What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 149 Stack Exchange communities.

At smaller sites (up to 50-100 questions a day average) closing a question takes hours or even days.

  • As an example, this SO question has got 5 close votes in 15 minutes, while its twin at Programmers has been struggling to get 3rd vote for over 10 hours. "Twin" here means, same question has been cross-posted at both sites, is equally inappropriate at both (tool recommendation), has comments explaining the issue for readers and is downvoted rather heavily, to -4 / -6.

Given above, suggest to take into account when there is a solid evidence that question may turn out inappropriate for the site and exercise more caution about promoting such questions in the hot list.

Penalize smaller sites questions having 3-4 close votes - say, cut their hotness score by half or drop them off the list completely. This will increase chances for less troublesome questions to enter hot list.

Indiscriminately advertising questions that are on their way to closure, with only 1-2 votes left to go hurts site community morale and dilutes the Stack Exchange brand:

watching the 'hot network questions' leads me to think that several of the recent sites are gigantic broken windows of self-absorbed subjectivity... The questions are whiney, the answers are opinion...

Consider also disappointment and frustration of new users coming to sites to ask questions similar to ones advertised in the hot list only to discover that these are voted down and closed as inappropriate (broken windows effect).


Related:

  • Avoid tweeting bad questions

    we won't tweet questions with close votes

  • Reorder questions picked for hot list...

    Close votes on the question should reduce the total hotness of the question by 20% (so a closed question would have a 0 hotness score)

  • Better criteria for the hot questions list

    Weight Flags/Close Votes Against Questions... 30% of our hot questions over the past month all got closed, which is less than ideal, because we want the attention to go to good polished questions rather than works in progress... If close votes count against the hotness score, then we can vote to close while trying to improve the post, and not worry about it getting a dozen answers in 24 hours as the entire SE network flocks in.

share|improve this question
5  
"while its twin at Programmers has been struggling to get 3rd vote for over 10 hours" Not anymore... ;) –  Yannis Jul 31 '14 at 19:20
1  
@Yannis I hate you mods for breaking my nice experiments. :) If seriously, thank you! –  gnat Jul 31 '14 at 19:22
    
for a recent example: question with 4 votes down and close at #7 in the hot list, gaining lemming upvotes from all around SE network, fully exploiting The Trouble With Popularity. How does it feel like for community regulars who voted down and close? –  gnat Aug 1 '14 at 22:10
    
...another recent example: 3 votes to close, 2 down, #19 (of 100) in the hot list with "32.619 hotness points". "...watching the 'hot network questions' leads me to think that several of the recent sites are gigantic broken windows of self-absorbed subjectivity" -- sure why would it be different when system works like that –  gnat Aug 1 '14 at 22:57
    
I like the idea, but the question might be asked: "Since we are talking about smaller sites, why isn't a pro-tem mod closing the question if the community already has 3-4 close votes on it?" In other words, the problem might not be so bad... assuming the pro-tem mods are reasonable judges of the question's worth –  Mike Pennington Aug 5 '14 at 10:00
8  
@MikePennington this would essentially mean moderators have to carry a 24x7 watch on questions that may suddenly make it into hot list, along with getting enough close votes, this is not how system is expected to work. I am also uncomfortable with the idea of moderators intervening depending not on site community feedback (flags, chat, meta) but on not particularly relevant cross-network feature (hotness formula) –  gnat Aug 5 '14 at 10:30
    
There is already moderation. I think community must learn and grow on it's mistakes, what here is suggested is authoritarian approach where few people knows what's best and can shape new community in diligent little SE, will certainly save purity of site, but as a result, even with growing popularity you will not see new people becoming active contributors. –  nes Aug 9 '14 at 6:20
1  
@nes observations show that hot questions don't really help site community grow, see: Can we track the positive effects of a popular question? It's more like "flash in the pan". Longer term observations at Programmers show pretty similar effect: "Tons and tons of people visited the site... but very few decided to stick around..." –  gnat Aug 9 '14 at 7:35
1  
...regarding community growth, it is also worth reminding that hot questions are simply not intended to help in that. Their only purpose officially stated so far is to show entertaining / interesting stuff, nothing else –  gnat Aug 9 '14 at 12:40
    
Why does it matter if a site is small? Bad questions are bad questions! –  bjb568 Aug 26 '14 at 13:37
3  
@bjb568 per my observations it matters indeed (check the example with "SO-Programmers twin"). I definitely won't mind if larger sites questions are penalized that way, I only don't expect this to make much difference. Think of it, inappropriate SO question at sidebar attracts eyeballs of, say, 100-200 users with close privileges at SO - how long would you expect it to hang in there before getting closed and wiped out? –  gnat Aug 26 '14 at 13:44
    
Does related #2 mean this is effectively implemented? 3-4 close votes drops the hotness by 60-80%, far more than the 50% drop requested. –  doppelgreener Oct 15 '14 at 8:00
2  
@doppelgreener correct. If the related #2 was implemented (at least in that part), there would be no need for this one. But it's status-declined wholesale and I "extracted" the part I frankly like most into this request –  gnat Oct 15 '14 at 8:04
1  
related: Avoid sharing bad questions at Facebook - 'It is very "helpful" when a question on its way to closure gets 45 likes and 2 shares (many of whom probably have association bonus to vote up what they like / share)...' –  gnat Nov 27 '14 at 9:14
    
@bjb568 FWIW there is a similar request at MSO: How to prevent close-able questions from being listed under “Hot Network Questions” –  gnat Mar 19 at 9:45

2 Answers 2

One way to achieve what is requested here is to modify a step at which hot questions are shuffled to pick ones to go to sidebar.

At this step, questions with many votes to close could get "lower weight" in shuffling so that these will have less chances to be displayed at sidebar compared to less troublesome questions.

Performance wise, this doesn't look like a heavy load. The only noticeable difference from the way how things work now seems to be the need to additionally query small fixed amount of questions (100) to find out which of these have many votes to close.


Another option is for system to pick more than 100 (110, 120...) "candidates to the list" and then use data on close votes to decide which questions get to "final list" of 100. Performance load wise, this seems to be similar to the option discussed above.

share|improve this answer
3  
however simple it is, I don't expect this (or any other solution for reported issue) to be implemented in foreseeable future, because its impact on Stack Overflow seems to be negligible. Substantial, direct benefit of proposed feature would likely be achieved only on smaller sites –  gnat Feb 2 at 22:02

When you put it simply, it makes sense and is a good idea.

However, looking at it from an exploitation perspective, it could cause more system-gaming.

For example, a few months ago I edited a good question aiming for it to appear on the hotlist, and provided a good answer. Shortly after it appeared on the hotlist, the experienced user who asked the question chose to accept my answer, knowing that it would then lose some hotness rating (or so it's been said on the grapevine). It quickly disappeared off the hotlist, as I believe the OP wanted. The OP doesn't like me.

Now given that childish feuds happen between users and probably always will, giving users the power to affect hotness ratings with a few close votes would make it very easy for malicious actions to be taken.

Therefore I believe it introduces more potential problems than it solves. I agree something needs to be done about the issue, but not this.

share|improve this answer
1  
worth noting that accepting an answer officially doesn't impact hotness score: "Note that accepted answers weight not at all in hotness. This is intentional, as I feel accepted answers are a fine social contract, but not a good data point for question or answer quality..." –  gnat Jun 22 at 20:55
    
@gnat Even if it's not a part of the calculation, it's well known that accepting an answer discourages new answers, thereby affecting the hotness rating as less answers are given. –  Dom Jun 23 at 14:35
    
I see. Yeah, accepting is known to have some "cooling" effect on typical ("normal") questions. However, the kind of questions discussed in this request is not normal. In close-worthy, bikeshedding questions everyone is interested only in sharing their two cents and nobody really cares about whether it has accepted answer or not –  gnat Jun 24 at 10:53
    
@gnat that's just an example to reinforce my statement that childish feuds happen between users. The essence of the answer is that the idea proposed in the question would make it a lot easier for users to act maliciously. –  Dom Jun 24 at 13:52
    
understood now, thanks for explaining. What you fear though, seems to be extremely unlikely, if you take into account that current system already gives users means to manipulate hotness rantings with relative ease - by voting question and answers up and down. Given that so far voting up/down hasn't been known as too problematic, it is unlikely that voting close would be noticeably more dangerous. This is especially so if you notice that as opposed to votes up/down, close voting is not anonymous and as such it quite easy to monitor –  gnat Jun 24 at 14:01
    
Regardless of likelihood, or your estimation of "extremely unlikely", it's still an additional vector for attack alongside the ones you've already mentioned. No malicious behaviour on the network is "quite easy to monitor", it takes a lot of effort, even if you don't see it. –  Dom Jun 24 at 14:27
3  
with all due respect, I still believe that matters of likelihood and ease of monitoring really matter here. If one is going to consider implementing proposed feature, it looks only natural to assess expected gains (avoid promoting inappropriate questions) against risks (likelihood of "attacks"). And just in case if you aren't aware, partly this assessment and monitoring is already done - closed questions (those taking 5 votes, not 3-4) are already excluded from hot list, which apparently suggests that associated risks are considered acceptable for these –  gnat Jun 24 at 14:36
    
With all due respect, there are too many aspects of your arguments to respond to in a comment, and we're just going round in circles. I've clarified my statements sufficiently, so now we're just arguing opinions. –  Dom Jun 24 at 14:52
1  
what you seem to miss is that votes to close don't come from thin air. These are from users already trusted to decide what content is appropriate for the site. The question is not whether system should trust them (because it already does) but how much. Currently, system completely stops advertising when 5 of these users vote, the question is should it somehow account when "just" 3-4 of them already voted –  gnat Jun 24 at 18:36
    
A user with 2 or 3 sock puppets and a grudge can stop their adversaries from getting questions on the hotlist without closing the q, and just let those CVs age away. That'd be real hard to track. Or even just a group of users with a grudge against a particular user. –  Dom Jun 24 at 19:13
1  
are you aware that voting to close requires 3K reputation. If it would be possible to get "2 or 3 sock puppets" each having 3K rep, this site would have much much bigger problems than advertising close-worthy questions. Generally, just keep in mind the simple fact that no matter what, system already is designed to trust five users with 3K rep each to decide on what questions to close (and remove from hot list). Only question is, should it somehow account when 3-4 of these users reach agreement –  gnat Jun 24 at 19:17
    
@gnat you make a lot of assumptions and statements without evidence. In answer to the only question, for the final time, no it should not count with 3 - 4 CVs, because that makes it a lot easier for unhappy users to take undesirable actions. –  Dom Jun 25 at 9:19
1  
it's a straight and simple fact that 5 votes of 3K users close the question and drop it off the hot list –  gnat Jun 25 at 12:51
2  
@Dom "you make a lot of assumptions and statements without evidence." - Uh, what assumptions, what evidence? That closing requires 3K rep or that closed questions don't appear on the hotlist? Those are straight-up facts. But well, for a start take a look here. –  Christian Rau Jul 7 at 15:12
1  
I'm trying to imagine how smart (and jobless?) a user has to be to get their sock to 3k rep. –  inɒzɘmɒЯ.A.M Jul 11 at 19:26

You must log in to answer this question.

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