A Statistical Approach to the Pro Football Hall of Fame

The Pro Football Hall of Fame, located in Canton, Ohio, is the most selective museum of its kind for any North American Sport. It houses the stories of the greatest North American football players of all time. The goal of my project is to use predictive statistics and modeling to create a way to predict whether or not a player will one day be considered for the Hall of Fame. While statistics do not tell the whole story of a player, no one has a chance of making the Hall of Fame without being significantly better than his peers. This summer I want to measure how much “better” than average a player must be in order to be considered for the Hall, as well as which statistics can most accurately predict a player’s success. I want to apply my findings both historically, to learn about stories of players who had the “numbers,” but were overlooked, and predictively, to determine which players are most likely to have Hall of Fame statistics at the end of his career.

This project was inspired by the time I spent in my Business Statistics class during my first semester at the College, and there is a business angle to this project as well. Many companies choose to endorse professional athletes to help market themselves and improve their image. Endorsing a player early on that goes on to have a Hall of Fame career can be a real difference-maker for any company, and the capital investment is enormous. Endorsing a young rising star means that every fan will see your name next to a star player for a family, but paying a player who goes bust and never makes it is a huge waste of time and money. I want to create a model to find the players that will have the biggest impact on the field and in the commercial studio.

Comments

  1. This is a interesting topic. Are you planning on factoring in how off the field issues can affect a player’s chance of being inducted. Many voters thought Cris Carter, should not be in the Hall of Fame because of the many off the field issues he had, despite retiring as the second most prolific receiver in NFL history. He was not voted in between 2008 and 2012, though he finally received enough votes in 2013.

    Do you think that the use of media representatives as voters increases bias and makes it harder to find a statistical model to predict induction?

  2. mattk26 says:

    I am so glad you brought up Cris Carter– he is one of my favorite players ever. I am planning on taking a historical perspective in my second post and looking at exactly what you are alluding to. I discovered that career length is probably the most important statistical predictor of whether or not a player is a Hall of Famer, and Cris Carter’s drug problems, as you know, almost ended his career after just 3 seasons. Longevity relates to much more than injuries, although that is important as well.

    That is an interesting point about the media selecting the players for the Hall of Fame. I think that it adds a few wrinkles, but I am hoping to find exactly which qualitative factors, if you will, such as postseason success, can predict induction into Canton. But some are probably impossible to quantify, like personality. I know some voters probably keep another of my favorite players, Ken Stabler, out of Canton because of some of his off the field activities.