In my previous post (seekingalpha.com/insta...) I discussed some ideas about improving the Seeking Alpha user/comment rating system. I'd like to take it a step further and look at some numbers actually relating to the issue.
One commenter who seems to generate a significant impact in the comment sections is Cetin Hakimoglu, so let's start with him. A quick comparison:
Betweenthenumbers: 81 comments (in about 4 months), 205 votes, net +131 rating, 82% thumbs-up
Cetin Hakimoglu: 1262 comments, 9215 votes, net -4775 rating, 24% thumbs-up
John Lounsbury: 1095 comments 3256 votes, net +1666 rating, 76% thumbs-up
I'm getting about 2.5 "thumbs" per comment, which is pretty comparable to John L. (#1 commenter ATM) getting just shy of 3.0 votes per comment (and I am a nobody, and he's saying interesting stuff). I'm also comparably postively recieved by the community at an 82% rate compared to his 76% rate.
This is in comparison to Cetin, who is raking in 7.3 votes per comment, with more than three out of four votes going against him. Though his posts are getting significantly more "thumb"-action, he is still getting fewer positive votes (1.75 thumbs-up per post) than John (2.3), or even myself (2.1).
What I would like to be able to tease out (and what my proposed meta-rating system would provide) from the data is the difference between a poster with quality content but with a contrarian viewpoint, and a poster with poor quality and a popular viewpoint. Reading Cetin's comment stream makes me certain he is the latter (bullish cheerleading is popular, but one-sentence contentless posts aren't), but other posters are more difficult to determine. John L has long and well-thought posts, but he doesn't get the net positive ratings nearly as much as others in the top five (John L at +1.5 net thumbs-up per post, compared to prudentinvestor at +3.0/post and CautiousInvestor at +3.1/post).
Additionally, we can use the rating system to eliminate or minimize the contributions from those posters who have consistently shown little value to the community. By being adaptive, a meta rating system will still enable one to see viewpoints the user prefers (even Cetin has 84 followers compared to my 1), while the community as a whole will be spared to some degree when not expressing a preference.
The most valuable articles/posts are the ones that generate meaningful thought and ideas from someone not already inclined to agree with you. We can better separate those who are appealing to a wide range of users and providing substantial content from those filling space in an attempt for self-promotion.