The analytics movement in basketball is a perpetual motion machine. Cycles of statistical insight, implementation, analysis, reflection and adjustment are being filtered through games on a nightly basis. Although we might not be able to see this process across a Wednesday night double-header on ESPN, all you have to do is dial up a game from five seasons ago on YouTube to know that the league is changing, fast. These changes are being pushed on the public side by academia, media outlets and a legion of amateur analysts and researchers. Of course, all of that pales in comparison to the proprietary work that is being done at the behest of individual teams behind closed doors.
With all of this motion, not all in the same direction, it’s easy to get lost or confused about what is known, what is unknown and what is assumed. One of our core values here at The Nylon Calculus is accessibility, not just of our content but of the guiding ideas of basketball analytics. What We Know is a series that aims to do just that. We want to press pause, take a deep breath, and recap the ground that has been sprinted over in the past few seasons. This is not about formulas or specific statistics, this is about the big ideas that statistical analysis has brought to the table.
One of the biggest and earliest ideas put forth by basketball analytics is the concept of efficiency. Although this had been well-ingrained in certain coaching trees for years, it was Dean Oliver who really pushed this idea into the popular consciousness with his 2004 book, Basketball on Paper.
Efficiency itself is not a new idea, chances are you spend a good portion of your day chasing it. If you have seven errands to run at locations all around town, you probably spend a minute or two thinking of the quickest possible route between those seven locations, the one that will waste the least of your precious free time. Efficiency is about the cost of things — trying to find a way to spend the least while gaining the biggest return.
In basketball efficiency is all wrapped around the nugget of a possession. Each time a team gets the ball, a new possession begins. A possession can end with a turnover, a trip to the free throw line or a shot attempt (the majority of statistical systems treat an offensive rebound as a continuation of the previous possession as opposed to a new possession). Possessions are the currency each team gets to spend in the pursuit of points. By looking at different statistics in the context of possessions we can see not just what teams accumulated, but what the cost was.
For an example of this idea, compare Dirk Nowitzki and Paul George. Both played in 80 games this season and their regular season point totals were separated by just two points — 1,737 points for George, 1,735 for Nowitzki. But Nowitzki is a much better shooter and turns the ball over far less often, meaning he wastes less possessions. If we look at the shot attempts, turnovers and trips to the free throw line we see a fairly wide chasm in the cost of their points — 1,790 possessions for George to 1,555 possessions for Nowitzki. Each player provided their team with essentially the same quantity of points throughout the season but it took George 235 extra possessions to get there, three extra possessions per game.
We can see the same phenomenon at the team level as well. Last season the Houston Rockets’ offense scored nearly 500 more points than the Heat’s offense — 8829 to 8380. Per game that works out to averages 107.7 for the Rockets and 102.2 for the Heat, a difference of more than five points. Simply looking at those quantities would leave you with the impression that the Rockets’ offense was far superior to the Heat’s. But the difference is that the Rockets played an aggressively uptempo game, looking to shoot early in the shot clock before the defense was set. Playing at this pace meant there were more possessions in each game, more opportunities for shots, turnovers and trips to the free throw line — essentially more currency for them to spend. If we take the point totals scored by each team and calculate how many they scored for each 100 possessions they used we’d see that the two offenses were separated by just a tenth of a point — 111.0 points per 100 possessions for the Rockets, 110.9 for the Heat. In terms of cost, the Heat offense generated the same amount of offense for their possessions, the Rockets just played in a way that ensured they had a lot more currency to spend.
This idea of efficiency is incorporated into a host of statistics, from team Offensive Rating and Defensive Rating (points scored and allowed per 100 possessions) to player statistics like Turnover Rate (turnovers per 100 possessions). But what we’re really focused on here is how we see this idea of efficiency playing out in front of us, and the evidence is everywhere.
Efficiency is the reason three-point shots are increasing league-wide (a chance at three-points on a possession instead of two), particularly the ubiquitous corner-three which is slightly closer than other shots beyond the arc, and thus marginally more efficient. It’s also the reason more and more teams are placing value on offensive players who shoot a high percentage from the field and don’t turn the ball over. A focus on efficiency is why more teams are playing at a faster pace offensively (better scoring odds against a defense that isn’t set). It’s one of the reasons why you hear less and less about players “who can get their own shot” as teams recognize that all but an extreme handful of players are less efficient in an isolation situation than a well-run offensive set. It’s why defenses are increasingly sagging into the paint and inviting opponents to take mid-range jumpshots.
The truth is most of the analytics movement at this point is about finding advantages on the margin, small holes that can be exploited. Many of the best teams in today’s NBA adhere strongly to a specific process. Although things might night work out in their favor in one specific circumstance, these systems are set up to maximize the advantages there team has available to them over the long run. Often these narrow advantages make little to no difference in a single quarter or even across an entire game, but stretched across an entire season they can be the difference between the lottery and the playoffs, great and just very good. Efficiency is about weighing both quantity and quality, recognizing that results are only important in the context of what those results cost.
Ultimately, almost everything that’s going on with basketball analytics can be traced back to the hunt for efficiency, squeezing the most out of what you have.