One of the factors that I've noticed lately is that due to the quantity of new submissions, it takes only 30 minutes or so for an article to drop off the first "new" page.
I try to check the "new" page periodically, but I must admit, I don't scroll beyond the first page-- so, the article referenced above I missed entirely, and it seems like I wasn't the only one.
Which means that the problem may not be that "nobody is reading the new page much any more" as much as "things age off the new page too quickly for most people to notice them."
As for potential fixes: off the top of my head, I haven't the foggiest.
It's obvious: feature a random article from the new page on the front page like Reddit does (a different one each time you load the page), so it gets some exposure.
But that would have the unintended side-effect of making the site much more attractive to spammers because they'd be almost guaranteed some homepage exposure.
If you want to modify the probability based on karma points, I would suggest giving larger probability to those users with fewer points, as those with more points (ostensibly) would already get more attention given to their submissions anyway.
Dave won't get anywhere with either this scheme or the one I proposed, at least not immediately. But that's the idea: it will be a little bit harder for Dave to gain traction, but it will demonstrate that he's not a spammer, which should be a reasonable price to pay for ensuring the site isn't overrun with garbage (any more than some claim it already is).
Worse things have happened. You could easily cap the "considered" karma at like 500. Or some value that is a function of the total karma in the system.
EDIT: or cap the "considered" karma at the median for all users.
Since the signal:noise ratio on the /new page is pretty darn good right now, I don't think that featuring one of the new submissions on the homepage would change that much.
Also: this proposal would only lead to more impressions for the spammer if the homepage received more than 30x the traffic of the new submissions page. That may be the case, but how would a spammer know that?
Comments
One of the factors that I've noticed lately is that due to the quantity of new submissions, it takes only 30 minutes or so for an article to drop off the first "new" page.
I try to check the "new" page periodically, but I must admit, I don't scroll beyond the first page-- so, the article referenced above I missed entirely, and it seems like I wasn't the only one.
Which means that the problem may not be that "nobody is reading the new page much any more" as much as "things age off the new page too quickly for most people to notice them."
As for potential fixes: off the top of my head, I haven't the foggiest.
It's obvious: feature a random article from the new page on the front page like Reddit does (a different one each time you load the page), so it gets some exposure.
But that would have the unintended side-effect of making the site much more attractive to spammers because they'd be almost guaranteed some homepage exposure.
Feature a random article selected from new articles submitted by users with at least 10 points?
Feature a random article selected from a probability distribution based on user karma.
e.g.
Dave's new article will never get front page exposure. Charlie has a 50% chance, Bob a 33% chance, and Alice a 16% chance.If you want to modify the probability based on karma points, I would suggest giving larger probability to those users with fewer points, as those with more points (ostensibly) would already get more attention given to their submissions anyway.
Dave won't get anywhere with either this scheme or the one I proposed, at least not immediately. But that's the idea: it will be a little bit harder for Dave to gain traction, but it will demonstrate that he's not a spammer, which should be a reasonable price to pay for ensuring the site isn't overrun with garbage (any more than some claim it already is).
That wouldn't work either, it would simply cause a karma feedback loop.
Worse things have happened. You could easily cap the "considered" karma at like 500. Or some value that is a function of the total karma in the system.
EDIT: or cap the "considered" karma at the median for all users.
Would it really?
Since the signal:noise ratio on the /new page is pretty darn good right now, I don't think that featuring one of the new submissions on the homepage would change that much.
Also: this proposal would only lead to more impressions for the spammer if the homepage received more than 30x the traffic of the new submissions page. That may be the case, but how would a spammer know that?
Try looking at the newpage with 'showdead' set to 'on'.
Right now I count 8 'dead' links, but it's not rare at all (especially when it's quiet) to see 15 or more.
Perhaps: autoscroll the pages (of the "new" pages), a-la duckduckgo.com ?