14

The Hot Network Questions (HNQ) list is a list of questions drawn from across the network. It appears in the right column on the front page of the site and on all question pages. An algorithm computes "hotness" based on velocity (votes and answers), scores, and newness. New, controversial, active questions make the list. This has led to a lot of complaints from across the network, and SE is starting to think about how to change it. Any changes will take quite some time, of course; we don't have a proposal or design yet.

Meanwhile, in the wake of recent events that led to another site being removed from the list, moderators have now been told that sites can petition to opt out if we show strong community support for doing so.

I propose a trial run. Do we, as a community, want to opt out of having our questions shown in the Hot Network Questions list for 90 days? After 90 days we would evaluate the effects and decide to either continue or ask to be reinstated. If something dramatic should happen before then, like they change the algorithm or our traffic drops by an order of magnitude, we can decide as a community to end the experiment early.

Opting out means that questions from The Workplace wouldn't appear on the list. It does not mean that we would stop seeing the HNQ list here. We'll still see the HNQ list with questions from across the rest of the network.

We won't do anything for at least a week, to allow time for people to respond, counter-propose, vote, and otherwise weigh in. After that, if we have clear community consensus we'll proceed with a request to the CMs; if we don't have clear consensus we'll make no changes.

Please indicate your support or opposition through answers (and votes on answers). I'll be adding an answer to express my own position, and I'll leave it to others to express other positions in different answers.

  • There has been a lot of posts about changing HNQ (both in the last years an recently [as you know]). I think it would be better to link and reference to the proposed modifications in order to let the community decide what is better – llrs Nov 23 '18 at 21:23
  • 3
    I did - the link is in the first paragraph. If you had something different in mind, please add. Thanks. – Monica Cellio Nov 23 '18 at 21:44
  • 5
    Are we a "community"? Or just a place where people occasionally drop in if they see an interesting HNQ? I'm sure someone has the statistics. – Michael Kay Nov 23 '18 at 23:54
  • 2
    Once we start trying to force peoples voting habits then I think we're getting swelled heads, which I have noticed with some people. People should be allowed to vote how they want and given every opportunity to do so except for a tiny minority of abusers, it's their site. HNQ is not even a problem to TWP, lets not make it one. And lets not make an embarrassing to everyone involved twitter-whatever-you-want-to-call-it that had nothing to do with TWP even become an issue to TWP. I don't give a crap about it and I would think very few people do. – Kilisi Nov 25 '18 at 15:15
  • 3
    If HNQ = entertainment sometimes... so what? But thats just a false dilemna anyway, current HNQ for us is workplace.stackexchange.com/questions/123237/… which looks fine to me. – Kilisi Nov 25 '18 at 15:21
  • 1
    Great idea ! – Fattie Nov 29 '18 at 12:36
  • 1
    Another way of handling this matter would be (another) queue for that. When the bot handling this detect a possible hot question it wants to share with the network, it put the question in a queue and there the community have ... let's said 24h only (it has still to be "hot" after all) to eventually block the question from going into the HNQ list as the question might be considered as not representative of the site. Also delaying the publication of a question by 24h is a good way to fix some problems before it goes live. At least that make a fair compromise between SE and the community. – Walfrat Dec 3 '18 at 9:43
  • 4
    FWIW I understand some problems of the HNQ being a high rep user over on SFF where vote exaggeration (TM) is a big problem once something hits the HNQ. However, the vote problem is usually only very localised to that specific post and the extra views it draws in to the site as a whole is more beneficial than a "few" extra votes here and there. Whilst I am mainly a drive by HNQ user here I haven't seen anything overly problematic that you would need to opt out though of course I only see a narrow window of the site. – TheLethalCompany Dec 3 '18 at 10:00
  • 1
    @TheLethalCompany Our problems are less about extra votes and more about the extra discussion. The Workplace tends to have more subjective questions, so everyone wants to add their own two cents, even if it's been said 20 times already. Throw in the fact that a lot of HNQs have controversial topics, and you get a lot of heated exchanges and personal attacks in the comments. – David K Dec 3 '18 at 17:41
  • 1
    The meta post linked by Monica says that the HNQ problem is merely one of "scale" (see the "what's broken" section). Looking at the language here, it appears folks have wildly different and very incoherent ideas about what the problem actually is. But the thing is, stackexchange is a media product like many others. One can't rely on automated tallies of "hotness" to decide what deserves promition. Doing so is a recipe for clickbait. As much as it tickles me to see SE purists get bent out of shape, HNQ definitely needs some curation and that may mean different things to different sites. – teego1967 Dec 4 '18 at 11:06

18 Answers 18

55

I found out about this part of Stack Exchange through HNQ so I'm a bit biased. But if you're looking for more people, that's an easy way to "advertise".

I'm new here, so don't really know the needs of the "community". Maybe this sort of person is not what you're interested in (at least after the redesign).

But being new, I'm sharing my perspective, as someone who was aware of the "workplace" category mostly from HNQ.

  • 1
    Thanks for sharing! We do want new people to join us. We also want to avoid repelling people with provocative-sounding questions that hit that list. It's hard to know how much this happens, but people complain so I proposed the experiment. – Monica Cellio Nov 26 '18 at 1:23
  • 3
    My own experience is that I see these questions and think "that's a good question" -- though maybe I'm just seeing less controversial questions. – Nate 8 Nov 26 '18 at 1:26
  • 1
    I'm glad to hear that! We do try to address the wilder ones when we find out they're on the list (ideally by editing the titles to be more informative and less sensational), but we don't find out in advance and have no way to directly control it, so it's kind of hit or miss. – Monica Cellio Nov 26 '18 at 1:33
  • 2
    I, too, came here via HNQ so I'm probably biased - I get the downside but I also see the upside in terms of attracting new members. – dwizum Nov 26 '18 at 19:05
  • 2
    I probably would also have never found TWP, if not for seeing it in HNQ. – Time4Tea Dec 7 '18 at 17:59
40

Why fix something that isn't broken?

  • 8
    Well, HNQ is broken, so we are trying to fix it. – David K Nov 26 '18 at 13:15
  • 18
    @DavidK it's not broken, a few individuals got caught up in something from another site and suddenly it's a broken big deal here... na... let them sort their own crap without involving us I reckon. I see no need for us all and TWP to be roped in to someones agenda. Potential issues, but no benefits for us. I've been quietly watching, these same people won't lift a finger on our behalf and we have more important unsolved issues to look at. – Kilisi Nov 27 '18 at 0:53
  • 3
    You might think that HNQ is a problem, but many people disagree, and have since long before this latest incident. You can see all the questions tagged hot-questions here, and the discussion has been going on even longer at the main Meta site. Many people here have wanted the HNQ to be changed for quite a while, but it's just that now is the first time we are actually given an option to do anything about it. – David K Nov 27 '18 at 12:47
  • 1
    Err, "you might not think HNQ is a problem". Darn edit time-limit. – David K Nov 27 '18 at 13:19
  • 4
    @Kilisi Well said. As usual. – Lumberjack Nov 29 '18 at 18:19
  • 2
    I don't think HNQ is as much of a problem as it's sometimes made out to be, but passive-aggressive insults (in your reply comments) don't make for a very compelling argument. – V2Blast Dec 1 '18 at 10:45
  • 1
    It's a discussion. if we start dancing around the bush it's not a discussion, it's a disco. So I'll just speak my mind and leave the PC ballet to anyone who wants to pirouette around in a frilly tutu. I'm serene on the possibility of people going to all the trouble of upsetting themselves somehow. – Kilisi Dec 2 '18 at 4:44
  • 1
    @DavidK Some small percentage of questions end up in HNQ. Some small percentage of that set of questions have titles that some tiny percentage of users may find offensive in some way. If that's your criteria for "broken", we may as well just shut the whole SE network down. – berry120 Dec 11 '18 at 12:15
  • @berry120 This has nothing to do with the offensive titles - the recent twitter incident just finally caused SE to allow sites to do something about HNQ. The problem is about creating lots of low quality content. I don't want to rehash everything here, but you can see my arguments on this other thread. – David K Dec 11 '18 at 13:16
18

I came to this site through a HNQ link and have since become an active member. I feel the additional traffic gained makes it worth the general level of quality in that traffic.

At any given time this community only has a few dozen members that are active on a daily basis giving quality answers. We need replacements for enderland and Maskedman and whoever else will quit over the next 1, 2, or 5 years. By closing off a common entry point to the site I'm very concerned that the short term effects are of little importance compared to the long term.

14

I am not really a part of this community, since the only thing I do here is flagging spam.

TL;DR first, I'm against getting Workplace out of HNQ.

A bit of background of myself. I'm a university student as of August 2017, majoring in computer science. I have been actively paying attention to existing working environments in this industry, which is primarily working in front of computers in offices (not carrying bricks on ports and docks).

I am very interested in what I will encounter in my future working place. And since it's neither a direct nor an upcoming case, I don't and have no reason to spend hours a day preparing for this.

On the other hand, I am very active on some other sites like Stack Overflow and Meta. The HNQ is my primary entry to finding interesting content on sites where I'm not active. It is the reason that I joined Programming Puzzles & Code Golf, as well as Information Security.

My objection is that I see no reason for Workplace to opt out of HNQ. I understand that Interpersonal Skills was kicked out for randomly having "potentially NSFW" question titles (see the top few images of this answer). However, this is not the case for Workplace, as Workplace is... well, for questions in the workplace.

As a conclusion, see the 2nd paragraph.

5

I think Workplace should stay in.

There's another set of users that get missed in questions like this, users like me: users who do not browse the site regularly, who are based in another site primarily, but are reminded to come over and browse the site when there is a HNQ that looks interesting to me (in fact, why I'm answering this now).

HNQ isn't perfect. But it's better than not having it, in my opinion. (And I say this as a mod on a small site that has a lot of work any time we get a HNQ.)

5

I only have two posts on this site, but I've upvoted several questions over time as I come from HNQ links. I find many of these questions intriguing at the least and enjoy reading them. While I do understand the network-wide needs to improve what is a HNQ, I think the current cost is a worthy one that brings in many legitimate members.

5

I'm only here because of HNQ. Frankly, my participation in this site would drop drastically if we opted out. I think we need to realize that this community is generally more chill than Stack Overflow. I for one think improvement here means being more open in regards to industry and paradigms. I'm fairly certain opting out of HNQ will be a detriment to that end.

Opting out of HNQ would be bad for this community imo.

4

I feel that I agree to this on a strictly trial basis and run some analysis on how this affects question quality and engagement of new users. There’s certainly a balance between salacious questions designed to attract clicks and gaining new users with good questions.

I suggest that we wait until after the holiday period so that we have a daily good idea of how this change affects the quality of the site. Otherwise the cause of any dip in question asking rate might be ambiguous between hnq and the holiday season dip.

  • 1
    I agree. We should see how this works and after that decide on a more permanent call – DarkCygnus Nov 23 '18 at 23:00
  • 4
    How many good questions actually show up on the HNQ list? I'm not sure goodness is part of the criteria for selection. – Joe Strazzere Nov 24 '18 at 0:23
  • 1
    @JoeStrazzere aren't they all good questions until proven otherwise via votes? My understanding is HNQ generates some traffic... also a good thing. Seems a bit paranoid to think that people will deliberately try and abuse it, and even if they do it's a minor issue compared to loss of traffic.... I could be wrong of course. – Kilisi Nov 25 '18 at 14:51
  • 2
    @Kilisi - votes aren't the proper measure of goodness, IMHO. I see lots of stupid questions get lots of votes. I see some excellent questions that get few votes. And my goal here isn't to maximize traffic. I don't think I'm paranoid. But I guess by definition, that's for others to judge. – Joe Strazzere Nov 25 '18 at 16:15
  • @JoeStrazzere you might be right, but votes are the only measure we have. Traffic isn't my goal either, but it's essential for the site. – Kilisi Nov 25 '18 at 16:35
  • 2
    Regardless of when we do it, we should look at year-over-year stats, not just the immediate before and after. All SE sites have fluctuations throughout the year. – Monica Cellio Nov 25 '18 at 16:42
  • @Kilisi - one should construct their measures based on their goals. Not the other way around. Just because it happens to be easy to measure votes, that doesn't mean it has any value if "goodness" is your goal. – Joe Strazzere Nov 26 '18 at 1:20
  • 1
    @JoeStrazzere I've never been one to follow others agenda blindly. I'm just about helping people, no traffic, no people to help. Don't really care about site vision except where it coincides with mine. Those that do seem to go off the rails over it. – Kilisi Nov 26 '18 at 2:03
  • 2
    @Kilisi - I think we agree on most points. I'm guessing we'll still have traffic without any HNQ. But that's what a good experiment will confirm or deny. – Joe Strazzere Nov 26 '18 at 12:25
  • @JoeStrazzere FWIW, IPS lost a good 60% of it's traffic since then. Relatively minor changes in # of questions asked, major dip in answers, no change in downvotes, major dip in upvotes, 66,67% of all questions asked in the past 30 days closed. – Magisch Nov 26 '18 at 13:46
  • @Magisch - sounds good to me (except for the Closed part) – Joe Strazzere Nov 26 '18 at 13:48
  • @JoeStrazzere that's compared to about half from before then, but that number is likely skewed by the organized cleanup effort of old questions and answers that has been going on since then. – Magisch Nov 26 '18 at 13:54
  • @Magisch this number looks skewed by some historic cleanup indeed: when I studied hundred of their recent questions mentioned in my answer here close rate looked noticeably lower, like 30-35% – gnat Nov 26 '18 at 16:28
  • When does the holiday period end? 7th of January? – Peter Taylor Nov 29 '18 at 10:43
  • @PeterTaylor: When does the holiday period end? -> within 6/8 weeks. – OldPadawan Dec 7 '18 at 7:44
3

The Hot Network Questions list has been a source of complaints here throughout the years. Specifically, questions -- usually sensationalist or controversial questions, because that's what HNQ optimizes for -- get onto the list that's shown in the right sidebar all across the network, attracting vastly disproportionate voting and lots of comments. We also get more than a few provocative questions here that seem designed to try to get onto that list.

Let's try some time away. I think it'll be good for us. If we want to better advertise (the good parts of) The Workplace across the network, we can design some ads and submit them to other sites.

  • By submitting adds do you mean the Community Adds that one includes as an answer on the meta post, and if it reaches X UV it gets displayed? Which site(s) do you think would be worth "promoting" us in? – DarkCygnus Nov 23 '18 at 22:37
  • 5
    @DarkCygnus I mean the community ads, yes. If we wanted to advertise on other sites I think we could make decent proposals for many of the professional sites -- Software Engineering et al but also non-software sites like Electrical Engineering, Home Improvement (aim for professional contractors), Motor Vehicles (mechanics), Personal Finance, Freelancing, Project Management, Writing (if they ever let it shake the beta mislabel)... somebody would have to design and make the ads, though. – Monica Cellio Nov 24 '18 at 23:26
  • 1
    @MonicaCellio This sounds like a discussion for a new Meta post, but I think we could probably come up with a generic one-site-fits-all ad that could be proposed on a large number of the other sites. We wouldn't necessarily have to specify which ones; if a user thinks it would be worthwhile on a site they can propose it there themselves. – David K Nov 26 '18 at 13:19
  • @DavidK I encourage anybody who's interested to kick off that discussion. Let's see what people think about intentional advertising on other sites -- where (or whether) we should, what we want to convey in ads, design ideas, etc. – Monica Cellio Nov 26 '18 at 16:21
3

From the responses over the last nearly three weeks, it seems clear that the community wants to stay on the Hot Network Questions list. If somebody wants to re-raise the issue in the future feel free, but for now we won't request any changes.

The primary tool we have in response to a "bad" entry on the HNQ is the edit, and it's something all of us can do. If you see a title that looks bad out of context, see if you can make it better. Ditto if a popular question isn't as clear as it could be and you're able to see how to improve it. Even if the question isn't currently hot, these kinds of edits improve the quality of our site. Please help keep The Workplace looking good and professional.

  • 1
    Wow, that's disappointing. It became clear to abandon the experiment because of a few answers here in Meta? Or did folks reach out to you in private? I'm just wondering how one concludes the feelings of such a vast community. – Joe Strazzere Dec 13 '18 at 14:38
  • 1
    @JoeStrazzere nothing private; I'm reacting to the voting on all the answers here. The community doesn't seem to be behind the experiment. – Monica Cellio Dec 13 '18 at 14:51
  • 1
    Thanks. I don't agree with the conclusion, but I understand how you reached it. Thanks anyway for the proposal and attempt. – Joe Strazzere Dec 13 '18 at 15:19
2

I will agree to a trail as long as at the end, we can get some numbers (posted in META) about how it went. So things like:

  • Overall site traffic (before and after)
  • Average question view for a sampling of question
  • Average number of answers with a positive rating
  • Average number of answers with a negative answer

It doesn't need to be exactly this criteria, but just something to illustrate the pros and cons after the trial. I found TWP through HNQ, so while I am willing to have us removed, I want to see some stats before a final decision.

  • 1
    If we decide to proceed, I (or someone else who beats me to it) will ask a separate question about what exactly we want to measure. Evaluation criteria should precede the experiment. – Monica Cellio Nov 26 '18 at 16:23
  • @MonicaCellio i'd be cool with that – SaggingRufus Nov 26 '18 at 16:26
  • Please propose things we should measure (before and after) in answers to this question. – Monica Cellio Nov 27 '18 at 3:40
2

I came to Workplace via HNQ. It doesn't seem like Workplace gets all that much traffic that we'd want to close off one of the ways that people come to the site. I visit fairly often (ok, too often) and I don't feel like the site is swamped with bad questions and answers.

We're balancing two things here: getting people to visit Workplace so it doesn't die, and keeping the riffraff out. My feeling is that we don't have a riffraff problem, and if we didn't have HNQ, we might suffer a population collapse. So I'd vote that we keep HNQ until it's clear that it's causing a problem for us.

0

Do we, as a community, want to opt out of having our questions shown in the Hot Network Questions list for 90 days?

I never knew that was a possibility.

Yes - do it yesterday.

I'm not sure what the whole point of "Hot Network Questions" is, but I'm pretty sure it's aim is not to help people navigate the workplace.

  • 1
    The only ones that make it to HNQ are the salacious or controversial ones. It doesn't help the site, and just ends up getting us swamped with 101 rep users who clog up the place with annoying comments. – Retired Codger Nov 24 '18 at 4:17
  • 14
    "I never knew that was a possibility." -- it wasn't; communities that asked got told no. Then somebody with a lot of twitter followers ranted about a couple questions, and in under an hour a site had been removed from HNQ across the network. Surprise -- turns out they can do it! And so people from other sites started asking too. Since, in the wake of the tweetpocalypse, I imagine that smart, motivated users could otherwise plot to plant questions that would get sites booted, I'm hoping SE realized the wisdom of doing it the nice way instead. :-) – Monica Cellio Nov 24 '18 at 23:21
  • 1
    note about the aim of HNQ appears to be correct: last time I checked it was officially stated as entertainment – gnat Nov 25 '18 at 9:09
  • 2
    @gnat - okay. We should opt out of the entertainment goal. – Joe Strazzere Nov 25 '18 at 12:07
  • @MonicaCellio - I don't follow twitstorms. I'm glad there is a possibility of doing things to shape the direction of The Workplace without resorting to silly resistance tactics. – Joe Strazzere Nov 25 '18 at 12:09
0

I support this because proposed experiment would make a great opportunity to learn about how things work. To start with, it could help us learn if Stack Exchange would do something about requests which aren't posted on Twitter.

...At first I was going to add here that from this perspective it would make useful knowledge even if they tag it declined (because that would prove that they care about meta). And I was going to make this a comment because it is rather tangential to discussed matters. However, after studying things a bit further I changed my mind and decided that I am really interested in proposed experiment...


What made a difference to me was a visit to site that was blocked at HNQ about a month ago. I checked about hundred of their recent questions (all posted long after HNQ block) and was quite surprised to see that they appear to be doing totally all right.

There was a decent amount of questions, answers, votes, views - everything I checked looked pretty OK. Everything I saw indicated a solid, active, healthy site. As if HNQ didn't matter at all.

What made me worry though (and what made me wish to try this at our site) is that I couldn't reasonably compare it to how it was prior to their HNQ block. I mean, okay, they look good now - but do they do better or worse than before? I totally lacked experience with their past questions and had no idea about their topics and quality norms to make my mind about that.

Granted, I also found one of their meta posts analyzing how different metrics changed after HNQ block. Differences in numbers were quite noticeable but again, lack of experience with past content, site topics and quality norms made it too difficult for me to interpret their stats and decide whether similar changes at Workplace would be for better or for worse.

Summing up, studying things at IPS made me very curious about how it would go here, and proposed experiment looks like a great way to find it out.

0

The up side is the site gets more traffic.

The down side is I think the site gets more trolls trying to make HNQ.

I would like a rule cannot make HNQ unless the OP has rep of 1000.

-1

While I do think the HNQ concept has value - at the moment it's fundamentally broken at best. So until they actually make it into something that isn't the SE equivalent of the Daily Mail's Sidebar of Shame I think we'd be better off out.

-2

I would say we should at least trial opting out as long as we have the choice. As alluded to in Monica's recent MSE question, the risk is there for a similar incident like on IPS to happen here - some of our question titles are just as incindiary and/or look just as bad out of context. There but for the grace of pure RNG, the site that started this storm might have been TWP instead of IPS.

Other then that, the HNQ has been a pain point and has had a quite mixed effect on question quality here over the years.

  • We're far better at patrolling titles. I seem to recall warning against the titles elsewhere and was pooh poohed and told I was making a big deal of nothing. Now, I no longer issue warnings about anything. I just pop some popcorn – Retired Codger Nov 26 '18 at 13:40
-5

Any time a Q gets to HNQ, we get a bunch of inane comments from 101 rep users who never again participate.

It's a disruption and should die a painful death.

Kill it with fire.

in the meantime, let's opt out.

  • 1
    Were you ever a 101 rep user? – Mooseman Dec 8 '18 at 16:11
  • 3
    Well, that proves my point – Retired Codger Dec 8 '18 at 17:30

You must log in to answer this question.

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