Wikipedia talk:Decision Making Process

Trying to understand the process
I'm probably going to come at this at a totaly different perspective from everyone else. I don't realy understand the need for this decision making process in the first place. As long as a majority of the people have broadly the same goals for wikipedia the "good" edits will always fix anything that goes wrong in the first place, and in areas where people realy care about something they can always go change things themselves. If a majority of people have "bad" intentions you're screwed anyway.

A rough consensus, a bit of civility, and an expectation that if you want something you'd better do it yourself should be all that's required.

Why am I saying this? I don't realy want to participate in the "community" or anything like that, I'm probably never going to join any mailing lists and I only just got a user account because I'm a bit twitchy about my IP being displayed. I just want to get on with working on the 'pedia and I don't want to worry about politics, rules, and regulations, or whether I did the formatting for the article the "right" way. And I especialy don't want to have to make an effort to keep up with them. From where I stand things look fairly simple, follow NPOV, use common sense, and when in doubt follow previous examples on how things are done.

So to summarise, I don't see why a rough consensus, common sense, and an expectation that if you want things done a specific way, you'd better do it yourself can't work (In other words making a decision isn't going to stop morons, trolls, or crap flooders but fixing their errors however will). It certainly appears to be working a whole lot better than the Nupedia way. -- v


 * I strongly agree with all of that. I vote for no voting on Wikipedia :). Enchanter

okay, as much as I am personally convinced that a decision making process is necessary, I play the advocatus diaboli now (feel free to move stuff into the article)

that's it, for the moment. --Elian
 * first, this page belongs on Meta, not on the English wikipedia.
 * second, Wikipedia is not only a social system but also a technocracy, meaning that the developers constitute a part of the executive. A special solution has to be found for decisions which involve programming effort. What if the community decides upon something which none of the developers wants to program? Should we force them? How is enforcement done?
 * third, a veto right should be installed - and also an instance to refer to if a decision is taken which contradicts one of the essential rules of wikipedia in the opinion of some people.
 * fourth, a body of unalterable things (a wikipedia constitution ;-)) should be defined which cannot be changed by vote (NPOV for example)
 * fifth, polls should be announced at announce-l, otherwise they are not valid.

Our goals
1) Our goal is to arrive at a final decision about the process, made by Jimbo. I'm not sure to which extent such a decision should/can be made for the non-English Wikipedias, but they can certainly adopt it.

2) This is a minor issue -- if nobody wants to implement the system we agree on, we can always pool money to pay someone to do it.

3,4,5) These belong as proposals in the article.

3) Except for Jimbo and possibly other people at Bomis, who should be able to veto what and why?

4) If we use voting, the "body of unalterable things" can still be alterable, but simply with a significantly stronger majority. I don't think unalterable rules are ever a good idea.

5) I have no problem with posting polls to announce-l, but there should also be a "Recent polls" list on Wikipedia itself for those who don't want to subscribe to the list. --Eloquence 16:19 Nov 15, 2002 (UTC)

That comment ''In other cases, a small minority might manage to get a particular change implemented for the simple reason that most people do not take notice of it and therefore do not decide to oppose it. The majority may then, eventually, be surprised by a change that nobody really wanted. ''...

...should not be seen as a particular drawback of consensus making process. It is related to poor information, not to this specific process. A vote which is not advertised, or with a very short voting time, can lead to making a decision which was not desired by the majority just as well. This issue can be solved by making clear what discussions are currently under way, and when they seem to be coming to a final consensual decision. This comment is "not" a valid argument against consensus ahma.


 * I have modified the wording slightly, but I will have to think about that in more detail. --Eloquence


 * Sure. But, then


 * The majority may then, eventually, be surprised by a change that nobody really wanted.  is no more valid, since the point is that people did not care rather than take notice. The fact they don't care to give their arguments does not imply they will be surprised by the change.

Why on Wikipedia and not on Meta==

Why in the world is this on the English Wikipedia and not Meta when this grew from a post from the general policy mailing list? Wouldn't such a process be used by the entire project or just en.wiki? --mav 21:09 Nov 15, 2002 (UTC)


 * Slow down, cowboy. The "general policy mailing list" was still the English Wikipedia mailing list 2 days ago and is probably still perceived by many of its subscribers as such. The reason I put this here is that I suggested to Jimbo that we work on ideas, and that he should eventually declare, as benevolent dictator, which decision making process we want to use. It is my understanding that he agreed to this proposal, he said he would also add his own arguments to the page.


 * The problem is that I do not know whether he wants to declare a binding process for all Wikipedias (and whether their respective maintainers would want that to happen), or just for the English one. For the time being, I consider this a project that is the result of the chaos on the English Wikipedia list in the last weeks, and unless Jimbo says otherwise, consider the results of our discussion primarily relevant to the EW, with adoption of our conclusions by the other Wikipedias being an option. --Eloquence 22:50 Nov 15, 2002 (UTC)

Polls
Perhaps polls on binding policy issues should be left open indefinately and people allowed to change their registered vote. This would allow newcomers an immediate voice while also providing long term flexibility to allow the community to drift appropriately with the prevailing weather.

The results of these polls should be summarized concisely and presented to newcomers and users as per Mav's and Axel's suggestions regarding the edit submittal form. Both a concise form and the long form could be required to initiate the poll so people know exactly what they are voting for and how it will be presented to the community and public at large.mirwin

In favor of democracy
I'm in favour of democracy with a meritocratic flavour. Voting power of a person could be weighed. A contributor who added/enhanced lots of articles and hopefully doing so has gained some wisdom about do's and dont's could have a larger say in matters of policy. Of course number of edits does not per se imply high standards and quality, but in most cases there will be a correlation, since most wikipedians whose work is not appreciated will improve their conduct or leave. So on average veterans will have a more balanced/mature view of what if good conduct and what is not. At the bare minimun I think that only registered users should be allowed to vote. This normal practice all over the web, and in society as a whole. If registering is too much trouble, tough luck. Erik Zachte 03:01 Nov 16, 2002 (UTC)


 * I agree with the "only registered users" part, everything else would be far too easy to abuse. You have to realize something, though: While most of us may like voting, Jimbo doesn't. So we should try to find arguments against the alternatives, and to refute the arguments against voting that can be made, otherwise we may be stuck with the ";decision by consensus" process for years to come. --Eloquence 03:27 Nov 16, 2002 (UTC)

What types of decisions?
What types of decisions would this be for? I think that dramatically affects what type of process we should use. We've pretty much already decided the process for deciding what's in most articles--the process we call wiki. For decisions about formatting pages, et cetera, we use straw polls on policy pages to create guidelines which can be ignored in favor of common sense and any attempt to have strict votes and strict enforcement would ruin the wiki(pedia) way. But for other decisions, like banning users, we need some other way. But why only one method for all types of decisions? Banning a user is different from deciding whether September 11, 2001 Terrorist Attack/Memorials belongs in the : namespace and I don't see why both of those should have the same process.DanKeshet

Middle way between consensus and majority vote
There is a middle way between consensus and simple majority vote. Surely consensus put to the extreme will mean no decision is taken at all, when one or two stubborn users disagree ad infinitum. (What about two vandals pleading for each other?).

In practice consensus means most of the time: a vast majority supports a certain decision. This might be formalised and rendered more objective by stating that say 3/4 of votes should be in favour of a certain solution to make it a decision, debate continues till such a clear majority arises. Of course a preset voting period should be observed, lest the first three voters determine the outcome right after voting has started. When more than two alternatives exist, voting might be done in two rounds, only best two of first round proceed to second round. Like any formal scheme this proposal contains loopholes and can be refined. Someone may vote under several accounts. Erik Zachte 14:45 Nov 16, 2002 (UTC)

In some consensus processes, there can be some people given specific roles to hold. A facilitator, a secretary (who take some time to report and summarize the outcome of a discussion - notice that reporting and summarizing is different from setting up a poll), a time guardian (in case there is a time issue when a pb must be solved rather quickly), a peace guardian (to avoid insults fusing or to help stop them), a moderator maybe (in case there is a troll in the group, to prevent people from disrupting discussion).

In some cases, setting up these roles (more or less officially) can help the process to run better, and avoid getting stuck. Of course, a moderator or a peace guardian should not be too much personnaly involved in the outcomes of the decision. Some of the latest discussions would have gained from having a peace guardian or simply a secretary (cf the "portal")

Designated voter
I think when voting we should have the option to have a Designated or representative voter for instance, let's say I didn't consider myself knowledgeable about chess. In such a case, rather than voting myself, I would designate RAM-MAN and whatever he voted would be worth twice as much. Lir 21:40 Nov 20, 2002 (UTC)


 * I agree, and in fact have already used this device. It's called a "proxy". I assigned my proxy in a vote that should interest you, it was the naming convention for countries on Wikipedia! If you're interested, I can show you how to set up a Wiki-election. --Ed Poor

As for this consensus notion, I find it is little different from good ol totalitarian tyranny. There are a few elites who declare that their viewpoint is the majority and they dismiss all dissent as lunatic crankery unworthy of a vote. Lir 21:51 Nov 20, 2002 (UTC)
 * be logical Lir. Consensus does not necessarily imply vote. So you can't say it is dismissing dissent as rant unworthy of a vote


 * Lir, it's called a proxy vote, and it is done a lot at the corporate level in the US.
 * I would personally not mind seeing wikipedia incorporate as a non-profit organization, like kuroshin.org, with real bylaws and real membership, legally binding in a court of law in the United Sates. I think a lot of what we argue about would just go away. Christopher Mahan

Classifying contributions
Just curious, how about automatically classifying contributions as approved and unapproved for major edits, distinguishing between the two based on whether or not the piece has reached a threshold of positive review. Presley H. Cannady

An argument in favor of the status quo
User "v" makes a compelling argument for not implementing a formal decision-making process at Wikipedia. Of the many options for a formal decision-making process that might be considered, one of the worst would be one that involves voting on proposed policies; tied for worst would be a decision-making process that is placed in the hands of an officer or a committee like the Wikiquette Committee recently appointed by Jimbo Wales.

Notwithstanding the recently appointed Wikiquette Committee, Wikipedia has very few official policies and no real systematic mechanisms of enforcement for the few official policies that it does have. Rather, individual Wikipedians propose, commment on, and enforce various policies as they see fit, and there are very few situations where this informal structure does not work exceptionally well. A noteworthy exception is the edit war phenomenon, but the best solution for the edit war phenomenon is a software fix, not a formalized decision-making process. Such a software fix would automatically protect articles that had been reverted three times, forcing a cool-down period of 48 hours, absent the intervention of a sysop. -- NetEsq 21:48, 18 Dec 2003 (UTC)


 * Actually, that would make the problem worse, as people would now try to circumvent the software by making edits during the edit war, which would then give them justification for continuing it ("I put additional work into this revision, you can't just revert it to yours" - "I did, too!"). The solution to edit wars is a firmly enforced reversion policy, not a Slashdot style "lameness filter" that will only frustrate them.


 * As for decision making, voting should always be an option if no consensus can be reached. But consensus should always be the primary goal.&mdash;Eloquence

''<< The solution to edit wars is a firmly enforced reversion policy. . . >>''

Any policy that needs to be "firmly enforced" would be anathema to Wikipedia. In any event, I really don't understand how automatic page protection could make matters worse when it comes to edit wars. Its twofold purpose would be to protect edit histories from unnecessary corruption and enforce automatic cool-down periods, and it would be subject to the intervention of sysops. -- NetEsq 22:40, 18 Dec 2003 (UTC)


 * NPOV is a policy that needs to be firmly enforced, and it is. Wikiquette is a policy that needs to be firmly enforced, and it isn't. We have banned quite a few users for violating either one persistently, though. To me, edit wars are a subset of Wikiquette.


 * Of course banning should only be the last resort. But it should be the last resort. The problem with software barriers is that they only incite users to circumvent them (trivial in this case as software can't distinguish between hostile and non-hostile alternating edit sequences -- as soon as you don't have an exact revert, the barrier fails), when the same users might tolerate a policy on the matter. As long as people can get away with edit wars, they will try to.&mdash;Eloquence

What are our policies?
One thing in particular that I find frustrating and troubling now is that even the policies that we do have are sometimes not clearly stated as policies. For example, Policies and guidelines is the main entry point into the Wikipedia policies. There you will find that "No personal attacks", "No offensive user names", and other items that seem to have the force of policy are listed under "Specific guidelines to consider", where you will also find "Ignore all rules" and "state the obvious".

I believe that the time has come to firmly state exactly what policies we have, and clearly separate them from "suggestions". We still want to keep the list of policies as short as we can, but no one is served by the confusing mish-mash that exists now.

I suppose what is required first is for someone (Jimbo?) or some group to define exactly how policies are set and reviewed. (I've heard that 80% is considered valid consensus for a policy, but where is that stated? How is the polling to be conducted?)

I'm not sure how the Arbitration Committee is proceeding where policies are vague, but it seems to me that their efforts would be only aided by any clarification of what our policies are.

-Rholton 02:56, 16 Mar 2004 (UTC)