This week's column explores NBA offenses by play types such as isolations, off-ball screens, pick-and-rolls and post-ups. The focus is on two specific data sets -- how many 'points per possession' (PPP) each team averaged with a particular play type, and what percentage of each team's offensive possessions was accounted for with a particular play type.
To take a random example, the Hawks averaged 0.86 PPP on post-up plays during the 2013-14 season, which was 12th best in the league. However, post-ups only accounted for 6.2% of their total offensive chances (resulting in a shot, turnover or free throws), which ranked 28th in the NBA ahead of the Thunder (5.6%) and Pelicans (5.1%). Atlanta's efficiency in post-ups, therefore, was negated by limited attempts. That should come as no real surprise -- Al Horford (torn pectoral muscle) was out of commission and Pero Antic very rarely posted up, leaving Paul Millsap as Atlanta's only reliable option. I was delighted to learn, contrary to my expectations, that only 8.1% of Elton Brand's offensive possessions ended as post-ups, whereas 54.7% ended as a roll man (28.0%) or off cuts (26.7%).
The 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. That caveat doesn't prevent the data from showing trends and tendencies throughout NBA offenses, however, while providing useful perspectives on offensive schemes, play calling, and matchups.
This column will be followed next week by a similar analysis of each team's defense against play types, and I may also bring shot distances into the mix. Having done so, I hope to overlap the two data sets and see what happens. It took a long time just to compile these spreadsheets, and I couldn't have done it without the glorious information available only at Synergy Sports and on NBA.com/stats. If you notice anything interesting or surprising in the numbers, or have any specific questions, send me an email or a Direct Message on Twitter.
You can view the spreadsheets I compiled on Google Sheets via this link...unfortunately it messes with some of the chart details and eliminates the conditional formatting which I find very helpful.
Alternately, you can view and/or download my Excel file on Dropbox to preserve the original formatting and play around with the file in any way your heart desires.
To begin with the broadest view possible, the following charts reference unweighted league-wide statistics.
Source: Synergy Sports
Source: Synergy Sports
Notice that while spot-up attempts account for the largest chunk of offensive possessions across the league, pick-and-rolls would be the most prominent if Synergy didn't divide it into two categories (ball-handler and roll man). Spot-ups prove to be relatively efficient at 0.98 PPP, and it's no surprise that the most efficient play types are cuts, transition opportunities and offensive rebounds.
One thing which did surprise me was the relative inefficiency of plays finished by pick-and-roll ball-handlers, who averaged only 0.79 PPP across the NBA last season. This play type accounted for 15.1% of offensive possessions league-wide, but in the Mavericks' offense that jumped to a whopping 21.4%, far more than any other team. Dallas used it well with Jose Calderon running the show, resulting in 0.82 PPP, but that fails to justify their league-high concentration of plays ending with the ball-handler in pick-and-rolls.
Taking that example as our cue, let's whisk through each play type to examine outliers for points per possession, percentage of offensive chances, or both. Play types are arranged from most to least frequent. (There is a ton of information available here, and I'm not pretending to get at half of it. In the interest of space, time and my sanity, I leave the bulk of the analysis and inferences to the reader.)
NBA offenses used spot-ups for 18.6% of all offensive opportunities last season. No team used it more than the Heat (24.1%), who were followed by the Nets, Jazz, Hawks and Knicks.
The teams least likely to use spot-ups were the 76ers, Thunder, Nuggets, Kings and, least of all, the Warriors.
The woebegone 76ers were also the least effective team in spot-ups, producing 0.84 PPP (the league average was 0.98). They were joined in the cellar by the Pistons, Grizzlies, Celtics and Bulls.
I've already mentioned the inefficiency of this play type, as well as the Mavericks' surprisingly heavy reliance upon it for offense. The Raptors, Cavaliers, Thunder and Trail Blazers were also very prone to plays ending with the ball-handler in pick-and-rolls. The Thunder and Cavaliers made the top-10 for PPP in the category, but the Raptors and Blazers were more middle-of-the-pack.
The least efficient teams were the Celtics, 76ers, Jazz, Timberwolves, Wizards and Lakers. Note that all of these teams featured PGs who struggled with their shooting percentages last season:
Rajon Rondo - 40% from the field, 29% from downtown (Phil Pressey was even worse)
Michael Carter-Williams - 40%, 26%
Trey Burke - 38%, 33%
Ricky Rubio - 38%, 33%
John Wall - 43%, 35%
Kendall Marshall - 41%, 40%
The Nets used this play type the least (11.3% of their offensive possessions), another symptom of Deron Williams' faded glory.
Transition plays were the third-most common among NBA offenses last season, behind spot-ups and pick-and-rolls. They also yielded the second-most PPP behind cuts. The Suns (18.2% of their opportunities), Rockets (17.9%), Nuggets (16.9%) and Clippers (16.5%) were all near the top of the league in transition opportunities. Nevertheless, the team which used the play type most was awful last season -- that would be the frenetic, D-League-heavy 76ers, who also ranked dead last in PPP for transition opportunities (an abysmal 0.99 PPP).
On the other end of the spectrum, the teams least inclined to transition opportunities were the Knicks, Mavericks, Grizzlies, Bulls and the Raptors (if you're keeping track, the Knicks led all teams in isolations and were last in transition). The Raptors were the sixth-most efficient team in this play type, however, averaging 1.17 PPP in transition, so they made the most of their limited opportunities. The Mavs were the most efficient of all teams in transition, followed by the Clippers, Wizards and Heat.
It makes sense that the Grizzlies used post-ups far more than any other team (17.5% of their opportunities) but I was mildly surprised to see them ranked 16th at only 0.85 points per possession. NBA teams do have scouting reports and Memphis ranked dead last in 3-pointers made per game, so defenses simply packed the paint against Zach Randolph and Marc Gasol.
The Pacers, Kings and Hornets also leaned heavily on post-ups for their offense, but only Charlotte cracked the top-10 in PPP. The Heat led all teams in PPP for post-ups, followed by the Mavericks, Nets and Clippers. LeBron James (1.1 PPP) was the sixth most efficient player in post-ups, Dirk Nowitzki was 14th, Joe Johnson was 26th, and Blake Griffin was 34th -- and unsurprisingly, all four of them relied heavily on post-ups for their offensive possessions.
The Pistons, Wizards, Rockets and Bucks were all near the bottom in PPP for post-ups. The Bulls, who ranked last in isolation efficiency, were also dead last in this play type (as it happens, post-ups accounted for 38.6% of Pau Gasol's possessions last season).
Across the league, isolation plays yielded an average of 0.84 points per possession. However, the Thunder (0.95), Clippers (0.95) and Heat (0.94) were all far more effective using isolations -- it's no surprise that they also feature preeminent talents in Kevin Durant/Russell Westbrook, Chris Paul/Blake Griffin, and LeBron James/Dwyane Wade. Whereas Miami and OKC relied heavily on isos as part of their offense, Doc Rivers' Clippers were right around the league average (7.9% of their offense).
The Bulls were easily the worst in this category at a mere 0.68 points per possession, followed by the Grizzlies (0.76), Nuggets (0.77) and Pistons (0.77). The Grizzlies and Bulls were wise enough not to rely on isos for their offense (see spreadsheet), but the Nuggets and Pistons weren't as savvy or strategic. The four teams most heavily reliant on isolations were the Warriors, Nets, Pelicans and the league-leading Knicks, whose offense was 13.4% isolations -- that's sure to change now that coach Derek Fisher is installing the triangle offense.
Cuts accounted for 7.7% of all NBA teams' offense last year, which made it the sixth most common play type in Synergy's database. Cuts also proved to be the most efficient play type at a stellar 1.2 PPP. The Bulls haven't fared well in other categories during this analysis and they prove to be a mixed bag here as well -- despite easily leading the NBA in cuts (11.3% of their offensive chances) they ranked 28th in PPP. That still put them at 1.13 points per possession, an excellent number compared to other play types.
The Jazz, 76ers, Kings and Cavs also struggled to convert off cuts, and the category's PPP leaders were all playoff contenders -- the Heat, Rockets, Blazers and Clippers. Gregg Popovich's Spurs used cuts for 9.6% of their offensive possessions, trailing only Chicago, whereas the teams least likely to use cuts were the Blazers (negating their efficiency), Clippers (ditto), Magic, Knicks and Suns.
Pick-and-roll (roll man)
Plays ending with the roll-man off screens were, to nobody's surprise, the fourth-most efficient play type behind cuts, transition plays and offensive rebounds. They accounted for 6.7% of the league's total offensive opportunities, and no teams used it more often than the Mavericks, Magic, Hawks and Blazers. The Mavs were also the fourth-most efficient team, notching 1.1 PPP via roll men, making it a very potent source of scoring for Dallas. The Heat were the most efficient at 1.18 PPP, followed by the Suns and Rockets.
In spite of their efficiency, however, the Rockets' offense finished with a roll man just 3.7% of the time. That was easily the lowest percentage of any team's offense last season, and preseason talk of Dwight Howard being featured in pick-and-rolls proved to be wishful thinking. Side note: Dwight was tremendously effective in a few offensive categories including transition (1st in PPP across the entire NBA), cuts (2nd) and as a roll man (3rd). Unfortunately, those categories combined were only 21.3% of his offensive opportunities. Fully 53.1% of his chances came in post-up situations, where he averaged only 0.77 PPP (128th in the NBA). I doubt those numbers will change significantly at this point in his career, even if he works out with Hakeem Olajuwon 12 hours a day this summer.
The other teams least likely to finish plays with a roll man were the 76ers, Bucks, Suns and Pistons.
According to Synergy, "This category includes things like last second heaves from the back court and when a player is intentionally fouled and sent to the free throw line." Such random plays accounted for 6.4% of the league's total offense and yielded the worst PPP of any play type at just 0.49.
The Rockets were the logical leaders in PPP for the category due to Dwight Howard being intentionally fouled -- even his 54.9 percent FT shooting would shatter the league-wide PPP for 'other' plays. For whatever reason this category accounted for the largest offensive share with the Pacers, 76ers, Suns, Bulls and Nuggets. Reading into this category is not a wise use of your time.
The Pistons' offense finished with offensive rebounds 8.8% of the time, easily the highest mark in the NBA. They were followed by the Kings, Wolves, Grizzlies, Bulls and Bucks. Of the six teams mentioned so far, only the Timberwolves and Kings made the top-10 for PPP in the category -- the Bulls were mediocre, while the Pistons, Grizzlies and Bucks were all in the bottom-third of the league. It must be said that this is an extremely efficient play type, with a range from 1.2 PPP to 0.95 PPP, so for offensive rebounds frequency is more important than efficiency. That said, the teams with the lowest percentage of plays concluding with OREB are the Nets, Hornets/Bobcats, Heat, Spurs and Hawks.
This play type constituted only 5.1% of NBA teams' offensive opportunities last year, but for the Milwaukee Bucks it was a healthier 8.5% chunk (see below for a major caveat). They were followed by the Pacers (7.3%), Clippers (6.5%) and Raptors (6.4%), with a steady decline all the way down to the Lakers (3.5%), Pistons (2.9%) and Rockets (2.4%).
Despite their infrequent use of off-ball screens, Houston still ranked second with 1.03 PPP, just behind Portland and ahead of Phoenix. The Nuggets proved to be the least effective team with this play type, managing a paltry 0.76 PPP. As alluded to above, the Bucks ran plenty of off-ball screens but they didn't do it very well -- their 0.77 PPP ranked 29th in the league.
This play type accounted for a mere 3.1% of offenses last year, on average, so it's not worth laboring over the data...there's enough to consider without mulling the reasons why Miami was by far the team least likely to finish plays with hand offs (0.7% of their offensive chances) or why Philadelphia used it so frequently (6.0%). As usual, if you want to know more you can browse the spreadsheets at your leisure.
Note: If you're still making sense of this summer's flurry of free-agent movement, check out my freshly updated columns on that topic. The first three 'Overview' columns summarize every FA who has (or hasn't) agreed to sign with a team.
Overview: Point Guards & Shooting Guards
Overview: Small Forwards & Power Forwards
Overview: Centers & Restricted Free Agents
Analysis: Free Agent Updates 1
Analysis: Free Agent Updates 2
The remaining pool of unrestricted free agents isn't pretty. Here are the bulk of them, with restricted FAs excluded:
Point Guards/Shooting Guards: Mo Williams, Ramon Sessions, Ray Allen, E'Twaun Moore, MarShon Brooks, Toure' Murry, Toney Douglas, Jamaal Tinsley, Othyus Jeffers, Jared Cunningham...
Swingmen/Small Forwards: Shawn Marion, Al-Farouq Aminu, Dahntay Jones, Metta World Peace, Michael Beasley, Francisco Garcia, Rasual Butler, Chris Douglas-Roberts, Hedo Turkoglu...
Power Forwards/Centers: Andray Blatche, Elton Brand, Earl Clark, Dante Cunningham, Jermaine O'Neal, Greg Stiemsma, Byron Mullens, Greg Oden, Emeka Okafor, Nazr Mohammed, Ekpe Udoh, Gustavo Ayon, Ryan Hollins, Tyrus Thomas, Kenyon Martin, Charlie Villanueva, Antawn Jamison, Al Harrington, Ivan Johnson, Marcus Camby, Andris Biedrins and Andrew Bynum...
That concludes this edition of 'The Numbers Game.' Check back next week for what I hope will be a companion piece to today's column, with a focus on team defenses.