Producing many of the same crap won't increase quality. Producing one crap, testing it, and improving it will. This is almost definitional. The state of "many

Producing many of the same crap won't increase quality. Producing one crap, testing it, and improving it will. This is almost definitional. The state of "many things" is irrelevant, it's the process that links those things together that makes the difference. The ability to learn from the results of the first crap is crucial. There are too many examples of processes that result in the same crap over and over even though the process is iterative. The "quantity vs. quality" reasoning is sort of a red herring here, the real discussion is about how to maximize learning while minimizing investment and baseless assumptions.