NBA Analytics Research

This repository contains the complete codebase, datasets, and analysis for our NBA analytics research paper examining team performance and ranking methodologies across multiple seasons (2016-2024).

🏀 Overview

Our research applies advanced statistical modeling and machine learning techniques to analyze NBA team performance, with a focus on logistic regression models for ranking and classification tasks. The study encompasses comprehensive data from regular seasons, playoffs, and play-in tournaments.

📁 Repository Structure

Core Files

• README.md - This documentation

Data

Game Logs (gamelogs/)

• nba_games_cleansed.csv - Cleaned NBA game data
• nba_games.csv - Raw NBA game data

Standings (standings/)

• standing_2016.csv to standing_2024.csv - Season standings by year

Code

Data Cleaning (cleaning_code/)

• parse.ipynb - Data parsing notebook
• preprocessing.ipynb - Data preprocessing
• scrape.ipynb - Web scraping scripts
• sort_by_year.ipynb - Temporal data organization

Statistical Models (models/ and tables/)

🔬 Research Components

Data Collection & Processing

• Web Scraping: Automated data collection from NBA sources
• Data Cleaning: Comprehensive preprocessing and validation
• Temporal Organization: Multi-season data structuring (2016-2024)

Statistical Modeling

• Logistic Regression: Team ranking and classification models
• Pseudo R-squared Analysis: Model performance evaluation
• Group-based Analysis: Hierarchical team performance modeling

Analysis Scope

• Regular Season: Complete game logs and standings

🔄 Reproducibility

All analysis is fully reproducible:

1. Clone this repository
2. Install requirements
3. Run Jupyter notebooks in specified order (need to change the directories accordingly)
4. Results will match published findings

📈 Data Sources

• NBA official statistics
• Historical game logs (2016-2024)
• Regular season standings

🏆 Results

Our analysis provides:

• Team ranking algorithms with statistical validation
• Performance prediction models
• Comprehensive statistical tables
• Visualization of trends and patterns

📝 Citation

If you use this research or data in your work, please cite:

[Li & Jabbari] (2025). NBA Analytics Research: Advanced Statistical Modeling 
for Team Performance and Ranking. GitHub Repository.
https://github.com/JerryLi24/nba-analytics-research