Think in problems

Is it in certain phases of a project better not to think in solutions, but rather in problems?