January 12, 2017 by J.H. Yeh
Scoring is the best way to see a player’s offensive prowess. Below is a list of the top 30 scorers in the NBA as of January 9, 2016. In this list, it’s easy to see Russell Westbrook is leading the league in scoring with average of 31.2 points per game. Now dig in a little deeper and we can see Westbrook is shooting 42.7% overall from the field and needs 23.8 shots and 10.5 free throws every night to get to 31.2 points. On the other hand, LeBron James only needs 18.6 shot attempts and 7.1 free throws to get to 26.1 points per game. It’s quite obvious that James is the more efficient player. Incidentally, the inclusion of the free throws validates what true shooting percentage reflects as it takes free throws into consideration when calculating how many total shots it takes to get to certain points. Click here for more details.
Efficiency is the true measure that determines a player’s value to the team. It is also the main factor that influences how teams value each player differently based on how efficient their performances are. There are various ways to produce points. If we go back to some old game logs, we can see that in the first round matchup between the Golden State Warriors and the Portland Trailblazers in last year’s playoffs, Damien Lillard needed 18 3-point shots and a total of 30 shots to get to 36 points. He was 9 for 30 and shooting 30.0% from the field. On the other hand, Stephan Curry came off the bench and scored 40 points on a 16-32 shooting performance. Similarly, on January 6, 2005, in a matchup between the Memphis Grizzlies and the Detroit Pistons, Richard Hamilton led the low-scoring Pistons with 14 points. He achieved that by shooting 0 for 10 from the field and making all his 14 free throw attempts. Earlier this season, James poured in 44 points against the Charlotte Hornets. He did that with just 24 shots as he put on an incredible shooting display by going 17 for 24 and making 70.8% of his shots. And the results? The more efficient sides won all those games.
Firstly, to measure a player’s offensive efficiency, we need to find out the points generated (PGEN) when the player terminates a scoring possession by either directly scoring or making a pass that quickly leads to points scored. PGEN is the amount of points a player produces each time he or she terminates a scoring possession. Points generated are calculated from points a player scored and the number of passes that leads to points being put on the score board.
Points per assist (PPA) is a good metric to see how many points a player generates per assist. It is the total points a player has assisted divided by the number of assists by that player. For example, if Kyrie Irving has 6 assists on two 2-point field goals and four 3-point field goals, then his PPA, or points per assist, will be 2.67 points. [( 2 x 2 ) + ( 4 x 3 )] / 6 = ~2.67
A PPA of 2.67 is pretty high. It indicates most of these assisted points come from 3-point field goals. Also note that PPA will always be between 2 and 3. Now we have PPA, we can define PGEN, or points generated, as follows:
PGEN = Pts + ((Assisted Points / Player’s Tot Ast) * ( Ast + 2nd Ast + FTAst ))
PGEN = Pts + PPA x ( Ast + 2nd Ast + FTAst )
2nd Ast is secondary assist. It is the pass that comes before an actual assist. FT Ast is free throw assist. It is the pass that leads to a player who is immediately fouled while attempting a shot and makes at least one free throw. And-1s are not included in FT Ast. According to the NBA, a free throw assist is recorded when the receiving player is fouled in the act of shooting within 4 seconds and 2 dribbles after receiving the pass and makes at least one free throw. Both sub-types of assists measure effective offense that are inevitably left out by the traditional assist. Click here for the article about these assists. Below is a list of all the top scorers’ field goal attempts, assists, secondary assists, and free throw assists per game. The ranking is sorted out accordingly by scoring average. All the traditional stats in this article are courtesy of the player tracking provided by NBA.com and are as of January 9, 2017.
With all these numbers lined up, we can compute the PGEN. Below is the list of the points generated by each player per game. Sorted by PGEN, we can see James Harden leads the top 30 scorers in points generated every time he terminates a scoring possession. This is not surprising consider that all the Rockets players are shooting 3-point shots. The Rockets are attempting an unprecedented 46.0% of their shots from beyond the 3-point arc, which subsequently contribute to Harden’s high PGEN.
PGEN may seem unfair to some players such as Jimmy Butler, who has teammates who are struggling to shoot. In fact, any time when a metric based on any type of assists is largely dependent on the quality of teammates and can put some players at a disadvantage as certain players have better teammates. However, at the same time, teammates also often depend on the player to be able to produce those assists and point scored. For example, Tyson Chandler largely rely on Eric Bledsoe and Devin Booker’s scoring and playmaking abilities to score as the two players can drive and attract defense towards them and leave a wide open Chandler to score from an uncontested dunk. As a result, the player depends on teammates to produce assists, points, and good PGEN, and teammates also depend on the player to produce in those same categories. For another word, a player needs good teammates and teammates also need a good player in order for them to mutually benefit from each other. Generally, guards tend to have higher PGEN because of their playmaking ability.
Next, we will look into net possessions terminated (NPT). PGEN only includes scoring possessions, which are actions that lead to immediate scoring such as points and free throw assists. On a broader spectrum, NPT takes into account of all the actions that end a possession. There are five ways for a possession to end: 1) field goal attempts, 2) assist of any type, 3) shooting fouls, 4) offensive rebounds, and 5) turnovers. Below is the definition for NPT:
NPT = FGA + Ast + 2ndAst + FTAst + ( 0.44 x FTA) – ORB + TO
We add all the actions that can terminate a possession together to get NPT. It is called “net” possessions terminated because offensive rebounds are subtracted. The reason to remove offensive rebounds because they extend a possession. In addition, a player can also get his or her own offensive rebounds which can further confound the result. Therefore, offensive boards are subtracted from calculating the NPT. Moreover, we use 0.44 as an approximation for the number of free throws taken by each trip to the free throw line since players go to the free throw line to shoot anywhere between one to three free throws. If a player only shoots just two free throws each time and never attempts one or three free throw shots each time he or she is fouled, then we would simply use 0.5 as the approximation. For another word, 0.44 reflects the occasional And-1 and fouled on a 3-point attempt.
Below is updated table with the NPT. The ranking is sorted according to NPT. Players’ respective scoring averages and PGENs are also included.
NPT tells us how often a player terminates a possession each time he or she touches the ball. As expected, guards tend to have higher NPT because they have more control of the ball and they pass and drive often, which lead to higher turnovers.
Now finally, to find out the individual offensive efficiency (IOE), we divide PGEN by NPT. The formula is:
IOE = PGEN / NPT
The list yields a very good look at how efficient the players are. Surprisingly (and perhaps not), Kyle Lowry comes up atop by leading all 30 scorers in individual offensive efficiency. The next four players are James Harden, Stephan Curry, Giannis Antetokounmpo, and LeBron James. Like mentioned earlier in the article, Westbrook is ranked 16 on the IOE and is considered inefficient despite his monstrous numbers. Largely this can be attributed to the fact that he is the sole offensive focal point for the Oklahoma City Thunder and regularly draws the toughest defenders on him due to his versatility.
Furthermore, like any metric and category used to evaluate a player’s performance, it is not accurate to gauge a player by looking at just one season because performance can vary each season. As stated earlier, performances are largely dependent on the teammates, and vice versa. Players have different roles and teams have different players and situations each year. As a conclusion, when evaluating a player’s overall efficiency, it is better to look at the general efficiency as opposed to the occasion elite seasons. With that said, doing so can prevent excluding established star players such as LeBron James and Kevin Durant from being considered a less valuable offensive player than Lowry. This is in no way to discount Lowry’s offensive efficiency.
The list definitely offers a better look from a very different perspective as we delve more into analyzing a player’s offensive efficiency. It is no surprise to see all three LeBron James, Kyrie Irving, and Kevin Love making the top 15 because all three players are very efficient themselves (in different ways) and are also benefiting a lot from each other and their strong teammates, which in return, reciprocate and make them better and more efficient offensively. This is also the same reason why some teams don’t value Carmelo Anthony and DeMarcus Cousins as much as They value James or Durant. On the contrary, players such as Andrew Wiggins and Harrison Barnes are in the bottom because they are young players and are inconsistent and inefficient (in elite’s standard). It is only through time and experience for them to learn how to approach the game more effectively and efficiently. Furthermore, Antetokounmpo and Karl-Anthony Towns are considered very exciting budding stars and franchise-caliber players because they mature very fast in terms of their developments and possess an overall very good instinct for the game which successfully translates to their good offensive efficiency at such young age.
— J.H. Yeh
(all tables are created by me, and all sources of stats are courtesy of NBA.com and BasketballReference.com as of January 10, 2017)