2

I'm aware "recycled Questions" appearing from Community are the system's attempt to cull hopeless questions from the stack. And I've read the "formula" for culling unanswered questions.

As one single example of how ineffective this formula is, I submit this recently recycled question. Screenshotted here for your convenience:

example of cluelessness

I can appreciate the difficulty of algorithmically classifying a question as hopeless and unanswerable. I suppose that difficulty occurred to whoever (@JeffAtwood ?) developed the "formula". But this one "recycled Question" seems to bear witness to the inadequacy of this formula.

It's too bad the "formula" can't be loosened a bit:

  • As I read through the "formula", I'm struck by how unlikely it is that those AND'd conditions will occur in practice. In this case, the Question got a vague (IMHO) answer, which may mean that the question will never be deleted?!?!

  • On the same day the Question was posted, the OP made a couple of comments to the answer, but since then (now more than 18 months ago) the OP never followed up, never up-voted the answer.

  • No one else has up- or down-voted the Question or the Answer... until now, when I just down-voted it.

  • It's now been over 9 months since the OP has visited (logged in?) to this site:

user history

OK, long-winded rant, so let me finally get to my Question: What, if anything, can we here in RPI do to move questions like this toward the "purge" bin?

| |
2

I'm with you on the dubious value of auto-dredged-up questions. If I owned the universe, that whole front page queue would be gone. That includes the queuing of things just because they were recently edited (which the difference between the front page and the "Questions" new stuff only page is, AFAICT, the presence of recently edited + Community dredged questions).

What I'd replace it with would be a new queue proper (as, like the review ones where it takes five votes to close a question) for things that are potentially to be injected into the front page. That could auto-include things that have been recently edited and Community dredged stuff, but there would also an option (with a rep bar, like close and delete voting) for users to nominate an older question for injection.1 This way nothing except new questions would make it to the front page without being looked at and approved by a person (usually at least several people).

What, if anything, can we here in RPI do to move questions like this toward the "purge" bin?

The obvious things are vote to close and vote to delete. Note that (the details here are from memory so may be flawed) closed questions with no upvoted or accepted answer, and (?) no positive score on the question itself, are reaped regularly by the system, meaning it goes from closed to deleted. So in many context close voting initiates that process.

I don't actually look at the close queue for reasons I've explained here before, namely that I'd rather see something go down with 5 community votes than 1 mod hammer. There is a bit of an issue here in that we don't have a large number of regulars with the rep to close vote.2

I do close things all the time, but again, I don't even look at the close queue. I may start to do that with an eye out for this stuff.

There's also flagging. If you flag something "in need of moderator intervention" and indicate you think it is driftwood, if I agree I will close or delete it. You can flag closed questions too, in which case the suggestion would be to delete it completely (which as mentioned above the system often does with old, closed, pointless junk).


  1. Which might be redundant if all it takes is an edit, but never mind the nitty-gritty.

  2. Right now there's 50 3000+ members (after 7+ years). I doubt more than a fraction of those look at the review queues regularly. This is a corollary to the community's stinginess with upvoting, which also helps to create scenarios where we have perfectly good, even exceptionally presented, questions that have the same rough metrics as total crap: 0, which perhaps shadows the problem from question here...

| |
  • As I've taken more notice of the posts Community dredges up, I think I am seeing that they are generally ignored by many of the more frequent answerers here. Got any stats on answer rates for these vs. "fresh" questions? – Seamus Feb 29 at 23:27
  • There is an online, publicly accessible interface to a (cached weekly) version of the SQL database which backs the network, although if you don't know SQL it is either a learning opportunity or pretty much useless. – goldilocks Mar 1 at 15:20
  • Okay, so there is definitely a way to do this. There's a record of the events that you can find in a post's "timeline": data.stackexchange.com/raspberrypi/query/edit/1204158 (you have to hit "Run Query" at the bottom). They include CommunityBump, presumably the "bumped" by Community event (eg.), You'd need this, and the timeline of votes (there is a Vote table as well with post Id and a timestamp). – goldilocks Mar 1 at 15:27
  • However, my SQL is too rudimentary to cook that one up easily; it would take some work (I am positive it is possible). There are a lot of ready made queries on the Data Explorer site (I believe it saves them all much like exchange questions, so a lot); if you click the "Queries" tab you can search them but "votes community bump" turned up nothing. – goldilocks Mar 1 at 15:34
  • Thanks for the information, and it's interesting that all of this is available. But at least for now, I'm not prepared to divert any energy to SQL; my query will remain a point of speculation for a while longer :) – Seamus Mar 1 at 19:14

You must log in to answer this question.

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