How can we automate our test generation so that our covered score goes up more
quickly? One way to do it is with randomization seeds. Randomization seeds are
basically the initial state of your randomization engine. So if you have the
same test and you give it different seeds, you get actually different results.
This way, you get much more bang for the buck from your tests and we use this
method extensively. This will help you to create higher coverage scores, but it
goes only so far. Giving a thousand seeds for the same test is not very
effective, the effect levels off pretty quickly. Therefore, we want to use
randomization seeds extensively. It's much better to use computers than humans
to drive up your coverage scores.