Virtual and real archery shooting: Dal 2024 study

Twenty-two university students with no archery experience, eight women and fourteen men aged between 20 and 24, shot ten real arrows at 18 metres and ten arrows in virtual reality, using a headset and a game, while their heart rate, heart rate variability and breathing rate were recorded. Each did both, in counterbalanced order, after a resting measurement.

The two situations do not resemble each other as the authors expected. Mean heart rate was 109.1 beats per minute shooting the bow and 97.0 in the headset, a rise of 32.5% over rest against 17.5%. RMSSD, the variability index most closely tied to vagal tone, was 38.2 milliseconds with the bow and 30.9 in the headset: relative to rest it rose by 10.3% when shooting for real and fell by 19.8% in virtual reality. The ratio of low to high frequency power was higher in the virtual condition. Breathing rate, the standard deviation of the intervals between beats, and low and high frequency power taken on their own did not differ.

On the self-reports: no difference in stated anxiety, 4.76 in both conditions on a ten centimetre scale, and none in confidence. Shooting the bow was rated more difficult (5.43 against 4.00) and more effortful (4.86 against 2.90). The score, by contrast, was higher in virtual reality: 86.9 over ten arrows against 78.8 with the bow.

A repeated measures design within the same people: three periods of four minutes each, rest, real shooting and virtual shooting, with the order of the two shooting sessions counterbalanced across participants and fifteen minutes between them to change equipment. In each session ten arrows at 18 metres at an 80 centimetre target face, preceded by six familiarisation shots to adjust the sight or settle into the headset; pacing was self-chosen, around 20 to 25 seconds per arrow according to a pilot on five novices, and participants were told to use the same technique in both settings. Anyone who finished the ten real arrows inside four minutes shot more, but only the first ten counted.

Heart rate was recorded by ECG in a lead II configuration with three electrodes at 1,024 Hz, and breathing with a belt at umbilical level. The intervals between beats were visually inspected for artefacts and irregular beats, then analysed in the time domain (NNmean, SDNN, RMSSD) and in the frequency domain by Fourier transform; the power values were log transformed before comparison. Anxiety and confidence were reported before each session, difficulty and perceived exertion immediately after, on ten centimetre visual scales. The comparisons are paired t-tests, first on absolute values and then on percentage change from rest. The virtual application was a commercial game on an HTC Vive Pro headset, and a simulated archer appeared in the virtual environment, which participants were told to ignore. All sessions ran between noon and three in the afternoon.

These are twenty-two beginners, not archers, and the authors put that first among their limitations. They had completed a fourteen week course, and the real shooting session was their final examination, with a minimum number of points needed for course credit: the virtual session carried no marks. This difference, which the authors do not discuss, is enough to account for much of the extra difficulty and exertion reported, and perhaps for the higher heart rate too, because an examination is not a shot.

The two scores are not on the same scale. How the game awards points is not stated, nor whether the virtual environment reproduced the same target face and the same distance, so the comparison between 86.9 and 78.8 does not measure accuracy across the two settings. The bow has to be held at draw weight, the headset does not: the physical demand differs whatever instruction is given about technique.

The recordings last four minutes and ten arrows, which means heart rate variability computed over short windows with the movement of the shot inside the recording. Several comparisons sit close to threshold, RMSSD at p 0.04 with a medium effect, and no adjustment is made for the number of comparisons. The rise in RMSSD during real shooting, alongside a higher heart rate, runs against what the authors themselves predicted, and the explanation they offer, a rebound effect during the pauses between arrows, is a hypothesis advanced after seeing the data. The caption to figure 5 gives perceived exertion as higher in virtual reality, contradicting both the table and the text. The standard deviation of the participants' height, 1.8 centimetres across twenty-two people, is not credible.

For an archer the useful finding is that a virtual reality simulator does not reproduce what the body does during a shot, at least in beginners. Heart rate rises by less than half as much as it does when shooting for real, and the index tied to vagal tone moves the other way. Anyone thinking of using a headset to train the management of arousal, the part of the shot that runs through breathing and heart rate, would be training something else.

What remains is that the comparison was made under very different conditions, with the bow under examination and the headset not, on people who had been shooting for fourteen weeks. The study says nothing about whether a simulator helps in learning the sequence of the shot or in practising when the range is out of reach, which is where the authors began, and it says nothing about experienced archers: they name that comparison themselves as the next step.