You know what Wikipedia needs? Game mechanics. Give every registered user x number of commits/edits/reversions. Jack up x based on participation and commits/edits/reversions that "stick". This way, new people can still add stuff (because editors can't infinitely revert) but can't just make a bunch of spam accounts to flood the site.
Obviously, you still do the IP address/superuser stuff as additional checks, but this increases the penalty for adding useless stuff or deleting useful stuff just because you didn't do it.
I feel like Wikipedia's latest move is just a bandage solution to a more fundamental problem. A bandage that causes Wikipedia to be even less inclusionist. (See http://news.ycombinator.com/item?id=761589 for some discussion on deletionist and inclusionist).
I believe Jimmy Wales has indicated that he would like to just stop editing of Wikipedia at some point and save that as version 1.0 and any later edits are a 2.0. This is just a less extreme version of that.
Edit: I also wanted to add that it seems like Wikipedia and Wales is intent on changing. As the saying goes, "where there is change, there is opportunity." I believe this now is when the seeds of the company to overtake Wikipedia are sowed.
How about a more indirect approach. You could make a Yahoo Answer's type site using Wikipedia. People are encouraged to ask questions about particular sections that they don't understand or are lacking information and then points are given for answered questions with citations.
They already have this and, according to some insiders, it's part of the problem.
Basically the claim is that quantity of edits makes it more likely for you to "level up" within the Wikipedia community. This count includes reverts of perfectly good info as when you optimise for any variable, people find ways to play the system.
They also do simpler things like force you to sign up to create a new page, as opposed to edit an existing one.
I'm not advocating tossing all the other stuff out the window. Whatever mechanism they use now to stop spam accounts, keep it. I'm suggesting a different way to aportion the ability to make lasting changes to the site's content.
Jack up x based on participation and commits/edits/reversions that "stick"
I would rather a more sophisticated metric than this. The metric of stickiness seems to work off of the idea that, in the average case, every reversion is a net win for content quality. My intuition might be serving me wrong, this sort of game mechanic seems prone to causing all sorts of ill-effects, gaming and odd edge cases.
For instance, who gets credit for a reversion of a reversion? How can you keep someone with a sufficient value of x from camping on an article and reverting whatever changes a well-intentioned and clueful but relatively new user posts? Since that new user has no more x to spend, the user camping gets rewarded for having his trollish reversion "stick", and the new user gets disenchanted and leaves.
Comments
You know what Wikipedia needs? Game mechanics. Give every registered user x number of commits/edits/reversions. Jack up x based on participation and commits/edits/reversions that "stick". This way, new people can still add stuff (because editors can't infinitely revert) but can't just make a bunch of spam accounts to flood the site.
Obviously, you still do the IP address/superuser stuff as additional checks, but this increases the penalty for adding useless stuff or deleting useful stuff just because you didn't do it.
I feel like Wikipedia's latest move is just a bandage solution to a more fundamental problem. A bandage that causes Wikipedia to be even less inclusionist. (See http://news.ycombinator.com/item?id=761589 for some discussion on deletionist and inclusionist).
I believe Jimmy Wales has indicated that he would like to just stop editing of Wikipedia at some point and save that as version 1.0 and any later edits are a 2.0. This is just a less extreme version of that.
Edit: I also wanted to add that it seems like Wikipedia and Wales is intent on changing. As the saying goes, "where there is change, there is opportunity." I believe this now is when the seeds of the company to overtake Wikipedia are sowed.
How about a more indirect approach. You could make a Yahoo Answer's type site using Wikipedia. People are encouraged to ask questions about particular sections that they don't understand or are lacking information and then points are given for answered questions with citations.
Or you could do something like this http://www.cs.washington.edu/homes/raphaelh/publications/chi... and then reward answers with points.
Check out Mahalo Answers? (http://www.mahalo.com/answers/) I think this is basically exactly what you are describing.
I do thank you for the link to this Kylin project. That was a confounded read but this team is definitely trying to do something amazing.
They already have this and, according to some insiders, it's part of the problem.
Basically the claim is that quantity of edits makes it more likely for you to "level up" within the Wikipedia community. This count includes reverts of perfectly good info as when you optimise for any variable, people find ways to play the system.
They also do simpler things like force you to sign up to create a new page, as opposed to edit an existing one.
I'm sorry, I must have missed something. With these game mechanics, why exactly can't new people make a bunch of spam accounts to flood the site?
I'm not advocating tossing all the other stuff out the window. Whatever mechanism they use now to stop spam accounts, keep it. I'm suggesting a different way to aportion the ability to make lasting changes to the site's content.
Jack up x based on participation and commits/edits/reversions that "stick"
I would rather a more sophisticated metric than this. The metric of stickiness seems to work off of the idea that, in the average case, every reversion is a net win for content quality. My intuition might be serving me wrong, this sort of game mechanic seems prone to causing all sorts of ill-effects, gaming and odd edge cases.
For instance, who gets credit for a reversion of a reversion? How can you keep someone with a sufficient value of x from camping on an article and reverting whatever changes a well-intentioned and clueful but relatively new user posts? Since that new user has no more x to spend, the user camping gets rewarded for having his trollish reversion "stick", and the new user gets disenchanted and leaves.