I couldn't agree more with the basic sentiment of this post. I'm so sick of how knowing how to solve these test-the-basics questions is supposed to determine wh

I couldn't agree more with the basic sentiment of this post. I'm so sick of how knowing how to solve these test-the-basics questions is supposed to determine whether a person will do a good job for the company. This approach favors ambitious geeks who like to study tons of these problems before hand, and only mildly helps those who are really bright. Rather, the author has it quite right. "Talk about projects, see if the guy gets animated. Combine that with some pre-submitted code samples, and you can get a genuine idea of how suited the candidate is." This favors candidates who are genuinely interested in the particular job. The best way to determine the aptitude of a candidate is to see how enthusiastic he/she is about providing solutions to basic problems the company faces. With this, I'm of course not implying that your company should hire any random idiot, so spare me that. But I am saying that the virtues of the traditional Spolskyan model are highly exaggerated.