What's the name for an exploratory step before writing a failing test in TDD?

Where you play with arbitrarily changing existing values/throwing exceptions from the implementation learn from what tests fail, and why.

Comments