Process by which a large audience Is sampled to achieve results different than a random sampling.
This convoluted selection strategy is basically random, if random were stupid.