Females are high 30's to 40ish% and rising - I am sure they would like to get it to 50%. Minorities are about 35% and stable across the last few admission cycles, but you could see them rise if there are sufficient qualified applicants. These numbers demonstrate the difficulty of obtaining an appointment to CGA. Including CGAS, of the 280 who accept appointments, approximately 140 are athletic recruits, so if you aren't an athletic recruit, you are already down to 140 available slots, which includes CGAS spots. So for the major demographic groups the available slots would be as follows:
36 white female,
20 minority female,
54 white male,
30 minority male
My simple math assumes that the percentages of female and minority appointments for recruited athletes and those coming in from CGAS is the same as it is within the general appointment pool. I have no idea whether that is actually the case, but it probably doesn't vary much. My math also assumes that appointments from CGAS are also 50% athletic recruits. I would guess the number is actually higher, but that's just a guess. If CGAS is 75% recruited athletes, and the proportion of females and minorities remains the same, available slots to non athletes would be approximately
40 white female,
22 minority female,
60 white male,
33 minority male.
Presumptions about the percentage of athletes in CGAS and gender and minority proportions within the CGAS and athlete pools aside, the data shows that each applicant is competing for only a handful of slots, as the individual candidate simply isn't eligible for a demographic they don't belong to.
I should also note that these numbers are based on the 280 accepted appointments. There are close to 400 offered appointments, including CGAS. We know that about 310 are offered a full appointment, and 220 accept, along with another 60 coming in after completing the CGAS program. Unfortunately, we don't know anything about the demographics of those who were offered an appointment, but declined. We also don't know what percentage of them were recruited athletes. Further extrapolating numbers with so much guesswork could be done, but isn't very helpful.