NBA Over/Under Prediction

An NBA Over/Under prediction model built using machine learning techniques to predict a statistic of an NBA player based on past statistics and other game-related data. This project uses logistic regression to classify whether the player's target statistic in his next game will go over or under a given betting line. It is intended for sports analysts, bettors, and anyone interested in analyzing NBA games.

Getting Started

Prerequisites:

• Python 3.7+
• Java 21.x+ (if using the GUI)

Installation:

1. Clone this repository: git clone https://github.com/yourusername/projectname.git
2. Navigate to the project directory: cd projectname
3. Create a virtual environment to isolate dependencies: python3 -m venv venv
4. Activate the virtual environment:
   • For Linux/macOS: source venv/bin/activate
   • For Windows: venv\Scripts\activate
5. Install the required dependencies: pip install -r requirements.txt

Usage

Key Parameters

1. First Name: The first name of the player (as found at https://www.nba.com/players)
2. Last Name: The last name of the player (as found at https://www.nba.com/players)
3. Target Stat: The statistical category you want to predict for (MIN, FGM, FGA, FG3M, FG3A, FTM, FTA, OREB, DREB, REB, AST, TOV, STL, BLK, PTS, PLUS_MINUS).
4. Money Line: The betting line (e.g., 25.5).

Through GUI

1. Start the server in the backend folder. From the root project directory, run:

   • For Linux/macOS/Windows:

     fastapi dev backend/main.py

   • For Windows:

     fastapi dev backend\main.py

2. Compile the Java file in the GUI/MoneyLinePredictor/src folder. From the root project directory, run:

   • For Linux/macOS/Windows:

     javac GUI/MoneyLinePredictor/src/PredictorApp.java

   • For Windows:

     javac GUI\MoneyLinePredictor\src\PredictorApp.java

3. Run the Java GUI application by executing the following command from the root project directory:

   • For Linux/macOS:

     java -cp GUI/MoneyLinePredictor/src PredictorApp

   • For Windows:

     java -cp GUI\MoneyLinePredictor\src PredictorApp

4. In the name box: enter the first name and last name of the player.

5. In the target stat box: enter the desired statistic.

6. In the money line box: enter the betting line.

7. Click the predict button to see the prediction.

Troubleshooting

If you encounter any issues, here are a few things to check:

1. Ensure that you've activated the virtual environment and installed the dependencies using pip install -r requirements.txt
2. Ensure you're using Python 3.x. Verify the version by running python --version
3. Make sure you have Java 21.x+ installed for the GUI. If there are issues compiling and running the .java file, try reinstalling Java.

Acknowledgments

This project uses the following libraries and resources:

• scikit-learn: A machine learning library for Python. (https://scikit-learn.org/)
• pandas: A data manipulation and analysis library. (https://pandas.pydata.org/)
• nba_api: Python API client for accessing NBA.com data. (https://github.com/swar/nba_api/)