I shouldn't have said statistic and I think you misinterpreted. My friend submitted the SAME application with literally identical essays and everything except for the race. In other words, my friend submitted two applications to the same college, Harvard.
Ah, I got you. But I believe I asked the question do you know what the total pool of applicants looked like?
How do you know (again) 20 whites and 20 blacks applied (in total) and 10 whites and only 5 blacks (which includes your friend's submission) happened to get accepted? If you don't know what the total pool looked like the example above doesn't demonstrate a whole lot.
Also, how do you know the two applications weren't both "right on the border" and by 50/50 chance the black guy just happened to make it through and the white guy didn't? The application reviewer might of had a bad day when the white one rolled in and a good one when the black guy's was reviewed..
You know what I mean? There's so many variables here making this exercise nothing more than fruitless speculation.