Hi forum, I'm investigating the weird collision behaviour of a bowl game, and I'm starting to think it has something to do with angular velocities. But I found something weirder in the process. The "angular velocity" of the ball (myRigidBody.angularVelocity, expressed in rad.s-1) behaves like it was the derivative of the actual angular velocity (expressed in rad.s-2) : - it has a minus sign along the axis where it should be positive given the direction of the ball's course, - it stays perfectly constant during most of the course of the ball on the ground, while the speed of the ball decreases constantly (e.g., ball is rolling along the X axis, angular velocity stays constant at (0; 0; -9) while the speed drops slowly from 3 to 0,3 m.s-1), - it finally decreases and drops to zero at the very end of the course when the ball completely stops. So, it looks like the derivative of the angular velocity, showing the friction of the ground constantly slowing down the ball. Is that a known behaviour, or am I sorely missing the point ?