User talk:Scs/Wiki karma proposal

Rationale
Personally, I don't like any scheme that has an arbitrary, binary distinction between "administrators" versus "users", or between "privileged" versus "ordinary" users, or between "draft" versus "approved" articles. I don't think a binary scheme can be flexible enough; I think the qualities we're measuring here -- experience of an editor and maturity of an article -- are continuous variables. Also there's eternal bickering over those labels and statuses and who deserves them or not. Also any system that requires human intervention to approve or endorse or accept (a user or an article or a change to an article) doesn't scale well. Also these approval mechanisms represent a psychological barrier to entry, as opposed to a scheme that appears to be completely open, whose limitations on new users are applied somewhat more quietly and implicitly, as part of the infrastructure. Steve Summit (talk) 00:23, 9 February 2006 (UTC)

Comments
I am concerned that such a system would drive off new editors. Who wants to have to make lots of minor, insignificant edits just to be allowed to add real content to articles they care about? For example, I have contributed significantly to a featured article and started editing it heavily from around my fifth edit. If this system had been implemented when I finally got around to editing I'd have been bored to death and lost interest rather quickly.

I think such a system would really do away with the 'anyone can edit' philosophy that has made this site so successful as the vast majority of articles would be cut off to the vast majority of people.

And having editors rate each other will open doors to some rather massive abuse, I forsee. Kamryn · Talk 14:41, 5 August 2007 (UTC)

And this really concerns me:

"Our experience -- sorry to be blunt about this -- is that "reasonable" editors don't have these problems. We're sorry if our definition of "reasonable" does not match yours, but this wiki can only work if all of its participants share a definition of "reasonable behavior" which allows them to collaborate on the editing process in a spirit of mutual trust and cooperation."

See no true scotsman. I would be vastly against a policy which tells people that get bullied: "tough shit". Kamryn · Talk 14:44, 5 August 2007 (UTC)

My comment.
Yuck.

One sign of baddness in a proposal is when it has lots of complicated heuristics which lack objective data to support their correctness. There are an infinite number of similar but clearly distinct possible proposals, and we have no means to choose among them. Instead of this proposal why not set the karma to equal the root of the sum of the squares of the probabilities of each word added to Wikipedia? ohoh and multiple that by log10(days since account creation) and and...

This proposal also suffers from the common misconception that we can accurately determine if an edit survived. I've seen a number of attempts to automatically attribute text in Wikipedia articles and most fail spectacularly, some of the best just fail amusingly. Complicated multi-edit reverts, text moves, and re-phrasings make this a really hard problem. If the proposal was just advisory the lack of accuracy might not matter much, but since the ability to edit depends on the users score this is an important problem.

It also doesn't take much time looking at RC to see that a lot of perfectly good edits are made by new users, even to well established pages.

To me this seems to be a proposal which would substantially constrain who is allowed to edit, damaging openness just as the proposal accuses other proposals of doing.

The leading proposals for Wikipedia are for stable versions. As proposed, stable versions doesn't limit who can edit at all... in fact it enables, more openness by removing most of the need for protection. What stable versions does is it decouples "who is allowed to edit" from "who is allowed to publish the article for viewing by hundreds of members of the public per minute". It does this so we can be more confident that someone with at least a little trust has reviewed an article before pushing it public.

If this proposal were instead re-factored into a proposal on how to determine who can push a new stable version rather than edit a page, my view on it would likely be very different. I think it's more acceptable to have fuzzy (and inaccurate) rules for less important tasks. Though if such a refactoring were to happen, I think that page-view rate rather than article maturity should be the controlling factor for the adjustable per-page threshold.

Thanks for the food for thought. --Gmaxwell 15:16, 5 August 2007 (UTC)