Question about msd command

Dear LAMMPS users,

I want to study the mean-squared displacement of polymer chains during deformations.
I want to use compute msd command, but i have a question: in LAMMPS documentation i read that the displacement of an atom is from its original position at the time the compute command was issued. Original position is intended t = 0 (at starting MD) or a generic timestep precedent???

Thank you

Compute msd simply calculates a displacement. It pays
no attention to time.