FWIW, this is erosion happening. There aren't, fortunately, that many differences between hn/news and hn/classic, but this is one. A lightweight post that has been voted up almost entirely by accounts less than a year old. I'm not sure what to do about these. It's not offtopic enough to kill, but it doesn't deserve to be ranked #2, as it currently is.
The good news is, these are fortunately still quite rare. If they become more common I suppose I'll finally have to implement vote weighting.
I apologize for submitting this article. If it is viewed by the community as sub-par or unworthy please know it was unintentional. My thought was that this take on the article helped put the numbers into perspective relative to other classes of trade. It probably would have made a better comment than submission.
I have a huge respect for what has been created at HN and though my account is less than a year old I've been reading for far longer.
Two thoughts on things that could be done other than weighting (which I'm actually a fan of).
1. I would make this guideline a bit more explicit "Please submit the original source. If a blog post reports on something they found on another site, submit the latter. -- Reference replies to the primary source in comments unless the analysis is equal to or greater in quality than the original"
2. Create a 2 column approach for comments with "new" on the right and sequential, or popular on the left. This would help some comments get a fair hearing in heavily threaded stories.
However neither suggestion will take the place of a strong community. Thanks for calling me out.
Don't worry about that. The problem is not so much that it got submitted-- all kinds of stuff gets submitted-- but that it got so many votes so rapidly. Could just be a random statistical outlier, come to that.
If you aren't yet treating the community like they will aggressively abuse and spam the system, I think you should start.
Are you looking at any other information about how this story got voted up so fast? I haven't looked at the HN code at all, but I wonder if there is anything sophisticated for stopping people from colluding and spamming. Do these accounts come from the same geographic area? ISP? Do they vote like each other in the past?
Vote-weighting doesn't even really seem like it's going to be the solution to the problem. Well, it might end up being the solution if you give a great deal of weight to trusted accounts, but you might in the end discourage new users from contributing.
I think maybe you consider attacking spammers head on. I really don't think the problem is a bunch of newbies up-voted this story. HN is just easy to game.
You see spam on HN, but it is hard to believe this is a case. The story about profit per company had been up for about 10 hrs by the time I posted. A respected blogger with impeccable tech/startup credentials (founder of Tabblo acquired by HP) posted a thoughtful reply showing how poor PPE was at some large tech companies relative to other companies. It is related and interesting, if a bit mislocated.
It's a decent (though light) link-- it doesn't deserve #2 status, but I still upvoted it. It made me think a bit about what our own company focuses on for measuring ROI.
In the book Good to Great, there's discussion on finding a good "denominator". They give Walgreens as an example of maximizing "profit per customer". Any denominator choice tends to have different effects on your business style and this is another way to optimize. In a lot of workforce-intensive industries, I assume it really just mirrors your overall profit as labor costs could be your number one expense.
This was an interesting followup to the tech company rankings -- but doesn't itself have the weight to justify high billing. Perhaps posting such followup links in comments to the original story could be given higher (or folded-under) placement -- keeping the topics together, giving the original story a greater lifetime, reducing clutter, unifying discussion. (That is, posters or editors could fold stories into existing threads.)
Also: can we get comments ranked in classic fashion? Or maybe display two numbers for a comment's score: all/classic? I think I'm seeing more cases where the absolute sign of the score would be different for early-vs.-recent voters.
I think new users are especially inclined to vote up headlines related to previously popular articles, regardless of the new article's standalone value. They may think it contributes to the discussion and deserves the same attention. Then suddenly one topic dominates the top slots and most links are junk.
Usually, the new article is a hurried blog post that links back to the old, popular one, often in the first paragraph. One rudimentary way to stop this behavior would be to check if the new article contains a very recently popular link.
Does the specific rankings matter? I'd thought the only metric that matters is on/off frontpage.
My assumption(perhaps woefully assuming) is that people scan the front page for new/interesting titles and read them. Not that they think "Oh, I'm not reading that it's only at #13, maybe when it breaks into the top-ten."
Is there any feedback to the submitter why a story is deleted? I've had a couple of stories deleted, one yesterday. Sometimes it's obvious, but this one wasn't to me. Having said that, I'm still in favor of more aggressive pruning.
Edit: I'll take that back. It disappeared because nobody cared. :-( Lesson to me.
I came to the comments just to see what was the reaction to this article which was up early in the morning and still is on the front page a few hours later, even though there's not much content to it.
What about only counting "orange" users for the frontpage?
This seemed a very strong measure of user quality, and it was only reverted because of social effects - which wouldn't be present for frontpage votes. It might also give evidence of whether it really is a strong measure of user quality or not.
You have said that some sites like xkcd have a permanent downweight. Is it only a domain by domain thing? If it can be story by story then you or the editors can just flag stories as lightweight to tweak the front page at will, or push a second type of flag to the community at the expense of some extra complexity.
Goldman Sachs is an interesting case: in 2007, it made above a 3 million per employee profit. Of course, not quite as hot this year. But you can see why young people are so interested in finance.
The good news is, these are fortunately still quite rare. If they become more common I suppose I'll finally have to implement vote weighting.