Wholistic Approach at Analyzing NFL Players

Author: Lawson Porter   Major: Industrial Engineering

Lawson Porter

My name is Lawson Porter and I am finishing up my senior year in May here at the University of Arkansas. I am pursuing a degree in Industrial Engineering within the Bell College of Engineering as well as a minor in Finance within the Walton College of Business. I have been in the processing of completing my thesis since the Spring of 2020 and have been guided by my research mentor Dr. Manuel Rossetti, also of the Industrial Engineering department, who has provided valuable insight and direction for me throughout the process of my research. Throughout my research experience, I have been able to take interesting and unique looks at something that I was already interested in, that being NFL statistics, and it has made my research journey even more rewarding so far.

The main goal of my research is to analyze different position groups in the NFL using a methodology called MODA, or Multiple Objective Decision Analysis. MODA is actually a methodology developed by another University of Arkansas Industrial Engineering professor, Dr. Greg Parnell, and has many applications in different industries. For the case of my research, MODA takes in a bunch of different statistics or value metrics for a position group and evaluates players within this certain position group based on how they perform in the different statistics or objectives. The MODA methodology then scales these scores for each player and weights are developed for each objective. The weights can be established in a number of different ways for decision-making models like MODA, but in the case of this project, we will be using the input of people knowledgeable in the field of football to establish these weights in the model. Once a group of players is evaluated within the model using the proper statistics and weights, their “value” can be established and their overall performance is compared with the price associated with a certain player’s salary in order to demonstrate players that may be over or undervalued for a general manager making these hiring decisions.

This topic of Multiple Objective Decision Analysis within an NFL context was suggested to me by my faculty mentor Dr. Rossetti. In the Industrial Engineering curriculum, we have an honors class that provides those of us interested in completing a thesis with the opportunity to listen to various faculty member’s research to establish and identify a mentor that we would be interested in working with. Dr. Rossetti’s research generally revolves around simulation models for various industries, but in talking to him initially he mentioned the idea of decision modeling in the NFL and this idea really sparked my interest. Choosing a topic in an industry that was interesting to me was very important, and as a big NFL fan, it has allowed me to be able to see more direction for my research having some interest/knowledge in the subject already. One of the biggest aspects of research to me is the ability to explore new directions and choose for yourself different ideas to try out. Research is a lot of trial and error, and I think it has made me both a more independent and creative thinker in the process, something that has carried over to my internships as well. I have come across many different challenges throughout my research experience so far, everything from struggling to find the exact data I want to have to learn topics outside of my normal curriculum, but with the aid of Dr. Rossetti I have been able to overcome many of these initial challenges. Part of my research also revolves around some data analysis techniques using a programming language called R, a skill that I was not super familiar with but with the help of Dr. Liu, another Industrial Engineering professor, I was able to get a much stronger grasp of this and apply it to my own research. I will continue to develop the methodology for this research over the course of this spring semester and will be polishing up my thesis over the next few months!