Altitis

Wolfshead’s excellent blog (on which I can no longer comment because I’m sitting behind a proxy) took the opportunity of a sensationalist news report about Wikipedia losing contributors to revisit the Threshold fiasco, and it’s interesting to go back to what I wrote then, having followed this fascinating train wreck as it unfolded.

For those unwilling to read up the entire background, in January, after a long edit war on the Wikipedia entry for Threshold, one of the more well-known MUDs ever produced, the article was proposed for deletion for lack of notability. Followed two bitter weeks of fighting and concerned / angry / outraged posting throughout the online gaming blogosphere, including reactions from people like Richard Bartle, Raph Koster or Scott Jennings.

Many posts painted a picture of a Wikipedia subjected to the whims and tantrums of the so-called Deletionists, a dark troop of book-burners intent on building reputation by deleting thousands of articles on the flimsiest grounds, about to doom the free encyclopedia with their antics.

I was largely under that same impression – for instance I commented on Wolfshead’s initial post with “The problem I have with the whole matter, like Wolf(shead) pointed out, is that based on the Threshold (and other similar matters, eg the webcomics purge of 2007), I have a hard time conceiving the displayed behaviour as an exception rather than the norm.”

I believe this isn’t too different from what most of the gamers reading about this at the time were thinking.

It’s also pretty much entirely wrong.

In the 10 months since, the English Wikipedia has passed 3 million articles. The Threshold article has remained entirely untouched by anyone since March, and is since then getting between 200 and 300 clicks per month.

And from a critic, I actually turned into an editor myself. See, after the Threshold fiasco passed and my temper cooled off, I read a gaming article, found something that was not correct, created an account, and fixed it. 5’000 edits later, and I’ve found that I was wrong back then.

Deletionists, and their counterparts inclusionists, may make for good headlines because their clashes are always loud, spectacular and sensational, but they are a tiny, tiny minority among regular contributors. There are far more people called deletionists in anger (it’s Wikipedia’s equivalent of Godwin’s Law) than true deletionists among regular editors.

The claim that deletionists build their cred with the internal Wikipedia cabal by getting stuff deleted is patently false. Since I started editing, about 150 editors submitted themselves to the community’s appraisal through a “Request for adminship”, to become administrators with the tools to, among others, delete articles. Among these 150ish editors, how many true deletionists  have been promoted to administrators?

Zero.

Yes, not a single of those decried book burners has been successful in getting access to the tools that would allow them to delete more content. In fact, during all these months, the two most important criteria the Wikipedia community has looked at before voting whether a candidate should get access to administration tools (beyond deletion, they also include the ability to block users and protect pages from editing) were their interactions with other users, in particular in conflictual situations, and the amount AND quality of their content contributions. Or, in other terms, what they have done to build the encyclopedia rather than destroy it.

In 10 months of contributions, I haven’t run into any of the Threshold antagonists, I could probably count the interactions I had with true hardcore deletionists on one hand, and I’ve had exactly two articles I contributed to deleted, on what I found to be valid grounds in both cases.

What is true is that the inclusion standards are strict and become stricter as time passes, and these rely on the four core content policies: that an article be written in a neutral tone, its content verifiable through independent, third party reliable sources and that they contain no original research. The oft-decried notability criteria isn’t even a policy (and is unlikely to become one for quite some time), and in general, pure rationales of “it’s just not notable” aren’t considered in a deletion discussion at all.

Notability itself is a bit of a misnomer anyway. What the notability guideline tries to do is to specify some (relatively) objective criteria a topic must adhere to in order to be written about, most importantly, that the subject has been noticed and commented about by these famous reliable sources in a sufficiently non-trivial fashion that a neutral article can be written on it. If nobody ever heard about a topic before, Wikipedia cannot, and should not, be the first place to talk about it.

Threshold was a typical example of the kind of train wreck that happens when the need for reliable independent third party coverage clashes with a topic that is extremely familiar to a (all things considered) small group of people, where most coverage is available only on specialized, user-generated sites. Many MUDers of old and most people who dabbled in MUD content design or coding are intimately familiar with TopMudSites or TMC, probably used its forums, and have a relatively decent notion of what MUDs were / are imp0rtant to the genre. The older among us will also remember using Usenet rec.games.mudding.* discussions and will definitely be familiar, for instance, with Michael Hartman and Threshold from these days. We probably all remember the MUD Commandos and the design discussions we held there, and from that, we have a good feeling of what is / was key to our hobby and what isn’t.

The problem is, none of the above fit Wikipedia’s criteria for Reliable Sources. And that’s where culture clashes happen, and Threshold is a perfect example of that. The thing is, when trying to build an online, entirely volunteer-based encyclopedia, you need to draw a line somewhere on what kind of information you base your articles. Stupid marketing slogans put aside (If you’re really trying to find “The sum of all human knowledge”, Google + archive.org are probably the closest you will get), should Wikipedia have an article on my pet cat or the great maths teacher I had in 3rd grade (OK, I lied, he was anything but great but you catch my drift)? Should this blog have an entry on Wikipedia? Of course not. My own personal delusions of grandeur put aside, nothing I did so far has been noticed and commented upon sufficiently to warrant an article in Wikipedia, and that’s probably unlikely to change for that matter.

But back to the original topic, or rather, this new round of “look at the disaster that is Wikipedia, people are leaving in droves”. Well, first of all, this statement is pretty much similar to all the doomsday prediction we get when WoW’s overall activity level decreases. Yes, people stop contributing, and yes, you could very easily measure the decay then make a projection on when Wikipedia will die off if you speculate that the curve is entirely linear, and like WoW keeps proving, those predictions would probably end up quite wrong.

That being said, the learning curve for a newbie is steep, in particular since while Wikipedia can be edited by anyone, creating a new article for the first time is usually a quite unpleasant experience.

What typically happens to an article created by a new contributor is:

• It gets first checked by a bot against obviously copy / pasted material, if the bot comes up with a positive result, the article will probably be deleted very swiftly. Wikipedia does its best to weed out copyright issues as soon as it is made aware of it, and many people simply do not understand that just because a text has no copyright notice that doesn’t mean it is free for taking. On average, between 30 and 80 copyright issues are identified and manually verified every day, of which about 50 concern new articles exclusively. This is a first deterrent to new contributors, but one that is difficult for Wikipedia to bypass, because of the legal ramifications.

• Next, it will be visited by the so-called New Page Patrol, and this is where the user experience can become negative pretty fast. The average new user will start an article and write a couple of sentences, then save it, and continue working on it, in a quite natural editing process. Between the first and the second save, though, chances are that a patroller will have visited the article, and, in the best case, placed a series of maintenance templates highlighting its most glaring defects (it’s a draft after all), or nominated the article for speedy deletion. When that happens, the article is often gone before the new user has finished reading through the warnings (provided he got some). This is definitely where the process can be improved, and a discussion is currently underway to seek venues to improve that. Wikipedia has also recently created an Article Wizard which guides new users through a series of key questions to cover in order to build an article that has staying power.

One other reason for the drop in active contributors is that with over 3 million articles, topics with abundant third party coverage become more rare. Also, undeniably, when I talk about a steep learning curve, the trend is definitely worsening as the years pass – at the same time, we’re now in an age where less and less people bother to RTFM before they do anything (I still remember the old Netiquette recommendation that people lurk for a while before posting on a new forum, it seems outlandish nowadays).

Wikipedia has its problems, and among them, the biggest is that when you purport to be the free encyclopedia that anyone can edit, anyone does, and that includes teenage pranksters who believe it’s great to insert swearwords into articles or make up stuff in biographies of personalities. It also includes politically motivated people who will try to influence the perception of the side their backing or of their opponents, it includes PR representatives from small or new companies who want to use Wikipedia as an extension of their advertising campaign, and many people editing sensitive stuff in which they are way too emotionally involved, like national topics or, let’s say, their own MUD’s entry. But protecting biographies of living people is currently the biggest issue, and it’s one of the core reason why the New Pages and Recent Changes Patrols exist.

Because of the fallout generated by vandalism, regardless of pranks or pure malice, many well-meaning contributors get turned away at the door as well. As indicated above, in order to ease the learning curve a bit for new contributors, there’s the article wizard (Wikipedia’s newbie zone if you will), there is also a feature called Articles for Creation where people can start on their drafts and get reviews before their article goes live. In a Blizzardish “soon” way another feature is due to be piloted called Flagged Revisions, which will hide from public view any changes to an article as long as they have not been reviewed by a human – to ensure some of the most well-known fiascos and defamation scandals that have happened in the past will no longer be repeated, and to provide a better layer of protection when events like Thierry Henry’s infamous hand action in a highly disputed qualification match for the 2010 FIFA soccer world cup happens.

The other reality is that in the grand scheme of things, you cannot rely on Wikipedia alone for information. Some articles are excellent, some are abysmally poor. In any case, you should stay well clear of re-using it in a school work for instance. Two things are valuable in any case, though: the footnotes on a well-referenced article will allow you to cut short a bit on bibliographic research and provide you with a good starting point on any topic. And on most articles’ talk page, you will find a quality assessment of the article that should help you decide whether you’re dealing with something Wikipedia’s community has self-identified as decent work or not. Articles are classified on a scale which starts as a Stub (a couple of lines, barely more than a draft), then from Start class to C, B, A, then Great Article and Featured Article. The last three quality classes are subject to multiple reviews and more stringent criteria, and while far from perfect, will give at least some measure of the article’s content quality.

To conclude this overly verbose post, let’s go back to what Wolfshead had written in the immediate aftermath of the Threshold trainwreck:

“The other problem is that the fastest way to earn points is to delete articles, which has the deleterious effect of removing information from Wikipedia instead of adding to it. It is far easier to destroy then it is to create. A project like Wikipedia that deems to hold the public interest with regard to being a virtual library of the “people” like Wikipedia should be held to a higher standard then it currently is.  Their current admin policy which incentivizes and rewards destruction and penalizes creation is woefully flawed and should be overhauled immediately.”

In over 10 months of contributing, I’ve found this assessment wholly without merit. Despite my own initial impressions, generated in the same context, and pretty much in line with Wolfshead’s sentiment at that time, the Threshold case, from what I’ve since experienced, is a rare exception rather than the norm, and the current community’s overall sentiment is, on the contrary, very unforgiving of the few people who try to earn points through deletion of information. Once a contributor is identified as a true deletionist, he has less than a snowball’s chance in hell to get the community’s trust for any additional responsibilities.

The reality is that the majority of the regular contributors are reunited around a single goal, to build the online encyclopedia along the above mentioned content criteria, to remove the “suck” still remaining in way too many insufficiently referenced pages, to weed out plagiarism, copyright violations, vandalism and libel wherever they find it.

They strive relentlessly to produce and improve content, driven with one unified passion, to share knowledge. And they all could do without the drama of these rare, but visible, exceptions like the Threshold affair.

On Similar Matters

• No similar posts.