RUNNING WITH DATA
Predicting marathon times
Regis University | Masters in Data Science | Practicum II
ABOUT
Can I use my past running data to predict my marathon time? Running is a passion of mine, and like most marathon runners, one of my goals is to qualify for the Boston Marathon. As of 2020, the Boston Marathon standard qualifying time for women between 18-32 is 3 hours and 30 minutes and the qualification time accepted is 3 hours and 28 minutes and 21 seconds. This is not an easy goal to achieve, and as a person who loves data, I’m intrigued to see if my running data can help with my performance. My goal for this research is to first see if I can predict my marathon time based on my training data from the 2019 Chicago Marathon. Once I accumulate those results, I will analyze what influenced my faster run times and what made me slow down, in hopes to modify my training plan for optimal time.
THE DATA
For this research I am using my own data gathered from my Garmin Forerunner watch along with Strava. The data being used is collected from May 2019 – October 2019 while I was training for the Chicago marathon. Additionally I utilized Pete Riegel’s marathon prediction formula to determine my marathon prediction time. Pete was both a marathon runner and an engineer who developed this formula:
T2 = T1 x (D2/D1) 1.06
T1 = given time, D1 = given distance, d2 = distance to predict a time for, T2 = calculated time for D2
GITHUB
Github repository of data can be found here