This made me laugh. Also, I knew someone would do this the second Eliezer proposed new boundaries for Lesswrongers.

The only problem I can think of with this experiment is that your post could have been deleted for one of your more overt offenses, but it took until the time it was actually deleted for someone to actually get around to deleting it, especially with all the controversy. You have evidence that it was attacking Eliezer that broke the camel's back, but maybe not strong evidence. I don't think you can get anything conclusive from this.

What you should have done is post several different posts each violating one of the above rules and wait for something to happen, then post your conclusions in a day or two. (sigh) I guess it's too late for that now.

When you have power, you generally don't have the option to not be seen as having power.

If you have the power to stop X, and you do nothing, then you are observed to tolerate X. That is what "tolerating" is — "I could do something about X, but I don't" — whether X is "proposals of violence on my web forum" or "squirrels raiding the bird feeder in my back yard".

If you delegate capability Y to Alice, and Alice does action A with that capability, and you do nothing about it, then you are observed to accept A being done. This includes where "A" is "post threatening things on the web" and Y is "the ability to submit posts to your web forum".

You cannot beg off responsibility for power that you actually do possess. "Alice put a post on your web forum saying that all green-eyed, black-haired people are dirty wiggins and maybe we should bisect them all! Are you really OK with that!?" While you are running the forum, you do not have the option of credibly saying, "That's nothing to do with me." You either tolerate the proposal to bisect wiggins, or you do not tolerate it — in which case you delete the post, ban Alice, or something of the like.

As far as I can tell, you would do well to realize that there is a whole Web out there that is not Less Wrong, and you can post your ideas on quite a lot of it. What you are doing here is what the Wikipedia folks call disrupting the project to make a point, and indicates that you value your freedom to post violent speculations on other people's web forums much more highly than their freedom to have a web forum that does not endorse or promulgate your violent speculations.

In the Supreme Court case Jacobellis v. Ohio, the courtroom judges were to rule on whether or not obscenities were protected by the 1st Amendment. The resulting decision was that all speech should be protected in the public square except for hardcore pornography. Justice Potter Stewart, in writing the concurring opinion, when writing about what exactly constitutes hardcore pornography, stated this:

I shall not today attempt further to define the kinds of material I understand to be embraced within that shorthand description; and perhaps I could never succeed in intelligibly doing so. But I know it when I see it, and the motion picture involved in this case is not that.

And thus, "I know it when I see it" became a legal standard.

I've noticed this when I started moderating a subreddit on Reddit.com. You can create a set of rules to follow, but it's entirely possible to follow those rules while creating a hostile atmosphere that goes against the culture that you're trying to cultivate. As an example, running a community that's dedicated to pointing out the flaws of a particular philosophy (for instance, feminism) will attract a lot of people who are belligerent towards that philosophy. You may expect the userbase to post intelligent and/or funny article relevant to critiquing feminism, but no specific set of concrete rules will prevent people from posting something that expresses belligerent hatred. Ultimately, good moderation requires judgment calls, and trust from your moderators.

I think that the same thing is at play here. There is no precise measurement for how intelligent a particular comment is (well, maybe there is, but it would be virtually impossible to implement practically), and so the standard for appropriateness ultimately requires moderator judgment. For a good community to thrive, there is no way around the "I'll know it when I see it" standard.

With that being said, I'm willing to guess that the reason for the rule against "violence" and "illegal activities" is that I've seen forums blow up with drama over certain topics: rape, sexual assault, violence, etc are all good ingredients to create forum-based drama. Ultimately, this boils down to "I'll know it when I see it", but I think that the heart of the law is to avoid sensitive topics that will destroy this forum's intellectual paradise.

With that being said, I'll await your solutions.

Downvoted for bragging about trolling here - that's not something I want to see encouraged.

It's also a form of dishonesty to request public feedback on a policy issue, then systematically ignore all feedback that disagrees with your predetermined decision.

Come on, his "request for public feedback" was explicitly phrased as only asking for things he didn't know:

In other words, the form of this discussion is not 'Do you like this?' - you probably have a different cost function from people who are held responsible for how LW looks as a whole - but rather, 'Are there any predictable consequences we didn't think of that you would like to point out, and possibly bet on with us if there's a good way to settle the bet?'

... in other words, he was warning you that if all you do was whine about how "censorship" was a word with negative connotations, he was going to ignore you because he was well aware of that. Nothing dishonest there.

I've designated kodos96 as a troll (note: user confessed to trolling) and will delete future posts and comments at will.

You present your table of results as if it is uncontroversial data based on which a reasonable unbiased person would draw the conclusion that the deletion was caused by the accusation of dishonesty. But the table already smuggles in your preferred causal hypothesis, representing the deletion as a reponse to the accusation.

A better way to present the data would be to have a column for your actions, and then a second column giving the amount of time between that action and the deletion. By your account, the deletion occurred about a day after you posted trollish comments and 60 seconds after you accused EY of dishonesty. You obviously think it is much more likely that the mods will act immediately than that they will act with a one-day lag, but I don't buy this, especially considering the fact that the one day during which the post wasn't deleted was Christmas. The mods reading, processing and deleting a thread based on a comment within 60 seconds of posting seems less likely to me than the mods not having noticed the multiple violations in that thread until the day after Christmas and deleting based on those violations.

Moreover, even if I concede that the likelihoods skew the other way, my prior for the "deleted due to violations of policy" hypothesis is substantially greater than my prior for the "deleted in a fit of pique" hypothesis, so even a greater likelihood for the latter hypothesis doesn't imply a greater posterior.

I would have preferred if you had simply stated that you suspect the deletion was due to your accusation of dishonesty. I would have disagreed, but wouldn't have downvoted. But I don't like the fact that you've constructed this pseudo-rational argument aimed at convincing us that your causal hypothesis is best supported by the evidence. So downvoted.