Maybe the way would be to correlate all admissions with success, and add a feedback loop.

I read somewhere that people who graduated at the top of their class generally became average with respect to success.

Also, I suspect success has to be quantified, which might be hard.

> Also, I suspect success has to be quantified, which might be hard.

I wouldn't say hard. It's expensive, time consuming, and the people who can perform qual to quant conversions usefully need to have a foot firmly planted on each side of the subject matter fence.

More to the point, nobody's really interested in compiling this kind of data. Adding dimensions beyond income to your definition of "success" would result in e.g. revealing there isn't anyone from your school successfully practicing family law.