So maybe have 2 tests. A second one that is harder and spreads students out more. I was just questioning the value of a test that 'normalizes' student scores having no distribution at the top end.