I have just completed and written up a little research project of mine: http://www.gwern.net/In%20Defense%20Of%20Inclusionism#the-editing-community-...
Summary:
1. Talk pages are where references/links/citations go to die; less than 10% ever make it back 2. In just the sampled edits, millions of page-views are affected 3. Conclusion: putting references/links/citations in an Article's Talk page is a bad idea (compared to External Links)
Numbers, source code, and lists of edits are provided in the link.
This article starts as a complaint about external links being moved to talk pages and never making it back to the main page, and then becomes a rant against deletionism.
About external links, the real question is: what is a good number of links to have at the end of an article? Everyone will surely agree that an article with 100 external links at the end is not ideal. What people want from Wikipedia is a site where others have sifted through the chaff to present the most relevant information. What article needs more than about 5 to 10 external links to cover the issues that haven't been addressed in the inline citations and the text?
As for deletionism, I understand that it's a serious issue, but for a quick and dirty random sample, let's take a look at the last 20 closed deletion discussions:
Deleted: * Bill Batstone - non-notable musician * Stefan Duncan - non-notable visual artist * Jason Gagliardi - non-notable author * Bosnianism - supposedly original research. I know nothing about the topic, but it sounds dubious at best. * Anton Strastev - No "keep" opinions - "Poorly written page on non-notable person." * Hoarding (Psychology of) - duplicates info in "Compulsive hoarding", not notable on its own. * Jamie Hanley - failed political candidate who didn't achieve public office, not notable in any other sense. * Esh (Unix) - non-notable minor unix shell. * Hannibal Reitano - socialite journalist * Byron Rakitzis - programmer, musician, student, one-time winner of Obfuscated C contest. * Trent Evans - "A person whose only claim of notability is that he put a coin under the ice before a hockey game. " * Reading My Eyes - A song which never charted. * Jorge Castro (actor) - A Puerto Rican Theatre actor. Doesn't have an article on es.wikipedia. * BRINK (magazine) - No comments in support of keeping the article, apparently fails notability criteria.
Kept: * Yaesu FT-1000MP - discontinued amateur radio receiver: no consensus for deletion.
Merged / Redirected: * Martin County Sheriff's Office - redirected toMartin County, Kentucky * Ladies Masters at Moss Creek - wrong name, relisted at "redirects for discussion" * No More Sorrow (Linkin Park) - Song was never released to radio: redirected to album article. * Animals on the Underground - Merged to "Tube Map" * List of historically significant Michigan Wolverines football games - Merged to Michigan Wolverines football.
I think this is an example of a working immune system. Which of the deleted articles do you think we needed to preserve on Wikipedia, as opposed to someone's blog?
2011/12/22 David Richfield davidrichfield@gmail.com:
This article starts as a complaint about external links being moved to talk pages and never making it back to the main page, and then becomes a rant against deletionism.
No, it does not 'start' as that; the complaint is a subsection and case-study into one deletionist practice (deleting external links). Feel free to ignore the 'rant' part and deal with the observed facts.
About external links, the real question is: what is a good number of links to have at the end of an article? Everyone will surely agree that an article with 100 external links at the end is not ideal. What people want from Wikipedia is a site where others have sifted through the chaff to present the most relevant information.
I would not agree. On an extremely complex topic, perhaps 100 links is perfectly justifiable. Figure 5 sub-divisions, that's only 20 links a piece. (No one looks at an article with 5 sections with 20 references a piece and goes 'everyone will surely agree this is not ideal!') Context is king, and you are immediately trying to make dangerous generalizations.
So tell me, what failure rate would you find acceptable? You apparently are not disturbed at a >90% failure rate to use external links; would you be disturbed at 95%? At 99%? Before trying to put me onto a slippery slope, explain where on the original topic you would finally agree, 'yes, this is too bad a failure rate, something must be done'. Until you present some principled reason or specifics, you read like a blind defense of the status quo.
What article needs more than about 5 to 10 external links to cover the issues that haven't been addressed in the inline citations and the text?
Any article where the editors are largely absent and will not use even gift-wrapped excerpted references; as is the case for >400 articles with hundreds of thousands/millions of readers, which I just spent a great deal of time demonstrating.
On Thu, Dec 22, 2011 at 7:33 AM, Gwern Branwen gwern0@gmail.com wrote:
So tell me, what failure rate would you find acceptable? You apparently are not disturbed at a >90% failure rate to use external links; would you be disturbed at 95%? At 99%? Before trying to put me onto a slippery slope, explain where on the original topic you would finally agree, 'yes, this is too bad a failure rate, something must be done'. Until you present some principled reason or specifics, you read like a blind defense of the status quo. -- gwern http://www.gwern.net ------------------------ ------------------------
This rate, without additional context, is meaningless. As Rob pointed out, there are many different reasons for moving references/links/citations from an article to a talk page, and unless you have more information about why people are moving these to talk pages, the rate at which they move back doesn't really mean anything. By labeling this rate a 'failure rate' you are strongly implying that success would be keeping the link in the article. I don't believe this is right - I believe that 'success' is doing what's best for the article.
Even if 99% of things that were moved to talk pages were not subsequently returned, I would not find this at all disturbing without evidence that a large portion of the removed things should not have been removed. Frankly, I would be surprised if 10% of things that I personally moved to talk pages were moved back in to the article space. Generally if I move something to a talk page it's because it's not fit to be in the article and I don't see an easy way to make it fit to be in the article but I think that they may point the way to a resource that should be in the article.
Your observed facts are interesting, but they do not (sufficiently) support your conclusion.
---- Kevin Gorman User:Kgorman-ucb
About external links, the real question is: what is a good number of links to have at the end of an article? Everyone will surely agree that an article with 100 external links at the end is not ideal. What people want from Wikipedia is a site where others have sifted through the chaff to present the most relevant information.
I would not agree. On an extremely complex topic, perhaps 100 links is perfectly justifiable. Figure 5 sub-divisions, that's only 20 links a piece. (No one looks at an article with 5 sections with 20 references a piece and goes 'everyone will surely agree this is not ideal!') Context is king, and you are immediately trying to make dangerous generalizations.
So tell me, what failure rate would you find acceptable? You apparently are not disturbed at a >90% failure rate to use external links; would you be disturbed at 95%? At 99%? Before trying to put me onto a slippery slope, explain where on the original topic you would finally agree, 'yes, this is too bad a failure rate, something must be done'. Until you present some principled reason or specifics, you read like a blind defense of the status quo.
As Kevin said, this is in no way a failure rate. An external link provided as a formatted inline citation to support or expand on the text of the article is very helpful to the reader. A huge list of external links at the end of the article is just "here's a bunch of stuff you might like to read". It's unlikely to be well used or maintained, and quickly becomes a magnet for spam.
What article needs more than about 5 to 10 external links to cover the issues that haven't been addressed in the inline citations and the text?
Any article where the editors are largely absent and will not use even gift-wrapped excerpted references; as is the case for >400 articles with hundreds of thousands/millions of readers, which I just spent a great deal of time demonstrating.
So maybe what you actually demonstrated is that dumping a site onto "external references" is much less useful to readers or other editors than finding a place in the text where it would actually be relevant and typing <ref>[http://www.example.com/index.html The editing community is alive and well - Example.com]</ref>
wikimedia-l@lists.wikimedia.org