Improved accuracy and reliability of sweepback angle, pitch angle and hand velocity calculations in swimming

The estimation of forces in swimming using the quasi-static approach (Schleihauf, In: J. Terauds, J.P. Clarys (Eds.), Swimming III, International Series on Sports Sciences. Vol. 8, University Park Press, Baltimore, 1979, 70-109) has been popular in recent years as propulsion is an important determinant of performance. The aim of this study was to establish the accuracy and reliability of current and newly proposed procedures for the reconstruction of hand velocity, sweepback angle and pitch angle from underwater three-dimensional video analysis. A full-scale mechanical arm capable of simulating a controlled and highly repeatable underwater phase of the front-crawl stroke was filmed for a set of five trials. A seven-point model of the arm and hand was then digitised at 25 Hz. Hand velocity, sweepback angle and pitch angle were calculated using the procedures of Schleihauf (1979), Berger et al. J. Biomech. 28 (1995) 125-133 and a newly proposed procedure (Lauder). Statistical comparisons were made between procedures to establish their relative accuracy and reliability throughout the stroke. The mean absolute error in measurement of hand velocity between points on the hand was very small (+/- 0.04 and +/- 0.06 m s(-1) in the x and z directions, respectively). The mean errors in sweepback angle and pitch angle were, respectively, 9.3 degrees and 7.6 degrees (Berger), 10.1 degrees and 8.1 degrees (Schleihauf and 10.7 degrees and 7.0 degrees (Lauder). Agreement between procedures showed the standard error between Schleihauf and Lauder to be the least (Schleihauf and Lauder, 0.4 degrees; Berger and Schleihauf, 1.3 degrees; Berger and Lauder; 1.6 degrees). The use of four points in the reconstruction of the orientation of the hand (Schleihauf and Lauder procedures) was shown to be less sensitive to errors in the digitising procedure. The reconstruction procedure proposed in this study (Lauder). further reduced the sensitivity to digitising error in the reconstruction of sweepback and pitch angles in swimming.