There comes a moment on the range when someone behind you says it: "your arrow is wagging". And they are right, because it really does show: a tail flapping sideways, or a shaft that seems to swim to the target rather than fly there. The usual reaction is to shrug and check the score, which wastes an opportunity, because arrow flight is a diagnostic channel in its own right: reading the target tells you where your shots are going astray, and the flight helps you understand why.
How much a well-flying arrow oscillates
Before hunting for faults it helps to know what is normal, because every arrow oscillates, and in two different ways. The first is flexing. As it leaves the bow the shaft bends and snakes around the riser, which is the archer's paradox, and the registry rates as corroborated that the arrow flexes as it leaves the bow instead of leaving straight. Once free of the string the shaft keeps vibrating very fast, between 75 and 90 times a second in high-speed footage, so the naked eye barely sees it.
That vibration dies away slowly, though. In two different laboratories a carbon shaft set vibrating at rest showed minimal damping, and in one of the two tests it took about a second and a half to lose almost two thirds of its amplitude, longer than the flight time at competition distances. Hence the claim, reported in the registry, that carbon arrows damp their vibration very little and are still oscillating when they reach the target; the second half is a deduction, because that calculation leaves the air out. The comparison with other materials rests on even less: a single laboratory test, on a handful of shafts, finds that carbon damps vibration less than aluminium and wood, with values that partly overlap.
The second kind is slower and involves the whole arrow, which sways about its centre of mass. The centre of mass sits towards the front and the vanes sit at the back, so as soon as the shaft tilts against the air a force appears on the tail that brings it back into line, carries it past and pulls it back again. According to calculations and paper tests at several distances, the arrow completes one cycle every fourteen to twenty-two metres, and the sway shows mostly over the first stretch: in simulations of a recurve arrow over seventy metres it is obvious in the first twenty and then fades, and in the measurement made with a compound it died away over about twenty-six metres of flight. In still air the largest angle stays between a quarter and half a degree, which is why an arrow that leaves cleanly seems to fly straight. The registry rates this claim as reported and no higher, because the measurements come from a single researcher and a compound and concern the horizontal plane: for the recurve and the vertical plane the figures are still those of the models.
How quickly the sway dies away also depends on the fletching, because the type of vane changes how quickly the arrow settles in flight: in the compound test, vanes 24% larger shortened the cycle by 12% and damped it sooner. I compare vane models in Vanes, feathers or Spin Wings.
The loads on the arrow in flight stay small, after all. An electronic point screwed onto a compound arrow went outside its ±16 g range only at launch, estimated from video at about 1,100 g, and when it stopped in a bag target, at about 3,700 g, while it stayed inside that range for the whole flight: this is the basis of the reported claim that the arrow is loaded hardest only at launch and at impact. The air works steadily too: over its flight the arrow loses a noticeable share of the speed it started with, about 8% over 55 metres in an indoor test with a sensor inside the shaft, and I explain how much and why in How much the air slows an arrow down.
In still air, then, a wide undulation, the kind the naked eye picks up, is most likely the same oscillation starting out larger, because the arrow left the bow already askew. Wind is another matter, since in the calculations a crosswind of three metres per second takes the angle up to three degrees. The symptom to read is that starting undulation, and depending on how the tail moves it goes by one of two names.
Fishtailing: the tail wagging sideways
The first flaw lives on the horizontal plane: seen from behind, the flight shows the arrow's tail snaking left and right, like a fish's tail, hence the name. The point more or less holds its line but the tail does not, so the shaft flies skewed to one side, and the wider the sway starts, the longer it stays visible.
The causes all sit on the horizontal plane of the system, and they come in three families. The first is spine that does not match your bow. Every model that has looked at it finds that arrow spine has to be matched to draw weight and arrow length, and the registry rates this as reported: in the calculations a shaft that is too weak ends up slapping the plunger and one that is too stiff leaves the string with a sideways jerk at the nock, so either way it leaves the bow already unsettled.
The second is clearance: the arrow or its fletching touching the rest, the plunger or the window on the way out, a contact that spoils the flight at its source. The margin is small and varies from archer to archer. In footage of sixteen world-class recurve archers the tail cleared the plunger by anything from about 2 to 25 millimetres, and in the archer followed frame by frame the point where the arrow passed closest to the bow was not its tail but a section about fifteen centimetres from the nock.
The third is the release: fingers leaving the string in a rough or asymmetric way add an extra sideways push to the tail, and no amount of tuning can fully erase it. How fast the fingers open, on the other hand, does not seem to count as much as you might expect: in a shooting robot with a competition recurve, opening its fingers at three different speeds, the arrow's flex did not grow in proportion to the speed, and the differences came to a few tenths of a millimetre.
One clue to tell the three apart is consistency: a spine or clearance flaw is systematic and shows up identically arrow after arrow, whereas a rough release is variable, with some arrows flying clean and others not, and it can get worse as fatigue sets in. So if the fishtailing comes and goes, suspicion falls on your shot before it falls on the equipment.
Porpoising: the tail waving up and down
The second flaw lives on the vertical plane: seen from the side, the flight shows the arrow waving up and down, the tail rising and falling like the back of a surfacing dolphin, hence the name. With a finger release it shows up less often than fishtailing, and a model suggests why: in its calculations the arrow's flex in the vertical plane is about ten times smaller than its sideways flex. It is also less ambiguous, because tuning manuals tie it mainly to one cause: a nocking point set too high or too low.
The mechanism is straightforward: if the point where the nock sits on the string is too high or too low, the string's push does not travel in line with the arrow's axis, so the tail is kicked upwards or downwards at launch and the shaft carries that vertical wave through its flight. More rarely a nock that grips the string too tightly, one that is loose on it, or a poor vertical seat on the rest can contribute; on the loose nock, though, there is only a hypothesis: nobody has measured whether a nock that rattles on the nocking point changes the impact height. The first place to look stays the same: when you see the dolphin, check the nocking point.
A machine test shows that the arrow's height on the string matters a great deal: raised by ten millimetres on the string, at two metres the arrow tore the paper over four centimetres, against less than a millimetre with the nocking point set. The same paper, however, takes the edge off the idea that you need millimetre precision. In its model, which covers a single bow, a millimetre either side of the best height changes very little, which is why in the registry the claim that nocking point height needs to be set to the millimetre is contradicted: a centimetre out shows at once, a millimetre out, in the model, hardly at all.
Vertical first, then horizontal
Manufacturers' technical manuals write into their protocols a rule that saves a good deal of time: you fix the vertical plane first, then the horizontal one, which puts the nocking point before spine and clearance. It is a rule of practice that no study has put to the test, but there is diagnostic logic behind it, and it becomes clear once you look at the knock-on effects of a vertical flaw.
According to the manuals, a vertical flaw muddies the reading of everything else, because an arrow kicked out of line on the vertical plane flies unsteadily, touches what it should not and can produce sideways symptoms that look like a spine problem but are not. If you correct the spine while the nocking point is too high or too low, then, you are tuning the shaft to a fault, and as soon as you fix the nocking point you will have to start again. The manuals' order avoids that wasted round, and the model mentioned above points the same way: moving to the next spine shifts the best nocking point height by only 0.2 millimetres, so a vertical set at the start stays valid after later adjustments to the spine. The same principle runs through the whole tuning process, where the steps come in a set order and reversing them leads to improvements that are only skin-deep.
The confirmation: the bare shaft, read on two axes
However well trained, the eye remains an imprecise instrument: it sees the big flaw but cannot measure the fine one, and with adrenaline running it also sees things that are not there. The firmer confirmation comes on the target, with the bare shaft test: you shoot fletched arrows and unfletched shafts from the same distance and look at where one group lands relative to the other. Without fletching to straighten the arrow and mask its faults, the bare shaft shows the problem as it is. The test comes from coaching practice and manufacturers' guides, and as a procedure nobody has checked it with archers shooting; two machine tests, though, point its way. In one, with the nocking point set, fletched and bare shafts landed less than two centimetres apart vertically at twenty metres; in the other, with a right-handed bow, the stiffer shafts ended up on the left. Like the flight, the bare shaft is read on two separate axes.
-
Vertical axis
Nocking point
With the nocking point too high the tail gets kicked upwards, the arrow travels point-down and the bare shaft lands below the fletched group. With the nocking point too low, the reverse: bare shaft higher. You correct it in small steps, closing the gap between the two groups.
-
Horizontal axis
Spine
For a right-handed archer shooting with fingers, a bare shaft to the left of the fletched group means the arrow is too stiff, and one to the right means it is too soft; for a left-handed archer the directions are reversed. You correct it with point weight, arrow length or draw weight, one change at a time.
Mind the order here too: first you bring the bare shaft to the same height as the fletched arrows by working on the nocking point, and only then do you read its sideways offset to judge the spine. Reading both axes together, with the vertical still out of place, is the fastest way to adjust the part that was never the problem. And the usual rule applies: one change at a time and then shoot again, because changing the nocking point and the point weight in the same afternoon means never knowing which of the two did the work.
When it is not the equipment
What I say about reading the target holds here too: not everything that flies crooked is the bow's fault. Flight reads the whole system, and you are part of that system. A release that changes from arrow to arrow produces flights that change from arrow to arrow, and no tuner can chase them: if the flaw is inconsistent, a screwdriver will not help much and it is the shot that needs work, perhaps with a properly filmed video so you can actually see it.
The way to tell the two apart is the same as in all shooting diagnostics, systematic versus variable: equipment fails the same way every time, the shot fails differently each time. Fishtailing that is identical across twenty arrows belongs on the workbench; fishtailing that shows up on three arrows out of ten, perhaps late in the session as fatigue builds, belongs in front of a mirror. And when doubt remains, the bare shaft settles it: it carries your release with it too, but over a full series it averages that out, and the average tells you more than any single arrow.
The same criterion comes back when you turn to the score. In the only direct test, with a compound on a shooting machine at thirty-five metres, halving the sideways sway did not change the group, which stayed within a circle a centimetre and a half across, because the point, much heavier than the rest of the arrow, strayed less than three millimetres from a straight line while the tail swung. In the registry, then, the idea that an arrow that oscillates less in flight groups more tightly is contradicted, with a condition set by the author himself: the sway does no harm as long as it stays the same from shot to shot. Nobody has run the test with a recurve, a finger release and seventy metres, but the result fits the criterion above: on the target, what seems to weigh less is how large the sway is, and what seems to weigh more is how much it changes from one arrow to the next.