4
$\begingroup$

Meta is used to decide site policy how should we use the results of questions? Is the highest voted answer immediately taken as a rule or is it a suggestion? If it is a rule then what consensus is needed to set a new rule? Would a vote of 1 on an answer be enough or does there have to be a large number of up votes? How large is that number?

$\endgroup$
8
$\begingroup$

It depends, and it's necessarily vague.

Suppose a site's community, via meta, voted overwhelmingly to, say, outlaw question closure. That would be interesting, but it wouldn't change anything. At the other end of the spectrum, if a site's community votes to change the name of its chat room, or to define a tag synonym, or to reopen a question, there's probably no reason not to do that.

A community's custodians, including its moderators, cannot have their hands tied by a meta consensus if there are reasons to not implement it. But, by the same token, it would be unwise to ignore that meta consensus -- something is driving it, and chances are there's an underlying issue that we need to address. So maybe it's not even about the specific proposal; perhaps that's a symptom of something else and we need to talk about that.

Sometimes meta is data not democracy. Other times it's democracy, and if there's clear consensus and no objection on a site-internal matter, then it'll probably happen when meta activity has died down (people have had a chance to object) and somebody gets around to doing whatever the thing is (merge those tags, update the sandbox rules, start the monthly challenges, whatever).

I can't give you a set of rules. I think our community has been pretty good at both reaching consensus and reaching consensus that we've reached consensus. If you see something that seems to be languishing but also seems to have consensus, bring it up somehow -- comment on the post, bring it up in chat, etc.

$\endgroup$
2
  • 3
    $\begingroup$ Ah, but have we ever reached consensus about reaching consensus on having reached consensus? $\endgroup$
    – Frostfyre
    Jun 19 '17 at 19:01
  • 2
    $\begingroup$ @Frostfyre I don't think we have reached consensus on the necessity of reaching consensus about reaching consensus on having reached consensus. I propose a Meta debate to clarify the need for such consensus, but because there is as of yet no consensus on such a need, maybe we need to discuss that latter first. $\endgroup$
    – user
    Jun 20 '17 at 4:14
1
$\begingroup$

As Monica already said, it depends. I've certainly seen it go both ways on Meta Stack Overflow and Meta Stack Exchange...

Some proposals are incredibly popular, yet never really get full support or implementation​, while others seem to make the grade and are eventually added as policy.

One workflow that seems to work well for a lot of issues and one that I support would look something like this:

  1. Ask a general discussion question to see if the community recognizes that this is an issue that needs addressing.

  2. If it looks like people generally agree that there's a problem ask a question about potential solutions.

  3. If there seems to be a well supported solution escalate to a or

  4. If the feature gets implemented or the added then you have solid confirmed consensus.

It's a little harder to gauge community consensus here than it is on some of the larger sites where you tend to see a little more Meta traffic... I think we should be careful to keep in mind that some of our higher voted posts here on Worldbuilding's meta only tend to see a handful of votes and may not always be fully representative of the larger community. Honestly I think we could do more to encourage people to get involved in the process, but even then a slow steady work flow should keeps things on track.

Perhaps adding to the featured on Meta list could draw some more eyes before we move forward with new FAQs?

$\endgroup$
3
  • 2
    $\begingroup$ I think #1 and #2 can be readily combined into one, with just some careful phrasing of the question. For example, instead of first asking "is X a problem" and then later asking "what should we do about X", consider asking both at the same time on a form like "how should we handle X". If you phrase it in a way that allows answers including "X is perfectly fine" or "X should be killed with fire and brimstone", then the community can express their feelings about the subject and propose ways to handle it at the same time. This also ensures that you don't get different people voting on the two. $\endgroup$
    – user
    Jun 20 '17 at 4:14
  • $\begingroup$ @MichaelKjörling I thought about that, but often people trip over that careful wording part. Ideally I would prefer to see the two combined. $\endgroup$
    – apaul
    Jun 20 '17 at 4:26
  • $\begingroup$ I don't see any problem with being explicit about allowing both, if that's preferred. $\endgroup$
    – user
    Jun 20 '17 at 4:36
1
$\begingroup$

What if the highest voted answer has 3 votes?

I think there should be some sort of minimum (overall) participation before a policy decision is made and perhaps some special tag for "imminent policy change" that we can monitor.

When there is such a policy change vote on the board, one of the answers could be "ABSTAIN", so we can at least see people have seen the imminent policy change and don't care -- thus knowing we have reached a sufficient level of exposure to conclude that a consensus has been reached.

$\endgroup$
4
  • 2
    $\begingroup$ The views count on the question can be useful there. A question where the highest voted answer is voted +3 and the question has ten views probably can't be taken for much; a question that has a views count in the thousands can probably be taken to mean something even if the votes are few and far between. Exactly what it means would obviously be very context-dependent. $\endgroup$
    – user
    Jun 27 '17 at 12:39
  • $\begingroup$ @MichaelKjörling It becomes a question of political philosophy: What does it mean we should do, if nobody really cares enough to vote? Is that equivalent to assent either way? Who knows. Close ties, likewise: Suppose irreconcilable answers are 999 to 1000? Mathematically, implementing either option is unstable; Majority opinion could change by 3% or more within a month (esp. if one is implemented and some outcomes unexpected). But we cannot poll and change every month. To be stable, change should require at least 60% with hundreds of voters; to reduce the likelihood of a majority reversal. $\endgroup$
    – Amadeus
    Jun 27 '17 at 13:05
  • 1
    $\begingroup$ "To be stable, change should require at least 60% with hundreds of voters; to reduce the likelihood of a majority reversal." I kind of doubt even Meta Stack Overflow has hundreds of active users, let alone what we have here. $\endgroup$
    – user
    Jun 27 '17 at 13:21
  • $\begingroup$ @MichaelKjörling I am unaware. If that is the case, we should have 30, or if that is still less than 5% of the pop, perhaps move to the Student's t-Distribution and increase the % of votes that constitutes a super-majority. In any case, I think the gist of my argument holds; that probabilistically speaking, a vote to make a change should be expected to remain stable for at least a year or so. Votes and views give us the probability of someone holding a categorical stance, and what is too evenly divided to be long-term stable, or if too many don't care enough yet to warrant a permanent change. $\endgroup$
    – Amadeus
    Jun 27 '17 at 13:52

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .