Earlier this month, the College Board announced sweeping changes for the SAT. Reverting to the old 1,600-point scale, the new SAT will eliminate the essay, outdated vocabulary words, the “guessing penalty” for wrong answers and complex math problems. The overhaul is designed to better align the test with high school coursework. According to the president of the College Board, the old SAT had “become disconnected from the work of our high schools” and will improve with these changes.

The influential position of the SAT makes the College Board one of the most powerful forces in higher education. The perennial anxiety that the SAT inspires in students and parents makes determining the test’s primary purpose incredibly important. What is the SAT supposed to measure? Knowledge gained in high school, preparedness for college or some more elusive, innate intelligence? The SAT’s changes suggest that its creators want the test to measure high school learning.

But pinning the SAT to high school learning means that testable material can be learned, even coached. Criticisms that affluent students with hired tutors can game the exam do not disappear with these new changes. In fact, the changes might exacerbate the problem. Narrower vocabulary, less abstract and more predictable math problems and an optional essay prompt now released beforehand may give students with tutors an even bigger edge. The College Board claims that the new SAT will be harder to game but they provide no data to support their point.

Designing the SAT to measure some natural aptitude is even more problematic. Does aptitude even exist? How might we define it? We know that intelligence comes in a variety of forms, functions and flavors. Indeed, intelligence’s many manifestations might be so diverse as to make measurement a hopeless task. The elimination of the old SAT’s writing section—a change we applaud—seems to concede this point. “Good writing” is a hard enough goal for Duke’s Writing 101 program, even though it has a 16-week syllabus, PhD-credentialed instructors as well as personal writing tutors through the Thompson Writing Program. The old SAT, which ultimately pandered to the lowest common denominator, had no chance.

We might be asking the SAT to do too much. The question is not so much “What should be on the SAT?,” as it is “How should the SAT be fairly and accurately used to admit students to college?” One of the immediate challenges facing elite university admissions is that the new SAT will be, by most accounts, substantially easier than the old SAT. In an already hyper-competitive admissions pool—most of Duke’s serious applicants hover in the 2100 neighborhood—an easier test will make differentiation harder. In the longer term, the growing rifts between community colleges, lower-ranked state schools and elite liberal arts institutions may demand different standardized tests altogether. No longer will the SAT be an all-purpose test for every type of university.

The more we meditate on the SAT, the more we recoil from its nearly impossible task and the more we discourage its use as the gold standard of college admissions. At a moment when the SAT drives college rankings, fuels high school anxieties and spawns huge tutoring industries, maybe the time has come to ask whether or not we ought to use it all.