Weight set-up for barebow bows: Pietra 2023 study

The study looks for the best weight set-up for barebow, where stabilisation is limited to weights screwed onto the riser at two attachment points, centre and lower. Eleven configurations were tested: the bare bow, a single centre weight of 200, 250 or 350 g, six two-weight combinations and, for comparison, a recurve V-bar. Seven archers from a Johannesburg club rated them without seeing the weights, scoring each from 1 to 5 on three aspects: balance while aiming, vibration on release and the bow's behaviour after the shot. The same configurations were measured with an accelerometer at five points on the bow and a motion sensor next to the pressure button.

Two-weight configurations were preferred to single weights, with the heavier weight at the centre. With the signal filtered below 500 Hz, measured vibration tracked the ratings: near the bow hand, for instance, the correlation between overall amplitude and rating was -0.895 (less vibration, better rating). The rating for the bow's behaviour followed mainly sideways displacement, perpendicular to the plane of the bow (r = -0.763), and rotation within the plane of the bow (r = -0.598). The duration of vibration made no difference. The authors name as best 250 g at the centre with 200 g below, and 350 g at the centre with 200 g below; the 350 and 250 combination, also well rated, was dropped as too heavy for a long competition.

Score decided nothing. One archer shot 30 arrows at 50 metres with each configuration, scoring from 228 (bare bow) to 262, but according to the authors' simulation differences of that size are within normal variation, and comparing configurations on score would need 900 to 1,050 arrows each.

The bow was a 2019 Gray AIX riser with long Win&Win Wiawis NS wood limbs at 32 lb nominal, tuned with the recurve procedure; the weights were Gas Pro at 200, 250 and 350 g. Shooting was three fingers under the nock with gap aiming. The PCB triaxial accelerometer, sampled at 12,000 Hz for 15 seconds, was placed in turn on the upper limb, near the upper limb pocket, by the pressure button, near the lower limb pocket and on the lower limb, with six shots per position. Each signal gave peak-to-peak amplitude, RMS value and a fatigue damage measure, first on the full signal and then filtered at 500 Hz, the band the human arm is sensitive to.

The motion sensor, an MPU6050 on an Arduino board, recorded displacements and rotations over four shots per configuration. Ratings and measurements were compared with Pearson correlation across the 11 configurations. The score is from 30 arrows at 50 metres on a 122 cm face, shot by a single advanced archer.

Each correlation rests on 11 points, one per configuration, using mean ratings from seven archers that often coincide, and the authors calculated dozens of correlations and highlighted the strongest; some are positive, meaning more vibration and a better rating, and are explained by saying the archer is not sensitive to that component. The 500 Hz filter was introduced after the full signal gave weak links. The paper does not say who shot during the vibration and motion measurements, or whether it was always the same archer. The motion sensor values (displacements over 100 mm and rotations over 100 degrees for the bare bow) are very large for a bow moving in the hand after the shot, and suggest the drift typical of a low-cost sensor; the authors do not discuss this, and it is my own reading.

The bow was tuned as a recurve and shot with gap aiming, not with the string walking many barebow archers use; with another riser, other limbs or another technique the preferences might change. Dampers, allowed in barebow since 2022, were left out by choice. The score, as the authors acknowledge, supports no conclusion. The paper appeared online in 2023 and was later placed in a 2025 issue.

For barebow archers the study offers a starting point, not a rule: with that riser and those limbs, two weights were preferred to one, with the heavier at the centre, and the preferred combinations were those that vibrated least near the bow hand. This is measured as feel and as vibration, not as accuracy: with 30 arrows per configuration no difference in score can be shown.

Perhaps the most useful finding concerns method. Comparing two set-ups over a few ends, as is often done on the range, is not enough by the authors' calculations to reveal differences of a few points, which would take 900 to 1,050 arrows per configuration. The authors also note that the archer adapted to each configuration within a few arrows, and from this they suggest that good technique makes up for the bow's shortcomings; they add that total mass matters for fatigue in a long competition.