Thank goodness we don't have downmods on HN

74 points by raganwald ↗ HN
From time to time I have wanted a downmod button. Today there was some linkbait on the front page--something about real programmers not using abstractions or what-not--and my first thought was that the best thing to do is to ignore these posts.

And then it hit me: downmodding is not ignoring. Neither is upmodding everything else. Sure, there's some imaginary value of pushing it off the front page quickly, but if people naturally upmod the things they like, it will drop off the front page in due time.

Trying to accelerate that process for things I don't like is getting emotionally invested in them. Why give them the time of day? The very best use of our time is working on the things that matter. Downmodding stuff is not working on things that matter.

Thinking about good posts isn't either, but it's a lot closer to things that matter.

I'll stop now. My thesis is this: not having a downmod button is a good thing if you believe that you should be channeling your time and energy into things that matter.

64 comments

[ 3.3 ms ] story [ 196 ms ] thread
PG has remarkably good instincts about feature choices.
Funnily enough, when I wrote my own reddit clone (back when every Lisper was doing it), I also reduced the choices from two to one. But I did the opposite - I removed the up-arrow and only had the down-arrow. Maybe I have remarkably bad instincts ;-)
Upmodded for candour, not because I think you were wrong. In fact, I think you simulated most corporate environments: peopel are silent when they agree but argue when they disagree :-)
why does PG leave out the search feature on HN?
maybe because it's not so hard to use "site:..." on google.
Using site: on google is a terrible alternative to search on YC. Google's ranking algorithms might work for the web at large but it doesn't work so well for searching HN. SearchYC does a pretty good job at searching YCombinator and I would like for such a feature to be integrated into this site, but I'm guessing PG has better things to do with his time.
Also, searching via google doesn't always work well because there are redundant copies of most comments on both the relevant post and the commenter's thread page.
I still think a "Search" link at the bottom of the front page (or a note in the FAQ) would be good, whether it points to searchyc.com, "google site:news.ycombinator.com", etc. People probably only need to be told how to search once, but searching for how to search is a catch-22.
That's absolutely horrible, since Google indexes individual replies here and often times you end up clicking 10 links that all belong to one post.
I don't think PG wrote this app the way it is today. He used iterative development and listened to suggestions from his users. Even on days he did not make changes, he might have thought certain things were fine. Over time, different pains became apparent.
I don't agree... I think it's better to have good content in the sites I visit, than to get less than a few seconds a day back for downmodding links. I prefer going to a good library, with many good chosen titles, than to a terrible library, where I have to do the full filtering myself, wasting more of my time by reading tons of useless book titles only to get one or two good ones.
It's not obvious to me that if everyone has a downmod button the library will be better. Somebody is upmodding books you don't like. How do you know they aren't downmodding the books you like?

I suggest that if the library has a few titles you don't like, going without a downmod button is a good thing, just ignore them and move along.

But if the library is overrun with bad titles, giving everyone a downmod button won't solve the problem, because the problem is caused by the librarians.

Now, that's a good argument for not having a downmod button. But not having a downmod button for avoiding wasting the people energy? That's what I don't agree with.
Well, the two motivations are orthogonal and neither is really right or wrong. So I can appreciate your point of view.
My point of view is that I only care about the quality of articles on the front page. If a downmod button would increase it, I would agree with such proposal, but I'm not sure (as you said) that such would be the case, and anyway I think the flag button does already work in a way like a big down button.

What I do believe is that content can get better if people actively submit more articles (so spending a little more energy in the site), not only by voting the articles in the new queue.

Anyway I have the feeling that the quality of articles in the site isn't always getting worse... these days I'm quite happy about it.

Also, this eliminates the gaming of downmods. A few downmods early in the life of a headline are very significant, so it's easy to game this with just a few cohorts. A few upmods are also significant, but this is often dwarfed by the deluge of upmods on a genuinely popular link. The action of the crowd doesn't have as strong a self-correcting effect on downmod conspiracy, because the crowd's exposure to the downmodded links is lessened.

There is a parallel in behavioral psychology: punishment can be a powerful tool, but can often backfire. So positive reinforcement is seen by many as the preferred tool.

That's probably the "real" reason we don't have downmods. Thanks for the insight, that hadn't occurred to me.
I am certain that it happened on reddit. In fact, there were occasions where the conspirators would go to the list of your past submissions and give you a little extra punishment to further discourage you from posting. I only had a few submissions at the time, and they were months old, so I am certain that this is what happened.

Apparently a lot of this has been addressed on reddit with some clever code. I think it's far cleverer to just leave the feature out!

I stopped using reddit after several such incidents with my raganwald account. After disagreeing with certain users there, suddenly all of my comments and posts would get downmodded in my history.

At first this would cause a karma hit, so I tried a workaround: after my submissions peaked, I would delete them and this would present a smaller surface area for revenge attacks.

This turned out to be a bad idea because it deprived reddit of the history. So I switched to a new tactic: I simply stopped participating in the forums and started submitting things under a nom de plume.

I think that they have changed things so that revenge downmods do not affect karma, but it doesn't matter to me. I stopped one day and asked myself why I was bothering to worry about my reddit karma?

Then I asked why I was bothering with reddit?

Then I closed up the nom de plume and HN is the only such site I visit. And let me tell you, I am a far, far happier man for it :-)

I note that HN also discourages historic downmods. I can't downmod any posts older than a day (but I can upmod them).
Yup, I only visit reddit now to see what the masses are up to.
If you're so worried about downmods early in the life of a submission, how about this:

Between P and P + H, where P is when the submission was Posted and H is some (constant?) number of hours, any downmods on the submission increment the downmod count but this has no effect whatsoever on the ranking. When we hit P + H, the downmods are made effective. An upmod on a story cancels a downmod.

Then if a story has lots of inappropriate early downmods but is actually interesting, it will get high on the front page and the mass of legitimate upmods will cancel the illegitimate downmods.

Showing the number of effective downmods (abs (- downmods upmods)) on a story might or might not be appropriate.

The simplest thing that culd possibly work: just leave the feature out. Less code that way.
Implementing that feature would likely take less characters than it took me to explain it. It's more a matter of is it a good idea at all, so unless you can come up with a scenario where we end up worse with this feature than without...
This strikes me very strongly as looking for justification for something beyond our control. "I can't change this, so I'll convince myself that I like it."

Let's look at the ability to upmod a submission. Upmodding a submission expresses approval of the submission, and signals to other users that it might be something that they would find interesting. Ultimately, the goal of upmodding is to increase the perceived quality of the site.

Downmodding expresses disapproval of a submission, and, again, its ultimate goal is to increase the perceived quality of the site.

While downmodding is vulnerable to a certain amount of gaming, that is a problem with its mechanics, not its principle, and upmodding is vulnerable to the exact same form of gaming.

If poorer quality items tend to get upmodded, and if there's no way to counter that, then the site's quality falls and eventually the site itself becomes the sort of waste of time that you're talking about.

Regardless, downmods aren't likely to be implemented at any point in the near future, and this entire discussion is a waste of time. I'm a little annoyed at myself for participating, but ... y'know ... someone on the internet was wrong. [http://xkcd.com/386/]

[Edit] P.S.: You miss your blog, don't you? :-)

You're arguing the value of downmods from the perspective of improving the site's quality, which is not what I argued. I am simply expressing approval with respect to my personal happiness.

As to convincing myself, I am now convinced :-)

[Edit] P.S.: You miss your blog, don't you? :-)

No. Exchanging ideas like this is not the same thing as blogging, no matter how much Web 2.0 people call the Internet a "conversation."

> Ultimately, the goal of upmodding is to increase the perceived quality of the site.

> Downmodding expresses disapproval of a submission, and, again, its ultimate goal is to increase the perceived quality of the site.

Don't have much experience with people, do you?

People have multiple, often inconsistent goals. Improving someone else's site may be one of them, but it's rarely the highest.

It all depends on the algorithm used behind the scene. Why would 1 downmod + 1 upmod = 0? Calling them up/down might be part of the framing issue that we're stuck with since the reddit days.

Fun property of a system with upmod and downmod: If every user mentally assigns a target score to a post (eg.: that comment should be worth 7 points) and votes always in the direction to reach the target, then the resulting score will be the median of all the targets. Statistically. That seems like a desirable goal, so maybe one should encourage people to vote that way.

[downmodding things I don't like] is getting emotionally invested in them.

I have this experience too, even for comments. On reddit, where downmodding a comment makes its down-arrow blue, it increases my attachment to it (partly because it makes them more visible - unlike HN's greying out).

Every so often on reddit, I'll downmod to express my frustration, but then later put it back to neutral, because I don't want to be attached to it. I rarely downmod - and it even more rarely stays. I think this approach is not common.

Is the issue of downmodding front-page because of the quality of recent HM articles? I'm finding HN less interesting over the past week or so (one symptom is the many recent meta-comments - including this one, unfortunately). I was thinking it's due to the US holidays...

[thesis]: not having a downmod button is a good thing if you believe that you should be channeling your time and energy into things that matter.

I agree. It's a bit like positive thinking, in that by attending to things that can help you, when you need something you have at your fingertips useful things - as opposed to filling your mind with things you don't like, when you reach for something, all you have is rubbish (or worse, arguments for why something isn't worthwhile).

But then there's the problem: what if HN starts suggesting many articles that don't help you?

Yes, by downmodding you help discourage people who post less HN-like stuff. I guess we're relying on PG to bring his methods up to date.
I think pg's solution is to ignore what you don't like:

Please don't submit comments complaining that a submission is inappropriate for the site. http://ycombinator.com/newsguidelines.html

He's talking about "inappropriate for the site" submissions, not "poor quality" submissions - but I think the same reasoning applies.

Perhaps the solution isn't a downmod button, but for the community to change its response? To ignore bad submissions instead of commenting on how bad they are. Commenting just draws attention to the submission and contributes to the problem.

I think there's a deeper philosophy here, that it would be worth hearing more about...

I wouldn't mind a placebo downmod. Just to get those feelings out.
That's not ignoring. So you are disagreeing with Raganwald, but haven't said which of his points you believe is mistaken.
(comment deleted)
Why to have downmods on comments then?
Good question. Perhaps trolling and hostile (or at least "uncivil") comments are better dealt with by different methods than spamming and irrelevant submissions?

Their motivations are probably different, at least, and comments are moderated by other people interested enough to read a particular thread's comments, while front page posts affect everyone equally.

You just can't stop "blogging" can you? :)
See my comment above, chatting with people that have similar interests is not blogging. And had I put this in my blog, it would have been off-topic.

I probably would have written it up as a draft, left it for a few hours, then deleted it as not being a net positive for what I was trying to accomplish.

I guess we have differing definitions of blogging; I consider a blog as achieving the same goal of information transmission. From that perspective, your posting could be considered a blog post that has eliminated one click in route, but I can appreciate people having different perspectives of things, blogging being one among them.

And about up and down buttons, current social news system are still inefficient and primitive. Consider this scenario: HN has two non-overlapping audiences, one English-speaking and the other Japanese. Now consider every front-page being equal parts Japanese and equal parts English, with the same articles but in both language. Can you provide a good reason for having up/down buttons or not having any? Any group of a considerable size will find itself coalescing into mini pools of interest, no matter how narrowly you define the group interest (Hackerese, in this group's case). The most desirable scenario is probably one that has the right balance of topic diversity, excluding irrelevant (unless you are interested in learning Japanese) ones. In that case, we are talking about some form of personalized news...so yeah, up and down vote buttons in their current state are like the reply button when browsing using an email client.

I consider a blog as achieving the same goal of information transmission. From that perspective, your posting could be considered a blog post that has eliminated one click in route, but I can appreciate people having different perspectives of things, blogging being one among them.

Well, I have never stopped "transmitting information," there is just a change in my strategy.

(+ Sometimes there are simply not enough parenthesis :)
I'd like to have a button that says "hide from my view".
I don't think that would serve the audience/community of Hacker News well at all. If you're hiding stories that are popular, the bad content is still existing in the community and shaping the brand and therefore it's current and new audience.

The benefit of Hackers News is the /users/, one only has to look at Digg to see where that goes once you open the flood gates beyond the social geek core.

.. as opposed to what? I'm not sure I get you. You sound like you're arguing for a true downmod, to help hide it from /everyone/'s view.
Upmod the good stories and ignore the bad ones. Eventually the bad ones will get filtered out of view as the good stories rise.
Agreed, a "Hide this crap" X would be very useful.
or ignoring upmod altogether, it's faster

i rarely use point because it's an easily gamed metric

nowaday i filter news based on #comments (and title)

and only follow comments with <= 3 levels deep (5 if interesting)

and start searching /pg /paul /nickb if it's a >50+ comments

I think downmod button is necessary but in the right hands, People who care about the good content. I would say, people with more karma points can have have the downmod rights. This is not democracy per se since some users have more rights than other users but it is a free market. People who have worked more harder and earned more karma points tend to have more rights than early users on HN.
I agree. I for one don't insist on having democracy here. If some other system provides better overall quality, it's fine for me.
I propose: The first rule of Hacker News is you don't talk about Hacker News.

Maybe I'm alone in not finding a lot of value in meta-discussion unless pg initiates it, though, because I see the parent post has quite a few points.

I suggest that for every user of HN, there is at least one post that makes it to #1 that the user feels should not be part of HN.

I know it is true for me, there are lots of popular posts that suprise me. That being said, I have no problem whatsoever with adding your suggested rule to the HN rules.

It would be much easier to understand than the curent "piques your intellectual curiosity and is not limited to hacking and startups."

A 1-week delay between creating an account and being able to submit a story would go a long way to cutting out spam, and not penalize those who haven't had time to build karma yet but still want to contribute.
I agree. I think dzone is an example of downmod gone bad as well. There's so many people who prefer technology X as opposed to Y that they downmod Y into oblivion all the time. It's made the index pretty much useless to me.
I actually think that in some ways having a sucky post on the homepage can be a good thing. For a newer developer reading the comments of their peers and heroes smack down a turd like the "real programmers" post is a good thing.

If a terrible idea/post/rant makes it on to the frontpage there are obviously some people that need the wisdom of this crowd to set them straight.

I love that HN is about positive energy. We don't downmod we explain why something shouldn't be upmodded any more.

PG should add rel="nofollow" to the link anchor for each comment... Then google search would hypothetically just work.
I've been noticing this log time since I join Reddit. In no time I was confronted with the "injustices" of downmodding. You summarized very well, it saves lots of energy, and creates a much more friendly community. I still laugh when I read all the mod dramas on reddit comments... It's just savage...
Not having downmods lets new stories have a better chance of survival. Reddit started going downhill when they failed to control downvoting. Stories posted from outside the American time zones are especially vulnerable, since a couple of downmods can push them into oblivion. I think PG made a very wise choice in not having downmods. More importantly it poisons the atmosphere. Reddit in the first few months was just plain incredible, much better than even the current hacker news.
(comment deleted)
I use Reddit a lot and I never use the downmod button but for spam (which I should be using the `report` feature for). I think HN's decision to just have the up-mod button is a incredibly intelligent one.

(But spam does still get to the front page of HN)

I'd rather have a Bayesian filter than a downmod. Heck, it might even pick up on self-referential forum language.