but how do you know this isn't confirmation bias? The guy that got in with the 3.3/23, at least, has the usual pedestrian set of EC's, nothing amazing nothing terrible. It is easy to assume that borderline applicants that got in must have had amazing EC's, but when you compare them vis-a-vis with other applicants with higher numbers, they really don't stand out as much as you think. And they do not carry as much weight as most people think, either (unless we're talking publications, peace corps, teach for america, etc.). 9 times out of 10, these "underdogs" that get in, do so because their numbers are more competitive than everyone thinks, not because their EC's were "amazing".