It eliminates the need for a control group.

It reduces variability by grouping similar subjects together before random assignment.

It ensures that the sample is always larger.

It removes the need for blinding.