Introduced on day 6

MCAR (missing completely at random)

Missingness unrelated to any observed or unobserved value; complete-case analysis stays unbiased and only loses precision.

In detail

Every row has the same chance of being missing. Rarely true in operational data; testable against observed variables (missing rates should not differ by segment).