Acceptance rate and yield are very closely linked. Yield is one of two independent variables in acceptance rate for residential colleges (which are the ones we care about for these purposes). The other independent variable is number of applications received. Since the number of available spaces is a constant for any particular college, acceptance rate is essentially a function of how many applications it receives and how many offers it has to make in order to fill its slots:
Acceptance Rate = Acceptance Offers / Applications
Acceptance Offers = Slots / Yield
So . . .
Acceptance Rate = Slots / (Yield x Applications)
@gibby : You are seriously over-reading that blog post. Yes, May 1 is the most important date for determining yield. But colleges generally don’t report yield until late summer or the fall (if they report it separately at all). Most yields are calculated from the CDS, and it will be Enrolled Students / Acceptance Offers. I have never heard of anyone backing out waitlist acceptances from that calculation. And it would be difficult to do that accurately, since not all colleges follow the somewhat nefarious practice of only making waitlist offers to applicants who have indicated in advance they will enroll. (Harvard, for instance, at least in the past, would offer admission to waitlisted students without extracting a commitment to enroll.)
It has been common for years to criticize one college or another for trying to manipulate its acceptance rate and yield by admitting an artificially low number of applicants in March, then filling the rest of the class with waitlisted students who are pre-qualified so that their yield is similar to the yield of ED applicants. It’s like Late Decision – a chance to fill slots at a 100% yield, thereby reducing the acceptance rate necessary to fill the college’s class. That wouldn’t work if the statistics were locked in the first week in May. But they aren’t.