This article made me realize, part of the quality of the conversation on HN comes from the ability to downvote.
I’ve never thought about this before, but one of the effects of Twitter and Facebook only having positive reinforcement is that you don’t get the “disapproving stare.” It’s possible for a person to “like” with some degree of anonymity and no effort, but not possible to do the reverse. Worse, since the Algorithms heavily bias content based on “likes,” there’s no recognition of content that is actually broadly liked versus content that is polar.
Also, for individual posters in places with downvoting, the little dopamine hit you get from upvotes is heavily countered by the bad feeling of downvotes.
I know that in my own participation on HN I will occasionally think of a snarky thing to say, but usually don’t post it because I know it’ll get downvoted, which I don’t like, and that moment of hesitation is usually enough for me to think “yeah that is actually unhelpful and not worth posting.”
This all probably very obvious to most of this crowd, but I honestly haven’t considered it before. How different would the world be if Facebook and Twitter merely had dislike buttons to go along with likes?
Most high quality communities have active moderation. And many times, the vast majority of moderated activity is as simple as the mod privately telling a member to chill out or adjust their attitude. Don't need bans or downvotes even, most people will happily comply with a mod request simply because the request was made.
This is a striking point. In most "real life" communities, there will be someone who gets the final word on whether something said is acceptable. Think mom or dad at the dinner table, the CEO at a private company, or the principal at a high school. They are effectively the moderators of the public forum they physically inhabit.
We do, however, structure this moderation as a hierarchy and it self organizes. The mom has the final say on acceptability, but in her absence, the older sister takes this role. We rarely see this structure in online communities. Should downvotes be weighted by the karma of the voter?
> Should downvotes be weighted by the karma of the voter?
While a good idea in theory, I'd hate for people to have yet another incentive to want "good karma". Would prefer tiered levels of moderation where long standing community members get promoted to 1st-level moderators. Their actions get moderated by the next action, etc. etc.
I remember being an @op at an IRC trivia channel when I was ~13 years old or so, back when we had dial-up. I played in it daily, for a lot of time during weekends, was nice to everyone, gave tips to newbies, hung around. Was asked to join the team, as a "level 5" op. Loved the "behind the scenes" so much, the op community, fell in platonic love with a chick 14 years older than me (actually met her IRL 6 hours away by car about a year later, so cringe!), kicked trolls and spammers... then later as I learned mIRC scripting helped write a new trivia script, a bunch of tools, etc. Ended up being "promoted" to level 10, started training new operators, then 20, started making longer plans for the channel, then 50, 100, then "founder" status. All of a sudden I was calling shots at 15 years of age, with ~500-1000 concurrent users in the channel during peak activity. That experience of ensuring people were having fun, cultivating a community of friends among users and operators, teaching new people how to be moderators, etc. was incredibly rewarding and left such a huge imprint in my life. There were no points to be earned, no karma to show anyone... You wouldn't even know the channel unless you were on BrasNet (unlikely!) but I was eager to contribute everyday just because there was a real community behind it.
That sort of community is all but gone from the internet with its fake internet points
You can say the OP status was a points system of sorts, but not really... Users didn't know my level, and you couldn't click on my username to see how "big" I was. Most importantly, one couldn't click on one's own nickname and get a dopamine hit from seeing a number next to it like one can when their post makes the front page of HN, reddit or any other platform, or when they get 100 likes on FB an IG
I think this is right, but would add that it is also key that the vast majority of here have a lot of trust in YC + moderators and nearly universally view them as good faith actors. This wouldn’t work on CNN.com.
FWIW I also understand how valuable moderation is and I think it is more impactful than downvoting. I just had never considered before that downvoting was, at least potentially, a valuable tool to increase civility.
I've come to the opposite conclusion about downvoting. To me, downvoting is the most useful to people with an agenda to push. They can "push" the agenda by both upvoting posts that support it and downvoting posts that are against it, regardless of post quality. Either action is equally beneficial for the narrative. It essentially gives people two upvotes. Normal users, not trying to push an agenda, do not have this "two upvote" advantage, since they are generally not upvoting/downvoting content in a narrative-consistent way.
If you take that view, then it appears downvoting is much more potent than upvoting. An upvoted post would at most rise to the top of its tree, but a downvoted post will eventually disappear.
I think with hn, it's more than just the downvote, it's the subtle greying out of the comment each vote below zero until finally being killed at -4.
This is a bit different than say reddit where downvotes can just keep going on and on and a comment with -120 votes is just as visible as a comment with -1 and only a quick click away from being as visible as positively voted comments.
Hn's comment greying reminds me more of just silently being shut out or ignored and gives you immediate visual feedback as to the relevance, or lack thereof, of what you said.
I see the value in having a limit to down votes, but that shouldn't elminate the comment. Like every community HN has norms. If you have an opinion that's not popular - valid or not - it can get down voted. That doesn't mean the voice shouldn't be heard.
Mind you, if something is a violation then let the moderators handle it. But there is a difference between unpopular and violation, even on HN.
> If you have an opinion that's not popular - valid or not - it can get down voted. That doesn't mean the voice shouldn't be heard.
HN does have the "vouch" mechanism to compensate at least somewhat for this. Someone with enough karma to vouch can give a downvoted post that they think deserves to be heard a second chance.
That and there is always show dead mode for those that have an account and wish to see those comments. They're just generally hidden from casual browsers and must be turned on by those who want to see them.
I'm much more likely to upvote a post if it's grey, and much less likely to downvote it. I think you'd see a lot more downvoting if you removed the grey effect.
In my personal experience, the Reddit comment-hiding system only gets triggered for particularly horrible comments, not just controversial ones. It must take into account the upvote-to-downvote ratio.
There's always going to be, subjectively, a lot of overlap between those two categories in looking at the comments of others.
I know I'm not a troll, so obviously if someone downvotes me it's because they disagree and therefore unfair. But maybe not from their perspective.
You can also downvote and reply; I don't know how common that is. But given the number of people who say they don't know why they are downvoted, it seems to me like a reasonable thing to do.
Yes. "Look at this cool thing MS/Google is doing!" and "Look at this evil thing MS/Google is doing!" can both be seen as trolly agenda comments and relevant ones, depending on where you stand.
Would be cool if "downvote"(for relevance)and "disagree" would be mutually exclusive (either/or).
You either decide this post is not relevant to the discussion at all or you disagree with the argument.
"Upvote" for relevance and "disagree" could be combined.
You decide this post is relevant to the discussion but you disagree with the argument.
Of course everything depends on users using the tools in good faith. But it would be an interesting approach to decouple "downvoting" from "disagreement".
> This article made me realize, part of the quality of the conversation on HN comes from the ability to downvote.
It’s a little more nuanced than that. The downvote alone gave rise to the “Bury Brigades” and the ensuing Groupthink that killed Digg. HN used to suffer the same problem, though not fatally so, and I think @dang has made some code mods to curtail that. Plus his awesome efforts as our moderator.
Reddit has downvotes as well and its conversations are uniformly bad. Slate Star Codex has a creaky old antique of a comment system without voting at all, but its conversations were of the highest quality. So are LWN's. I don't think voting mechanics affect discussion quality as much as you think. What matters a lot more is a sense of mutual respect, shared purpose, and, honestly, a certain level of average intellectual horsepower.
I'd argue that it's the norms that a community develops.
LWN and SSC had excellent discussion because their respective communities focus on both promoting quality and policing bad behavior -- go off the rails, and you'll get a bunch of folks trying to help you get back on.
I'm a huge fan of Scott's writing, but the forum discussions at SSC did not seem particularly good (for any value of good - niceness, erudition, signal-to-noise, etc) to me. Certainly they were much worse than HN on average.
Maybe they seemed especially bad next to Scott's essays which were all very well thought out in comparison. I do agree though that I've found them more disappointing than HN comments on a pg essay.
If the community is bad, it doesn't matter what tools you use, people are going to write and upvote bad stuff, and sometimes downvote good stuff if such option exists, flagging will be misused, etc. There is no way to conjure a good discussion out of bad actors.
On the other hand, any community will contain a few bad actors. If not now, then definitely when the community becomes famous. The good guys may try to stop them, but it depends on user interface how comparatively easy it is to wreak havoc vs police the behavior. If the good members outnumber the bad ones 10:1, but it is 100 times easier to post bad stuff than to remove it, the website is doomed. Ideally, fixing problems should be so easy that moderators won't get tired doing it.
Related problem is that an online community cannot work on pure democracy, because it is simple to create sockpuppet accounts, and also unless the community is huge, it can be easily overrun by hordes of real people coming e.g. from Twitter. Again, a reasonable user interface will assign greater powers to established users over the newcomers. But it needs some metric to do so, for example the total karma gained. Again, if the core members are bad, they will abuse the rules for petty reasons.
When people complain about some features of the interface, typically about downvoting or even karma as such, they usually correctly point out how the feature can be abused, but forget to consider how a lack of the feature could be abused. For example, downvotes can be used to suppress dissent, and karma can lead to hive minds. However, websites without downvotes are easy to spam (whether by literal ads, or just a small group of people pushing their agenda everywhere, upvoting each other). The standard solution to spam is blocking, but then we have gone full circle, because blocking can also be used to suppress dissent, except it is done per user instead of per comment. (A good website needs both mechanisms. Sometimes there is a bad comment from otherwise good user, but also some users consistently post lots of bad stuff.)
Also, ultimately there is no way to build an online community that would make everyone happy, even if we exclude the obvious trolls. People have different opinions, duh. Most people don't like the environments where the opposite opinion is a norm; many dislike even when the opinions are balanced; and for some even hearing the opposite opinion, no matter how small the minority that supports it, is beyond the pale. Deep in our hearts, most of us want to live in a bubble; we just don't want to retreat to the bubble and give up the places we enjoy, we would rather prefer the bubble to grow and swallow the places we enjoy. We all can't have it this way, so someone is going to complain no matter what.
I’ve never thought about this before, but one of the effects of Twitter and Facebook only having positive reinforcement is that you don’t get the “disapproving stare.” It’s possible for a person to “like” with some degree of anonymity and no effort, but not possible to do the reverse. Worse, since the Algorithms heavily bias content based on “likes,” there’s no recognition of content that is actually broadly liked versus content that is polar.
Also, for individual posters in places with downvoting, the little dopamine hit you get from upvotes is heavily countered by the bad feeling of downvotes.
I know that in my own participation on HN I will occasionally think of a snarky thing to say, but usually don’t post it because I know it’ll get downvoted, which I don’t like, and that moment of hesitation is usually enough for me to think “yeah that is actually unhelpful and not worth posting.”
This all probably very obvious to most of this crowd, but I honestly haven’t considered it before. How different would the world be if Facebook and Twitter merely had dislike buttons to go along with likes?