31

The Close-Vote queue is large and growing. I searched for definitive information on how that queue is sorted for reviewers, and the nearest I could find is that it is randomized - but that post was some years old.

Would it be possible to sort on order of votes already received? That is, all questions with 4 votes to close are surfaced first (in random order) to reviewers? Questions with 3 votes would come up before ones with 2, etc. In that manner a given question gets out of 'will it close?' limbo faster.

This feature request would, in my opinion, be supportable if the majority of reviewer votes were going to single-vote questions. I cannot (I think) actually find these numbers. If this is the case, the majority of the work is going to an ever-increasing queue, and proportionally fewer votes are being resolved.

Note there are other distributions that could help get in this direction without being a strict ordering, like weighting the chance a given question is surfaced to a reviewer by the number of votes. However, for discussion purposes I thought I'd keep it simple.

Usual apology if this has actually been addressed. Search is sad-making.

  • 1
    I think sort order, and being able to filter by # close votes or # of answers, could be useful in conjunction with the existing filter and perhaps on the same interface. The idea is to make the time of the viewer more effective, which is often only a few minutes here or there. In the off topic category, I'd be willing to quickly go over the 0 answer questions. In the "based on opinion' category, I'd be looking for questions with too many answers (to close to stem the tide, since the good answers are probably already out). In either case, being able to sort/filter is useful. – Paul Aug 24 '13 at 2:03
18

The queue already works this way... Sorta: the more previous reviews a given task has had, the closer to the top of the queue it'll be. Note that this doesn't necessarily mean that questions with the most close votes will be given priority - "Leave Open" reviews count toward this too.

However, preference is also given to the most recently queued items, particularly when filtering by tag; depending on your preferences and time of day, this might end up giving you newer (rather than more-reviewed) items.

  • 7
    Is there a reason that the ordering of the Close Votes queue isn't more straightforward? I assume there is data behind the decision to make it work this way (because that's how you guys tend to roll), and I'd be interested in knowing the justification. – Josh Darnell Oct 8 '13 at 18:56
  • 1
    Would it be better to put task that are closer to be completed at the top, e.g. when there is a lot of agreement on “close / leave open” and they have been lots of reviews. Including the number of views a question has got in the last week may be useful, as more viewed bad questions are more important to close. – Ian Ringrose Feb 20 '14 at 10:42
  • @Ian - see: meta.stackexchange.com/questions/199879/… – Shog9 Feb 20 '14 at 17:21
  • Shog, consider sorting to additionally favor close votes / flags cast by tag badge holders in their tags, as proposed here – gnat Oct 10 '14 at 10:53
  • This is as clear as mud Shog.. it really gives us no true idea of how it is sorted. – Yvette Colomb Dec 17 '15 at 11:36
  • Considering the primary factor is # of existing reviews, the rest doesn't much matter; it's not like you can get a sorted list of pending reviews. – Shog9 Dec 17 '15 at 11:49
  • 1
    After more than 2 years, this answer seems to contradict the [status-completed] in the question. I mean, the queue is very clearly not sorted by votes. But it should be! Or at least, that's a viable request. I know that there are a lot of questions with many close votes that are routinely not being shown to me in review before my votes run out. Is there any rationale against an optional sort according to votes? It really does seem plausible that it would help in parsing the CVQ more quickly. And if "Leave open" doesn't dequeue a post, then there's no use showing fixed posts to everyone... – Andras Deak Apr 7 '16 at 14:31
  • 3
    At this point, the primary blocker for any tweaks or enhancements to /review is simply that it's 4 years old, no one who originally wrote it is able to work on it, and it's showing its age, @Andras. – Shog9 Apr 7 '16 at 14:36
  • Oh, that is quite unexpected:) I'll just bear with the situation then, until further notice. Thank you! – Andras Deak Apr 7 '16 at 14:42
  • Aren't you a software developer? This is pretty much par for the course, @Andras! Watch out for second-system syndrome... – Shog9 Apr 7 '16 at 14:54
  • Firstly, you got me: I'm a physicist, and not a developer in the usual sense:) Secondly, I always tend to falsely assume that SO has so much thought and effort put into it, that it's the epitome of a uniformly well-tended system. So far I've only expected oddities to arise from conceptual disagreements between SO staff and me, rather than "nobody will touch this part of the codebase". – Andras Deak Apr 7 '16 at 15:15
  • 1
    Not so much that no one will touch it as that its basic architecture was designed for a time when SO was getting a lot fewer posts per day, when the biggest challenge was getting anyone to even look at review, when the suggested edit queue routinely got backed up and blocked editors until it was cleared, when moderators still handled most of the close flags, etc. You can, in theory, retrofit fuel injection on a 1979 Camero... In practice, that's probably not something you're gonna try if it's still your daily transportation, @Andras. – Shog9 Apr 7 '16 at 15:21
  • recent complaint at MSO made me feel that this way to prioritize close queue may be somewhat troublesome for questions asked on weekends – gnat Mar 30 '18 at 22:18
2

I'd think that random sort order would give the benefit of the doubt to the person posing the question.

If the queue is sorted by number of close votes, highest first, then that would encourage a quicker closure overall. If the ordering is random, then it gives the asker more time (on average) to revise his/her question.

But the really poor questions would be closed quickly anyway, just from regulars reading the question in the feed.

  • This is another question I don't have data to answer, but; what is the average time from the first close vote to a question being closed out? Instinctively I doubt that questioners don't have time to fix questions, but I could be wrong. – Nathaniel Ford Aug 23 '13 at 22:58
  • 2
    It is not a race to fix the question before it gets placed on hold. The whole point is to place the question on hold so that it can be fixed. Once it is fixed, it goes into the reopen queue. (Oh great, another queue!) @nath – Cody Gray Aug 24 '13 at 6:34
  • @NathanielFord, the reopen queue works so well these days that it is not an issue. – Ian Ringrose Feb 20 '14 at 10:43

You must log in to answer this question.

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