Unsophisticated applicants can be at a disadvantage under manipulable and hence strategically demanding school choice mechanisms. Disclosing information on applications in previous admission periods makes it easier to asses the chances of being admitted at a particular school, and hence may level the playing field between applicants who differ in their cognitive ability. We test this conjecture experimentally for the widely used Boston mechanism. Results show that, absent this information, there exist a substantial gap between subjects of higher and lower cognitive ability, resulting in significant differences in payoffs, and ability segregation across schools. The treatment is effective in improving applicants' strategic performance. However, because both lower and higher ability subjects improve when they have information about past demands, the gap between the two groups shrinks only marginally, and the instrument fails at levelling the playing field.