Fly patterns break into two categories, realistic and impressionistic. The impressionistic ones are what most anglers actually fish, and it's always been that way.
Which is another way of saying most fly patterns are based on a bug that doesn't exist.
The Adams isn't a mayfly. Griffith's Gnat isn't a cluster of midges. The Prince Nymph isn't anything you'll ever turn over a rock and find. The Royal Wulff has a red band around its waist. Nothing in a river has a red band around its waist.
Match the hatch is doctrine, so this reads like heresy.
Every fly fishing book since Halford has told us to identify the bug and then imitate the bug. And yet the flies that fill most working fly boxes look nothing like specific insects.
There's a reason for that, and it's the reason those flies keep working.
The classic examples
When people first started tying flies, way back when, they didn't have the materials or the knowledge we have today.
What they had was proportion, color, and a rough sense of what a bug looked like moving through water.
Take Sawyer's Pheasant Tail. When Frank Sawyer tied it on the Wiltshire Avon in the 1950s, he used two materials. Pheasant tail fibers and copper wire. No thread. No dubbing. No separate legs. Just a slim brown shape with a bit of weight.
He called it a mayfly nymph, and it does the job across half a dozen species he never bothered to name. It doesn't look like any one of them. It looks like the idea of them.
The Diawl Bach takes the same principle further. It is a Welsh reservoir pattern that was tied deliberately to represent nothing in particular, or more fairly, to represent the general idea of several bugs at once.
The tie is a peacock herl body, brown hackle, and a bit of red at the head on some versions. It doesn't imitate a chironomid, a small nymph, or an emerger, but it works on trout that are eating any of them.
Then there are the ones every reader already carries. The Adams was tied by Leonard Halladay in Michigan around 1922 for a friend named Charles Adams.
The Griffith's Gnat is meant to be a cluster of midges balled up on the surface. Trout eat it during single midge hatches, during small mayfly emergences, and when nothing much is coming off at all.
What actually drifts past them is a small dark shape with hackle sparkle.
The Prince has white biots on a peacock body and brown hackle. Match that against anything living. The Royal Wulff carries a band of red silk around its middle. Nothing in nature does.
Why it works anyway
Materials
More times than not, fish don't make out the pattern of a fly the way we do at the vice.
For dry flies especially, a trout is looking up through its window, seeing a silhouette against the sky. Not color, not texture, not the count of hackle turns. Silhouette, size, and the disturbance the fly makes on the surface film.
The Elk Hair Caddis is the honest example. It doesn't look like a caddis you'd pull out of a hatch net, but the tent-wing silhouette on the water is close enough that a trout sees "caddis" and commits.
You're not fooling the fish with detail. You're giving it the outline it already decided means food, then getting out of the way.
This is the case for most patterns. They're far better off representing a category of bugs than an individual bug.
One Adams covers half a mayfly hatch and a caddis emergence at the same time. One Pheasant Tail covers most of what's crawling on a streambed in April.
Whereas the ultra-realistic tie that copies one specific species will only get eaten during that species' hatch, and often not even then, because the fish never gets the long clean look you assumed when you were tying it.
The impressionistic fly has range. It covers insects it was never designed for, water types it was never tied for, and days when you can't tell what's on.
What this means for your box
I've said it before and I'll say it again. Ugly flies catch fish.
So your fly doesn't have to be ultra-realistic. It has to hit the trigger profile that's working today.
An ideal fly box is built by trigger, not by species. Small dark nymphs for fish grubbing along the bottom. Buggy medium nymphs for general searching. A few flashy attractors. Big moving streamers for when nothing else is on.
That covers most of what any river will throw at you across a season.
Most of the new patterns that come out every year are cosmetic variants of one of those profiles.
Change the hackle color, swap in a different flash material, give it a name, put it on YouTube. The trigger profile underneath hasn't moved. You already had the fly, it just had a different name.
I like tying semi-realistic flies. There's a satisfaction in getting the segmentation right or matching a specific mayfly's leg count. But most of the time, that's the tying hobby taking over.
It doesn't make the fly catch better. It makes the tyer feel better. Those are two different things and it's worth being honest about which one you're chasing when you sit down at the vice.
So next time you tie on an Adams, remember what you're actually doing. You're not fishing a mayfly. You're fishing a shape that a fish, somewhere, once decided looked enough like food to eat.
Tie for the trigger, not the textbook.