Archers are now offered machines that promise to train directly what happens in the body before the arrow: headbands with electrodes that turn brain waves into a sound, chest straps linked to an app that paces the breathing, computer programmes in which you follow moving targets with your eyes, headsets for shooting in virtual reality. Anyone who sees them for sale, or hears them pitched on a course, has a simple question: do they help you shoot better?
The studies give a less simple answer, but one thread runs through almost all of them. Whatever the machine trains usually moves: the chosen brain rhythm rises, heart rate variability goes up, concentration on the test improves. The score, or the anxiety it was meant to lower, moves much less, and when it does move it does so in small studies or in a journal with questionable editorial standards. There are few experiments on archers, and several of the claims in circulation rest on other sports or on people who do not shoot.
How training with a feedback signal works
A sensor measures something the archer cannot feel unaided, such as a rhythm in the brain's electrical activity, the variation from one heartbeat to the next or a slight sag in the drawing hand, and turns it in real time into a sound, an image or a number; by seeing or hearing the signal, the person learns to push it in the desired direction. When the signal comes from electroencephalography the method is called neurofeedback, and when it comes from the heartbeat, the breathing or the muscles it is called biofeedback.
For any of this to reach the target, two separate steps are needed: the training has to change the signal, which happens often, and the changed signal has to lead to better shooting, which is where the evidence thins out. Computer-based perceptual exercises and simulators do not read the body, but they are offered to the same archers with the same promise, and that is why they are on this page.
Neurofeedback: the signal rises, the score does not
The most direct controlled trial comes from an Indian university. Twenty-four university archers, with about four years of shooting behind them, were split into two groups of twelve: the first did twelve neurofeedback sessions over four weeks to raise the sensorimotor rhythm, between 12 and 15 hertz, and lower theta, while the second did nothing. Before and after the programme everyone shot a 72-arrow competition at 70 metres.
The trained signal moved: the ratio of sensorimotor rhythm to theta rose from 0.331 to 0.395 in the neurofeedback group and barely shifted in the controls, and after the programme the difference between the groups was significant (p = 0.018). The score did not follow. The trained group went from 527.3 to 529.9 points and the control group from 526.6 to 526.7, and the gap between the groups after the programme, just over three points, does not stand up statistically (p = 0.561); in the trained group the number of Xs did not change either, 5.17 before and after. In their abstract the authors write that neurofeedback makes scoring more regular, but their tables show no difference between the groups in score or Xs.
A 2024 review of neurofeedback in precision sports paints a more favourable picture. It reports an early-1990s study of twenty-four pre-elite archers in which feedback tied to the left hemisphere raised the score and feedback tied to the right lowered it, and this is the only favourable result in archers; it comes second-hand, though, from the review's description, and the same review describes the Indian study as an improvement in accuracy and consistency in twenty-eight archers, whereas the original paper has twenty-four archers and no difference in score. In other sports the review finds mixed results: in twenty-eight elite biathletes frontal theta rises with training but shooting does not improve relative to the controls, while in thirty-two women darts players neurofeedback improves performance as much as motor imagery and physical practice do.
That is why in the registry the claim that EEG neurofeedback improves an archer's score is contradicted: the only controlled trial on archers whose data can be read goes against it, and it is a single small study in a regional journal. What EEG studies of archers measure, and why the sensorimotor rhythm is the signal most often trained, is covered in the article on EEG and neuroscience in archery.
Heart rate variability biofeedback
The biofeedback most widely used by athletes works on the heartbeat. A chest strap measures the interval between one beat and the next, an app draws it on the screen and paces the breathing towards the rate at which the swings in heart rate become widest. The mechanism is well described: breathing at around six breaths a minute brings heartbeat and breathing into phase and variability peaks, and in one controlled study four sessions were enough to raise it, which muscle biofeedback combined with relaxation did not do. The claim that breathing at around six breaths a minute brings heart rate variability to its maximum is transferred, because it rests on a review of fifteen studies of healthy adults and two narrative reviews, the exact rate varies from person to person, and nobody has looked for it in an archer.
Whether raising variability lowers the anxiety of people who have to perform is a separate question. Almost all the evidence comes from a 2018 systematic review of slow breathing, which gathers several biofeedback studies on healthy people, and the three that measure anxiety point in two directions. In dance conservatoire students, ten sessions lowered anxiety compared with a group that had no intervention, and variability rose only in those doing biofeedback. In healthy young adults, on the other hand, biofeedback done before bed raised variability during sleep over the following nights without changing anxiety, and in another study six sessions changed neither variability nor the questionnaires.
Biofeedback, then, changes the heartbeat fairly reliably and anxiety only sometimes. The claim that heart rate variability biofeedback reduces the anxiety of people who have to perform is contradicted, because the indirect evidence there is leans against it: the favourable study is a single one, outside archery, and does not concern the anxiety of a competition.
Target shooting offers a closer clue, about the body rather than reported anxiety. In five young pistol shooters, after eight weeks of mental training that included relaxation through breathing, the ratio of the LF to the HF band of variability fell from 3.4 to 1.0 while they imagined a crisis in competition, and from 1.9 to 1.0 while they imagined a successful performance; at rest nothing changed, and the other indices, RMSSD included, stayed put. These are five people with no control group, and the condition that changed most was also the one furthest out before the programme, so part of the fall may be a return towards the mean. The claim that after weeks of mental training the body reacts less when the competition is imagined is transferred, and it rests on an index that needs to be read with great care.
Which number to look at on the screen
Heart rate variability apps show plenty of numbers, not all of which mean what they seem to. According to the reference methodological review, from 2017, RMSSD, the square root of the mean of the squared differences between successive beats, estimates vagal tone and is little affected by breathing, and two minutes are enough to compute it. The ratio of the low band to the high band, by contrast, was long read as the balance between the sympathetic and parasympathetic systems, but the low band mixes sympathetic, vagal and baroreflex activity, and the authors consider its meaning uncertain, even though about 65% of papers still base their conclusions on it. The high band, in turn, only works as a measure of vagal tone if breathing stays between 9 and 24 breaths a minute, and the slow breathing of biofeedback takes it outside exactly that range.
In archers the indices have never been tested against each other, but two studies show why the rule matters. In twenty-two university students who had never shot before their course, shooting ten arrows with a bow and ten in virtual reality, the band ratio changed from rest so differently from one person to the next, -114 ± 476% with the bow and -58 ± 368% in the headset, that it could not tell the two conditions apart, while RMSSD could. In eighteen elite Turkish archers measured seated before a training session, RMSSD was 82.8 ± 52.9 milliseconds, a standard deviation of almost two thirds of the mean, so the comparison that makes sense is with your own values and not with anyone else's.
That is why the claim that to measure an archer's heart rate variability you need RMSSD, not the ratio between the bands is transferred: it holds as a methodological guideline borrowed from psychophysiology, and in archery it has yet to be tested. The same review adds two conditions that the shooting line does not offer, since vagal tone cannot be read while the body is moving and a comparable measurement needs the same time of day, the same sleep, and no coffee or hard training beforehand. How heart rate variability fits into the rest of training is covered in the article on HRV in archery.
Computer-based perceptual training
NeuroTracker is a programme in which, wearing 3D glasses in front of a television, you follow a few moving targets among identical ones. In archers there is a single experiment, reported in two papers: twenty archers from an Indonesian university club, ten adding twelve sessions over four weeks to their training and ten doing the club's usual training. The dose is small, twenty trials of eight seconds per session, about 32 minutes of task time across the whole month.
In the first paper the outcome is concentration, measured with the grid test, and the computer group gains 8.5 points against 4.5 for the group on its usual training. In the second the outcome is the score at 30 metres: up 32.9 points against up 19.7. There are three caveats. The control group also improves a great deal, so the part attributable to the computer is about thirteen points; the controls do not spend an equivalent time at a screen on a dummy task, so novelty and the attention received weigh alongside the exercise; and the paper on scores sits in a journal with questionable editorial standards. The registry reports the result on scores but does not use it to decide the status, and the claim that computer-based perceptual-cognitive training improves the score remains not studied.
A signal while you shoot
Simpler signals, which arrive while you shoot, have also been tried on archers with a bow in hand, and two of them feed back something from the body. The first, in a 1993 study, was a tone that sounded when the drawing hand sagged past a threshold in the final phase of the draw: with the tone the draw became steadier in five archers out of six, while in the sixth the mean distance from the trend line rose from 2.53 to 4.27, and the score did not improve significantly. The authors suggested that the extra information drew attention away from other things, and the series with the tone always came after the one without, so part of the gain may simply be practice. The claim that a sound cue that flags collapses in the draw makes the draw steadier is contested, because the same study answers yes for five archers and no for the sixth.
The second comes from Japan, and it is the closest thing to biofeedback that anyone could try. Thirty-six experienced archers shot six arrows at 50 metres twice, with and without their own heartbeat in headphones, in counterbalanced order. With the heartbeat the score rose from 42.3 to 44.4 out of 60, shot timing became steadier from arrow to arrow, with the coefficient of variation falling from 16.7% to 12.5%, and the release fell more often between beats, on 139 arrows out of 216 against 102. A metronome at a fixed rate, tried on twelve university students, changed nothing. The claim that listening to your own heartbeat while shooting raises the score is reported, and the picture is fragile: two tests of six arrows, a regional journal, no replication, and six archers out of thirty-six found the heartbeat in their headphones a nuisance.
The virtual reality simulator
The headset is offered on a similar argument, that it trains the mental side of the shot even without a bow. The only direct comparison was made in twenty-two students at the end of a fourteen-week archery course, who shot ten arrows at 18 metres with a bow and ten with a headset and a video game, in counterbalanced order. The body responded differently: with the bow heart rate reached 109 beats a minute against 97 in the headset, 32.5% above rest against 17.5%, and RMSSD rose by 10.3% when shooting for real while it fell by 19.8% in virtual reality. Real shooting was rated harder and more effortful, while reported anxiety was identical in the two tests, 4.76. The score was higher in the headset, 86.9 against 78.8 over ten arrows, which the authors read as a sign of an easier simulation.
Two things weigh on the reading. The bow test was the course's final exam, with a minimum score needed for the credits, while the headset test did not count; and for someone who has been shooting for a few months, drawing a bow is still a real physical effort, which a controller with no draw weight does not ask for. The claim that shooting in virtual reality does not load the archer the way a real bow does is transferred, because it comes from students at the end of a course rather than from trained archers, and the claim that scores are higher in virtual reality than with a real bow is reported, for novice archers and a single system.
Where the studies stop
Taken together, this research shares one limit: it measures the signal well and the shooting much less. In archers there is one controlled trial of neurofeedback and one experiment with a computer, with twelve and ten archers per group, and in both the control group does no dummy activity of equal length. Heart rate variability biofeedback has not been put to the test on archers, and the variability indices have never been compared with each other on the shooting line. Finally, none of these studies follows people for more than a few weeks, so nobody knows whether an effect, when there is one, lasts.
What I take to the shooting line
I do not conclude that the machines are useless, but that they should be judged on what they have shown, and the first point is to keep the signal and the shooting apart. Nearly all these tools show a graph that improves session after session, which is the signal they train; whether the scores change can only be seen by comparing enough arrows before and after a period of use, in the same conditions, and those arrows are what I would judge the tool on, not the app's graph.
The second point concerns heart rate variability. If I were measuring it, I would read RMSSD at rest, at the same time of day and in the same conditions, and compare it with my own values from other days, not with other archers'; the band ratio, which many apps present as a stress index, I would not read that way. Slow breathing paced by biofeedback changes the heartbeat fairly reliably, and anyone who finds it useful in their routine can keep it for that reason; the studies do not show that it lowers competition anxiety.
The third concerns signals during the shot. The heartbeat in headphones and the tone on the sagging draw are the two body signals tried on archers with a bow in hand, and their results are small, unreplicated and not the same for everyone: with the tone, the draw became less regular in one archer out of six. An extra signal is also extra information to handle while shooting, so I would try it only in training, watching the group on the target as well as the feel.
Finally, the headset does not put the body in the conditions of real shooting, and the score it gives cannot be compared with the one on the target face. It may help to rehearse the shot sequence when you cannot shoot, but nobody has checked whether training in a headset changes the score with a bow.