Well said. Community obviously is a big part of it. Implementing a penalty rating system may help.
But the real issue is really the combination of a weak community and weak expectations.
On HN, the weight of the community likely outweighs any negative (not intellectually stimulating) submissions. Interesting articles are up voted. Link-bait is down voted (or removed). With fewer people here at DT, we can only give so much weight to the better articles.
The bigger issue though is the weak expectations. HN has outlined guidelines about what to submit (though it is fairly general). I haven't been at HN since the beginning, but I feel its fairly easy to get a sense of what is a good submission and what is not after following the site for a week. That is not true at all here. A new user (or even me, who has been following this site for awhile), has NO IDEA what a good submission is here. Thus, they may find anything even tangentially relevant to data science a submission. Then if it is bad, we lack the community power above.
How to solve these issues? Im not quite sure, but a couple wacky ideas:
1. Only 10-15 articles on the front page. This will give weight to the "better" articles.
2. Approval process for new user submissions; needs to be approved by 5-10 older users.
There is no glue in the community right now. It is too weak and disconnected, which results in a lack of signaling to new users. We need to change that.
