Pattern comments on A simpler way to think about positive test bias