A fast start is considered to be crucial in skeleton with marginal gains in start performance perceived to make meaningful improvements to overall chances of success. Currently, knowledge surrounding the underlying determinants of start performance is sparse and training practices are based on limited scientific evidence. A series of investigations were conducted to advance this understanding. Initial observations revealed similarities between dry-land push-starts and those on ice tracks. However, the number of steps taken before loading was adjusted to seemingly accommodate unique track profiles and appeared to be influenced by physical capacity. Consequently, skeleton athletes completed multiple two day testing sessions (four physical tests and biochemical analyses) across two seasons, alongside dry land push track tests. Additionally, body composition was assessed either side of selected training and competition blocks. Three independent physical factors (countermovement jump, sprint and force-power characteristics) were identified as fundamental to a fast push start and a regression equation comprising these variables provided an accurate prediction of start ability (R^2 = 0.86). Testosterone appeared to influence push track performance and lean mass accrual, however, retrospective biochemical analyses were deemed to have limited utility in applied practice. Conversely, the importance of monitoring body composition, particularly across competition seasons, was apparent and dual energy X ray absorptiometry is an appropriate tool to detect meaningful changes. A continuous sled velocity measure confirmed the contribution of physical capabilities to both the distance and velocity attained before loading. Importantly, loading phase success appeared independent of physical ability, perhaps warranting specific loading technique training. Finally, a trade off between pre load velocity and load effectiveness was evident, and experimentally modifying loading distance provided a promising approach to improve performance in developing athletes. This thesis has informed skeleton training by identifying factors which contribute to performance, alongside approaches to thoroughly evaluate athlete progression and has introduced processes through which start performance can be enhanced.

Skeleton is one of the three Olympic sporting disciplines to be held in the manmade bobsleigh tracks. The sport of skeleton uses a one-man sled, on which the athlete travels headfirst down a mile long track reaching speeds of up to 147 km/h. As with many sports the engineering of the equipment is playing a greater role in the overall performance of the athlete. Although the sled alone cannot win medals a poor choice of equipment can be the difference between winning and losing. The primary focus of this research is on the trajectory and response of the sled frame and how these relate to athlete perception during a descent and overall performance. Sleds were instrumented with accelerometers and strain gauges that enabled the mechanical behaviour of the sled to be determined quantitatively. Qualitative data comprised of athlete training logs (mainly from the author), provided information about the feel and perception of the run. Tests were made on whole tracks, dedicated push-tracks and in the laboratory. In addition this PhD has touched on aerodynamics and runner-ice interaction. The thesis is split into three main sections: (1) The initial push phase of a descent was investigated at the Torino Sliding Centre and Calgary Olympic Park with a sled instrumented with an accelerometer. Using a single axis in the forwards direction of the sled determined the sensitivity of the measuring and acquisition device along with the capabilities and quality of information gained. Through analysis it is possible to identify the dynamics that occur during a push start and how to interpret them in order to improve athlete performance during the push start. (2) A whole descent at the Koenigssee International Race Track was measured using a three axis accelerometer. The dynamics at specific track locations were examined in detail and linked with athlete perception. Comparison of multiple descents enables the sled trajectory to be quantified to determine the overall success of the resultant trajectory. This analysis shows there is scope for maximizing athletic performance in conjunction with quantitative instrumentation of the equipment. (3) Complete descents at the Lake Placid Olympic Park were made on a sled instrumented with rosettes of strain gauges. The strain gauges were calibrated in the laboratory. Analysis of strain gauge data from the track showed the extent of deformation of the frame upon entering and exiting curves and while under the g-forces experienced, again this data is compared with athlete perception. Consideration is briefly given as to how these dynamic measurements can be used to evaluate current and future frame designs.