Recreational running and training have been on the rise and therefore insights into training data can be very helpful for the runner. This study tries to provide certain insights such as predicted race time finish and fatigue levels for a certain distance. Leveraging a comprehensive dataset from Strava, a popular physical activity tracking application, the research employs machine learning techniques to predict race times and assess injury risks. Specifically, regression models to forecast race times, while classification algorithms determine runners' fatigue levels, categorized into low, medium, high, and very high. The analysis not only demonstrates the models' precision in estimating race times but also identifies training patterns that contribute to fatigue, a precursor to injury. The findings offer valuable insights for the running community by providing personalized race time predictions and understanding fatigue-levels, thereby enhancing training strategies and reducing injury rates among recreational runners.