Part of the issue is the test design. If the university takes the top 1% of performers, then it could mean you need 99% correct answers. But students scoring 97% or 99% correct on a test already know all the material. The difference is in luck and accuracy, not intelligence.
The solution is to write a harder test that is only meant for the top 10% of students to take and actually contains harder questions.
The solution is to write a harder test that is only meant for the top 10% of students to take and actually contains harder questions.