Introduction

The sport of ultra-marathon running is one of the fastest evolving sports on the globe. Races of 100 miles or more are becoming normalized, prize money is skyrocketing, and sponsorship dollars are flooding in. Each year more and more people are lacing up a pair of trail running shoes. An ultra-marathon is defined as any foot race longer than a traditional marathon: 26.2 miles. Some of the most common ultra-marathon distances are 50 kilometers, 50miles, 100 kilometers, and 100 miles, but races in excess of 500 miles exist. While the label of ultra-running relates only to distance, the races are primarily held off the road on hilly trails. A global sport, regions like France and Colorado hold some of the largest and longest standing races. However, new races are being founded every year in the United States and around the world. While the sport’s explosion in popularity is relatively recent, ultra-athletes have been existing on the fringes of society and competing in ultra-marathons since the 1800s. No longer existing on the fringes in 2025 companies like Nike and Adidas are pouring billions into the sport.

As of 2022 the global trail running shoe market was valued at $7.3 billion. This value is expected to grow to $13 billion by 2032. Each shoe company is scramble to gain control of the market. In-depth information and analysis of what was once a niche activity is now becoming necessary for fortune 500 companies. By analyzing decades of results from thousands of races insight can be gained into the secrets of ultrarunning. These insights may lead to the creation of the shoe that finally gives someone control of the $10 billion dollar industry. Perhaps analysis enables the signing of an undercover athlete who is primed and ready to become the next big thing. In addition to product creation, marketing, and sponsorship insight, analysis of ultrarunning data impacts an even larger industry: gambling. As the sport continues to grow, and fans increase into the hundreds of millions large gambling platforms have joined the scene. Fans can now place bets on their favorite athletes in the biggest of races and gambling organizations require every scrap of data they can manage.

To glean valuable insights into the world of ultra-running this project will apply a range of data analysis techniques to multiple relevant datasets. In conjunction with every current race record available data has been gathered through multiple APIs. This additional information includes weather, elevation, and location data. By integrating multiple data streams, a clearer picture of ultra-running trends can be developed. Through the application of machine learning techniques, including unsupervised methods such as Principal Component Analysis (PCA), clustering techniques, Association Rule Mining, valuable insights into the history and the future of ultra-running performances will be derived. The analysis covered in this project will uncover trends and patterns within the sport. More specifically, clustering techniques such as K-means, Hierarchical, and DBSCAN can categorize athletes based on performance similarities, revealing trends among elite runners and amateur participants. Furthermore, Association Rule Mining discovers hidden relationships between race conditions, athlete characteristics, and finishing times. Through this process, valuable information and prediction capabilities will be available to inform corporations, race organizations, athletes, and fans alike. This analysis will return exciting results from a sport full of exciting results!