I honestly wonder sometimes how it’s physically possible to do what dang does, it seems like such an enormous and Sisyphean task. Sad to see twitter replies complaining about the moderation, because if it wasn’t for the moderation the quality here would be much lower, the moderation is the whole reason those people are attracted to participating here. Respect and many thanks indeed.

As someone that moderated a … pithy … community site, I am all too familiar with that type of complaint.

I would regularly get called a “nazi,” because I would step into catfights, and break them up, or “nip them in the bud.” Once you’ve been doing it a while, you learn to smell trouble brewing, and it can escalate rapidly.

Many folks chafe at restrictions, and they may actually be correct. They, themselves, may be entirely trustworthy; but others may not.

Generally, we’re all for rules and structure, as long as they apply to others (especially those with whom we disagree), but they shouldn’t apply to us.

I participate in HN, quite a bit. I basically avoid all other social media.

A big reason is the decorum, here. If it dissolves, then I will no longer participate.

I’ve been “danged” a few times. In some cases, I felt that it was not what I considered “reasonable,” but I accept the slaps anyway.

When in an Italian city, wear a bedsheet, and all that…

> Many folks chafe at restrictions, and they may actually be correct.

We're at least a couple decades beyond the point where a reasonable person could possibly entertain such a thought. Lack of restrictions always turns communities into cesspools. They don't necessarily die, but the people that hang out there are the ones that enjoy spending time in cesspools, and that's usually different from the vision the creator of the site had in mind.

There is however a big difference in the legitimacy of restrictions. The reality is that their are many unreasonable people on the Internet. Most unreasonable people however think of themselves as reasonable and other similarly unreasonable people as reasonable as well. There are also some reasonable people on the Internet. So how do you know which one you are?

When we can't assume one or the other we have to look at something else where we might reach a conclusion. If someone says they are oppressed we can look at what would hinder their oppression. Arguably this would be things like transparency, participation and equality. As those things make oppression harder.

So if someone says they are being silenced a reasonable person might look if there are multiple known moderators who can check each other, a moderation log to see how moderation is happening or some voting system to reflect the opinions of the community.

In most cases where people are says they are being silenced these things doesn't exist. As such it would be unreasonable to dismiss their opinion as there seems to be an argument for the restrictions not being legitimate. Otherwise you just create a different cesspool of unreasonable people.

It would basically be impossible for a site like HN to silence anyone. There are many other highly-visible places to break news even if dang wants to cover something up. It's not remotely the same as if the government were to impose restrictions on speech.

I don't think it is impossible. If I did I wouldn't have made the argument I did. I also didn't say anything about the government. The same situation applies to for example a company. A company that wants to find out how things are going is going to insist in multiple interests in important meetings, documentation on decisions taken and things like employee surveys.

Presumably you disagree, but then you disagree with the far majority of well regarded implementations in the real world. I am about to buy an apartment and it is the same thing. You need a witness, a contract and approval from the housing organization. Because that how you reasonably decide what happened.

And yes, people break the news all the time. For example on Twitter on this very post. That is the point. When they do how do you know who is reasonable? Which is what I have argued.

I only recall getting danged once and it was spot-on. I wasn’t acting like a full on flaming ass, but there was at least a glowing ember there and he phrased his response in a way that made it easy to accept.

One of the moderation techniques I don’t really understand is the “We detached this subthread” move. I don’t know what it accomplishes, besides narrowly targeting a (sometimes highly-upvoted) subcomment to specifically bury it. It’s harsher than even flagging because it nukes the whole subthread full of replies, too. Often no explanation. Just “We detached this subthread from [url]” and that’s that. Seems like something you'd only want to do to the most heinous of posts.

I recall having a +80 upvote or so reply “detached” in this way, but can’t find the link.

Yes, it's understandable that that moderation practice is kind of obscure and can be annoying. I wrote an explanation of it here: https://news.ycombinator.com/item?id=27132402.

I looked for posts that match your description and found these three:

It looks like we detached https://news.ycombinator.com/item?id=27058506 from https://news.ycombinator.com/item?id=27058435 because it was offtopic - and probably also several other of the reasons I listed at the above link.

Looks like we detached https://news.ycombinator.com/item?id=28695222 from https://news.ycombinator.com/item?id=28694236 for similar reasons.

Looks like we detached https://news.ycombinator.com/item?id=30749836 from https://news.ycombinator.com/item?id=30748070 for similar reasons.

In all 3 cases, the original parent was a massive subthread, probably at the top of the page. When child threads develop under top-heavy subthreads and cause discussion to drift in a generic direction, away from the original topic, that's a problem for thread quality. This is one of the most common times when we detach subthreads. All 3 examples look like classic cases to me.

Moderating top subthreads in this and other ways is probably the single biggest thing that moderators do here to promote thread quality. It makes a huge difference. We've learned a lot about this in the last few years.

By the way, none of this is necessarily a problem with the detached comment, nor does it imply anything bad about your (i.e. the child commenter's) intent. Often the child comment has been posted for exactly the right reason—thoughtful conversation. But if it takes a step away from the original topic in a generic direction, or a flameprone direction, upvotes and replies often start pouring in, skewing the thread as a whole in a way that makes it less interesting.

In other words, this is not just the fault of the child commenter (and maybe not at all). It's a co-creation along with all the other repliers and upvoters.

The more we learn, the clearer it gets that we need to moderate not specific posts, but subthreads. Subthreads are the unit of HN comment moderation. (Edit: doesn't that mean—someone will ask—that you could maliciously derail/destroy a subthread by adding bad comments to it? No, because we can detach those. That's case (2) in https://news.ycombinator.com/item?id=27132402.)

Understanding that online discussion gets less interesting as it gets more generic is one of the bedrock principles here. https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

Wow, thanks for taking the time to make that massive reply. This kind of Inside Baseball is facinating and really helps to take the mystery out of what goes on here. Surprised that this "tree balancing" really makes much of a difference.

EDIT: I wonder if trolls deliberately wait for and then target top main-threads, in order to maximize the blast radius of their trolling.

Lots of people do this without necessarily being trolls because they feel their comment would be lost, especially if some giant topcomment is dominating the discussion.

People also do the opposite - writing a toplevel comment to discuss (berate, more often than not) multiple other comments - 'I can't believe the nattering nabobs of negativism that have been commenting here, etc'.

In both cases, writing a good, non-repetitive, non-meta toplevel comment is probably better but it's more work and potentially unsatisfying - a lot of the time, a comment like that will in fact be lost in a big discussion.

As I've responded elsewhere in this subthread, a good top-level post can often succeed in gathering enough votes / mod attention to float up.

Probably not on the busiest posts (though those are something of a conversational lost cause already), but on most typical discussions without excessive comments, yes.

It’s amusing to me that this submission (from pg’s 2-year-old tweet) serves to make more work for dang on a Saturday.

Searching for "by:dang <typical moderation phrase>" is a handy way for seeing what actions are taken, in what circumstances, and why.

I'll admit to tailing dang's comments periodically, or searching for specific terms when people challenge / inquire about my own takes on comments. (Most often in the case of dupe / non-dupe article submissions, occasionally others.)

Initial conditions, both over the lifetime of a post, and of a reader's encounter with a post, both have a huge impact on discussion direction. Much of HN's moderation seems to operate with this in mind.

Clickbait titles are de-baited. Indirect links are disintermediated. Divergent distracting comments are downweighted.

Much of my own interaction with the HN mod team (ask dang, I email fairly frequently, probably a few times a week) are with regard to link or title issues.

I'll also occasionally try to steer a discussion back on track with a top-level comment I hope more directly addresses the post than many early takes do. This isn't always successful, but I've been surprised several times where my own late-in-the discussion comment ends up highly-placed in the thread. One key is to really fight back against expressing frustration (the comments are usually born from an excess of that), and just lay out a strong case for an alternative take on what seems significant.

I actually did express that frustration out loud in this thread ... and dang commented on the initial-conditions aspect:

https://news.ycombinator.com/item?id=26824383

On an older account I once had a factual comment detached whilst the other person his comment (which pretty much stated the polar opposite) stayed under the main post. Looking at which other comments stayed or got detached it was quite easy to see an enormous bias. Pointing that out, all I got was a vague “I understand you are passionate”. I think that’s the only time I’ve ever seen him miss his stride, though.

Link?

> Once you’ve been doing it a while, you learn to smell trouble brewing, and it can escalate rapidly.

I'm not actually proposing this, but that makes me curious how well that could be modeled via deep learning.

I run an online community, with my own developed software. We use lots of different tools to prevent spam, harassment, etc. We use different APIS to check IP, ASN, email for previous spam behaviour, use of disposable email domains, etc.

We have a team of moderators and as mentioned before you can "smell" something brewing. The words, to whom they are directed, etc.

A couple of years ago we started using Perspective API [0], a Google service "Using machine learning to reduce toxicity online". We use the toxicity model, trained by Google and The NY Times. Basically we submit the text of posts and replies and receive a score. We do some account history calculations and if it's past a certain threshold then moderators receive a notification - we can't read all posts in our platform all the time but this helps a lot.

The community itself helps too, and the reporting tool is used when things happen that we aren't aware.

[0] https://www.perspectiveapi.com/#/home

There are a lot of projects out there that are attempting to do this because of the impracticability of a human team to deal with the entire torrent of information.

For example https://playvalorant.com/en-us/news/announcements/valorant-v...

> As part of a larger effort to combat disruptive behavior, Riot Games recently updated its Privacy Notice and Terms of Service to allow us to record and evaluate in-game voice communications when a report for that type of behavior is submitted—with the goal of kicking this off in VALORANT first.

I believe it would be quite possible. Text is a bit dryer to work with but has the advantage of being all text.

I'm pretty sure that's exactly what FB and Twitter are trying to do.

Not sure how well it's working out.

I hear constant stories about how "blunt instrument" these are, and how it's impossible to appeal.

I wonder if this is a sign that they need more moderators. Given how much moderation is done by the community, I'm surprised dang needs to work as hard as he seems to.

i wonder what tool dang uses to see comment replies and respond ?

the ycombinator web UI does not encourage real-time or even near-time responses.

To my knowledge, there is a combination of specific views on the site that highlight important features of the site (an obvious example might be the list of flagged comments; a less-obvious example might be https://news.ycombinator.com/topcomments which shows the top-most comment of each front page item) as well as email alerts from a variety of sources, some of those being user-driven.

in order to read your reply to my comment, i had to :

- click on my user name

- click on comments

- search/scroll to this thread and see your reply

I don't get notified of comment replies like on reddit. there is no slack/email page of DMs/PMs.

You can just click the 'threads' link at the top left and it'll show you your comments and nested replies.

I had no idea about this until someone mentioned it. Prior to that I couldn't be bothered very often to check my comments for replies. It was too many steps and made me feel narcissistic for doing all that just to see.

edit- spelling.

You can refresh your replies list over time, or you can sign up for Notify HN or any similar service that notifies you via email.

I like to think he's some kind of monk, a zen master of human forums, and a genuinely good person. At least any exchange I had with him or observed with all kinds of people have been full of humanity, while firm in his mission. I don't know how he does it, and I wish we'd celebrate him and people like him far more.

I posted a link elsewhere in this thread to this article [1] - The lonely work of moderating Hacker News (2019) - and you might like its author's insights/views, plus there are plentiful quotes from Dang and his colleague, Scott Bell. [edit to add: a comment in reply to this says that Scott no longer works at HN. Surely Dang can't be doing this solo nowadays?]

[1] https://www.newyorker.com/news/letter-from-silicon-valley/th...

It's just Dan now, I think. Scott no longer works for HN.

Thanks for this. It makes me hopeful that we can find people and a process to keep things civil and somehow engaging in such a forum.

Sure HN has its pains (and I'm an unconditional of the late n-gate site) but they don't come from moderation shenanigans.

I also appreciate the pedagogy I often see. Some participants here have never been actively moderated and dang takes the time to explain as if the moderated person was one of the 10000 of the day.

I've always just hoped that he's not just one person.

Echo'd.

Respect and many thanks, dang.

Paul really should fund an AI startup focused on chatboard moderation based solely on learnings-from-dang.

Is the one thing we’ve learned that you need multiple human (or one dang) mods for a successful social media sites?

Yep, dang is amazing.

I get a very different sense from the reading of his comments to users; curt and inconsiderate. He tends to make declarations about behavior that sound objective but are really just his very casual read of a situation, and there’s no appeal or redress possible.

Additionally, he tends to favor the original aggressor, ignoring the concept of instigation to fix the proximate cause of the negative interactions here. Don’t get caught punching second, in other words.

It’s classic utilitarianism; if 1% of users get shat upon by dang, and 99% in any way benefit from his negative behavior, I suppose the folks here would be alright with that?

> Additionally, he tends to favor the original aggressor, ignoring the concept of instigation to fix the proximate cause of the negative interactions here. Don’t get caught punching second, in other words.

Can you give an example? I just went through dang's comments in the last few days and found nothing of the sort, and it has not been my experience at all, unless posting an unpopular opinion counts as instigation. Dang's pattern, every time I've been witness to his moderation, is to cut the first comment that descends into incivility or flaming.

Conversations on HN sometimes get heated. Some people take people disagreeing with them as a personal attack and quickly descend into actual personal attacks in response. Those people may feel like they're being singled out by dang, but it's because they actually are misbehaving.

Look at his post history, you can see plenty of examples.

A pattern he follows very commonly is seen here [0]: "Can you stop doing $thing?" However, it's dang's opinion that $thing is occurring, and he makes zero effort to understand why $thing could even be happening, and further entertains zero challenges to his interpretation of what has taken place. It's, "stop doing $thing, which is definitely happening, or else I will ban you."

Here's another good example of dang handling moderation poorly [1]: he spends zero time explaining the situation, zero time listening to the perspective of the "accused", and simply, curtly, dismisses that person from the site.

Ironically, dang's intervention is often the most toxic behavior in a comment chain, and these examples are what I found in just the past two hours.

There are thousands, or more, people treated this way, and instead of concern, praise is constantly dumped on him, like what he's doing is some incredible or even desirable behavior! The assumption is the site is better off without these people he treats poorly on it, but I doubt any rigor has been placed on that assumption whatsoever.

[0] https://news.ycombinator.com/item?id=32039759

[1] https://news.ycombinator.com/item?id=32039501

Your complaints are totally unreasonable and sound ridiculous to me. "dang's intervention is often the most toxic behavior in a comment chain" is laughable. If you look more into dang's comment history, he has often explained at length about the kind of issues you link to. Often his comments link to a whole lot of previous long explanations. I've spent hours down those rabbit holes, and it's the most impressive, insightful writing I've seen in my years on HN.

I was 'danged' once, after I angrily responded to someone who made an outrageous comment to me, attacking me strongly for saying cheating was bad. Dang objected to my comment. I said, what the other guy said was much worse! Dang hadn't even read the other guy's comment, which seemed so unfair.

He explained to me that it doesn't matter what the other guy said. "Punching second" or in any order, is not ok on here. Read the comment guidelines. He said that being on HN is a continual exercise is not being triggered, in not commenting while angry. They were wise words which I think greatly improved my HN comments, or at least I don't do those triggered, snarky, cutting responses any more, or almost never. No matter what the other guy said. Silence is often the best response, or the only good one. That someone else is being an asshole on here is no reason to join them.

You are apparently fairly new here, and seem to have little idea of the massive number of comments written on HN each day. I looked at the first couple of pages of your recent comments, and most of them have a very unpleasant, hostile tone. You are in no position to criticize dang. Read the "In comments" section of the guidelines, the part starting "Be kind. Don't be snarky." A lot of your comments break them.

https://news.ycombinator.com/newsguidelines.html

Haha, your comment is exactly the comment I'd expect to get moderated, but won't because it's long and dang won't read it (and he is spread too thin).

Any comment citing "newguidelines.html" should probably get flagged for review, for example.

I also am not new here, but imagine why I might seem to be...

It should be easy for you to link to examples substantiating this?