You say that people's soft factors with lower numbers (at Yale) can get them chosen over candidates with higher numbers. So in this situation, good enough soft factors override numbers. Yet in a situation where numbers are equal - soft factors can't override numbers? In this situation, Yale would choose numbers, even though there is no large number difference between the candidates, but choose soft factors when there is?
No, that is not what I'm saying at all. What I am saying is that there could be a hundred reasons why a 175/3.9 is rejected from Yale. Lacking soft factors could most certainly be one of them, but very far from the only reason. He could be an Engineering major and the Yale adcoms felt like they already admitted too many with engineering background, and the list goes on like that.
So a 176/3.8 with amazing soft factors would be less desirable to Yale than a 176/3.9 with nothing but 3.6/167 with amazing soft factors would be chosen over a 3.8/170? (for example).I just don't think this is right.
No, of course not, where did you get that from? I certainly said nothing like that. As said before, there's a large number of reasons why schools like Yale might reject you. There's only one reason why a school like Yale would admit a 160/3.5 student. My entire point was that you can't look at a rejection without knowing anything about that person and automatically chalk it down to a lack of soft factors, no matter how you define the term soft factors. You simply cannot know why
the adcoms rejected that candidate without knowing everything they knew about him (or her).
I don't think you can 'prove' what it is that makes a school take a candidate but you can make an educated guess. Schools say they value soft factors, outliers often get through on soft factors - but talking about soft factors is 'lazy'? It could be that all the candidates reviewed after lunch are dinged because the adcoms have indigestion. Just because we can't disprove it doesn't make it the most sensible theory. Just because adcoms may ding on completely random things we can't measure doesn't make a theory based on soft factors stupid or unhelpful. Not sure why you think it does.
So, you actually agree with me.