‘Tis the season for NBA writers (and fans) to begin the annual discussion of who could/would/should make the NBA All Star Game. Which brings us to Dr. Dick’s Mid-Season Analysis of Player Productivity.

So today I will first briefly explain my methodology, then create All Star teams for each conference. The NBA currently has 12 member All Star teams, but this antiquated number does not reflect the growth in the number of teams, nor the fact that individual teams can now carry 15 player, thus the All Star teams announced in this space will be comprised of 15 players for each conference. So, without further ado, here is the methodology:

To be included in the analysis a player must play approximately 20 minutes per all games his team has played. So for this Mid-Season Analysis 800 minutes played is the threshold. Players just missing this somewhat arbitrary cut off are Andrew Bynum (798), Lamar Odom (796) Antonio McDyess (794), and Fabricio Oberto (792). Perhaps they will play enough/avoid injury and be included in the year end analysis.

After identifying the 173 players qualified to be ranked, statistics in 10 categories are considered. A quick look at the ten categories:

- Minutes Played: Captures durability and coach’s willingness to play the player.
- FG Percentage: Proxy for accuracy and shot selection.
- FT Percentage: Can he make the freebies?
- Total Rebounds: Each represents a posession (whether offensive or defensive).
- Blocks+Steals: Only tangible measures of defensive contribution.
- Three pointers made: Indicates an ability to spread the defense.
- Points per shot: A measure of offensive efficiency.
- Turnovers (negative weighting): Each represents a posession lost.
- Points: The way most GMs, announcers and fans value players.
- Assists: Most nebulous in terms of direct contribution. But NBA types track it, so we’ll work under the assumption that they impart something positive to a team.

For each of these categories, the average is calculated for the 173 players being evaluated. The standard deviation is also determined. (Briefly, a large standard deviation indicates that the data points are far from the average and a small standard deviation indicates that they are clustered closely around the average. Follow this link for a more in depth explantation and the formula).

Then for each player in each category the average is subtracted from his production and then divided by the standard deviation, resulting in a z-score. Here is an illustration using LeBon James and Minues Played.

(Lebron’s Minutes – Average Minutes) / Standard Deviation = Z Score

(1587 -1142.2) /214.8 = 2.070

Translation : LeBron plays minutes that are two standard deviations above the average for the 173 players. A score of zero means the player is right at the average for the measure. In this group, Samuel Dalembert played 1142 minutes, the average was 1142.2, so his Minutes Z-Score is -.001, as the negative sign indicates he is juuuuuust *below *the average.

This same calculation is completed for each player in each of the ten categories. The results are then summed for each player to arrive at his overall Score. Continuing with LeBron his z-scores look like this:

- Minutes Played: 2.070
- FG Percentage: .165
- FT Percentage: -.415
- Total Rebounds: .741
- Blocks+Steals: 1.255
- Three pointers made: .760
- Points per shot: .485
- Turnovers (negative weighting): -2.318
- Points: 2.738
- Assists: 1.684

Looking at these ten scores, it is apparent that LeBron is better than the other 800+ minute players in eight of the categories (positive values) and below average in two maeasures – Turnovers and FT Percentage (negative values). It can also be noted that in two categories he is two standard deviations above the average player. This means he is better than 95% of players on each of those measures.

Adding all of these scores together, LeBron’s overall Score is 7.166. In the next post, we’ll see where he ranks among his peers so far this season.

## Leave a Reply