4 ms·
We've seen this result before, from Wikipedia: http://thedartmouth.com/2007/10/22/news/wiki http://thedartmouth.com/2007/10/22/news/wiki They found that there
by roguecoder 14y ago
We've seen this result before, from Wikipedia: http://thedartmouth.com/2007/10/22/news/wiki http://thedartmouth.com/2007/10/22/news/wiki
They found that there were two patterns of content contribution: people who wrote one review and people who wrote lots. For people who wrote one review, anonymous contributions were higher quality than signed, and those single reviews were higher quality than people who contributed multiple reviews.
Chances are good your most valuable reviews are coming from the long tail and your least accurate reviews are coming from people who are posting lots, but potentially with sockpuppet accounts. Tossing out your most accurate reviews in an attempt to suppress the latter seems counterproductive.
- shangrila 14y agoThis needs to be higher. This is exactly the reason why what the article calls "Assumption X" is flawed.