Astroturfing and troll farms are different from friends and people on your side saying their opinion
This is correct. What I am talking about is brigading.
Astroturfing and troll farms are only similar in the mechanism behind their ability to distort public opinion. That mechanism is: People are influenced by the tone and volume of comments they read.
Are you saying you’re against people being allowed to tell their friends and supporters about something they consider to be unethical and encouraging them to vote and comment according to their conscience?
Yes, this is brigading. There are things you can do mitigate this brigading effect, for example: (1) Begin comments with “I am here from …” or “This post was shared by...”. (2) Commenters acknowledge, when asked, that the post was shared with them.
Take your case (i.e. Ben Pace’s post on nonlinear), neither (1) or (2) was done. In fact, I found myself needing to comment alluding to this effect, after I confirmed this was the case with one of your collaborators.
Inviting participation from people who are not part of the relevant community is clearly brigading. Unless they abstain from voting and clearly disclose their origin, they would be masquerading as community members and giving a false impression of the community’s views.
Inviting participation from people who are part of the relevant community presents a closer question. There’s still a risk of creating a misleading impression of the community’s views, but there isn’t the astroturf-like presentment of inauthentic views as community-member views that classic brigading presents.
For instance, I would generally view reaching out to a reasonable number of active Forum participants individually as not brigading. This is less likely to create a sufficient mass effect to mislead observers about the community’s range of views. This may be contingent on those individuals following the next bullet point.
I suggest that people who have been “recruited” in this manner should ordinarily refrain from strong-voting on the post, either up or down. A small group of strong-vote wielding users can significantly effect the course of discussion on the Forum through their voting. If someone is close enough to the subject of the post to solicit for support without constituting brigading, they are likely too close to the subject to be casting strong up/downvotes on that post.
I would generally view reaching out to a reasonable number of active Forum participants individually as not brigading. This is less likely to create a sufficient mass effect to mislead observers about the community’s range of views.
I think about it this way. If a post was written critically about me, I would suspect 5-10% of people that know me in the community to see it, and 0.5% to comment. If I reach out to everyone I have ever been friendly with, I expect these numbers would be 50% and 5%, respectively. In other words, there would be 10x more comments defending me if I reach out to friends than if I don’t.
I think for independent observers, reading comments in fictional scenario above, it’s useful to know whether comments were of unsolicited or not.
A small group of strong-vote wielding users can significantly effect the course of discussion on the Forum through their voting.
Totally agree. Why would the same not be true of comments?
I think “a reasonable number of active Forum participants individually” is doing some real work here—“everyone I have ever been friendly with” would not count. I think there is usually value in having people who know the subject well participating in the comments, and by your math there is a good chance that zero or one of the ten people best positioned to provide a sympathetic perspective would even see the post organically. A reasonable number would depend on the circumstances, but I was thinking more ~3 acceptances?
One could argue that these individuals should disclose their status as solicited commenters. But people comment on the Forum for any number of reasons, obvious and inscrutable, so I can’t find a sufficient rationale for singling out a few solicited commenters. There’s no norm, for instance, for friends of a post’s author to self-identify themselves as such.
I am relatively less worried about a few commenters skewing the course of discussion (as opposed to strongvoters) for two reasons. The first is that comments have substance that can be evaluated as convincing or non-convincing. The popularity of that substance can be evaluated via up/down and agree/disagree voting, which provides some check on unrepresentative comments appearing to be consensus. Second, at least regulars have a decent sense of who is who; if someone who is an infrequent commenter starts on a commenting spree defending person X, we have a pretty good idea that they are motivated by some sort of external reason and can adjust accordingly.
I think the qualifier “a reasonable number of active Forum participants” in my comment is doing some real work and wouldn’t be met if you asked “everyone [you] have ever been friendly with”—even if we add in an implied limitation to current Forum participants.
Let’s take a case in which I invited my hypothetical friends Abel, Baker, and Charlie to participate in a thread that was critical of me. I think there is value in having some people who know the person well who is subject to the controversy present in the conversation. An invitation increases the likelihood of having those voices present; if the base rate of people even seeing the post is 5-10% per your example above, there’s a good chance that zero of the ten community members best situated to provide a favorable perspective on the subject will even see the post—much less decide to comment.
On the whole, I think the presence of Abel, Baker, and Charlie in the comments would be net positive. I’m sure it is exhausting to feel the need to respond to a post that is critical of you and all the comments thereunto, and asking for help can be appropriate. Even if all accept, it’s only three voices, and the community is capable of evaluating the substance of what they say and reacting accordingly. In contrast, with votes there is no ability to evaluate whether the votes are based on solid reasoning or instead represent a voter’s predisposition toward the subject of the post.
I see the point that Abel, Baker, and Charlie could say that I asked them to comment. However, I think they should be part of the conversation, and expecting them to flag themselves gives the impression that they are true brigadiers. People have all sorts of incentives and motivations for posting, and I’m not convinced this motivation should be singled out for per se special disfavor.
In this particular case, most active Forum participants would have seen the post given the prominence of the Owen situation and the Time article. And participation rates are likely to be high due for various reasons. So I see concerns about a few “recruited” members of the community potentially upsetting the balance as less pressing than in the hypothetical case in which the post was critical of you.
Admittedly, I have seen some possible evidence of strategic voting here (e.g., a strong downvote on a post with lots more agrees than disagrees is a yellow flag for me, certain posts got a good bit of karma quickly and were then voted down to near-zero or even negative karma). But that would likely be coming from people who are pretty active on the Forum anyway (as evidenced by their powerful strongvotes), so the nexus between “recruitment” and that strategic voting is doubtful to me.
OK, that seems more reasonable. Not sure I agree, but at least this seems doable. Before it just seemed like you were saying that people shouldn’t be allowed to share a post with friends and say to vote and comment according to their conscience.
This is food for thought. I will think about it and may update my policy.
This is correct. What I am talking about is brigading.
Astroturfing and troll farms are only similar in the mechanism behind their ability to distort public opinion. That mechanism is: People are influenced by the tone and volume of comments they read.
Yes, this is brigading. There are things you can do mitigate this brigading effect, for example: (1) Begin comments with “I am here from …” or “This post was shared by...”. (2) Commenters acknowledge, when asked, that the post was shared with them.
Take your case (i.e. Ben Pace’s post on nonlinear), neither (1) or (2) was done. In fact, I found myself needing to comment alluding to this effect, after I confirmed this was the case with one of your collaborators.
There are some grey areas here:
Inviting participation from people who are not part of the relevant community is clearly brigading. Unless they abstain from voting and clearly disclose their origin, they would be masquerading as community members and giving a false impression of the community’s views.
Inviting participation from people who are part of the relevant community presents a closer question. There’s still a risk of creating a misleading impression of the community’s views, but there isn’t the astroturf-like presentment of inauthentic views as community-member views that classic brigading presents.
For instance, I would generally view reaching out to a reasonable number of active Forum participants individually as not brigading. This is less likely to create a sufficient mass effect to mislead observers about the community’s range of views. This may be contingent on those individuals following the next bullet point.
I suggest that people who have been “recruited” in this manner should ordinarily refrain from strong-voting on the post, either up or down. A small group of strong-vote wielding users can significantly effect the course of discussion on the Forum through their voting. If someone is close enough to the subject of the post to solicit for support without constituting brigading, they are likely too close to the subject to be casting strong up/downvotes on that post.
I think about it this way. If a post was written critically about me, I would suspect 5-10% of people that know me in the community to see it, and 0.5% to comment. If I reach out to everyone I have ever been friendly with, I expect these numbers would be 50% and 5%, respectively. In other words, there would be 10x more comments defending me if I reach out to friends than if I don’t.
I think for independent observers, reading comments in fictional scenario above, it’s useful to know whether comments were of unsolicited or not.
Totally agree. Why would the same not be true of comments?
I think “a reasonable number of active Forum participants individually” is doing some real work here—“everyone I have ever been friendly with” would not count. I think there is usually value in having people who know the subject well participating in the comments, and by your math there is a good chance that zero or one of the ten people best positioned to provide a sympathetic perspective would even see the post organically. A reasonable number would depend on the circumstances, but I was thinking more ~3 acceptances?
One could argue that these individuals should disclose their status as solicited commenters. But people comment on the Forum for any number of reasons, obvious and inscrutable, so I can’t find a sufficient rationale for singling out a few solicited commenters. There’s no norm, for instance, for friends of a post’s author to self-identify themselves as such.
I am relatively less worried about a few commenters skewing the course of discussion (as opposed to strongvoters) for two reasons. The first is that comments have substance that can be evaluated as convincing or non-convincing. The popularity of that substance can be evaluated via up/down and agree/disagree voting, which provides some check on unrepresentative comments appearing to be consensus. Second, at least regulars have a decent sense of who is who; if someone who is an infrequent commenter starts on a commenting spree defending person X, we have a pretty good idea that they are motivated by some sort of external reason and can adjust accordingly.
I think the qualifier “a reasonable number of active Forum participants” in my comment is doing some real work and wouldn’t be met if you asked “everyone [you] have ever been friendly with”—even if we add in an implied limitation to current Forum participants.
Let’s take a case in which I invited my hypothetical friends Abel, Baker, and Charlie to participate in a thread that was critical of me. I think there is value in having some people who know the person well who is subject to the controversy present in the conversation. An invitation increases the likelihood of having those voices present; if the base rate of people even seeing the post is 5-10% per your example above, there’s a good chance that zero of the ten community members best situated to provide a favorable perspective on the subject will even see the post—much less decide to comment.
On the whole, I think the presence of Abel, Baker, and Charlie in the comments would be net positive. I’m sure it is exhausting to feel the need to respond to a post that is critical of you and all the comments thereunto, and asking for help can be appropriate. Even if all accept, it’s only three voices, and the community is capable of evaluating the substance of what they say and reacting accordingly. In contrast, with votes there is no ability to evaluate whether the votes are based on solid reasoning or instead represent a voter’s predisposition toward the subject of the post.
I see the point that Abel, Baker, and Charlie could say that I asked them to comment. However, I think they should be part of the conversation, and expecting them to flag themselves gives the impression that they are true brigadiers. People have all sorts of incentives and motivations for posting, and I’m not convinced this motivation should be singled out for per se special disfavor.
In this particular case, most active Forum participants would have seen the post given the prominence of the Owen situation and the Time article. And participation rates are likely to be high due for various reasons. So I see concerns about a few “recruited” members of the community potentially upsetting the balance as less pressing than in the hypothetical case in which the post was critical of you.
Admittedly, I have seen some possible evidence of strategic voting here (e.g., a strong downvote on a post with lots more agrees than disagrees is a yellow flag for me, certain posts got a good bit of karma quickly and were then voted down to near-zero or even negative karma). But that would likely be coming from people who are pretty active on the Forum anyway (as evidenced by their powerful strongvotes), so the nexus between “recruitment” and that strategic voting is doubtful to me.
OK, that seems more reasonable. Not sure I agree, but at least this seems doable. Before it just seemed like you were saying that people shouldn’t be allowed to share a post with friends and say to vote and comment according to their conscience.
This is food for thought. I will think about it and may update my policy.