Things of note in the Avon police blotter – Browns Friday Fumble
October 23, 2015Smite the Knights: Ohio State – Rutgers Preview
October 23, 2015The 2015 NBA Finals between the Cleveland Cavaliers and Golden State Warriors was fascinating for a variety of reasons. We (America/the World/basketball-watching vertebrates) had the reigning best basketball player in the world (LeBron James) facing the reigning NBA’s Most Valuable Player (Stephen Curry), who both double as two of the most popular sports personalities in the world.1 We had two teams in historic title droughts (the Warriors’ previous championship was 1975, the Cavs previous championship was never) vying for the NBA’s greatest prize.
We also had two star-laden teams, we had two rookie head coaches, we had (at their best) two of the most exciting teams in the NBA, we had multiple players capable of mesmerizing basketball-related feats, we had a veteran playing their career’s best basketball on its biggest stage (Andre Iguodala), third-rate guards captivating the nation’s imagination (Matthew Dellavedova), we had J.R. Smith arriving for games on a remote-controlled scooter which was presumably an impulse buy from an airport Brookstone, we had ABC seeing how many Pixels promos it could run before I was tempted to take my own life, we had LeBron James attempting to win a title for his desperate home region, we had the Warriors attempting to cement itself as one of the best teams in NBA history, and we had the Cavaliers trying to overcome some historically cruel injuries. It was fun.
If the Warriors had the offense of the Space Age, the Cavs had the offense of the Industrial Revolution. Offensive determinism versus basketball barbarism. The Finals was more than a battle of two teams. It was a clash of ideologies.
The obvious question that follows is, Does ideology even matter? There’s a school of fans who assume a team with LeBron James need not even have an offensive philosophy; that the Stephen Currys and LeBrons and Michael Jordans propel their teams to victory on sheer talent. Yet some wacky coaches and GMs insist on implementing their player talent according to a specific design. But are successful teams more likely to deploy a certain offensive style? Does winning basketball games require winning philosophies? Can we model the league’s offensive styles in a way that confirms what is seen by our eyes which so frequently deceive us? Does a coherent basketball identity do anything other than help coaches sleep at night? Are teams trapped in existential crises trapped at the bottom of the standings?
These are important questions, people; not just about basketball, but about life. Because how you feel about midrange jumpers and corner threes tells me as much about you as who you vote for or what you do for a living.
Modeling the NBA’s Offenses
To help answer some of the questions about the importance of philosophy and style in basketball, I attempted to diagram NBA offenses in two-dimensional space in a way that was moderately useful. There weren’t a lot of rules about the model, but I knew it had to be simple, visually digestible, and the Charlotte Hornets shouldn’t be anywhere near the Golden State Warriors.
There are a lot of ways to evaluate NBA offenses. Scoring, obviously, is an important basketball measurable. But points scored don’t tell us much about the manner in which that scoring is done. Other ways to numerically evaluate NBA offenses are speed of play, passing frequency, rebounding frequency, types of rebounds, passing quality, efficiency, shooting distance, type of shots taken, who’s doing the shooting, play type frequency, and many others. Some of these are easy to statistically evaluate, some are challenging. Data on some of these qualities has been available since basketball games had scores. Data on others has only recently become available due to advances technology and the generosity of those willing to share it (mostly the NBA itself), and data on the rest is either not yet possible or only possible through great effort.
The most logical starting point for a model of NBA offenses is pace, or the speed of play, best expressed by the number of possessions in a period of time.2 Pace is something that a casual spectator is likely to notice. Even the most uneducated basketball fan notices when one team runs a lot more than the other. It doesn’t take much to see that Magic’s Lakers play a different brand of basketball than the Wisconsin Badgers.
Furthermore, a lot of offensive philosophies are predicated on pace, and it’s such an important part of basketball that advanced stats usually normalize statistics based on the pace of play — using points per possession instead of points per game. Teams that want to run run a lot, and therefore have a lot of possessions. Teams that want to slow the game down run less, and tend to have less possessions. Pace has problems,3 but it’s still the most obvious place to start and demands inclusion in any model differentiating between NBA offensive styles.
Isn’t this kind of … arbitrary? Absolutely. I was trying to find a way to quantitatively express how I felt — and how many other basketball fans feel — about the league’s diverse offenses. So it’s largely an ad hoc, capricious model. However, the model still accomplishes the following: it helps categorize NBA offensive philosophies based on what we know from watching basketball, but also relying on numbers independent of human judgment (pace and Style Points).
The style model also has the following attributes: 1. It’s is somewhat intuitive (if flawed);4 2. It’s relatively simple and not overly convoluted; and 3. It’s completely transparent. There’s no behind the scenes noodling, no shoulder-shrugging over parameters that my supposedly omnipotent proprietary software regurgitated, and no assurances that the details are something you don’t need to worry about it. In that way, it’s more accessible than things like ESPN’s RPM or other advanced stats. It’s a little arbitrary and imperfect, but keep in mind that assists are still tracked by scorekeepers: fallible human beings who may or may not be eating nachos during the game.
That’s how I set about finding philosophically significant numbers, so that we can categorize every NBA team by offensive philosophy. In most cases, it’s fairly accurate and intellectually rewarding. In few cases, it’s slightly misleading and harder to explain or rationalize. However, this may say as much about the teams without an offensive identity as it does the model.
NBA Offensive Styles (It’s like Vogue, but with fewer makeup ads and more Gregg Popovich)
Pace and Space
Teams: Boston Celtics, Houston Rockets, Golden State Warriors
Criteria: More than 85 Style Points, Pace over 97 possessions per 48 minutes
“Pace and space” basketball is the current trend in the NBA — many of the “smart” teams are attempting to play this way. These offenses emphasize playing fast, uptempo basketball, spacing the floor with shooters, and using rapid ball movement to find the best available shot, which is often a three-pointer. They keep defenses covering as much of the court as possible, thus creating uninterrupted, high percentage shots near the hoop. This is the modern style of basketball, and some of the league’s best teams and offenses have fully embraced the pace-and-space philosophy. Foot-stomping calculator-phobes would say pace and space play is merely a fad, and other philosophies are just as or more effective. Other skeptics would argue you need exceptional personnel to duplicate this success, or a freak of nature like Stephen Curry is required to dominate with this style. Will more team start drinking the Kool-Aid? The Warriors and the Rockets were both top-five scoring offenses, and both attempted a bonkers 27 three-pointers per game. The pace-and-space teams attempted five more three-pointers per game than the league average. The Boston Celtics are a great test case to determine if a team can will its way to success using the pace and space style.
Sharing Is Caring
Teams: Atlanta Hawks, Cleveland Cavaliers, Los Angeles Clippers, San Antonio Spurs
Criteria: More than 85 Style Points, Pace below 97 possessions
These teams are slower pace-and-space teams, and you wouldn’t be wrong to characterize some of them (especially the Hawks and Clippers) as such. An alternative way to describe the sharing is caring teams is “best available shot” teams, where “best” means “most efficient.” These teams use the three-point shot as a weapon, but don’t rely on tempo as much to create their open shots. Like, what’s the hurry anyway? It makes sense that the Cavaliers would end up here, as I think the best examples of these teams are LeBron-fense teams — teams with deliberate, methodical offenses, that will either penetrate the lane using (often slow-developing or multiple) pick-and-rolls or empowering players in the post (such as Tim Duncan and Blake Griffin) as means to create open three-point shots. LeBron’s Miami Heat teams from 2012-14 likely fit this style. All the teams in this category were top top-seven offenses from an efficiency standpoint last season.5 As a group, they average more than 25 three-point attempts per game. The Hawks, Clippers, and Spurs all average over 57 points per game created by assists (trailing only the Warriors).
Run and Gun
Teams: Dallas Mavericks, Denver Nuggets, Philadelphia 76ers
Criteria: Between 75 and 85 Style Points, Pace over 97 possessions
These teams play with pace, but don’t loiter at the perimeter or in the paint (effectively, anyway). If these teams had better shooters or ball movement, they may have been pace-and-space teams. They want easy baskets (but not as badly as the next category of teams does). All three run-and-gun teams were top-ten in offensive transition possessions and top-eight in number of drives per game, according NBA.com’s tracking data. The Nuggets and 76ers were bad.
Fast Break
Teams: Phoenix Suns, Oklahoma City Thunder, Sacramento Kings
Criteria: Fewer than 75 Style Points, Pace below 97 possessions
These teams want to go fast but aren’t necessarily known for their shooting or stellar passing. The Kings’ inclusion here is a little surprising, given that fast break specialist George Karl didn’t show up until later in the season. When imagining one of these teams, one has visions of Russell Westbrook taking the ball the length of the court in .5 seconds and trying to inflict as much harm on the rim as humanly possible. All three of these teams were in the top half of the league in points in the paint, and the Suns and Thunder trailed only the Warriors and Rockets in fast break points. All three were in the bottom five in assist rate. The Kings and Thunder attempted over 24 free throws per game. I anticipate that Billy Donovan and Kevin Durant will have the Thunder in a different category this season.
Ball Control
Teams: Miami Heat, New Orleans Pelicans, Miami Heat, Detroit Pistons, Indiana Pacers, Chicago Bulls, Washington Wizards, Milwaukee Bucks, Portland Trail Blazers, Toronto Raptors, Utah Jazz
Criteria: Between 75 and 85 Style Points, Pace below 97 possessions
This is the most crowded category, for a variety of reasons. Most of these teams have indistinct offensive identities. The teams here are some of the best defenses (the Bucks, Wizards, Pacers, and the Jazz in the second half of the season) in the league and worst offenses (the Knicks, Bucks, Pacers, and Heat). If you want to discern more about these teams’ offensive philosophies, you’re going to want to look more closely at what they do with the ball. The good defenses may lack explosive offensive personnel or emphasize floor balance over offensive rebounding. Five of the teams in this category obtain more than 19 percent of their points from midrange jumpers (the Knicks and Wizards being the biggest culprits) and six of these teams rank in the bottom third of the league in fast break points. The ball control teams tend to yield few points off turnovers and play decent transition defense: half of the league’s best 10 teams at preventing fast break points are in this category, and none of are in the bottom third of fast break defenses.
Ground and Pound
Teams: Brooklyn Nets, Charlotte Hornets, Los Angeles Lakers, Minnesota Timberwolves, Memphis Grizzlies, Orlando Magic
Criteria: Fewer than 75 Style Points, Pace below 97 possessions
These are the teams who want to slow down games, give it to their big guys in the paint, and turn games into a ground war. It’s understandable that some of these teams would want to depress tempo, given their best scorers feel at home on the block (Al Jefferson on the Hornets, Brook Lopez on the Nets, Nikola Vucevic on the Magic, Zach Randolph and Marc Gasol on the Grizzlies). The Lakers are here, who are coached by Byron Scott, renowned grouch and hater of three-pointers. Do we attribute some of this to an ineffective philosophy? All of these teams are in the bottom third of the NBA for three-point attempts. These offenses are also generally bad: only the Grizzlies are in the top half of the league in offensive efficiency. The Grizzlies were also the only team to win more than 38 games. Do we attribute some of this to an ineffective philosophy? All the ground and pound teams rank in the bottom half of the league in fast break points. Also, they are all in the top half of the league in post-up frequency, with the exception of the Magic, who were more drive-oriented last season. There are an offensive amount of midrange jumpers going on here as well.
Questions, questions, and more questions
This is a strange exercise, but I think the preceding explanations are mostly accurate and largely reassuring. As is to be expected, teams adhere to the generalizations to varying degrees and most incorporate elements of all styles in their offenses — this is just the resulting composite picture. Some teams don’t seem to fit where they’re classified, but most teams do. Some of the questions raised by the groupings say more about those teams and their offensive philosophies (or lack of) than they do flaws about the model.
Let’s take one team as an example. Why are the Kings one of the most frequent post-up teams in the league, but also a team that plays at a relatively brisk pace at over 97 possessions per game, and will probably only try to play faster with Karl at the helm? Is pace forced on them by Pacific Division opponents like the Warriors, Suns, and Clippers (who actually play a slower pace than the Kings)? Is this a deliberate style adopted by the Kings, or are they losing the ideological battle? How does Rajon Rondo fit into this? If they want to play with pace, is DeMarcus Cousins, a dominant post presence, compatible with that choice?
The chart above also raises some interesting questions about league trends. Is the success of the pace and space teams a result of the style they play, or is the style a byproduct of their talent? Will the Boston Celtics make a leap toward 50 wins by imitating the pace and space teams or sharing is caring teams that have prospered? Are difficulties listing distinguishing characteristics for some teams a failure on my part (or the model’s part), or a failure by the teams that never established a well-defined offensive identity? Can teams like the Wizards, Bulls, and Raptors be taken seriously as contenders unless they develop more modern offenses? Can the Heat compete in the East at the lollygagging pace they play at, and will Goran Dragic quicken their pace? With Memphis the only team thriving in the ground and pound category, what the hell are the other teams doing there? Is a ground and pound style a good strategy to have in the modern NBA? I think the chart concisely raises all these issues — illustrating many questions NBA fans, general managers, and coaches are asking themselves all the time.
Putting This Nonsense to Use: What happened when the Cavs played like the Cavs, and what happened when they played like the Warriors
Using the Cavs as a case study demonstrates that there may be some actual value to using this crude model to classify teams based on offensive philosophy.
As stated above, I classified the Cavs as a “sharing is caring” team, in the same category as the Spurs, Clippers, and Hawks. The Cavs play at a slow pace, and literally the slowest team in the league in average speed. They don’t move the ball the best or generate assists at the highest clip: but they attempt a ton of threes and knock down their shots at an impressive rate. This is the value in surrounding LeBron James with shooters. The Cavs also had an embarrassing number of isolation possessions (leading the league with 993 isos), but were able to succeed doing so because James, Kyrie Irving, and even J.R. Smith are exceptional iso players. The Cavs are primarily offensive opportunists. But it’s possible their success may be anomalous because of their rich offensive talent, and less because of their stylistic tendencies or offensive ideology.
My theory was that the Cavs were damn good when they played like the Cavs — whenever they practiced their patient, don’t-shoot-til-you-see-the-whites-of-their-eyes offense. I also suspected that they’re even better when they played a pace and space style — when they played more like the Warriors or Hawks. After all, I spilled more than a few thousand words urging the Cavs to amp up their fast break offense last season.
By charting every Cavaliers game on this imaginary philosophical coordinate system, we see the Cavs’ wins and losses from last season and how closely they played “like themselves” in each contest. With a bit of geometry we can also find their record in contests in which they conformed to their offensive style. When the Cavs played their average style of basketball,6 they were 15-8, which projects to a season win total of 53.47 games — basically their resulting win total from last season. Not bad.
What if they played like someone else? Were the Cavs better when they played like the Warriors? What about the Grizzlies? When the Cavs played Warriors-style basketball, they went 8-4 (which projects to 54-55 wins). They also went 11-5 playing like the Hawks (56-57 wins) and 18-6 playing like the manic Westbrook-led Thunder (61-62 wins). On the other hand, they were a mediocre 13-14 when playing like the New York Knicks (39-40 wins) and an egregious 9-15 when playing like the Grizzlies (30-31 wins). It’s starting to look like style matters.
The table below summarizes the Cavs record for each stylistic category for the model. The Cavs were 10-2 when playing like a pace and space team and 25-7 when playing their sharing is caring style, but only 9-8 when playing a ground and pound style. As far as pace alone goes, they were 13-5 in games playing at pace faster than 98 possessions/48 minutes (.722 win percentage). The Cavs were 13-4 when their Style Points exceeded 110, which prorates to 61-62 wins. These figures suggest that the Cavs should play faster, and foster an offense philosophy predicated on ball movement and opportunistic three-point shooting. How David Blatt exercises whatever stylistic influence he has on the offense, how the Cavs offense evolves with LeBron James, Kevin Love, and Kyrie Irving doing the basketball-ing, how the team integrates new pieces like Mo Williams and Richard Jefferson, and how much the team adheres to or drifts from their tendencies last season will be interesting to follow throughout the year. With some tweaks, they may be able to improve from what was already the fourth most efficient offense last season.
Play With Style
So did this exercise accomplish anything beyond wasting an inordinate amount of my and your time, and convincing myself that there’s any order in the universe? It appears that the model may be useful for ongoing analysis throughout the season. A model that quantitatively evaluates something abstract like “style” or tries to put numbers on a team’s offensive sensibilities may be a futile endeavor. But it’s probably no more flawed than an evaluation based solely on how many points a team scores or how many assists it earns. It’s crude, but effective, and suggests that offensive philosophies matter, and that the best way to succeed in the NBA is to marry talent, personnel fit, and philosophy in a winning combination. That or continue to finish near the bottom of the standings — and losing has never been in style.
- #Brands! [↩]
- Predominantly per 48 minutes, the length of a regulation NBA game. Pace measured this way is also nice in that it produces a friendly, non-threatening number in the neighborhood of 80-110. [↩]
- For example, pace is not measured or calculated uniformly, and numbers may vary from source to source. I used NBA.com’s numbers for the 2014-15 season, which can be found here. Pace is also dependent on other teams, in that a team is going to have more possessions when it plays a “fast” team whether it wants them or not. For that reason, teams in the Western Conference have slightly inflated pace numbers, because more teams want to play uptempo basketball than their Eastern Conference peers. (The Western Conference teams average 96.8 possessions per 48 minutes, while the Eastern Conference teams average 95.5.) [↩]
- For example, if I were to do the same thing with college football, I would probably put the number of receivers per play on one axis (to reflect formation/personnel), and some calculation involving pass play percentage on the other axis. You’d have a Mike Leach offense in one corner, and the Georgia Tech triple option in another corner, and everything else scattered around them. Kind of intuitive, right? [↩]
- Based on NBA.com’s offensive ratings, which represent points per 100 possessions. [↩]
- Defined as any game whose stylistic data point fell within a 10 unit radius of their season average. [↩]
6 Comments
Power packed story Kyle seriously well done. I look back at the finals and even though the Cavaliers lost 4-2 the series as well all know was much closer. And that’s remarkable given just how banged up and undermanned the Cavs were vs the Warriors. Point of fact the Cavs were beat up all season it just went to an unbelievably horrible level in the playoffs. Cue the changes made to the bench. I also think David Blatt probably learned a lot his rookie season. We won’t have long to find out. Cavs face Chicago, Memphis and then Miami to in four days and besides missing KI & Shumpert there’s a real chance, I hope not, LBJ may not play in one or more of those games.
Do you want to know something really interesting that is worth paying your attention right now,a fabulous online opportunity to work for those people who want to use their free time so that they can make some extra money using their computers… I have been working on this for last two and half years and I am making 60-90 dollar/ hour … In the past week I have earned 13,70 dollars for almost 20 hours sitting ….
Degree ,Specific qualification or skills is not necessary for this, just keyboard typing and a good working and reliable internet connection ….
Boundations or time limitations to start work is not required … You may do this work at any time when you willing to do it ….
Do you want to know how I have been doing this?…..….see this {Iink} on my !|profile|!` to know how I am working` on this`
gfvgfg
SAC is a mess. Cousins is one of the slowest big men but that crazy owner wants them to play fast pace.
The Cavs basically tanked those 2 games at year’s end vs. BOS so they’d easily be 12-0 vs. pace/space by your metrics
Now that Blatt’s seen the league, I don’t think the style will be so much an issue vs. timing. Trying not to be lame about it, but from Mozgov trade until the point they rested starters (i.e. Boston tank jobs), Cavs lost 2 games that weren’t back-end of back2backs. The wacky HOU overtime game and some game vs. the Nets.
Personally, I think the games we are at most risk are games vs. a fast scoring PG that abuses Kyrie and gets the Cavs outta their defensive sets.
I Don’t know style Points. What is style Points?