If the main Less Wrong codebase offered an option to hide the names of contributors and non-negative karma point totals (because negatives you may want to let your eyes just skip over), not revealing them until voted on (including a neutral vote)...
...then what would you think of only giving half a karma point for any votes made without obscuring the name and current total? (Likewise any votes changed after the reveal.) Fair? Or just annoying?
The half-point idea is an interesting one. I find it hard to think of genuinely good reasons for wanting to know how positive a score is before deciding whether to vote up or not, and I haven’t heard any good explanations from anybody else yet either. That’s to be contrasted with the multitude of bad reasons for wanting to know the score beforehand, which exploit common cognitive biases that we are all at least partially subject to. I do find the idea a bit heavy-handed, but I think it would make the community slightly less herd-like and a bit more rational.
Of course, you’d also have to change the code base so that once an initial blind vote has been made, any further changes would only register at the lower value.
As to whether it’s unfair, you could explain it as getting an EXTRA half point for your willingness to make a small sacrifice for the betterment of the community at large. You could even make it 1 point versus 1.5 points.
IMHO, unfair. But it would be interesting to have an option to see such a total, or to see how much karma is from info-obscured vs info-revealed votes.
Though I frequently change my votes as I reflect on them, so even reading anonymously I’d probably end up with all devalued votes. And this process of reflection would eliminate any debiasing effect of reading anonymously.
Of course, I put a lot of stock in character, so I don’t see the problem with occasionally voting up comments by better people.
Then add in an autocomplete box that I can use to guess the author’s name based off the comment. It’s fascinating trying to pick out the fist of an author.
Simply annoying from a usability point of view. It requires you to know too surely which posts you will want to vote on and which authors you’ll want to know; if you care about the value of your karmic vote you’ll wind up interfering with your enjoyment of the posts to preserve its value; etc.
If the main Less Wrong codebase offered an option to hide the names of contributors and non-negative karma point totals (because negatives you may want to let your eyes just skip over), not revealing them until voted on (including a neutral vote)...
...then what would you think of only giving half a karma point for any votes made without obscuring the name and current total? (Likewise any votes changed after the reveal.) Fair? Or just annoying?
The half-point idea is an interesting one. I find it hard to think of genuinely good reasons for wanting to know how positive a score is before deciding whether to vote up or not, and I haven’t heard any good explanations from anybody else yet either. That’s to be contrasted with the multitude of bad reasons for wanting to know the score beforehand, which exploit common cognitive biases that we are all at least partially subject to. I do find the idea a bit heavy-handed, but I think it would make the community slightly less herd-like and a bit more rational.
Of course, you’d also have to change the code base so that once an initial blind vote has been made, any further changes would only register at the lower value.
As to whether it’s unfair, you could explain it as getting an EXTRA half point for your willingness to make a small sacrifice for the betterment of the community at large. You could even make it 1 point versus 1.5 points.
that’s brilliant and hilarious. We should implement more de-biasing programs by exploiting biases.
Though I’d say by this scheme it should be 1 point versus 2, just to preserve the punchline.
Maybe have endorcements for the program by respected members? Pictures of attractive women marvelling at Bayes on little adverts? Make it opt-out?
IMHO, unfair. But it would be interesting to have an option to see such a total, or to see how much karma is from info-obscured vs info-revealed votes.
Just annoying.
Though I frequently change my votes as I reflect on them, so even reading anonymously I’d probably end up with all devalued votes. And this process of reflection would eliminate any debiasing effect of reading anonymously.
Of course, I put a lot of stock in character, so I don’t see the problem with occasionally voting up comments by better people.
Good idea Eliezer.
Then add in an autocomplete box that I can use to guess the author’s name based off the comment. It’s fascinating trying to pick out the fist of an author.
Simply annoying from a usability point of view. It requires you to know too surely which posts you will want to vote on and which authors you’ll want to know; if you care about the value of your karmic vote you’ll wind up interfering with your enjoyment of the posts to preserve its value; etc.