Free explainer2 min read
Random allocation avoiding regression to the mean
Understand the model, recognise it in a stem, separate the look-alikes, then apply it.
01Memory anchor
Extremes drift back on their own — only a randomised control arm proves the treatment did it, not chance.