Basketball is a data-driven sport where insights into team performance, player contribu ons, and evolving game trends enhance both analysis and fan engagement. With rich NBA data of across mul ple seasons, effec ve visualiza on is crucial to transform raw sta s cs into meaningful insights.
ABOUT THE DATASET
This project uses five structured CSV datasets related to the Na onal Basketball Associa on (NBA). Together, these datasets capture informa on about teams, players, games, standings, and individual game performance, allowing both team-level and player-level analysis across seasons.
- Teams Dataset (teams.csv) - The teams dataset contains sta c and historical informa on about NBA franchises.
- Players Dataset (players.csv) - The players dataset lists NBA players and the teams they played for during a specific season.
- Rankings Dataset (ranking.csv) - The rankings dataset provides team standings for a given season and date.
- Games Dataset (games.csv) - The games dataset contains game-level informa on for NBA matches. Each record represents a single game.
- Game Details Dataset (game_details.csv) - The game details dataset provides player-level sta s cs for each game.
The datasets are connected through shared iden fiers like, TEAM_ID links teams across all datasets, GAME_ID connects game-level data with player-level sta s cs, PLAYER_ID uniquely iden fies players across games and seasons. Together, these datasets form a rela onal structure that allows comprehensive analysis of NBA teams, players, and games across mul ple seasons.
ABOUT THE OPERATION
The whole analysis has been done in 4 parts :
PART A (FOUNDATIONAL ANALYSIS) - Includes Trends: Season-wise and game-wise scoring & performance trends, Rankings: Team rankings and win–loss pa erns over me , Team & Conference Analysis: Team comparisons and East vs West insights, Home vs Away performance comparison.
PART B (INTERMEDIATE ANALYSIS) - Includes Player Insights: Key player stats, consistency, standout, Performances, Consistency & Reliability Analysis: Iden fica on of the most consistent players and teams using game-to-game variability, highligh ng dependable contributors versus vola le performers, Bench & Rota on Impact: Analysis of starter vs bench contribu ons to team success, including scoring share and impact on win probability.
PART C (ADVANCED ANALYSIS) – Includes Advanced Performance Metrics: Custom player and team impact scores combining scoring, efficiency, playmaking, and defense to go beyond basic box-score analysis. Winning Factors Analysis: Sta s cal comparison of wins vs losses to iden fy which performance metrics (shoo ng efficiency, assists, rebounds) most strongly correlate with winning games. Home-Court Advantage Index: Quan fica on of home-court advantage by team using win percentage and scoring differen al, with conference-level comparisons, Clutch Performance Proxy: Evalua on of player and team impact in close games using plus-minus and narrow margin outcomes to iden fy clutch contributors.
PART D (STRATEGIC ANALYSIS) – Includes Team Iden ty Profiles: Offensive and defensive team “fingerprints” based on shoo ng efficiency, ball movement, and rebounding to classify team play styles. Momentum & Streak Analysis: Tracking win/loss streaks and rolling performance trends to highlight momentum shi s during the season. Franchise & Contextual Insights: Explora on of long-term factors such as franchise age and arena characteris cs in rela on to performance and home success.
This analy cs report represents a sophis cated transi on from Descrip ve Analy cs (what happened) to Diagnos c and Predic ve Storytelling (why it happened and what it means).
TECH STACK
Python, pandas, numpy , matplotlib, plotly, sckit learn
This repository contains the 5 csv datasets, the jupyter notebook file where the operation was performed and a fully documented report of the analysis..