About This Project

This project was inspired by a Nationals fan exploring her own FitBit a week or two after the World Series, then noticing similar patterns on her friends' FitBits.

This project was built with roughly 1.5GB of FitBit heart rate and sleep logs obtained in json and .csv format from 35 fan volunteers. The data was cleaned and normalized using Python and Pandas in Jupyter notebooks; the visualizations were built using Bokeh.

Noteworthy game events were culled from Baseball Reference's game log then cross-referenced with Twitter, using the earliest event-related tweet to establish the event's timestamp. The two datasets for the game log-style charts were selected based on completeness; many fan logs for each day contain large blocks of missing data, making timeline visualization difficult. The specific games were chosen for their expected high emotional value among Nats fans.