What happens when the trolls are gone?


Continuing the discussion from Articles on Comment sections:

Trolling is a problem, and civil comments ™ are surly doing a good job getting rid off it. But what happens after the trolls are gone? I will tell you. I can tell you, because I have been part of the nearly zero troll community of Krautreporter. After the troll is gone the commenters start demanding attention from the editors. Which is in many ways far worse for the common editor than the trolling problem, because of two reasons: resources and criticism.

Full disclosure: I do not work in journalism, I am not even a experienced commenter (I do comments for around a year now). So maybe I do not see the whole picture here and resources and criticism are not really big issues of the comment section. Also, English is not my native language, so please excuse my writing. Please PM me for grammar or spelling.


From the perspective of a newspaper’s manager the phrase “don’t read the comments” will remain after the trolling problem is gone, because she cannot afford her staff to argue in the comment section. No newspaper has the resources to let editors debate for hours with the commenters, even if the discussions are fair and reasonable.


As I stated here, I suspect that editors are not willing to debate a conflict of opinion in public. Not with another editor and not with the commenters. Jennifer Brandel makes the case that editors should engage with the readers before they write an article, because in that state editor and the reader share a common goal: find out about something. And she is right. In comment section editor and commenter do not share a common goal. The editor has already delivered and wants to go on, the commenter has just received and wants to debate. The same problem exists in many professions: physician & patient, writer & critic, web developer & newspaper publisher :wink:. It is a common thing, nobody likes their work critized. That is why they usually communicate via intermediaries, like lawyers, customer support or in our case community/engagement manager.


If both problems are real and we still want a debate between editor and commenter, we need a mechanism to (#1) force the editor into a debate that (#2) does not take forever. My solution would be a voting system to promote a commenter, who is then authorized to have a public discussion with the editor. The main thesis of this discussion may also be publicly decided. Maybe the commenter should candidate with a certain thesis for the voting procedure.

This solution might be good compromise between the editor’s resources and the commenter’s right to hold the editor accountable. Also, the editor cannot avoid a deep discussion by selecting the comments worthy of her attention (like in an AMA). Last but not least, this solution would be an incentive for quality argumentation, because commenters have to demonstrate good argumentation skills to get elected (not quite like, but similar to Reddit CMV).

Any comments? :blush:

Meet @argubot – the civil discourse bot
Designing our TALK product
What configuration options should we provide?
Going Beyond Hearken
From comments to conversations

This is a really interesting (and good) “problem” to make for ourselves. The way I like to think about this is by putting equal importance on moderation (aka, getting rid of trolls) and “engagement” (aka, becoming part of the community and reacting to what happens there.)

I take this so far as to resist building a “moderation” interface and instead focusing on a “curation” interface that includes both the ability to do away with the bad and the tools for time-restricted members of the news room to engage their communities effectively.

The idea of the community ‘electing a voice’ and expecting a response is very interesting. I can see it working well in an experimental contest where the editorial staff puts out a call for votes on topics and commits to engaging in a certain kind of debate, or perhaps commits to giving some real estate on the site to the winner(s). (With legal disclaimer of course :slightly_smiling:).

I think this could be even more powerful as a kind of community education tool. If the editorial staff is looking for a certain kind of community activity, making that part of the guidelines would be a great way to nudge their community to identify and elevate that kind of activity.


I would like to give more power in the hands of the reader/commenter. I fear that the editorial staff might want to avoid certain debates. I imagine a community that has some rights in the editorial decisions. The task would be to avoid that those rights can be exploited. For instance, in articles with low community interaction there shouldn’t be an election. Maybe an absolute threshold of votes has to be passed in order to grant the privilege of a public discussion with the author. BTW: this (voting on candidates) might be a way to include the lurkers more.


I like the election idea. I am wondering if the discussions can be two part curated. So in addition to an elected voice and kind of discussion, the process of this is curated as well. If trolls are gone then this could be an excellent project in making communities dynamic and self governing. Eg: “We don’t let you talk about this because of *insert reasons” as the beginning of negotiation. Where users and staff work out what the expectations are and what common ground they can find.

It isn’t mentioned often enough is that threads can be derailed by floods of people who are not strictly trolls . People reiterate and emphasize in hope of being heard . I would love to see a community work that out publicly and record that process.


I hear what you’re saying. The idea of certain thresholds is a must for this to make sense.

Concerning “giv[ing] more power” to the community, the editorial team needs to be the one to grant this. If it were baked into a piece of software without the ability to configure / buy in, I expect that it wouldn’t be used/respected.

For me, the setting of expectations and committing to engagement up front by the editorial team is the really interesting model. Doing this sets to the tone and encourages engagement, then rewards it in a predictable way.


Hi @sydette and @David,

thank you for your feedback and your interest in the commenter election idea. The idea came to me while reading and writing on this forum. So for me this forum has already paid off :heart_eyes:

This is a very important point. The editorial team has to be willing to discuss with the commenters. As @asuozzo pointed out here, the authors are not very keen to jump into the comment section of their article. This is also my experience. When the author actually does respond to a comment, its usually one of the following two: #1 a thank you notice for a detail correction, like a wrong date or a bad spelling. #2 a reply to an obviously bad argument, an argument so weak, it collapses on the slightest puff. In both cases the author’s intention is partially to never have to reply to this thread again.

If this does sound familiar to you, maybe we can agree that the author generally does not wont to get sucked into a in-depth discussion. When the trolls are gone, the author is not suddenly going to respond to the more subtle and sophisticated argument. He does not do it now, and he won’t do it then.

So how can we get the authors to take part in a in-depth discussion with a commenter about their articles? We want this to happen, right? As @David pointed out, it is not an option to bake this into a piece of software, because then the software would not be used at all. So maybe it can be done optionally (I know, I argued against it here, but I irred). Give the author the opportunity to grant a “commenter election” on her article, and maybe even let her choose to define a threshold of a minimum amount of votes.

Would the authors use this option? I am skeptical about it, but I might be wrong. Maybe some will use it and this would put some pressure on the others to use it as well. But I know, I wouldn’t be satisfied with the idea that the author can choose which articles she likes to discuss and which not, because I think that readers and authors do not share a common goal here. The goal has to be to formulate a standard that does not depend on the author’s capriciousness. But I see that the optional approach might be a way in to get the editorial team on board.

Also, a financial incentive would help. In the initial phase of the magazine Krautreporter.de it was part of the author’s job description to take part in the comment section. The same might be true for piqd.de (where I am commenting currently). So there is a good chance that the initiative has to come from the publisher.

Articles on Comment sections

I am not sure, I understood you correctly. The community justifies their reasoning for letting someone not talk about something – when and why do they do it, and to whom?

Yes, derailing is a mean problem. You cannot force someone to respond to your point. Except maybe in a 1:1 discussion, if you are an experienced discussant. That is partially why I like 1:1 discussions so much. If your partner tries to derail (which often does not happen on purpose), you can point it out and keep on track. In a discussion with multiple participants that does not work, because others are often more than eager to take the new track :trolleybus: (this icon is called trolleybus, perfect metaphor :stuck_out_tongue_winking_eye:).


###Q: What happens when the trolls are gone?

###A: They remain above the comment section.

What are your ideas to make the authors part of the community?


Thanks Christoph, I’m glad you liked the talk :smile:


When the trolls are gone, then you have the space for community hosts, reporters, editors and thought leaders to engage in thoughtful conversations. If, as Andrew often says, communities are places that news editors and reporters consider to be separate spaces, over there, it’s because they’ve historically been out of control. When the trolls are gone, we can extend the internal ladder of engagement in newsrooms and get real crosstalk with the community. I see this happening on specialty sites like ProPublica, and with select journalists. When the trolls are gone, we can really talk!


In my research on online deliberation we have never generated enough attention to attract trolls, nonetheless our groups of around 200 people suffered a lot of distortions while discussing with time centric tools such as a forum. So I am super interested in the post troll discussion. Informational cascade, scattered content, low info to noise ratio are just a few of the problems we have seen and the literature discuss (See Mark Klein papers). Ideation software a la idea scale solve the issue of filtering out redundancy and noise via adding a voting mechanism on the comment itself, but this beauty contest approach generates its own biases by highlighting the most fashionable topics and strengthening the voice of those that are already in the majority. In my field we care a lot about diversity of opinions for epistemic reasons and about capacity building of minorities for democratic reasons, thus the beauty context approach is not viable. I fear that the proposal to vote a person would generate a similar distortion. The voice of the majority would be constantly reinforced. The elected person would never represent the minority, effectively silencing it. I am wondering if an alternative could be a whitelist approach and a playful mechanism that allow community members to enter the whitelist by doing a capacity building mini class, or performing other-regarding and community enhancing activities. Then once a member is in the whitelist they have automatically a certain number of tokens they can spend to request an answer from the authors/editors. In order to gain more tokens they have to do more other regarding and community enhancing activities. There could be all sort of schema to donate tokens and to reward the donation of tokens toward people that have never generated a discussion but could be good candidates with the objective to promote engagement among lurkers. For example if I donate my token to a person that generates a comment that is judged interesting by the editor/author I would get back two tokens. Another simpler approach is to simply draw randomly 10 people from the whitelist that represent a variety of opinions.