Who (Really) Writes Wikipedia

Nico ? La ! writes "Aaron Swartz questions Jimbo Wales' (Wikimedia's founder) belief and evangelized truth that only around 500 people are the most important contributors to Wikipedia. Whereas the truth is that they probably are the people who do the most editing. From the post: 'For example, the largest portion of the Anaconda article was written by a user who only made 2 edits to it (and only 100 on the entire site). By contrast, the largest number of edits were made by a user who appears to have contributed no text to the final article (the edits were all deleting things and moving things around).'" Which ultimately means that Wikipedia in some ways much more closely mimics a real encyclopedia, with many contributors writing the bulk of the content, but a small group massaging that text to insure standards compliance with the overall work. Interesting thing there and worth your time, although the super-computer thing doesn't make a lot of sense to me.

More Statistics & What I Expect

[eldavojohn]

But when you count letters, the picture dramatically changes: few of the contributors (2 out of the top 10) are even registered and most (6 out of the top 10) have made less than 25 edits to the entire site. In fact, #9 has made exactly one edit -- this one! With the more reasonable metric -- indeed, the one Wales himself said he planned to use in the next revision of his study -- the result completely reverses.

With the shear mass of people writing on Wikipedia, I think you'd be able to find examples of the articles that were heavily contributed to by a large group of people (say things like Fermat's Theorem) and also things that are primarily the work of one person. Frankly, that's what I expected of Wikipedia.

What about these statistics [wikimedia.org]? Could Wales perhaps post average number of edits per page with a standard distribution? What about the same for average number of users contributing to page? What about statistics for average number of characters changed per edit?

Things that have many books written about them are going to be edited by a lot of people that read those books (like The Beatles [wikipedia.org]). But if I want to read up on Procul Harum [wikipedia.org] (A not-so-well-known rock band), I'm assuming that there is some die hard nutjob out there with two children named Procul and Harum that filled in most of the information in that page.

Is this a good thing? Well, yes and no. I think The Beatles' entry holds to more rigorous standards than Procul Harum's on account of the possibility of one person unintentionally inserting their personal views into Wikipedia. For instance, "Known as the World's Greatest Rock Band" may be appropriate for The Beatles' page but not for Procul Harum's. Yet, we all know how insane fans treat their favorite bands. Passion and emotion are not useful tools when authoring Wikipedia or history in general. And that, in my opinion, is Wikipedia's greatest hinderance.

Wiki Language Issue

[eldavojohn]

They point out that it's only a 500 people editing most of Wikipedia. But are they talking only the English version. Something to note is that many of the Top Users [wikipedia.org] of Wikipedia are language bots concerned with propagating interwiki information accross languages. Number one is Hashar for the French language. I imagine the strategy is that getting a bad translation only requires someone fluent in French to correct for it to be a good article -- that person doesn't need to know the information because it's pretty much already there in broken French.

Another large contributor by number of edits is GuanoBot who's only job is to bypass redirects.

Are these bots that are helpful skewing the statistics because they are needed for maintenance?

Editcountitis

[Dan East]

As with anything that keeps stats or running totals, there are those that seek to achieve the highest count possible. Wikipedia is not immune to this. There are those that will make 50 small, distinct edits to an article (each comprising minor changes, like punctuation, formatting, spelling corrections, etc) to increase their edit count.

It is my personal experience that those with the highest edit counts peruse any and all articles applying Style Guidelines. This results in changes like like correcting capitalization of headers ("External Links" -> "External links"), placing bullets in front of external links, formatting dates, wikifying appropriate words, updating links that redirect, etc. Once a person becomes familiar with the guidelines they can easily nitpick pretty much any article and find something to correct (or at least change to their personal preference).

Also, don't forget those that run bots. That's a very easy method to rack up edit points.

Dan East

Much like Digg

[otisg]

This is not a surprise. That is simply another example of nature's laws on the web. This is not much different from the now well known fact that most stories on Digg are submitted by a handful of people (see: Top 100 Digg Users Control 56% of Digg's HomePage Content [seomoz.org]).

Re:Wiki Language Issue

[hansamurai]

Bots have to be labeled as bots, one way or another, when created and editing articles. So bots are not included in top editors, etc. usually unless you want to see them specifically. I highly doubt Wales included bots in his top editor count.

Who really vandalize Wikipedia.

[Anonymous Coward]

The trend is also for the vandal "community". I am a retired vandal (as of September 2006). What the Wikipedia community doesn't want you to know is most of the vandals are actually ex/undercover Wikipedians who vandalize either to get revenge or to test the system. The vandals are actually very technical in their nature. and some even exploit vulnerabillities in the mediawiki code. The most infamous attacks happened in February 2005 (goatse on the main page), August 2005 (Willy on Wheels replacing the Wikipedia logo) and October 2005 (thousands of randomly generated accounts attacking pages with SUPER COOL on them).

Reecently, a lot of the vandal community has been under attack due to an eassy known as WP:DENY (basically WP:OFFICE for vandals). While loads of categories have been deleted off the Official "face" of Wikipedia, the vandal community are still at large, even when they are just lurking rather than actually vandalizing. It is the fear of vandalism that is worse than the results.

Even Willy on Wheels is harmless, he just likes to put as many pages "on wheels", and due to his method, to be "on wheels" is like the vandal version to Wikify. A point of contention is the "anti-vandal" community, which discredits a lot of Wikipedia. The CVU aka counter vandalism unit is an embaressment and just encourages vandals to play games instead of going to work on serious projects. In fact their latest "logo" features a anime girl with a riot shield. Any vandal who sees that will just laugh.

If Wikipedia wants to eliminate vandalism, they need to get to the core. Stop being "on wheels". Actually enforce policy, be liberally inclusionist, allow all points of view be addresses fairly and do not bite users and replace AFD with something more sensible, like outsourcing topic scope to Wikiprojects instead of having random people say keep or delete which is utter ignorance.

I am a vandal, I do not like vandalizing, but until Wikipedia gets the point and respects its users its going to keep attacting new ones. It will only be a matter of time before a vandal gets shell access and runs a SQL statement that vandalzes millions of pages, which will shut the Wiki down for quite a long time before it is fully restored.

Re:Define: Important

[pmc]

Wikipedia is trying hard for quality, hence the importance of copy editors - those quick edit users who do a lot of banging articles into shape.

Depends what you mean by quality - grammatically correct and logically ordered is one thing. Actual content - possibly in rough and ready form - is another.

What JW is (apparently) arguing is that what he considers the contributions to wikipedia is best measured by "edits", and that by this metric there is a hardcore of 500 users that do most of the work.

What the article argues is that the a better measure of of contribution to wiki is raw material, and that far from 500 people doing it, it is actually orders of magnitude more than this.

My opinion is that anyone treating all edits are equal and using that to derive a metric for measuring user contributions to the site is using a seriously flawed method. Selling the output of this method as "The Truth about who Created Wikipeida (or the Tale of the Noble 500") is just trying to invent history to fit their preconceived notions.

Nobody is the most important

[wackymacs]

I am a Wikipedia contributor, with around 5800 edits. I started editing in 2005. Most of my edits are minor ones, cleanup, etc. Although I have contributed a lot of text to many articles, I also enjoy contributing free-use images and providing references for articles in order to improve the reliability of an article; too many articles on Wikipedia do not cite their sources. Jimmy Wales should not state that only several hundred users are the most important editors on Wikipedia, because in reality everyone on Wikipedia who is contributing, whether by adding text or by cleaning up articles or doing maintenance, they're all helping out in one way or another. Most people all work on different stuff, so nobody can be called the "most important", its too broad a term.

Re:Duh, they design it that way

[voice_of_all_reason]

"The vandal hordes" knocking at the gates is a self-sustained myth. Take a look at the history for Steve Irwin's page and see how many of the good-natured edits revolving his death were reverted. First blatantly as "vandalism" then repeatedly as "unsourced" by every article provided until it showed up on an american news portal (cnn or abc). Even harmless edits like tributes ("you will be missed, steve"), while really not helpful, are classified as "vandalism" when people who don't understand Wikipedia keep putting them back in.

So? What's the point?

[junglee_iitk]

I personally find that one cannot just go and pour in all the information in as much as 1000 of pages. Anywhere on the internet. Wikipedia is no exception. Any serious editor who wants to contribute positively knows that most of our knowledge is prejudiced (received-knowledge). Most of us when get down to write something which needs accountability, we just write what we are absolutely sure of. Of course there are different levels of how-much-surety-means-i-am-correct for every individual, but that is why one cannot have 1000 edit counts (let alone edits on 1000 different pages) all referring to content addition. Those edits include spelling corrections, reversions after vandalism, discussions and many other things.

Number of edit counts do show one thing though: your obsession with Wikipedia - to stay there and do something. Having higher priority in discussions, becoming an admin, etc. are all rewards for that, and come on guys, it is nothing unexpected! Wikipedia has long passed the point of just being a site anyone can edit, it is more of a community.

PS: Slashdot is different because it demotes accountability and promotes heated discussions. Wikipedia is just opposite of it.

Re:Editcountitis

[Chacham]

It is an encyclopedia

It *pretends* to be an encyclopedia. It is not, however. Due to its very nature it is a mixture of current thought and old grudges, with the latter being less influential. Calling it an encyclopedia is like mixing all Slashdot comments into one comment and modding it informative.

Alternatives to edit counts

[Animats]

The "edit count" fanaticism is indeed a problem for Wikipedia. The edit histories of some of the editors with the highest edit counts are disappointing. Some of them never actually write anything; they just make administrative edits. Others make vast numbers of very minor edits.

Better metrics are possible. A metric like "number of words added which stayed in an article for at least 30 days" would measure useful contributions.

But this isn't the real problem with Wikipedia. The real problem is "churn". Articles do not steadily improve over time. They typically reach about 80% of the "good article" level, and then slowly change over time, with edits of varying quality.

For a striking example of this, see Horse [wikipedia.org]. Take a look at the article at three month intervals. The article is so heavily edited that it changes almost completely every few months. Yet today's version is really no better than the versions from three and six months ago. That's churn.

Or ...

[Martin Spamer]

... more likely to suffer from Group-Think [wikipedia.org].

Re:It's true, though they're still important edits

[greenreaper]

You can type on a keyboard while wearing a fursuit?!!? I use a mime-to-text converter. :-)

Re:Your voice is still not being heard

[DerekLyons]

If you are no longer contributing, then your voice still won't be heard. Reputation on Wikipedia is immensley important. That makes sense: why would policy decisions be made by those who haven't proven they understand the goals of the project or those who don't have a track record of improving the site by contributing articles?

You miss the OP's point - reputation isn't made by proving an understanding of the goals, or by contributing articles - it's made by amassing a large number of edits. They key problem with this is that and edit can be anything - even a quick spelling correction earns you a line in your earns you a line in your 'user contributions' log, and creating a new article equally gains you single line...