News for hackers. Stuff that matters.
Re:Experiment: No Comment Scores by pg on Saturday September 26, 2009 @01:15AM (#845077)
Voting patterns showed a lot of users voted to get a comment to what they felt was an appropriate score: they wouldn't up- or downvote something unless they felt its current score was too low or high respectively. But if comment scores aren't displayed, you won't need to anymore.
Re:Experiment: No Comment Scores by tc on Saturday September 26, 2009 @01:31AM (#845101)
Hmmm... I always considered that a feature rather than a bug. Only voting when the scores looked "wrong" was an optimization. How do I decide when to vote now? I have a feeling I'm going to settle into a pattern of voting to reorder comments and that I'm going to vote less overall.
Re:Experiment: No Comment Scores by nandemo on Saturday September 26, 2009 @03:09AM (#845209)
I also considered that a feature. I have 3 choices: -1, 0 and 1, and almost always I choose 0.

I do think that people make too much of karma and comments scores. But I don't understand why pg thinks this particular aspect is an improvement.

Re:Experiment: No Comment Scores by sev on Saturday September 26, 2009 @03:45AM (#845237)
I can see why the change is for the better: my vote (-1, 0, 1) should not be skewed/altered based on past events (others voting). Also, an "appropriate" total vote count is not something that an individual should be determining; it should be a collective thought. The only thing an individual should decide on is whether to award the comment a -1/0/1 purely based on the comment, nothing more. This change removes everything else (or so we assume: but considering the fact that the position on the page is still modified, it's not perfect, but it's close enough in my book).
Re:Experiment: No Comment Scores by borism on Saturday September 26, 2009 @04:49AM (#845265)
"The only thing an individual should decide on is whether to award the comment a -1/0/1 purely based on the comment, nothing more."

exactly

Re:Experiment: No Comment Scores by jacquesm on Saturday September 26, 2009 @06:53PM (#846231)
Since global karma (your collected karma over time) does not have any real function in the site as far as I can see you could lose it completely and not much of value would be lost, the way to look at it now is more 'relative', if you see a comment thread with a comment directly above the comment you're looking at that you think contributes less than the comment below it vote to lower one up.

After all, the biggest function by far of the votes is to influence the sort order, so the good stuff gets read by the people that skim. A vote is like a little bit of editorial power, you get to have a say in what goes 'above the fold'.

Re:Experiment: No Comment Scores by skolor on Saturday September 26, 2009 @10:32AM (#845504)
I always did the same thing, and I (so far) have been up/down voting far more. Instead of keying in on a specific score I feel appropriate for that comment, I'm keying in on just the three options: Up, Down, No change. Without having a visible score, I am no longer constrained by "If I up/down vote this, it will take it out of the range I want it to be in". Instead I'm just up/down voting as I want to.
Re:Experiment: No Comment Scores by gojomo on Saturday September 26, 2009 @02:25AM (#845161)
I think your previous guidance helped create the "vote toward a target score" behavior. You've spoken against piling-on voting, and your clipping of the negative range advances the idea: once a signal is 'far enough' in one direction, people should hold off.

I can see ways hidden scores may work for or against extreme totals. On the one hand, without the satisfaction of seeing a vote take immediate effect, overall voting may go down. But on the other, without the indication that a comment is already 'far enough' in a desired direction, charged comments may rack up unprecedented positive or negative values. I hope you can share info about which effect predominates.

I will miss the chance to quickly 'rescue' slightly-downvoted comments I don't particularly like or agree with, but feel have been unfairly squelched into nether-ratings by hair-trigger down-voters.

Re:Experiment: No Comment Scores by kniwor on Saturday September 26, 2009 @02:59AM (#845195)
In fact I will go as far as to say that the ideal interface would be one that let me specify what the ideal score of every comment should be and my vote should be counted as an upvote if its below that value and a downvote if its above that...
Re:Experiment: No Comment Scores by tome on Saturday September 26, 2009 @06:17AM (#845305)
Yes I've had this idea before too. I also think that the votes should be applied to the score in reverse order of the time at which they were made. This means that those who vote earlier end up having more effect on the final score, rather than none at all, as would be the case otherwise.
Re:Experiment: No Comment Scores by gdee on Saturday September 26, 2009 @04:50AM (#845267)
That would be a star rating system then. Supposedly they work worse.
Re:Experiment: No Comment Scores by tome on Saturday September 26, 2009 @06:15AM (#845304)
No, star rating systems generally average. The GP's proposal is not the same.
Re:Experiment: No Comment Scores by blasdel on Saturday September 26, 2009 @01:40AM (#845113)
You're discounting the effect of the age factor -- I don't downvote high-position low-score comments for reordering purposes because I know it'll probably decay down anyway.

I don't want to be the asshat to drop a comment below 1, just because there's a better comment that's currently below it (but probably with more karma). At least from my usage, you'll see a net gain in karma inflation.

Re:Experiment: No Comment Scores by scott_s on Saturday September 26, 2009 @02:06PM (#845802)
I only did that with comments that went below 1 when I felt it was valid. Is that not expected behavior? (Although one can still do this because text color indicates it's gone negative.)
Re:Experiment: No Comment Scores by netsp on Saturday September 26, 2009 @01:27AM (#845093)
You don't think that this is a good thing?