[Rain Man]

Y'know, if you built a mathematical model to predict a player's odds of success, you could then optimize your selections based on the points in the draft model.

For example, if your top QB candidate has a 60 percent chance of success and costs you 2,200 draft points, and if your fourth-favorite QB candidate has a 20 percent chance of success and costs you 480 draft points, you could then compare that to a couple of LB candidates who has a 90 percent chance of success and a 10 percent chance of success at the same draft positions. It then becomes a simple optimization to see which you should pick where.