Analysis, visualizations and KMeans clustering using data from submission grappling matches of ADCC events from 1998 to 2022
Submission grappling (aka no-gi brazilian jiu jitsu) is a combat sport rising in popularity all over the world. Despite that, there's still a lack of publicly available data and data based analysis and studies. A training partner managed the daunting task of web scraping BJJ Heroes website for data on ADCC matches from 1998 to 2022 and through extensive cleaning and treating of the data, some interesting stats and trends could be identified and displayed graphically.
Deepest thanks to https://github.com/bjagrelli/adcc_dataset_analytics for the web scraping step and making this possible.
This is an interactive Python notebook, which is also available on Kaggle where it can be directly run using any web browser.
Code cells are collapsed by default there, but can be expanded with a single click if you're interested in the code.
If you just want to see the results from my analysis, the notebook is well organized for casual viewers.
Since the plots there are interactive, you can mouse over any portion of them to reveal details about it, zoom in or out and redefine the scope of what's being displayed.
Link to Kaggle notebook: https://www.kaggle.com/code/albucathecoder/adcc-fighters-eda-and-clustering-kmeans
If there's any specific visualization or analysis you want to see done but don't know how, feel free to leave a comment there.
If you want to download the data to analyze it yourself, it's also available for free at the Kaggle link under 'Input'.
Thanks for the support!