Seems like a lot of confusion could be straightened out if they'd just use percentages among peers. Perhaps not all tests show a percentage?