The idea is that often a breakthrough in mathematics isn't achieved by tackling a problem directly, but converting it to a simpler problem, then solving that one.

Disclosure: I'm a software developer.