[meta] Future moderation and investigation of downvote abuse cases, or, I don’t want to deal with this stuff
Since the episode with Eugine_Nier, I have received three private messages from different people asking me to investigate various cases of suspected mass downvoting. And to be quite honest, I don’t want to deal with this. Eugine’s case was relatively clear-cut, since he had engaged in systematic downvoting of a massive scale, but the new situations are a lot fuzzier and I’m not sure of what exactly the rules should be (what counts as a permitted use of the downvote system and what doesn’t?).
At least one person has also privately contacted me and offered to carry out moderator duties if I don’t want them, but even if I told them yes (on what basis? why them and not someone else?), I don’t know what kind of policy I should tell them to enforce. I only happened to be appointed a moderator because I was in the list of top 10 posters at a particular time, and I don’t feel like I should have any particular authority to make the rules. Nor do I feel like I have any good idea of what the rules should be, or who would be the right person to enforce them.
In any case, I don’t want to be doing this job, nor do I particularly feel like being responsible for figuring out who should, or how, or what the heck. I’ve already started visiting LW less often because I dread having new investigation requests to deal with. So if you folks could be so kind as to figure it out without my involvement? If there’s a clear consensus that someone in particular should deal with this, I can give them mod powers, or something.
Thanks for your work! Much appreciated.
Thanks!
Given that Kaj_Sotala was apparently the only active moderator, it probably makes sense to at least try to get a new one. We should nominate new moderators, and one of the current moderators can confirm that the people are interested and appropriate for the role.
I would nominate Viliam_Bur. I went through the recent comment history of the all the top 30 contributors, and Viliam has many recent comments and a polite style of commenting that I think is a marker for someone being a good moderator.
I agree. Viliam seems to be currently the most active of the people who would fit the role, if he’s willing to take it.
Thanks for the trust. I can imagine doing unplesant decisions based on data. I am not sure how to get those data; but I guess I would speak with developers and ask them to make me some database reports. Or Kaj would explain me what he did.
I accept the nomination. I am okay with doing this either alone or with someone else—would slightly prefer to have a second opinion, but could act without it, too.
I’ve exchanged e-mails with jackk and asked him to pull out voting data for me. In other words, spoken with developers and asked them to make some database reports. :-)
Here is a link to the start of an exchange I had with Kaj_Sotala:
http://lesswrong.com/lw/kfq/moderator_action_eugine_nier_is_now_banned_for/b2f2
My request is pretty simple: I am pretty clearly the victim of a mass downvoting; I would like to have the downvoting reversed. For reasons which are not entirely clear, Kaj_Sotala wouldn’t do it. Thank you for your assistance.
Thank you for the report. I have already received two reports (via private message), so you are the third.
If, and only if, a formal official policy is hammered out and clearly stated somewhere, I’d be willing to volunteer to enforce it.
Actually, I was going to recommend against a detailed formal policy. Codified rules seems like an invitation for passionate arguments about Stuff That Doesn’t Matter™. If we need a rule, how about:
If you are acting in a way incongruent with the greater harmony of LessWrong, a moderator may private message to ask you to stop. If you don’t stop, you may be banned.
Also have a stated social norm:
Don’t downvote comments because of the author. There’s no hard and fast rule, but if you’ve downvoted someone more than 8 times in one day or read through someone’s comment history and downvoted past the first page, you are doing something wrong.
In the specific case of mass downvoting, if a moderator gets a complaint I think the typical procedure should be:
Moderator private messages the alleged downvoter a non-accusatory FYI letting them know that someone felt they were heavy on the thumbs-down and it might be a good idea to take a breather
If vindictive downvoting continues, look into the logs, do some quantitative analysis, and evaluate. Most probable conclusion is that the downvoter isn’t being vindictive. But if, regrettably, the downvoter appears to have a bullying voting pattern, the moderator should private message them “Really, you should stop. Even if there are good reasons for all of your downvotes, you shouldn’t continue just because it looks bad. There are plenty of other people are around who can pick up the slack if you bow out of downvoting.”
If abusive downvoting continues, give a final warning, then ban.
But before you complain to a moderator: Please consider the possibility that your comments are just bad, that someone might disagree with you for legitimate reasons, and that negative feedback is information positive.
It’s not wrong if most of what the user writes is bad. The wrong thing is to act on an incorrect judgement, something that won’t be supported by idealized community. The heuristic you suggest limits influence, which guards against consequences of overconfidence. In some cases, you can see that there is no mistake.
I tend to agree with the first statement as a rule of thumb. If you’re reading and downvoting 8 postings in a day that you think are not worth reading (apparently), it seems like you’re taking it upon yourself to punish that person, whereas I think it is better if we try to read what we consider to be valuable, and if we happen across some writing that seems bad, sure, critique it with a −1 if that seems worthwhile, but don’t go on a jag reading all the bad (by your standards) writing you can find and downvoting it.
Harmony is near the bottom of the list of values I want to see enforced. Probably below oral hygiene.
I invite suggestions for improved rule phrasings which avoid objectionable or controversial words like harmony :-)
I invite suggestions for improved rule phrasings that avoid objectionable or controversial words like harmony :-)
I am not opposed to a moderator having power like what you describe, but I wouldn’t want to be them.
Could you imagine writing a proposal for a policy that you would like to enforce yourself?
I specifically don’t want to write it myself. I would like to minimize angles from which I may be accused of self-serving moderation.
This should happen regardless.
How about this as a policy:
1) You can’t advocate violence or criminal acts.
2) You can’t engage in what the moderator determines to be massive down voting of a single person.
3) You can’t create posts that the moderator determines are not beneficial to the LessWrong community.
4) The moderator has the right to ban posts or users that violate (1), (2), or (3).
I would add that massive downvoting is about voting down a post based on who wrote it instead of voting it down based on the content of the post.
I do think that there are valid discussion about the usage of drugs such as modafinil that edge on the border of illegality. The LW scholar thread that allows requesting scientific papers could also be interpreted as violating copyrights and therefore fall in that category. I don’t have trouble with someone advocating to drive faster than the speeding limit. Having private contractors that fight in wars can violate military laws that all combatants have to wear military uniforms. Homosexuality happens to be a criminal act in some countries.
Instead of saying “criminal act” I would speak about actions that are likely to be persecuted as felonies in Western states if they become known to the authorities.
How about “violent, abusive, or fraudulent acts”? That seems to cover most of what we’d want to ban.
I wouldn’t be as happy with limiting it to felonies; first of all, I know just enough about law to know that what is or isn’t a felony is a complicated and potentially counterintuitive question and likely varies substantially by jurisdiction; and second, there’s a lot of things that I wouldn’t want people advocating but aren’t violent or felonious. Doxxing your pseudonymous critics, for example, is legally a fairly light gray but strikes me as very bad behavior when seen through the lens of blog policy.
Starting a war is certainly a violent act. Do you want to prevent that kind of discussion on LW? No one is allowed to write a post that the US should start a war against Iran?
Well, you could modify that with “personally” if you want to be able to advocate state violence. Honestly, though, I’d be just as happy if that was forbidden; the recent discussion of the violence in Gaza certainly hasn’t seemed very productive.
The problem is when you forbid exactly one side of the argument. Forum rules that allow criticism of a Israeli attack on Gaza but that don’t provide the ability to defend it, can lead to trouble when some people find their views unfairly censored.
On the other hand I think the incident that lead to us having a rule against advocation of violence is not covered in the provision of “personally”.
“Violent crime, abuse, or fraud”, then.
Though I’d note that when you’re discussing a war or violent political conflict, supporting any acts of any side can be construed as supporting violence, so this doesn’t look too politically slanted. It bars advocacy of aggressive war, terrorism, or violent revolution, but given that this isn’t a politics board trafficking in expansionist nationalism or exotic revolutionary ideologies, and that one of the fastest ways to mindkill someone is to advocate physically killing them or their buddies, I’m okay with that. (It’s also obviously noncentral to the rule, and I maintain that some degree of vagueness in forum policy is necessary if you want to get stuff done without every policy issue degenerating into unproductive trivia.)
You don’t need to support Hamas to criticize Israeli action in the region.
Okay, that looks fine to me.
It makes sense to imagine what the rule actually does in practice. There are certain actions like killing your neighbor where it’s perfectly fine to allow moral arguments about why killing your neighbor is bad but still forbid people from advocating killing your neighbor. That’s because we have a consensus that killing your neighbor is bad.
If you start banning the advocation of violence in a political debate where one side favors violence and the other isn’t you are set up for drama.
No, but—to move back to something a little less topical—you may recall that criticism of American action in Afghanistan and Iraq circa 2001 − 2011 was seen in certain circles as implicit support of Islamist violence. It wasn’t, of course, but if you’re trying to avoid drama you need to take perception into account as much as reality.
In this case, though, the spirit of the rule is less “avoid political drama”—we have a weaker norm against politics for that—and more “don’t advocate things that make us look like we’re all about to go Ted Kaczynski on someone’s ass”, which is why I feel that discussing war in its context is noncentral.
If you start banning people on forum for positions that they don’t explicitly argue but that you think they argue implicitly because of tribal associations than you have problems.
Yes. It wasn’t by any reasonable rational standard that a forum moderator is supposed to use to make moderating decisions. Don’t let yourself be mindkilled. Arguments aren’t soldiers. It’s quite easy to make an argument against invading other countries without arguing in favor of violence.
On reflection, you’re right; a prohibition on advocating violence doesn’t extend that far. Though I’d appreciate not having memes from the politics sequence flung at me.
Points 1 and 2 are reasonably clear. Point 3 is unhelpfully vague. If I were moderator, I would have no idea how far that pushes, and as a commenter I wouldn’t have a lot of insight as to what to avoid.
I don’t mind giving a catch-all authority to a moderator, but if there are specific things you have in mind that are to be avoided, it’s probably better to enumerate them.
I would add an explicit “nothing illegal, nothing personally threatening” clause. Those haven’t been problems, but it seems better to remind people and to make clear we all agree on that as a standard.
NancyLebovitz seems like she would make a good moderator.
Well, it is clear that modding a forum is not for you, thanks for trying, if ineffectually (Eugene never stopped downvoting), to make this place better. I appreciate you stepping in when no other mod would, and stepping down when many others in this situation would keep making mistakes. Administration and content creation are completely different skill sets, and I hope this negative experience does not deter you from posting more insightful content here in the future.
Thanks again!
Thanks!
He should have: banned users have been blocked from voting as of a patch deployed on July 24th. (I also verified that the patch works by creating a new account, banning it, and trying to vote with it.)
Thanks! It is perfectly conceivable that he is not the only one like that (defection is contagious), or that I made a mistake.
You are selling yourself short. The fact that you have a 94% positive karma rating counts in your favor. Being a MIRI research associate also means that you have a former affiliation which qualifies you.
But if you don’t want the job, I think the straightforward way is to have a procedure where differently people volunteer to be moderators and then Eliezer as main person responsible for this website picks on of those people to transfer to them moderator powers. Preferably a person with >1000 karma and a high positive karma rating.
Good idea. But this volunteering shouldn’t happen here in some subthread. Instead somebody—probably Kaj—should create a volunteering post where anybody can volunteer and with one top-level comment to discuss the voting or candidates.
If you do this, you’ve just declared people downvoted by Eugene to be ineligible for moderator.
I was one of Eugine’s targets (though to a lesser extent than some). I have ~6200 karma and ~90% positive.
I think David Gerard was one of Eugine’s targets. He has >8000 karma and ~75% positive.
Stuart Armstrong was (so at least one of his comments suggests) one of Eugine’s targets for a bit. He has >13000 karma and ~90% positive.
Cyan says he was one of Eugine’s targets. He has ~5000 karma and ~90% positive.
Ialdabaoth was one of the first people (maybe the first?) to complain of Eugine’s abuse. He has ~2000 karma and ~75% positive.
In the course of looking out prominent contributors who were hit by Eugine, I didn’t find anyone who looked like they might have been high-karma and high-positive-proportion without Eugine’s attacks.
[EDITED to add: For the avoidance of doubt, the above is not intended to be a complete list, nor a list of the highest-karma, best-positive-fraction, most-abused-by-Eugine, or any other such selection. It’s just some examples I noticed.]
I currently have 685 karma at 65% positive. Assuming 200 downvotes from Eugine, upvotes of X, and legitimate downvotes of Y, then X/(X+Y+200) = 65% and X-Y-200=685. Solving gives Y=599, X=1484. So without Eugine I would have had 885 karma at 71% positive.
If I had 300 downvotes from Eugine, that would be Y=511, X=1496 which is 985 karma at 75% positive.
Not necessarily. There is only so much influence a surge of undeserved downvotes can have over a consistently high-quality, long-term contributor. After all most people do vote more or less fairly, otherwise we wouldn’t bother with a karma system.
As gjm summarizes that’s not the case. I don’t think it should automatically be the case the the people downvoted by Eugene are ineligible for moderator. However having been the candidate of a downvote attack doesn’t make you the best person to objectively decide how to punish the next mass downvoter, so it’s okay that people downvoted by Eugene score less well on that metric.
I want a moderator who’s not a controversial figure and who’s trusted to make objective decisions.
Do you mean a hypothetical future mass-downvoter, or actual Eugene?
Either one. Of course, I have no way to know if the statement is literally true, since I don’t have access to other people’s karma scores and downvote records, but Eugene downvoted enough that the effect is plausible. You should at least consider, when selecting moderators, whether the potential moderator was mass-downvoted and adjust the karma and karma ratio cutoffs appropriately.
Then commit to never working on another investigation, no matter what the request.
Yeah, this post is kinda my way of doing that.
I think it depends. If he’s going to commit to stop moderating, that’s fine. Moderators are volunteers, after all. But we should recognize that if he does this this basically means we have no moderator, and we should treat it as urgently as having no moderator should be treated. Because dealing with mass downvoters is only a tiny sliver of a moderator’s duties, it’s tempting to say “we don’t need to do anything about this because the moderator’s good most of the time”, which is like having a fire department that refuses to fight fires that take place in buildings over 10 stories tall.
Remember to give Kaj an upvote for this post :-)
Speaking of moderation. Anybody know why the “Radical idea: take over sick countries” thread was deleted?
If I remember right it was mostly downvoted, which make it a valid target for deletion. Taking over a country is also advocating violent war and we have rules against advocating practical usage of violence.
That said I don’t support it’s deletion.
I suggest giving clear and prominent notification on the consequences of mass downvoting in any “you have joined the list” message, and a similar notification email to existing members.
Also, update the “How should I use my voting powers?” section of the FAQ with the same information.
An “Offenses Warranting Banning” section would be appropriate as well. It should state that it is not intended to be exhaustive.
Strawman moderation process:
Get a complaint about vindictive downvoting
Do quantitative analysis on the downvoting
Based on the quantitative analyses, send a short message to relevant parties. Example note: “I confirmed that 12 of your comments were downvoted by Username123 on August 17th, which feels a bit much. I analyzed the voting logs, and I didn’t find enough evidence in Username123′s voting patterns to convince me they they are going after you personally. That doesn’t mean that Username123 is not going after you personally, but the evidence I see does not meet the high bar required for me to do anything moderatory about it. If you want I can send a private message Username123 that the 12 votes downvotes on August 17th seem a bit excessive.”
If the aggrieved party does not feel that their grievances have been addressed, try to talk it out. You are under no obligation to make sure they leave happy. Some of your decisions will have to be unfair and arbitrary, and that is life.
If Username123 is acting unacceptably, message them repeatedly and more and more explicitly. After a few messages, tell them that if they continue this behavior they will be banned. If they continue, ban them.
Keep the criteria for what constitutes abusive behavior consistent and quantitive, but opaque. The criteria are neither based on your arbitrary whims nor transparent enough to be gamed or become a rathole of debate. Anyone approaching the line will have received enough warning messages to know and turn away.
From the previous discussions of this, I still maintain that a possibly minor code change would be worth investigating to reduce the moderator load: only allow upvotes to affect user karma, instead of both upvotes and downvotes. The overwhelming complaints seem to be not so much that comments are downvoted, but that user karma can drop precipitously as a result of a downvoting campaign. Preventing negative votes from affecting karma greatly reduces that.
Where we disagree is interesting for discussion’s sake; where we agree is interesting for community building.
That said, I don’t have a clear idea on whether we need to do something about ‘inability to post due to negative karma’ if this change is implemented.