Advanced Stats 101: Defense

August 8, 2008

Most recently in this continuing space, Advanced Stats 101 has looked at APBRmetrics for Player Shooting as well as exploring Team and Player Offensive Ratings. Given the old adage that defense wins championship, it seems only appropriate to follow with a look at two key defensive stats and their application.

Measuring defensive prowess and success has long been a difficult thing to accurately project, particularly for individual players. Generally, we use dated measures such as: Opponent Field Goal Percentage, Opponent Points Per Game, Rebounding Rates and Turnovers Against.

While these all have their place, every dedicated fan knows that much of what makes a great defender cannot be expressed in pure numbers alone (something Hoops Addict’s Floor Burn Tournament is currently showing us). We all accept that Bruce Bowen is an all-world defender based on watching his tactfulness and tenaciousness first-hand, yet he is not one to lead the league in steals or blocks.

Likewise, when we consider great defensive teams, we have a few standard measures to point to as a means of validating our assumptions about the greatness of team units like Boston, Detroit and San Antonio. But as with the offensive statistics we have already looked at, APBRmetrics have advanced how defense can be studied too.

Consider the most basic of team defensive measures, Opponent Field Goal Percentage:

Traditional Opponent FG%
1 Boston 0.419
2 Houston 0.433
3 Detroit 0.437
4 Dallas 0.443
5 San Antonio 0.444
6 Los Angeles 0.445
7 Orlando 0.446
8 Portland 0.451
9 Chicago 0.453
10 Indiana 0.454

This list largely reflects what a subjective viewing of the game tells us, as the usual suspects of defensive intensity are all listed. Yet, as explained previously in Advanced Stats 101, there exists a more complete shooting measure. Effective Field Goal Percentage (eFG%= (FGM + .5*3FGM)/FGA) adjusts basic field goal percentage to reflect the difficulty and importance of the three-point shot. Since it provides a more complete and accurate picture of a team’s shooting efficiency, it can easily be used to assess a team’s defensive efficiency as well. This measure sees that teams with defenses susceptible to threes have this weakness expressed.

Opponent Effective Team FG%
1 Boston 0.457
2 Houston 0.465
3 Detroit 0.470
4 Dallas 0.474
5 San Antonio 0.477
6 Orlando 0.484
7 Los Angeles 0.485
8 Phoenix 0.488
9 Portland 0.491
10 Cleveland 0.494
League Average 0.497

In and of itself, the league top 10 reveals only small, insignificant changes. Boston and Houston still dominate, and the majority of the other teams fall in around the same order. But let’s look slightly beyond the top 10.

There is some very interesting movement amongst playoff teams not listed above, all of whom place below the League Average for OEFG%. Consider that Toronto drops from 15th to 22nd (.505 OEFG%) based on their inability to defend the three. Even more incriminating is the case against Washington, who plummets from the already-low-for-a-contender 18th slot all the way to 26th (.513 OEFG%) placing the Wizards amongst the likes of New York, Miami and Minnesota.

Alternatively, some playoff teams see a benefit of considering the importance of defending threes. Atlanta jumps up from 21st to 16th (.501 OEFG%), and equally impressive is Utah’s bump from 20th to 17th (.501 OEFG%).

Certainly defensive effectiveness cannot be measured by opposing field goal percentage alone. Another traditional, standard defensive metric is Opponent Points Per Game. The reason for using this is simple: a team that allows its opponents fewer points is more likely to win. Here is a list of the league’s best in this category:

Traditional Opponent Points Per Game
1 Detroit 90.1
2 Boston 90.3
3 San Antonio 90.6
4 Houston 92.0
5 New Orleans 95.6
6 Dallas 95.9
7 Philadelphia 96.2
8 Portland 96.3
9 Cleveland 96.7
10 Toronto 97.3

Again, this ranking reveals many of the teams one expects when considering great defense. However just as we made the case for Offensive Rating previously, we must do the same for Defensive Rating. Defensive Rating is a way of measuring success based on the idea that total points against is meaningless unless considering for the same number of possessions, given that each team plays at a different pace.

Defensive rating then measures points allowed per 100 possessions. This standardizes how we measure defense and does not reward teams such as Cleveland that play its games at a very slow pace. The rating was developed by Dean Oliver, a leader in APBRmetrics movement whose work focuses largely on evaluating players and teams on possession and situational analysis. The top 10 for Defensive Rating is below:

Defensive Rating
1 Boston 98.9
2 Houston 101.6
3 San Antonio 101.8
4 Detroit 102.9
5 Los Angeles 105.5
6 Orlando 105.5
7 New Orleans 105.7
8 Philadelphia 106.0
9 Dallas 106.1
10 Denver 106.3
League Average 107.5

The most striking revelation comes from Denver, who is generally considered a very porous and uninspired defensive team (even if they had/have strong individual defenders such as Marcus Camby and Kenyon Martin). The Nuggets are 29th in the league using points against per game at 107, yet when we account for Denver’s fast pace and high number of possessions by looking at the measure on a per possession basis, they leapfrog all the way to 10th overall, ahead of the league average as well are defensive-minded teams such as Cleveland, Utah and Chicago. This is a considerable achievement. Another team that benefits positively is Los Angeles: the Lakers are 19th in opponent scoring (101.3) but rise up all the way to fifth in defensive rating.

On the other side of the coin, Portland drops considerably from eighth down all the way to 17th (108.4 D-Rating). This measure too spells trouble for the Wizards. Just as with the shooting statistics, this measure shows just how ineffective the Wizards’ defense is on a per-possession basis, as Washington falls from 12th to 24th (109.6 D-Rating). Washington is the lowest among all playoff teams in this respect.

Finally, one might be interested to examine Defensive Rating for Individual Players to see who allows the least amount of points per 100 possessions against them. It should surprise no one that Boston dominates the list, with three of the top five in the league. It is also predictable that perhaps the two best defensive power-forwards of this (or any) era sit atop the list:

Player Defensive Rating
1 Kevin Garnett-BOS 93.8
2 Tim Duncan-SAS 96.6
3 Chuck Hayes-HOU 96.7
4 Kendrick Perkins-BOS 97.3
5 James Posey-BOS 98.0

Photo Credit: Icon Sports Media



7 Comments on “Advanced Stats 101: Defense”.

  • Garron

    I was just wondering how did you work out that last defensive rating for a player per 100 posessions? Is this by matching the teams defensive rating to the time when the player is on the court?

    August 8, 2008 » 5:43 PM »

  • [...] Holler at it! [...]

    August 8, 2008 » 6:31 PM »

  • [...] ready to take on the world New Haven Register Former Huskies back to be with Calhoun HoopsAddict Advanced stats 101: Defense ABS-CBN It’s all about the money The Examiner Pro basketball team coming to Ford Arena [...]

    August 9, 2008 » 9:46 AM »

  • Kevin Garnett is a top defensive player in the NBA of historic proportion, that much is a given.

    However, despite his many accolades, and my personal affinity for Timmy D, as an elite level NBA player and a person, IMHO, he has always been highly over-rated as a Defensive Stalwart.

    Yes, he is solid, as 1-on-1 defender on the Block; and, yes, he is a decent shot-blocker … but, he simply does not cover enough ground coming from off the ball to block the shots of incoming players AND do a good enough job of moving his feet laterally on the perimeter vs smaller players to be considered amongst the very best Defensive Players in the NBA.

    re: Chuck Hayes, Kendrick Perkins and James Posey

    Solid players all … but nowhere near versatile enough to be considered as legitimate threats in this specific category when you’re talking and measuring them against men like Kobe Bryant (G/F), Bruce Bowen (F), Tyson Chandler (C) or Andrei Kirilenko (F).

    August 14, 2008 » 10:50 PM »

  • Trevor Smith

    Garron- I had meant to include this explanation earlier, but yes you are exactly right. Individual Defensive Rating is determined based on the team’s rating only during those minutes the player is actually on the court. So if Boston’s rating suffers when Garnett is off, his own rating is still the same (almost like a +/- stat)

    August 18, 2008 » 9:07 PM »

  • Greg

    Boston 09′

    August 22, 2008 » 12:01 PM »

  • [...] Holler at it! Share and Enjoy: [...]

    January 28, 2009 » 11:18 PM »

Have Your Say:

© Copyright 2013 HOOPSADDICT.com.
Logo Designed by Matt Dyck Studios.
Here's our Privacy Policy.
Good Times brought to you by Matty Morgs™ & The Brosociates.