Assignment of trial participants to treatment arms using a chance mechanism to reduce bias and balance confounders.