Relating to Randomization, Which Statement Is Inaccurate

Relating to randomization, which statement is inaccurate?

A) as the number of exceptions to random assignment increases, the statistical equivalence of the experimental and control groups decreases
B) when exceptions to random assignment are made, no bias enters the selection process
C) randomization in experiments for evaluation purposes are best suited for programs where exceptions are at a minimum
D) Random assignment of people to receive some especially desirable or punitive treatment may not be possible for legal, ethical, and practical reasons.