If those were numbers based on random chance, I would agree with you. However, poor applicants tend to apply more broadly than good applicants and that fact is missing from your analysis.
Being in-state vs out-of-state isn't a factor in weighing candidates. Because more out-of-state candidates are applying broadly to account for their lower-quality application, we accept fewer of those.
This person applying isn't a roll of dice or a random card draw, and the strength of their application is not a random chance event. You're interpreting those numbers as if it were.