What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 122 Stack Exchange communities.

The more I see how the new review system works in practice, the less I like it.

StackOverflow is not just a single community, but many semi-independent communities roughly organized by tags. Users have favorite tags, and ignore others. Nobody is a member of all communities equally.

However, the review system doesn't respect this, as far as I can tell. It encourages and enables people to review questions and answers outside of the communities in which they are active, respected and knowledgeable. Is that a good thing? Keep in mind that large communities can easily swamp small ones through sheer numbers.

As an example, here are four recently closed Haskell questions:

Other than myself, I couldn't find a single of the close-voters who had any Haskell points.

The problem isn't that the questions should or should not have been closed; it's that the issue was decided 'for' the Haskell community by outsiders. Although I personally felt the questions should have been closed, I'm used to the old system in which it took 5 votes to close a question; I almost feel as though I single-handedly closed the questions by casting the first vote, which put the posts into the review queue and attracted the attention of users who would otherwise probably never have come across these questions. My soul is filled with regret ( boo-hoo-hoo :( ), because (except for the last one) I cast the first close vote.

I have a couple of ideas for addressing this:

  1. get rid of the review queue. If there's no review queue, users won't have any reason to look at posts. Meanwhile, people who are knowledgeable in their favorite tags are the ones who will be active there, closing/reopening/editing posts.

  2. segregate review tasks by tag, similar to how tag wiki voting is done.

  3. get rid of parts of the review queue. Or restrict access to parts of it. Perhaps higher rep limits, no close vote review, etc.

Are there any other thoughts on this? Anybody else frustrated by this?

share|improve this question
8  
I voted to close two of those. I can assure you I read them carefully and considered whether they should be closed before I did. I do not believe that in the cases I voted on, knowledge of Haskell was required. –  Andrew's a Unitato Dec 17 '12 at 17:30
    
@AndrewBarber I apologize if it seems like I'm calling you guys out; really, I'm not. As I stated in the question, The problem isn't that the questions should or should not have been closed; it's that the issue was decided 'for' the Haskell community by outsiders. –  Matt Fenwick Dec 17 '12 at 17:33
8  
I understand. My point is that I don't think being 'outsiders' is relevant here. –  Andrew's a Unitato Dec 17 '12 at 17:35
    
I guess the take-away message for me is "don't use close votes anymore". –  Matt Fenwick Dec 17 '12 at 17:50
    
I'd like to second what @AndrewBarber said, that is I too reviewed two of those posts and considered whether they should be closed or not. I do encounter every so often posts where I think I don't have sufficient domain knowledge to review them and in those cases I skip them. –  Jack Dec 17 '12 at 21:56
add comment

1 Answer

up vote 9 down vote accepted

It encourages and enables people to review questions and answers outside of the communities in which they are active, respected and knowledgeable. Is that a good thing?

In my opinion, yes, it is. It is specifically one of the reasons the review queues were created in the first place.

In many of the small communities you refer to, it's often difficult to get enough traffic to close questions, despite the fact that (barring certain types of exact duplicate questions) little to no knowledge of the topic is required to know if it should be closed.

Suggested edits likewise rarely require knowledge of the content to be reviewed, but having lots of suggested edits sitting around for long periods of time waiting for a reviewer from that sub-community to see it would be harmful.

As for first/last posts, they were specifically created because those posts often benefit substantially from some attention from more experienced SO users (sometimes from within the same tag communities, sometimes not) to help newer posters get used to the rules/guidelines of SO. For these queues more can be done (especially with answers, but also for questions) when you know the subject material, which is why the queue will, to the best of it's ability, prefer giving you questions in tags your active in. There are still beneficial activities that can be done in these queues without knowledge of the content, but it's admittedly less.

share|improve this answer
    
Although I disagree that it's helpful, at least now I know that it's the intended effect, and not merely a side-effect. Thanks. –  Matt Fenwick Dec 17 '12 at 17:49
    
For a bit of context: The haskell tag is an unusually active "small community"--there are more than enough people willing and able to use close votes who check it daily-ish. There are other niche tags with similarly engaged communities, but they're certainly not the norm. –  McCannot Dec 17 '12 at 17:52
add comment

You must log in to answer this question.

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