Project Video
Watch the full project presentation and demo on YouTube.
Interactive Demo
Select tracks to change the BPM. Watch how the runner's cadence and speed respond to the beat.
About This Project
This data science project investigates whether the tempo of music played during running workouts has a measurable effect on athletic performance. Using the Spotify Web API, we extract audio features -- tempo (BPM), energy, danceability, and valence -- from thousands of running playlists and correlate them with running metrics including cadence, pace, and heart rate data collected from fitness trackers.
Our analysis reveals a statistically significant positive correlation between music BPM and running cadence, particularly in the 150-180 BPM range where entrainment (the tendency for movement to synchronize with an external rhythm) is strongest. Machine learning models trained on this data can predict optimal playlist orderings that maximize performance improvements across different runner profiles and training intensities.
Audio Feature Extraction
Leveraging Spotify's audio analysis API to extract tempo, energy, danceability, and spectral features from track catalogs.
Cadence Analysis
Statistical modeling of the relationship between musical tempo and running step frequency across diverse runner populations.
Correlation Modeling
Regression analysis and machine learning to quantify BPM-performance relationships and identify optimal tempo zones.
Playlist Optimization
Algorithmic playlist generation that sequences tracks to match target cadence profiles for interval training and tempo runs.