This repository is a space where I provide a peek into the machine-driven processes of performance analysis in professional football. Check the files above for any post that I have not updated as a description in here.
I hope this repo provides a good insight into programming & automation in the space of sports performance analysis in India.
Let's go!
In modern day performance analysis, there is massive emphasis on data analysis because performance analysis practitioners are looking to improvise and take the next step. Everyone has access to video, hence the focus is on working smartly to extract the most from video. This is where the edges are. Every performance analysis practitioner strives to create a workflow that makes gaining edges consistent. This can be achieved by setting up initial filters to video analysis tasks with the help of data analysis. Not every detail is important, but nothing is unimportant. The ability to pick up the relevant bits of information is at the core of analysis and the main driver of good performance and success on the pitch.
As Rob Carroll puts it in a blog titled Automating Match Reports, "Nothing pains me more than to see people constantly doing manual tasks when there simply is a better way," automation is no longer a choice for performance analysts, it is mandatory. Otherwise the risk of getting left behind or being slowed down is very big. This is very pertinent in teams that don't have massive resources to invest in big analysis departments.