Monitoring and modelling swim performance

The capacity to objectively track, predict and evaluate changes in competitive performance is a critical component of an effective athlete development programme. This thesis focused on developing objective analytical tools for monitoring and assessing performance progression in swimmers using the mixed modelling procedures in the Statistical Analysis System (SAS) software. In the first study, a systematic review of estimates of age of peak competitive performance of elite athletes from a variety of sports established the need to generate progression-monitoring tools specific to the sport of swimming. Four original-research studies focused on different aspects of athlete development are then presented: benchmarking, talent identification, career training and performance-enhancing strategies and interventions. First, I developed quadratic trajectories to track the career development of Olympic top-16 swimmers using their annual best performances. These trajectories provided event-specific progression benchmarks that can be used to monitor and assess performance changes of developing swimmers. Secondly, a comparison of the accuracy of four methods for predicting future performance provided some evidence that this trajectories method may be useful for identification and early selection of swimmers tracking towards Olympic-qualification standards. Thirdly, I used a novel application of mixed modelling to quantify the relationships between swim-specific and non-specific career training hours and performance in competitive swimmers. The focus of the final investigation was the development of a method to assess the effects of strategies and interventions on swim performance progression. By quantifying the deviation of top swimmer’s performances from their individual trajectories after they joined the centralised elite squad, we showed that Swimming New Zealand’s centralisation strategy took several years to produce substantial performance effects. This method also produced annual performance estimates for New Zealand swimming as a whole, and for each New Zealand swimming club, creating a tool that can be used to evaluate performance progression at both the national level and the club level. In this thesis, I have demonstrated that mixed modelling can be used to provide objective solutions appropriate for monitoring and assessing the performance progression of swimmers. Prospective longitudinal studies are required to improve our understanding of the multiple factors affecting career progression of swimmers, while further research is also needed to adapt the models presented here for other sports