One click to Pitch Control - Running Spearmans Pitch Control model in Google Colab.
I show how you can run William Spearmans Pitch Control model implemented by Laurie Shaw directly in Google Colab. No instalation needed, it just works.
The content the guys from Friends Of Tracking Data (#FoT) are putting out there is growing and recently Laurie Shaws shared his implementation of William Spearman’spitch control model from 2018. Also recently Devin Pleuler shared his Analytics Handbook which consists of a bunch of jupyter notebooks that you can run directly in Google Colab. This means you do not need to install anything, using the code is just one click away. I think that every repository should be build in this way as you won’t have any overhead if you just want to play around.
I thought it would be really cool to also be able to just run Spearman’s Pitch Control model in this way. Therefore, I’ve restructured Lauries repository to allow for direct Google Colab integration.
| Notebook | Colab |
|---|---|
| Basic Plotting of Event and Tracking Data |  |
| Advanced Plotting and Summary Statistics | |
| Pitch Control | |
You can get the code on Github: https://github.com/seidlr/LaurieOnTracking
Just click on the badge next to Pitch Control and you are good to go. Then:
- Run each cell in the notebook and ignore the warning about the unknown environment.
- The data is downloaded directly from Metricas repository.
- Give it some time when it says “Reading team home” in a cell.
If you are interested in the details what I needed to change to be able to run the code in Google Colab, you can find all code in my fork of Lauries repository on Github. Here are the new benefits.
- A clean structure, where notebooks and modules are separated.
- The Metrica sample data is directly read from their github repository.
- You can run the notebooks directly in Google Colab.
- It is pip-installable. Just run
pip install git+https://github.com/seidlr/LaurieOnTracking.gitAnd you are able to use the implementation anywhere in your projects. I.e. you can read in the metrica data of sample match 2 in a dataframe with this code snippet.
import friendsoftracking.metrica.IO as mio game_id = 2 # let's look at sample match 2 # read in the event data events = mio.read_event_data(game_id) events.head()