Ayres and Brooks eliminate the admissions effect of affirmative action by sending all the black students, on paper, to the set of law schools that white students with the same entering credentials attend. (Like Sander, they divide all the schools into six tiers, grouped roughly by selectivity.) Then Ayres and Brooks compare the bar-passage rates of the reshuffled black students in each tier with the passage rates of the black students who really did attend those schools. Sander makes a counterfactual comparison like this, too, but he compares his reshuffled black students with the white students in the black students' new tiers. Ayres and Brooks say that's a mistake because, unfortunately, out in real life, white students on average do substantially better on the bar than do black students who attend the same schools. In other words, affirmative action isn't to blame for the bar-passage rate of black students, but yikes, is it low.