Why I think that Wikipedia is imperfect, and why the need for a place like Wikipedia Review:

All information is biased, all books are biased, all history is biased. But historians have been able to account for this bias, and to establish more and more accurate texts. It is a difficult job and we are always changing our mind as to what is the established truth, even for centuries-old history (for example the ancient Romans - if you know anything about history, you know about the lies that the ancient Romans created about history, which was discovered centuries later).

Now, we can, to some extent, account for these biases, even the outright lies. Compared to other texts, the lies become increasingly obvious. For example that it was the Greeks, not the Romans, that created the Olympic Games - this was easily provable once enough information was known. The biases can also be taken care of by knowing a bit about who had written the text and where they stood.

Once encyclopaedias, and texts written by groups of people with conflicting interests came about, accounting for lies and biases became more difficult. I can account for biases written by someone who is a member of Al-Qaeda quite easily, but how can I account for a bias in an encyclopaedia? In theory there are no biases, but of course we all know that there are. Thankfully, however, encyclopaedias print who the people who are involved in the collaborations are, and we can research them, and try to discover their slants. All published encyclopaedias have been researched to determine their slants. Whether their slants are nationalistic, political, or simply a combination of the individual biases of the people concerned, if you study it well enough we can account for this.

When Wikipedia first came about, one of its first rules was the issue of NPOV (neutral point of view), something created by Larry Sanger. Now, I don't know why he decided to create this rule, but I for one significantly object to it. Nobody is ever actually neutral, and the problem with aiming for neutrality is that you remove obvious bias and instead create hidden bias.

Now, truth that is biased is still very useful. If I read something written by Osama bin Laden, I know the bias, and I can account for it, but it is still very useful. Similarly if I read something written by George W. Bush, I know its bias. Just because something is biased doesn't make it untruthful, or useless. Indeed, even outright lies can be useful, as beneath the lie can be an important truth.

The big problem with Wikipedia is that, because of the aim for NPOV, they create hidden bias. An individual article may be written by hundreds of people. Anyone reading the article has no idea who these people are, what biases they have, and what kinds of slants and untruths they have added in. We don't know how to take it. We don't know how we can trust it.

Analysis of Wikipedia's articles demonstrates this problem on a higher level - you can separate articles by type, and generally by the degree of accuracy and reliability.

The most accurate types of articles on Wikipedia, in a general sense, are fan-fiction type articles. Spongebob Squarepants, Simpsons, South Park and the like are better sources of information than anywhere else on the internet. There are few fights, and everyone is keen to dig deeper for more information. Whilst articles about various actors and singers may occasionally have minor controversy, generally these are of the highest degrees of accuracy too.

The next most accurate types of articles on Wikipedia are purely factual articles. Scientific, medical, mathematical and the like generally have few arguments, and people try very hard to be completely 100% accurate about them all. The issue of bias is hardly relevant as they can be neutral very easily. So long as you accept that 1+1=2 you are fine. These articles are as good as what exists anywhere on the internet.

The third most accurate types of articles on Wikipedia are on long-established historical truths. Whilst in the past we did have problems with lies by Ancient Romans and the like, we have now fairly reasonably established truths about such things, and we can all pretty much agree. If we can't, most of us aren't knowledgeable enough about them to argue anyway, so all is fine. These articles are of a reasonable level of accuracy. If only we had more historians on Wikipedia, they would be better. They are not quite as good as established encyclopaedias, but they are good enough.

The fourth most accurate types of articles on Wikipedia are on current events, recent history, politics, and living people. Such things as George W. Bush, someone who recently died, anything happening in Iraq, 9/11 and the like, have many problems. With these kinds of articles there are multiple valid viewpoints, and the problem with Wikipedia's NPOV is that it only allows 1 article on it. You can't have "9/11 according to Republicans" and "9/11 according to Democrats" and "9/11 according to people in Iraq" etc. You just need one article. And these points of views aren't necessarily lies, but the problem is that Wikipedia has to at some point make judgement calls. Do they allow merely one point of view, or do they talk about all of them, but about one mainly? The problem with all of this is that by having multiple authors, they cannot ever have an accurate article on this kind of thing. The articles end up being ridiculously long, overly complicated, ever changing, and nobody can account for the biases, as one paragraph might be written from one point of view, the next from another, and so forth, until it all ends up looking like complete and utter hogwash. In spite of being one of the most viewed articles on Wikipedia, the article on George W. Bush has been analysed as one of its least accurate articles. And this is not something that can be fixed by having more people edit it. It can be fixed by having less people edit it. This is where Fred Bauder's idea of Sympathetic Point of View, which he created on his Wikinfo site, is very useful. There we can have 10 lots of George W. Bush articles. Unfortunately, Wikinfo never really hit off, but it is a great idea.

The fifth, and worst type of article, are highly controversial topics. These can include well known recent murders, political lies (weapons of mass destruction, children overboard), mysteries, and so forth. On these kinds of articles, there is no established truth. It isn't so much a matter of biases, it is a matter of truth. A number of legal cases have suppression orders over them, about what can be said. And there are groups of people who work to try to uncover the truth, to investigate things, and so forth. Now, when Wikipedia comes to write articles on these types of articles, they fail abysmally. Wikipedia as a general rule completely ignores all investigations, and instead decides to focus on the tiny amount of information that is officially released. Then they have a point of view on it. The end result of all of this is that Wikipedia's articles on these kinds of topics are primarily full of invented facts, absolute lies, and things which have no relevance to anything. And herein lies the problem - on a number of these topics, thanks to suppression orders, Wikipedia remains the only comprehensive coverage of these topics. This means that then news reports often are forced to use Wikipedia as a source. The Wikipedia article can then use these news reports as proof for their article, hence justifying a lie. Now, this has happened in the Port Arthur massacre article. I believe it has happened on many other articles, including the Pan Am Flight 103 article, although I don't have sufficient knowledge of the topic to know for sure.

Now, if everyone out there knew how accurate Wikipedia was, there would be no problem. We could all use Wikipedia as a resource, as a first step, as about as useful as a Google search, then check our facts, and so forth, then everything would be fine. If people used it factually on the types of articles which it is most accurate with, then everything would be fine. It wouldn't really matter.

The problem is that Wikipedia is taken far too seriously, when it really shouldn't be.

This is where places like Wikipedia Review come in.

As one person, I am only really knowledgeable about a handful of topics. I can comment on these, but I know very little about others. So others can collaborate, and sort out what are the danger areas of Wikipedia.

There is no reason at all that a place like Wikipedia Review cannot in some way work in collaboration with Wikipedia. The aim of Wikipedia Review isn't to destroy Wikipedia, it is to fix it, and to educate people about how to use it effectively.

The greatest fear that I have is that if we end up trusting Wikipedia too much, all paper encyclopaedias will be thrown out, and all books, other than fiction books like Harry Potter, will be thrown out too, and we will be left with one central pool of knowledge.

This is all fine and good if Wikipedia is 100% accurate. But can anyone honestly say that it is?

I don't think that any serious critic ever worries about vandalism as a problem, by the way. Vandalism is not the problem - it can be fixed in seconds. The problem is established lies and unaccounted biases.