Why Eliezer Yudkowsky receives more upvotes than others
One of the reasons for why Yudkowsky is being drastically upvoted is of course that he’s often, fasten your seatbelts, brilliantly right (whoever reads my comments knows that I am not really the most frenetic believer, so I think I can say this without sounding cultish). But others are too, so is Less Wrong a cult? Nah! There is a simple explanation for this phenomenon:
As you can see, there are already 13 people who subscribe to his Less Wrong feed via Google Reader. And there are many other ways to subscribe to a RSS feed (which is not the only way to follow his mental outpourings anyway), so the number of people who follow every post and comment is likely much higher.
That’s why most of his comments receive more upvotes than other comments. It is not because he is a cult leader, it’s just that his comments are read by many more people than the average comment on Less Wrong. There are of course other causes as well, but this seems to explain a fair chunk of the effect.
Also consider that I’m often upvoted (with a current Karma score of 1959) and I do not keep quiet regarding my doubts about some topics directly related to Yudkowsky and the SIAI. How could this happen if Less Wrong was an echo chamber?
I just wanted to let you know, because I have been wondering about it in the past.
People subscribing to Eliezer’s comments is what you would expect if they were cultists that wanted to make sure they don’t miss a word from Mr. cult leader!
I think LW is not a cult, but please don’t make bad arguments for why LW is not a cult.
Agreed, I think there is a little bit of a personality-cult effect. I don’t think it’s the main contributor though, and you have to consider that EY has a better incentive than most of us to carefully proofread his comments before submitting them. Plus, he’s been on the site for longer.
Is it in fact true that his comments are unusually upvoted? A quick look at Anna’s comments shows at least a sufficiently similar picture to doubt the outlier status you assume in this post. I suggest writing a script to collect data from user pages to learn more before taking this hypothesis for granted and trying to explain it (this will also give a better quantitative picture).
Note in addition that such analysis should be limited in scope by time. Early posts and comments would completely obliterate anything that could be learned about upvote patterns if they were considered.
Posts, early or not, shouldn’t be considered at all in estimating comment karma.
Yet serve as a ‘container’ for comments.
I question the explanandum. It wouldn’t surprise me to learn that I was more upvoted, but it wasn’t my impression that I was.
Slightly, but not excessively. Certainly not in disproportion to comment quality.
This effect should be at least somewhat attenuated by the activation energy required to click through to the actual discussion.
Eliezer also has an often recognisable writing style, meaning that the antikibitzer doesn’t work.
Also, this post doesn’t appear in the recent posts column. Why?
I always assumed it was because he was a malevolent AI.