In the derived category $D(C)$ of an abelian category $C$, one formally inverts quasi-isomorphisms. In the context of model categories, one inverts weak equivalences.
What does one gain by doing so?
Is there a big-picture way to think about what is being done?