What should be avoided when first encountering a complex problem?
Breaking it down into simpler parts
Jumping straight into writing a solution
Forming a hypothesis
Planning the experiment