[Rudy tossed tigger's salad]

make the mock draft performance based using draft value projection and +/- differential. Bottom 4 are removed for next year along with people who drop out. Top 15 picks will be safer in that they are less likely to pick a Clausen, but also won't benefit from an Alualu. After that it's all pretty even I'd think.

use the +/- for the first 3 or 4 rounds...or maybe even all of them. Then take that number and divide by number of picks and lowest 4 numbers are out. All undrafted players used in mock would result in negative difference between mock pick and last pick + agreed upon number as penalty (just 1 could work)

plus it'd be funny to look back on and see someone booted who actually made good picks