When we talk about a players efficiency, a debate often arises discussing a players usage. Usually the debate is, well so and so is only efficient because they don't shoot much, the defense doesn't concentrate on them etc, and the high usage guys are punished by taking shots late in the shot clock.
So I decided (because I'm clearly a loser) to do a little research on efficiency and usage, trying to see if there was much information out there that would provide a correlation. I came across this from the Sloan conference.
One part stood out to me
So I decided, why not apply this to the Raptors. Who, given their usage, was the most efficient?
A couple boring notes:
1)I had no interest in going through and trying to calculate the details and subjectivity of a 'half court' offense. I'd assume perimeter players would use more 'non half court' offense because of leak outs, steals, fast breaks etc. But then again, scoring straight off of offensive rebounds may apply, and the Raptors seem to have a couple good floor running bigs. All in all, its way too much work so I ignored it. Given the Raptors played at a slower than average pace, it may be insignificant anyways.
2)i) 0.0025 to 0.006 is quite a range (I know it seems small, but its relatively significant - the top range is almost two and a half times the lower range) But the article also claims perimeter players usage curve (ie. efficiency vs production) is flatter. So I did this 2 ways. I used the average of the given range (0.00425) and I also used the extremes for each type of player (big at 0.006 vs perimeter at 0.0025). This result is much more harsh on the big men (each increase in usage by 1% is a 2.4x drop in efficiency vs a perimeter player) - except Bargnani, who I gave the benifit of the doubt and called a perimeter player.
2)ii) I actually found a couple less reliable sources (not unreliable, just less reliable) that come to a similar conclusion (although theirs wasn't a range, rather it was 0.0025 drop in pts for a 1% increase in usage). So it helps verify this information. However they also found a low correlation between usage and efficiency.
3) I left out the less relevant players (Kleiza, Telfair etc) and players no longer with us (Jose and Ed) *edit I added Jose for part.
4) I'm no statistician, nor am I a doctor (except perhaps of love), so feel free to take what I'm about to write with a grain of salt should you so choose.
Part 1 - the Basics
So lets start with each players usage and pts/possession (PTS/FGA + FTA(0.45) + TOV)
R. Gay 29.1% 0.895
Demar 24.2% 0.946
Andrea 23.8% 0.865
Lowry 20.3% 0.887
Alan A. 23.6% 0.912
Amir J. 16.1% 1.006
T. Ross 19.6% 0.88
Jonas 16.9% 1.017
Fields 12.9% 0.832
for reference an average NBA Player in the 2012/13 season - 20% 0.921
A quick review shows Jonas and Amir are far and away the most efficient scorers, they are however lower usage players. Demar is an above average scorer on and above average usage. Poor Landry....
*Note: I thought its worth noting that using the above pts/possession to judge scoring efficiency is perhaps slightly unfair to Lowry. It includes TOVs, and since there is a correlation between assists and turnovers, and PGs tend to pass more (and Lowry's assist% was significantly higher than anyone else), Lowry's #s will reflect alot more passes that lead to turnovers than scoring possessions that lead to turnovers than anyone else on the team.
Part 2 - the Meat
Lets get down to the nitty gritty shall we. What I did was estimate a teams #1 option would have a usage of around 30%. Honestly the %number is irrevelant, but it seemed relatively common. I then subtracted a players usage, multiplied that difference by the weighting (0.0025 or 0.006 and 0.00425), then subtracted that result from their average pts/possession. In the end we have 2 results - 1) pts/possession given the usage change of bigs vs perimeter players 2) pts/possession given average usage change
R. Gay 0.893 - 0.891
Demar 0.931 - 0.921
Andrea 0.850 - 0.839
Lowry 0.863 - 0.846
Alan A. 0.896 - 0.885
Amir J. 0.923 - 0.947
T. Ross 0.854 - 0.836
Jonas 0.939 - 0.961
Fields 0.789 - 0.759
Quick review shows us Jonas comes out as the 'theoretical should have been #1 option' under both variables (I should note, the opposite of the assist% that applies to Lowry may apply to Jonas. His low assist% means less passing and therefore he was likely to turnover the ball less on non-scoring possessions). Demar is #2 if we apply a difference in usage cost to bigs vs perimeter players, Amir is #2 if we use average.
Andrea was used despite the hell and highwater that existed.
Poor Landry. Poor, poor Landry......
Part 3 - Conclusion
So the title may be a bit deceptive. I did try and esitmate who the teams #1 should have been, but I also don't think its a perfect representation. But it does perhaps give us an idea of who should have been (or should be in the future) used more often, and who should have been (or should be in the future) used less often.
- The Raptors didn't use their REAL big men enough. They were 2 of the lowest usage players, yet 2 of the most efficient. If the theory above is true, an increase in usage would still keep them as 2 of the more efficient scorers on the team.
- Demar came out as the biggest suprise to me. He had above average efficiency to start, and would in theory maintain an average to above average scoring rate if used like a #1.
- Rudy Gay was used way to much. Hopefully his eye surgery changes things. But given the $s he's getting.....
- Trade Bargnani
- I think fans have been a bit unfair to Anderson. He doesn't come out as an attractive 'scorer', but given the other options available to the team, he probably shot roughly as much as he should have. That ofcourse assumes a team that was trying to win now
- did anyone check to make sure Fields' arm wasn't amputated during surgery?
- Ross was definetely not 'NBA ready'
- Lowry is probably unfairly punished in the analysis, but at the same time he's got a job to do regardless. So I decided to run Jose's numbers (while on the Raptors), and he blew Lowry out of the water. To be fair he was ahead of everyone when we factor in position at 0.975 (this despite the potential disadvantage in this analysis of being a PG - although that itself may be because he is so good at passing without turning over the ball) and only 2nd behind Jonas when just using an average factor (0.954).
I think this verifies some previously held beliefs.
1)This organization, under Colangelo, has valued shot taking over efficiency. There is a pretty consistent relationship between high usage players on this team and lower efficiency results. Lets hope this changes with new management.
2) Just because a player is a low usage player, it shouldn't be assumed that he wouldn't be a high efficiency player if used more. An increase in usage may lead to a decrease in efficiency, but that doesn't mean that player isn't a potentially better choice to use (vs the league or teammates)
3) Give Jonas the ball!!
I also need to show some humility and give Demar credit. This analysis definetely goes against my belief that Demar is an inefficient chucker. His usage may be above average, but he is at the very least relatively efficient at it.
Final disclaimer: I'm not claiming this is perfect. Its only 2012/13 data, doesn't factor in injuries, and there are probably numerous variables that could be considered. I'm also not going guarantee I did everything accurately, or even interpretted it accurately for that matter, but I feel rather confident about the results.
So I decided (because I'm clearly a loser) to do a little research on efficiency and usage, trying to see if there was much information out there that would provide a correlation. I came across this from the Sloan conference.
One part stood out to me
A player who wants to use an additional 1% of his team’s half-court offensive possessions, will see his overall efficiency drop by .0025 − .006 points per possession. This estimate holds teammates constant
A couple boring notes:
1)I had no interest in going through and trying to calculate the details and subjectivity of a 'half court' offense. I'd assume perimeter players would use more 'non half court' offense because of leak outs, steals, fast breaks etc. But then again, scoring straight off of offensive rebounds may apply, and the Raptors seem to have a couple good floor running bigs. All in all, its way too much work so I ignored it. Given the Raptors played at a slower than average pace, it may be insignificant anyways.
2)i) 0.0025 to 0.006 is quite a range (I know it seems small, but its relatively significant - the top range is almost two and a half times the lower range) But the article also claims perimeter players usage curve (ie. efficiency vs production) is flatter. So I did this 2 ways. I used the average of the given range (0.00425) and I also used the extremes for each type of player (big at 0.006 vs perimeter at 0.0025). This result is much more harsh on the big men (each increase in usage by 1% is a 2.4x drop in efficiency vs a perimeter player) - except Bargnani, who I gave the benifit of the doubt and called a perimeter player.
2)ii) I actually found a couple less reliable sources (not unreliable, just less reliable) that come to a similar conclusion (although theirs wasn't a range, rather it was 0.0025 drop in pts for a 1% increase in usage). So it helps verify this information. However they also found a low correlation between usage and efficiency.
3) I left out the less relevant players (Kleiza, Telfair etc) and players no longer with us (Jose and Ed) *edit I added Jose for part.
4) I'm no statistician, nor am I a doctor (except perhaps of love), so feel free to take what I'm about to write with a grain of salt should you so choose.
Part 1 - the Basics
So lets start with each players usage and pts/possession (PTS/FGA + FTA(0.45) + TOV)
R. Gay 29.1% 0.895
Demar 24.2% 0.946
Andrea 23.8% 0.865
Lowry 20.3% 0.887
Alan A. 23.6% 0.912
Amir J. 16.1% 1.006
T. Ross 19.6% 0.88
Jonas 16.9% 1.017
Fields 12.9% 0.832
for reference an average NBA Player in the 2012/13 season - 20% 0.921
A quick review shows Jonas and Amir are far and away the most efficient scorers, they are however lower usage players. Demar is an above average scorer on and above average usage. Poor Landry....
*Note: I thought its worth noting that using the above pts/possession to judge scoring efficiency is perhaps slightly unfair to Lowry. It includes TOVs, and since there is a correlation between assists and turnovers, and PGs tend to pass more (and Lowry's assist% was significantly higher than anyone else), Lowry's #s will reflect alot more passes that lead to turnovers than scoring possessions that lead to turnovers than anyone else on the team.
Part 2 - the Meat
Lets get down to the nitty gritty shall we. What I did was estimate a teams #1 option would have a usage of around 30%. Honestly the %number is irrevelant, but it seemed relatively common. I then subtracted a players usage, multiplied that difference by the weighting (0.0025 or 0.006 and 0.00425), then subtracted that result from their average pts/possession. In the end we have 2 results - 1) pts/possession given the usage change of bigs vs perimeter players 2) pts/possession given average usage change
R. Gay 0.893 - 0.891
Demar 0.931 - 0.921
Andrea 0.850 - 0.839
Lowry 0.863 - 0.846
Alan A. 0.896 - 0.885
Amir J. 0.923 - 0.947
T. Ross 0.854 - 0.836
Jonas 0.939 - 0.961
Fields 0.789 - 0.759
Quick review shows us Jonas comes out as the 'theoretical should have been #1 option' under both variables (I should note, the opposite of the assist% that applies to Lowry may apply to Jonas. His low assist% means less passing and therefore he was likely to turnover the ball less on non-scoring possessions). Demar is #2 if we apply a difference in usage cost to bigs vs perimeter players, Amir is #2 if we use average.
Andrea was used despite the hell and highwater that existed.
Poor Landry. Poor, poor Landry......
Part 3 - Conclusion
So the title may be a bit deceptive. I did try and esitmate who the teams #1 should have been, but I also don't think its a perfect representation. But it does perhaps give us an idea of who should have been (or should be in the future) used more often, and who should have been (or should be in the future) used less often.
- The Raptors didn't use their REAL big men enough. They were 2 of the lowest usage players, yet 2 of the most efficient. If the theory above is true, an increase in usage would still keep them as 2 of the more efficient scorers on the team.
- Demar came out as the biggest suprise to me. He had above average efficiency to start, and would in theory maintain an average to above average scoring rate if used like a #1.
- Rudy Gay was used way to much. Hopefully his eye surgery changes things. But given the $s he's getting.....
- Trade Bargnani
- I think fans have been a bit unfair to Anderson. He doesn't come out as an attractive 'scorer', but given the other options available to the team, he probably shot roughly as much as he should have. That ofcourse assumes a team that was trying to win now
- did anyone check to make sure Fields' arm wasn't amputated during surgery?
- Ross was definetely not 'NBA ready'
- Lowry is probably unfairly punished in the analysis, but at the same time he's got a job to do regardless. So I decided to run Jose's numbers (while on the Raptors), and he blew Lowry out of the water. To be fair he was ahead of everyone when we factor in position at 0.975 (this despite the potential disadvantage in this analysis of being a PG - although that itself may be because he is so good at passing without turning over the ball) and only 2nd behind Jonas when just using an average factor (0.954).
I think this verifies some previously held beliefs.
1)This organization, under Colangelo, has valued shot taking over efficiency. There is a pretty consistent relationship between high usage players on this team and lower efficiency results. Lets hope this changes with new management.
2) Just because a player is a low usage player, it shouldn't be assumed that he wouldn't be a high efficiency player if used more. An increase in usage may lead to a decrease in efficiency, but that doesn't mean that player isn't a potentially better choice to use (vs the league or teammates)
3) Give Jonas the ball!!
I also need to show some humility and give Demar credit. This analysis definetely goes against my belief that Demar is an inefficient chucker. His usage may be above average, but he is at the very least relatively efficient at it.
Final disclaimer: I'm not claiming this is perfect. Its only 2012/13 data, doesn't factor in injuries, and there are probably numerous variables that could be considered. I'm also not going guarantee I did everything accurately, or even interpretted it accurately for that matter, but I feel rather confident about the results.
Comment