Ari Caroline (@aricaroline), Director of Quantitative Analysis and Strategic Initiatives at Memorial Sloan-Kettering Cancer Center, returns to his amazing work in analyzing how NBA players get playing time using the lens of Wins Produced.
The Center is Off-Balance
If you recall this work started examining how teams misplayed centers. Astute readers made the point that minutes per-game was a better measurement than total minutes. Here’s the chart remade that way.
A few players shift around but the same problem remains – some really good centers are underplayed and some really bad centers are overplayed — no such luck for Chris Kaman getting a reprieve.
Here’s an updated rundown:
- Marcus Camby
- Chris Andersen
- Kosta Koufos
- Omer Asik
- Ben Wallace
- Tiago Splitter
- Greg Stiemsma
- Samuel Dalembert
- Andrea Bargnani (bad)
- Chris Kaman (bad)
- Andray Blatche (bad)
- Mehmet Okur (bad)
A Better Shot with Shooting Guards
Let’s break down the same problem with Shooting Guards.
Good news! In terms of how shooting guards are played, the NBA is much better. Let’s look at the outliers.
There’s even better news. When it comes to bad shooting guards getting too many minutes, only Jordan Crawford shows up — a fun note, Cory Higgins resulted in the negative scale getting stretched even further. In terms of good shooting guards not getting enough minutes we’re left with James Jones, Mike Miller, Tracy McGrady and Thabo Sefolosha. Also, where did Brandon Rush come from this season?
Staying on point with Point Guards
Finally we get to point guards. The relationship between playing time and Wins Produced is now at its strongest. What’s more, there are no outliers! Basically, it would appear the skills valued by Wins Produced are valued by NBA decision makers, when it comes to point guards! In fact, over half of playing time in point guards can be explained by Wins Produced. Compare this with the noticeably absent power forwards, which have no statistical significance in regards to Wins Produced.
Challenges to the Reader
Without cheating, which of these four statistics significantly explains variations in playing time for each position? In other words, what do coaches seem to value the most position by position? Bonus points for getting the sequence right as well.
- Points per 48
- Rebounds per 48
- Assists per 48
- WP48 (acknowledging that this one, with some twists, is somewhat inclusive of the three above)
Also, we’d love to hear your thoughts on why the NBA seems to have figured out the point guard and still needs help on the rest.
Coming Next: More Answers!