Category Archives: Football

2021 NFL Big Data Bowl Submission

A Framework for Accessing Individual Defensive Performance in Coverage

1. Introduction

Current defensive coverage metrics are only calculated on plays where that defender’s assigned offensive player is the one targeted. This approach is flawed and incomplete. On any given play, there are around five available receivers, and the best coverage may prevent the quarterback from even throwing the ball to a given receiver. Further, these metrics identify coverage assignment simply by considering the responsible defender to be the nearest one at the time the ball arrives. Thus, this methodology only gives us coverage assignment for one player per play, and even that assignment is questionable; the closest defender at one particular moment isn’t necessarily the assigned defender.

The work presented here illustrates a framework we’ve developed to address these shortcomings. Our approach begins by identifying the responsibilities of each defender on a given play. If a player is in man (everywhere) coverage, he is assigned a specific offensive player whom he is to cover until the ball is thrown. On the other hand, if a player is in zone coverage, he is instead responsible for an area and any receivers who enter that area. Because the responsibilities of a player differ in each coverage type, the evaluation of performance should be different for each. Once we identify the type of coverage being played by each defender, we then consider two separate time points for evaluation. For one, we consider the time frame after the pass is thrown. This is the most straightforward but limits us to only the targeted receivers, as players tend to abandon their assignments and converge to the ball once it is thrown. To include all defenders in our analysis, we also consider the time frame between the snap and the pass. Figure 1 below illustrates our proposed framework…….

Read our full paper on Kagge here: https://www.kaggle.com/model284/a-defensive-player-coverage-evaluation-framework

2019 Wide Receiver Draft Model

Our Wide Receiver draft model incorporates player and team level college statistics and NFL Combine metrics to generate predicted probabilities of possible outcomes for each player’s NFL career. We focus on each player’s likelihood of making a Probowl, becoming a starter, becoming a role player (3rd – 5th wide receiver), or not making it in the NFL. Continue reading 2019 Wide Receiver Draft Model

2018 NFL Game Predictions

The following table contains our spread and total predictions for each NFL game. We will update this post with new predictions/matchups each week and keep track of how our models perform against the closing Vegas lines for each game. Our models are built using a combination of variables, including our Elo Ratings, Vegas spreads/totals, situational factors, team statistics (from a team’s last 7 games played), historical QB rating of the starting QBs, and more. Our predictions are generated as a composite (i.e., the average of multiple models). We are working on a more detailed write up on our modeling methodology that we will eventually link in here.
Continue reading 2018 NFL Game Predictions

Fantasy Points Above Expectation (FPAE)

Fantasy Points Above Expectation (FPAE) is a metric we created to better capture how much fantasy production each team is giving up. FPAE measures how many fantasy points a team gives up to each position relative to what their opponents had averaged going into each matchup. For example, if GB’s FPAE against QBs is 5.0, that means they have given up an average of 5 points more than expected to QBs. In this context, “expected” is referring to their opponents’ average fantasy points heading into the matchup. We give a detailed explanation of FPAE here. FPAE values for each position group are shown below and were calculated using each team’s last 7 games (PPR scoring).
Continue reading Fantasy Points Above Expectation (FPAE)

2018 Fantasy Football Rankings

FOOTBALL IS FINALLY BACK! And since there is such a shortage of Fantasy Football content out there, we thought we needed to give the people some material to prepare for their drafts. The following rankings are loosely based on model predictions for each position, which use historical player/team data to predict fantasy points for the coming season (i.e., separate models for QB, RB, WR, TE, K, DST). That said, there are plenty of factors that any model will have trouble capturing perfectly (injury status, team depth charts, suspensions, QB/Coach changes, etc.), and thus we have made some subjective adjustments to the rankings where necessary. For example, if a model weighs last year’s cumulative statistics too heavily, Odell Beckham Jr. is not going to come out very high since he only played 4 games. Our rankings can be found below, along with a short description of the model we used for each position. All rankings reflect PPR scoring.

To download our rankings, click here Continue reading 2018 Fantasy Football Rankings

My Model Monday: Modeling NFL Injuries

Injuries are inevitable in a game as physical as NFL football. Every season, numerous star players and important contributors are sidelined, leaving their fans and fantasy owners disappointed. Injuries appear to strike at random; an elite athlete can have his knee blown out in one cut like Dalvin Cook last season. However, is there a way to identify if certain players are more injury prone than others? I dug into the data to find out how well we can predict injuries among skill position players in the NFL.
Continue reading My Model Monday: Modeling NFL Injuries

My Model Monday: NFL Coaches Exceeding Win Totals

Every year in the barren wasteland of the NFL offseason, pre-season win totals are set by oddsmakers for each team, serving as a rough expectation for how many games a given team will win in the coming season (e.g., the New England Patriots currently have a win total of 11 for 2018). I took a look at how coaches have performed against those pre-season expectations, with the idea being that a “successful” season is one in which they exceeded their expected wins. As an example, the 2017 Los Angeles Rams had a pre-season win total of 6, but ended up winning 11 games, so that gives a +5 mark for 1st-year head coach Sean McVay. For coaches with multiple seasons under their belt, we can do that calculation for every season and roll-up the results to give a view of how a given coach has performed against expectations over a long period of time.
Continue reading My Model Monday: NFL Coaches Exceeding Win Totals

My Model Monday: Wide Receiver Draft Model

This year’s class of players entering the NFL draft lacks real star power at the Wide Receiver position. At first glance, there doesn’t appear to be a cant-miss player like a Julio Jones or Calvin Jonson. However, it’s important to remember that the most dominant receiver in the game today, Antonio Brown, was considered a below average prospect prior to the 2010 draft, and wasn’t picked until the 6th round, after 21 (!) wide receivers already went off the board. It is certainly possible there is another hidden gem hiding in this years crop of Wide Receivers.

Continue reading My Model Monday: Wide Receiver Draft Model

NFL Combine Tool

To use the Combine Tool, enter the name of a player and click go. The charts display that player’s scores in the 6 main NFL Combine drills: 40-yard dash, Broad Jump, Vertical Jump, Shuttle Run, 3 Cone Run, and Bench Press. The Yellow distribution curves show the distribution of other players’ scores from the selected player’s position. The numbers above each chart are the given player’s score for the drill and the percentile that that score falls in relative to other players at that position. The data comes from www.pro-football-reference.com and nflcombineresults.com. Our sample consists of all NFL Combine Invitees from 2005 – 2018. Look up some current and former players, or some players that will be coming off the board in this year’s draft!
Continue reading NFL Combine Tool

My Model Monday: NFL Team Complain-Ability

Ever since the first Olympic games were recorded in 760 B.C., sports have had an important social component in society. Whether its Baseball in Japan, Soccer in England, or Hockey in Canada, it’s almost impossible to live in a major sports market without being somehow affected by the local sports. While these celebrations of competition and the strive for excellence have brought neighbors together for centuries, there is another tradition that bonds us in a way that is much stronger: the tradition of complaining about our teams. It doesn’t matter if your team is a dominant force, or a bottom-feeding afterthought; we will always find something to complain about. But is it all warranted? Surely there are teams whose suffering has afforded their fans the right to complain more than all the others. In this week’s My Model Monday, I attempt to quantify the factors that lead to this allowance of complaining for each NFL team, and rank each fan base on their comparative level of what I like to call “complain-ability.”
Continue reading My Model Monday: NFL Team Complain-Ability