Revitalized Data Science Club Eyes Insightful Statistics for NJIT Athletics
Sports provide an everyman's opportunity to explain statistical analysis, and the rejuvenated NJIT Data Science Club is all-in.
The club already has more than a dozen members and is led by action-hungry graduate students Jake Byford and Parth Patel, president and vice president respectively, who both stepped up when most members of the previous Machine Learning Club graduated last year.
Byford and Patel will both earn M.S. degrees in data science. Byford is starting by taking math courses, while Patel will graduate in December. Their plans for the club are what you'd expect — collaboration with other computing-oriented clubs, community data projects, guest speakers, social events, study sessions, perhaps their own data-centric hackathon — but they're happiest using data science to gain insights combined with muscle and sweat, not just algorithms and debugging.
Patel, a basketball fan before he even came to America last year, said he became interested in sports statistics as a cricket fan and amateur player in India. Upon arriving at NJIT, he offered to volunteer for women's basketball Head Coach Mike Lane, who welcomed the help. The team wasn't doing much with analytics but already understood the value of it because of former player Ivana Seric, who earned a Ph.D. in math here and became a data researcher for the NBA's Philadelphia 76ers. Seric is now a senior researcher at basketball development league Overtime Elite.
"I get the raw data and I crunch numbers. I give some percentages. If there is a game tomorrow, I give the coach a pre-game report," Patel explained. He met Seric, who taught him how to obtain latent data.
It helps that the players themselves are already comfortable with aspects of data science, which is part of nearly every major at a school like NJIT
Patel quipped that he's in no danger of becoming a good hoops player — "I had an idea to practice with the girls but they'd kick my butt … Stick to numbers," — but the five or six hours per week that he spends on statistics like assists, blocks, field goals, free-throw percentages and rebounds definitely helped the team, Coach Lane agreed. The evidence is in their results, as the team finished 15-15 this year and reached the semifinals of the America East tournament after not playing .500 ball since the 2012-2013 season. They won a close Nov. 18 game against Long Island University in overtime, which Lane said could certainly be attributed to Patel's efforts if an extra basket was made or blocked based on data analysis.
Lane said he would like Patel or future student-analysts to evolve from looking at team data to looking at individual player data. It helps that the players themselves are already comfortable with aspects of data science, which is part of nearly every major at a school like NJIT, not just the Department of Data Science in its Ying Wu College of Computing.
Byford hopes to follow Patel's path by working with the men's baseball team, which reached Cinderella status in the NCAA regional tournament last year. He is just as interested in projects beyond sports, observing that data science touches fields such as autonomous vehicles, COVID research, gaming, social media, robotics, virtual assistants and voice recognition.
"I think it's a good way to involve students in data science and expose them to a practical application. The movie Moneyball comes to my mind," club advisor and Associate Professor Usman Roshan added. "Basic statistics gives you player performance but advanced methods can forecast a player's performance across time. Even more advanced and we get into methods to help players with their training."