Ever since the eponymous “Moneyball,” sports data has become the secret sauce of major sports teams. Denison students teamed up to analyze football and basketball data in a recent Sports Analytics conference at Ohio State University.
The conference provided an outlet for university students of all levels to exhibit their own original research, learn marketable data skills, and learn from industry leaders.
Becca Skolnick, Tamara Dzolic, and Rediet Habtegebriel presented “Finding the Next Defensive Player.”
Their abstract: Predicting the efficacy and quality of a defensive player in the NFL from their college and combine performance is something the draft and scouts aim to do on a regular basis. In our study, we utilize the combine and statistics from each player’s senior season of college, and the NFL 2019 season in an attempt to expose a relationship between an individual player’s growth and the team’s success. Success in our study is defined by the team’s ranking in the top or bottom three teams of the NFL as of Super Bowl LIV. Using a principal component analysis coupled with a linear regression model, we attempt to find a relationship that links the explanatory variables: percentage of solo tackles, quarterback hits, 40 yard dash, 20m shuttle, bench press, broad and vertical jumps, to the success of the NFL team associated with each player. The ultimate goal of this study is to identify a relationship between our response and predicting variables so as to predict the ability of a defensive player to contribute to a team’s success on gameday.
Hong Jui Shen and WenYi Shi presented “Linear regression for basketball team scores analysis.”
Their abstract:
At present, there is a lot of prediction for team scores in sports. Therefore, we selected the data in the basketball reference website. Mainly focused on the overall game data of the Los Angeles Lakers, Milwaukee Bucks, and Houston Rockets from 2019 to 2020. And through the average performance of the entire team’s players on the field from 2019 to 2020 to predict the next game’s score. We will use Principle Component Analysis to make the composition of the main factors of a game. Through these factors, use linear regression to predict the team’s next game score.
