Compound has one anchor that matters, and it isn't on your face
Ask ten compound archers where their anchor is and you'll get ten answers involving the jaw, the corner of the mouth, a kisser button, the string touching the tip of the nose, the knuckle behind the earlobe. Ask them which of those is doing the aiming and most will not have an answer, because the question sounds strange. Anchoring is anchoring. You put your hand in the same place every time. Everyone knows this.
Everyone knows it because recurve archers worked it out over a very long time, and they were right — for recurve. Compound archery inherited the doctrine wholesale, the way it inherited a great deal else that didn't survive contact with the hardware. What almost nobody noticed is that compound arrived with a new anchor built into it, one that is more precise than every facial contact point combined, and that the old advice is not merely redundant on a compound. Past a certain point it is actively harmful.
This article makes an argument that will sound wrong at first: on a compound bow, the peep is the anchor. Everything else is stabilization. Those are two different jobs, they have different requirements, and confusing them is why a lot of archers have a beautifully consistent anchor and inconsistent arrows.
There is a practical instruction at the end for archers who just want to know where to put their hand. It's worth reading the reasoning first, because the instruction only makes sense once you see what the peep is actually doing.
Why recurve needs many anchors
Start with the system compound borrowed from, because the recurve doctrine is correct and understanding why it's correct is what reveals that it doesn't transfer.
A recurve archer aiming at 70 meters has, in the best case, a sight pin and a string. There is no peep — World Archery rules prohibit any aperture, lens, or magnification in the recurve division. The aiming reference is therefore assembled out of the archer's own body: the string aligned against the riser or the sight housing in the peripheral vision, the hand indexed against the jaw, the string touching the chin and the tip of the nose, sometimes a kisser button between the lips.
None of these references is precise. The jaw is soft tissue over bone; a hand pressure change moves it. The string touching the nose is a contact that can shift a few millimeters without registering. The string picture against the riser is a peripheral, out-of-focus alignment that the archer is reading with the edge of their vision.
The recurve solution is redundancy. No single reference is accurate enough, so you stack several imprecise ones and let the errors partially cancel. If the jaw contact drifts but the nose contact and the string picture both say the head is where it should be, the archer catches the drift. This is a genuinely elegant answer to a hard problem, and it is why recurve coaching is emphatic about multiple simultaneous anchor points and about checking all of them. Remove one reference from that stack and accumulated error has nowhere to be caught.
That is the logic. Note what it depends on: the absence of any single reference precise enough to do the job alone.
What the peep actually is
A compound archer has something a recurve archer is not permitted to have. A small aperture, fixed in the string, that the eye looks through.
The site has an entire article on peep apertures working through the geometry, and it establishes a number worth carrying into this discussion. At a 25-inch sight radius, shooting at 20 yards, the X ring gives you an error budget of about 0.013 inches of lateral drift at the peep. That is the total misalignment the system tolerates before the arrow leaves the X — thirteen thousandths of an inch.
Read that as a statement about precision rather than as a warning. A reference that reports misalignment on the order of a hundredth of an inch is an extraordinarily sensitive instrument. When the peep and the scope housing are concentric, the eye is confirming an alignment to a tolerance that no facial contact point can approach. Skin compresses more than that. Jaw position varies more than that between the first arrow of the day and the sixtieth.
This is the fact the anchoring conversation keeps missing. The peep is not one reference among several. It is more precise than all the others by an order of magnitude, and it is the only one located on the aiming axis itself. The jaw, the nose, the kisser — these are references near the aiming axis that correlate with it. The peep is the axis. You are not inferring your alignment from a proxy; you are looking down the actual line the arrow will follow.
Once one reference in the stack is dramatically better than the others, the recurve logic inverts. Redundancy exists to catch errors that no single reference can catch. When one reference catches everything the others would, the additional references stop adding information — and start adding constraints.
The test that settles it
Here is an experiment any compound archer can run in five minutes, and it is the fastest way to see the whole argument at once.
Draw your bow. Instead of bringing your hand to your face, stop with your release hand roughly six inches out from your jaw — nowhere near your face, no contact with anything, no kisser, no nose touch, nothing. Now find the peep with your eye, center the housing, and shoot at a target at ten or fifteen yards.
You will hit it. Not as well as usual, but far better than the doctrine predicts. Most archers trying this for the first time are startled by how well it works, because the anchor they have been told is the foundation of accuracy has been entirely removed and the arrows are still in the target.
What you lose is not alignment. It's stability. The float is bigger and messier. The shot is harder to hold together, the bow wanders, and executing cleanly is difficult. But the arrows go roughly where the pin was, because the eye-to-peep-to-pin line is still intact, and that line does not care whether your hand is touching your face.
That single observation separates the two jobs anchoring has been doing under one name:
Alignment — is the aiming axis pointed correctly? The peep answers this, completely, by itself.
Stability — is the whole structure quiet enough to execute a shot without disturbing that axis? The peep does nothing for this. Your body does it.
Recurve merges these, because in recurve the facial contacts genuinely perform both. On compound they came apart the moment someone put an aperture in the string, and nearly all the confusion in compound anchoring advice comes from continuing to treat them as one thing.
The rotational system
This is the part that turns "the extra anchors are redundant" into "the extra anchors can be harmful," and it is the argument this article most wants you to take away.
At full draw your aiming system is a straight line with three points on it: your eye, the peep, and the scope pin. The peep is attached to the string. The scope is attached to the riser. Those two things are connected but they are not rigid with each other — the string is at your face, the riser is at arm's length.
Now change distance. Your 50-yard mark sits below your 20-yard mark, which means at 50 the pin must be higher relative to the target, which means the entire bow rotates. The riser tips, the sight arm swings, the geometry between your bow hand and your string hand changes.
The question is: what does that rotation happen around?
There is exactly one correct answer. It must rotate around the peep, because the peep is co-located with your eye, and your eye must stay on the aiming axis. The peep is the center of the rotational system. Everything else in the shot — the riser, the sight, the bow arm, your grip, the angle of the string across your face — swings around that fixed point.
Run the numbers for a typical setup. Moving from a 20-yard mark to a 60-yard mark is on the order of two inches of sight travel at a 25-inch sight radius, which rotates the aiming axis by roughly 4.6 degrees. Not much. But it has to go somewhere.
If the peep is the pivot, the answer is trivially fine: your head stays still, your bow arm raises, the peep never leaves your eye, the sight picture stays concentric at every distance. Nothing is induced.
Now suppose your head is locked rigidly to the string — kisser button between the lips, string hard against the nose, jaw pressed into the hand, all of it clamped simultaneously. The string can no longer move across your face, because you have fastened your face to it. The pivot migrates from the peep to the facial contact patch. And the peep, which sits a few inches from that patch, gets displaced as the system rotates.
How much? At 4.6 degrees of rotation with a nose contact roughly 2 inches from the peep, the peep moves about 0.16 inches. With a jaw contact 3.5 inches away, about 0.28 inches.
Set those against the error budget: 0.013 inches. The displacement is twelve to twenty-one times the entire tolerance the X ring allows.
In practice archers don't experience this as a catastrophic miss, because they compensate — they let the nose contact slide, or shift their head, or accept a sight picture that isn't quite concentric, or unconsciously re-index everything. Compensation is exactly the problem. The archer has built a system whose parts fight each other, and resolves the conflict differently on different shots. That is a variance source, and it grows with distance in two ways at once: the rotation is larger, and the same peep displacement projects into a bigger miss downrange.
This is the mechanism behind a complaint that turns up constantly: an archer who shoots well indoors and inexplicably falls apart outdoors at long yardage. Indoors at 20 yards the bow is nearly level, the rotation is small, and an over-anchored system has very little to fight about. Push out to 50 or 60 and the same setup is being asked to rotate several degrees while three facial contacts insist nothing may move. Something has to give, and what gives varies shot to shot.
What the other contact points are actually for
None of this argues for shooting with your hand floating in space. The six-inch test loses stability for a reason, and stability wins tournaments. The argument is about what job each contact is doing, because a contact assigned the wrong job is worse than no contact at all.
The correct frame: every contact point other than the peep exists to stabilize, and must be compliant enough to permit rotation about the peep.
That single sentence resolves most of the apparent contradictions in anchoring advice. Contacts are not references. They are structure. And structure that cannot yield along the rotational axis is structure that fights the aiming system.
Which brings up the claim that goes too far in the other direction.
The release hand is not optional
There's a strain of modern compound advice — usually reacting against exactly the over-anchoring described above — that says the hand position on the neck or jaw doesn't matter at all, that only the peep matters. That is part right and substantially wrong, and it's worth separating carefully.
Right: the release hand is not doing meaningful aiming work. The peep has that covered, and asking the hand to serve as an alignment reference is asking a soft-tissue contact to hold a tolerance it cannot hold.
Wrong: the release hand is doing critical structural work, and it is doing more of it than any other contact on your body.
The hand at the back of the bow is the termination of your entire draw-side structure — the load path running from the string through your hand, wrist, forearm, elbow, and into the back muscles that are holding the shot together. It is the point where holding force is transmitted and where back tension is expressed. That is not a small job and it is not one the peep can absorb.
More specifically, the hand position determines the geometry of that load path. Move it and you change the forearm angle, the elbow position, the line the draw force takes through your skeleton, and the muscles being recruited to hold it. An inconsistent hand position is an inconsistent structure, and an inconsistent structure produces an inconsistent float and an inconsistent release — which will show up as inconsistent arrows even with a perfectly centered peep.
So the release hand must satisfy two requirements that sound contradictory and aren't:
It must be stable. Same place, same pressure, same forearm geometry, every shot. Not because it aims, but because it is the foundation everything else is built on.
It must permit rotation. As the bow tips through the yardages, the relationship between your hand and your face must be able to accommodate that without forcing your head off the peep.
The reconciliation is that these constrain different things. Stability is about the hand's relationship to your own draw-side structure — forearm, elbow, back. Compliance is about the hand's relationship to the bow's rotation. A hand that is consistently placed relative to your own skeleton, and not clamped rigidly to your face, satisfies both. It is only when the hand is being used to pin the head to the string that the requirements collide.
Kissers, nose contact, and the rest
Same test applies to every remaining contact: is it stabilizing, and does it permit rotation?
A kisser button is a light, single-point contact confirming vertical string position. As a soft touch it costs nothing. As a firm bite that locks the head to the string, it becomes a competing pivot. The distinction is pressure, not presence.
String to the tip of the nose is the one most likely to cause trouble, because the nose sits close to the peep and firm contact makes it an effective pivot. Light contact as a soft reference is defensible. Pressed contact — especially maintained identically across all distances — is precisely the rigid facial pivot described above.
Hand or knuckle indexed against the jaw or neck is fine and useful, provided it is understood as structure. Consistent placement, yes. Clamping to prevent all relative movement, no.
The test for any of them, and it's a good one: if I change from my 20-yard mark to my 50-yard mark, does this contact have to change in order for my eye to stay behind the peep? If yes, it is fighting the rotation, and it will not be the peep that loses — it will be your sight picture, silently, on some shots and not others.
So how do I anchor? A compound-only method
Everything above is diagnostic. Here is the constructive version, for an archer setting up from scratch or rebuilding after concluding their anchor is a mess.
1. Set draw length first, then peep height — in that order, and don't compromise either for the other.
The most common anchoring failure isn't a bad anchor. It's a peep at the wrong height, with the archer distorting their head position on every shot to find it. If you have to tilt, crane, or duck to see through your peep, you do not have an anchor problem. You have an installation problem wearing an anchor problem's clothes, and no amount of anchor discipline will fix it. Install the peep properly, at the height your natural head position actually puts your eye.
The order matters because draw length determines where your string hand lands, which determines where your eye ends up, which determines correct peep height. Set peep height against a compromised draw length and you'll be rebuilding both later.
2. Find your natural head position, then bring the bow to it.
Stand as you would to shoot. Look at the target with your head level and neutral — not tilted toward the string, not craned forward. That is your head position, and it should not change for the rest of your archery life.
The instruction "bring the bow to your head, not your head to the bow" is old and correct and constantly reversed in practice. Your head is the stable element. The bow is the thing that moves and rotates. Every distance, every shot.
3. Let the string hand arrive where the structure puts it.
Draw with your back, and let your release hand land wherever your skeletal alignment naturally terminates. Under the jaw, along the jaw line, against the neck — the specific landmark matters far less than whether it's the natural endpoint of a correctly aligned draw. Do not reach for a landmark you were told to hit. If you're forcing your hand to a spot, you're distorting the structure that spot was supposed to indicate.
Then note where it landed, and repeat that. Consistency of the structure, not conformity to a diagram.
4. Confirm with the peep. Only the peep.
Come to full draw and check the sight picture: peep and housing concentric, pin where it should be. That is your alignment confirmation, and it is complete. If it's centered, you are aligned — regardless of what your nose is or isn't touching.
If the peep is not centered, something in steps 1–3 is off. The fix is upstream, in draw length, peep height, or draw structure. It is not to shove your face around until the picture looks right, which is the reflex and which trains a compensation you'll be undoing in a year.
5. Add contact points only for stability, and keep them light.
Once alignment is confirmed by the peep, add whatever soft contact makes the hold quieter — a light kisser touch, a gentle nose reference, hand indexed against the neck. Add them one at a time and keep asking: does this make my hold steadier, and can it still move when the bow rotates?
Anything that fails the second question comes out. Anything that doesn't measurably improve the first isn't earning its place.
6. Verify across the full yardage range.
This is the step almost nobody performs, and it's where over-anchoring gets caught. Shoot your shortest and longest marks in the same session, paying attention to what your face is doing at each. Your head position should feel the same. Your string contacts should be free to sit slightly differently. The peep should be centered at both.
If your 20-yard anchor and your 60-yard anchor feel like different anchors, your system is fighting its own rotation. The contacts are the thing to loosen. The peep is the thing to keep.
Related: The peep sight is an aperture, not a peephole works through the error geometry the budget figures come from. Installing a peep sight covers getting the height and rotation right on the bench. What compound inherited from recurve covers the rest of the borrowed doctrine.
Published 2026-08-31 · Axial Bowstrings