I disagree with the article. If his point has merit it might be that the data for the Netflix algorithm is the wrong or, more probably, incomplete. That doesn't

I disagree with the article. If his point has merit it might be that the data for the Netflix algorithm is the wrong or, more probably, incomplete. That doesn't invalidate it, rather it means the factors are more complex. Netflix can't know that he leaves Chaplain DVDs on his player unwatched for a week, but it does know that he returns low-brow stuff faster. That faster turn-around probably indicates a behavior. If that data wasn't exposed in the prize, then maybe Netflix will look at exposing it for the next prize, or adding it themselves.