47

In direct contrast to this post:

Close Votes review: I'm going on a strike!

I'd like to say: I'm not going on strike!

... and I'd like you to do the same.

But... why?

Yes, I am frustrated with the SE staff (seemingly) not doing anything about the queue size.

But simply stopping reviewing doesn't help all that much.

There's a whole bunch of questions that needs to be closed, and someone needs to close it.

Just about everyone's demotivated by the queue size, so it's very probable that there's a whole lot less reviewing happening than will happen if the queue is almost empty (which is somewhat ironic) (for one thing, we have a few people on record saying they're going on strike for this very reason, and how many more simply stopped without saying anything). So perhaps simply clearing the queue once will relieve the problem forever (or, well, the symptom).

I'd like to take a metaphor from one of the answers to the above post, just with a more logical point:

Say there are 2000 people and 1600 miles of beach needs to be cleaned.

It does sound like a lot, but each person doesn't need to do all that much work, and they just need to see that they don't.

The problem comes in when 1800 people give up, and there are 200 people left to clean the entire 1600 miles of beach. Now it's no longer a seemingly daunting task, but rather a basically impossible one.

So, to the 1800 people who gave up: Come back, just keep your head down and clean a little part of the beach. When you look up, you may find the whole beach clean.

What you need to do

Until the end of the year (31 December 2013), commit to reviewing items from the close vote queue.

Either a full 40 reviews a day, or simply however much you're willing to commit to.

If you weren't reviewing before, perhaps now's the time to start.

If you're on strike, just get back to reviewing for just over a month.

If we see that we can make a dent in the queue, this in itself should provide a bit of motivation by showing that it's actually possible to clear the queue.

Progress report

The queue is, at the time of writing this (23 November 2013), at 94,443.

6 January 2014: Ok, the queue has been steadily increasing (as far as I've seen) since this has been posted, apart from a drop caused by a change in when votes expire (currently at 100,943). But we can just attribute this to not enough people taking part. Well, that was my attempt... (though you can feel free to keep your head down until ... whenever)

Improve this question
17
  • 3
    And if you're into the collective, organized effort, there's a chat room ( related discussion ).
    – Bernhard Barker
    Commented Nov 23, 2013 at 13:15
  • 12
    I am doing my best to get to 3k rep so that I too can join the task force! :)
    – shad0w_wa1k3r
    Commented Nov 23, 2013 at 13:16
  • 6
    "I am frustrated with the SE staff not doing anything about the queue size" ... that you know of.
    – Bart
    Commented Nov 23, 2013 at 13:47
  • @Bart Indeed, added "seemingly" for correctness.
    – Bernhard Barker
    Commented Nov 23, 2013 at 14:40
  • 16
    We are doing something about it, it's just a very hard problem to solve. Shog is going to be answering a bunch of the suggestions we've received while working on this next week - in short, we may have an interim fix. Remember, vetting possible changes involves a lot of data crunching and that's what we've been doing, this has gotten the full attention of a few of us, sorry that we didn't make that more obvious.
    – user50049
    Commented Nov 23, 2013 at 15:28
  • 1
    @TimPost This post is basically for a summary of all staff feedback, right? (instead of us having to keep track of the feedback on the dozens of posts about this) (I keep hearing it's a (very) hard problem, but that just makes me think you're overcomplicating it)
    – Bernhard Barker
    Commented Nov 23, 2013 at 15:36
  • 1
    FWIW, it's at 96,894 just four days later. That's about 600+/day. Meaning it needs an additional 15 reviewers capping per day to break even, assuming that's a relatively stable number. Not criticizing, just giving some numbers to go on.
    – Geobits
    Commented Nov 27, 2013 at 15:32
  • @Geobits Yeah, it's not looking all that well. My guess is the 42 upvoters are some combination of users with < 3k reputation, hypocrites or those too busy. Well, that's my attempt, I guess it's all up to the SE staff now.
    – Bernhard Barker
    Commented Nov 27, 2013 at 18:22
  • I upvoted this originally but lost interest. It's hard to care about the CV queue. There are loads of bad questions around but the only effect of closing them is to give negative reinforcement to the asker, which can be done more simply by downvoting the question. Duplicates or migrations are worth sorting out, the rest isn't. If anything, the enormous size of the queue means not that there aren't enough reviewers, but that there are unnecessary close reasons, or people are overusing them, or the site UI is failing to explain to new users how to ask useful questions in the first place.
    – Boann
    Commented Dec 2, 2013 at 23:42
  • @Boann The problem with simply a downvote is that 2.5 upvotes cancel out 1 downvote, and many of these questions are plenty upvoted. And we want them gone, we don't want them sitting around like bad apples, infecting the place. Well, we don't really want them to get answers in the first place. The only way to achieve either of that is to close them quickly, which can't really be done consistently while there's 100k questions in the close vote queue. The enormous queue size could've been caused by one or more of multiple factors, but your last point definitely has a lot of merit.
    – Bernhard Barker
    Commented Dec 3, 2013 at 0:01
  • For the first few days of Winterbash 2013 the CV queue was getting smaller, but that trend seems to have reversed itself. I guess a bunch of people got the hats and then quit... What can you do?
    – dmckee --- ex-moderator kitten
    Commented Dec 26, 2013 at 20:23
  • "Will update occasionally until the end of the year, and at the start of the new year."
    – Boann
    Commented Jan 6, 2014 at 8:19
  • @Boann Updated. I didn't really want to occasionally update it with bad news.
    – Bernhard Barker
    Commented Jan 6, 2014 at 12:40
  • 2
    "Either a full 40 reviews a day, or simply however much you're willing to commit to" — that is the problem right there. I for one am willing to commit to more than 40 reviews a day, but the system won't let me. I did 40 reviews a day for months on end. You think I wouldn't have done 41? 50? 80? Think again. I have tens of thousands of reviews and edits across the entire network. Why am I forced to take a break while the queue is not empty and in fact is still filling up? I gave up reviewing on SO because of that and that alone.
    – ЯegDwight
    Commented Jan 6, 2014 at 15:04
  • Seems it had a little impact at least. :) The queue was going down a little bit and back up and down again ... Never seen that before. Only growing so far until a few days ago.
    – Devolus
    Commented Jan 6, 2014 at 15:29

5 Answers 5

Reset to default
17

Say what you will, but I doubt that it was a coincidence that there was a response from a community manager the day after the strike was proposed...

I guess I can see both points of view. Going on strike may not solve the problem, but it did seem to shake things up enough to get the management to respond.

All in all it seemed to do what strikes are intended to do.

Improve this answer
4
  • 1
    It may just have been the last straw, or one of the last straws (there has been dozens of discussions and suggestions on the topic). Side note - the need for strikes usually points to a broken system.
    – Bernhard Barker
    Commented Nov 23, 2013 at 18:24
  • 1
    To be fair, Shog has been banging the drum about the close queue for as long as I've been working as a CM. He and a few others on the team have been working on a number of solutions to the problem. I'd say the "strike" was convenient excuse to focus on our close review deficit.
    – Jon Ericson Staff
    Commented Nov 24, 2013 at 7:23
  • 1
    Strike, "...a work stoppage caused by the mass refusal of employees to work. A strike usually takes place in response to employee grievances... Strikes are sometimes used to pressure governments to change policies... A notable example is the 1980 Gdańsk Shipyard strike led by Lech Wałęsa. This strike was significant in the long campaign of civil resistance for political change in Poland, and was an important mobilized effort that contributed to the fall of the Iron Curtain and the end of communist party rule in eastern Europe."
    – gnat
    Commented Nov 24, 2013 at 11:52
  • 1
    As always, the more... enthusiasm for something on Meta, the more it becomes clear that it merits at least an in-depth review. BUT, before you think I'm endorsing the effectiveness of walkouts, don't forget the Z-factor: The strike timing, like our internal increase in urgency, were both likely affected by our dismay at the arbitrary but painful notion of the queue hitting 100k. Stay tuned, we're on it. For real.
    – Jaydles Staff
    Commented Dec 5, 2013 at 20:36
14

I don't know how many others view the situation this way, but I don't see myself as having just "given up". I've grown to be opposed to the current SE software and policies as it relates to close-worthy questions. With all due respect to the well-meaning SE staff, I consider the software/policies to be extremely wasteful of people's time. I've reluctantly concluded that the growing queue size is actually beneficial at this point because I think it will help force a change in the software/policies for the better that will ultimately get us to a cleaner beach, faster and with less total effort.

The situation we're in right now is that of a factory with huge, growing piles of work-in-process inventory. The reason we have this inventory is because our manufacturing process has uneven flow, with the associated bottlenecks and delays. We can work overtime in the hopes that we will get the inventory down and it's possible that the inventory itself was the primary bottleneck, but that's highly speculative on both counts. In any event, by temporarily clearing/hiding the queue, we can see and focus on handling the incoming questions in timely/effective manner as discussed in https://meta.stackexchange.com/questions/208311/regarding-the-stack-overflow-close-review-queue/208635#208635. Another way of looking at this is that if we get our cycle time down, our WIP will stay down and our throughput will go up. That's Little's law:

The long-term average number of customers in a stable system L is equal to the long-term average effective arrival rate, λ, multiplied by the (Palm‑)average time a customer spends in the system, W; or expressed algebraically: L = λW.

Although it looks intuitively reasonable, it is quite a remarkable result, as the relationship is "not influenced by the arrival process distribution, the service distribution, the service order, or practically anything else."

As for the pizza box answer, that's a valid concern. But note that we have to be committed as an organization to absolutely minimizing the pizza boxes in the living room. Without that, we will fail under any scenario. And if we succeed in that, we can readily deal with the pizza boxes moved to the other room. Fortunately, in our case, there is no physical decomposition taking placed with the "hidden" boxes/questions.

Improve this answer
5
  • If I understand your answer correctly, you want less questions to be closed. In the current state of affairs, how would that correlate with a cleaner beach? I'm afraid the exact opposite would happen.
    – Frédéric Hamidi
    Commented Nov 23, 2013 at 16:43
  • We don't know whether there's something wrong with the system (software / policies), because it's intended to work with an empty / near-empty queue.
    – Bernhard Barker
    Commented Nov 23, 2013 at 16:45
  • @Dukeling Assuming the software/policies were introduced with a non-empty queue, that's a fair point and would justify changing things to allow that to be tested. I don't think the current process for getting to an empty queue is going to get us there, or at least not at reasonable cost.
    – Peter Alfvin
    Commented Nov 23, 2013 at 16:48
  • 4
    @FrédéricHamidi No, I want less of these questions to be created in the first place, more of them to be improved and more of them to be closed. I also think there are a lot more that haven't even been marked for closing that should be closed. It's just that I don't think our current software and policies will achieve that.
    – Peter Alfvin
    Commented Nov 23, 2013 at 16:55
  • 4
    With that latter clarification, your answer doesn't sound all that unreasonable. Perhaps it's a good idea to include that information into your answer and emphasize it. It might stop some of the downvotes you seem to be getting.
    – Bart
    Commented Nov 23, 2013 at 17:17
6

One thing that helped me to make this task less daunting was to set the filter to only one tag. I set mine to and I'm slowly moving ahead. Because I see questions for only one tag, I can see my progress: few days ago I was doing questions from about 6-7 months ago, now they are from 1 year ago. The whole queue is growing but my personal queue is shrinking.

That gives me motivation to continue. I hope one day I will go through all questions and then I'll switch to another tag.

Improve this answer
7
  • 1
    The sql-server tag might be shrinking, but that probably won't happen in the javascript or php tags.
    – John Dvorak
    Commented Nov 24, 2013 at 7:25
  • 1
    I'll try android next, that one will be probably harder too. Anyway, that's just one way of making the task look easier...
    – Szymon
    Commented Nov 24, 2013 at 7:27
  • Another way to look at this is that you'll soon run out of things you can easily review yet the queue as a whole won't be affected the least bit. (sorry for pessimism; I'm playing a devil's advocate here)
    – John Dvorak
    Commented Nov 24, 2013 at 7:31
  • 4
    Butt the queue will be affected @Jan. Theoretically if 5 people do exactly what Szymon has done then all SQL Server questions get "reviewed" instantly. You can then spread this out to other tags.
    – ben is uǝq backwards
    Commented Nov 24, 2013 at 7:36
  • I've reviewed all that I can on primarily opinion based in the JavaScript tags, but not sure how long that's going to last...
    – Qantas 94 Heavy
    Commented Nov 24, 2013 at 12:37
  • I didn't even know that you can use tags in the review as well. Today my 40 are exhausted, but tomorrow 'l check that.
    – Devolus
    Commented Jan 6, 2014 at 15:32
  • @JanDvorak, as some php quesions will also be taged sql-server it will help a bit with the php quesions.
    – Ian Ringrose
    Commented Feb 4, 2014 at 10:51
2

I was on a strike, but recently I decided to get back to working with Close Votes Queue. I left out duplicates using the filter, because I find all other types easy to diagnose most of the time. However I'm again at the precipice of a close-vote-mental-breakdown. Despite using all 50 votes a day, the queue systematically grows 1k every few days.

I'm not giving up, but plz halp! Moar people need to join reviewing!

Improve this answer
1
  • If you don't want to feel alone when reviewing drop by in so close vote reviewers chat, no obligations, just for a heads up.
    – rene Mod
    Commented Jan 9, 2014 at 15:08
2

Surely review queue is a dirty job, and the interface for review is very poor, giving very limited possibility to interact with question while reviewing. Controversial, badly chosen edits doesn't make that job grateful either.

However, there's something worse. If you only review queue, you start to see only that, what is worse on StackOverflow. Every question is a potential candidate for close or downvote. Every asker a help vampire. It makes you burn out. I've also noticed that constant reviewing makes me more aggresive when it comes to downvoting and closing, and we should, anyway, try to be nice to new users.

Therefore I'm reviewing the queue, but I'm also taking vacations from it. For healthy balance I need to cast more upvotes than downvotes and closevotes.

Improve this answer

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