The physics of flipping

[lemiant]

Today I was thinking about how acceleration/deceleration can tip a robot. I know that somehow acceleration acts on the center of mass to flip the robot and this is my theory as to how it works:

The effects of inertia act at your center of mass. They are equal in magnitude, but opposite to the direction of the acceleration of your robot (when you accelerate forwards you will flip backwards).
Mathematically if you add the inertial vector (equal an opposite to acceleration), to the gravity vector, you will get the net vector of your robot. When you project this vector from your center of mass you can determine if a robot will flip. If the projection lands inside the wheelbase you will stay on the ground, but if it lands outside of your wheelbase... prepare to flip. (I have attached some diagrams illustrating this).

Could one of the physics people on these forums, confirm or deny that method (in a strange twist I cannot find the right search term to take me to lessons about this). Also I'm wondering what is a ballpark value for the maximum acceleration a FRC robot sees during a match. And if someone wants to rephrase the above paragraph to use the actual terminology I would be immensely greatful.

Thanks,

- Alex