Before we release a program, we should perform a range of tests. Why might this be important?
Great! Even though we won't be able to find every bug in a program, we shouldn't release a program with severe bugs.
Can we really take every possible outcome into account?