Beta waves and archery accuracy: Gaffar 2026 study

Twelve young archers from Surakarta, Indonesia, shot 15 times with an electronic bow at a target shown on a television screen while wearing an EEG headband. The authors wanted to know whether beta power, which they treat as an index of concentration, went together with score and group size. No such link appeared: the correlation with score was 0.145 (p = 0.652) and the correlation with shot dispersion was -0.327 (p = 0.300), both far from significance.

The mean score was 131.17 out of 150 (standard deviation 11.23), ranging from 104 to 144. Dispersion, the average distance of each shot from the centre of that archer's own group, averaged 11.00 cm (standard deviation 5.17), from 5.17 to 22.25 cm. Mean beta power was 39.18 µV² (standard deviation 2.63). Both correlations point the way the authors expected, with more beta going with a slightly higher score and a slightly tighter group, but with twelve people coefficients of this size cannot be told apart from chance, and in their discussion the authors themselves call any reading of the direction highly speculative. In the discussion the authors point to an archer at 42.24 µV², which they describe as the highest value, who was below average for both score and grouping, while the one with the best results on both measures sat at 38.65 µV². The results section, however, gives the highest value to a different archer, at 43.61 µV².

The participants, 6 boys and 6 girls aged 12 to 15 (mean 13.92), were recruited from schools in Surakarta among those with at least one regional competition behind them, right-handed and with normal eyesight; reported experience ranged from 2 to 10 years, with a mean of 5. Shooting took place in a laboratory room measuring 2.5 by 1.6 metres. Each archer used a smart bow (Wonderfitter's Houyi 2), a device with motion and position sensors linked to an app, and shot 15 times at a 43-inch television 1.5 metres away showing a virtual 122 cm face, allowing each archer up to 30 seconds a shot. No arrows were released.

The score is the total of the 15 shots as calculated by the app. Dispersion comes from the coordinates of each shot on the virtual face: for each archer the authors found the mean centre of the shots and then the distance of each shot from it, in centimetres on the simulated target. EEG was recorded with a Muse 2 headband at 256 Hz, using only two frontal electrodes (AF7 and AF8), throughout the series; the signal was filtered between 1 and 30 Hz, cleaned of artefacts with independent component analysis and by hand, and reduced to mean power in the 12.5 to 30 Hz band. Each archer therefore contributes one EEG value, one score and one dispersion figure, and the Pearson correlations rest on these twelve values per measure.

The biggest limitation is that this is not archery in the usual sense. An electronic bow aimed at a television a metre and a half away has none of the distance, draw weight, arrow, wind or release of a real shot, and the paper does not explain how the device turns the movement into a point on the target or how accurately it does so. The mean of 131 out of 150 and the dispersion in centimetres are therefore measures of the simulator, and cannot be compared with a score at 18 or 70 metres.

The sample is twelve people, and with twelve people a correlation has to exceed about 0.58 to reach significance: a moderate relationship would have gone unnoticed, so the null result rules out very little. Beta power comes from two electrodes on the forehead, where facial muscles and the movement of the draw easily contaminate the signal, and it is averaged over the whole series, with no separation of aiming from the rest of the shot and no resting baseline. That frontal beta power measures concentration is an assumption of the authors, not something the study tests.

There are also some inconsistencies. The highest beta value is given to two different archers, 43.61 µV² in the results and 42.24 µV² in the discussion. The archers are called novices, under an age-based definition, yet report five years of experience on average. The abstract presents the opposite direction of the two correlations as confirmation that score and grouping depend on different mechanisms, while the discussion, more soundly, calls it highly speculative; in any case the two correlations have opposite signs simply because one measure rises as shooting improves and the other falls.

For anyone who shoots, the study does not say that concentration is unimportant. It says that a crude measure of frontal electrical activity, taken over fifteen shots on a simulator, did not separate the youngsters who scored more or grouped tighter. That is a reason for caution about training EEG headbands sold as concentration metres: in this study one of the highest readings belonged to an archer who shot below average.

The useful part lies elsewhere and has nothing to do with EEG. The authors keep score and group dispersion apart, and use a tiny example from their data (135 points with 9.24 cm of dispersion against 136 points with 9.64 cm) to show that a tighter group does not guarantee a higher score, because the group can sit off centre. Anyone who analyses their own shooting will know the distinction: where the group lands depends largely on sight adjustment, how big it is depends on how repeatable the shot is, and a single score blends the two.