Re:Vote early and often. (Score -1)

Serendipity.

There are algorithms already written that can incorporate serendipity (academic and corporate I'm sure). What this does is brings out the fray that may not have as many ratings, but has high ratings.

You can also throw collaborative filtering into the mix - "people who rated these items high have also rated XYZ high".

Pretty cool stuff, but it's expensive and takes a lot of ratings before things start to connect right. Not to mention the cost of running the database of ratings and then chugging real-time calculations on that data.