Real reviewers also write fake-ish reviews. It's often sufficient to criticize minor points and rate it "not recommended for publication" to get a paper rejected. The more prestigious journals get so many submissions that a paper with one bad review doesn't stand a chance. It's a way to keep the competition down.