Sooner or later almost every archer wonders about it: is it better to have one finger above the nock and two below, three below, or just two fingers? Anyone who watches traditional or Asian archery also wonders what changes when the string is held by the thumb. On the range there is no shortage of answers. Three fingers are said to hold the string steadier than two, the thumb is said to tire the hand less, and the grip is said to need setting up properly from the first lesson because it costs points later.
Measurements of the string grip are few, and nearly all of them come from a single experiment in 1990, in which nine international-level archers tried four different grips. Around that study there are a few others, each holding part of the answer: a laboratory task without a bow on how steady the fingers are, a comparison of two grips in people who had never shot, the wrist forces of a horseback archer who releases with the thumb, and the finger joints of a recurve archer filmed at 500 frames per second. Taken together they answer some questions, and on one of them, whether three fingers are steadier, they reach opposite conclusions.
The study that compared four grips
The experiment comes from a Belgian group, and it is the second of three reported in the same paper. The nine archers were FITA competitors with international experience, all used to the vertical three-finger grip and with no experience of the others. Since learning a new grip properly was not realistic, and the athletes themselves did not want to, the authors chose a standard procedure. Each grip had a quarter of an hour of practice without a target, then three ends of three arrows to settle the anchor, and finally three shots recorded with electromyography of six muscles of the draw arm. There were four grips: three fingers, two fingers, thumb and a reversed grip with the hand pronated. Precision meant grouping, the contour in centimetres of the three arrows of each end, not score.
Between three and two fingers the study found no significant difference. On the third end the group contour was 14.1 centimetres with the usual grip and 11.1 with the two-finger grip, which was new to all of them, and the statistics could not separate the two. The two-finger grip was also the quickest to learn. The picture changes completely with the thumb and the reversed grip: on the first end the group measured 175.7 and 179.4 centimetres, and over nine arrows it came down to 42.0 and 53.2. The improvement was fast, but after nine arrows those groups were still three to four times wider than with the finger grips.
That is why the registry rates the claim that a two-finger string grip is about as accurate as a three-finger one as reported: there is a result on archers pointing that way, but it is a single one with clear limits. Grouping was measured over nine arrows per grip, the archers had only just tried the two-finger grip, and precision was measured as grouping rather than as a score over a round. The study says that a strong archer does not group worse with two fingers than with three straight away; it does not say what happens after months of practice, or over a whole competition. Nor does the study say whether the two fingers were both below the nock or one above and one below.
Three fingers and a steadier hold
The idea that three fingers hold the string steadier than two has an intuitive mechanical basis, because the load is shared by one more finger. Twenty-seven years after the Belgian study, a Korean group put it to the test in the laboratory, but without a bow and without archers. Eight right-handed men, whom the study does not describe as archers, pressed with two, three or four fingers on force sensors mounted on a simulated string, held the force for a few seconds while aiming at a target on a screen, and then let go; the computer worked out where a virtual projectile would have landed.
With three fingers the force error during the aiming phase was lower than with two on all three axes, and the virtual firing point varied less from trial to trial. It did not land closer to the centre, though: the accuracy index was the same in every condition. The whole gain came with the third finger, while the fourth added nothing, and with four fingers the consistency could no longer be told apart from two. The duration of the simulated release did not change with the number of fingers either. The same study also shows how much each finger does: in the maximal four-finger press the index finger reached 29.8 newtons, the middle finger 16.5, the ring finger 11.1 and the little finger 14.9.
Here the two studies pull in different directions. The laboratory task says that three fingers give a more consistent hold than two; the only comparison made with a bow in hand, on international-level archers, finds groups with a contour of 11.1 centimetres with two fingers and 14.1 with three, with no significant difference, and so no advantage for three fingers where the arrow is concerned. The two measurements are not the same thing: the first looks at how much the force of a hand on a sensor fluctuates, the second at where the arrows end up. The claim, however, promises something to an archer holding a bow, and on archers there is a piece of data: so the registry cannot rate it as transferred, and since that data goes against it, the claim that holding the string with three fingers gives a steadier hold than with two is contradicted. The contrary evidence is thin and indirect, a single study that looks at the group rather than the hand, and the status says exactly that.
The thumb: less muscle activity, more difficulty
The Belgian study also measured the muscles of the draw arm: the trapezius, anterior deltoid, biceps, brachioradialis, finger extensor and flexor digitorum superficialis, all expressed as a percentage of maximal contraction. Taking the six muscles together, the thumb was the grip with the lowest average activity, 53.4%, against 59.3% with three fingers, 62.4% with two and 56.7% with the reversed grip. The authors judge it the most economical of the four grips, and put it down to a better lever arm.
Muscle by muscle, though, the picture is more modest. Compared with three fingers, the trapezius, finger extensor, biceps and deltoid did not change significantly with any grip; only the flexor digitorum superficialis and, with the reversed grip, the brachioradialis dropped. The standard deviations with the thumb reached 12 points, about twice those with three fingers. And all nine archers, once they had settled the anchor, found the thumb the hardest grip. The authors draw a hypothesis from this rather than a conclusion: practised long enough, the thumb might be the most economical grip.
The second piece of data on the thumb comes from an American ergonomics conference in 2007, and it concerns people starting out. Sixteen adults aged 21 to 25, eight men and eight women, who had never shot, were randomly assigned to one of two string grips, the thumb or the reversed grip, and to one of four foot positions; after forty minutes of instruction and thirty practice arrows they shot eighty measured ones. With the reversed grip the trapezius and finger extensor worked harder, and fatigue arrived sooner and more strongly, with the drop in score that goes with it; those using the thumb scored 77% more points on average than those using the reversed grip.
The two studies point the same way, but neither says what an archer would want to know. The first measures muscle activity over three shots with a grip just tried, not fatigue over a session; the second compares the thumb with a grip no target archer uses, and three fingers were not in it. That is why the claim that a thumb grip on the string is less tiring than a finger grip is reported, on the strength of the one study on archers, and the claim that the way the string is held changes a beginner's score is transferred: the registry counts those sixteen adults, at their first session, as the general population rather than as archers. For someone starting out the grip seems to matter, but nobody knows how much it matters among the grips they would actually be taught.
The thumb release and the bow arm
People who shoot traditional bows with a thumb draw often say that most of the work is done by the bow hand, which keeps the wrist still while the string goes. The only measurement comes from a Malaysian laboratory, on one person: a horseback archery world champion with a 45-pound traditional bow and the khatrah release of the Mameluke technique, in which the string is held by the thumb and, at release, the bow wrist deviates towards the little-finger side while the hand punches forward and turns the bow. Electromyography of three muscles per wrist, converted into forces using muscle cross-sections taken from the literature, gave 176 newtons at release for the flexor carpi radialis of the bow arm against 132 on the string arm; the extensor carpi ulnaris reached 88 against 60, the finger extensor 50 against 21.
Between full draw and release the flexor carpi radialis of the string arm held level, from 136 to 132 newtons, while that of the bow arm kept rising, from 159 to 176. The detail most closely tied to the technique is the extensor carpi ulnaris of the bow arm, which at release worked almost twice as hard as the finger extensor and peaked in the follow-through, as you would expect from that wrist deviation. In 2020 the same data, extended to ten muscles and published in a journal with questionable editorial standards, add that the trunk worked harder on the drawing side. An older finding comes from kyudo, where the string is held by the thumb inside the glove: in three Japanese archers, during the first half of the draw, the bow wrist was fixed by the extensor carpi ulnaris and the string wrist by the flexor carpi radialis, so the two wrists had different jobs, but the forces on the two sides were not compared.
The claim that with a thumb release the bow arm works harder than the string arm is therefore reported, with three limits that carry weight: a single person, forces estimated rather than measured, and no comparison with a three-finger release. It is not known whether the difference comes from the thumb, from the bow or from the wrist rotation specific to the khatrah.
How the fingers open in a three-finger grip
For the three-finger grip, the one used by almost every recurve target archer, we know precisely how the fingers open, but for one archer only. In 2009 an Austrian group used eight infrared cameras at 500 frames per second to film the string hand of an archer from the Austrian B team: aged 49, with an indoor record of 1131 out of 1200 and 56 usable shots at 18 m. The joints at the base of the fingers barely moved, between 6 and 9 degrees, while most of the opening happened at the two outer joints. The index finger moved most, 81 degrees in all, followed by the middle finger with 58 and the ring finger with 48, and the leading joint changed from finger to finger: the end joint in the index and ring fingers, the middle joint in the middle finger.
The same study also looked at speed. With the arrows split into those in the X and those scoring eight or less, the peak speed at which the middle finger's middle joint opened was similar in both groups, and it was not linked to score as a correlation either. Two years later the same authors re-read the same arrows, splitting them into good and poor ones: on the poor arrows the curve of the middle finger's base joint strayed further from the group average than on the good ones, so the very joint that moves least was the one that changed most between a successful shot and a poor one. In the registry the claim that at the release the fingers open mainly at the middle and end joints, and barely at the knuckles is reported, and so is the claim, written in the direction of the result, that the speed at which the fingers open does not separate good arrows from poor ones. It is a careful description of a single three-finger release, and it does not say whether it holds for other archers, other levels or other grips. I cover what happens to the string in the milliseconds that follow in the article on the release.
Where the measurements stop
Everything known about comparing grips comes from people who had been using a new grip for a few minutes: experienced archers in 1990, beginners at their first session in 2007. That is the biggest limit, and the first study shows it on its own: with the thumb the group went from 175.7 to 42.0 centimetres in nine arrows, and there is no way of knowing where it would have settled after a few months. None of these studies compares archers who have each shot their own grip for years, or the same people before and after a period of training with a new grip.
The question a recurve archer asks most often, one finger above the nock or three below, has never been put to the test either: none of the grip studies compares these two versions of the three-finger grip. The same goes for score over a full round, for fatigue over a whole session, for the steadiness of the hold measured with a bow in hand, and for the thumb release compared directly with a three-finger release on the same bow.
What I take from it for the range
These measurements do not tell you which grip to choose, and they do not point to one grip that is better than the others across the board. They do say a few things that are useful to anyone thinking of changing, and I keep them in this order.
The first concerns the cost of the change. Nine international-level archers, with the thumb or the reversed grip, shot their first three arrows into a group whose contour averaged more than 1.7 metres; with two fingers, the grip closest to their own, the group was already similar to their usual one on the first end. Anyone who changes grip should therefore expect a period in which the arrows spread, longer the further the new grip is from the old one, and judge it over many arrows and several sessions, not on the first end.
The second concerns two fingers. On these data there is no reason to regard them as a fallback grip: in the only comparison with a bow in hand they did not group worse than three. Then again, an archer who shoots three fingers and is happy with it has no reason in the measurements to switch to two, because the measurements find no advantage either way, and the two-finger grip was also the one with the highest muscle activity. Bear in mind that the data come from strong archers, over a few arrows.
The third concerns the thumb, and it is for those who shoot traditional or Asian bows. The lower muscle activity is an average over six muscles, and muscle by muscle it is significant only in the flexor digitorum superficialis, and the same archers found the thumb the hardest grip. With a thumb release, moreover, the only archer measured loaded the bow wrist more, so with this grip I would watch the hand that holds the bow as well as the one that releases. A slow-motion video helps you see both hands together, and check that the anchor, which changes with the grip, stays the same from one arrow to the next.