Sample ratio mismatch

A data-quality check for online experiments that tests whether the traffic actually assigned to each variant matches the intended split, usually with a chi-squared goodness-of-fit test.