First, not everyone’s stats are the same. Among applicants who do have similar stats, the differentiating factors IN MY OPINION are
-
likelihood of enrolling
(based on big-data algorithms that can include legacy status, sibling attending, other connections to U-M or Ann Arbor, history of yield from applicant’s HS, demonstrated interest through visits/ other outreach to U-M, and grades/scores in the sweet spot that makes the candidate strong enough for U-M but an unlikely prospect for HYPSM. -
non-eliminative recommendations and essays --meaning that they don’t necessarily have to make the applicant stand out from the crowd, but they have to be “good enough” and cannot include any info that would eliminate the candidate.