Replace the total karma score by the username in the upper right of the page with the average that's already visible on the profile page.
An average is a better indicator to the user of how well they are engaging with the site. It only goes up if the user consistently posts worthwhile comments, which rewards being selective about when you post and thoughtful about what you post. Fewer and better is encouraged. Consistently good is encouraged. Lousy comments (like one-on-one arguments and "Me too!") are discouraged even if nobody votes them down.
This bridges the gap between two existing forms of feedback:
total karma (which nobody really cares about, even though it's on every page) and the voting on individual comments (which is the closest thing we have to an indicator of comment quality), and gives the user continuous feedback on how much value they are adding to the site. It makes it much easier to know if your commenting is getting better or getting worse. That is, it is much easier to tell that my average is up or down from the last time I looked than it is to detect the fluctuation in the rate at which my karma accumulates.
Even if lousy comments get voted up--as happens when, for instance, someone first-posts an opinionated quip on a controversial submission and people swarm to vote in agreement--the impact on the user's average karma (which better reflects their value to the community) is much lower than the immediate spike in their total karma (the feedback they currently see). You don't have to stop people from writing such comments--just give them a smaller cookie when they do.
Obviously you can't make anybody alter their behavior over a number, but I think most here would if given a number that they could trust as representative of the shared goal of keeping this place from sucking.
tl;dr average karma is better feedback than total karma
Lousy comments (like one-on-one arguments and "Me too!") are discouraged even if nobody votes them down.
Comments which few people will notice are also discouraged. Most of my 1-point comments are the only comment on their respective articles, and many of those are 'Ask HN' posts. I don't want to find myself thinking "I could contribute here, but doing so will drag my average down".
The ideal measure would be number of upvotes of a comment divided by the number of views of the comment. That way you wouldn't get punished for making comments late in a thread or on a less well read post. Unfortunately measuring this statistic would require a lot of work to implement.
You could hook some javascript to the page's onscroll event. The javascript would send ajax calls back to the server that record which comments have actually been seen (although there is no way to judge if the comment was actually read). It would be a bunch of work though, because the raw log would be enormous and so you'd have to create a special batch processing job to get the view counts split out for each separate comment.
You could probably reasonably approximate this without too much effort.
The number of votes an article has and the number of comments are both going to be strongly correlated with the number of views per comment. So would the mean and range of the comment scores. You could also take the time since the comment was posted into account.
You could have the visible "raw score" for each comment and an invisible "quality adjusted score" that is used to give each user a quality rating.
It also encourages you to upvote people who are wrong so that more people will see it when you correct them. Similarly, it encourages people to upvote bad stories so that more people will see your comments.
Comments which few people will notice are also discouraged.
I think this is a feature, generally, but I get what you're saying. With any criteria you present for deciding what is worth posting you run the risk of it being the only criteria. Ideally, we'd all only care about our karma insofar as it helps us avoid making lousy posts, and ignore it entirely otherwise, but idealism doesn't scale.
"How will this influence my average?" is definitely an oversimplified heuristic for deciding what is worth contributing, but it's reasonably accurate and I think it's the best one we've got. (Better than total karma at the very least.)
As I said in the above comment, the voting on comments is the closest thing we have to an indication of their quality. It obviously doesn't always reflect reality, but what that points to is that some comments--the ones that we want around here but that would bring your average down--are undervalued. Those comments should result in positive feedback (via average karma or otherwise), but they don't.
That is to say: The site has a reward mechanism (karma). There is desirable behavior that is not being rewarded or is under-rewarded (e.g. replying to less popular Ask HN posts). Is there some way to reward that desirable behavior?
FWIW -- and it appears that pg has just implemented exactly this -- I decided to play a game with my "average" a while back, back when it was a reasonable value.
I found that it didn't improve my interaction with the site, it just changed the way that I "played the game". Not literally playing a game, but there was a clear difference in motivations for commenting.
For example: thread is more than 6 hours old? Don't comment, nobody will ever see it or upvote, so my average will get dragged down. Don't reply to any comment at the bottom of a long thread. Use short, witty comments, because those take the least amount of effort for others to upvote (I'm not good at making comments like that, though). Etc.
Interesting. What about a moving average? One that ranks you based on the last ~30 days of activity?
How about the ability to shut off karma for certain posts? One exception if it's down-voted by more than ~10pts. Then the user gets that subtracted from their score, preventing abuse when posting mean/hateful comments.
An average is a better indicator to the user of how well they are engaging with the site. It only goes up if the user consistently posts worthwhile comments, which rewards being selective about when you post and thoughtful about what you post. Fewer and better is encouraged. Consistently good is encouraged. Lousy comments (like one-on-one arguments and "Me too!") are discouraged even if nobody votes them down.
I disagree. I like it when people make a post because they have something to say to someone else. I like it when I reply to someone and they reply back and it starts a conversation. Maybe it leads to an interesting train of thought that I might not have otherwise come up with, maybe it does the same for the other person. I wouldn't like to see HN losing that.
And possibly like this post. I don't know how many other people will find it interesting or will agree with me. But, if I cared about average, I might be hesitant to make this post.
Lets keep the up/down voting arrows, and let the karma operate in the background in the usual way (percolating good comments/submissions to the top, and letting bad comments fall through). But lets just not show karma. Anywhere. Thus keep the quality of submissions/comments the same, and keep the eyeballs focused on what really matters: good articles and good comments.
I think this is the best approach, too, and is one of the tactics I'd decided to use when building a forum one day.
If we assume that any displayed score is going to trigger some amount of gaming motivation in most people, then displaying scores of any kind contributes to a rise in game behavior versus social behavior.
Extending this idea a bit, add a second statistic which indicates what percentile in the overall HN bell curve you fall into.
For the sake of example, imagine an evenly-distributed bell curve. The high end averages 10 karma's per post, and the bottom end 0 karama's per post. If your average is a 6, you'll be in the 60th percentile. A little "60%" could be displayed next to your username.
This allows you to quickly compare your level of contributions to those of others.
A lot of people probably like logging in and instantly seeing their 'karma results' from whatever they recently posted... but I guess that's kind of addictive behaviour we'd probably be better off without (in terms of spending too much time on HN), even if the change causes some irritation.
This is probably due to the fact that some people spend a lot of time commenting on threads that don't get much viewership. I doubt there's much harm in showing the average as a reminder to try to keep the quality high, but I think there are probably better ways to encourage it.
Is there any way I can see other people's average karma?
The individual users pages (except my own) still seem to display total karma.
The leaderboard is still ranked by total karma. If we are adopting (temporarily perhaps) average karma as a metric , it would be great to see who is ranked highest on it or even just how much average karma a user has, just as we can check total karma of a user today
I also suspect that the frequency of average karma calculation may need to be increased.
Agreed - people whose behaviour might be changed by the move from total to average are people who care about how they look, so if no-one else can see it they won't care, surely.
Showing (total, avg) may be a good compromise. Showing the average is an incentive for / against certain kinds of comments (i.e. against answering questions in old threads), showing both might soften that. Also, I use changes to my karma as a quick proxy for if I have more comments in any threads.
Come to think of it - a "new responses" link on the profile page could be very helpful.
Is there a reason flag isn't shown on comments until the reply page?
This definitely discourages posting things that won't be highly upvoted.
The problem I have with this is that absolute number of upvotes is not equal to quality of a comment.
A good comment posted under less popular news item will get eyeballs, less upvotes, and lower poster's average.
Similarly participation in long thread (even if that's constructive thread, not a flamewar) or response far down the page is unlikely to get many upvotes, regardless of quality.
Comments
Replace the total karma score by the username in the upper right of the page with the average that's already visible on the profile page.
An average is a better indicator to the user of how well they are engaging with the site. It only goes up if the user consistently posts worthwhile comments, which rewards being selective about when you post and thoughtful about what you post. Fewer and better is encouraged. Consistently good is encouraged. Lousy comments (like one-on-one arguments and "Me too!") are discouraged even if nobody votes them down.
This bridges the gap between two existing forms of feedback: total karma (which nobody really cares about, even though it's on every page) and the voting on individual comments (which is the closest thing we have to an indicator of comment quality), and gives the user continuous feedback on how much value they are adding to the site. It makes it much easier to know if your commenting is getting better or getting worse. That is, it is much easier to tell that my average is up or down from the last time I looked than it is to detect the fluctuation in the rate at which my karma accumulates.
Even if lousy comments get voted up--as happens when, for instance, someone first-posts an opinionated quip on a controversial submission and people swarm to vote in agreement--the impact on the user's average karma (which better reflects their value to the community) is much lower than the immediate spike in their total karma (the feedback they currently see). You don't have to stop people from writing such comments--just give them a smaller cookie when they do.
Obviously you can't make anybody alter their behavior over a number, but I think most here would if given a number that they could trust as representative of the shared goal of keeping this place from sucking.
tl;dr average karma is better feedback than total karma
Lousy comments (like one-on-one arguments and "Me too!") are discouraged even if nobody votes them down.
Comments which few people will notice are also discouraged. Most of my 1-point comments are the only comment on their respective articles, and many of those are 'Ask HN' posts. I don't want to find myself thinking "I could contribute here, but doing so will drag my average down".
The ideal measure would be number of upvotes of a comment divided by the number of views of the comment. That way you wouldn't get punished for making comments late in a thread or on a less well read post. Unfortunately measuring this statistic would require a lot of work to implement.
A lot of work, and it might even be impossible. If someone loads a page with 200 comments, how many of the comments do you judge that he has read?
That said, upvotes per page-load-including-the-comment is probably still better than the raw number of upvotes.
You could hook some javascript to the page's onscroll event. The javascript would send ajax calls back to the server that record which comments have actually been seen (although there is no way to judge if the comment was actually read). It would be a bunch of work though, because the raw log would be enormous and so you'd have to create a special batch processing job to get the view counts split out for each separate comment.
You could probably reasonably approximate this without too much effort.
The number of votes an article has and the number of comments are both going to be strongly correlated with the number of views per comment. So would the mean and range of the comment scores. You could also take the time since the comment was posted into account.
You could have the visible "raw score" for each comment and an invisible "quality adjusted score" that is used to give each user a quality rating.
It also encourages you to upvote people who are wrong so that more people will see it when you correct them. Similarly, it encourages people to upvote bad stories so that more people will see your comments.
Comments which few people will notice are also discouraged.
I think this is a feature, generally, but I get what you're saying. With any criteria you present for deciding what is worth posting you run the risk of it being the only criteria. Ideally, we'd all only care about our karma insofar as it helps us avoid making lousy posts, and ignore it entirely otherwise, but idealism doesn't scale.
"How will this influence my average?" is definitely an oversimplified heuristic for deciding what is worth contributing, but it's reasonably accurate and I think it's the best one we've got. (Better than total karma at the very least.)
As I said in the above comment, the voting on comments is the closest thing we have to an indication of their quality. It obviously doesn't always reflect reality, but what that points to is that some comments--the ones that we want around here but that would bring your average down--are undervalued. Those comments should result in positive feedback (via average karma or otherwise), but they don't.
That is to say: The site has a reward mechanism (karma). There is desirable behavior that is not being rewarded or is under-rewarded (e.g. replying to less popular Ask HN posts). Is there some way to reward that desirable behavior?
FWIW -- and it appears that pg has just implemented exactly this -- I decided to play a game with my "average" a while back, back when it was a reasonable value.
I found that it didn't improve my interaction with the site, it just changed the way that I "played the game". Not literally playing a game, but there was a clear difference in motivations for commenting.
For example: thread is more than 6 hours old? Don't comment, nobody will ever see it or upvote, so my average will get dragged down. Don't reply to any comment at the bottom of a long thread. Use short, witty comments, because those take the least amount of effort for others to upvote (I'm not good at making comments like that, though). Etc.
Interesting. What about a moving average? One that ranks you based on the last ~30 days of activity?
How about the ability to shut off karma for certain posts? One exception if it's down-voted by more than ~10pts. Then the user gets that subtracted from their score, preventing abuse when posting mean/hateful comments.
what if it were average of: (comment score/parent score) ?
An average is a better indicator to the user of how well they are engaging with the site. It only goes up if the user consistently posts worthwhile comments, which rewards being selective about when you post and thoughtful about what you post. Fewer and better is encouraged. Consistently good is encouraged. Lousy comments (like one-on-one arguments and "Me too!") are discouraged even if nobody votes them down.
I disagree. I like it when people make a post because they have something to say to someone else. I like it when I reply to someone and they reply back and it starts a conversation. Maybe it leads to an interesting train of thought that I might not have otherwise come up with, maybe it does the same for the other person. I wouldn't like to see HN losing that.
And possibly like this post. I don't know how many other people will find it interesting or will agree with me. But, if I cared about average, I might be hesitant to make this post.
Lets hide all karma.
Lets keep the up/down voting arrows, and let the karma operate in the background in the usual way (percolating good comments/submissions to the top, and letting bad comments fall through). But lets just not show karma. Anywhere. Thus keep the quality of submissions/comments the same, and keep the eyeballs focused on what really matters: good articles and good comments.
I think this is the best approach, too, and is one of the tactics I'd decided to use when building a forum one day.
If we assume that any displayed score is going to trigger some amount of gaming motivation in most people, then displaying scores of any kind contributes to a rise in game behavior versus social behavior.
http://news.ycombinator.com/item?id=844979
This is really the best solution because it creates disincentive for upvoting "on the bandwagon," as so often occurs.
Extending this idea a bit, add a second statistic which indicates what percentile in the overall HN bell curve you fall into.
For the sake of example, imagine an evenly-distributed bell curve. The high end averages 10 karma's per post, and the bottom end 0 karama's per post. If your average is a 6, you'll be in the 60th percentile. A little "60%" could be displayed next to your username.
This allows you to quickly compare your level of contributions to those of others.
A lot of people probably like logging in and instantly seeing their 'karma results' from whatever they recently posted... but I guess that's kind of addictive behaviour we'd probably be better off without (in terms of spending too much time on HN), even if the change causes some irritation.
Looking at the leaderboard, the relationship between high ranking hn members and average comment quality seems fairly weak to me:
http://news.ycombinator.com/leaders
This is probably due to the fact that some people spend a lot of time commenting on threads that don't get much viewership. I doubt there's much harm in showing the average as a reminder to try to keep the quality high, but I think there are probably better ways to encourage it.
Ok, let's try that for a while.
Is there any way I can see other people's average karma?
The individual users pages (except my own) still seem to display total karma.
The leaderboard is still ranked by total karma. If we are adopting (temporarily perhaps) average karma as a metric , it would be great to see who is ranked highest on it or even just how much average karma a user has, just as we can check total karma of a user today
I also suspect that the frequency of average karma calculation may need to be increased.
There is now, on the profile page.
Agreed - people whose behaviour might be changed by the move from total to average are people who care about how they look, so if no-one else can see it they won't care, surely.
Showing (total, avg) may be a good compromise. Showing the average is an incentive for / against certain kinds of comments (i.e. against answering questions in old threads), showing both might soften that. Also, I use changes to my karma as a quick proxy for if I have more comments in any threads.
Come to think of it - a "new responses" link on the profile page could be very helpful.
Is there a reason flag isn't shown on comments until the reply page?
I usually end up posting on old threads (since I only check once or twice a day), which means I rarely get more than 1-2 upvotes per comment (if any).
I've got .77, which seems a lot lower than I would expect. I rarely end up in negatives on my comments. But when I do, the negative is a high one.
This definitely discourages posting things that won't be highly upvoted.
The problem I have with this is that absolute number of upvotes is not equal to quality of a comment.
A good comment posted under less popular news item will get eyeballs, less upvotes, and lower poster's average.
Similarly participation in long thread (even if that's constructive thread, not a flamewar) or response far down the page is unlikely to get many upvotes, regardless of quality.