So, you remember back when we read that glowing review from Rolling Stone about Ima Robot's album and it read like a 4-star review despite the fact that it was awarded only 3? This bit was on The Velvet Rope today and it explains how it is a common occurence for the review writers to have very little do do with eventual star ratings:

From Chicago Magazine.

Tribune rock critic Greg Kot, who contributes occasionally to Rolling Stone, revealed an interesting nugget about the rock magazine’s star ratings, in the first of a two-part interview with teen magazine New Expression. (Part two will appear later this month.) The interview was conducted by Joseph Struck, of Providence Catholic High School. We pick up the interview about midway through the first installment. Edited slightly for clarity.

Kot: It helps credibility or whatever to have pieces published in Rolling Stone, and I like it because they’ve never changed a word that I’ve written. I have always said exactly what I wanted to say in reviews. Occasionally, I’ve been surprised to see that the star ratings get changed, but that’s the editor’s decision.

Struck: They change your ratings?

Kot: Yeah, they do. They have. It is weird. Occasionally they’ll bump up a record that I didn’t feel was as good, and they don’t change a word of writing, so it’s like weird, it’s like you read the review and it reads like a two-star review but there is a three-star rating on it. So it’s just kind of strange, but the editors do have the right. I guess it is in the fine print of the review section that the editors determine the star ratings and not the writers, but often that gets misinterpreted. The writers are blamed for everything [laughs]. So it’s like, “How did you give that thing three stars?” And I say, “I didn’t. I wanted two stars on it.” So that gets a little frustrating. But I have to say that the editors there have been great for the most part.
ChicagoMag.com