I think their system works as long as the player produces at the top % of players in that position. I look at it like this: WR #A is the top of the market and he plays at the top of the position, then the ROI is a wash. He's the #1 guy and has the #1 production, so everything is fine. If they pay a guy 75% of the market and he produces in the top 95%, then that player has a very high value. If a player is paid 50% of the market but only produces 60% of the market production, that player still has better value than say the #2 WR who performs at 80% of the market but is paid at 99% of market for the position. Follow? I feel they don't want to be paying 90-100% of the market when they are almost assured that player will underperform the contract, whereas the guy paid at 50% of the market has a very good chance of being on the right side of production. All this is skewed by Brady's performance, because he was able to raise guys up and get them to a higher % of ROI. I think further that they say that X position has a budget of a certain number, and they aren't going to exceed it, period, doesn't matter what the player is offered elsewhere. It's no different than any other job. The mid level manager gets paid X# of thousand of dollars per year, and that's what that job pays. If the employee performs above expectiation, then the employee can earn more, but the cap will be at a certain level. Once that employee reaches the cap, he/she is either promoted, or let go and someone else is paid less. There's always someone who will do your job for less money, right?
Defensively, Belichick was the best at coaching up DBs, so they knew that if they paid a guy 90% of market, he'd still be a good value because he'd probably be the top performing in the top 5% of DBs anyway, so it'd be a win.
I don't think they look at stats, per se. I think they look at the whole picture, real money ball analysis type stuff, not just fantasy football numbers. I think they analyze everything.
@Ian I would like your opinion on this one.