Thanks! Hm, it occurs to me that I maybe could have multiplied the relative movement of the bar w.r.t. the movement of the ball, so it could stay inside of the screen.
