Defensive Metrics

Overview

Defensive metrics datasets track individual player defense including opponent shooting when guarded by the player, rim protection, defensive field goal percentage (DFG%), and defensive impact statistics. This data enables analysis of defensive effectiveness and shot contestation.

Data Files

Defense Master

Comprehensive defensive statistics

defense_master.csv (Regular Season)
defense_master_ps.csv (Playoffs)

Opponent FG%

Overall opponent shooting statistics

dfg.csv / dfg_p.csv

Rim Defense

Rim protection metrics (< 6 feet)

rimdfg.csv / rimdfg_p.csv

Rim Accuracy

Opponent accuracy at rim when contested

rim_acc.csv / rim_acc_p.csv
rimfreq.csv / rimfreq_p.csv

Schema: Defense Master

File: defense_master.csv / defense_master_ps.csv
Generated by: defense.py + organize_defense.py
Records: ~20,000+ player-season records
Source: Combined NBA.com and pbpstats.com data

Core Fields

Identifiers
Overall Defense
Rim Protection

nba_id

integer

required

NBA.com player ID

team_id

integer

NBA.com team ID

PLAYER

string

required

Player name

Team

string

Team abbreviation

year

integer

required

Season ending year (2014-2025)

Sample Data

nba_id,team_id,PLAYER,Team,overall_dfg%,all_dfga,all_dfgm,dif%,year,rim_dfg%,rim_dfga,rim_dfgm,rim_dif%
201599,1610612746,DeAndre Jordan,LAC,45.5,1375,626,-2.4,2014,56.4,621,350,-3.7
201577,1610612757,Robin Lopez,POR,43.4,1315,571,-4.4,2014,46.7,642,300,-12.7

Schema: Opponent Field Goal %

File: dfg.csv / dfg_p.csv
Generated by: defense.py
Source: NBA.com leaguedashptdefend endpoint

PLAYER_ID

integer

NBA.com player ID

PLAYER

string

Player name

TEAM_ID

integer

Team ID

TEAM

string

Team abbreviation

POSITION

string

Player position

AGE

integer

Player age

integer

Games played

integer

Games started

FREQ%

float

Frequency of defensive matchups (0-100 scale)

DFGM

integer

Defensive field goals made (opponent)

DFGA

integer

Defensive field goal attempts (opponent)

DFG%

float

Defensive field goal percentage (opponent FG% when guarded)

FG%

float

Expected opponent FG% (league average for those shot types)

DIFF%

float

Difference between DFG% and expected FG% (negative = good defense)

year

integer

Season year

Schema: Rim Defense

File: rimdfg.csv / rimdfg_p.csv
Purpose: Tracks opponent shooting at the rim (< 6 feet) when contested by player Fields identical to dfg.csv but filtered for rim attempts only.

Schema: Rim Accuracy & Frequency

Files:

rim_acc.csv / rim_acc_p.csv - Opponent accuracy at rim
rimfreq.csv / rimfreq_p.csv - Frequency of rim attempts allowed

Source: pbpstats.com API (on-off splits)

pbpstats.com Fields

Name

string

Player name (lowercase)

Team

string

Team name (full)

Year

integer

Season ending year

Season

string

Season format: “2023-24”

AtRimAccuracyOpponent

float

Opponent shooting accuracy at rim when player on court

AtRimFrequencyOpponent

float

Frequency of opponent rim attempts when player on court

Understanding Defensive Metrics

Negative DIF% is Good Defense

DIFF% = -5.0 means opponents shoot 5% worse than expected
DIFF% = +3.0 means opponents shoot 3% better than expected (poor defense)

Key Concepts

Defensive Field Goal % (DFG%)

The field goal percentage of shots taken when a specific player is the closest defender. This measures how well a player contests shots.Formula: DFG% = DFGM / DFGA * 100Lower DFG% = Better defense

Differential % (DIFF%)

Compares opponent’s actual shooting percentage to their expected FG% based on shot type and location.Formula: DIFF% = DFG% - Expected_FG%Negative differential = Player forces worse shooting than expected
Positive differential = Player allows better shooting than expected

Rim Protection

Defensive metrics specifically for shots taken within 6 feet of the basket. Elite rim protectors have:

Low rim_dfg% (< 50%)
High rim_dfga (high volume of rim contests)
Negative rim_dif% (force misses)

Usage Examples

Find Elite Rim Protectors

import pandas as pd

df = pd.read_csv('defense_master.csv')

# 2024 season, minimum 200 rim contests
elite_rim = df[
    (df['year'] == 2024) & 
    (df['rim_dfga'] >= 200)
].sort_values('rim_dif%')  # Most negative = best

print("Top 20 Rim Protectors (2024):")
print(elite_rim[['PLAYER', 'Team', 'rim_dfga', 'rim_dfg%', 'rim_dif%']].head(20))

Overall Defensive Impact

import pandas as pd

df = pd.read_csv('dfg.csv')

# 2024 season, high volume defenders
impact = df[
    (df['year'] == 2024) & 
    (df['DFGA'] >= 500)  # Min 500 contests
].sort_values('DIFF%')

print("Best Defensive Impact (2024):")
print(impact[['PLAYER', 'TEAM', 'DFGA', 'DFG%', 'FG%', 'DIFF%']].head(20))

Rim Protection vs Perimeter Defense

import pandas as pd

df = pd.read_csv('defense_master.csv')

# Compare rim defense to overall defense
df_2024 = df[(df['year'] == 2024) & (df['rim_dfga'] >= 150)].copy()

# Calculate perimeter defense (non-rim)
df_2024['perimeter_dfga'] = df_2024['all_dfga'] - df_2024['rim_dfga']
df_2024['perimeter_dfgm'] = df_2024['all_dfgm'] - df_2024['rim_dfgm']
df_2024['perimeter_dfg%'] = (df_2024['perimeter_dfgm'] / df_2024['perimeter_dfga'] * 100).round(1)

print("Rim vs Perimeter Defense:")
print(df_2024[['PLAYER', 'rim_dfg%', 'perimeter_dfg%', 'overall_dfg%']].head(20))

Track Individual Matchup Performance

import pandas as pd

df = pd.read_csv('dfg.csv')

# Look at specific player's defensive evolution
player_name = 'Anthony Davis'
player_defense = df[df['PLAYER'] == player_name].sort_values('year')

print(f"Defensive Evolution: {player_name}")
for _, row in player_defense.iterrows():
    print(f"{row['year']}: {row['DFGA']} contests, {row['DFG%']:.1f}% DFG, {row['DIFF%']:+.1f} DIFF")

On-Court Rim Defense (Team Context)

import pandas as pd

rim_acc = pd.read_csv('rim_acc.csv')
rim_freq = pd.read_csv('rimfreq.csv')

# Merge rim accuracy and frequency
merged = rim_acc.merge(
    rim_freq[['Name', 'Year', 'AtRimFrequencyOpponent']], 
    on=['Name', 'Year']
)

# 2024 data
merged_2024 = merged[merged['Year'] == 2024]

print("Rim Defense Impact (2024):")
print(merged_2024[['Name', 'Team', 'AtRimAccuracyOpponent', 'AtRimFrequencyOpponent']].head(20))

Defensive Position Analysis

import pandas as pd

df = pd.read_csv('dfg.csv')

# Compare defensive metrics by position
by_position = df[df['year'] == 2024].groupby('POSITION').agg({
    'DFG%': 'mean',
    'DIFF%': 'mean',
    'DFGA': 'mean'
}).round(1)

print("Defensive Metrics by Position (2024):")
print(by_position)

Data Collection Scripts

defense.py

Collects DFG% data from NBA.com and pbpstats.com

organize_defense.py

Combines defensive datasets into defense_master.csv

Important Notes

DIFF% is the key metric: Measures defensive impact relative to expected performance
Volume matters: High DFGA indicates a player who defends frequently (usually starters)
Context: Big men typically have lower DFG% due to rim protection, guards contest more perimeter shots

Defensive metrics don’t capture help defense, rotations, or team schemes. They only measure the closest defender at time of shot.

Combine defensive metrics with tracking data (deflections, steals, charges) from hustle.csv for comprehensive defensive analysis.

Get Started

Data Collections

Scripts & Automation

Data Schema

Overview

Data Files

Defense Master

Opponent FG%

Rim Defense

Rim Accuracy

Schema: Defense Master

Core Fields

Sample Data

Schema: Opponent Field Goal %

Schema: Rim Defense

Schema: Rim Accuracy & Frequency

Understanding Defensive Metrics

Key Concepts

Usage Examples

Find Elite Rim Protectors

Overall Defensive Impact

Rim Protection vs Perimeter Defense

Track Individual Matchup Performance

On-Court Rim Defense (Team Context)

Defensive Position Analysis

Data Collection Scripts

defense.py

organize_defense.py

Important Notes

Build docs developers (and LLMs) love

Get Started

Data Collections

Scripts & Automation

Data Schema

Documentation Index

​Overview

​Data Files

Defense Master

Opponent FG%

Rim Defense

Rim Accuracy

​Schema: Defense Master

​Core Fields

​Sample Data

​Schema: Opponent Field Goal %

​Schema: Rim Defense

​Schema: Rim Accuracy & Frequency

​Understanding Defensive Metrics

​Key Concepts

​Usage Examples

​Find Elite Rim Protectors

​Overall Defensive Impact

​Rim Protection vs Perimeter Defense

​Track Individual Matchup Performance

​On-Court Rim Defense (Team Context)

​Defensive Position Analysis

​Data Collection Scripts

defense.py

organize_defense.py

​Important Notes

Build docs developers (and LLMs) love

Overview

Data Files

Schema: Defense Master

Core Fields

Sample Data

Schema: Opponent Field Goal %

Schema: Rim Defense

Schema: Rim Accuracy & Frequency

Understanding Defensive Metrics

Key Concepts

Usage Examples

Find Elite Rim Protectors

Overall Defensive Impact

Rim Protection vs Perimeter Defense

Track Individual Matchup Performance

On-Court Rim Defense (Team Context)

Defensive Position Analysis

Data Collection Scripts

Important Notes