One of the things left to determine is how many votes it should take to close.
Thoughts?
The following should all be configurable per-instance (which becomes default per-community in that instance) and per-community, eliminating any need to decide now:
We need to decide early whether any particular action will even be possible - e.g., whether there should be such a thing as “vote to close” - but not any of the numeric values.
We have not, for example, decided whether to have a dupe-hammer. Maybe we want a different approach to expedited close/reopen, or maybe we don’t want it to exist.
No we don’t need to decide on the defaults right now, but on the other hand, it can be helpful to put a stake in the ground as we talk about various parts of the system. I think a default of 3 votes in either direction makes sense.
I think it would be helpful to have default number, and the first thing we probably setup will be replacement for this forum (to start dogfooding) and it would be good to have the numbers decided.
A Stack Exchange (SE) annoyance was diamond moderators being unable to cast ordinary close votes. SE refused to implement ordinary close votes, going against the community and diamond-moderator requests. Maybe this can be addressed?
Personally, I think 3 is too few—5 feels more like a community effort. But this very much depends on the site.
I would say it not only depends on the site but also the tagging (assuming there’s tagging similar to what’s used on SO). Tags with a lot of traffic could stand to have a higher barrier than those with low traffic.
This could, perhaps, be less of a problem if the review queue size stays low. But on re-open posts it’s definitely a factor. It’s much easier to decide a question is missing something than to be able to judge whether it now has enough information to be answerable in a specific technology.
One may even need to consider a different number of votes to re-open than to close for low-traffic tags. Or an analog to the “dupe hammer” for re-opening in such tags.
3 feels like the right number for any communities smaller than SO. And I haven’t been a frequent SO user, only SE user, so I can’t say if SO scale would benefit from an increased number of required votes.
Since we’re not planning to start with a 100’000 daily user visits on day one, I think we should be good with 3 for a while.
3 is good enough for starters. This will be configurable per community anyway.
This is yet another value that should at least somewhat track site activity. For example, 5 close votes may be just fine for a site that gets 200 question-views per day, but way too high a bar for one that gets 15 question-views per day.
Perhaps we could use a scaling system? Use 3 for default, 4 if the site has X avg views per day, 5 if Y or higher.
I would suggest though not making it automatic. Start a new site out low (e.g., 3) and revisit the settings over time as sites grow or shrink. There are too many variables to make this automatic in any practical way.
Plus it’d be confusing for users. “Wait, why did my third close vote work today when it didn’t last week? Is this a Christmas special?”
It wouldn’t be daily fluctuations anyway but presumably some sort of rolling average over weeks or months, but the point remains – there will be some point in time when the threshold changes and doing it silently is confusing. Even if we generate an auto-notification, I’d still prefer the decision be in the hands of users informed by data. The system can call attention to the changed conditions without acting.
Here is an idea. Make closing and reopening both a routine action that can be triggered with one vote (like protecting questions on SE).
In the background, the person who closes and person who reopens are both silently logged as a vote to close and a vote to reopen.
After a question is reopened, further votes to close and reopen are tallied against each other. Majority vote wins. Ties lean in favor of opening (so that reopening, like closing, can also be done with one vote). Let people log silent votes for opening or closing that maintain the status quo, for use in the tally. Make it easy and transparent to transfer vote from one side to another.
I don’t think that’s such a good idea. I can think of at least two users off the top of my head who would vote to reopen any closed questions simply on religious grounds. The total required votes has to be high enough to get closer to the average will of the users.
There is also the problem that new users don’t understand question closing. I know I didn’t when I was new. I was used to forums, and the whole idea of closing questions and them being off topic due to asking where something could be bought was foreign to me. It didn’t seem right. Eventually (6 months to a year) I understood and saw the logic behind it. I didn’t go around trying to subvert the system because I recognized I was new, but unfortunately a lot of people don’t do that. They barge in and flail against anything that appears wrong, without any respect to those who are already there and the norms and traditions that have been established, often via hard lessons.
I agree with @Olin.
Giving only one vote to open/close assumes a lot of trust. Some people, as you have mentioned, will open/close questions based solely off their own believes, not the will or opinion of the community. This will also enable people to open/close questions for “fun”, causing unrest because they can. The amount of votes required should reflect the opinion of the users, not just one judge, jury, and executioner.
Suggestion:
Using as example values ratio 2:1 and minimum 3:
This would eliminate one pain point especially, which is drive-by-duplicate-voters. Happened to me a couple of times that people obviously only read the title and voted duplicate - question was closed, and comments certainly didn’t do jack about reopening. Especially if it’s not a grand question that gets hundreds of views.
One issue with that kind of system is that controversial posts would be kept open for longer, and as they’re controversial posts they’re even more likely than normal to get bad answers. Controversy should be settled by discussions on Meta, not ever increasing vote targets. I could maybe see it working for reopening only, but I’d guess that most users would find that confusing as well as kind of ugly philosophically.
Frankly, that was your fault then. If the title and the first sentence sound like a dup, then I can’t blame anyone for closing it as one. It’s your job to make it look obviously like not a dup. Remember that people reading and reviewing questions are volunteers too. First impressions are sometimes all you get.
This is a good point. If there is much controversy, the question is borderline, and we should err on the side of closing. Good questions are clearly so, with even a single close vote being rare on a good question. If you don’t like your question getting closed, don’t try to skirt so close to the edge. Stay out of the gray area. Stick to the solidly white area.
I did that, as much as I could - but the subject matter was just similar. The answer gotten somewhere else, however, was not.
I doubt that volunteers who only read half the title before hammering the post are still adding value… but that’s not really the subject of this thread.
Using a fixed number of votes to close/reopen still suffers from this to some extent, since a controversial question will naturally tend to suffer close/reopen wars, thus allowing bad answers in at intervals. Fortunately, there’s at least some delay between cycles that reduces question visibility at least a little.
Starting off with, say, 3 votes to close/reopen, and raising that threshold every cycle, would both directly slow this down and serve as an immediately visible warning of the question’s controversial status.