Though in fact I do have a new plan for dealing with cases like Valleywag (suggested by Nick Grandy of Wundrbar): instead of simply banning linkbait sites, I'm going to try semi-banning them by requiring them to get more points to make it onto the frontpage. That sounds right: they're semi-spam, so semi-ban them.
- At one end: the posts from new/under-participatory users, where spam and linkbait is more likely.
- At the other: the posts to popular domains.
What I love about HN is that it manages to uncover interesting content from all over, but I am inclined to believe this is due to its current user base. With an algorithm in place that reinforces this behavior, I think we can continue to grow in the right direction.
I believe the first point may already factor in, but does the second?
The solution could be relatively simple: the more "popular" a domain name is, the more votes a submission requires to become "relevant."
Popularity could be determined a number of ways, at least using the number of submissions under the same domain, and perhaps the average number of votes the domain receives. The actual formula would take some tweaking over time.
In theory, a site with multiple submissions, like Valleywag or TechCrunch, would require more universal and active interest to warrant a front-page appearance. Valleywag (the more volatile of the two), would probably fade away, whereas the interesting TechCrunch articles would still come through.
I like the idea of requiring sites like XKCD, Valleywag, TechCrunch, and others to get more votes to make it to the front page. When that's implemented, will you then un-ban Valleywag?
So, like Paul said, I guess there IS a time when censorship in the answer, and it's for obvious spam.