One of the best achievements of the LessWrong community is our high standard of discussion. More than anywhere else, people here are actively trying to interpret others charitatively, argue to the point, not use provocative or rude language, apologise for inadvertent offenses while not being overtly prone to take offense themselves, avoid their own biases and fallacies instead seeking them in others, and most importantly, find the truth instead of winning the argument. Maybe the greatest attribute of this approach is its infectivity - I have observed several newcomers to change their discussing habits for better in few weeks. However, not everybody is susceptible to the LW standards and our attitude produces somewhat bizarre results when confronted with genuine trolls.

Recent posts about epistemology1 have all generated large number of replies; in fact, the discussions were among the largest in the last few months. People have commented there (yes, I too am guilty) even if it was clear that the author of the posts doesn't actually react to our arguments. After he was rude and had admitted to do it on purpose. After commiting several fallacies, after generating an unreasonable amount of text of mediocre to low quality, after saying that he is neither trying to convince anyone nor he is willing to learn anything nor he aims for agreement. In short, perhaps all symptoms of trolling were present, and still, people were repeatedly patiently explaining what's wrong with the author's position. Which reaction is, I must admit, sort of amazing - but on the other hand, it is hard to deny that the whole discussion was detrimental to the quality of LW content and was mostly a waste of time.

So, here is the question: why didn't we apply the don't feed the troll meme, as would probably happen much sooner on most forums? I have several hypotheses on that.

1. We are unable to recognise trolls for lack of training. The first hypothesis is quite improbable, given that the concerned troll was downvoted to oblivion2, but still possible. There are not many trolls on LW and perhaps it is difficult to believe that someone is actively seeking that sort of confrontation. I have never understood the psychology of trolls - I try to avoid combative arguments instinctively and find it hard to imagine why somebody would intentionally try to create one. Perhaps a manifestation of the typical mind fallacy combines with compartmentalisation here: although we consciously know that there are trolls out there (as this is hard to ignore), when meeting one our instict tells us that the person cannot be so much different from us.

2. We are unwilling to deal with trolls. The second theory is that although we know that a person isn't sincere, we cherish our standards of discussion so strongly that we still try to respond kindly and maintain a civil debate, or at least one side of the debate. If it is the case, it is not automatically a bad policy. Our rationality is limited and we always operate under the threat of self-serving biases. Some quasi-deontological rule of kindness in debates, even if it is an overkill, may be useful in the same way presumption of innocence is useful in justice.

3. Sunken costs. Once the debate has started, our initial investments feel binding. It is unsettling to quit an argument admitting that it was completely useless and we have lost an hour of our life for nothing. Sunken costs fallacy is well known and widespread, there is no reason to expect we are immune.

4. Best rebuttal contest. An interesting fact is that not only the number of replies was fairly large, but also lot of replies were strongly upvoted. It leads me to suspect that those replies weren't in fact aimed at the opponent in the discussion, but rather intended to impress the fellow LessWrongers. Once the motivation is not "I want to convince my interlocutor" but rather "I can craft an extraordinarily elegant counter-argument which until now didn't appear", the attitude of the opponent doesn't matter. The debate becomes an exercise in arguing, a potentially useful practice maybe, but with many associated dangers.

5. Trollish arguments are fun. I include this possibility mainly for completeness since I don't much believe that significant number of LW users enjoy pointless arguments. But still, there is something fascinating in fallacious arguments. They are frustrating to follow, for sure, especially for a rationalist, but I cannot entirely leave out of consideration the appeal of seeing biases and fallacies in real life, as opposed to mere reading about them in a Kahneman and Tversky paper.

Whatever of the above hypotheses is correct, or even if none of them is correct, I don't doubt that on reflection most of us would prefer to have less irrational discussions. The karma system works somehow, but slowly, and cannot prevent the trollish discussions from gaining momentum if people continue in their present voting patterns. One of the problems lies in upvoting the rebuttals which gives additional motivation for people to participate. There seem to be two main strategies of voting: "I want to see more/less of this" and "this deserves more/less karma than it presently has". The first strategy seems marginally better for dealing with trolls, but both strategies should work better when applied in context. Even a brilliant reply should not be upvoted when placed in an irrational debate: first, it is mostly wasting of resources, and more, we certainly want to see less irrational debates. I don't endorse downvoting good replies, if only because the troll could interpret it as support for his cause. But leaving them on zero seems to be a correct policy.


1 I am not going to link to them because I don't want to generate more traffic there; one of those posts figures already on the 4th place when you Google lesswrong epistemology. Neither I write down the precise topic or the name of the author explicitly, which I hope decreases the probability of his appearing here.

2 In fact, the downvoting, even if massive, came relatively late, with the person in question being able to post on the main site after several days.



New Comment
247 comments, sorted by Click to highlight new comments since: Today at 3:30 PM
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

So, here is the question: why didn't we apply the don't feed the troll meme, as would probably happen much sooner on most forums?

To your question, my own question: Why would we apply the "don't feed the troll" meme? There is no reason to believe it works. There's no academic study of least trollish places on the web, nor are there reams of data with which we can analyze best responses. Interestingly, the most trollish places on the internet tend to rely entirely on the "don't feed the troll" meme. While the least visibly trollish places on the web all use distributed moderation systems along with an active defense by all the posters there to combat trolling.

Using 4chan's method of ignoring trolls when we could be following Stackoverflow's method of obliterating trolls is insane. It's like asking for advice on passing our driving test from the girl who failed the test 12 times. Sure she's got more experience than anyone, but that's only because she so bad at it. We have no reason to believe that "don't feed the trolls" works. Going further, I'd say we have good evidence to believe it doesn't work. We should first determine what deters and eliminates trolls best, instead of aspiring to an untested ideal.

"Don't feed the trolls" is just a meme that may or may not work. Nothing more and nothing less.

I agree that 'don't feed the trolls' is an untested folk theory that we shouldn't necessarily assume works, but your argument seems to assume that the expected result of not feeding trolls is no trolls at all, which I'm fairly sure is a strawman. The point of not feeding the trolls is, as far as I can tell, to minimize how disruptive they are when they appear. It's more analogous to "if you cut yourself, apply first aid and consider seeing a doctor" than "if you want to avoid scurvy, make sure to get some vitamin C in your diet".

For what it's worth, I know several people who love to troll and feedback definitely seems to encourage them. This [] is also a relevant discussion.
I have no experience with 4chan. But from what you say follows that regular users of 4chan actually don't feed (i.e. communicate with) the trolls and the trolls are still there. How does such a discussion look like? Do the trolls interact only with each other? By which I partly mean that the meme obviously doesn't work when people fail to apply it. There are more direct ways to obliterate trolls, but they come with some costs.
4Chan is probably not the best choice of a site to emulate in this context; it's clearly extremely successful at community-building, but its values concerning discussion could hardly be more different from ours. Describing it as "trolls trolling trolls trolling..." is almost a cliche, and one more or less borne out by my limited experience with the site. On the other hand, I'm certain that the "don't feed the troll" meme predates it considerably.
Goes back to Usenet at least, I'm pretty sure.
This [\]#Troll_sites) might be enlightening. Also see this comment []. Being a troll sort of is a prerequisite to join 4chan.
I didn't understand the OP to be suggesting that we not use the existing distributed moderation system. I understood him to be comparing ("don't feed"+distributed moderation) to ("continue to encourage"+distributed moderation). And, sure, I'm all for gathering data about what works. That said, I'm reasonably confident that "don't feed" works better than "continue to encourage," all else being equal.
Why do you believe what you believe? I too have been told not to feed trolls in the past. I have heard that meme so often that I have formed an automatic pattern. Yet there continue to be trolls. At this point, I am convinced that "don't feed the trolls" is pure superstition. Imagine if you will, a universe where the "don't feed" doctrine actually worked. "Don't feed" is already a common meme, so trolling in all parts of the internet would be obliterated in a handful of hours. The "don't feed" meme itself is only invoked when one spots a troll, which means that people would go without hearing it and gradually forget it. The meme would die out of memory as it killed off it's own reproduction vector, the trolls. We'd only have small, isolated flare ups of trolling as people re-invented trolling, then others remembered or re-discovered the cure to trolling. Is a version of reality that the world actually resembles? Now imagine a universe in which "Don't feed the troll" is useless or worse, it emboldens the trolls and causes them to act out more. How would that look? One of these possible universes is much, much more similar to reality than the other. I think that says a lot about the efficacy of "Don't feed the trolls" as a policy.

The problem on most forums is that people say "don't feed" while continue feeding (not necessarily the same people are engaged in both parts). I believe that the not feeding policy works because (1) I don't remember seeing a really obnoxious trollish exchange not feeded by non-trolls (itself a weak argument, since I don't frequent troll habitats often), (2) it corresponds to my proto-model of troll motivation, which is seeking attention (also a weak argument, I don't really understand trolls) and (3) the trolls need something to react to, and the responses to their debate contributions provide more material and thus opportunities (this is a bit stronger argument, it seems almost self-evident).

The continued simultaneous existence of trolls and the no-feeding policy doesn't say much and is well compatible with the policy being effective. Note that:

  • The claim isn't that non-feeding is capable of eliminating all trolls in any situation, but only that it reduces the negative effects of trolling.
  • The non-feeding policy, although well known, isn't universally applied, and trolls can easily thrive on places where the local debaters lack discipline and engage them.
  • To show that t
... (read more)
Trolling that isn't being replied to is simply called spam. Unfed trolls aren't trolls by definition. That leaves open the question whether a policy of not feeding trolls reduces the the total volume of trolling + spam aiming to troll. The answer seems likely to be yes, but might depend on how rigorously the policy is followed (I can imagine unsuccessful admonitions not to feed them encouraging trolls).
I believe it because in all the epic trollings I've seen, there have been locals who have engaged with the trolls throughout. I can't remember the last time I saw a troll simply monologuing into the silent ether at length. I also believe it because I've experienced people in real life who seem motivated by the desire to get a response from others, but who don't seem differentially motivated by different kinds of responses. But, all of that said, I certainly agree that none of that is definitive. I could easily be wrong. And it doesn't matter too much for my own behavior... I mostly don't talk to trolls because I don't enjoy it. If you are getting good results from talking to trolls, that's a fine reason to keep talking to them.
I confirm that your interpretation is correct.

The poster in question at least wouldn't be immediately obvious as a troll to outsiders reading only a small part of the discussion, more proactive ways to deal with posters like that would seem to carry a serious risk of making less wrong appear even more cultish (group-think).

For whatever measures we take we should first consider how much ammunition against less wrong they offer, how likely they are to cost us genuinely valuable contributions due to seeming closed to dissent, and whether the expected magnitude of intended effects is worth that.

That includes using the word "troll". After the various facts about their behavior and motivations there is no additional fact as to whether they are a troll. Using the word troll might easily lead people who only took a quick look to come away with the impression that we generally dismiss non-bayesians as trolls, whereas talking about how to prevent endless discussions not aimed at resolving disagreements seems less dangerous that way.

That includes using the word "troll". After the various facts about their behavior and motivations there is no additional fact as to whether they are a troll. Using the word troll might easily lead people who only took a quick look to come away with the impression that we generally dismiss non-bayesians as trolls, whereas talking about how to prevent endless discussions not aimed at resolving disagreements seems less dangerous that way.

Indeed. I'm not even sure the user in question was a troll by intention (even if they were one functionally) - being persistent and dense beyond reason is a highly plausible trait of participants in Internet philosophy discussions, after all, particularly when the participant has their very own site all about what they're talking about.

That is, the label "troll" assigns intent in a way that is not actually all that relevant to the problem, which is the behaviour, when you can accurately describe the problematic behaviour.

"I prefer trolls to cranks, because trolls sometimes rest." - Alexandre Dumas (fils) (loose translation)

This is a valid point. Approximately for these reasons I have limited my suggestions to altering our individual voting policies, which seems reasonably safe - just don't upvote a comment if it appears in a nowhere-leading lengthy debate, even if the comment itself is well written and sound. I agree that more proactive methods carry risks which we may not have enough reasons to take now.
Very, very good point. I apply a zero-tolerance approach on my own blog, banning people who even faintly smell of troll because they're not worth my mental energy, For a site like this (where the deletion of one post caused such an uproar) it would be counterproductive at best,
Re the word troll: Yes, I could have written that without using the word. The reason I did use it was a need for a label, since trolling is not an easily describable phenomenon (it implies not only endless discussions not aimed at resolving disagreements, but also a kind of rude behaviour, use of fallacies, personal attacks on the opponents and excessive amounts of generated text). Since there was a standard label for that kind of behaviour, I used it. I couldn't think about other standard label with approximately same meaning which lacks the negative connotations, else I would have used it instead.

I don't think the "trolls" started that way - in fact claiming to be trolling (or that they don't care, etc) is a common defensive response, especially on the internet, after which appearances and cognitive dissonance do lead to some actual trolling.

They were just people who disagreed with with the majority, with the unfortunate need to frame everything in terms of their favorite topic. Sure their arguments were well below normal LW standards, but we have high standards. They were normal, pre-rationality-training arguers, and not talking to normal people doesn't appeal to me at all.

Neither does to me. I also agree that at the core there was a genuine disagreement which later evolved into the entrenched defense we have seen. However, what do you suggest to do in such situations?
Tough question, let me think. It's possible that pushing particular LW posts at people when we see a problematic argument would help, e.g. resolving arguments about definitions, being charitable to the other person, arguments as soldiers, that sort of thing. Someone asking the other person to write a longer, more general discussion post outlining their views actually helped me a lot - previously I didn't understand just what they was being argued for. Once people are already committed to not listening to us, feeding them or not feeding them doesn't seem to make a difference (unless some person goes a step further and start being intentionally disruptive, but then they'd just get banned, hopefully), but maybe we could "not feed the trolls" as a proactive strategy against having bad discussions. If someone seems to be getting into a bad pattern, we could stop all confrontational discussions and try to only have more cooperative, fact-finding discussions. Discuss the problem before proposing solutions (and link to that too). If they are confrontational, turn the other cheek, unless they violate very basic (e.g. not lying) ground rules for discussion. This plan should be carried out in a way that does not fail if they are right, so discussion should keep going as long as there you can muster genuine curiosity, and not much longer. For ending discussions, even if you feel tired of the other person, maybe try to ramp up the politeness - "thank you for talking about this" "I'm sorry we couldn't find more common ground," that sort of thing, and then link them to an interesting place in the sequences if they want to read more. Not sure if that would actually work, but I'll try to give it a shot next time.

Your first suggestion, "we are unable to recognize trolls", is a valid one. I personally didn't think of the downvoted user as a troll, just as someone who supported an incorrect epistemology and had a couple of confused ideas. It was only after said user failed to learn from several clear, well-written replies that it became clear that ver presence on LW was a bad thing.

It wasn't immediately clear to me either. But the discussions had continued well after said user failed to learn from that replies. And I have written at least two responses to his posts when I already knew it was going nowhere, and it required non-negligible willpower to give up after that.

A really good troll would be able to maintain a high karma level even while wasting Users' time on trivial matters -- and this forum has been able to fend off even that kind, so I would not regard it as worrisome that an occasional less-obvious troll like User:[withheld] succeeds.

I don't know whether to upvote this or downvote it. I recognize that I am confused. I shall reward your assistance in that matter by buying an extra set of paperclips tomorrow. Given my general habit of misplacing small objects in both my apartment and my office it is likely that these paperclips will stay intact and unused for some time.
You're a good human!
A hundred thousand upvotes would not be enough!

How about 6.: arguing for the lurkers' benefit? That argument was voluminous, repetitious, and random sampling of the low-voted comments turned out to only be useful for verifying that their scores weren't capricious, but a few of the upvoted comments were worth my time to find and read; if a few dozen other people felt the same then they might have also been worth the authors' time to write.

It is true that some people can benefit from such arguments. My claim is that a voluminous repetitious heated debate is a rather inefficient way to convey the benefits. If I want to provide the lurkers with a valuable counterargument against a plausible position, I would better write it somwhere else, in a separate post with proper explanation, where the readers would not be distracted by the noise invariably present in the trollish debates. There is a second related point. It seems to me that if LW is intended as a place to have arguments, the arguments ought to satisfy some minimal quality requirements. There are many places over the internet where you can encounter debunking of common non-sequiturs, red herrings, ad hominems and strawmen; I just think LW should be one level above that - we should take for granted that the debaters here don't commit those things, at least not systematically. You may possibly argue that if some people wish to engage in fallacy busting, let them go; that LW can have a low-quality debate now and then. The problem I have is that those debates are distracting. I feel an urge to react to fallacious arguments, to support people arguing against rude opponents, and often don't resist and do that. But afterwards, I almost universally regret participating in these nowhere-leading discussions. I don't think that I am so extraordinary in this respect and suppose thus that people can be attracted to such debates "against their will". And finally, such debates feel bad and damage the atmosphere of friendliness and mutual trust which LW has.

For what it's worth -- I had one exchange with the person in question pretty early on, decided based on their response that the conversation wasn't going to go anywhere useful, and dropped it.

But I mostly refrained from downvoting them initially because there were people I respected who were continuing the discussion with every indication that it was being productive, and I value productive discussion even if I'm not getting anything out of it personally. (There are a lot of exchanges on this site that kinda sound like gibberish to me, and in at least some cases I'm fairly confident that this is because I don't understand the issue well enough to participate usefully in the discussion even as an observer, not because the participants are in fact spouting gibberish.)

After a very short while I stopped reading any of the comments on that thread except to sample them every once in a while to see if they were going anywhere, and it seemed pretty clear that they weren't. At that point I started downvoting everyone involved on the grounds that I want less high-volume discussion that makes no progress.

No general lesson here, just another data point.

In the cases where you believe people are talking gibberish, I'd suggest adding a comment saying that you're confused. It lets people know which topics people find confusing. And if you're lucky you might get some links to background material which will help things make sense.
I do that sometimes, but in general only when I'm prepared to dedicate some effort to making sense of an explanation, should someone provide one. "I'm confused, and choose to stay that way for now" seems like an inane thing to say.

As someone who was probably doing more feeding than most, I'd like to apologise here.

In my case it was primarily the 'didn't recognise troll' problem, I'm not very good at distinguishing the more eloquent and seemingly reasonable type of troll from honest commenters who disagree with me. I also have a strong aversion to just walking out on a debate without reaching any kind of agreement, mainly because it annoys me a lot when other people do it.

I will try not to fall into a similar pattern again, but since I'm not all that good at noticing it any 'don't feed the troll' PMs are appreciated.

I'm in a similar situation - I can, technically, recognize that someone fits the profile of 'troll', but my brain doesn't like to actually use that information for anything. (It's not just an issue with trolls, it's a general inability to use different scripts for people based on categorizing them.) What I've found works, in my case, is to be more aware of subtleties in peoples' behavior, rather than trying to categorize them. I still wind up feeding trolls sometimes (and in fact tend to enjoy doing so in those cases), but if someone is being logically rude or otherwise offensive, that's a thing worth noticing and reacting to whether they're 'a troll' or not.

Another trick that's useful sometimes is backing out of the argument a bit and looking at the bigger picture. Sometimes that shows patterns that aren't otherwise obvious, and that can make it clear that the person isn't worth continuing to deal with. For example, in the most recent case, the individual was defying the evidence in an irrational and not very obvious way. Once I noticed that, it seemed obvious to me that without a more thorough understanding of when that is and isn't a reasonable thing to do, they weren't going to stop doing it to any evidence that we gave them, and thus the argument at hand was not going to resolve anything, so it was pointless. It's a lot easier to walk away in a situation like that, when you can see that it's impossible to actually get the goal you were aiming for.

I'd like to apologize as well. I have a bad habit of not knowing when to stop talking to someone. I agree with you and several commenters about about not being able to immediately recognize trolls, although I think it's good in the long run that we are slow on the uptake rather than trigger-happy. There is, of course, an important difference between leaving a debate unfinished and walking out on one that's a lost cause. But as you pointed out, it's hard to distinguish between the two from an inside perspective. What I'm going to try to do in the future is simply ignore people when they reuse previously refuted arguments, as this is generally a sign of dishonesty or denial.

1. We are unable to recognise trolls for lack of training.

2. We are unwilling to deal with trolls.

3. Sunken costs.

4. Best rebuttal contest.

5. Trollish arguments are fun.

It's good to recognize how tempting these are. In fact, my impression was that the exact same mental bugs were motivating the "troll" in this case to continue with the conversation. As far as motivation for participating in that specific conversation was concerned, the only difference between the "troll" and his interlocutors was that the interlocutors knew for sure that they had a large approving audience.

Good observation.

There's another issue here- a lot of my comments in reply (almost all of them) were upvoted. Some were upvoted quite high. I interpreted this as a sign that people were interested in the subject when it seems in retrospect I should have interpreted those upvotes more as "agree" or "well-argued" and not "want more". Being more clear about what people mean when they upvote might help.

I'm not sure clarity is the issue. Note that this comment [] explicitly endorsing such replies has itself been moderately upvoted, which might reflect a more general endorsement of such replies. So it's possible that the upvotes really were expressing "I want more of this" from an interested minority, as you initially proposed. (And as you convinced me was plausible initially.)

Upvoted for admirable restraint in not linking or naming.

Several times the troll mentioned he was forced to slow down posting by the site. Was this because of low karma? If so, can we just penalise people more for massively negative karma?

Apparently there is a 10 minute countdown between comments if you have negative karma.

That might not have helped in this particular situation [] .
1Paul Crowley12y
OK we can't hope for a fully attack-resistant trust metric, but I'd like to do a little better than this.
Inability to create discussion posts for users with negative or zero karma was already proposed when we were experiencing attacks of spambots. I don't know whether it has been implemented already, but suppose it hasn't.
It has been. It isn't clear how they were able to post the top level post so late. Some have suggested that they made additional accounts to vote up their older posts but I don't know of any evidence of that. Unfortunately, there's very little in the system that makes detecting that sort of thing very easy. (ETA: By "they" I mean the potential troll, not the spambots.)
Spam stopped immediately when having positive Karma became a requirement. Only a few spam messages appeared in comments after that, AFAIK.
Yes, sorry, by "they" I meant the troll in question, not the spambots. Bad wording on my part.
I'd be surprised if that helped very much in cases like the one under discussion, given that nothing stops people from creating new accounts. That's enough to stop casual spammers, which is great, but I'd expect anyone willing to sink hours into writing comments to also be willing to create new accounts on demand when their karma got too low. More generally, I'd be surprised if any change to the karma system itself rendered us significantly less vulnerable to that sort of dedicated resource-grab without introducing negative side effects. My own feeling is that the best immune system is cultural, here. To the extent that LW members find participating in discussions like that one valuable (there's a defense of that position here [], for example), we will continue to periodically experience such discussions.

Replying to people who are wrong or systematically wrong is not a problem, so long as we keep the good Less Wrong tradition of addressing the statements with excruciating seriousness, like this.

The problem appears when people who systematically produce wrong or low-quality content don't slow down in modes in which they get downvoted. Global Karma level and 10 minute delay don't address this problem directly.

Perhaps downvotes should act as a cooler measure, temporary ban points, for example:

  • Count the total Karma K of all comments published within the las
... (read more)

That system would also ban people for some time after they set up a poll. Perhaps it would be better to let K bet the total Karma of all comments, not just negative Karma comments, so good comments could offset bad ones.

This could just make implementation of a polling system a dependence. Total Karma probably wouldn't work, most of the edge cases where it's not totally obvious that the user should be removed allow for a positive average balance. Also, one of the use cases applies to established contributors going into a wrong mode, in which case they'd be quite capable of offsetting the downvotes.
I agree, provided we really do add a poll feature first.
My $0.02 - I'd be OK with extending the current "impose delay on negative karma" policy into a tiered solution with longer delays for more negative scores (either using the algorithm you sketch or some other), if someone were highly motivated to code that, but I don't think it would be a particularly valuable use of anyone's time.
I worry that a measure like this would encourage trolls or people with trollish tendencies to start PMing their interlocutors, which would be a public good but a private nuisance with no outlet for moderation.
Throttle PMs too, then.
There is a "report" link on PM's. What does it do? We could also add a feature to allow a user to block another user's PM's.
Goes straight to Santa Claus [].
I'm not sure. When a comment is reported, I can see how many reports it's accumulated when viewing the comment in some ordinary context, and then I can "ignore" the reports (making them evaporate), or ban the comment, or leave it alone for someone else to deal with. It doesn't send me a message notifying me that a report has been lodged. I can't see other people's PMs so I'm not sure how I'd become aware of reports on them.
These links collect reported items on main section and in discussion (they work for moderators, not visible if not logged in, not sure about other users): * [] * []
I get the message "The page you requested does not exist".
That is because you are not Englightened.
Oh, neat, thank you.
Well, there goes my apparently silly theory that moderators would have access to a list of pending reports, which would obviously let them see reported PM's that they wouldn't be able to see otherwise.

... all symptoms of trolling were present, and still, people were repeatedly patiently explaining what's wrong with the author's position. Which reaction is, I must admit, sort of amazing ...

Amazing only if you assume that educating that trollish author is the purpose of the response. Frequently, though, one responds imagining an audience much larger than a single troll. Sometimes one writes experimentally, for oneself, seeking feedback from the community as to whether one's own viewpoint finds resonance with other people.

it is hard to deny that the

... (read more)
Yes, but at the same time, people who enjoy that should realize they might be damaging LW's very good signal to noise ratio. I certainly was guilty of that by repeatedly replying.
At the risk of sounding glib, one man's signal is another man's noise. About half of those threads consisted of reasonably good and interesting arguments. And the other half included some links to good ideas expressed less shallowly.

I'm unfamiliar with this particular instance, but I've engaged trolls on Fark before, for these reasons:

1) Some trolls just give up when taken calmly seriously at face value, as opposed to getting hit with indignation.

2) Bored at work.

3) They provide such wonderful strawmen against which to clarify my own thoughts and sharpen my arguments.

4) In some discussions (political ones especially), Poe's Law applies, oftentimes real people hold real opinions that are as blunt as trolls' opinions.

In the discussed case I have little doubts that the said troll indeed held most of the opinions he has expressed here. To clarify, by "troll" I mean someone who argues for fun of arguing and without caring a bit about the standards of a reasonable discussion; a troll needn't to pretend holding opinions which he in fact doesn't share. (My use of "not being sincere" in the OP might be misleading in this respect.)

Responding to curi was good training. I found myself discovering, for myself, what levels parts of my understanding were at - automatically seeing exactly where he went wrong, or being able to derive from my web of beliefs a principle that was violated... and I could go on.

Point is, I fed the trolls because it was good exercise.

I recommend numbering your hypotheses, for easy reference.


There are some more constructive reasons for arguing with the incorrigible. One is to persuade the audience, rather than your interlocutor -- anything you say on a public stage is addressed as much to the general readership as to the individual being explicitly addressed. Another is to practise your skill at arguing the material.

Not that these justify going on and on indefinitely, but they are worth putting in the balance.

  1. Trollish arguments are fun.

As something of a troll, depending on what is meant by that at least, it is often quite interesting to debate with people that do not hold your point of view and are capable of making decent arguments. Too often one runs into arguments where both sides don't know what they are talking about in the slightest, which is terribly frustrating.

As for the Karma, I have had for a while enough karma to make a top level post, getting it isn't particularly hard. I would have much higher karma if I didn't occasionally post controversi... (read more)

Also one could probably create an semiautomatic spammer as well that had the purpose of defeating the karma requirements without much difficulty. Basically create a core of accounts that is the size of the minimal number of Karma needed to make a top level post. Then one of those accounts needs to get one Karma, which hopefully can't be automated. Then one has to create some banal discussion that includes one post from each of the core accounts which then gets upvoted in a cycle. Then start creating spam accounts, have it post randomly something like "I agree" on something and have the core accounts upvote that one post so the spam account can post the spam on the top page. Rinse and repeat. It would be important to avoid detection to have a scripted discussion between the core accounts and to not have the spam accounts post on that discussion except according to the random algorithm that is being used. There are probably ways to make this simpler to create and there may be easy ways to defeat such a program. This is just an idea.
We can implement a karma minimum for upvotes if that becomes a problem.
One interesting "feature" of the karma system that makes this a lot easier is that fact that upvotes/downvotes of deleted comments still contribute to your karma.
On the topic of karma, why are you downvoting every post I make regardless of content?
(Downvoted for making an accusation without presenting evidence.)
It's a long story, starting with Eugine publicly declaring that he was downvoting the comments I made that he disagreed with, which has seemingly escalated to downvoting every comment I make even where I'm just conducting meta-housekeeping and the like. I'm not commenting to shame or accuse, I'm trying to understand his motivations.

Good post, thanks.

I note you don't list as a hypothesis "even though the interlocutor was a troll, their arguments had enough merit to be worth the effort of rebutting". I didn't read most of the threads; I take it if I had, I'd know that wasn't a good hypothesis? Or does this come under the "Best rebuttal contest" heading?

It was rather an unstated assumption that the arguments have not enough merit. Of course a troll can make few good arguments, and perhaps there were few in those threads, but most of the debate wasn't centered around reasonable arguments. The worth rebutting property itself can possibly be further reduced with respect to the motivation of the debater. Arguments may be rebutted for pure intellectual curiosity, just as "let's look whether I can find a hole in that reasoning", they may be rebutted to ensure that other people don't fall victim to fallacious deductions, or they may be rebutted to show one's intellectual superiority or other social signalling purposes. Those are different mechanisms and an argument worth rebutting for one reason may not be worth the same for other reasons (in fact, presence of the signalling reasons depends more on the context than on the argument itself). I don't endorse the rebutting for signalling, as it is in fact what the trolls generally aim for. When other motivations are present, a better way is to create a separate post to discuss the problem, ideally after waiting for a while until the troll disappears. Instant rebutting will likely be contaminated by trollish distractions and thus suboptimally productive.
3Paul Crowley12y
Yes, that makes sense. Thanks.
That hypothesis is actually the most reasonable one. The troll's arguments didn't have any merit - they were all perfect examples of every bad argument going - but they were the arguments one sees time and again from people who aren't (consciously) trolling.
I think this is quite a large part of it. I have several times on Less Wrong followed discussions that seemed to be headed towards trollishness, and then all of a sudden someone changes their mind, updates, and everyone moves on. It is one of the things I love about this website, and I would be sad if an anti-trolling sentiment led to these sort of discussions being abandoned before they concluded. Sometimes persistence is a waste of time, but sometimes it makes a difference.
1Paul Crowley12y
I'm glad I didn't read it now - thanks!

Under the usual rules, trolls are to be treated like zombies: they emit messages, but their words don't reflect what they actually think, but a sort of fake-thinking designed to deceive you. Or they are outlaws: responding in good faith towards them is considered bad behavior, under the "don't feed the troll" principle.

"We are actors; we are the opposite of people. So? We need an audience." — Rosencrantz and Guildenstern are Dead

If what you care about is arriving at accurate beliefs about the world under discussion rather than about the... (read more)

We could hide downvoted posts from "recent posts" (make them available only to those who already know the url) and hide downvoted comments from "recent comments". And hide all children of downvoted posts/comments from "recent comments" too. That would discourage feeding.

Undesired side effect.
Why? We already automatically collapse downvoted comments along with all their children, to discourage people from writing more comments in these threads. My proposal has similar intent.
Because there are many valuable discussions that have happened to occur below one downvoted comment. I do not desire any additional penalties to those comments. Deal with the problem of trolls by some other more direct mechanism.
Hiding downvoted posts is already available as an individual preference, which lots of people have turned off. Do not want.
I want to have a possibility to see all recent posts without knowing the specific url, if only to be able to find it easily if I want to look at the comments at some later moment. But I agree that it may be useful to have an option to switch the display of downvoted posts off, analogously to how it works for comments. It should even be the default setting.

So, here is the question: why didn't we apply the don't feed the troll meme, as would probably happen much sooner on most forums? I have several hypotheses on that.

There have been a few comments of that type, and I've sent private messages to some participants saying just that. Public "don't feed the troll" can be an encouragement.

Part of this is probably due to the feeling that "we can't let this outrageous statement stand". We feel the need to respond to a bad argument, even if there is no particular benefit to doing so.

I know. Never said it was rational. Upvoted (that's a particularly good one)

Which is how these disucssion always go.

If you knew in advance that you were wrong and your comment was meaningless, why did you make it?

[+][anonymous]12y -16

New to LessWrong?