The "possession" is arguably the single most integral concept of the advanced statistics movement in basketball. Explained brilliantly in 2002’s Basketball on Paper by current Nuggets analyst Dean Oliver, a new possession essentially begins every time the ball switches hands. Its importance in quantifying the sport cannot be overstated. It is the great equalizer for players that inflate their numbers through excessive shooting, teams that amplify their point totals by running up and down the floor, defenses that suppress their conceded points by slowing the game to a crawl, and everything in between.
The commonly used estimate of team speed, pace factor, derives from the possession statistic. Team pace counts how many possessions a team uses over 48 minutes. And for the most part, it does a great job of quantifying how quickly teams play. If you were to name some of the NBA’s fastest teams off the top of your head, the Phoenix Suns, New York Knicks, Golden State Warriors and Denver Nuggets would probably come to mind. Sure enough, each of those ranks among the league’s top five teams in pace factor.
But pace factor has its shortcomings. Since the ball doesn’t switch teams on offensive rebounds, a sequence of multiple offensive boards and shot attempts simply counts as a long, drawn out possession. As Henry Abbott has noted at True Hoop, turnovers can also lead to artificially increased "pace," even in the absence of increased speed of play.
More importantly, shrewd observers will have noticed that the average NBA game is contested by two teams. If one of those teams slogs it out in the half-court on offense and the other looks to push at every opportunity, is it really fair to slap a single number on the game and call it the "pace" of both teams?
An alternate way of measuring team speed is via the shot clock. Specifically, how much time is left on the clock on average when a team attempts its field goals, and how much remains when an opponent’s attempts go up? 82games.com categorizes all team shots into four groups- 0-10 seconds (clock used), 11-15, 16-20, and 21+. Those numbers can, in turn, be used to develop an alternative method of measuring team pace -- one I’ve dubbed "Speed Index."
Speed Index sets the average league "speed" to 100, with a standard deviation of 15. The higher a team’s number, the faster it plays. Since 82games provides data in those four categories only (and not discrete shot clock values), slight deviations from these values should be expected. Overall, Speed Index provides us with some fascinating points of analysis. Importantly, it can be separated into offensive and defensive components. First up, Offensive Speed Index, or how quickly in the shot clock teams attempt shots.
For reference, an Index value of 100 corresponds to approximately 11.4 seconds left on the shot clock. One standard deviation is equivalent to about 0.65 seconds.
At first glance, the majority of the order agrees with conventional wisdom about team speed. Phoenix plays fast. New York plays fast. Boston and New Orleans tend to primarily play in the half court. We can more scientifically assess the differences between Speed Index and Pace Factor by plotting them against one another. The next graph plots the difference between team Speed Index and Pace Factor (where Pace is normalized around 100 as well). Click the image for larger text on the team axis.
Some of the main takeaways from the above graph:
· * As Henry Abbott has often pointed out, Portland’s high offensive rebounds and low turnovers mean they don’t play nearly as slowly as Pace Factor indicates. Their Speed Index value supports this notion; by Speed Index, they’re still in the bottom half of the league, but they’re much faster than their Pace lets on.
· * As fast as Phoenix is by pace factor, possessions/game may not actually capture how fast they play.
· * San Antonio ranks as the league’s most overstated "fast" team. The Spurs rank below league average in shot clock usage but above league average in pace factor. The reason for this disparity is covered next.
As noted earlier, teams play different speeds on offense and defense. The next chart is the defensive counterpart of Offensive Speed Index. How quickly do opponents shoot in the shot clock against NBA teams?
At this juncture, a number of intriguing details present themselves. First of all, while a number of elite defensive teams force opponents to shoot late in the shot clock (New Orleans, Orlando, Chicago, Milwaukee), many great defensive teams do not do this (Miami, San Antonio, Lakers, Boston). In fact, there is not a statistically significant relationship between defensive efficiency and defensive speed index.
But the low speeds of certain teams offer explanations as to their defensive footprints. The New Orleans Hornets, the league’s slowest defensive team, utilize heavy doses of zone defense and aggressive rotations to force extra passes and minimize opponent drives. This naturally extends the time in the shot clock at which opponents finally attempt shots.
The craziest data point, of course, belongs to the San Antonio Spurs. Opponents shoot earlier in the shot clock against the Spurs than against any other team in the league. It’s difficult to specify exactly why this is. The Spurs are a great defensive rebounding team and thus do not give up many put-backs early in new shot clocks. They’re about league average in fast break points allowed. They foul less often than any other team in the NBA and thus aren’t sending opponents to the line in penalty situations. Moreover, the Spurs rank in the top five in lowest field goal percentage allowed on shots taken early in the shot clock.
So what gives? Your guess is as good as mine. But the overall point is San Antonio’s increased pace this year is largely a function of defensive speed index as opposed to offensive speed index.
We can extend the question "is a team’s speed driven more by their offense or their defense?" to the rest of the league as well.
Orlando’s team speed is more driven by its defense than any other team in the league. (In other words, they’re considered a "slow" team by pace factor and speed index due entirely to their defense). On the flip side, Minnesota’s and Phoenix’s team speeds are driven entirely by their speedy offenses.
One common talking point among commentators is the idea of "controlling tempo." Measuring team speed as disparate quantities on the offensive and defensive ends allows us to more accurately determine which teams control tempo and how they do it.
Teams like the Suns try to dominate game tempo, but it's primarily an offensive phenomenon. We don't have game by game variance data, but the Suns' complete lack of defensive success suggests that while they do succeed in controlling game speed on one side of the ball, they have little say on what happens on the other. The Spurs, in many ways, appear to be the anti-Suns. For mysterious, unexplained reasons, they rapidly drive up defensive tempo while maintaining sub-league average offensive speed. Finally, a team like New Orleans dictates, on a nightly basis, the tempo on both ends of the floor.