What process do you go through before you start the experiments? Do you use a hypothesis?
Our experiments are of two broad types. In one you start with a hypothesis and design an experiment to check it. In the other, you design an experiment to look at something no one has ever seen before to explore new "territory." The first type of experiment usually leads to steady progress whereas the second can sometimes lead to unexpected new discoveries. Both types are important to overall progress.
Nathan Isgur, Theory Group Leader (Other answers by Nathan Isgur)