Always try to do this first step of requirement analysis regardless of what you think the final solution is: Understand the problem as best you can. The amount of effort and time invested in understanding and documenting the requirements will depend on the importance of the system, complexity of the situation, and the possible downside or risk of requirement misspecification.

Do not advocate or be a slave to any particular method or notation scheme for what you discover or how you discover it. Too often, methodologies become mythologies and people come to believe that just because a specific method or style is used, the solution and answer are correct. Wrong. It simply means that a level of consistency exists and whatever solution was derived looks like other solutions. The quality of the solution is more dependent on the individual problem solvers than on the method. There is good in methods though. Various methods do help contain and control group dynamics, and help facilitate the discussions. Other methods help document different types of problems. Having a consistent method can help reduce issues with communication, but there is no one method or one way of doing problem-solving or requirements analysis that will work in all situations. Excellent problem solvers have multiple tools at hand, know when to use them, know the strengths and weaknesses, and know when to run like hell away from them. They are not caught in the religious cult of a particular school of thought at...