Too many dubious "facts" get put into Wikipedia. Later someone notices it is dubious, but is reluctant to delete it, so tags it [citation needed]. I think this is the wrong approach. If something looks dubious, it probably is. And in that case, the burden of proof is with the "fact". So if I see something which I know is dubious, I have no hesitation in deleting it. I rarely if ever attach [citation needed] to dubious statements. I just delete them.
Wikipedia editors need to be very careful with their facts, because their work is plagiarised, oops I mean used, all over the web, creating authoritative-sounding sources and obscuring the real sources.
Example: back when I created the Tracey Wickham article on 22-Feb-2006, I wrote that she was the youngest ever Australian recipient of an MBE. Unfortunately I neglected to cite my source. When this fact was queried on 4-Jan-2007, I tried to find my source. Alas, I found lots of sources - but none which I was sure were not simply ripped off the Wikipedia article. No good citing them, that's a circular source. So I deleted the claim.
Moral: Wikipedia is widely cited, so it needs to be thoroughly cited itself.
Anonymous authors do more harm than good. Wikipedia is big enough now that it doesn't need them. I would not object if the default was to Semi-Protect articles.
I despise most external links. I don't mean footnotes or references. I mean the cluster of links in the "external links" section of an article. These are external articles which are tagged "sort of relevant", but are not relevant to be used in a citation or footnote. They are a happy hunting ground for wikispam: people increasing the hits on their page by putting it onto Wikipedia.
If an external link adds nothing, or even little, I am happy to delete it. I particularly despise external links created with Wikipedia text. I often (and probably should always) delete them without hesitation, regardless of what good content is in them.
The only external links that really deserve to be on a page are "official" links, such as the official page of the subject of the article. Anything else, if it cannot warrant being a footnote or reference, should probably be deleted.
If I was setting policy on external links, it would be this: "Every external link should be accompanied by a short sentence explaining its relevance." Later comment: in fact Wikipedia:External links#How to link includes the sentence, If you link to another website, you should give your reader a good summary of the site's contents, and the reasons why this specific website is relevant to the article in question. So the guideline exists, it's just usually ignored. Peter Ballard 02:29, 28 May 2007 (UTC)
I love links which are relevant. Click on a related topic, and read more about something. Links are what made HTML - and hence the World Wide Web - great.
But what I despise are links to words which everyone understands, or to years (as in 2006) or days (as in December 7). They just add pointless clutter to an article, and obscure the important links.
I know it's in the point above, but it deserves a separate section. Links to dates are pointless and add nothing to an article except for unnecessary blue clutter.
However, in my opinion, links to bare years (like X was born in 1977) are pointless and should never be inserted. The manual of style discourages them, WP:DATE#Autoformatting and linking saying, "Wikipedia has articles on days of the year, years, decades, centuries and millennia. Link to one of these pages only if it is likely to deepen readers' understanding of a topic."
After seeing disputes and arguments over whether certain chess players really are Jewish or not (especially when said players don't make a big deal of their religion or ethnicity themselves), I've come to the conclusion that racial and religious categories (like "Jewish chess players") are A Bad Thing. And the same goes for ethnic categories. For the purposes of Wikipedia, I don't care what race or religion a person is(*), unless it directly relates to their field (a Jewish theologians category, for instance, is quite appropriate). The categories seem to only exist for racial and religious point scoring ("See how many eminent people in field X belong to my race/religion!")
p.s. I've just discovered this is a known issue. See Wikipedia:Overcategorization (first created 23-Nov-2006)
99% of these are Stoo-pid!! If a person, who is already famous, gets mentioned in a TV show or song, that isn't noteworthy. Like Trivia sections, these sections are just a gathering ground for random factoids that can't be fit into the article in a meaningful way. These "References in Pop Culture" sections trivialise the article and actually detract from Wikipedia.
These sections are also inherently WP:Original Research, because it is up to Wikipedia editors to decide which items belong in the lists.
Pop culture references are only really relevant for otherwise unnotable things which have passed into pop culture. Claytons is an example of an article which deserves a pop culture section - because that's the only reason anyone remembers Claytons at all!
I think a good rule of thumb for popular culture references is this: they only makes sense when referenced by something or someone more famous.
Some editors like to add a paragraph to a person's article every time they make the news. Bad idea. Some news items turn out to be irrelevant to their career as a whole. It seems to me that a good rule of thumb is to wait a week or two before putting news stories into a biography (except for things that are obviously significant, e.g. election results for politicians). The guideline I would suggest, I christen the "one month / five years rule:
On a related note, and I see this especially with politicians, is the temptation to only insert negative news items about a person. Are they accurate and referenced? Absolutely. But are they notable, in the context of the person's career? Often not.
Or to paraphrase Animal Farm, "edit once good, edit twice bad".
Many articles duplicate material from other articles, rather than (or as well as) putting in a link. The problem is that any fixes then have to be done twice. Articles should be succinct, with lots of links, rather than trying to duplicate what is in those links.
I don't like heading names with implied judgements ("the youngest ever XXXX") or trying to convey a sense of drama ("Success, failure, success"). Headings should be staid and informative, giving a guide of how the article is divided up.
Wikipedia help is a mess. I see a guideline once, then can never find it again.
Wikipedia help needs a comprehensive Table of Contents. (Help:Contents/Site map and Help:Contents are not comprehensive).
If Wikipedia help has an index or a contents, I've never been able to find it. Which is just as bad as not having one at all.
Every now and then, when a good reference is not available, a dubious one is used. Sometimes, a little deduction - which some people class as WP:Original Research - is better than using the dubious reference.
Example 1: Should Garry Kasparov be in the Category:Jewish chess players? In the absence of a statement from the man himself, I did a little research. Kasparov's father was Jewish, his mother is not - those facts are not in dispute. Nor is there any evidence that he identifies as a Jew. According to Who is a Jew?, all significant branches of Judaism do not consider such a person a Jew. But because someone could find an internet source calling him a Jew, my argument was dismissed as WP:Original Research. Sigh. I only dropped the matter because I decided #Racial and Religious categories are A Bad Thing.
Example 2: How many people play chess on the internet? The chess article used to say 285 million, based on a single reference: http://www.turowski.com/chess/world2005.html . I checked the membership of the largest online Chess servers, and concluded it would struggle to exceed 1 million. Thankfully this time common sense prevailed, and my WP:Original Research overruled the reference.
Usenet is generally not regarded as a WP:Reliable Source. But sometimes it is. In particular, from the late 1980s to late 1990s, news items were often posted on Usenet, and these postings (archived in Google Groups) are often the only online reference available. So, sensibly used, Usenet can be a good and useful source.
This is for my own personal reference: useful pages I sometimes watch, but which I don't want cluttering my watchlist:
Articles I want to add to when I get the time or find the answers.
