When I first saw Michael Buckley plead for ratings I thought it was odd. Then two things happened. First, I realized that ratings generally drove a video higher in the charts. Second, I’m now realizing that if someone rates, it’s usually a 5.
As Google’s blog shows, “five stars” (the highest) rating are the vast majority of ratings, followed a distance by one star (the lowest). Since only a very small portion of viewers rate, begging for ratings can shape things dramatically.
Many YouTube partners have loyal fan bases that immediately 5-star their video… Buckley and others have trained their audiences to do that.
But here’s the thing. This is friggin’ GOOGLE. Once it starts to see that ratings are not reflective of the quality of the content, will it still weigh that heavily in what videos get “love” (on spotlight rotations, related videos, or “bestest” charts). I think not.
I’ve found that neither “highest rated” or “most viewed” is any indication of quality. Even “most popular” is somewhat “gamed.”
So what’s the right way to rank videos? Easy… average duration of view based on similar videos of that duration. If the vast majority of people started and finished a video, it’s probably good (or at least consistent with expectations, in the form of title or thumbnail). If the video has a lot of dropoff in the first seconds, then it’s failed the viewer and YouTube since an ad exposure won’t likely count).
GREAT SCOTT I just figured out the answer from last post! THAT’S how Trippy and Shaycarl are getting top popularity ratings for their 10-12 minute vlogs. Their viewers are far more likely to watch the significant portion of their videos relative to similar view rates of long videos like that.
You sneaky bastards. I wouldn’t have given away your secret if you hadn’t made me figure it out. Stay tuned for the 30-minute UncleNalts video to test my theory.