On the overarm throw: Evaluating technique with a novel metric for performance.

Abstract

Throwing is a highly complex movement pattern. Being refined over millennia (Lombardo & Deaner, 2018a), the body of research surrounding the throwing motion is extensive. However, a predominant focus on baseball pitching may have led to a distortion in the way we interpret similar throwing skills. Furthermore, performance is often considered as ball velocity, or accuracy, with the two rarely being considered together. The aim of this work was, therefore, to investigate the throwing skill in a more general context, and develop a measure for performance that incorporates both ball velocity and throw accuracy. Aim 1 sought to establish the movement patterns utilised for this general throwing skill. Kinematic data for 28 experienced positional baseball players revealed that the pattern of movement was largely similar to that which has previously been reported for American football (Fleisig et al., 1996b) and handball (van den Tillaar et al., 2011). Initial rotations away from the target were followed by rapid accelerations towards it, with joint angles and velocities comparable to these similar skills. Stride length was considerably shorter at 34-38% standing height than had previously been described, suggesting that participants may have adapted their movement due to task constraints (Cook et al., 2000). Coefficient of variation analysis later indicated that the early establishment of a stable throwing foundation is key. Lower-extremity kinematics generally varied less during this time, with variability in distal kinematics reducing as the motion progresses. Aim 2 sought to develop and validate the accuracy-integrated metric (AIM) for objectively evaluating throwing performance. AIM demonstrated a capability for considering all aspects of the throw, and was therefore deemed a viable option for evaluating throwing performance. Weightings for ball velocity, accuracy, and throw distance ensured that faster, more accurate, or longer throws were rewarded providing other aspects remained the same. Consequently, a holistic assessment of any given [throwing] movement pattern should now be possible. The final part of this work sought to identify relationships between technique characteristics and AIM. A mixed-model analysis revealed that only forward trunk flexion at ball release (BR) was statistically related to AIM score, however, it appeared to lack practical significance. For every 1° increase in trunk flexion, AIM score was only expected to increase by approximately 1 point. It was, therefore, concluded that additional variables might better explain AIM than those included in the present work. As such, these should be considered in future investigations of throwing performance. A foundation for future throwing research to build upon, and a performance metric for objectively assessing performance has been provided. It is hoped that these approaches will be adopted in a variety of sports where velocity and accuracy are key components of performance.