ImageImageImageImageImage

Players' individual defense rating on a 1-on-1 basis

Moderators: Danny Darko, Kilroy, TyCobb

Kilroy
Forum Mod - Lakers
Forum Mod - Lakers
Posts: 21,520
And1: 12,222
Joined: Jul 10, 2006
Location: The Motel 9 in Vegas
       

Re: Players' individual defense rating on a 1-on-1 basis 

Post#41 » by Kilroy » Fri Sep 20, 2019 4:00 pm

There's a point at which analyzing statistics can become more about the act of analyzing numbers and less about using them to assess player performance... I think we passed that line a state or 2 ago...

Time to reground yourselves in the reality of what you're seeing on the floor.

To start the season last year, Rondo was borderline brilliant for us. He didn't really start sucking bad until he was included in the failed Davis trade.
KCP takes 2/3s of a season to get into full efficiency... By the ends of the last 2 seasons, he was a great defender for us and a solid all around player.
Caruso is a great bench guy who came up big for us after the season was already a total loss... He's a low risk/high reward player both in what we put into him and what he brings to the court.

The numbers Ball So Hard posted simply tell us that there's some logic to either Rondo or KCP starting... Nothing more... This isn't a slam-dunk kind of discovery. Just some odd numbers here and there that support the idea that they could succeed as starters.

Guys like AB and Cook are unknowns... Because we don't have data about how they perform next to either Lebron or Davis...

Point being, don't rat-hole on the analysis... It's all just detail wrapped up in what will amount to an interesting factoid once the season starts anyway.

The reality is none of these guys have played with both LeBron and Davis on the court at the same time. It's going to take some trial and error to figure out who will perform the best in that situation.

And I think everyone should be keeping an Eye on D Wade... I think it's looking like pretty much a lock he's going to move to LA and be in P&G by Christmas at the latest.
Which, if it happens, will make for some interesting trade scenarios with the guys we're discussing now.

In fact, I would love to see a BSH breakdown of what Wade has been doing defensively in these same situations.
Never have rice at Hanzo's house...
Landsberger
General Manager
Posts: 9,031
And1: 1,969
Joined: Jul 04, 2016
 

Re: Players' individual defense rating on a 1-on-1 basis 

Post#42 » by Landsberger » Sat Sep 21, 2019 2:35 am

Kilroy wrote:There's a point at which analyzing statistics can become more about the act of analyzing numbers and less about using them to assess player performance... I think we passed that line a state or 2 ago...

Time to reground yourselves in the reality of what you're seeing on the floor


As someone that does data analysis for a living I couldn't have said it better..... :clap:

There is a stark difference in data analysis and predictive modeling.
tamaraw08
Analyst
Posts: 3,207
And1: 1,235
Joined: Feb 13, 2019
     

Re: Players' individual defense rating on a 1-on-1 basis 

Post#43 » by tamaraw08 » Sun Sep 22, 2019 11:47 pm

Kilroy wrote:There's a point at which analyzing statistics can become more about the act of analyzing numbers and less about using them to assess player performance... I think we passed that line a state or 2 ago...

Time to reground yourselves in the reality of what you're seeing on the floor.

To start the season last year, Rondo was borderline brilliant for us. He didn't really start sucking bad until he was included in the failed Davis trade.
KCP takes 2/3s of a season to get into full efficiency... By the ends of the last 2 seasons, he was a great defender for us and a solid all around player.
Caruso is a great bench guy who came up big for us after the season was already a total loss... He's a low risk/high reward player both in what we put into him and what he brings to the court.

The numbers Ball So Hard posted simply tell us that there's some logic to either Rondo or KCP starting... Nothing more... This isn't a slam-dunk kind of discovery. Just some odd numbers here and there that support the idea that they could succeed as starters.

Guys like AB and Cook are unknowns... Because we don't have data about how they perform next to either Lebron or Davis...

Point being, don't rat-hole on the analysis... It's all just detail wrapped up in what will amount to an interesting factoid once the season starts anyway.

The reality is none of these guys have played with both LeBron and Davis on the court at the same time. It's going to take some trial and error to figure out who will perform the best in that situation.

And I think everyone should be keeping an Eye on D Wade... I think it's looking like pretty much a lock he's going to move to LA and be in P&G by Christmas at the latest.
Which, if it happens, will make for some interesting trade scenarios with the guys we're discussing now.

In fact, I would love to see a BSH breakdown of what Wade has been doing defensively in these same situations.

More about analyzing numbers.....I'm sorry but I don't know what that even means. Yes, I look at stats more bec some fans here and there, even at my work would love to play experts using the lazy way of the eye test to proclaim certain players as great shooters and prolific rebounders bec according to them, they saw several games. I then cringe after realizing that actual stats like shooting % and rebs/game actually proved the contrary.
Am I too obsessed with stats? maybe, but I am not the one who basically disregarded the advanced stats formulated by sports professionals at ESPN, basketballreference and geeks from cleaningtheglass etc. so he can do another extensive research to come up with a separate conclusion. The 3 websites point to KCP as a below average defender but you have every right to believe BSH's results on Rondo, even after he himself said
it doesn't tell the whole story
.
I have repeatedly stated that ALL defensive stats have their own flaws but I must confess I tend to side with 2-3 other entities using a variety of different markers that arrive with the the pretty much the same conclusion. And BTW, BSH's results are not even done yet, what if his results reveal that Danny Green is a bad defender while Cook and Troy Daniels turned out to be great stoppers? I am not saying this will be the case but it can reveal more flaws to the formula. But yes, I am willing to concede that KCP may not be the horrible defender that they are describing him to be but you used the words like "borderline brilliant", "great defender and solid all around player" on these players, I really want to say let's agree to disagree but I just can't help but wonder why 29 other teams didn't bother to tender them offer sheets so they tap into their greatness, (albeit that you mentioned for short periods of time}?
Also, what does your eye test reveal when KCP hit his career 3pt% of 34.7% 10th lowest among SGs last year.
I guess I am too crazy to believe that it's imperative that this team lead by Lebron must hit those 3's in order to be efficient on offense based on his success with his past teams (Top 5 in ORTG when his team hit a high % from the 3pt area)..As a Laker fan I want every player on this team to succeed, heck I advocated for them to resign Rondo after they acquired Davis.
User avatar
zimpy27
Forum Mod
Forum Mod
Posts: 38,867
And1: 36,510
Joined: Jul 13, 2014

Re: Players' individual defense rating on a 1-on-1 basis 

Post#44 » by zimpy27 » Wed Oct 30, 2019 6:59 pm

Ball so hard wrote:I wanted to see how some of our players did when defending certain positions. Specifically I wanted to see who would best defend PGs and therefore IMO should start since we most need someone who is competent at the PG position. There are obviously various metric used to rank defense, many of which factor in team defense, which i'd agree is far more important. I performed an analysis using Player PTS DIFF. Player PTS DIFF is an offensive player’s Points per 100 Possessions in a specific matchup compared to his season average Points per 100 Possessions. I checked Kuzma's Player PTS DIFF result and noted the result moved in the same direction as DWS and DBPM. For example: Kuzma's 2018 DWS and DBPM were 1.9 and -1.3, respectively. His 2017 DWS and DBPM were 2.3 and -1. My analysis of his PTS DIFF revealed an overall net +71 and +14 for 2018 and 2017, respectively. In other words, Kuzma's 2018 numbers got worse in both scenarios.

What exactly I did

I essentially took all the numbers from NBA.com and throw them into a spread sheet, then apply weighting to Player PTS DIFF relative to total possessions. For example, 1 possession had a Player PTS DIFF of +154; this would obviously skew the result if I kept as is. Therefore, I calculated a weighted Player PTS DIFF. For those who are pedantic... a + number is bad lol

Source data: https://stats.nba.com/player/203484/matchups/?sort=PLAYER_PTS&dir=1

I will be posting the results of various players throughout this thread. I will begin with Kuz (copied from another thread).

Kuzma's Weighted Player PTS DIFF by position is as follows:

2018
C - 175 possessions. Centers scored a combined net +122 points more than their season average.
PF - 699 possessions. PF combined under performed their their season average by -60 points.
SF - 485 possessions. SF combined under performed their their season average by -18 points.
SG - 303 possessions. SG combined under performed their their season average by -28 points.
PG - 164 possessions. PG scored a combined net +56 points more than their season average.

2017
C - 164 possessions. Centers scored a combined net +2 points more than their season average.
PF - 773 possessions. PF scored a combined net +51 points more than their season average.
SF - 431 possessions. SF combined under performed their their season average by -67 point.
SG - 268 possessions. SG combined under performed their their season average by -18 point.
PG - 113 possessions. PG scored a combined net +47 points more than their season average.

- Like many of us suspected... he's terrible defending centers.
- Got considerably better his second year defending PF
- Solid both years defending SF
- I'm kind of mystified by his numbers against SGs

For my sanity check on his 2018 SG numbers:

Elite SGs he had a positive impact while defending (Total possessions 74; -33 points):
Luka
Klay
Eric Gordon
Donovan Mitchell
Gary Harris
Booker
Buddy
Redick
Brogdon

Negative impact while defending (Total possessions 33; + 30 points):
Derozan
Butler
Beal
Lou


Did you ever do this for LeBron?
"Let's play some basketball!" - Fergie
Ball so hard
Sixth Man
Posts: 1,996
And1: 677
Joined: Jul 04, 2017
     

Re: Players' individual defense rating on a 1-on-1 basis 

Post#45 » by Ball so hard » Thu Oct 31, 2019 3:47 pm

zimpy27 wrote:
Ball so hard wrote:I wanted to see how some of our players did when defending certain positions. Specifically I wanted to see who would best defend PGs and therefore IMO should start since we most need someone who is competent at the PG position. There are obviously various metric used to rank defense, many of which factor in team defense, which i'd agree is far more important. I performed an analysis using Player PTS DIFF. Player PTS DIFF is an offensive player’s Points per 100 Possessions in a specific matchup compared to his season average Points per 100 Possessions. I checked Kuzma's Player PTS DIFF result and noted the result moved in the same direction as DWS and DBPM. For example: Kuzma's 2018 DWS and DBPM were 1.9 and -1.3, respectively. His 2017 DWS and DBPM were 2.3 and -1. My analysis of his PTS DIFF revealed an overall net +71 and +14 for 2018 and 2017, respectively. In other words, Kuzma's 2018 numbers got worse in both scenarios.

What exactly I did

I essentially took all the numbers from NBA.com and throw them into a spread sheet, then apply weighting to Player PTS DIFF relative to total possessions. For example, 1 possession had a Player PTS DIFF of +154; this would obviously skew the result if I kept as is. Therefore, I calculated a weighted Player PTS DIFF. For those who are pedantic... a + number is bad lol

Source data: https://stats.nba.com/player/203484/matchups/?sort=PLAYER_PTS&dir=1

I will be posting the results of various players throughout this thread. I will begin with Kuz (copied from another thread).

Kuzma's Weighted Player PTS DIFF by position is as follows:

2018
C - 175 possessions. Centers scored a combined net +122 points more than their season average.
PF - 699 possessions. PF combined under performed their their season average by -60 points.
SF - 485 possessions. SF combined under performed their their season average by -18 points.
SG - 303 possessions. SG combined under performed their their season average by -28 points.
PG - 164 possessions. PG scored a combined net +56 points more than their season average.

2017
C - 164 possessions. Centers scored a combined net +2 points more than their season average.
PF - 773 possessions. PF scored a combined net +51 points more than their season average.
SF - 431 possessions. SF combined under performed their their season average by -67 point.
SG - 268 possessions. SG combined under performed their their season average by -18 point.
PG - 113 possessions. PG scored a combined net +47 points more than their season average.

- Like many of us suspected... he's terrible defending centers.
- Got considerably better his second year defending PF
- Solid both years defending SF
- I'm kind of mystified by his numbers against SGs

For my sanity check on his 2018 SG numbers:

Elite SGs he had a positive impact while defending (Total possessions 74; -33 points):
Luka
Klay
Eric Gordon
Donovan Mitchell
Gary Harris
Booker
Buddy
Redick
Brogdon

Negative impact while defending (Total possessions 33; + 30 points):
Derozan
Butler
Beal
Lou


Did you ever do this for LeBron?


Nope. I didn't get a chance to look at his. I can try to get to it at some point this weekend if I find time.

Return to Los Angeles Lakers