Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The rise and decline of Wikipedia [pdf]

www-users.cs.umn.edu

Post reply on HN
Print view View original

31–40 of 156 posts

Re: The rise and decline of Wikipedia [pdf]

#31

Post by Retric »

lawpoop wrote: post #20
placeybordeaux wrote: post #14

Earlier quoted context omitted.

The decline in number of editors seems to have started at 2007 given figure 1. I wouldn't tie the amount of negative press to the decline in editors. The point about there being less new stuff to write about seems apt, what do you think about the counter points that the article had? "First, as noted above, the vast majority of articles in Wikipedia are still below community standards for “good” articles. Second, unde…

Not OP, but 1. Isn't this a moving target? In the beginning, all you had to do was throw up text; standards evolved (and probably continue to, I don't follow it), and now you have to cite references, adhere to standards, etc. As more articles are reviewed, more probably "become" substandard. 2. Is this really a decline issue? I guess it is if you are considering decline not as failure to maintain but failure to grow.

I contributed to several Wikipedia articles over the years (~50-100). It's gotten to the point where it's pointless to improve even terrible articles like stubs as basically everything is automatically reversed. At best a mod may decide to add your content back, which means at this point it's basically stopped being a public wiki.

Which is fine, IMO they should just change the name to Webpedia or something if they don't want new edits.

PS: As to being 'full' they have 2.7 million stubs, that's a lot of missing content. https://en.wikipedia.org/wiki/Wikipedia:Statistics

Re: The rise and decline of Wikipedia [pdf]

#32

Post by Animats »

So the peak was in 2007. I think I know why.

In 2006-2007, there was an organized push on Wikipedia by pro-Israel editors to make Israel-related articles favorable to Israel. This provoked some major controversies, especially because at least one of the pro-Israel editors was a Wikipedia admin and abused their authority. (Eventually, they were "desysopped" for that, but it took years.) For a sense of this, see [1], including all the archived talk pages.

Wikipedia dealt with the problem by strictly applying the "reliable sources" rule. It became the practice that every single fact in Israel-related articles must be cited to a reliable source and properly footnoted. Anything that wasn't was deleted as "uncited" or "synthesis" by one side or the other. Until that time, articles often had text which wasn't cited to a specific source. The tough citation requirements anchored the dispute to factual information, and, though the process was painful, worked. The controversies settled down and the edit warring slowed to a crawl.

Those strict standards gradually spread to the rest of Wikipedia. Now, everything had to be footnoted like a doctoral thesis. Few people outside of academia write that way. This results in a big hammer coming down on someone new wanting to write an article about their new favorite DJ. Does he have mainstream press coverage that isn't from a press release, and did you cite it? If not, he's out of here.

Until you learn to write in that strict style, you can't contribute much to Wikipedia. For people used to blithering on blogs, Facebook, and Twitter, this is hard. That's what's discouraging new editors.

Writing for Wikipedia is now like contributing to a major open source project like the Linux kernel. There's a big body of code, and your changes have to fit in well and be justified. That's necessary. Wikipedia faces a constant incoming stream of PR, vandalism, and general crap. (Check out the ever-changing list of recent changes.[2]) The petty vandalism is mostly reverted by AIs now; ClueBot, driven by an ANN and trained on known vandalism, is quite good. The other problems take human attention, by volunteers tired of seeing similar annoyances over and over.

For the other extreme, of writing about anything with no quality control, see 4chan.

[1] https://en.wikipedia.org/wiki/Talk:Israeli_West_Bank_barrier [2] https://en.wikipedia.org/wiki/Special:RecentChanges

Re: The rise and decline of Wikipedia [pdf]

#34

Post by rloc »

runn1ng wrote: post #18
rloc wrote: post #13

I tried once or twice to edit articles but was quickly rejected by editors that usually believe they sort of "own" the content. I've quickly become demotivated as explained in the analysis and never contributed again. Another problem is that when you think you've made an important and useful update you're never sure when someone will roll it back. Some are spending a very big chunk of their time to make sure articles…

I have learned that in order to actually contribute to Wikipedia, you have to treat it less like a collaboration and more like a battle. You can't just contribute and consider it done. You have to be prepare to fight your way through, engage in edit wars, learn about the battlefield (the bureaucracy, the article sourcing) and in the end, if you fight long enough, you usually win. You have to repeatedly, over and over…

Yes that's exactly what I experienced. A knowledge repository built upon who fights harder and who gives up first (in most cases due to lack of time arguing for every sentence) is kind of broken.

Re: The rise and decline of Wikipedia [pdf]

#35

Post by mtw »

I rely on wikipedia like everyone on hacker news but I can't help thinking that changing their status to benefit corporation like Kickstarter would improve the structure and also their finances. Non-profit closes a lot of doors, especially growth opportunities. A renewed status could bring new mobile apps, tie-ins with media companies, perhaps an open API available to third-party developers etc.

Re: The rise and decline of Wikipedia [pdf]

#36

Post by js8 »

I think the reason for decline (and bureaucracy) is actually a lack of democracy. For example, there is no way to recall administrators. Also, opinion of majority is sometimes dismissed as not being good consensus, which only helps people in established cliques. Democracy is a rule that ensures fairness, but senior administrators do not always want it.

Secondary problem is IMHO the caste nature of administrator privileges. I think if the privileges were more separated and it would be possible to recall them, this would reduce the ossification of the hierarchy.

I also think Wikipedia could benefit from having a stable branch into which the changes from unstable branch could be regularly reviewed for addition. It would be nice if unstable branch could return to the idea that someone can add things even if they are not perfect or immediately sourced; IMHO experts can always add sources later for content that is already present, and it would give them some incentive to contribute.

That being said, Wikipedia is amazing! I would really like to thank to people who put some work into it.

Re: The rise and decline of Wikipedia [pdf]

#37

Post by Balgair »

runn1ng wrote: post #18
rloc wrote: post #13

I tried once or twice to edit articles but was quickly rejected by editors that usually believe they sort of "own" the content. I've quickly become demotivated as explained in the analysis and never contributed again. Another problem is that when you think you've made an important and useful update you're never sure when someone will roll it back. Some are spending a very big chunk of their time to make sure articles…

I have learned that in order to actually contribute to Wikipedia, you have to treat it less like a collaboration and more like a battle. You can't just contribute and consider it done. You have to be prepare to fight your way through, engage in edit wars, learn about the battlefield (the bureaucracy, the article sourcing) and in the end, if you fight long enough, you usually win. You have to repeatedly, over and over…

I do not think you are at all lying. However, why in the name of all that is holy would I ever do that? Literally, cleaning my toenails is a better use of my time than that. Doing my neighbor's taxes is a better use of my time. Getting high and watching the news is better use of time. It just seems so utterly hopeless and pointless to even try from what you say.

Why do you do this?!

Re: The rise and decline of Wikipedia [pdf]

#38

Post by Marcomasino »

digler999 wrote: post #11
_pmf_ wrote: post #10

I think Wikipedia will become the best case study in how bureaucracy cripples a working society. 80% of the effort of creating new content is following byzantine processes and procedures. While I admit that these processes work for consistency of content, I cannot in the least observe any noticeable quality improvements, and the reason is that valuable contributors with deep domain knowledge will not invest the time…

> OCD of some arbitrary editor suffering from Asperger. Is it really necessary to use mental illnesses in a pejorative way to make your point ? How do you think it makes people feel who legitimately suffer from those illnesses ?

I know exactly what _pmf_ means. I used deal with a self declared Asperger who was top class in his field. The thing is whenever I asked him do you know how to fix this problem. The answer was invariably, 'sorry I don't know how to do that'. Of course I should have phrased it as, I am having difficulties in solving this problem and was wondering if you would advise me on finding a solution.

Re: The rise and decline of Wikipedia [pdf]

#39

Post by sparkzilla »

An interesting paper that basically says that the number of editors is declining due to the difficulty in keeping and retaining new users. Any person who has tried to edit Wikipedia will know that aggressive rule-wielding wikifanatics immediately try to kill newcomer edits using whatever tools they have to hand. But it's important to realize that Wikipedia's ambiguous rules and arbitrary decisions are not a bug, but a feature, giving power to incumbent editors. While most new editors just give up, the ones who do stay on and fight become the same kind of fanatics.

Many people say that the toxic community is the problem, but the community issues IMHO are actually a symptom of poor software design. The wiki's open structure and article-based format, which allows the most powerful groups of editors to "own" pages, gives power to people who are interested in keeping and retaining control, rather than those who are interested in sharing knowledge. The power is centralized in these gatekeepers, leading to all the problems we see on the site.[1] This is also the reason the Gender Gap exists.[2]

The solution is to redesign the software to devolve power from entrenched groups, but this will never happen because there is no leadership, and the people who have the power are the survivors/fanatics that actually like the current system. Because they write the rules they will not give their power up easily.

As for the quality arguments below: without new users and continuous evolution in software development, content presentation, and editorial review processes, the site cannot maintain quality. I believe that it is impossible for any wiki system to maintain a high quality anyway: the amount of time to improve articles to high quality is exponential, while the number of editors and work available is linear. Even in the best circumstances, something has to give. If the number of editors is actually declining then quality problems may also increase exponentially.

[1]http://newslines.org/blog/wikipedias-13-deadly-sins/

[2]http://newslines.org/blog/the-sexists-at-the-top-of-wikipedi...

Re: The rise and decline of Wikipedia [pdf]

#40

Post by sparkzilla »

zkhalique wrote: post #29

Isn't this more describing a typical s-curve with regard to the information being mostly filled out? Wikipedia isn't a news site.

Actually most of the popular articles for readers and editors are news-based. However, the site, while barely adequate as an encyclopedia, is terrible for news presentation.
Post reply on HN