We've built elaborate tooling around code complexity metrics, to try and rein in the difficulties of managing complex codebases, but do those metrics actually correlate with programmer's brains working harder? It's a small sample in this paper, but so far the answer is: nope. Not at all.