NFL Data Analysis¶

Names¶

Alexander Scheibe, Addison Atkin, Seth Jordan, Henry Severson, Melanie Yadgar

Introduction¶

For our project we will be designing models to make predictions based on historical NFL game-by-game data. Our data consists of a wide variety of features ranging from football stats like passing yards, touchdowns, win rates, as well as other factors such as weather, player data and stadium data. For our analysis, we will not just look at game scores, but also these other various factors to see how and if it affects game outcomes. By using machine learning, we aim to transform raw data into meaningful predictions and uncover patterns that may not be immediately obvious. Ultimately, we hope that with our model we will be able to predict which team will win this upcoming Super Bowl, as well as learn more about Vegas betting odds accuracies in the world of football betting.

Dataset¶

https://uwprod-my.sharepoint.com/:x:/g/personal/avscheibe_wisc_edu/ERafpnPpvd1Omsku-j6QxaEBEfWoQehoGeiriStTU6dPOA?e=4oMMJi

Questions¶

  1. Player Performance Prediction: Can we predict a player's performance (e.g., passing yards, touchdowns) based on their historical data and game conditions?
  2. Betting Data Analysis: How accurately does betting data predict a team’s success on a given week? Does betting accuracy increase as the season progresses?
  3. Team Performance Analysis: How do environmental factors (e.g., weather, stadium) impact team performance (e.g. win/loss rate)?

Variables¶

Independent Variables:¶

  1. Player Performance:

    • Completions: Number of successfully completed passes by a player.
    • Attempts: Number of attempted passes, regardless of whether they were completed.
    • Passing Yards: Total yards gained through passing plays.
    • Receiving Yards: Yards gained by a player who catches a pass.
    • Rushing Yards: Yards gained by running with the football.
    • Touchdowns: Number of times a player scores a touchdown.

  2. Game Conditions:

    • Weather (Temperature, Wind MPH, Humidity): Respective atmospheric conditions during the game.
    • Weather Detail: Indicates whether stadium is indoor/outdoor
    • Stadium: Stadium name.

  3. Betting Data:

    • Over/Under Line: A betting line predicting the total points scored by both teams combined.
    • Team Favorite ID: The team expected to win a game as indicated by betting odds.
    • Spread Favorite: Surrendered point spread for favorite team.

Dependent Variables:¶

  1. Player Performance (as Dependent Variables):

    • Passing Yards: Total yards gained through passing plays.
    • Touchdowns: Number of times a player scores a touchdown.

  2. Team Record:

    • Wins: Total number of games won by a team.
    • Home Score: Points scored by the home team in a game.
    • Away Score: Points scored by the away team in a game.

  3. Betting Data:

    • Betting Accuracy: A measure of how accurately betting odds have predicted game outcomes in the past.

Methods¶

Player Performance Prediction (Q1):

  • Linear regression for predicting continuous performance metrics such as yards gained.
  • Soft-margin SVM for classifying players into distinct performance categories. May also be helpful for high dimensionality data which we anticipate.
  • Descision trees for identifying key factors influencing player performance.
  • Feature engineering for binning numerical performance data into categories.

Betting Data Analysis (Q2):

  • Decision tree classification using favored team by brinigng in the multiple betting data points in order to figure out what's most useful.
  • Logistic regression using the betting line for predicting the probability of a team's success.

Team Performance Analysis (Q3):

  • Linear regression for analyzing how environmental factors quantitatively impact team performance (win/loss rate).
  • Decision trees to understand the impact of categorical environmental factors (like type of stadium) on team performance.
  • KNN to predict team performance based on similarity to historical games in similar environmental conditions.
In [ ]:
import pandas as pd
pd.set_option('display.max_colwidth', None)
pd.set_option('display.max_columns', None)

nfl_data = pd.read_csv('nfl_data.csv')

nfl_data.head(10)
Out[ ]:
id name position team week season season_type completions attempts passing_yards passing_tds interceptions sacks sack_yards sack_fumbles sack_fumbles_lost passing_air_yards passing_yards_after_catch passing_first_downs passing_2pt_conversions carries rushing_yards rushing_tds rushing_fumbles rushing_fumbles_lost rushing_first_downs rushing_2pt_conversions receptions targets receiving_yards receiving_tds receiving_fumbles receiving_fumbles_lost receiving_air_yards receiving_yards_after_catch receiving_first_downs receiving_2pt_conversions target_share air_yards_share fantasy_points fantasy_points_ppr total_yards ypa ypc ypr touches count comp_percentage pass_td_percentage int_percentage rush_td_percentage rec_td_percentage total_tds td_percentage passer_rating rookie_season round overall ht wt forty vertical offense_snaps teams_offense_snaps snap_pct years_played Unnamed: 0 schedule_date schedule_season schedule_week schedule_playoff team_home score_home score_away team_away team_favorite_id spread_favorite over_under_line stadium stadium_neutral weather_temperature weather_wind_mph weather_humidity weather_detail team_home_abbreviation team_away_abbreviation
0 1 A.J. Brown WR TEN 1 2019 REG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 4 100 0 0 0 54 56 2 0 0.166667 0.372414 10.0 13.0 100 0.0 0.0 33.33 3 1 0.0 0.0 0.0 0.0 0.000 0 0.000 33.87 2019.0 2.0 51.0 Jun-00 226.0 4.49 36.5 25.0 60.0 0.42 1.0 12414 9/8/19 2019 1 False Cleveland Browns 13 43 Tennessee Titans CLE -5.5 44.0 FirstEnergy Stadium False NaN NaN NaN NaN CLE TEN
1 1 A.J. Brown WR TEN 2 2019 REG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 5 25 0 0 0 25 8 1 0 0.192308 0.168919 2.5 5.5 25 0.0 0.0 8.33 3 1 0.0 0.0 0.0 0.0 0.000 0 0.000 33.87 2019.0 2.0 51.0 Jun-00 226.0 4.49 36.5 27.0 59.0 0.46 1.0 12440 9/15/19 2019 2 False Tennessee Titans 17 19 Indianapolis Colts TEN -3.5 43.5 Nissan Stadium False NaN NaN NaN NaN TEN IND
2 1 A.J. Brown WR TEN 3 2019 REG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 5 4 0 0 0 69 -2 0 0 0.131579 0.163507 0.4 1.4 4 0.0 0.0 4.00 1 1 0.0 0.0 0.0 0.0 0.000 0 0.000 33.87 2019.0 2.0 51.0 Jun-00 226.0 4.49 36.5 39.0 80.0 0.49 1.0 12443 9/19/19 2019 3 False Jacksonville Jaguars 20 7 Tennessee Titans TEN -2.0 38.0 TIAA Bank Field False NaN NaN NaN NaN JAX TEN
3 1 A.J. Brown WR TEN 4 2019 REG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 3 94 2 0 0 43 51 3 0 0.136364 0.330769 21.4 24.4 94 0.0 0.0 31.33 3 1 0.0 0.0 0.0 0.0 0.667 2 0.667 33.87 2019.0 2.0 51.0 Jun-00 226.0 4.49 36.5 26.0 62.0 0.42 1.0 12461 9/29/19 2019 4 False Atlanta Falcons 10 24 Tennessee Titans ATL -3.5 46.5 Mercedes-Benz Stadium False 72.0 0.0 NaN indoor ATL TEN
4 1 A.J. Brown WR TEN 5 2019 REG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 2 27 0 0 0 25 2 1 0 0.095238 0.242718 2.7 4.7 27 0.0 0.0 13.50 2 1 0.0 0.0 0.0 0.0 0.000 0 0.000 33.87 2019.0 2.0 51.0 Jun-00 226.0 4.49 36.5 37.0 58.0 0.64 1.0 12486 10/6/19 2019 5 False Tennessee Titans 7 14 Buffalo Bills TEN -3.0 39.5 Nissan Stadium False NaN NaN NaN NaN TEN BUF
5 1 A.J. Brown WR TEN 6 2019 REG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 4 23 0 1 0 39 5 2 0 0.125000 0.138298 2.3 4.3 23 0.0 0.0 11.50 2 1 0.0 0.0 0.0 0.0 0.000 0 0.000 33.87 2019.0 2.0 51.0 Jun-00 226.0 4.49 36.5 38.0 66.0 0.58 1.0 12493 10/13/19 2019 6 False Denver Broncos 16 0 Tennessee Titans DEN -1.5 41.0 Sports Authority Field at Mile High False NaN NaN NaN NaN DEN TEN
6 1 A.J. Brown WR TEN 7 2019 REG 0 0 0 0 0 0 0 0 0 0 0 0 0 1 -2 0 0 0 0 0 6 8 64 0 0 0 56 17 4 0 0.275862 0.284264 6.2 12.2 62 0.0 -2.0 10.67 7 1 0.0 0.0 0.0 0.0 0.000 0 0.000 33.87 2019.0 2.0 51.0 Jun-00 226.0 4.49 36.5 39.0 62.0 0.63 1.0 12514 10/20/19 2019 7 False Tennessee Titans 23 20 Los Angeles Chargers TEN -2.5 42.5 Nissan Stadium False NaN NaN NaN NaN TEN LAC
7 1 A.J. Brown WR TEN 8 2019 REG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 3 11 1 0 0 32 1 1 0 0.090909 0.094955 7.1 9.1 11 0.0 0.0 5.50 2 1 0.0 0.0 0.0 0.0 0.500 1 0.500 33.87 2019.0 2.0 51.0 Jun-00 226.0 4.49 36.5 36.0 59.0 0.61 1.0 12530 10/27/19 2019 8 False Tennessee Titans 27 23 Tampa Bay Buccaneers TEN -2.0 45.5 Nissan Stadium False NaN NaN NaN NaN TEN TB
8 1 A.J. Brown WR TEN 9 2019 REG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 7 81 0 0 0 102 25 4 0 0.189189 0.310976 8.1 12.1 81 0.0 0.0 20.25 4 1 0.0 0.0 0.0 0.0 0.000 0 0.000 33.87 2019.0 2.0 51.0 Jun-00 226.0 4.49 36.5 50.0 72.0 0.69 1.0 12535 11/3/19 2019 9 False Carolina Panthers 30 20 Tennessee Titans CAR -3.5 43.0 Bank of America Stadium False NaN NaN NaN NaN CAR TEN
9 1 A.J. Brown WR TEN 10 2019 REG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 4 17 0 0 0 89 12 1 0 0.210526 0.354582 1.7 2.7 17 0.0 0.0 17.00 1 1 0.0 0.0 0.0 0.0 0.000 0 0.000 33.87 2019.0 2.0 51.0 Jun-00 226.0 4.49 36.5 47.0 50.0 0.94 1.0 12557 11/10/19 2019 10 False Tennessee Titans 35 32 Kansas City Chiefs KC -6.0 49.0 Nissan Stadium False NaN NaN NaN NaN TEN KC