A first look into KOE

As data analysis and NFL data becomes more and more available, information and analysis regarding flashy positions and plays constantly teach us new things about the game of football. While we have a pretty good idea of how the relationship between air yards and EPA works for instance, few analytical efforts have paid attention to the optimal kicking distance for a field goal. Kicking a field goal in the NFL is a very specific and straight-forward task, and KOE was built specifically to evaluate it as fairly as possible.

KOE Kicks Over Expected, or more properly and less catchy Field Goals Over Expected represents the points “added” by a kicker over those of an average kicker of the same sample, which in this context is a professional NFL Kicker. A better KOE will be better for a kicker in almost any context, as it will better reflect his contribution to the outcome of a kick. When a kicker is put in a position where almost anyone would fail, he will be more severly penalized, and the opposite will happen when he misses an easy kick.

A first attempt at drawing conclusions is trying to figure out if KOE will help separate kickers according to their value when kicking field goals. Bayesian bootstrapping was adjusted to estimate the average KOE of each kicker using the R package bayesboot. This approach in a way reveals each kicker’s true KOE, and as we gather more and more information about him, it should also become more accurate. You can see how the quantiles contract for veteran kickers.

It is always good when a model confirms the strongest priors, and offers valuable insight at the same time. Justin Tucker is in a league of his own regarding NFL Field Goals. He is significantly better than his strongest contenders, which coincidently are also pretty good. Josh Lambo is a name that also looks pretty good regarding KOE, however his floor appears to be the lowest of his class. Looking into his case, I learned he had a first strech playing for the Chargers where he wasn’t as good as he has been, so it makes sense to have some uncertainty to go with his high KOE.

Roberto Aguayo is also an interesting case, as he the worst KOE registered of all kickers considered since 2015. As a college player Aguayo broke accuracy records, however his career took a differnt path as a young professional kicker. It is perhaps because of his pedigree that he was allowed to continue kicking in the NFL far longer than kickers with his numbers. Aguayo’s case is even more surprising considering he kicked in kicker-friendly weathers, but maybe he could have chased a similar path to Vinatieri or Koo, who moved to kicking for a team with a dome after facing rough streches.

A kicker that will surely be in the hall of fame, Adam Vinatieri, is another interesting player. Even through all his games, Adam Vinatieri still stands without a negative KOE. He lived through different streaks, being top rated in KOE between the years 2002-2006 and 2014-2017. His numbers were very good, but rather inconsistent in New England, and then took off to a new level on his last years at the Colts. Vinatieri is also one of those data points that breaks scales for his longevity alone. It also suprising that he managed to retire with a positive KOE after 24 seasons (he was drafted in 1996 so most numbers you read on him could be incomplete).