What is the probability that your test data beats control? Make a solid risk assessment whether to implement the variation or not by comparing two continuous datasets.
Distribution of variants