70

I was browsing the flags queue today when I came across the following answer to the following question:

What is Trump's plan to get the United States out of trillions in debt?

enter image description here

That answer is nothing more than a sarcastic comment. It can be considered the quintessential example of a low quality answer.

It also has 11 upvotes.

This is not an isolated incident. It may be the most extreme example of this phenomenon that I've seen in recent memory, but low-quality posts get rewarded quite frequently, and this undermines the purpose of the voting system, which is to encourage quality posts, and to discourage noise.

Please don't up-vote low quality posts. Downvote them, and upvote the good ones. I can't force you to use your vote in a particular way. I can only encourage you to vote responsibly, and I hope that you do with the overall quality of the site in mind.

  • 9
    This Meta post is posted as a question, but it is rather an announcement. IMO, if it were re-formulated as a question, it became much more constructive. E.g., "What practical steps we can make to encourage users not to upvote the bad questions?" — and I have an answer then; we, the frequent visitors, must take the lead and not only downvote often, but also accompany our downvotes with comments telling why did we do that, and what's wrong with each post we think is bad. – bytebuster Jun 3 '17 at 3:01
  • 1
    … This would help elaborate the community consensus and help the "newbies" learn it. Keep in mind, new users can't even see the votes count and so they don't even know the post has received the downvotes (unless the total score is negative). – bytebuster Jun 3 '17 at 3:01
  • 8
    The problem here is probably that this was in Hot Network Questions. HNQ voters don't really know what the site standards are and posting in Meta won't help them learn. They aren't Politics.SE readers; they're from other stacks. We'd have to post in meta.SO, etc. instead. – Brythan Jun 3 '17 at 3:16
  • 6
    @Brythan In this case that's probably true, but I noticed similar issues in answers that definitely aren't HNQs. I think people here are likelier to upvote answers that match their political opinion, which is fine as long as the answer isn't poor quality. But it also happens with answers that are just unsupported rants or snarky comments, which does seem like a problem (one I don't think we can solve, except by editing or deleting such answers). – tim Jun 3 '17 at 10:20
  • It seems to me that this is an issue for the moderators to handle. Aggressive deletion of sarcastic non-answers will always be necessary, and when you're on HNQ you probably need a moderator to check every half hour at least if you want things to stay reasonable. – Joe Jun 9 '17 at 18:20
  • 2
    @Joe One of the core ideas of SE is that quality is community-moderated. That means that individual people control their own upvotes and downvotes and apply that pressure accordingly. As a moderator, I only delete questions and answers with serious quality issues or rule-breaking. I cannot control how people vote. This particular answer was delete-worthy, but not all low-quality answers are delete worthy. Sometimes down-vote is the correct course of action, but a critical mass of people need to do it well order for that to work. – Sam I am Jun 10 '17 at 16:52
  • @SamIam I am well aware of the community moderation principle and have had many years experience with it. However, there are times that cannot succeed - and trolling on HNQs is one of them. That's why moderators are chosen and are given the power to unilaterally delete answers, in large part. – Joe Jun 10 '17 at 18:19
  • 2
    @Joe The phenomenon is more widespread than just trolls on the HNQs. There are many cases of moderately low-quality answers that get rewarded with more upvotes than downvotes. It seems to me as though there's a culture here where people are upvoting and not downvoting out of politeness, rather than judging the quality of the question or answer. – Sam I am Jun 10 '17 at 19:06
  • this phenomenon... likely people getting a laugh out of something and simply giving it kudos with an upvote.... lol!!! This is not the proper forum for jokes.... it's politics.... ha!! Save the political jokes for jokes SE, right! – President Bernie Sanders Dec 11 '17 at 23:50
25

I was baffled by the comments and I'm glad I'm not alone

The question asks for Trump's plan, which doesn't exist, even in the most generous hypothetical sense. So this "answer," based on Trump's actual demonstrated business management model, seems as valid as any that can be offered as an explanation for something that does not exist.

And

I think that it is an answer. If we are to judge by Trump's past behavior, that is exactly what he does: BS people into investing in some project, rake off some of the money for himself, then declare bankruptcy and leave the other investors & lendors short.


This is Politics.SE and people can, and do, vote their political inclinations. But it's a cheap shot and not really an answer (both comments are a stretch in trying to make it such). The sad thing is he could easily have made it an actual answer (it's not unreasonable to assert something along those lines) but his comment made it clear it wasn't a serious answer. I was pleased that most reviewers were treating it correctly, tho.

Don't vote because of how you feel about Trump. Vote on the merits of the question or answer.

  • 8
    How can anyone review that as "Looks OK"?! :-/ – user11249 Jun 3 '17 at 16:00
  • 2
    "...it's not unreasonable to assert something along those lines) but his comment made it clear it wasn't a serious answer..." You mostly seem to complain about style and less about content, is that right? – Trilarion Jun 7 '17 at 7:19
  • 13
    It's not a question of style, but of substance. The post is a joke meant for an audience, playing on the fact that Trump actually did that in his private dealings. We know it's a joke because of the punchline He's a successful businessman (insert rimshot here). There's no way to take it seriously, and the comment the poster left made it clear it wasn't meant to be. – Machavity Jun 7 '17 at 12:14
20

People upvoting non-informative answers just because they agree with the conclusion is hardly a new phenomenon on this site.

The reverse is also true, by the way. People downvote informative answers because they disagree with the political topic the question is about. I believe that almost every single one of my answers on this site has attracted at least one downvote. I don't mind as such, or begrudge anyone for downvoting my content, but it certainly is a different experience from other SE sites.

Posts such as Sam's "please don't do it" are not likely to be very effective. This is not the first such post, and probably won't be the last.

The problem is ... I can't come up with any good ways to fix this, and it seems that no one else has either. I fear that at least for the time being we'll have to accept the site as-is, warts and all.

  • 2
    I completely agree with this point, but remember we are dealing with politics, which is literally the most subjective topic on SE – SleepingGod Jun 3 '17 at 16:52
  • 3
    "The problem is ... I can't come up with any good ways to fix this" The subjective voting, we can probably not fix. We just have to trust that the majority of voters vote reasonably. But we could be very strict with the guidelines that define a low quality answer or question and delete/close those. That may reduce the problem somewhat. – Trilarion Jun 7 '17 at 7:22
  • 1
    This topic is a very brave experiment, but I don't have a better suggestion other than we should expect it. It is really hard to be objective and impassionate on a subject that demands passion and partisanship in order to be informed. – Frank Cedeno Jun 16 '17 at 15:12
8

This issue permeates the site. Almost every question has a string of comments that are (at best) discussion, rather than suggestions for improvement or clarifying questions.

Many top users participate in these discussions, side arguments, displays of affiliation or viewpoint, etc.

If that is the kind of community we want to be, these kinds of answers and the support for them is a side effect.

If we don't want to be that kind of community, we need clearer guidelines, consistent treatment of chatter, and consistent behaviour from the top users.

  • 2
    We already have clear guidelines, tho. Answers are for answers. If you want to be snarky, clever, disagreeable, etc, that's what comments are for. The reason this guy posted it as an answer is because he was a new user and didn't have the rep to comment. – Machavity Jun 6 '17 at 12:50
  • this x 8000 (yep i know i am a problem user... but when in rome) – SoylentGray Jun 6 '17 at 20:39
7

I feel this point is moot, the answer was probably already down-voted by the people who read the meta i.e Politics.SE users. What probably occurred was this question started trending on the main Stack Exchange site, where random users from other SE sites came along and used their Association Bonus to be able to vote up the question. Which was compensated to some degree by normal politics.SE users down-voting it.

It is worth noting that I personally tend not to down-vote answers which are obviously up for deletion straight away (i.e some random first post which is obnoxiously racist and unsourced), preferring to flag it first and wait an hour or so before down-voting if it's still up.

  • 2
    While the "Hot Network Questions effect" probably didn't help, I feel it's not enough to explain the votes. I see plenty of really bad answers get voted on "the wrong way" (either up or down), and not all of them appeared in in HNQ. – user11249 Jun 3 '17 at 16:48
  • I think that's a common problem across all SE sites. What happens is that trending questions bring in all sorts of people not particular interested or informed about the particular SE's theme. So you get a lot of up-votes for things that are sometimes just plain incorrect. – user1530 Aug 9 '17 at 21:59
4

The voting mechanism (ironically) doesn't work too well on politics.se, because the questions are more "opinion-based" (in SE parlance) or based on evaluations and hard-to-adjust world-views than the more technical issues on other sites in the network.

People will therefore to some extent vote according to their political beliefs. This can undermine the quality of the answers if the politically inclined voting exceeds the quality-based voting. I realize that "quality" is essentially contested on a political site, but there are some non-political (if not objective) indicators, such as whether the answer addresses the question, whether it is soundly argued, and whether it is appropriately referenced.

The question then is how this can be fixed. I don't believe appeals on meta will help. Appeals to the common ground (everyone wants good answers) are reasonable, but voting is often rather impulsive than reflective; few users read meta; and "trending" questions attract even more opinionated and less informed users from other SE sites.

This leaves much responsibility in the hands of the moderators. Perhaps then politics.se needs more moderator intervention than other SE sites. This includes stricter enforcement of quality standards, e.g.

  • regularly pointing out (using boilerplate comments) what the quality standards are
  • editing out opinion-based and unnecessarily incisive parts rather than leaving a comment (e.g. anything that violates "be nice")
  • lower threshold (compared to other sites) before deleting answers and before moving comments to chat

It might also include tightening the quality standards for example by

  • generally requiring references (like on sceptics.se)
  • a rephrasing of "be nice" that discourages dismissive tone ("nonsense") or moralistic boo-words ("bleeding heart")
  • encouraging answers that discuss multiple perspectives or are explicit about their evaluations or both

Please not that I'm not trying to tell the moderators what their job is (thank you for you time and work!); I'm just submitting some ideas how the slightly disfunctional voting mechanism can be compensated for.

  • I myself have found such votes here irksome, but not to the point of undermining the system. It's not unfathomable that could happen though, at least on some topics. – Fizz Jul 11 '18 at 3:05
4

Perhaps one thing that could be done about this would be adding a rep limit for upvotes to mirror the limit on downvotes for this particular stack. That is, require 125 rep to upvote, so that one can't upvote from just the association bonus.

Part of the skew to the voting comes down to the fact that people who have never earned any rep at all on this stack (whether they come from HNQ or elsewhere) can upvote, but can't downvote. While I think that's generally ok for most stacks, I do think it contributes to this problem here (and I've also seen it be a rather large problem on Skeptics.)

A different possible solution would be to not include the association bonus when determining if someone can upvote on an HNQ question, just as is already done for answering a protected question. One easy way to implement that would be to make the association bonus not count for either answering or upvoting on protected questions and then automatically protect any question that hits HNQ for the duration of its time on HNQ. Actually, I really like that idea, as it would also cut out a lot of crap answers.

Most of my SE time is spent on Aviation and, even there where most stuff is pretty non-controversial, we end up having to protect a lot of the questions that hit HNQ. I don't think much would be lost on most sites by just automatically protecting HNQ questions.

  • Fully agree though this most certainly IS a problem on most stacks, not just here. Unfortunately, SE chose to summarily ignore any requests to fix this problem, including my proposal: meta.stackexchange.com/questions/238420/… – user4012 Aug 6 '17 at 11:48
  • 2
    @user4012 Yeah, it's not unique to this site, but it is worse on sites where more controversial topics are discussed. For example, at Skeptics, an upvote normally means "This is a well-researched, well-sourced answer and sound logic." When it hits HNQ, however, that tends to change to "This answer may not actually be supported by any facts at all, but it supports my general world views and biases." It seems that the same is true here. – reirab Aug 6 '17 at 17:55
  • I doubt this will be accepted by the powers that be. Without even being able to upvote, HNQ will mostly be a "spectator sport" for a large number of people. Even without downvotes, selective (enough) upvotes sort answers. In fact that's probably the main filtering mechanism of SE, downvoting is a secondary (and non-zero cost one). It was clearly meant to be that way... so any attempt to change those fundamentals seems unlikely to garner much traction from management. I do agree that Skeptics bears a similarity in some respects. – Fizz Jul 11 '18 at 3:02
  • @Fizz It's true that selective upvoting can (and does) sort answers. Where the problem comes in with HNQ is that the answers on controversial topics such as here and on Sceptics then tend to get sorted by how many people's world view and biases the answers agree with rather than the actual quality of the answer. Your concerns are a good point, though. Perhaps that could be resolved by only making the "association bonus doesn't apply to upvoting" rule apply to the more controversial sites while making the "HNQs get auto-protected" rule apply to all sites. – reirab Jul 11 '18 at 4:32
  • Actually I've seen a pretty eggerious answer more recently, so the problem here is probably more serious than I had thought... politics.stackexchange.com/questions/11168/… – Fizz Jul 11 '18 at 4:44
2

Removing the site from HNQ is the only legitimately impactful thing that can be done to improve this (it won't eliminate the problem wholesale but would at least scale it to smaller and much more manageable directions).

You must log in to answer this question.

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