Fantasy Football project using Sports Reference API

Since this project is still in early stages this README is mostly just for notes of my own at the moment

Helpful Links

• Pro Football Reference Glossary
• Sportsipy
• NFL Stadiums/Teams dataset

Current Objectives

• Continue data validation
• Setup Relational Database system with the current dataframes
• Start with some simple machine learning models training to predict the over/under

---

Data Notes

• Dates are all in Eastern time
• All columns are lower-case and have underscores instead of spaces --> team_abbreviation

Known Issues (Backburner)

• API does not account for players traded during a season (See the EDA workbook for more clear examples)
  • Correcting the age column for this by using the .fillna method in Pandas, this works because the row above has their age at the beginning of that season.

Future (lofty) Goals

• Get weather report as function for upcoming games during season
  • Open Weather Map
• ESPN API for things like images and rankings etc.
  • Could possibly get from Reddit API

---

Data Structure

df_boxscores

Each row is a particular games boxscore with stats from both teams along with additional data regarding the particular game

• Primary Key: game_uri Is set as the index. Example --> '200009030min'
• Foriegn Key(s): year, month, day_of_week

nfl_player_stats_by_career

Each row is 1 particular player that has played in the NFL within the last 20 years (if no stat exists for a column it is filled with a NaN value)

• Primary Key: player_id
• Foriegn Key(s): team_abbreviation, name (of player)

nfl_player_stats_by_season

Each row is 1 particular player during a particular season within the last 20 years (if no stat exists for a column it is filled with a NaN value)

• Primary Key: player_id
• Foriegn Key(s): team_abbreviation, year, name (of player)

player_game_stats

Index: Shows year, game_uri (string that is a primary key for a particular NFL game played during that year), team_abbreviation, name (Players name) Rows: Each row is 1 particular player that has played in the NFL within the last 20 years.

• Primary Key: game_uri
• Foriegn Key(s): team_abbreviation, year, name (of player)

team_season_stats

Each row is 1 particular team throughout 1 particular year in the NFL within the last 20 years.

• Primary Key: team_abbreviation
• Foriegn Key(s): year