Don't forget the issues related to momentum transfer and impulse in a VERY inelastic collision. Then add the time-domain issues of the rebound of the ball off your shooter...
There's many, many variables that COULD be added to the problem. I have strongly encouraged our programming team to consider the law of diminishing returns with regard to how much they try to analyze.
We are only shooting a ball 20' into a goal. We're not putting a hypersonic dart through a tank turret a mile away while the gun platform is bounding over terrain
Lots of physics going on, but we only need to know the parts that contribute to the shot within the limits of the robot's own tolerances.