My experience with karma-based community systems in the past leads me to believe frequently people upvote things on the basis of who wrote them. There does exist a kibitzer, but 1) one loses valuable information in the process and 2) the people swayed by political values like this are the not very likely to use it.
I suppose this post is evidence that the opposite direction is true as well.
Edit: I suppose “tribalistic” is a word with closer connotations than “political.”
I think there’s probably merit to your point, although I’d like to think that’s not as true for LessWrong. But I probably don’t have enough experience here to have a reliable estimate either way. And even if I did, causality is always difficult to determine, etc. But there are certainly examples where what you say is not the case, although they are obviously not the majority. Often I think that people are more inclined to downvote someone like Eliezer who has the Karma to spare it—had I made Eliezer’s comment, I highly doubt it would have been downvoted as strongly.
Edit: Perhaps “has the Karma to spare it” is not the best phrasing, but rather “Is also wildly upvoted all the time and won’t feel bad.”
Why all the karma bashing? Yes, absolutely, people will upvote or downvote for political reasons and be heavily influenced by the name behind the post/comment. All the time. But as far as I can tell, politics is a problem with any evaluation system whatsoever, and karma does remarkably well. In my experience, post and comment scores are strongly correlated with how useful I find them, how much they contribute to my experience of the discussion. And the list of top contributors is full of people who have written posts that I have saved forever, that in many cases irreversibly impacted my thinking. The fact that EY is sometimes deservingly downvoted is a case in point. The abuse described in the original post is unfortunate, but overall the LessWrong system does a difficult job incredibly well.
Often I think that people are more inclined to downvote someone like Eliezer who has the Karma to spare it
I’ve noticed a similar (or worse) effect in my own voting patterns. More than once I’ve neglected to upvote something I thought was good because it had very high karma and was posted by a very-high-karma user. My thought process is something like, “Lukeprog has already gotten more karma for this post than many users have altogether. Does he really need another upvote?”
The largest bias in voting I’ve noticed in my own thinking is when someone else has voted down a comment, and I see no reason it should be either up or down voted. It is very difficult for me to not upvote it to counteract the, in my opinion undeserving, downvote.
I’ve seen voting patterns like that and my example described elsewhere as being, not exactly biases, but a product of there being two or more ways people use the karma system. AFAICT, different people decide to upvote or downvote based on the answer to one of the following questions:
*Is this comment above my threshold of “good enough to upvote,” below my threshold of “bad enough to downvote,” or in between?
*Does this comment need more or fewer karma than it currently has?
*Does having posted this comment make the poster deserving of another karma point (or “more karma points than ey has already gotten from it)?
People who see the karma system more as a tool for ranking comments will probably use question 1 or 2. People who see the karma system as a tool for ranking users will use question 3. Also, people who ask question 1 will probably use the anti-kibitzer, while people who use question 2 probably will not.
I had considered one and two before, and strongly prefer one, but it hadn’t occurred to me that people might operate by three. And I don’t use the anti-kibitzer, mostly because I’m too lazy to get it working, and believe (probably erroneously) that I am not influenced by that. Also, I’d constantly be turning it on and off and then on again, so much so that it would only be frustrating.
I mostly use one, with occasional instances of two. For instance, I never downvoted the Popper troll from a few months ago, because every comment of eirs I saw was already at −20 and downvoting seemed pointless. My not upvoting Lukeprog is mostly two with a little bit of three. I used to use the anti-kibitzer, but it was preventing me from getting to know the other users, and long conversations with three people in them got confusing. I kept having to turn it on and off, and in the browser I had at the time that meant scrolling all the way to the top of the page and losing my place.
I suspect the same thing. Actually I think there are a couple of biases—the name brand association you describe and also an affect I have noticed in myself where I feel inclined to upvote posts that have been upvoted a lot. Following the herd I guess.
If a “famous” poster who regularly accrues a couple dozen or more Karma a day just from popular comments they post would post for a few days under a pseudonym (but otherwise do not post any differently then they would have done) we might get some data about the former affect as you describe it.
My experience with karma-based community systems in the past leads me to believe frequently people upvote things on the basis of who wrote them. There does exist a kibitzer, but 1) one loses valuable information in the process and 2) the people swayed by political values like this are the not very likely to use it.
I suppose this post is evidence that the opposite direction is true as well.
Edit: I suppose “tribalistic” is a word with closer connotations than “political.”
Aside: Every time someone says “would you elaborate,” I think of Derrida.
I think there’s probably merit to your point, although I’d like to think that’s not as true for LessWrong. But I probably don’t have enough experience here to have a reliable estimate either way. And even if I did, causality is always difficult to determine, etc. But there are certainly examples where what you say is not the case, although they are obviously not the majority. Often I think that people are more inclined to downvote someone like Eliezer who has the Karma to spare it—had I made Eliezer’s comment, I highly doubt it would have been downvoted as strongly.
Edit: Perhaps “has the Karma to spare it” is not the best phrasing, but rather “Is also wildly upvoted all the time and won’t feel bad.”
Why all the karma bashing? Yes, absolutely, people will upvote or downvote for political reasons and be heavily influenced by the name behind the post/comment. All the time. But as far as I can tell, politics is a problem with any evaluation system whatsoever, and karma does remarkably well. In my experience, post and comment scores are strongly correlated with how useful I find them, how much they contribute to my experience of the discussion. And the list of top contributors is full of people who have written posts that I have saved forever, that in many cases irreversibly impacted my thinking. The fact that EY is sometimes deservingly downvoted is a case in point. The abuse described in the original post is unfortunate, but overall the LessWrong system does a difficult job incredibly well.
I’ve noticed a similar (or worse) effect in my own voting patterns. More than once I’ve neglected to upvote something I thought was good because it had very high karma and was posted by a very-high-karma user. My thought process is something like, “Lukeprog has already gotten more karma for this post than many users have altogether. Does he really need another upvote?”
The largest bias in voting I’ve noticed in my own thinking is when someone else has voted down a comment, and I see no reason it should be either up or down voted. It is very difficult for me to not upvote it to counteract the, in my opinion undeserving, downvote.
I’ve seen voting patterns like that and my example described elsewhere as being, not exactly biases, but a product of there being two or more ways people use the karma system. AFAICT, different people decide to upvote or downvote based on the answer to one of the following questions:
*Is this comment above my threshold of “good enough to upvote,” below my threshold of “bad enough to downvote,” or in between?
*Does this comment need more or fewer karma than it currently has?
*Does having posted this comment make the poster deserving of another karma point (or “more karma points than ey has already gotten from it)?
People who see the karma system more as a tool for ranking comments will probably use question 1 or 2. People who see the karma system as a tool for ranking users will use question 3. Also, people who ask question 1 will probably use the anti-kibitzer, while people who use question 2 probably will not.
I had considered one and two before, and strongly prefer one, but it hadn’t occurred to me that people might operate by three. And I don’t use the anti-kibitzer, mostly because I’m too lazy to get it working, and believe (probably erroneously) that I am not influenced by that. Also, I’d constantly be turning it on and off and then on again, so much so that it would only be frustrating.
I mostly use one, with occasional instances of two. For instance, I never downvoted the Popper troll from a few months ago, because every comment of eirs I saw was already at −20 and downvoting seemed pointless. My not upvoting Lukeprog is mostly two with a little bit of three. I used to use the anti-kibitzer, but it was preventing me from getting to know the other users, and long conversations with three people in them got confusing. I kept having to turn it on and off, and in the browser I had at the time that meant scrolling all the way to the top of the page and losing my place.
I suspect the same thing. Actually I think there are a couple of biases—the name brand association you describe and also an affect I have noticed in myself where I feel inclined to upvote posts that have been upvoted a lot. Following the herd I guess.
If a “famous” poster who regularly accrues a couple dozen or more Karma a day just from popular comments they post would post for a few days under a pseudonym (but otherwise do not post any differently then they would have done) we might get some data about the former affect as you describe it.