Ivy League schools in the US have been doing this for rather a long time now. Whether they are any good at it is subject to significant debate, but they certainly like to pretend that they can evaluate it. Their evaluations tend to show a strong belief in the hereditary properties of "potential", which is not well established in actual objective research.
