Hyrox X Data Science
To best improve your Hyrox, you should train the exercises that will have the greatest impact on your overall score. This project aims to be the tool to help athletes identify exactly those exercises.
Big Shoutout to www.hyresults.com for inspiration.
Data collection
Hyrox results are posted on the official Hyrox website, which we scraped with python (BeautifulSoup4). Scraper code can be found here #TODO githublink.
Seasonal changes
The website interface changed between season 1-4 and season 5-8, and the number of timestamps collected per race differed (16 vs 30). Thus, the logic of the scrapers for season 1-4 and season 5-8 are quite similar, but not interchangeable.
Data clean-up
As our dataset contained ±600K entries, the scraped data was combined and stored into a duckdb database to ensure performant analysis. As we are mostly intrested in the “hyrox singles” scores, all doubles, duplicate values and invalid entries were omitted.
