Ryan Knaus

The Numbers Game

print article archives RSS

NBA Defenses by the Numbers

Monday, July 28, 2014


Last week I examined NBA offenses in light of two data sets -- how often a team's offense used different play types (spot-ups, isolations, etc.), and how many points-per-possession they averaged when using a given play type.

 

Today's column is a companion piece with NBA defenses under the microscope. I focus on points-per-possession (PPP), but I eschew team-specific data about how often a team faced a given play type. For instance, the Jazz gave up a league-worst 0.95 PPP against post-ups last year. You would expect opponents to exploit this weakness, resulting in more post-up plays vs. Utah's defense, but that level of analysis must be saved for another day (as it happens, Utah defended post-ups 10.1% of the time last season, only a sliver above the league average of 9.6%).

 

To take another example, the Bobcats/Hornets were dead last in defending against offensive rebounds on a points-per-possession basis, though interestingly they also gave up the fewest OREB per game (9.6).

 

That discrepancy highlights the importance of putting PPP data in context. The ability to defend pick-and-rolls is far more important than defending against 'hand-offs', which account for a much smaller percentage of a team's total plays. I have therefore arranged the play types below in descending order, from most to least frequent, and will begin with a reminder (from last week's column) of the frequency of each play type in NBA offenses last season.

 

Image and video hosting by TinyPic

 

Source: Synergy Sports

 

As I noted last week, "This data is drawn from the 2013-14 season and may have limited applicability for teams which have undergone major changes this summer. The Cavaliers are a perfect example, having replaced their head coach while adding a No. 1 overall pick and the game-changing presence of LeBron James." With that in mind, let's proceed to teams' defensive stats against the play types identified by the Synergy database.

 

Follow me on Twitter for NBA updates, stats, and occasional musings: @Knaus_RW

 

Spot-up defense

Image and video hosting by TinyPic

 

The Clippers allowed only 0.86 PPP on opponent's spot-up attempts last season, better than the vaunted defenses of the Pacers (0.91) and Bulls (0.93). Spot-ups accounted for more of the league's offense than any other play type (excluding the combination of pick-and-roll plays ending with the ball-handler or roll man), so L.A.'s proficiency in this category was fundamental to their defensive success.

 

A quick diversion into shot locations (which may be the focus of a future column): during the 2013-14 season, the Clippers limited their opponents to a league-low 36.1% shooting in the mid-range (inside the arc, outside the paint). They were similarly effective vs. 3-pointers from above the break (31.5%) and from the right corner (33.8%). Oddly, however, they tied with the Cavaliers for 29th in the NBA defending 3-pointers from the left corner, allowing opponents to make 43.8% of those attempts.

 

There is a gradual slope from the Bulls downward, as even the 27th-ranked Bucks allowed 1.02 PPP, after which there is a huge gap leading to the two worst teams -- the Kings (1.07) and Jazz (1.07). The Jazz also rank dead last in opponent field goal percentage last year (47.3%) so it's no surprise that they fare poorly against most play types.

 

Pick-and-roll (ball-handler)

Image and video hosting by TinyPic


The Kings generally lousy defense (23rd in PPP) was tough against pick-and-roll ball-handlers, ranking fourth in the NBA at a mere 0.75 PPP. Unsurprisingly, Isaiah Thomas held his own by limiting ball-handlers to 0.73 PPP -- that's a noteworthy stat, since 45.7% of his defensive possessions ended vs. pick-and-roll ball-handlers. The Kings' new PG Darren Collison was equally impressive in this play type, limiting opponents to only 0.72 PPP, but he was markedly worst vs. almost every other play type.

 

Miami's defense, by a very wide margin, was the stingiest in this category. Their 0.59 PPP was untouchable, followed by the Clippers (0.73), Wizards (0.73), Kings (0.75) and Pacers (0.77). LeBron James was effective when he faced ball-handlers, and Miami's imposing team defense deserves much credit, but Mario Chalmers was singularly impressive. His defensive possessions ended with ball-handlers in pick-and-rolls 31.1% of the time, yet he ranked No. 1 in the NBA by allowing a ridiculous 0.51 PPP. He wasn't nearly as good against most other play types, but at least he's got something to hang his hat on.

 

At the other end of the spectrum, the Knicks were incapable of shutting down pick-and-roll ball-handlers. They allowed a league-worst 0.90 PPP, trailing even the lackluster Pistons (0.87) and Bucks (0.84). Nearly half of Raymond Felton's defensive possessions came vs. this play type, and he unsurprisingly allowed 0.9 PPP (ranking 205th in the league). Pablo Prigioni (0.87 PPP) was hardly better, and the Knicks' profound struggles make new starting PG Jose Calderon (0.78 PPP) seem like a brick wall by comparison.

 

Transition

Image and video hosting by TinyPic


The Pelicans, Jazz, Kings, Magic and 76ers were the worst teams in transition defense on a PPP basis. Evidently, defensive awareness and quality shot selection trumps young legs when it comes to stopping the break. As I noted last week, the tanking 76ers led the league in offensive transition opportunities last year but they were hugely inefficient at a league-low 0.99 PPP.

 

The Pacers, who come out with the lowest weighted PPP of any team's defense, were the league-leaders vs. both cuts and transition plays. Their defense was easily the toughest to score against in the restricted area, where opponents shot 53.0% last season -- the Thunder ranked second at 56.6%. I don't want to say too much about shot locations in this column, but it's worth pointing out how awful the Kings (65.1%) and Timberwolves (64.9%) were at defending the restricted area.

 

Post-up

Image and video hosting by TinyPic

 

The Knicks ranked 29th in the NBA for points-per-possession (weighted) across all play types, yet they allowed a league-low 0.78 ppp on post-ups. In a bizarre twist, NBA.com/stats reveals that Andrea Bargnani held his man to 48.3 percent shooting at the rim last season, which ranked 14th among all big men who averaged at least 25 minutes and 1.0 blocks per game. Tyson Chandler ranked 23rd at 51.5 percent.

 

The Heat ranked 26th vs. post-ups, which isn't so much as surprise as something worth reiterating. Their small-ball approach left them susceptible to big men with post games, as owners in daily leagues are well aware, and opposing centers were nearly automatic starts last season. With Chris Bosh slated as Miami's starting center, that's unlikely to change in 2014-15.

 

The Bulls, Rockets, Nets and Pistons were also tough against post-ups, spelling trouble for players heavily reliant on post-ups such as Al Jefferson, Zach Randolph, David Lee and Pau Gasol. Interestingly, Joe Johnson's emphasis on post-ups paid off last year and he ranked fourth last season in made baskets within 5-9 feet of the hoop. He was the only guard or forward in the top-12 from that range, though Rudy Gay, Jeff Teague and Isaiah Thomas all made the top-20.

 

Isolation

Image and video hosting by TinyPic

 

The Lakers' defense ranked in the middle of the pack overall, but they had the second-best defense on isolations and the best defense against the roll man in pick-and-rolls (while being mediocre against pick-and-roll ball-handlers).

 

The Timberwolves and Pelicans also ranked in the top-five vs. isolation plays, even though both teams were in the bottom-third of the league on a cumulative basis. The Wolves were 29th vs. hand-off plays, and (as noted) the Pelicans ranked 30th in transition. The ability to defend against isolations, it seems, doesn't say much about overall team defense.

 

Cut

Image and video hosting by TinyPic

 

On a league-wide basis, cuts were the most efficient play last season. They yielded 1.2 PPP and few teams were particularly effective at shutting down cuts -- even the league-best Pacers allowed 1.07 PPP, followed by the Bulls (1.11), Trail Blazers (1.12) and Spurs (1.13).

 

Cuts accounted for only 7.7% of the league's possessions, however, so it's a category heavily skewed toward teams that utilize it frequently. Per last week's column, no team used it more than the Bulls (a whopping 11.3% of their offense), but the Spurs (9.6%), Grizzlies (9.2%) and Warriors (9.0%) also made it a staple. In spite of its universal efficiency, cuts were an offensive rarity for the Suns, Knicks and Magic.

 

Pick-and-roll (roll man)

Image and video hosting by TinyPic

 

The Clippers ranked 29th against roll men in pick-and-rolls, but they were the hardest team to score against via spot-ups. The Pacers, Bulls, Celtics and Wizards were also very stingy vs. spot-up shooters, though the departure of Trevor Ariza may make Washington more lenient this season. Against such teams, owners in daily leagues may want to avoid guys like Kyle Korver, Klay Thompson, J.J. Redick, Mike Miller, Wesley Matthews, Danny Green, etc.

 

In general, I remain surprised that play ending with a roll man only accounted for 6.73% of NBA offenses last season.  It's a relatively efficient play type, yielding 1.0 PPP across the league, but only the Magic, Mavericks and Hawks used it for more than 8.60% of their offensive possessions.

 

Other

Image and video hosting by TinyPic

 

Synergy explains this play type as, "Things like last second heaves from the back court and when a player is intentionally fouled and sent to the free throw line."

 

In other words, it's a capricious category which doesn't provide many insights into team defenses. We might ponder the reasons why three Atlantic Division teams (Toronto, Boston, Brooklyn) were the worst at defending this random play type. We might, but let's not.

 

Offensive rebound

Image and video hosting by TinyPic

Offensive rebounds accounted for 5.81% of the league's offensive possessions last season, but when they yielded an efficient 1.1 PPP. The Bulls (0.94) and Thunder (0.94) were particularly tough to score against, even after corralling an offensive board, but the overall range is fairly narrow -- the 3rd-ranked Pacers allowed 1.00 PPP, and the 30th-ranked Bobcats/Hornets allowed 1.19 PPP.


Since offensive rebounds almost invariably lead to high-percentage attempts, a more salient approach would be to see which teams allowed them most often. Instead of looking at raw OREB totals, which would be skewed by a team's pace, I looked at OREB per 100 possessions.

 

The five teams easiest to secure an OREB against were the Lakers, Clippers, Bucks, Mavericks and Rockets.

 

The five toughest teams to get OREBs against were the Bobcats, Kings, Wizards, Grizzlies and Cavaliers. I must say that I wouldn't have guessed most of those teams.

 

Off-ball screen 

Image and video hosting by TinyPic

At 5.06% of the NBA's total offensive possessions last year, off-ball screens were relatively scarce. Interestingly, the Bucks led the NBA by using off-ball screens for 8.5% of their possessions, yet they were also the easiest team to score against with off-ball screens (yielding 1.05 PPP). The toughest teams to score against with off-ball screens were the Nets, Jazz, Hawks, Raptors and Warriors. That's not a typical grouping of "top-five defensive teams," which is another reflection of the infrequent use of this play type.

 

Hand off

Image and video hosting by TinyPic

Hand-offs accounted for a mere 3.11% of the league's offensive possessions, so I won't go beyond my list of the top-5 best and worst teams.

 

Total PPP allowed (weighted)

Note: For this 'total' I have weighted each team's PPP in a given category by that play's percentage of the league's total offense. The use of team-specific data for percentage of plays defended might have yielded some interesting insights. It also would have been very labor intensive, and I didn't deem it worth the effort considering the relatively minor disparities between the two data sets.

Image and video hosting by TinyPic

 

Next week's 'Numbers Game' column may dive into shot locations, or I may bypass that in favor of a column combining the PPP data from last week (offenses) and this week (defenses). In the meantime, check out Mike Gallagher's 2014-15 rookie rankings, and be sure to follow the Rotoworld crew on Twitter:

 

Aaron Bruski: @aaronbruski

Dominic Ridgard: @DLRiddy

Ed Isaacson: @nbadraftblog

Mike Gallagher: @MikeSGallagher

Nick Raducanu: @ProjectRoto

Ryan Knaus: @Knaus_RW

Steve Alexander: @docktora

 

Finally, here is a link to a single Excel sheet with (most of) the data I’ve been using throughout this column: Enjoy!



Despite residing in Portland, Maine, Ryan Knaus remains a heartbroken Sonics fan who longs for the days of Shawn Kemp and Xavier McDaniel. He has written for Rotoworld.com since 2007. You can follow him on Twitter.
Email :Ryan Knaus



Highest Searched Players over the last 7 days



Video Center

  •  
    Monday

    Monday's Fantasy Minute
  •  
    Beer’s NFL 6 Pack - WR & TE

    Beer’s NFL 6 Pack - WR & TE
  •  
    RotoPat:: Week 3 Rankings

    RotoPat:: Week 3 Rankings
  •  
    Beer’s NFL 6 Pack - QB & RB

    Beer’s NFL 6 Pack - QB & RB
  •  
    Thursday

    Thursday's Fantasy Minute
  •  
    Levitan: Drop options

    Levitan: Drop options
  •  
    Levitan: Biggest flukes

    Levitan: Biggest flukes
  •  
    Wednesday

    Wednesday's Fantasy Minute