25
\$\begingroup\$

What's this? The Hot Network Questions (HNQ) list is that list of random questions from random sites you'll see in the bottom of the sidebar on most site pages. It displays questions that have attracted significant attention measured by an abstract “hotness” score based on how rapidly the question's vote count is accelerating. Since HNQ brings more attention and more votes, it frequently creates a self-perpetuating cycle where hot questions either fall off quickly or stay for days at a time.


Recent discussions on the network have been re-evaluating the Hot Network Questions feature, including what it does well and all the ways it doesn't work very well. Stack Overflow, Inc., the company that runs the Stack Exchange network, is evaluating ways to modify the feature or retire it to replace it with something else. There are positive things the HNQ theoretically accomplishes but also major problems.

In light of these discussions many communities including our own have been re-assessing their relationship to the feature.

I've reached out to find out if us opting out of the HNQ feature is among our options. It looks like it is an option if the community is on board.

Do we as a community want RPG Stack Exchange to opt out of the Hot Network Questions feature?

Here is what opting out of the HNQ would look like:

  • RPG Stack Exchange questions will no longer show up in the Hot Network Questions list.
  • The Hot Network Questions list will continue to show up on our site, presenting questions from other sites that are opted in.
  • At some point in the future, after the HNQ feature is revised or replaced, we will have the opportunity to assess whether we want to opt back in to the new system or remain opted out. (Honestly I can't promise it will unfold exactly like that—it might wind up that whenever the new thing is released, everyone automatically gets opted in again—but the point is we can probably evaluate whether we want to be in or out of that thing as well.)
  • If for some reason being opted out of the HNQ is really not working for us, we can opt back in. This wouldn't happen without community buy-in, just like this change.
  • According to site analytics, HNQ accounts for ≤5% of overall site visits, so we do not anticipate an overall significant drop in site visits.
  • HNQ is largely responsible for those questions that get answers scoring 100+ in less than a week, so that will probably stop happening. This is a minority of our questions.

There have been requests for feature changes to the HNQ system such as letting diamond moderators remove questions from HNQ, waiting periods and alerts before questions hit HNQ, etc. If these are implemented they'll be part of the new redesigned system we can opt back in to.

Here is the process of what we'll do here in this thread:

  1. We'll discuss here whether we want to opt out as a community.
  2. After 2 weeks (December 5th), if there is strong community support for opting out and not strong objection, the diamond moderator team will ask the CMs to opt this site out as described above. They will action that request.
  3. If there is not strong support, we will not opt out, and nothing will change from as they are now.

No action will be taken before 2 weeks have elapsed (i.e. before December 5th) so as to allow for enough time for discussion and voting.


Please show whether you are in support or against opting out with answering and/or voting like normal. I'm going to be posting my own answer representing my own position, and I'm going to leave it to others to provide their own positions.

\$\endgroup\$
  • 11
    \$\begingroup\$ Do we have any metrics on users whose first interaction with RPG.SE is on a HNQ that then go on to earn additional rep (say, at least 250) from other (non-HNQ) questions/answers? That is, how many actual contributors are we getting from HNQs, since we've already established it's not a major source of traffic in general? \$\endgroup\$ – Oblivious Sage Nov 21 '18 at 16:56
  • \$\begingroup\$ @ObliviousSage I'll see if I can find out the answer to that. I'm not sure if it's available via data explorer, so it might require a staff member to perform analysis and might take some time before we find out. \$\endgroup\$ – doppelgreener Nov 21 '18 at 16:59
  • \$\begingroup\$ Given that, by the looks of things, the result of this will end with "do not opt out", should we have a follow up question to this along the lines of "if we are to stay, how can we make it work for us?" Mainly trying to keep the "Subjective social questions" off the HNQ or something similar? \$\endgroup\$ – NathanS Nov 24 '18 at 18:07
  • 1
    \$\begingroup\$ @NathanS We have no tools to take subjective social questions off the HNQ list save for closing them until things cool down, which we're only going to do in exceptional situations. For the foreseeable future, expect that most social questions will hit HNQ. \$\endgroup\$ – doppelgreener Nov 24 '18 at 18:18
  • 1
    \$\begingroup\$ @NathanS I'm still happy for us to have that discussion, but I don't think I'm prepared to be the one to lead it or lay out the problems I see. \$\endgroup\$ – doppelgreener Nov 24 '18 at 21:25
  • 1
    \$\begingroup\$ @ObliviousSage Obviously this isn't statistically significant, but I joined this site via HNQ from Stack Overflow back in the day and probably wouldn't have found out about it otherwise. So at least one! \$\endgroup\$ – Please stop being evil Nov 25 '18 at 18:34
  • \$\begingroup\$ @thedarkwanderer You're not alone (as further answers here attest)! But I'm wary of drawing conclusions there: there's both serious selection and survivorship biases, and we lack the counterfactual as to whether HNQ-immigrants might have made their way through other channels. I feel ike it'd make a nice master's thesis for a data-science student interning at SE, though =) \$\endgroup\$ – nitsua60 Nov 25 '18 at 23:31
  • \$\begingroup\$ Do we have an idea when the alternative system might be introduced? \$\endgroup\$ – Pyrotechnical Nov 26 '18 at 5:02
  • \$\begingroup\$ @Pyrotechnical There's been no indication of a timeline for changes. Additionally although it appears to be important to SO Inc to update/replace HNQ, its development has to be prioritised against multiple other important projects. (Many of those are why the theme unification project is happening: Stack feature development was stalled by the heavily fractured codebase.) Realistically changes are probably months away, optimistically a single-digit number of months. \$\endgroup\$ – doppelgreener Nov 26 '18 at 11:40
  • \$\begingroup\$ I'm the guy who really wants to be able to filter rpg.stackange out of HNQ for certain reasons, and even I think this is a bad idea. You see, I might have come up with a way to turn the feature on and off based on time of day. \$\endgroup\$ – Joshua Nov 27 '18 at 2:56
  • \$\begingroup\$ I had to downvote this because it doesn't seem to be written in very neutral language. I could see your position well before I saw your own answer. \$\endgroup\$ – SeriousBri Nov 29 '18 at 8:59
  • 1
    \$\begingroup\$ @SeriousBri I did my best, but it remains I'm not neutral on the subject. \$\endgroup\$ – doppelgreener Nov 29 '18 at 9:03
  • \$\begingroup\$ @doppelgreener I hope you didn't take it as a dig, it was just meant as a comment because I know from experience how much the phrasing and tone of a question can influence an answer. \$\endgroup\$ – SeriousBri Nov 29 '18 at 9:44
71
\$\begingroup\$

Let's not opt out

The motivation for wanting out is perfectly valid, but I think the benefits of HNQ to tabletop players outweigh the problems that HNQ attracts.

The bad side first

I'm speaking of a trend here, not a universal rule: HNQ questions don't tend to be particularly good. The selection emphasizes scandal over substance, clickbait over clarity. We don't want the users to maximize those, but that's what HNQ rewards. Answers drawn in from outside the network also tend to be bad in quality because of a variety of reasons experienced users usually know to avoid --- such as promoting toxic behavior in tables or assuming that every RPG system either is or works like Dungeons and Dragons.

If it was only about the short-term well-being of our site and nothing else, I would vote to drop out of HNQ. However, I think there's a case to keep it for the visibility, because:

Our visibility helps the hobby as a whole

I had had a longstanding interest in tabletop RPGs through my youth and early adulthood, but never knew the slightest bit of how to play or who to play with. The idea of actually learning to play tabletop RPGs lied dormant in my mind, but emerged to me in 2013, when I was actively using Stackexchange --- where I saw lots of interesting RPG questions in the sidebar. Those questions led me to our site, which I hadn't known even existed before that.

I started reading RPG.stackexchange back then and the site gave me significant help in learning to play. It is a significant contributor to the fact that I am an active player of RPGs today. I don't believe I'm the only one who came into gaming through our community. I believe many people still find our hobby through our site. HNQ may account for only a small portion of page hits, but that's not relevant --- the important bit is how many new people find out about our site through HNQ.

In addition to helping people to start playing, RPG.stackexchange presents by and large a real, healthy culture of gaming. That is often not the case with other first contacts to tabletop RPGs --- video games based on RPGs, pop culture depictions of tabletop games and memes often present and reinforce a rigidly antagonistic relationship between the game/GM and the players, and omit things that are usually considered to be absolutely vital for a good RPG game like active dialogue between the players and the GM.

I cannot deny that HNQ does cause us problems, but by attracting new people to our site it helps spread the joy of both playing tabletop games and playing them in a friendly fun manner. I think the problems of HNQ are nothing but a nuisance compared to the benefit of attracting new people in.

\$\endgroup\$
  • 2
    \$\begingroup\$ I'd be interested to hear from a power-user: what concrete problems have you experienced from HNQ? \$\endgroup\$ – nitsua60 Nov 21 '18 at 17:24
  • 6
    \$\begingroup\$ @nitsua60 It got my best-voted question locked, for one (: ! The usual case doesn't get quite as extreme, though; the main problems are the occasional torrents of answers that aren't very good (either because they're not backed up by expertise or because they say nothing new) and runaway voting that messes up our normal quality metrics. I can't claim to know how badly, though. \$\endgroup\$ – kviiri Nov 21 '18 at 17:32
  • 1
    \$\begingroup\$ Thing is, it got closed, then reopened, and now you've got a high-voted question with an answer you like. There was a lot of activity around that question, but I don't see that as a problem. But this may be a place where reasonable people disagree. \$\endgroup\$ – nitsua60 Nov 21 '18 at 18:26
  • 3
    \$\begingroup\$ I don't see HNQ causing us problems that we can't handle;(minor quibble). +1 since this is a very good answer in terms of the breadth of coverage. \$\endgroup\$ – KorvinStarmast Nov 21 '18 at 20:16
  • 1
    \$\begingroup\$ @nitsua60 This question—with its original much more attention-getting title—hit the HNQ list, and the first answer (the user's only contribution to the site) started accumulating upvotes… yet that answer's useless to me and anyone who actually knows anything about the question. (Sorry, Peter S., I'm glad you got the rep, but it's true!) \$\endgroup\$ – Hey I Can Chan Nov 22 '18 at 0:27
  • \$\begingroup\$ This is also how I discovered RPG.SE too, so I can sympathise with the second half of this answer, but I believe that, due to the "clickbait-iness" of the questions picked, if we are to stay, we need to somehow prevent the "Subjective social questions" from getting in there, otherwise I'm with doppelgreener/mxyzplk on this one. \$\endgroup\$ – NathanS Nov 22 '18 at 12:56
  • 4
    \$\begingroup\$ +1. We shouldn't opt out HNQ, otherwise, I would never have known about this site, nor about RPGs that aren't computer games! \$\endgroup\$ – Ender Look Nov 22 '18 at 18:27
  • \$\begingroup\$ +1 I am a shorter way along the path you describe (who knows whether and when I'll diverge). I had heard of tabletop games/RPGs, and had a vague notion that I might want to find out more. That notion was fanned and enabled significantly by HNQ links from StackOverflow to this site. As a consequence I started reading a lot of material and co-founded a DnD (sigh yes, DnD) group with some of my colleagues. Difficult to reason about counterfactuals, but it actually seems rather unlikely to me that this would have occurred without HNQ. \$\endgroup\$ – Oly Nov 23 '18 at 13:18
  • \$\begingroup\$ +1, i'm right now in exactly the same situation you described. I was always interested in RPGs, but never got to play one. While scrolling SE i saw that there is a RPG community as well, started reading and immediately afterwards started my first game. I also found all the other SE communities i like through HNQ \$\endgroup\$ – WhiteMaple Nov 23 '18 at 15:16
28
\$\begingroup\$

We should stay in HNQ.

The arguments against HNQ: untrained users voting out of line with established community norms, new users submitting low-quality answers (especially GS/BS problems), comment argument, and then the inevitable backlash that we moderators sometimes face when new users bump up against this stack's notoriously-tight reading of some otherwise-looser Stackisms.

As a few-year active user and two-year moderator, I'll say I don't see those as serious problems. It seems to me that the structures in place handle them pretty well. You-all are good about downvoting and commenting on poor contributions, you sometimes protect questions, and you're pretty good about flagging things to our attention when it seems like more intervention's needed.

The argument for HNQ: it gets more eyeballs on the site. That's it. And I think that's important. Any community survives based on some mixture of retaining existing members and gaining new ones. Sure, plenty of people may come for an HNQ and never come back. But some will. And might become daily users. And eventually stand for election. And write this meta.

\$\endgroup\$
  • 2
    \$\begingroup\$ Two-year moderator?! I thought this election happened like several months ago... \$\endgroup\$ – Baskakov_Dmitriy Nov 24 '18 at 15:32
  • \$\begingroup\$ @Baskakov_Dmitriy Heh, temporal distortion is a thing. :) \$\endgroup\$ – KorvinStarmast Nov 25 '18 at 0:14
  • 1
    \$\begingroup\$ @Baskakov_Dmitriy well, one-and-three-quarters, I guess. [checks records] Hmm... maybe one-and-two-thirds. And maybe closer to one-and-a-half. Geez, time <strike>drags</strike> flies! \$\endgroup\$ – nitsua60 Nov 25 '18 at 23:29
25
\$\begingroup\$

I received though other channels the following from a high-rep active-most-days member; they'd rather not post under their identity for personal reason, and rep-requirements make sock-posting on meta cumbersome.

We shouldn't opt out.

I've thought for a long time that HNQ was damaging to the site. I've always been on the other side of this argument. We get bad answers, upvotes on bad answers, and so on. But then I read nitsua60's and kvirii's answers, and it got me thinking.

See, I found this site through HNQ, too. And I found it during a difficult time in my life, and it helped me through some tough years. The community here has helped me in ways I find difficult to express. I've made friends here, I've learned a lot, and I've gotten a nice sense of accomplishment out of my use of the site. Not only that, this site has made me a better gamer. If I'm being honest, it would be fair to say that it's made me a better person. I look back at some of my earlier interactions with this community and shudder at just how callow I was.

We have a great community here, and it's natural to want to protect it. I know I want to protect it exactly because it's great. The problem is, protecting it by taking away one of the main ways people find it is selfish. Yes, we get a lot of transient vote-and-forget visitors from HNQ. But we get community members too, and we owe it to them to be easy to find.

\$\endgroup\$
  • \$\begingroup\$ This is the sort of positive HNQ site interaction that seems to get lost in the sea of pitchforks. I lurk more than I post, but I'm also a HNQ convert, who said 'yes' to joining a DnD game because of previous interactions with this site, and am even now putting together my own adventure as a fledgling DM in my own right. IMHO removing HNQ wholesale and becoming more insular would do a lot more damage to the site than a few bad answers and long comment threads could ever do. \$\endgroup\$ – Robotnik Nov 29 '18 at 0:57
17
\$\begingroup\$

I like it being there.

As someone who often frequents Stack Overflow for work purposes, I enjoy seeing the occasional interesting thing coming up in HNQ.

Often, as a tabletop player myself, the ones that usually interest me are the mechanical questions. I've learnt quite a lot about edge-cases in D&D, for example, just recently, where something that is usually impossible can be worked around with the right equipment.

Some really interesting events have been put forth by other answers. Before I scrolled down that far I was simply thinking "but why would you do that?" but if the things mentioned in that ticket are a common reality, then I totally get it. It's a real shame - but I get it.

It'd be really nice if RPG could opt out - but individuals could opt back in. That way most of the riff-raff without a specific interest in RPG aren't encouraged to leave their opinionated and untested ideas, but it doesn't impact those with a specific interest.

\$\endgroup\$
  • \$\begingroup\$ I am a bit confused by this post expressing both the desire to stay in and opt out. \$\endgroup\$ – SeriousBri Nov 29 '18 at 8:46
  • \$\begingroup\$ @SeriousBri I think Shadow is describing a system where an individual user can choose whether or not particular stacks show up in the HNQ for them, as exceptions to a global default set of options. \$\endgroup\$ – Carcer Nov 29 '18 at 9:10
  • \$\begingroup\$ @Carcer I just wasn't sure if the overall position was for removal, or against \$\endgroup\$ – SeriousBri Nov 29 '18 at 9:39
  • \$\begingroup\$ @SeriousBri I'd probably say "stay in" unless such a system was developed. As in my first sentence, "I like it being there." \$\endgroup\$ – Shadow Nov 29 '18 at 22:05
2
\$\begingroup\$

Yes, we should opt out.

I believe HNQ does little to benefit our site overall, and regularly does significant harm to some of our questions and our site processes.

Over on the MSE discussion, Monica Cellio has higlighted that the HNQ feature optimises for controversy, not quality. I'll quote:

Sensationalist questions draw lots of rapid response, which feeds the HNQ algorithm, and then once it's on HNQ it gets even more rapid response, which keeps it there (and also distorts Q&A on that site).

From anecdotal experience, HNQ entries seem to fall into one of two categories:

  1. Interesting mechanical questions. The first group mostly falls off after a few hours and doesn't get much traction and voting.
  2. Subjective social questions.

It's this second group where the optimisation for quality hits us and hurts us. Difficult, controversial good subjective social questions need to be handled with care. They need our careful community evaluation, our expert answerers, and a crowd of highly knowledgeable, experienced RPG players to vote on what their experience and gut tells them will and won't work or be a good idea. They also need diligent community voting and curation to enforce that Good Subjective principles are being followed, including downvoting those that don't follow them. There also needs to be a possibility that latecoming answers will be able to compete with first-comers, overcoming the fastest gun problem.

However, once these questions hit HNQ, all of that goes out the window. We as a community become completely unable to provide any of that. It's been my consistent experience that such questions instead get a sudden influx of dozens of voters with plausibly no experience whatsoever in RPGs. They upvote answers in exponential proportion to whatever the scores already were, including if it's poor advice, amplifying the fastest gun and drowning out competing underdogs completely. This tide of upvotes overwhelms the community's ability to downvote answers not following Good Subjective.

This does an enormous disservice to everyone. It means people asking these questions don't get answers as judged by a community of knowledgeable experts with an avid interest in RPG gameplay. It means we are unable to enforce our community mechanisms of quality control properly: the community's capacity to downvote poor advice or answers that don't conform to Good Subjective is lower than HNQ capacity to upvote, and the nuclear option of deleting a post is something diamond moderators rarely do.

I really want us to be able to handle social questions without HNQ.

Opting out means that when a social question comes along, we won't get tides of upvotes. We'll get much more space to enforce our quality mechanisms, question posts, actually have meaningful downvoting, be able to provide more careful consideration and upvote late answers to outscore questions that came hours or a day earlier, like we do with mostly everything else.

I want to be able to provide that same experience with social questions that we provide for everything else, but HNQ throws that out of whack.

HNQ is bad for us, let's opt out.

\$\endgroup\$
  • 4
    \$\begingroup\$ Everything that I wanted to say here. We're a little insular, which can be viewed as a problem, but many of our topics require extremely specialized knowledge and experience. One wouldn't want a bunch of HNQ viewers providing input on the Physics SE, or one of the specialized language ones. \$\endgroup\$ – goodguy5 Nov 21 '18 at 16:15
  • \$\begingroup\$ "Category 1" HNQ aren't usually as bad. Would it be possible to let stack mods remove (or prevent) certain questions from appearing in HNQ? \$\endgroup\$ – MikeQ Nov 21 '18 at 16:18
  • 9
    \$\begingroup\$ @MikeQ “Let us take things off HNQ” is a popular feature request that may be part of the redesigned system. \$\endgroup\$ – doppelgreener Nov 21 '18 at 16:21
  • 2
    \$\begingroup\$ Can you flesh out a bit for the non-mods what sort of (concrete) problems you've experienced HNQ causing? \$\endgroup\$ – nitsua60 Nov 21 '18 at 17:25
  • 1
    \$\begingroup\$ @nitsua60 I believe I already have in the entire middle of the post. Do you mean elaborate on specific questions where there were problems? \$\endgroup\$ – doppelgreener Nov 21 '18 at 17:54
  • 4
    \$\begingroup\$ I think I'm looking at the last link in the chain of reasoning: you note that we often get 5+ crappy answers, but I'm not seeing the immediate jump from that experience to a problem for the site. Our users are pretty good about downvoting such things, they leave constructive comments, things that are totally out of whack get flagged... it just seems to me like "systems reacting normally." But maybe I'm not seeing the same things you are? \$\endgroup\$ – nitsua60 Nov 21 '18 at 18:30
  • 2
    \$\begingroup\$ @nitsua60 Ok, I'll delete that. It's not a problem such that it makes me want to leave HNQ, it's just a problem that exists. All the rest makes me want to leave HNQ. Our community's great at downvoting, but not in great enough numbers that we can overcome HNQ votes. \$\endgroup\$ – doppelgreener Nov 21 '18 at 18:37
  • 5
    \$\begingroup\$ I don't see HNQ causing any problem that we can't handle. Based on what I read over at SE/SO meta, and the 'maybe soon a feature' to let mods or very high rep users take certain Q's off of HNQ ... opting out seems an over reaction to me. We as a community become completely unable to provide any of that This kind of hyperbole strains the credibility of a pretty well organized answer. I have a very hard time accepting that, and thus your opening premise, as a valid position. \$\endgroup\$ – KorvinStarmast Nov 21 '18 at 20:18
  • 2
    \$\begingroup\$ @KorvinStarmast If that feature and when that feature does get implemented (there's no guarantees) it's a thing we can assess later. This is about the current version of HNQ. Don't bank on a feature that doesn't exist yet. \$\endgroup\$ – doppelgreener Nov 21 '18 at 20:34
  • 1
    \$\begingroup\$ @doppelgreener I agree, given how long and hard any number of folks from various stacks have lobbied for such a feature. I still don't think that the "problems" are stuff we can't handle. We can. We have a lot of folks who contribute to community moderation, and four solid diamond mods. We are in good shape. \$\endgroup\$ – KorvinStarmast Nov 21 '18 at 23:38
  • 3
    \$\begingroup\$ ‘More junk’ is not OK just because we have site mechanisms to combat it. +1, we should opt out. \$\endgroup\$ – mxyzplk says reinstate Monica Nov 22 '18 at 5:26
2
\$\begingroup\$

I Think HNQ Should Remain

For reasons that echo other answers: I think the pros outweigh the cons

Pro: It's good advertising for this stack inside the overall stack ecology. It is, in fact, how I found the site. (You are free to consider that a con, but I don't....) Although it's only prompted me to become and remain active on this stack, it's led to me taking a larger interest in several others and when my free time expands, I expect to be active in others.

Honestly, I scoffed at the site itself when HNQs first brought it to my attention. But after seeing several good questions, with good answers-- and finally one which I thought I could answer well-- that my mind was changed.

Pro: To a lesser degree, it may introduce people to the hobby. (I've been gaming for over 30 years, so this one isn't personal to me.)

Con: Occasionally we get trollish dogpiles.

But-- and this is key-- nothing the community and the mod teams have been unable to deal with. If I had metric evidence that this was becoming worse, in a way or at a rate that threatened to overwhelm the site, this would loom larger in my mind. But I don't, so it doesn't.

My considered opinion is that publicity is a double-edged blade, but that the community does a good job of sharpening the dulling the right edges.

A Modest Proposal

If that's all I had, I would vote accordingly but silently.

However, would it be useful/possible to auto-tag questions which become HNQs (either at that moment or after they leave that status) so that readers see that they were, in fact, HNQs?

(Indeed-- are they already? I took a quick glance at an active HNQ and they seem not to be.)

The idea is that if some subset of HNQs are troll-attractors, it might be useful to make people aware of this when they read through after the fact, and let that inform their responses. I am ambivalent about this idea-- I can probably be persuaded it's a good idea, a bad idea, or just a useless idea-- but I put in the public light for consideration.

\$\endgroup\$

You must log in to answer this question.

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