sports evaluation

Can a single efficiency stat tell the whole story?

When evaluating athletes, it’s commonly understood that more than one metric should be used to build any argument. There often isn’t a single answer to define whether one player is better than another or whether an athlete is playing well. However, in some circumstances, the simplicity of having one number to encapsulate the total value of a player can be very useful. That’s why plenty of research has gone into creating all-inclusive metrics that summarize contributions into a single statistic.

Many sports have tried to find these all-encompassing numbers with varying levels of success. For example, in baseball, the Wins Above Replacement (WAR) value is regularly used to identify how many additional wins the efforts of a single player bring to his team compared to a minimum baseline or replacement-level player in the same position.

Similar work has been done on evaluating a basketball player’s value compared to replacement players, but so far that seems to be targeted specifically towards single values, like points. However, there are three similar all-inclusive metrics that try to sum up a player’s overall value.

Three metrics to evaluate player efficiency

These combined metrics are called efficiency statistics, and they evaluate all the positive contributions a player makes on both offense and defense, while also considering the negative portions of a player’s game.

The NBA officially calculates and records a value known as efficiency (EFF), which is derived from simply adding the sum of Points, Rebounds, Assists, Steals and Blocks, then subtracting Missed Field Goals, Missed Free Throws and Turnovers. The downside is this calculated statistic skews in favor of offensive players compared to defensive players and does not take into account the number of minutes an individual plays.

To address some of the complaints about the NBA’s efficiency formula, John Hollinger, a former ESPN statistician now working for the Memphis Grizzlies, developed a metric of his own called the Player Efficiency Rating (PER). Although it still favors offensive contributions, PER factors in the minutes played, as well as the pace of the game. The values considered are weighted against each other to balance the different statistics used, so all factors are not assessed equally. The formula is calculated so that every season the league average is 15.00, which provides a good reference point to evaluate where individual players fall on the scale.

European teams use another version of an efficiency calculation created by the Spanish ACB League to help evaluate players for MVP awards. The Performance Index Rating (PIR) is not as complicated as PER but considers more statistics than the NBA’s efficiency value.

Each of these metrics has its own weaknesses and may fall short of providing a single inclusive number to reflect everything a player does on the court. But there is still plenty of value in using metrics like efficiency statistics. The simplicity of having a single value that represents a player’s worth is extremely useful for player comparisons, as well as trend analysis. They can also serve as summary numbers for more in-depth analysis into the individual numbers used to calculate the efficiency statistics.

How the NBA’s top players measure up

Here’s a look at the efficiency ratings of the top 10 players in the NBA for the 2016-17 season:

Player Name Team    PER    PIR    EFF
  Russell Westbrook   Oklahoma City Thunder   30.7   3000   33.8
  Kevin Durant   Golden State Warriors   27.68   2011   30.2
  Kawhi Leonard   San Antonia Spurs   27.62   2116   25.3
  Anthony Davis   New Orleans Pelicans   27.59   2633   31.1
  James Harden   Houston Rockets   27.43   2822   32.4
  LeBron James   Cleveland Cavaliers   27.11   2537   31
  Isaiah Thomas   Boston Celtics   26.59   2075   24.7
  Nikola Jokic   Denver Nuggets   26.4   1813   25.2
  Chris Paul   LA Clippers   26.25   1577   24.8
  Giannis Antetokounmpo   Milwaukee Bucks   26.13   2428   28.4

 

 

See what Kinduct can do for you!

Schedule a Demo