As others have said in the thread, it's noise over the signal. The top colleges only accept A students, don't they? If I take a class and clearly produce A quality work, I should get an A for it. The fact that all of my cohorts are also A students isn't my problem -- in fact it's no one's problem at all that we're all super good. Forcing a curve is just a deliberate attempt to screw with the data so it looks like it fits that curve better, and how does that help anyone?