66

Sometimes a question that is not representative of a site makes the HNQ list. This can lead to a large bump in traffic on a question the community may not be well equipped to handle. I propose that moderators and gold badge holders in the tag be able to set the "hotness points" to zero (or vote to reduce the points).

The idea is for the HNQ to represent questions that are both "popular" and a good (not just not bad) fit for the site.

  • 4
    Well, moderator can simply close the question in such a case. Closed questions don't appear in the Hot Network Questions list. (And if it's not really off topic it can be reopened after a day or so, then it won't make it to HNQ again.) – Shadow Wizard Sep 24 '16 at 17:46
  • 13
    @ShadowWizard but that seems like a hack to keep it off the HNQ list. – StrongBad Sep 24 '16 at 18:48
  • 7
    @Sha but then people would complain that the reason doesn't fit, or the closure is unwarranted, and suchlike. Regarding HNQ, there has always been two conflicting parties, one that endorsed them because they're good entertainment, and one that hated them because of the extra moderation load. I'm not saying it would never get out of hand because a mod might be really tired one day, but it's certainly an idea to consider. – Margarine Sep 24 '16 at 19:11
  • 7
    This is actually a great suggestion as I have felt embarrassed when I witnessed poorly-researched and unhelpful questions appear on HNQ. That proves voting, especially upvoting on SE doesn't work as well and effectively as it has been designed and hoped for. I don't see any harm allowing a moderator to delete their own HNQ question from the list if they want to. – Rathony Sep 25 '16 at 10:20
  • 2
    @ShadowWizard at smaller sites moderators can also proactively protect questions that enter or are about to enter HNQ. This approach appears to be popular eg at Physics.SE. FWIW it seems to be against this ancient guidance but to me it somehow feels fair to ignore it – gnat Sep 26 '16 at 9:16
  • @gnat yeah, I've also seen feature requests here asking to auto protect any question that becomes HNQ. – Shadow Wizard Sep 26 '16 at 9:18
  • 2
    @ShadowWizard Automatically protect questions that hit the Hot Questions list - they pretend that this is addressed by too-much-noobs auto-protection but per my observations this doesn't work particularly well – gnat Sep 26 '16 at 9:41
  • 1
    Editing and improving the question seems like a better approach than closing it. – Christophe Strobbe Oct 7 '16 at 20:36
  • 1
    @ChristopheStrobbe what about questions that are well written but only popular because of their strangeness and not for their actual usefulness? I.E. a question that is "popular" but not one that is a prime example of the community. – The Great Duck Jan 14 '17 at 20:38
  • @ShadowWizard Maybe also a golden tag badge holder could close the question as dupe, if he can find a similar one with a little search. I think the HNQ algorithm should in this case substitute the also the questions, what it may or may not do. – peterh May 18 '17 at 15:34
  • 1
    The sites with HNQ problems are mostly also the sites that don't have very many gold tag badge holders/30k+ rep users/etc. Of the usual suspects, only Roleplaying Games (34 gold badges), and maybe Worldbuilding (14) and The Workplace (11) might benefit from this. In contrast, Arquade's got five (four for games that don't tend to generate problematic questions), Skeptics has none, and Interpersonal Skills doesn't even have any silver tag badges. – Mark Dec 26 '18 at 23:10
  • @Mark so for small sites it is a mod only tool and for big sites there are a few more people who can help out. The key is to have a way to get things off the HNQ without taking it out on the person asking the question. – StrongBad Dec 26 '18 at 23:54
12

This is now implemented (at least for ♦ moderators):

Moderators have the ability to remove questions from the HNQ List.

There are times when the hotness formula selects a question that a site would rather not have featured. Up until now, the only recourse that was available was to close the question (which may be appropriate anyway but isn't ideal when done purely to manage traffic), or to do nothing. We're putting the power in the hands of our moderators to remove questions that don't set a good example for their sites.

49
+50

I believe your reasons for this are quite commendable. While it might be challenging to implement, it's not a bad idea at all.

A simpler way, perhaps, to deal with this is to simply increase the weight downvotes are given in the HNQ formula. At the moment, the formula merely looks at the total score. This is problematic because of the huge influx of 101 rep users and the asymmetric voting that leads to. It has been put quite excellently:

The asymmetry of hot list effect also comes into play here. Those coming from Hot Questions may recognize the question is crap, and perhaps 50% of them would downvote it under normal circumstances. But having only 101 rep, they can't. The 5% who want to upvote (for reasons that are difficult to fathom) can and do. Result: vote count that does not represent the opinion of either the particular community, nor the network at large.

It would be good to counteract this asymmetry. If not at the root, (which is the votes themselves) then at least at the undesirable results, one of them being the way it causes a snowball effect in the hotness points.

One such counteracting measure could be that downvotes on the question (not on answers) could be given a lot of weight in reducing the hotness points. So downvotes would take away considerably more hotness than upvotes add.

This would actually be allowing users to remove hotness points more effectively. But, of course, a much larger group of users. (Anyone with > 124 reputation.)

Mostly, the questions we would want to take off the HNQ list are the same questions that are worthy of downvotes. (Lack of research, extremely trivial, etc.)

Right now, very trivial questions often hit the HNQ. The questions themselves may "only" get like +25, but the answers often get +60 or more. This is a lot on many of the smaller sites. Due to the triviality of the matter, most of the HNQ visitors can easily recognize that the answers are correct, and thus upvote with a clean conscience. It's not those users' fault, it's fine to vote that way. But the highly upvoted answers do nevertheless contribute to the snowball effect on the hotness of that trivial/poorly researched question. Normally, such trivial content would perhaps not be upvoted so highly nor earn users so much reputation.

Figuring out how much weight to give downvotes could be done by trial and error. Start with one value. If it's not effective enough, then increase it, etc. This way, the sweet spot could be found. Personally, I imagine that quite a heavy weight could be implemented.

This might cause silly, trivial questions to be less likely to stay in the HNQ for a long time. And it would counteract the present effect in which their hotness is a self-fulfilling prophecy.

Would this ruin the HNQ? I don't think so. Rather, I imagine that it would save the HNQ. This way we could have questions featuring more quality and good content advertised network-wide, rather than a lot of the trivial, clickbaity questions we currently often see on the list.

Related: Prevent questions on Hot List from being upvoted by casual visitors (only rep is from association bonus)

  • 5
    This could actually well remedy codegolfs (note: I was never really active there but I also haven't been there in like a year so things may have changed) problem with bad golfing questions I saw on their meta. Point of complaint was drive-by upvotes due to HNQ but no downvotes because need 125 rep. – Jan Sep 24 '16 at 20:33
  • 2
    @Jan Yes. BTW, that asymmetry is discussed in detail here. – Revetahw Sep 24 '16 at 20:43
  • 1
    It seems a little unfair to down vote a question just to keep it off the HNQ. Down votes are for bad content, but I am worried about good content that just isn't a good advertisement for a site. – StrongBad Sep 24 '16 at 23:48
  • 3
    @StrongBad I understand your concern, but I believe the questions that are worthy of being downvoted are the same questions that are bad advertisement for the site. For example, questions that feature little or no research. (Like the downvote tooltip says.) Or are extremely trivial or easy to answer. Also, I believe that if questions have good content, they are also good advertisement for the site. What kind of questions are you talking about that have good content but are not good advertisement for the site? – Revetahw Sep 24 '16 at 23:55
  • 8
    I would much rather give downvotes greater weight in the hotness formula than saddle moderators/gold-tag-badge holders with the responsibility of keeping an eye on the site's HNQ representatives. At a basic level the "hotness" of a question should measure a community's response to it; that a question receives mixed votes should be an indication that it's not the best representative. – user642796 Sep 25 '16 at 14:20
  • 1
    @ifajra Yeah good points. – Revetahw Sep 25 '16 at 14:23
  • I think what you may be glossing over here is that things which don't work well as HNQs may still be fine for the site. Maybe they're trivial, maybe they're everyone's favorite topic, but that doesn't make them bad. So if people's behavior doesn't change, there won't always be enough downvotes to warrant keeping the question off HNQ. And if people's behavior does change, then people are now downvoting questions specifically to keep them off HNQ, which as StrongBad says seems quite unfair to the OP. – Cascabel Sep 27 '16 at 22:45
  • @Jefromi Yeah, I understand what you're saying. "Wanting to keep it off the HNQ" is not a good motive to downvote. Still, HNQs usually get a lot of upvotes anyway, and downvotes take away less rep, so it's not like OP is going to suffer or anything. Still though, I do agree that it would be unfortunate if people downvoted where they otherwise wouldn't have. It's one of the drawbacks of this proposal, indeed. (So maybe OPs solution is better, what do you think?) Also, what kind of questions do you think would be "bad" for the HNQ but "good" for the site? – Revetahw Sep 27 '16 at 22:51
  • One example from my site: cooking.stackexchange.com/questions/73396/… got a lot of duplicative pile-on "technically fruit is..." along with the inevitable "debates", some iffy voting, etc, but it's a legitimate question we were happy to answer. More broadly, if you agree that getting tons of rep for answers to really basic questions is undesirable, then those would be better off avoiding HNQ (but are fine for the site). – Cascabel Sep 27 '16 at 23:02
  • 1
    @Jefromi Well, OP got lots of upvotes and nice repuation from that question, so even if some people would have downvoted it "just to get it off the HNQ" OP would have probably been OK. Again, I do admit that it's not ideal if people downvote just because of that. So yeah, good example. I'll happily admit that that is one of the drawbacks to this suggestion. I personally think it's a small drawback, but alright. So, maybe OPs suggestion is better? I actually think OPs suggestion is quite alright, as I started this answer by saying. This was just meant as an alternative. What do you think? – Revetahw Sep 27 '16 at 23:07
  • I totally agree this is a useful idea - at the very least for questions that are already getting downvotes. It's just that if it's the only mechanism to avoid HNQ, then it's going to encourage downvotes - so it'd be best if it weren't the only one. For example, if protected questions were excluded (or mods had some other way to exclude them), then users could request that rather than downvoting. – Cascabel Sep 27 '16 at 23:35
  • 1
    @Jefromi Yeah, another way to remove them from the list would be good, if it was implemented well. I don't like the "protected" suggestion, since most popular NQs get protected anyway, regardless of quality. – Revetahw Sep 27 '16 at 23:36
  • somewhat similar feature request that suggests giving some weight (which is currently zero) to negative feedback from site regulars in form of close votes: At smaller sites, penalize hot questions having 3-4 close votes – gnat Oct 4 '16 at 16:34
  • 1
    @gnat Agreed, that is a somewhat similar. – Revetahw Oct 4 '16 at 16:35
19
+300

I agree with the suggestion to take downvotes into account in the hotness formula. If that doesn't produce good-enough results, here are some other things I would try, in order:

  1. Adjust formula for downvotes, per that other answer.

  2. Exclude protected questions from HNQ. This, like the downvotes suggestion, puts the control in the hands of the community instead of requiring mod intervention.

  3. If adjusting the formula for downvotes and/or excluding protected questions isn't sufficient, then allow mods to mark a question as "no HNQ". This would be another option on the mod menu. (Like locks, annotations, and protection, it should be a toggle.)

  4. If all of that fails or is rejected, then make "exclude from HNQ" a high-rep privilege like protection. This has more UI impact and carries a user-education burden, which is why I'm putting it last.

I think tying it to tag badges misses the mark; a high tag score doesn't seem to me to make one more qualified to judge hotness than a more-diversified user with 15k rep. On some sites there are no, or very few, gold tag badges, either because of the number of users or the number and variety of tags. Sites shouldn't adjust their tag collections just to defeat the HNQ list.

  • 2
    The idea is not that the gold tag badge holder judges "hotness", but that they know what is representative in that tag. A 15k rep SO user may have no idea what is representative on the Matlab tag. – StrongBad Sep 26 '16 at 22:04
  • 2
    @StrongBad but on SO, there are tons of other high-rep users who do have an idea. Maybe SO is different; on the sites where I'm active, there are experienced users who are well-qualified to judge what should be hot even though they don't have gold badges. In some cases, the ones with more tag diversity are even in a better position to judge it, because they see more of the site instead of taking a deep dive into only specific tags. (Again, SO is different that way; nobody reads everything on SO. But there are people on Workplace, Worldbuilding, and Mi Yodeya who do.) – Monica Cellio Sep 26 '16 at 22:07
  • Huh. I just noticed that posting altered my numbering. I'd written this starting with 0 (to further clarify that the first step is to follow the other answer), and it got rewritten. – Monica Cellio Sep 27 '16 at 2:46
  • @StrongBad your idea (as probably any other feature request about hot questions) applies to all SE sites except for Stack Overflow (I would recommend editing your post to clarify this in order to avoid further confusion). Thing is, the way how hot questions work at SO is totally different than at all other sites in the network. This is because of special adjustment that pushes SO questions out of hot list after 7 hours. Because of this adjustment SO doesn't experience any of the problems that other sites have – gnat Sep 27 '16 at 7:54
  • 1
    +1 for almost everything, but I'm not sure I like the "protected" part. Almost all popular HNQs get protected, right? Wouldn't all those be removed, then? – Revetahw Sep 27 '16 at 23:37
  • @Fiksdal do they? I know some do, but I don't think it's a large proportion on the sites I frequent. I'll try to pay more attention. Unfortunately, we can't science it because hotness data isn't in SEDE. – Monica Cellio Sep 28 '16 at 2:20
  • 3
    @MonicaCellio Take a look at the current HNQ list. Look at questions that are older than 40 hours. The vast majority of them are protected. – Revetahw Sep 28 '16 at 17:56
  • 1
    Possibly 40 hours is too long for any question to be hot, so that might be a self-correcting problem. :-) – Monica Cellio Sep 28 '16 at 18:31
  • 1
    @MonicaCellio Are you saying no question should be "hot" for that long? Well, I think really epic questions with great answers may deserve to be hot for three or four days. – Revetahw Sep 29 '16 at 13:32

You must log in to answer this question.

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