Gary LaFontaine said that the three main methods of deciding on a particular fly are empiricism (trial and error tempered by previous experience), generalism (actually a denial that fly choice is very important), and naturalism (bug watching).
The best hatch-matching anglers do not change flies every twenty seconds. If they see fish feeding, they spend the necessary time to learn what the trout are eating before they ever tie on a fly. Once they know what the trout are feeding on, they know what fly to use. They use a very calculated approach; they are naturalists because they study the water and the insects to determine the feeding activity of the trout. They are empiricists because they know from experience what fly to use when a specific insect is on the water. Fishing a hatch of pale morning duns on the Henry's Fork isn't different from fishing the same hatch on Fall River in California or Nelson's Spring Creek in Montana.
The problem that results from the debate between the overenthusiastic hatch matchers and the practitioners who depend on precise presentation with a few dependable patterns is the focus that is placed solely on catching trout. Both methods will catch trout and, more importantly, bring satisfaction and enjoyment to the angler. If you match the hatch, you'll learn to understand and appreciate the trout's diet and aquatic environment. The study of trout food can provide a real challenge. On the other hand, tying on a fly of your own choosing, without regard to what the trout is feeding on, can bring its own reward. Doing it your way and hooking a tough trout can be a most satisfying experience.
The best way to catch trout on spring creeks and get the most enjoyment from the experience is to be a generalist, an empiricist, and a naturalist.