/NBA

NBA Score Prediction

Primary LanguagePython

NBA Spread Predictions

An app for predicting the point spread for any given NBA basketball game.

Built off of swar's nba_api (1)

Occasional predictions found in DATA/Predictions. For predictions on today's games run below code:

Command Line
$ pipenv install
$ python -m NBA todays_bets --display True --save False
^^Incredibly slow due to how inputs to the model are currently being generated (i.e. many stats.nba api calls), see TO DO #6

Usage
$ pipenv install
$ pipenv shell
$ python
>>> from predcitions.predictions import GamePredict
>>> GamePredict('DEN', 'IND').predict_spread()

Output is in terms of the home team (i.e. prediction is DEN wins by 2.3)
-->[-2.3]

TO DO

  1. Command Line Interface
  2. Back test input generation method
    Currently using rolling average from last 6 games
  3. Scrape daily lines
  4. Write bet algo
  5. Method for recording bets and results
  6. Build database of past games and boxscores; configure methods to constantly update; will speed up prediction time