Analysis Sunday Stat Session Series - Updating Scoring::USG Efficiency

[scott]

Aloha friends,

On the other website I used to occasionally post some statistical things I find interesting or just feel people would benefit from understanding. We've done threads on understanding USG%, on the value of pick swaps (both individually and as value enhancers), trade compensation estimators, etc.

I thought it was a good time to bring this series back, and I thought best to start with one of my favorites: looking at the leagues top scorers and their efficiency when putting their scoring in the context of their USG. Recall, USG is a measure of how often a possession ends at the hands of a player, whether by a FGA, FTA or TOV (since these are all the ways a possession can end, other than time simply running out). It's not a measure of ball dominance. A player's ability to score relative to their USG is a good measure of their efficiency. PTS/USG is a measure of how efficiently a player converts their usage into points. Increased shooting efficiency and a reduction of turnovers pushes this number higher.

We can also look at PTS per 36/USG to normalize for playing time. I've done both.

When I last did this in the summer of 2024, Lauri Markkanen stood out the most efficient scorer/minute in the league in terms of usage. His 1.012 PTS/36 per unit of USG was 0.042 points higher than the next player (Shai). That gap was roughly the same gap between Shai and the tenth most efficient scorer (Steph Curry).

Here's how some (current) Spurs ranked back then, when I only looked at the Top 100 scorers:

Wemby - 0.807 PTS/36 per USG, 37th out of 40, right behind Ant and Randle and right ahead of Cade and Paolo
Fox - 0.858 PTS/36 per USG, 26th out of 40, right behind Murray and Mitchell, right ahead of Jaylen Brown and DJM

Updating the numbers for this season, I expanded to the Top 60 scorers (for a very specific reason) and the landscape looks like this:






SGA was the 2nd most efficient player in PTS/36 per USG in the summer of 2024, and he's the #1 player now. Here are the top 10.



This certainly isn't going to slow down any lust for Saddiq Bey

Here are the least efficient scorers so far this year:



Where do Spurs fall on this list?

Wemby - 0.920 PTS/36 per USG, 18th highest, right below Brunson and Maxey, right above Lavine and Jamal Murray
Fox - 0.863 PTS/36 per USG, 42nd, right behind Bane and Quickley, right ahead of Booker and Jalen Johnson
Castle - 0.768 PTS/36 per USG, 60th, behind Derrick White and Melo Ball

Where do the rest of the Spurs fall in terms of efficiency? Well, I went ahead and ran those numbers too. Please don't extrapolate this and infer that we should just force feed these guys, because their efficiency is in part because of their roll (this should be obvious). I just think it is interesting too look at their relative efficiency compared to top scoring players.

Korndog - 1.094 PTS/36 per USG. Would lead the league
Keldog - 1.010 PTS/36 per USG. Would rank 3rd in the league!
Julian - 0.936 PTS/36 per USG. Would rank 15th in the league, right behind Dono Mitchell and Demar, ahead of Kawhi and Brunson
Barnes - 0.907 PTS/36 per USG. Would rank 25th in the league, right behind Keyonte and MPJ, right ahead of Randle and NAW
Vassell - 0.889 PTS/36 per USG. Would rank 29th in the league, right behind Pritchard and Bridges, ahead of Okongwu and KAT
Harper - 0.769 PTS/36 per USG. Would rank 60th in the league, right behind Derrick and Melo, ahead of Castle by 0.001

Some things that stand out:

• Wemby has become significant more efficient from his rookie year
• Fox is slightly more efficient from who he was with the Kings, but his USG has reduced from 31% (11th amongst the Top 40 scorers) to 25.5% (which would have ranked 3rd to last in 2023-24)
• Castle is our least efficient scorer in the regular rotation

Some other random notes

• Frequently discussed TMIII ranks 12th in PTS/36 per USG at 0.952
• Old SOB KD is 11th at 0.952, only ahead of TMIII by thousandths of a point
• Luka leads the league in scoring, but only ranks 31st amongst the Top 60 scorers in PTS/36 per USG
• Cooper Flagg is off to a decent start for his career, 0.853 PTS/36 per USG
• What a breakout year for Jalen Duren... I haven't followed the MIP discussion closely but certainly he's gotta be in the mix

 

[Area 51.247]

Aloha friends,

On the other website I used to occasionally post some statistical things I find interesting or just feel people would benefit from understanding. We've done threads on understanding USG%, on the value of pick swaps (both individually and as value enhancers), trade compensation estimators, etc.

I thought it was a good time to bring this series back, and I thought best to start with one of my favorites: looking at the leagues top scorers and their efficiency when putting their scoring in the context of their USG. Recall, USG is a measure of how often a possession ends at the hands of a player, whether by a FGA, FTA or TOV (since these are all the ways a possession can end, other than time simply running out). It's not a measure of ball dominance. A player's ability to score relative to their USG is a good measure of their efficiency. PTS/USG is a measure of how efficiently a player converts their usage into points. Increased shooting efficiency and a reduction of turnovers pushes this number higher.

We can also look at PTS per 36/USG to normalize for playing time. I've done both.

When I last did this in the summer of 2024, Lauri Markkanen stood out the most efficient scorer/minute in the league in terms of usage. His 1.012 PTS/36 per unit of USG was 0.042 points higher than the next player (Shai). That gap was roughly the same gap between Shai and the tenth most efficient scorer (Steph Curry).

Here's how some (current) Spurs ranked back then, when I only looked at the Top 100 scorers:

Wemby - 0.807 PTS/36 per USG, 37th out of 40, right behind Ant and Randle and right ahead of Cade and Paolo
Fox - 0.858 PTS/36 per USG, 26th out of 40, right behind Murray and Mitchell, right ahead of Jaylen Brown and DJM

Updating the numbers for this season, I expanded to the Top 60 scorers (for a very specific reason) and the landscape looks like this:

View attachment 655

View attachment 656
View attachment 657

SGA was the 2nd most efficient player in PTS/36 per USG in the summer of 2024, and he's the #1 player now. Here are the top 10.

View attachment 658

This certainly isn't going to slow down any lust for Saddiq Bey

Here are the least efficient scorers so far this year:

View attachment 659

Where do Spurs fall on this list?

Wemby - 0.920 PTS/36 per USG, 18th highest, right below Brunson and Maxey, right above Lavine and Jamal Murray
Fox - 0.863 PTS/36 per USG, 42nd, right behind Bane and Quickley, right ahead of Booker and Jalen Johnson
Castle - 0.768 PTS/36 per USG, 60th, behind Derrick White and Melo Ball

Where do the rest of the Spurs fall in terms of efficiency? Well, I went ahead and ran those numbers too. Please don't extrapolate this and infer that we should just force feed these guys, because their efficiency is in part because of their roll (this should be obvious). I just think it is interesting too look at their relative efficiency compared to top scoring players.

Korndog - 1.094 PTS/36 per USG. Would lead the league
Keldog - 1.010 PTS/36 per USG. Would rank 3rd in the league!
Julian - 0.936 PTS/36 per USG. Would rank 15th in the league, right behind Dono Mitchell and Demar, ahead of Kawhi and Brunson
Barnes - 0.907 PTS/36 per USG. Would rank 25th in the league, right behind Keyonte and MPJ, right ahead of Randle and NAW
Vassell - 0.889 PTS/36 per USG. Would rank 29th in the league, right behind Pritchard and Bridges, ahead of Okongwu and KAT
Harper - 0.769 PTS/36 per USG. Would rank 60th in the league, right behind Derrick and Melo, ahead of Castle by 0.001

Some things that stand out:

• Wemby has become significant more efficient from his rookie year
• Fox is slightly more efficient from who he was with the Kings, but his USG has reduced from 31% (11th amongst the Top 40 scorers) to 25.5% (which would have ranked 3rd to last in 2023-24)
• Castle is our least efficient scorer in the regular rotation

Some other random notes

• Frequently discussed TMIII ranks 12th in PTS/36 per USG at 0.952
• Old SOB KD is 11th at 0.952, only ahead of TMIII by thousandths of a point
• Luka leads the league in scoring, but only ranks 31st amongst the Top 60 scorers in PTS/36 per USG
• Cooper Flagg is off to a decent start for his career, 0.853 PTS/36 per USG
• What a breakout year for Jalen Duren... I haven't followed the MIP discussion closely but certainly he's gotta be in the mix

Nicely done! Thanks for dusting off the Sunday Analytics series.

 

[Area 51.247]

Aloha friends,

On the other website I used to occasionally post some statistical things I find interesting or just feel people would benefit from understanding. We've done threads on understanding USG%, on the value of pick swaps (both individually and as value enhancers), trade compensation estimators, etc.

I thought it was a good time to bring this series back, and I thought best to start with one of my favorites: looking at the leagues top scorers and their efficiency when putting their scoring in the context of their USG. Recall, USG is a measure of how often a possession ends at the hands of a player, whether by a FGA, FTA or TOV (since these are all the ways a possession can end, other than time simply running out). It's not a measure of ball dominance. A player's ability to score relative to their USG is a good measure of their efficiency. PTS/USG is a measure of how efficiently a player converts their usage into points. Increased shooting efficiency and a reduction of turnovers pushes this number higher.

We can also look at PTS per 36/USG to normalize for playing time. I've done both.

What's a corresponding measure(s) for effectiveness?

In other words, is a player who can rack up efficient stats when the game's not on the line a more valuable part of the lineup than one who gets the crucial events to secure a win?

 

[scott]

What's a corresponding measure(s) for effectiveness?

In other words, is a player who can rack up efficient stats when the game's not on the line a more valuable part of the lineup than one who gets the crucial events to secure a win?

I wish I had a way to filter out low-leverage minutes from the stats, but unfortunately I don't.

But to answer your question... pretty much every player in the Top 60 scorers are guying who are going to be playing in medium and high leverage situations. When comparing bench players and how they would stack up against top scorers, you'd expect to see exactly what we see here - situations where Kornet ranks #1 in the league. I don't think any of us think that Korndog is poised to be the league's top scorer if he was just given the opportunity. I mostly included that part for curiosity (and I know other people might be curious), but it's not a good comparison.

Everyone in the Top 60, however, is generally expected to be carrying a significant scoring burden for their team, so comparing the relative efficiency between them is fair.

Not sure if that answers your question or not? But certainly I would not argue that Kornet is more valuable than SGA.