I've been, and still occasionally am, in both sides---sometimes in a position of selecting people, and more likely, in a position being selected. And I can assert the observation PG stated somewhere in one of his essays---in almost all circumstances there are a few stand-outs, a few obvious inferiors, and large number of the rest in the middle ground. Unless you're in either extremes, whether you pass or fail largely depends on small bits of random factors. And it doesn't help whether you know which of those small factors you missed. Because in highly competitive fields, you have to get into the few stand-outs eventually. Don't take the real life tests as if they are school exams where you can earn several points here and there to cover up some minus points elsewhere. You can tell excellent work from not-so-great work, can't you? Just aim at excellent work. If you think your work is less than excellent, then that's the reason you're rejected. If you think you did an excellent work and still rejected, then they are wrong.