Great point...but this is a fine line here with what's achievable. You can kind of see from our animation director (Simon Sherr) is that pretty much his whole reason for being is to find ways to reduce the amounts of animation to make features "work", and let the tools and code figure out the rest. This makes it so we can spend all our time on variety instead of on "coverage". The new tackling and catching systems are very good examples of this. But in the example I stated above about blocking, there isn't really anything close to a solution that could procedurally generate the movements needed for blocking animations - and honestly that'd be a LONG way out. Just think of all those subtle movements of edge rushing or hand fighting or pancakes or what have you...where we'll be able to attack procedural / runtime animation in that situation is with a new tech we're working on implementing for '10 called "interaction scaling", that will work to match up interactions based on player heights and scales. NBA Street had this so that a 5'11 guy and a 7'0" guy could use the same dunk animation and then scale at runtime to match their hands to the rim. You'll also see this a LOT in Facebreaker because their player sizes are so diverse...the punch animation may be the same but the GAME moves his fist to make contact with the other guy at the right spot, whether he's 3 feet tall or 7 feet tall.