Football Data

A collection of wrappers over football [*] data from various websites / APIs. You get: Pandas dataframes with sensible, matching column names and identifiers across datasets. Data is downloaded when needed and cached locally. Example Jupyter Notebooks are in the Github repo.

[*]	Soccer, if you're a heathen

Data sources:

fivethirtyeight.com

(https://projects.fivethirtyeight.com/soccer-predictions)

Season 2016-17 predictions and results for the top European and American leagues.

• Data no longer available on fivethirtyeight.com, I've added it to the repository

football-data.co.uk

(http://www.football-data.co.uk/)

Historical results, betting odds and match statistics for English, Scottish, German, Italian, Spanish, French, Dutch, Belgian, Portuguese, Turkish and Greek leagues, including a number of lower divisions. Level of detail depends on league.

clubelo.com

(http://clubelo.com)

First team relative strengths, for all (?) European leagues. Recalculated after every round, includes history.

Roadmap:

Add player stats, transfers, injuries and suspensions.

Installation

$ pip install footballdata

Dependencies

• Numpy
• Pandas
• Requests
• Unidecode

Usage

import footballdata as foo

# Create class instances
five38 = foo.FiveThirtyEight()
elo = foo.ClubElo()
mhist = foo.MatchHistory('ENG-Premier League', '2016-17')

# Create dataframes
matches = five38.read_games()
forecasts = five38.forecasts()
current_elo = elo.read_by_date()
team_elo_history = elo.read_team_history('Barcelona')
epl_2016 = mhist.read_games()

See the Jupyter Notebooks here for more elaborate examples: https://github.com/skagr/footballdata/tree/master/notebooks

Compatibility

Tested against Python 2.7 and 3.4-3.6

Licence

MIT