All scientific theories should have predictive powers, otherwise they wouldn't be falsifiable, and, as such, not science. Sometimes, as with your example, it's

All scientific theories should have predictive powers, otherwise they wouldn't be falsifiable, and, as such, not science. Sometimes, as with your example, it's less obvious, but say, if a new species is discovered, evolution predicts there will be genetic differences to its closest cousin or ancestor and that the differences will be related to its fitness to thrive within a particular environment. Also note that "predictive" doesn't need to be about future events. It can also mean new discoveries within existing records (like a feathered dinosaur).