By: Edward Egros

Advanced Modeling Techniques for Forecasting College Football Games

Pasted Graphic
This July, Charles South, an instructor at UT Southwestern, and I, gave a presentation at the R Users Group at the University of Dallas. In this talk, Charles and I discuss how to predict college football games. Using data from Clemson University and 247Sports, we used advanced modeling techniques to see what best predicts an out-of-sample set of games.

To see the powerpoint of our talk,
click here.

Analyzing and Aiding Sports Analytics in Television

Sports analytics serve a variety of purposes, from quantitative measurements for how successful an athlete or team is, to predicting the outcome of a sporting event. Much of the development of analytics comes from athletes and teams implementing these tools for a competitive advantage.

Unfortunately for the media, there are several outside forces preventing sports analytics from being used more frequently on television. Still, some sports journalists have found ways to use and enhance these tools. In this paper, three nationally televised sports shows are analyzed for how they discuss analytics for their respective sport.

Using text mining techniques, it is possible to see the frequency of sports analytics and the complexity of the tools used in broadcasting, as well as—ultimately its primary purpose of the paper—how sports analytics are used to supplement a preexisting argument. This paper concludes by looking at other ways sports analytics can be used to enhance journalism.

To read this paper, click here.