Long distance running insights - are you heading towards a new goal or a new injury?

Here is my first stab at doing something interesting with running and software.

This simple Streamlit app allows one to upload their Garmin extracted data and see some high level insights, four key things to look at:

  1. Cadence should strive towards 180 steps per minute.
  2. Stride length should strive towards cadence, most of us wouldn’t get anywhere near but anything above one meter is a good achievement.
  3. Vo2Max should strive towards 60 and higher.
  4. Resting heart rate should be lower than Vo2Max

In the attached sample data (same as print screen) you can see me running barefoot now VS kinda barefoot (Saucony Hattori) 3 years ago.

You can see that :

  1. Barefoot cadence got better.
  2. Barefoot stride length got safer - the drops in the stride length were not laziness but injuries.

Code is here : GitHub - Bihshtein/garmin: GARMIN insights for long distance runners

Please do reach out if you find this interesting!
Originally I wanted to use the Garmin data for useful predictions, but this is not so easy to formulate.
My intuition says that this would go around the stride length and predicting a manageable increase rate over time.

Me? I’ll choose ‘new injury’ for any value of distance > 0 :rofl:

One suggestion I would make is to take this content and put it into the app itself:

When I clicked through, it wasn’t immediately obvious to me what I should do with the app (because I only skimmed your post). If that content were in the app, or even better you used Python to check for those scenarios, then the tool would immediately tell you if an injury is in the near future.


1 Like

Thanks! I will definitely add some features to highlight the keys take aways for a given input data.
Probably will highlight some preset scenarios for now, an injury prediction as less trivial with heuristic and not ML.

Regarding yourself, i think that a good start could be to shrink your stride below 90cm and to focus on proper landing for every step, hopefully this should bring your distance significantly above 0

New version is deployed here with some trivial insights as per @randyzwitch suggestions

1 Like