I think I like this idea quite a bit. I don't know how many users there are with >1000 karma, but will they be motivated to keep endorsing everyone else's pending posts? Sometimes good discussions do happen on quieter threads, or way down the list that the 1000+ users might not see.
Well, I guess it really comes down to just how much (proper) endorsing ends up happening. One thing I like about HN is that it's open and fast to use. Having a pending mode on everyone's comments affects not only the troll-users, but many of the normal ones too who aren't abusing the system ><.
I was thinking the same thing. There's currently no known harm for upvoting, so we do it to promote good discussion, despite the fact that there's typically no personal benefit. But if endorsing posts not only provides no personal benefit but also bears the risk of possible harm, people might be too cautious with endorsements for this system to work.
One concern is that there's no direct feedback to the endorser, so he or she would have no sense of the relative importance of their endorsement to keeping the discussions rolling along. With visions of the Stack Overflow police, how high a bar should a comment have to pass? Although maybe that's a positive -- after all, we no longer directly see comment scores, only the derivative effect of thread reordering.
So, have a new meta-karma that is equal to the mods given to the posts you endorsed? (An alternative would be giving karma directly, or some fraction of karma).
Perhaps that could work, although it goes a bit down the road toward heavier and more explicit mechanics, as at Slashdot and Stack Overflow. In another comment somewhere in this thread, pg mentions that he'd like to keep it as simple as possible. Unspoken is the "...but no simpler" part.
It's fun to consider that a dynamic system could provide someone on the back end (or a smart algorithm) with a variable nozzle controlling comment flow.
But endorsing not at all guarantees loss of that privilege, in that you're never actually exercising it. If the penalty doesn't extend further, I'm not sure the cost will discourage terribly much.
>I don't know how many users there are with >1000 karma
There are about 5500 such users. There have been around 245,000 users to ever post on Hacker News and around 85,000 users have posted over the last year.
Source: I'm working on a fork of the Hacker News Karma tracker (not ready to be live yet) that uses Algolia's new HN Search API; I also downloaded every comment ever made similar to how minimaxir downloaded all of the submissions.
I wonder how many upvotes we would collectively have to give each day to bring comments out of pending?
Looks like in the last 10 minutes we've had 30 comments on HN - that's about 180 and hour - that's 4,320 comments in a 24-hour period.
Let's say 50% of those are worthy of being seen. If we assume it takes two upvotes per comment to bring it out of pending status then that group of 5500 people need to cast 4,320 upvotes a day collectively to to stay caught up.
Given the fact that A: it's unlikely that all 5500 of these users are still active, and B: it's very unlikely that they would be upvoting the same comments then it seems almost certain that there will be a SIGNIFICANT backlog of pending comments created each day.
Well, hopefully new users can hit the 1000+ mark quick enough to not have it affect them too much.
But, I wonder how the lack of visible scores on comments has affected the overall comment scoring rate. For the average new user today, how long do they have to wait to get to 1000 karma? How long was it a few years ago? These would be interesting questions to answer.
I joined more than 2000 days ago. I don't have 1000 karma. This way I'll maybe never will reach it. Seems a bit of a high threshold to me. But maybe my commentary just isn't good enough.
Karma per post is not well correlated with the actual quality of your contributions. You can get hundreds of karma for posting superficial observations early on a post that become popular. Someone that has observed HN for long enough could probably get 1000 karma in a day from just trying to optimize for karma instead of quality.
If you really do want to reach some karma threshold I would suggest trying to get it from submissions instead. You can easily get hundreds of karma per submission if you submit the newest release of some popular software product or the latest Zed Shaw rant with very little invested. This also has the bonus that you are not actually lowering the quality of your commenting for more karma.
That, and doing some basic research to post additional useful information (facts, numbers, stats, references, background info, etc) on a relatively new thread.
I'm mentioning this because I believe such posts are genuinely useful and surely deserve the points they get.
BTW I don't think I ever got near a hundred points for a comment. But I suppose it can happen in the above circumstance in a popular thread.
Maybe the easier solution would be to have two thresholds - one for un-endorsed posting and one for endorsing. Or having a heuristic that includes ave. comment score as a factor. Or... Or...
Honestly, the more I think about this, the more complicated it seems, which is usually not a good sign.
I'm not sure it is a worthwhile endeavor to hope that new users hit 1000 karma. I think the discourse would be better served by having a wide variety of people saying something worthwhile occasionally rather than trying to say something popular more often to play a karma game.
The best idea I've heard so far is a timeout, such that even if no one endorses your comment is to let you comment again anyway after a day.
I'm still not sure what the benefit of adding endorsements is compared to using a generic upvote as the endorsement.
Well, I guess it really comes down to just how much (proper) endorsing ends up happening. One thing I like about HN is that it's open and fast to use. Having a pending mode on everyone's comments affects not only the troll-users, but many of the normal ones too who aren't abusing the system ><.