I like the scientific method of knowing the world.
1. We put forward a hypothesis. We make it verifiable (scientific).
2. We conduct an experiment.
3. We accept the results of the experiment.
Sometimes my colleagues and I do experiments. Some of the results amazed even me. For example, testers solve problems for analysis better than analysts. In my talk, I will tell you the results of an experiment in creating software for bad requirements. And I'll try to show you how to make them good.