Coding for Basketball
By Rachel Marty
I’m Rachel, a PhD student at UC San Diego studying Bioinformatics and a Data Scientist for Noah Basketball.
I never anticipated becoming a computer scientist. I started my degree at UCSD with aspirations of becoming a doctor. I majored in biology but spent most of my time on the basketball court as a student-athlete. After some coaxing from my wiser-than-I-was-willing-to-admit father, I took a computer science class. It was VERY challenging and I often questioned if I was smart enough to continue. However, I could not deny the unique satisfaction I felt when I solved a problem. Not long after my first computer science class, I ended up switching my major to Computer Science — Bioinformatics.
Could I help millions of patients instead of just a couple by designing scalable treatments with the help of computer science?
I quickly became enthralled with the idea of studying genomes. Could I help millions of patients instead of just a couple by designing scalable treatments with the help of computer science? I have spent the past four years in a PhD program studying the role of genomic immune variation on cancer progression and discovering new biology through coding.
Outside of my PhD program, I am actively involved with Noah Basketball, which is a small startup that develops a basketball training tool that watches players shoot and measures precisely how every shot makes or misses. Three years ago, they needed a data scientist to drive the research on the massive amount of data they were collecting. Never losing sight of my love for basketball, I jumped at the chance to apply my bioinformatics-trained data science skills to a basketball analytics problem. Who knew that the same skill set could cross such different domains?
My research has resulted in several invitations to present research to NBA coaches. I guess you can never know where your coding will take you!
Noah Basketball has collected millions of shots taken by NBA, college and high school players. I use python to find interesting trends in the data and apply libraries like scikit-learn to train machine learning models. I recently presented a paper at the MIT Sloan Sports Analytics Conference about better predicting shooting ability based on this immense amount of 3D data. In a crazy turn of events, my research has resulted in several invitations to present research to NBA coaches. I guess you can never know where your coding will take you!
What tools and languages do I use?
- Python
- R
- Jupyter Notebooks
Rachel loves to bike, swim in the ocean and drink way too much tea.
