Evolution makes falsifiable predictions that keep being confirmed

Coyne stresses that evolution isn't merely a story fitted retroactively to existing data; it's a framework that generates specific, testable predictions before the confirming evidence is found. If evolution is true, certain transitional fossils should exist in particular geological layers, certain genetic similarities should track known ancestry, and certain vestigial structures should appear in specific patterns across related species. Time and again, Coyne argues, these predictions have been borne out by later discoveries rather than falsified by them, which is precisely the pattern expected of a genuinely correct scientific theory rather than an unfalsifiable belief. He treats this predictive power as evolution's strongest scientific credential, distinguishing it sharply from explanations that can accommodate any observation after the fact. Takeaway: a theory that keeps successfully predicting future discoveries is doing real scientific work.