Analyzing NBA Stats

When to use this skill

• •User asks for "player props" or "recent form" for NBA players.
• •User wants to analyze "box scores" or "advanced metrics" (PER, TS%, Usage).
• •Integrating new NBA data sources into the model component.

Workflow

• • Source Selection: Decide between nba_api (Python wrapper for stats.nba.com) or external scraping if needing prop betting odds.
• • Normalization: Map Player Names to IDs consistently. Handle "J. Brown" vs "Jaylen Brown".
• • Data Frame Creation: Always load data into a Pandas DataFrame for analysis.
• • Feature Engineering: Calculate rolling averages (last 5 games, last 10 games) as these are crucial for betting models.

Instructions

1. Using nba_api (Python)

The model directory is likely where this code lives.

from nba_api.stats.endpoints import playergamelog
import pandas as pd

def get_player_last_n_games(player_id, n=5):
    # Season '2023-24' needs to be dynamic
    log = playergamelog.PlayerGameLog(player_id=player_id, season='2024-25') 
    df = log.get_data_frames()[0]
    return df.head(n)

2. Key Metrics for Betting

Focus on these stats for prop bets:

• •PTS, REB, AST: Standard props.
• •Minutes Played: High correlation with output.
• •USG% (Usage Rate): Good for predicting high-scoring games.

3. Rate Limiting

stats.nba.com is strict.

• •Add user-agent headers if making raw HTTP requests.
• •Implement delays between requests in loops.

Resources

• •nba_api Documentation