Universities regularly "admit" more students than they can actually allow to "enroll". If a university can physically support a freshman class of, say, 1000 students, they will "admit" a number higher than that, say 1300, knowing that only a percentage of the students that they admit will actually end up enrolling (because very simply some of them choose to enroll at a another university that they were also accepted into).

Here at the university I work at they refer to this "admittance/enrollment" ratio as "yield" or "yield rate". The yield is tracked from year to year and the number of acceptance letters sent out is based on this historical data.

Sometimes, though, the University is surprised (like this past year here) and gets a much higher yield rate and ends up with more students than dorm rooms...

I think it's a value judgement. Is a lowered price of the majority of instances of a thing (flying or taking a class, in these examples) a worthwhile trade for the reward being a probability rather than a guarantee? Is the severity of the minority case (getting bumped) worth the trade? What actual probabilities for happy vs. sad path are acceptable to you?

Those are subjective questions with no inherently correct answers.