Parkinson's Disease is a devastating illness with no currently available cure. As the population ages, the disease becomes more common with a large financial cost to society. A rapid and accurate diagnosis, as well as practical monitoring methods are essential for managing the disease as best as possible. This paper discusses two approaches to discriminating movement data between healthy controls or Parkinson's Disease patients. One is a standard statistical analysis, influenced by prior work into classifying patients. The other is a programmatic expression evolved using genetic programming, which is trained to observe differences in specific motion segments, rather than using arbitrary windows of a full data series. The performance of the statistical analysis method is relatively high, but it still cannot discriminate as well as the evolved classifier. This study compares favourably to previous work, highlighting the usefulness of analysing a successful classifier to influence design decisions for future work. Examination of the evolved programmatic expressions that had high discriminatory ability provided useful insight into how Parkinson's Disease patients and healthy subjects have differing movement characteristics. This could be used to inform future research into the physiology of repetitive motions in Parkinson's Disease patients.