One of our missions here at Nylon Calculus is to help make basketball analytics accessible to anyone with interest. A big part of that mission is building a comprehensive and reliable glossary. My plan is to make this glossary different from some of the others around the internet in two ways. The first is that I’d like it to be comprehensive and sequential, like really, really, top-to-bottom comprehensive and sequential. You’ll see from the post below that I intend to start at the very beginning and build slowly. The thing that is often lost in discussions about basketball analytics is that each new statistic or technique is usually an adjustment to a previous model, trying to account for some hole or something that isn’t measured well in existing statistics. All basketball statistics have a lineage of mathematical models and I want to make sure that anyone with the time and interest can peruse the entire family tree.
The second thing that I believe will set our glossary apart is that I want to do more than define the statistics. For each I would also like to explain a little about what it says and what it doesn’t say — what purpose it serves in our analytic discussion.
We’re just getting started with our Glossary and there’s a lot more to build. Over the next few days we’ll be looking at some basic shooting statistics which are collected in the box score. Then we can begin moving into how we manipulate and adjust some of these statistics to compensate for their various shortcomings, trying to answer different analytic questions. I know this pace may feel too slow for some of you, or to basic for others of you. We’ll try to get to where you are as quickly as we can, but remember there are a whole slew of eager, interested basketball fans coming up this ladder right behind you.
The Basic Box Score
As a statistical measure, minutes are as straightforward as they come. Usually a player level statistic since teams play and equal number of games and, barring overtimes, each game is the same length; minutes show how long a player was actually on the floor. Since a player can’t contribute in any other measurable way unless they are on the floor, minutes essentially create the sample size for a player’s other statistical contributions.
Knowing about a player’s minutes can add context to their other statistical contributions and can sometimes be used to draw basic inferences about their performance. If someone plays a lot of minutes they are generally a good player (or at least viewed that way by their organization) and/or in some other way very important to the context of their team. For example, they may be the only player who plays at a certain position or who possesses a certain valuable skill. As with any other question of sample size, in general the more minutes a player has played, the more confident we can be that their other statistical contributions are an accurate representation of their ability to perform. Minutes are also sometimes used as a measure of durability, a player’s ability to stay healthy and avoid injury.
The most obvious and visible of basketball statistics, points are the most basic tool for measuring the offensive abilities of a player or team. You got two points for making a shot, three if it’s behind the three-point line and one if it’s from the free throw line. Although the relative importance of raw point totals as a measure of offensive impact has been largely discounted by the basketball analytics movement, they can’t be thrown out completely for the simple reason that, you still need to score more points than the other team to win a basketball game.
However, it has become increasingly obvious that looking at points alone does not adequately describe a player or team’s offensive abilities. Things like efficiency need to be taken into account, as well as how they are scored and from which locations, for the full descriptive value of points as an offensive statistic to be realized. In short, you can’t ignore points when trying to understand offense, but you can’t stop there either.
Rebounds come in two varieties — offensive and defensive. An offensive rebound is when the team that shot the ball recovers their own miss. A defensive rebound is when the defensive team recovers the miss. Each different kind of rebound is important and for different reasons. Offensive rebounds extend the possession for the offensive team, essentially wiping away a miss. They also often result with the ball in the hands of an offensive player who is fairly close to the basket, in good position to simply take another shot. The likelihood of a team scoring on an offensive rebound is much higher than the likelihood of a team scoring on any given possession.
Defensive rebounds are important because they end the possession and prevent the offense from getting a second, advantageous scoring opportunity like we mentioned above. Last season about three-quarters of the missed shots across the entire league were rebounded by the defense, so defensive rebounds are much more common than offensive rebounds. Interestingly, research has shown that height has a lot more to do with offensive rebounding than it does with defensive rebounding, highlighting the variability of rebounds at the offensive end. Also, where a shot is taken has a lot to do with how likely it is to be rebounded by either the offense or the defense.
For years, assists have been the standard statistic for measuring how well players and teams create scoring opportunities for other players. An assist, with some (occasionally abused) discretion from the scorekeeper, is awarded when a pass leads directly to a made basket. They are not recorded on passes that lead to passes that lead to made baskets (the hockey assist) or on passes that lead to shooting fouls (the free throw assist)[1. Although both of these types of assists are recorded in the NBA’s SportVU Player Tracking Statistics).
The problem with assists is that they require to actions — a successful pass and a made basket — but they award all the credit to the passer. There are many instances where an assist is literally the creation of the passer, a scoring opportunity that appears only because of their vision and ability to squeeze the ball through a tight spot. But there are also assists where the shooter does most of the work. For example, receiving a relatively simple bounce pass and proceeding to knock down a difficult fallaway three-pointer. Assists make no distinction between these two different types of plays always giving credit to the work of the passer. For that reason they can be a messy way of evaluating the passing or shot creation abilities of a player or team.
A steal, along with defensive rebounds and blocks, is one of the few defensive statistics that is collected in the basic box score and it is an incredibly valuable one. A steal is recorded when a defender takes possession of the ball from an offensive player. It ends the offensive possession without any points being scored and often puts the defense in good position to quickly score by running a fast break against disorganized defenders.
The problem with steals is that they represent a very small part of a player’s defensive ability. Deflections that disrupt a play but wind up back in the hands of the offense are not counted and, since the defensive player who ends up with the ball is credited with the steal, often the wrong player is recognized for having created the defensive play. However, steals are fairly rare occurrences compared to some of these other statistics and research has shown that the ability to generate steals has an outsized role in determining a player’s overall impact on a game.
A block occurs when a defensive player touches the ball either during an offensive player’s shooting motion or while the ball is in the air, preventing a shot or deflecting the ball away from the basket. Like steals they are an extremely important play. A blocked shot may be recovered by the offense, but it makes certain that the initial shot will not be made.
However, as a defensive statistic blocks suffer from many of the same problems as steals. A block is one outcome of good interior defense but many others aren’t counted in the basic box score. For example, times when an offensive player chooses not to shoot for fear of getting their shot blocked, or alters their shooting motion to avoid a block, making it less likely that the shot is made. Like steals, examining blocks as a measure of defense is a reasonable place to start as long as you are aware of all of the things that are being left out of the picture. Also, research has shown that the value of a blocked shot, in terms of the likelihood of the team scoring on the ensuing possession, can vary greatly depending on what type of shot it was, where it was taken from and which player blocked the shot.
Turnovers are a statistic that can be viewed offensively or defensively. A turnover is simply any offensive possession that ends without a shot attempt or a trip to the free-throw line (except when time runs out in a quarter or a half). A turnover can be because of a steal, an offensive foul, a ball dribbled or passed out of bounds, a double-dribble, a travel, a palming violation, etc. Turnovers are usually broken into two categories — live ball (steals) and dead ball (everything else). Live ball turnovers are labeled as such because play continues uninterrupted after the turnover is made. With a dead ball turnover there is a stoppage of play as the referee will need to blow their whistle to make the call. Live ball turnovers are vastly more valuable because they allow the defense to smoothly transition to the offensive end, usually against a defense that is not set. A dead ball turnover gives the defense a chance to recover into position and get set.
Fouls are often ignored as a box score statistic, other than looking at how close a player was to getting six, which leads to an ejection. But fouls can tell us quite a bit about a player. Fouls occur for a variety of reasons, usually for some sort of illegal contact, and can be charged to both offensive and defensive players.
On both ends of the floor a foul can be extremely costly. An offensive foul results in a turnover, meaning that team will have finished a possession without scoring. A defensive foul can give the offense a chance to reset or, if committed while an offensive player was in the act of shooting, result in free throws. In general, a trip to the free throw line is by far the most efficient way to use an offensive possession, almost regardless of which player is actually shooting the free throws. Fouls can also tell us quite a bit about a player’s defensive tendencies. A player who fouls a lot is usually either very aggressive and physical, or perhaps struggling with some sort of athletic disadvantage that causes them to end up in weak defensive position.