But even you are implicitly making the case that your cohort of Blind users is likely the same as in others. To whatever degree that Blind attracts a biased sa

But even you are implicitly making the case that your cohort of Blind users is likely the same as in others. To whatever degree that Blind attracts a biased sample within a company, that is true everywhere and even so Blizzards results are a strong outlier. Your excuse would only be plausible if the Blind participation bias only affected Blizzard in some way.