The dynamic of a cellular automaton (CA) evolution with a random initial condition can be studied by calculating the value of the mean, the total (sum), or the base conversion from binary to decimal for each step.

The evolution of CA rules class 1 tends to stabilize to just one value.

The evolution of CA rules class 2 tends to stabilize or to be periodic.

The evolution of CA rules class 3 tends to be chaotic (noisy).

The evolution of CA rules class 4 tends to be a mix of chaotic and periodic, with local repetitive distributions.

Transitory states can occur in the CA evolution depending on the initial condition.