A case for why labels should be removed
EDIT: I'm only talking about the Exemplary tag
Greetings. I'm Alexis, and I signed up for this site back on May 30th, when it was just beginning. However, I have returned and I see something that doesn't seem like that big an issue, but I fear it will lead to the same sort of 'circlejerk' that Reddit has.
The issue is with this comment. At first, it doesn't really seem that bad - it is a person (we'll call them Adam) replying to someone else (Barbara) who says the subject matter (Charlie) is using white nationalism as a means to cause chaos. Adam takes this to mean Barbara supports Neo-Nazis, as Barbara states the following:
Hate to say it, but neo-Nazism is better than this, if by a small margin: at least its followers have an ideology.
Adam asks what "lofty ideology do you think Neo-Nazis" have, including a long rebuttal of this miscommunication.
The real problem with this is the labels. On the top of Adam's comment: "2x Exemplary". I have not seen any labels whatsoever anywhere else on this forum when I have been browsing it. On a miscommunicative post. Adam seems to claim Barbara is downplaying the issue, despite the miscommunication.
To see why this is an issue, let's take a second post discussing a network of paedophiles on Youtube. As on 1:14 PM EST, Feb 21, 2019, not a single label is there. This is in spite of communication being just as civil and developed.
My theory is that the Exemplary labels were not used as a "Well done" to Adam's post, but rather as a "We hate Neo-Nazis as well", or "super-vote". What I fear is that this will lead to an echo chamber where Devil's advocates, such as Barbara, are shunned for things from a simple miscommunication to having 'wrongthink' and defending ideas that people think should not be.
Let me be clear: I am not conservative or a Nazi. If anything, I am a socialist - but it shouldn't matter. You should be allowed to discuss the pros and cons of relevant political views as long as it is in a civil manner.
I think that thread is the only egregious use of the Exemplary label I've seen so far. Until now, every comment I've seen flagged like that has been within my expected realms of what an Exemplary comment would be.
That said, I don't think one instance is enough for a policy shift. If anything, it's an exception that proves the rule—that specific exchange was a rather unorthodox use of the label and its use there is striking as a result.
I'd propose a wait-and-see approach before anything else after this thread. Often, communicating concern and having a conversations about expectations is enough to adjust how things can play out.
Thank you for your response. I definitely see the truth in "Exception that proves the rule"; however, it may also spur more instances of the label being used as such. I would agree on a wait-and-see for now, but it's definitely important to nip things in the bud before they spiral out of control.
I think the community is pretty small and I think a lot of the members are here because they were looking for or (interested in) a community which was (in some way) different than larger and more popular platforms. In a sense, I believe many of us have this in common.
As tildes continues to grow we will likely start getting users who do not share this backstory with us and at some point these users may become the majority of the tildes users. I think the tools these users will have access to will shape the future of tildes. I am a fan of the labeling system in tildes, but I feel the exemplary tag is an easy target for misuse as long as the tag remains visible. I feel that we don't see many examples of misuse now because the users who would misuse it aren't here yet. I think it can be reasonably compared with gilding in Reddit. It's a tool that allows for manipulation of discussions and users can use it to try to "win" a discussion which will most likely result in lower quality discussions.
I think another example of misuse is here. The top comment which was at the top for a long time kept on getting exemplary labels. I think continuously labelling the top comment which already has a few exemplary tags is a misuse as it is just a way of keeping it the center of the discussion and killing off any other strings of comments. I think this is something that can easily happen in polarizing topics. If the labels we're invisible, I think this effect would be significantly minimized.
I wonder if perhaps it's less the label, more the counter that's the problem. What if the posts simply said 'exemplary' instead of 'exemplary x9' - without displaying the counting behavior, there's less incentive to bump it again, I think.
I think part of the design ethos on Tildes is to use words rather than symbols (not sure where that is mentioned). While it's possible for people to learn what the highlight color means, a word is easier to understand initially.
Also, relying on color alone to denote meaning doesn't follow accessibility best practices; I doubt it gets picked up by screen readers, and depending on the theme the different colors might not be easily distinguished by folks who are colorblind.
https://docs.tildes.net/technical-goals#use-words-not-icons
Thanks!
I like that idea. It's a good middle ground between the current state and making it invisible.
The labels remind me of Slashdot, it'll be interesting to see how they function at larger scale and whether a circle-jerk effect can be avoided.
It wouldn't be a bad policy to try and remove the labels in certain groups to see if there's a difference.
In addition to moderation with labels, Slashdot also has (or had, I haven't been active there for years) so-called meta-moderation where each user had to verify the accuracy of others' modded labels in past conversations. I'm not sure what is done on the backend with this information, but I suspect that users who assign appropriate labels tend to get more opportunities to do so.
Without a budget for labeling, it will take a tremendous amount of collective self discipline to avoid labels just being super/down votes.
I am sure I have encountered it be used so before, once, IIRC in an thread about Trump's wall.
Just to provide a data point, not to support or contest the parent comment.
I'm fairly certain that any label, regardless of purpose, will be used inappropriately by some.
Even when it's not being used inappropriately, there will absolutely be people who will misinterpret it's use because they got a different message out of the labeled post or discussion.
I think exemplary labels, for the most part, have done a good job at highlighting posts, especially when they are several deep into a comment chain, as particularly useful posts. Yes, they are sometimes used as a "super-vote" but anything with extra visibility will be used in this fashion simply because people are emotional beings and want to feel a connection with others over thoughts or ideas.
A good example of this is people reporting other users in video games for cheating because they didn't like them or something they said. It's a misuse of the system but makes a lot of sense when you consider that humans are emotional beings.
I'm one of the people who pushed for invisible labels back when visibility was being experimented with, and I still think
Exemplarylabels should be invisible to everyone but the person who labeled the comment, and the commenter. I want to experience every comment with an open mind. I don't want to know, right at the top before I even read the comment (which is something I thought we were trying to avoid here, with the new comment box at the bottom of threads and the number of votes being below the comment) that someone else thought it was a good comment. I want to judge that on my own, uninfluenced by a label.It's easy to say "well, I won't let that color my thoughts on the comment," but in all honesty, it's very hard not to pre-judge with that label right there, especially if it's
Exemplary x5, letting me know in advance "hey, a lot of people really liked this comment." Let me see what I think about it first, then I can label it appropriately. The label is already influencing the sort positively; why let it influence the reader?My time is valuable and I don't want to waste it reading a thread where the same idea is posted 14 times in different words. I'd rather read the exemplary comment (or alternatively some of the highest voted comments) and if I feel I have the time to devote to the thread, to read the rest.
I understand wanting to keep things "pure" but the reality is not everyone has the time, nor wants to spend the time to consume everything.
I personally think it enhances voting because it can draw attention to late posts that are better worded than earlier posts or perhaps expand upon the same premise in ways that are important.
I think he's got a point, though.
Those are really the only people who need to see the labels. That preserves the social feedback on a one-to-one level, which is a feel-good phenomena. What it doesn't do is trigger the bandwagoning effect. I'm not convinced this bandwagoning bit is a real problem yet when it's focused on positive things, expecially since Exemplary can only be used once every 8 hours... but it could be. I think larger levels of user activity will tell us.
That said, Exemplary's true purpose is to apply a 50% boost to the vote weight. If there's a comment that has 100 votes, and you hit it with Exemplary, it's now treated as if it's got 150 votes in the default 'relevance' sort (and has no effect on the other sorts, as far as I know). This will push it up the page ahead of other content, which is the goal. Each additional Exemplary label on that comment adds another 50 to the weight. Each new vote it gets is amplified by the number of labels.
They are basically a 'to the top with you!' tool.
My point exactly. Since it's already receiving that boost in the sort, it'll rise on its own. I see no need to influence the reader with such an obvious label above the comment, when we even place the vote count below the comment in the name of "don't vote for things you haven't read." I want to read comments uninfluenced by labels.
That's a good point. While I disagree with OP's request for removal, placing the label underneath does seem more ideal.
That would be better, since we can see vote counts already. I'd be up for that
There is also something everyone seems to be missing here IMO... if the exemplary labels weren't publicly visible how would any of us identify their misuse in the rare case that they are? And I think while everyone here has some good points that are absolutely valid and worth considering (even making me reconsider my own wholly positive view of exemplary comments), this is honestly the only real instance I have seen of exemplary being used in a particularly bad way.
I think the solution is not to do away with the entire system, throwing the baby out with the bathwater, but perhaps simply punish the people who misuse the exemplary label... and that punishment could even be as simple as posts like this, essentially calling attention to the misuse and putting social pressure demanding better.
In the long term, that would be a moderator's role: someone would have a "label-review" ability in the same way we currently have a "tag-editing" ability. They would see the labels (good or bad) applied to a comment, and judge whether the labels were applied appropriately.
Right now, we can't see 'Offtopic' or 'Noise' or 'Malice' labels, so we're unable to identify their misuse. I don't think it's up to people without a hypothetical "label-review" ability to review labels.
I know that right now Deimos can (and does) track and review almost all the label usage... but that definitely won't work at scale and allowing trusted users to peak behind the curtain with regards to label use was discussed as a potential solution. So yeah, perhaps exemplary also only being visible to trusted users (and the original commenter) might be the best of both worlds; reduce bandwagoning but also still allow for review of use/abuse. The only problem is that with Tildes model, where ideally a huge percentage of the groups active users will be trusted, I wonder if all the benefits of hiding the labels at all will largely evaporate.
I wasn't suggesting that we make 'Exemplary' labels invisible. I was pointing out that the reason they're visible is not so that everyone can identify whether they're being misused. They're visible so everyone can see that the comment is considered exemplary by at least one person - just like we can see the vote counts on a comment. The visibility is for general readers to know which comments are considered good.
Identifying the misuse of labels would apply to a subset of readers: the hypothetical "label-reviewers". For their purposes, it wouldn't matter whether a label is publicly visible or not because their special ability would allow them to see labels anyway.
'Exemplary' labels aren't being made publicly visible for us to identify their misuse. If that was the purpose for making these labels visible, then we should logically make all other labels visible as well, so that we can identify their misuse as well. But we don't - because that's not why they're visible. It's only this one label, and only for one reason, which has nothing to do with misuse.
If the exemplary is a highly rated comment on a chain which starts with a lowly rated comment I probably won't be drawn, visually, to the comment in the same way that an exemplary tag will draw me in. It's a good way to skim the content that exists lower in a page with a lot of replies.
Perhaps this means the vote label needs to be redone and more visible, but even on reddit I find myself often completely ignoring conversations below a certain vertical set point on a page that I don't find myself doing on Tildes precisely because the exemplary tag makes them visible.
On Reddit we actually noticed this problem: if a thread has several gilded comments which rise to the top then people will:
This is part of the reason why, after a while, the phenomenon of gilded comments with negative scores became such a sight: people would try and pre-emptively legitimize comments by gilding them, knowing that it tended to attract more positive attention.
No system is immune to abuse. I think it provides more good than bad and I think there are reasonable countermeasures to prevent this abuse. However Tildes is not at the same scale as Reddit so we will have to see how it plays out.
There will be growing pains.
We're kinda operating on the theory that just as 1% of the users submit almost all of the content, only 1% of the users are troublemakers. That means setting thresholds for things 'activating' at a number that 1% of the users just can't reach. When they gang up to get past that threshold, all they are doing is identifying themselves as banhammer targets, and make no mistake, Tildes is dead serious about banning people who fuck with the place. While it's invite only, that ban ends the problem. It'll probably stay invite only until there's a better solution. That could be never, it could be next year - we'll know when we find it.
Which is pretty much what one of mine said in the thread.
I think people understand the value the label grants.
I can't but also think back to Hubski and how badges were used there: purely, and simply, to signify what were considered valuable, insightful, and/or funny comments. (Some comments badged were also deeply-personal and relevant to the people of the small community in a very human way.) Anyone could've gotten 'em. There were, however, a handful of people who'd never leave the "most badged" list simply because of the volume of good comments they'd left. (In fairness, I'd read many of those, and there were good. Some of the best read on the Internet came from those.)
For anyone who wasn't that one person, getting a badge was a great moment. A few? "Holy shit!".
There was also this thing on Hubski called Zen Mode. It turned off many of the community feature visibility for the user: how many shares anyone's posts/comments got, whether anything had been badged, how much activity until the next badge etc. It helped with my prima ballerina syndrome, 'cause it helped manage expectation of attention vs. delivering quality content.
Maybe introduce something like that. Let people choose not to engage with some of the brighter, shinier social stuff, and let the rest not have to worry about a thing. It sounds like the conversation here is a matter of one's perception, rather than value.
I agree with your first paragraph, but the sort boost should largely take care of that. I'd also counter that non-
Exemplary-labeled comments are not necessarily a waste of your time, or you'd do very little reading here. Maybe we need a way to filter for Exemplary comments, or to hide their labels if there's enough differing opinion on their visibility.I think it's important to point out that the sort boost is often completely irrelevant for everything except top-level comments. If you're a bit deeper into a comment tree where there are only a couple of replies to a specific parent comment and one of them gets labeled as Exemplary, all the sort boost can do is put it above its "sibling" reply, and it was very likely to have already been above it purely from votes.
If the presence of the label isn't displayed for a comment like that, it effectively does absolutely nothing to the comment's visibility at all.
Yeah, I thought about that. I don't have an elegant solution other than to keep that thread un-collapsed if the vote scores of the comments above it warranted collapsing.
I still think think that label causes premature judgement of the comment, though, despite how much we might want to tell ourselves that we'll note it but not let it influence us—and I've honestly tried. Maybe there should just be a general indication that "hey, this comment has drawn a lot of interest" so that we can go in not knowing what others thought, positive or negative. I think avoiding hive mind voting is important, especially in the long term as Tildes grows.
Isn't it already done by the sorting algorithm? Once a comment is marked exemplary, it gets pushed to the top. And this way, one exemplary vote doesn't have higher value than 50 regular ones. I think it is not necessary to display the Exemplary label to anyone else than the person that labeled the comment and the author of the comment.
On the other hand, if a comment buried down as reply to 246'th comment is tagged exemplary, we certainly need some feature that makes it more visible.
This doesn't make any sense. The entire purpose of a community like Tildes is to be influenced. That's not a bad thing: you are here because you agree with the rules, and you wanna know what cool things other people are posting and enjoying. Why would you use a social network if your goal is to not be influenced by it?
I use Tildes because I wanna be influenced by it instead of Facebook, Google, Reddit etc.
From my comment:
Does that clarify? I'm not talking about how Tildes would or wouldn't influence me; I'm talking about not pre-judging a comment based on a label, be it positive or negative. There was a time on Tildes when all labels were visible, so you'd see, at the top of a comment,
Joke x5orNoise x3. Do you really think you'd go into that comment with an open mind? Or would you instead, maybe only mildly, think "well, this is probably either a joke or just noise." I want to remove that prejudice.I'm not going into any comment with an open mind, but, as long as the rules are fair, I don't see a problem with it.
Okay, that would explain that. And it is also good that it can only be used once per eight hours. However, I still disagree with it being visible, as it is even moreso a super-vote. It could definitely affect group-think, for those on the fence. "I'm not sure who I side, with, but this commenter has two Exemplary votes so he's right."
Yes, but the concept of a super-vote is what I am against. As if it is visible, it's an easier indicator of what people think are extremely good or just agreed with. You may have two people with fairly equal points for their sides, but when someone Exemplifies only one of them, you have a percieved bias in quality of statements. Like how gilding on Reddit is a multiplier for good opinions, which makes things escalate.
Then put the
Exemplarylabel below the post at the very least, like theVotecount. It's been explained thatVoteis below the comments so as not to prejudice the reader into forming an opinion of the comment before reading it, yet we pop a bigExemplaryabove the comment that pre-influences thoughts on that comment. "How could I disagree with something so many people found 'Exemplary,' using their once-every-eight-hours label?" As I said above, the label is already positively influencing the sort. There's no reason it should influence the reader, before they've even read the comment.Honestly I don't think moving the label down to the bottom is going to help if the nice blue highlight on the comment is still there - it's quite eye-catching, which is the point.
One thing that might work we haven't discussed - trust-based label visibility.
We could adopt a policy of only sharing the visibility of the labels that people are not directly involved in (giving or receiving) with users that have earned a higher trust level. I don't mean like 'level one' either - access to the labels after 7 days kinda is level one right now. This would be more like only showing them to people who have been around for a couple months or hit certain participation thresholds (even just voting a lot, lurkers matter). This could be limited to the groups they participate in as well, not sitewide.
We can't do any of that right now since it isn't coded yet, but that's part of building the trust system. The question is if there's value in that. Will the trusted users still succumb to the same behaviors after having been around that long and learning the norms? Is this a universal behavior we can't just solve by acclimation? I don't know.
I agree that moving it down doesn't really solve the issue, which is why I advocate for removing it altogether, and making it only visible to the person who labeled it and the person receiving the label.
Your trust-based idea is interesting, but I think the answer to your penultimate sentence is likely "yes."
I don't know if it would take up too much scroll space, but maybe it could be place at the bottom as well as the label? Just a horizontal bar across the bottom of the comment?
it could, but does it? i don't think so anymore than vote totals period tend to. i've seen plenty of comments including ones i myself have made get an exemplary despite only having, say, five votes just as there are plenty of comments which get plenty of votes and get several exemplaries. if there's a causal relationship between getting an exemplary and agreement as expressed by vote totals, it's pretty hit and miss from what i've observed. also, even your example i don't think really follows quite what you're arguing here: i've been paying attention to that thread and the reply to that comment has half the number of votes on it and counting, while the comment being replied to has basically not moved in vote total.
also, just speaking for myself here as one of the two people who labeled it exemplary, i labeled it that specifically because it simply but clearly demonstrated how the argument being made about how neo-nazism was 'better' than to try the style of terrorist attack described in the WaPo news article is somewhat ridiculous, given the (1) clear link to white supremacism and white nationalism by the plotter; (2) number of significant terrorist incidents recently which have been perpetrated by white supremacists or white nationalists and their adjacent ideologies; and (3) fact that some of those incidents have been just as disorganized and ideologically scattershot as the plot in the article. for me, it had nothing to do with anything resembling the point you're talking about in the OP post; and even then, i put in the exemplary form you have to fill out for why you're giving one in the first place that the point could have been made in a nicer manner. so, i dunno.
That highlight will also draw people to be more critical, and make corrections if they can. The added visibility just means higher levels of scrutiny, and I think if something is labeled that way and full of misinformation, someone will make a reply (especially as the place gets busier) with corrections and clarifications. Those are also likely to be hit with exemplary by people who also notice the problems.
What this is really doing is creating a highlighted comment chain, not just the top. Again, that'll be more common when it's more busy and there are more Exemplary tokens to hand out.
Off-topic, but...
It's frustrating for me when people use short URLs. If they use the full URL, that incorporates the title of the post ("many_updates_to_the_feature_formerly_known_as_comment_tagging" in this case), and I can read that title and remember what the post was without having to open it. But, with short URLs, I have to open the post to see what it was.
Minimalism is a common trend on Tildes. There are frequent proposals to remove or hide voting, sorting or any kind of distinction. This reminds me of Cassandra, the last earthling from Doctor Who who made so many plastic surgeries that removed everything that made it human. But, in the case of Tildes, the goal is to excise everything that made it Reddit. Yes, a network without sorting, tags and labels might fulfill some dream of purity, but it would also be very boring. Every once in a while things will go wrong, and we'll just have to deal with it. Tildes is opinionated, and that's a good thing.
Somewhat in the same vein, I think Tildes needs to be careful of becoming an "anti-reddit." I tried Voat and Imzy and I noticed a similar mindset among users. I don't think you can build a product or community simply by being the antithesis of that other thing. I'm sure many of us come from reddit so it's natural to compare and contrast the two. But for Tildes to be successful, it needs to be able to stand on its own.
I think the examplary label is an important part of tildes. Here's the definition as found in the documentation:
This looks to me like "exemplary" comments should stand out in some way--it'll help other people find which comments are the best out there, and helps clear out threads. Of course, the eight hour limit is helpful, too.
I think the color bar and tag should stay on the comment for sure, though I think it might be a good idea to move or remove the tally of exemplary tags a comment receives. As others have posed elsewhere in this topic, maybe move the label and tally to the bottom? That should help encourage people to form more of their own opinion about a topic before seeing how many people used a super vote on it.
I like having positive labels. Right now we have three or four ways (depending of how you count "joke") of labelling a comment as "not good" but only one way of saying "this is an especially good comment." I get what you mean about how the label can be misused but I haven't really seen that much so far. Even looking at the recent Bernie announcement thread where there are 100+ comments and plenty of political debate on controversial topics, the "exemplary" tag isn't very common there and it doesn't look like it's being misused.
"Insightful" sounds like something to celebrate.
It seems like your problem is with the exemplary label only. If you are not talking about the other labels, maybe add a clarification to your post. While I am not convinced that exemplary is bad, I do see the potential problems you have outlined.
The other labels are for soft moderation and I think they have worked quite well. Malice in particular is the closest there is to a report button.
Ye\ah, I didn't realize the other tags were invisible. I'll go do that now, thank you though.
I am fine with having an exemplary label, but I don't like that it affects the sorting of comments. I think of exemplary as a big thank you. But everyone should have an equal say in sorting comments. Exemplary messes with that, gives too much power in some, non random groups hands (I don't think everyone uses them).
You do. We do. Everyone does an equal say in sorting comments (well, everyone with an account more than 7 days old). We can all apply 'exemplary' labels to any comment we want to (once every 8 hours). This is open to everyone. If people choose not to use 'exemplary' labels, that's their prerogative. But that doesn't mean they don't have an equal say; it means they're voluntarily choosing not to say anything.
If someone wants to have more of a say, then they should use the tools provided to them: they should apply 'exemplary' labels to comments they think are exemplary.
Not really: if I use the tag, I have less power to affect the sorting than others for 8 hours. If I choose to use the tag strictly as intended, then I have less power than those who choose not to.
You can say it is the same with voting, and that'd be part correct, but also part no: I have to choose who to tag, given it is a limited resource, whereas a vote is discrete.
But everyone has the same restriction. It's supposed to be a limited resource. It's equally limited for everybody.
And, if you truly want to give an 'exemplary' label to a comment but can't because you're in that 8-hour window, then bookmark the comment and come back to it when your 8 hours is finished.
You get 1 'exemplary' label every 8 hours, just like everyone else. Equally. There's no favouritism, no bias, no extra power being given to some people and not others (except for people with accounts less than 7 days old). We all get the same number of 'exemplary' labels to use, at the same frequency.
I'm going to append this as a comment, however: The top 2 non-offical post, "Where are you from?", and "Reddit — one of the world's most popular websites — is trying to cash in through advertising" do not have any applied labels at of 1:22 PM either.
I believe those are from before labels were added.
They are, labels are a pretty recent (couple weeks) addition so you wouldn't expect to see them that much.
It definitely hasn't only been a couple weeks.
They were present when the site launched for a while, then disabled for a while, and then re-enabled on September 7, and updated in a few ways to more or less their current form on September 26. That's about 5 months now.
I thought someone told me that they had been added fairly recently but I guess not.
New users can't see the labels show up for seven days right now - call it an acclimation period. That may be what you heard. After 7 days the 'label' button appears next to vote and they can apply them however they see fit.