8

Possible Duplicate:
Steward Badge gaming visible in the Review Close Queue

I've noticed that most people who /Review a lot of close votes on ServerFault do about 12-20 reviews a day. That seems like a reasonable number, it means they've cast a few votes while reading questions, then the rest when /Reviewing.

But there are a few people who are regularly doing 30+ or 40+ reviews per day (I think the most I've seen is ~50). That means a lot of their reviews are "Do Not Close," and I think some of those are iffy, i.e. questions with a couple downvotes and 4 other people voted to close but a "Do Not Close" from one of the accounts I'm wondering about.

Obviously, some Do Not Close votes are to be expected - people have different opinions. Since Data.SE is still not updated, I don't think there's any way to get accurate stats, and although I think I see a trend when browsing through my list of reviewed questions, I know that's pretty subjective.

But it got me wondering if there's any problem with people egregiously voting "Do Not Close" and if the system does anything to track /Review behaviour. I'd think that in an extreme case, (someone who blindly picks "Do Not Close" for everything), it'd be good to exclude them from reviewing.

Improve this question
5
  • 1
    I don't know ServerFault but on SO I'm more afraid of people fast confirming duplicates without even looking at answers. I've seen so many mistakes (questions closed with good answers while the "duplicate" was years old and obsolete or had only terrible answers). I agree that an automated check of voting patterns (mostly about speed I guess), if it's not in place yet, would be useful.
    – Denys Séguret
    Commented Oct 9, 2012 at 7:04
  • 8
    I'd much rather have more people click "Do Not Close" than blindly click "Close" on everything they see. Ever since we got the new review queue, a bunch of users consequently closes almost every question within a short time. But, without stats, who knows. It might even be different for SF/SU/SO.
    – slhck
    Commented Oct 9, 2012 at 7:10
  • On SF, our "Top 20" reviewers includes people who've only reviewed 35 close votes, so we have at least an order of magnitude less activity than SO.
    – Ward - Trying Codidact
    Commented Oct 9, 2012 at 7:14
  • See also: meta.stackexchange.com/questions/148018/…
    – martin clayton
    Commented Oct 9, 2012 at 7:40
  • 2
    I am definitely seeing a lot of "just click the easiest thing to click" reviews. Upvotes, Do Not Close votes, etc.
    – Andrew Barber
    Commented Oct 9, 2012 at 8:25

2 Answers 2

Reset to default
4

The root of the issue here is that we have a large (for Server Fault) backlog in /review/close combined with:

  1. Not a lot of people doing reviews;
  2. Each reviewer has only a very small number of close votes;
  3. The majority of the posts in /review-beta/close should be closed.

It seems that the original assumption behind the review badge was that people would click "Not Sure" most of the time; that badge (which has been redesigned) read: "1000 reviews, 200 actioned." This is almost certainly true of SO, where most people will likely be familiar with only a minority of the content being discussed.

Your assumption seems to be that people will act on a review most of the time, whether it is to "Close" or "Do Not Close." On Server Fault, this is probably much more true than on other SE sites (especially SO) due to the topic boundaries in place at SF.

I have never seen someone on SF who chooses "Do Not Close" for everything, though I have seen a couple of users who vote "Do Not Close" a majority of the time, maybe 75%.

I don't really see a problem with too many "Do Not Close" votes. It's not like a review is complete after one or two people look at it; each one is going to get a lot more eyeballs.

The review system also seems to be putting a priority on new posts in this queue, which is helpful.

Finally, the fact that someone disagrees with you doesn't necessarily mean that they are wrong. :)

Improve this answer
1
  • @PopularDemand Check out the SF review queue if you can. We have, at most, a dozen active reviewers, so 5 Do Not Close votes is probably a majority of the users who bother doing review, and in that case, the question definitely should be removed from the close queue.
    – HopelessN00b
    Commented Oct 25, 2012 at 18:18
2

I'm seeing a lot of questionable close votes in the Server Fault queue lately, and I don't see anything wrong with reviewing accordingly. If anything, I'm starting to lament the fact that voting "Do Not Close" doesn't actually have an effect (like cancelling out a close vote, or something).

If anything, we should be more concerned about people "egregiously" voting to close than people only voting not to close in review. If anything, clicking through and not closing everything is no different than just not casting close votes (no effect), except for the extra effort and potential for badge-whoring. Frankly, closing good questions is much worse than someone maybe getting a review-related badge that they "don't deserve," which is the only consequence I can see of blindly clicking "Do Not Close" a whole lot in review.

Improve this answer

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