 (UTC)

But we already have numerous policies and guidelines defining what contributions are acceptable and what are not. WP:NOT is the main list of them. Worldtraveller 19:19, 21 February 2007 (UTC)

That's different. WP:NOT, along with all policies, says "X, Y, and Z are not acceptable for Wikipedia". This says "Q is acceptable for Wikipedia, but only if you've done enough W". If a class of contributions is truly acceptable, and not just grudgingly tolerated, then you don't require other things as a condition. -Amarkov moo! 19:26, 21 February 2007 (UTC)

Stop focusing time and effort on inconsequential articles about things that do not exist!

Much effort is expended on documenting the intricacies of the Star Wars universe, or anime cartoons, or obscure characters from comic books, etc. This is not only amateurish for an encyclopedia, but the time spent making these articles and editing them could be better spent on articles that actually belong in an encyclopedia. — Preceding unsigned comment added by 71.131.43.48 (talk • contribs) 07:33, 21 February 2007 (UTC)

The obvious fallacy is the notion that those same editors who work on that sort of thing would spend their time writing scholarly material if we somehow got them to cut out the cruft. This is unrealistic. Friday (talk) 15:05, 21 February 2007 (UTC)

This is true, but if there were far fewer comic and cartoon articles, you might find that far more academics and experts were interested in getting involved. Worldtraveller 18:42, 21 February 2007 (UTC)

I doubt it. Wikipedia is Wikipedia, no matter how many articles like that there are. Academics and experts are still going to be skeptical of an encyclopedia whose fact checking method is hoping that if something is wrong, someone will wander by and correct it. -Amarkov moo! 19:19, 21 February 2007 (UTC)

The academics and experts I know and chatted about the wiki with never minded the pop articles. Their problem with the wiki is this.

Everyone bases their work on the work of others. In order to avoid being wrong, they can only

• make sure there is no fault in their original parts and
• make sure there is no fault in their choice of sources.

Sources can be wrong, and one can't check everything, so there is no guilt in falling for the mistake somebody else made. It is just that one's choice of sources has to be reasonable. They have to be things that, if wrong, "could have misled anyone", because otherwise. The wiki, not even claiming to be definite, cannot serve this purpose. So if any mistake from the wiki were to be propagated in scholarly work, the academic/expert himself would have made a mistake, in an environment where single mistakes can end careers. This is an unacceptable risk. The presence of pop articles is entirely unrelated to that. Denial 13:15, 23 February 2007 (UTC)

Wikipedia fail when ...

The wiki fail, when ...

• it is not a wiki.

The encyclopedia fail, when ...

• it is not an encyclopedia.

Wikipedia fail, when ...

• it fail to be wiki and encyclopedia at the same time.

It is not a wiki, when ...

• It does not have a freedom to edit and publish with good faith.
• The editors do not have a wiki mind or wiki spirit.

--Ans 03:10, 22 February 2007 (UTC)

The question is, is the wiki- bit subsidiary to or more important than the -pedia bit? If aspects of a wiki system conflict with aspects of an encyclopaedia building project, then what gives? Worldtraveller 18:35, 23 February 2007 (UTC)

If they conflict, then wikipedia fail :) --Ans 12:57, 26 February 2007 (UTC)

Why so? Worldtraveller 00:28, 27 February 2007 (UTC)