My Physics 101 course was taught by a boy-genius that had never taught at university level before. He was very smart and brilliant researcher I will grant but was a lousy teacher. He was very principled and a no curve policy in the syllabus and the first test given had an average score of around 30 out of 100 or so and all but 3 failed (I was #3 and had the equivalent of a D-). The test had no material that was covered in lecture or homework but was derivable from either the textbook or related concepts that the lecture covered indirectly.
He had no intention of changing his stance on that test and we had probably 15% of the class dropout but a week or so after he reversed that and skewed it so the average was a D and lowered its overall weight to final grade. Speculation was he was forced as it was very likely he would have failed 70+% of the class otherwise. It was very instructive to us and I think was a very good lesson as I took things much more seriously from then on and so did most of the class. So I thought that curve was not so bad but will tend to agree a strict curve is a bad thing especially if its not due to an unusual event like this.
Edit: This is far more adversarial than I intended, and realize it's not a direct response to my parent post (which I thought was saying that this test shouldn't really affect the curve at all, as everyone scored the same, besides the wolves). That being said, I'm always happy to rant against using a strict curve for grading.