Sometimes, when I'm in a hurry, I might link it for something that's so well known it can't possibly get it wrong. But as a definitive source, not even close. I'd never use it as a reference source, it carries zero weight with me.

I'd be curious to see examples of egregiously bad pages. I find that many people feel this way about Wikipedia on principle. My sense, however, is that it's better and more accurate than most people think, even for less trafficked subjects.

The high profile pages, like those on major political figures and current issues, are often surprisingly well-balanced, but if you're searching for bad articles the best bet is to seek out somewhat obscure topics that have a passionately devoted coterie who believe that the "truth" was fudged by plot, conspiracy or long-ago errors (emphasis usually on two former). More lucid, critical editors often don't bother with them, or have never heard of them, so the methodless cranks can have full sway. The page on Thomas H. Ince (1882-1924), for example, was obviously edited by somebody who passionately believes Ince was murdered by William Randolph Hearst. I'll sample one paragraph:

"Hearst took care of Parsons, apparently rewarding her for her silence. When the Oneida sailed, Parsons was a New York movie columnist for one of Hearst's papers. After the Ince affair, Hearst gave her a lifetime contract and expanded her syndication. Hearst also provided Nell Ince with a trust fund just before she left for Europe. In return, she refused an autopsy and ordered her husband's immediate cremation. Rumor also has it that Hearst paid off Ince's mortgage on his Château Élysée apartment building in Hollywood."

This has all the stigmata of a shit Wikipedia article. Unsourced facts, one of them admittedly just a rumor and so of questionable relevance. Vagueness, marked by a lack of clear dates. Non-sequiturs, which I've marked in italics, suggesting or asserting a quid pro quo that is not apparent by the bare facts even if they are accurate.

If you're well-read in obscurities and have an flypaper eye for the stuff that attracts the conspiratorial-minded you could probably compile quite a list.

I listed some egregiously bad pages in another thread, "Wikipedia not good for controversial or specialized subjects." I mentioned, for example:

The Japanese language. That is pretty good overall but it has some egregious mistakes, which I expect are there because not many English speakers know Japanese, and those who do (such as me) don't bother to correct the mistakes. Many English speakers know French, so I suppose the article about French is good. (I don't know French, so I wouldn't know.)

Cold fusion. Absolutely dreadful for many reasons, some of which I described in that thread.

Sure, I use Wikipedia all the time. I bet even professional scholars do. Also Google If I want to look up purely factual stuff, say the birth and death dates of King Louis IX of France, it contains tons of useful and reliable information, and is often the quickest ad easiest way to look it up, a whole lot easier than going to the public library. Plenty of their articles even have bibliographies tacked on the end, and that helps keep their contributors honest. Does it contain factual errors? Sure it does, but so what? So do standard printed reference works you'll find in any library. Once I made a project out of finding out as much as humanly possible about a particular English writer, and it quickly became apparent that the article about him in the Dictionary of National Biography, written by somebody with fairly impressive academic credentials, was crawling with factual mistakes. The quality control measures taken by the general editors of works like that simply aren't up to the job of weeding them out. When it comes to non-factual matters involving opinions or controversial subjects, of course, that's a completely different story. Caveat lector. But I bet that such stuff represents a fairly small percentage of the whole of Wikipedia.

cwilson:

I'd be curious to see examples of egregiously bad pages. I find that many people feel this way about Wikipedia on principle. My sense, however, is that it's better and more accurate than most people think, even for less trafficked subjects.

I never said there were bad pages, I simply don't view it as a definitive source. Especially not for contested or opinion based questions or facts.

For the most part, it probably is accurate. But it's user generated and not peer reviewed. It's had more than it's share of problems with bad edits over the years. They've addressed and are continuing to address many of these issues, and that's great. But as a definitive source, no, not yet.

How do you define "definitive source?" Because every reference text ever written has a percentage of errors in them, including encyclopedias and peer-reviewed scientific journals.

I keep reading this meme that because Wikipedia is written by anonymous writers it can't possibly be as good as other texts that are also written by anonymous writers but are published in paper form, so they must be Official and Correct. From my personal use of Wikipedia, it's just not the case. The site is a great way to get basic information about lots of things. If you want someone's opinion about people, events, etc. that's where it gets more dodgy, but that's true of any written source.

I agree with dobbsfox on this point--the references at the bottom are particularly useful. Very interesting stuff about conspiracy theories and complex topics like Japanese, though. I've heard alternative medicine entries are particularly dicey, as we might imagine.

Here's food for thought, though: Where Wikipedia doesn't stand up is in making the length of the entries proportional to the importance of the subject relative to other entries, the way an old-school encyclopedia does (to some extent). This was brilliantly demonstrated by Somethingawful with side-by-side entries on Modern Warfare and Lightsabers, God and Kevin Smith, etc. (Techcrunch calls this the "nerd bias.") I don't see this as a tremendous flaw or one that can really be addressed in any scalable way, but it is interesting.

cwilson wrote:

"Very interesting stuff about conspiracy theories and complex topics like Japanese, though."

The problems with the Japanese article are instructive. First, I am not sure if the problems are still there . . . But anyway, there were mistakes that seemed to be written by an enthusiastic person who has recently begun studying the language. He or she was trying to construct sample sentences in Japanese that were too much like English and that no native speaker would use. If you are going to use samples, you have to either find them in Japanese text somewhere or ask a native speaker. You might copy one from a highly authoritative source such as Martin, which has thousands of sample sentences, all carefully sourced.

The problem was not that the person was obstinate or aggressive. He just does not know enough about the subject to write about it.

There are some on-line resources for translators which are crowd-sourced, such as the Eijiro dictionary. This has tens of thousands of sample sentences. I have sent them a few, mainly for electrochemistry and for newly coined words such as the model name of the upcoming Toyota hybrid car. You fill in a form and supply a URL if you have one. You cannot edit the dictionary directly, but someone there gathers up the changes and implements them. So it is lightly peer-reviewed. A person could probably not add an amateur mistranslation, or a deliberate malicious change. There are also more formal on-line resources such as Prof. Breen's WWWJDIC. He has a form you fill in for suggestions and corrections. He is a nice fellow and he fixes problems promptly. You might say these are crowd sourced but everyone in the crowd is known to be bilingual and a translator.

dobbsfox wrote:

"How do you define 'definitive source?'"

That is a source that most experts in the field agree upon. Sometimes experts do not agree, so there is not definitive source. Of course experts can be wrong, but not often. Usually it turns they were not experts; they were people who thought they knew but it turned out they did not.

"Because every reference text ever written has a percentage of errors in them, including encyclopedias and peer-reviewed scientific journals. . . .I keep reading this meme that because Wikipedia is written by anonymous writers it can't possibly be as good as other texts that are also written by anonymous writers but are published in paper form . . ."

Encyclopedias are not anonymous. The authors and editors are listed. The ones I have are, anyway. So are the authoritative dictionaries such the Japanese English Kenkyusha. Also they are not paper. They are on the Internet these days, available by subscription. (I am a cheapskate so I use the 1954 printed edition instead.)

As I mentioned some of the electronic dictionaries nowadays are crowd sourced and therefore semi-anonymous, but someone checks new entries and corrections.

JedRothwell:

dobbsfox wrote:

"Because every reference text ever written has a percentage of errors in them, including encyclopedias and peer-reviewed scientific journals. . . .I keep reading this meme that because Wikipedia is written by anonymous writers it can't possibly be as good as other texts that are also written by anonymous writers but are published in paper form . . ."

Encyclopedias are not anonymous. The authors and editors are listed. The ones I have are, anyway. So are the authoritative dictionaries such the Japanese English Kenkyusha. Also they are not paper. They are on the Internet these days, available by subscription. (I am a cheapskate so I use the 1954 printed edition instead.)

I think you're missing dobbsfox's point that the authors of encyclopedias are effectively anonymous. Seeing an entry marked as being by "Dr. William H. Dagwood Jr., Associate Professor of Chemistry, Hyderabad University, India" only appears less anonymous than "JediKnight8177" because we believe that we can track down the former if we wish, and assume that someone else already has. In fact, reliance on this assumption is how impostors and frauds operate.

Oluseyi wrote:

"I think you're missing dobbsfox's point that the authors of encyclopedias are effectively anonymous. Seeing an entry marked as being by 'Dr. William H. Dagwood Jr., Associate Professor of Chemistry, Hyderabad University, India' only appears less anonymous than 'JediKnight8177' because we believe that we can track down the former if we wish, and assume that someone else already has."

Not really. If it is an article in a major reference book such as Britannica or a subject review in Scientific American about Japanese, let us say, it is likely that I will know of the author. I will have read his or her other works. The people they pick to write these things are well know to everyone in that specialty. And if I do not check the author, someone else in the field surely will, out of professional jealousy. Some professor will think: "They should have hired me to write that! Who is this joker William Dagwood? Never heard of him . . ."

Furthermore, nowadays universities all have their staff listed online so you can look up Hyderabad University and you will find there is no such prof. there. (I just checked.)

An obscure book from a small publisher by a faker might be published, but you will not find a faker in a major reference book. A major publisher will circulate the manuscript to other experts, asking for a friendly blurb. For example, anything about early aviation will be seen by Tom Crouch, who is the go-to expert. And if the book is written by a faker he will get call the publisher that afternoon saying: "Who is this William Dagwood?!? What does he know about Sopwith???"

Crouch is referenced extensively in the Wikipedia article on the Wrights, as you would expect.

cwilson:

I'd be curious to see examples of egregiously bad pages.

Look at the page for the site whatreallyhappened.com . But don't look too hard for it, because it doesn't exist. Wiki editors decided they didn't like that site's politics, so they not only banned it,they ban anyone from creating a page no that site. And as a final insult, they won't even allow a page of the creator of that site. Even he can't create a page about himself!

I'd say that passes for egregiously bad, wouldn't you?

It is peer reviewed. It's peer reviewed far more often than any paper publication, which is frozen on the date when it's pressed, and more often than most other online references.

I was checking out some page histories. 'Marble' has about 3 edits a day. Unfortunately, about 1/3 of those are vandalism, and another 1/3 is correcting the vandalism. And most vandalism actually was from anonymous contributors. And tends to be pretty obvious like "DANNY IS COOL". Danny was fixed 20 min later. Another was fixed in 25 min, one in 4 min, another fixed in 13 min. Uh, this one took over 4 hours to get fixed.

But there's lots of useful edits that come trickling in. One person adds a kind of marble to the list of marble kinds. Maybe it comes from an obscure town, or it's not in somebody's catalog of kinds of marble, so it wasn't listed before. Another adds a half dozen photos of marble, then the next editor removes most of them because they get in the way of reading the article. One adds a link. Another corrects the link spelling because it doesn't work. One changes 'Pale pink' to say 'Pale pink to cedar-red' somewhere. Another one interchanges a quote with a period. Another rewrites and condenses two paragraphs on the origin of the rock into one.

You can't get this kind of detail collection from a central editor - there's no time and no patience. The editor has an agenda. Sure, wikipedia editors also have an agenda, but it's much more democratic. Meanwhile, there's 3 million English articles.

So Badbone, what REALLY happened to whatreallyhappened? Is this about politics or about political enthusiasts refusing to live by the same set of rules everybody else lives by? Of course, it's the latter.

Wikipedia is a consensual democracy of ideas. This does not work if enthusiast groups wallpaper it with their own private truth sets. They already do this enough that snotty elitists enjoy insisting it's not "peer-reviewed" enough to be useful. To those people I say, if an article's bad, report it. if an article needs editing, edit it. If you're so smart, why are you sitting on your hands?

And to Badbone: an information commons only thrives if we respect each other. If truthers can't respect the normal rules of the road, they'll end up blocked by editors tired of cleaning up their messes. Grow the fuck up.