View previous topic :: View next topic |
Author |
Message |
durden-tyler Star Player
Joined: 22 Dec 2003 Posts: 1266
|
Posted: Wed Jan 17, 2018 10:33 am Post subject: |
|
|
GoldenThroat wrote: | Here are the Luck-Adjusted Net Ratings for the Lakers, per Nylon Calculus. If you want to know what the hell "luck-adjusted" means, click here.
Lonzo Ball........................+8.1
Kentavious Caldwell-Pope...+4.6
Brook Lopez.....................+3.4
Julius Randle....................+2.4
Kyle Kuzma......................+1.1
Josh Hart.........................+0.4
Larry Nance, Jr..................-0.8
Jordan Clarkson.................-2.8
Brandon Ingram................-3.7
Tyler Ennis........................-6.3
Corey Brewer.....................-8.4
Alex Caruso.....................-16.0 |
Thanks for the info. I'm no data expert, but he's saying that "individuals have next to no control of their opponent’s 3-point percent", I'm starting to read more about that, what do you (GT and others) think about that ? That really surprises me... |
|
Back to top |
|
|
BigGameHames Star Player
Joined: 24 May 2015 Posts: 7982
|
Posted: Wed Jan 17, 2018 10:41 am Post subject: |
|
|
GoldenThroat wrote: | Here are the Luck-Adjusted Net Ratings for the Lakers, per Nylon Calculus. If you want to know what the hell "luck-adjusted" means, click here.
Lonzo Ball........................+8.1
Kentavious Caldwell-Pope...+4.6
Brook Lopez.....................+3.4
Julius Randle....................+2.4
Kyle Kuzma......................+1.1
Josh Hart.........................+0.4
Larry Nance, Jr..................-0.8
Jordan Clarkson.................-2.8
Brandon Ingram................-3.7
Tyler Ennis........................-6.3
Corey Brewer.....................-8.4
Alex Caruso.....................-16.0 |
That terrifies me. The 3 other guys who consistently start are top 3 while BI is only ahead of bench warmers. |
|
Back to top |
|
|
epak Retired Number
Joined: 30 Aug 2005 Posts: 34147
|
Posted: Wed Jan 17, 2018 10:42 am Post subject: |
|
|
GoldenThroat wrote: | Here are the Luck-Adjusted Net Ratings for the Lakers, per Nylon Calculus. If you want to know what the hell "luck-adjusted" means, click here.
Lonzo Ball........................+8.1
Kentavious Caldwell-Pope...+4.6
Brook Lopez.....................+3.4
Julius Randle....................+2.4
Kyle Kuzma......................+1.1
Josh Hart.........................+0.4
Larry Nance, Jr..................-0.8
Jordan Clarkson.................-2.8
Brandon Ingram................-3.7
Tyler Ennis........................-6.3
Corey Brewer.....................-8.4
Alex Caruso.....................-16.0 |
+8.1 is about 34th out of players with 1000+ possessions.
Not too shabby.
So they consider FTs and 3PT% the aspects to consider as "luck." Interesting. |
|
Back to top |
|
|
GoldenThroat Moderator
Joined: 25 Jul 2005 Posts: 37474
|
Posted: Wed Jan 17, 2018 10:51 am Post subject: |
|
|
durden-tyler wrote: | Thanks for the info. I'm no data expert, but he's saying that "individuals have next to no control of their opponent’s 3-point percent", I'm starting to read more about that, what do you (GT and others) think about that ? That really surprises me... |
I'm not a math guy, so I can't explain it from that angle, but this isn't the first time I've heard that. It makes sense to me that a player wouldn't have a BIG impact on an opponents 3PT% simply as a function of opportunity, but I'd figure that they'd have a bigger impact than what the studies seem to suggest. |
|
Back to top |
|
|
BigGameHames Star Player
Joined: 24 May 2015 Posts: 7982
|
Posted: Wed Jan 17, 2018 10:57 am Post subject: |
|
|
GoldenThroat wrote: | durden-tyler wrote: | Thanks for the info. I'm no data expert, but he's saying that "individuals have next to no control of their opponent’s 3-point percent", I'm starting to read more about that, what do you (GT and others) think about that ? That really surprises me... |
I'm not a math guy, so I can't explain it from that angle, but this isn't the first time I've heard that. It makes sense to me that a player wouldn't have a BIG impact on an opponents 3PT% simply as a function of opportunity, but I'd figure that they'd have a bigger impact than what the studies seem to suggest. |
It’s hard to believe that length, properly closing out etc don’t have an affect on your opponents 3 pt %. I don’t think I buy that no matter what the studies say. |
|
Back to top |
|
|
GoldenThroat Moderator
Joined: 25 Jul 2005 Posts: 37474
|
Posted: Wed Jan 17, 2018 11:00 am Post subject: |
|
|
BigGameHames wrote: | GoldenThroat wrote: | durden-tyler wrote: | Thanks for the info. I'm no data expert, but he's saying that "individuals have next to no control of their opponent’s 3-point percent", I'm starting to read more about that, what do you (GT and others) think about that ? That really surprises me... |
I'm not a math guy, so I can't explain it from that angle, but this isn't the first time I've heard that. It makes sense to me that a player wouldn't have a BIG impact on an opponents 3PT% simply as a function of opportunity, but I'd figure that they'd have a bigger impact than what the studies seem to suggest. |
It’s hard to believe that length, properly closing out etc don’t have an affect on your opponents 3 pt %. I don’t think I buy that no matter what the studies say. |
Across a team, I would agree with you. But they're saying that one individual has little to no impact on an opposing team's 3PT%, which I think is defensible. |
|
Back to top |
|
|
BigGameHames Star Player
Joined: 24 May 2015 Posts: 7982
|
Posted: Wed Jan 17, 2018 11:06 am Post subject: |
|
|
GoldenThroat wrote: | BigGameHames wrote: | GoldenThroat wrote: | durden-tyler wrote: | Thanks for the info. I'm no data expert, but he's saying that "individuals have next to no control of their opponent’s 3-point percent", I'm starting to read more about that, what do you (GT and others) think about that ? That really surprises me... |
I'm not a math guy, so I can't explain it from that angle, but this isn't the first time I've heard that. It makes sense to me that a player wouldn't have a BIG impact on an opponents 3PT% simply as a function of opportunity, but I'd figure that they'd have a bigger impact than what the studies seem to suggest. |
It’s hard to believe that length, properly closing out etc don’t have an affect on your opponents 3 pt %. I don’t think I buy that no matter what the studies say. |
Across a team, I would agree with you. But they're saying that one individual has little to no impact on an opposing team's 3PT%, which I think is defensible. |
But if your team has length and closes out properly that means you have numerous individuals with length who close out properly. How can you have one and not the other? |
|
Back to top |
|
|
GoldenThroat Moderator
Joined: 25 Jul 2005 Posts: 37474
|
Posted: Wed Jan 17, 2018 11:11 am Post subject: |
|
|
BigGameHames wrote: | But if your team has length and closes out properly that means you have numerous individuals with length who close out properly. How can you have one and not the other? |
Across a team, sure. But how much of that is one guy responsible for?
Sacramento gives up the highest 3PT% to opponents (38.5%), and Washington gives up the lowest (33.6%). So the bandwidth that you're talking about is just 4.9%. |
|
Back to top |
|
|
Baron Von Humongous Retired Number
Joined: 02 Jul 2015 Posts: 32979
|
Posted: Wed Jan 17, 2018 11:17 am Post subject: |
|
|
GoldenThroat wrote: | Here are the Luck-Adjusted Net Ratings for the Lakers, per Nylon Calculus. If you want to know what the hell "luck-adjusted" means, click here.
Lonzo Ball........................+8.1
Kentavious Caldwell-Pope...+4.6
Brook Lopez.....................+3.4
Julius Randle....................+2.4
Kyle Kuzma......................+1.1
Josh Hart.........................+0.4
Larry Nance, Jr..................-0.8
Jordan Clarkson.................-2.8
Brandon Ingram................-3.7
Tyler Ennis........................-6.3
Corey Brewer.....................-8.4
Alex Caruso.....................-16.0 |
I'd be concerned that Brewer being so bad doesn't make Ingram look better than that, but I also wonder what Ingram's LANRtg(?) looks like during his recent uptick. |
|
Back to top |
|
|
levon Franchise Player
Joined: 11 Oct 2016 Posts: 10602
|
Posted: Wed Jan 17, 2018 11:30 am Post subject: |
|
|
GoldenThroat wrote: | Here are the Luck-Adjusted Net Ratings for the Lakers, per Nylon Calculus. If you want to know what the hell "luck-adjusted" means, click here.
Lonzo Ball........................+8.1
Kentavious Caldwell-Pope...+4.6
Brook Lopez.....................+3.4
Julius Randle....................+2.4
Kyle Kuzma......................+1.1
Josh Hart.........................+0.4
Larry Nance, Jr..................-0.8
Jordan Clarkson.................-2.8
Brandon Ingram................-3.7
Tyler Ennis........................-6.3
Corey Brewer.....................-8.4
Alex Caruso.....................-16.0 |
I like what this stat is trying to do but disagree a lot with its derivation. |
|
Back to top |
|
|
Inspector Gadget Retired Number
Joined: 18 Apr 2016 Posts: 46641
|
Posted: Wed Jan 17, 2018 11:40 am Post subject: |
|
|
Question for the old timers...
Is Ball a better defensive player at age 20 then Magic was coming out of Michigan? |
|
Back to top |
|
|
GoldenThroat Moderator
Joined: 25 Jul 2005 Posts: 37474
|
Posted: Wed Jan 17, 2018 11:43 am Post subject: |
|
|
levon wrote: | [I like what this stat is trying to do but disagree a lot with its derivation. |
Can you expound on that? |
|
Back to top |
|
|
levon Franchise Player
Joined: 11 Oct 2016 Posts: 10602
|
Posted: Wed Jan 17, 2018 11:55 am Post subject: |
|
|
GoldenThroat wrote: | levon wrote: | [I like what this stat is trying to do but disagree a lot with its derivation. |
Can you expound on that? |
The way I understand it, it punishes 3 point shot creators by regressing it to the average. I'm not entirely sure why the assumption is to take the mean of a player's on/off in terms of 3pt% without providing some kind of prior (like maybe the player's just good at creating 3 pt opportunities), but I suppose that's a lot of the issue with on/off metrics.
The obvious one is it also doesn't seem to account for lineups.
The way you're using it is perfectly informative though, which is within one team and with supplementary knowledge of roles on the Lakers. |
|
Back to top |
|
|
tox Franchise Player
Joined: 16 Nov 2015 Posts: 17876
|
Posted: Wed Jan 17, 2018 12:18 pm Post subject: |
|
|
BigGameHames wrote: | GoldenThroat wrote: | Here are the Luck-Adjusted Net Ratings for the Lakers, per Nylon Calculus. If you want to know what the hell "luck-adjusted" means, click here.
Lonzo Ball........................+8.1
Kentavious Caldwell-Pope...+4.6
Brook Lopez.....................+3.4
Julius Randle....................+2.4
Kyle Kuzma......................+1.1
Josh Hart.........................+0.4
Larry Nance, Jr..................-0.8
Jordan Clarkson.................-2.8
Brandon Ingram................-3.7
Tyler Ennis........................-6.3
Corey Brewer.....................-8.4
Alex Caruso.....................-16.0 |
That terrifies me. The 3 other guys who consistently start are top 3 while BI is only ahead of bench warmers. |
The weirdest thing about this stat is that you can argue part of why Lonzo is so high is because his backups are awful. When your backups at Ennis, Caruso, or JC (who has regressed in both senses of the word to sucking at point guard), then yeah your net rating will look better than if your backup were, say, Steph Curry.
But the same exact thing is true of BI. Corey Brewer is awful. Sure, sometimes Hart or KCP take the 3 spot but you'd think BI would just get a massive boost from the Brewer effect. (Note: I don't think this is a concern because BI was legitimately bad for two months or so. This is just another way of confirming that. I bet his recent numbers would look a lot better.) |
|
Back to top |
|
|
tox Franchise Player
Joined: 16 Nov 2015 Posts: 17876
|
Posted: Wed Jan 17, 2018 12:20 pm Post subject: |
|
|
BigGameHames wrote: | GoldenThroat wrote: | durden-tyler wrote: | Thanks for the info. I'm no data expert, but he's saying that "individuals have next to no control of their opponent’s 3-point percent", I'm starting to read more about that, what do you (GT and others) think about that ? That really surprises me... |
I'm not a math guy, so I can't explain it from that angle, but this isn't the first time I've heard that. It makes sense to me that a player wouldn't have a BIG impact on an opponents 3PT% simply as a function of opportunity, but I'd figure that they'd have a bigger impact than what the studies seem to suggest. |
It’s hard to believe that length, properly closing out etc don’t have an affect on your opponents 3 pt %. I don’t think I buy that no matter what the studies say. |
Because if you have length and close out better, offensive teams take fewer 3s (because you're running them off the line). When they do shoot 3s, they're still taking the same caliber of looks and so 3FG% isn't much different for a good defense vs. a bad one.
It might sound counterintuitive, but it's been shown in several academic basketball papers. Science is often counterintuitive. |
|
Back to top |
|
|
BigGameHames Star Player
Joined: 24 May 2015 Posts: 7982
|
Posted: Wed Jan 17, 2018 12:24 pm Post subject: |
|
|
tox wrote: | BigGameHames wrote: | GoldenThroat wrote: | durden-tyler wrote: | Thanks for the info. I'm no data expert, but he's saying that "individuals have next to no control of their opponent’s 3-point percent", I'm starting to read more about that, what do you (GT and others) think about that ? That really surprises me... |
I'm not a math guy, so I can't explain it from that angle, but this isn't the first time I've heard that. It makes sense to me that a player wouldn't have a BIG impact on an opponents 3PT% simply as a function of opportunity, but I'd figure that they'd have a bigger impact than what the studies seem to suggest. |
It’s hard to believe that length, properly closing out etc don’t have an affect on your opponents 3 pt %. I don’t think I buy that no matter what the studies say. |
Because if you have length and close out better, offensive teams take fewer 3s (because you're running them off the line). When they do shoot 3s, they're still taking the same caliber of looks and so 3FG% isn't much different for a good defense vs. a bad one.
It might sound counterintuitive, but it's been shown in several academic basketball papers. Science is often counterintuitive. |
Now that actually makes a lot of sense. They take less 3s because they have fewer good opportunities. How trustworthy do you consider the stat as a whole? |
|
Back to top |
|
|
yinoma2001 Retired Number
Joined: 19 Jun 2010 Posts: 119487
|
Posted: Wed Jan 17, 2018 12:25 pm Post subject: |
|
|
Lonzo has already missed 7 games this year (out of 43). Of course you can't extrapolate that out, but that's about a 13 missed games pace for this year. It's been what, the shoulder and knees so far? Has to get that durability built up. IIRC he missed a good number of games in the SPL/preseason too with injuries? _________________ From 2-10 to the Western Conference Finals |
|
Back to top |
|
|
GoldenThroat Moderator
Joined: 25 Jul 2005 Posts: 37474
|
Posted: Wed Jan 17, 2018 12:27 pm Post subject: |
|
|
yinoma2001 wrote: | Lonzo has already missed 7 games this year (out of 43). Of course you can't extrapolate that out, but that's about a 13 missed games pace for this year. It's been what, the shoulder and knees so far? Has to get that durability built up. IIRC he missed a good number of games in the SPL/preseason too with injuries? |
His durability is by far my biggest concern about him. |
|
Back to top |
|
|
danzag Franchise Player
Joined: 28 Apr 2013 Posts: 22313 Location: Brazil
|
Posted: Wed Jan 17, 2018 12:28 pm Post subject: |
|
|
yinoma2001 wrote: | Lonzo has already missed 7 games this year (out of 43). Of course you can't extrapolate that out, but that's about a 13 missed games pace for this year. It's been what, the shoulder and knees so far? Has to get that durability built up. IIRC he missed a good number of games in the SPL/preseason too with injuries? |
I thinks the Lakers are just being extra careful. No need to risk his career for this season |
|
Back to top |
|
|
yinoma2001 Retired Number
Joined: 19 Jun 2010 Posts: 119487
|
Posted: Wed Jan 17, 2018 12:32 pm Post subject: |
|
|
GoldenThroat wrote: | yinoma2001 wrote: | Lonzo has already missed 7 games this year (out of 43). Of course you can't extrapolate that out, but that's about a 13 missed games pace for this year. It's been what, the shoulder and knees so far? Has to get that durability built up. IIRC he missed a good number of games in the SPL/preseason too with injuries? |
His durability is by far my biggest concern about him. |
Yeah. Part of the problem with him fighting for rebounds. You run into those big tree trunks down low and he's a scrawny twig too, not like Slenderingram. _________________ From 2-10 to the Western Conference Finals |
|
Back to top |
|
|
tox Franchise Player
Joined: 16 Nov 2015 Posts: 17876
|
Posted: Wed Jan 17, 2018 12:32 pm Post subject: |
|
|
levon wrote: | GoldenThroat wrote: | levon wrote: | [I like what this stat is trying to do but disagree a lot with its derivation. |
Can you expound on that? |
The way I understand it, it punishes 3 point shot creators by regressing it to the average. I'm not entirely sure why the assumption is to take the mean of a player's on/off in terms of 3pt% without providing some kind of prior (like maybe the player's just good at creating 3 pt opportunities), but I suppose that's a lot of the issue with on/off metrics.
The obvious one is it also doesn't seem to account for lineups.
The way you're using it is perfectly informative though, which is within one team and with supplementary knowledge of roles on the Lakers. |
Lineups are definitely the real issue with this stat. If you play next to Brewer, Ingram, Nance, and Randle, the only thing unlucky about having a bad teammate 3FG% is your coach's lineup choices. It wouldn't be hard to find how many 3FGA each teammate had next to you vs. without you (the data is available on NBAwowy so clearly you can pull it from NBA.com programmatically).
I think you are overstating the effects of shot creators though. As mentioned above, the issue tends to be more that shot creators create a higher volume of 3s, not necessarily more open ones. This is most obvious when you take a look at a guy like Ingram. He basically only takes open 3s. So playing next to LeBron or Harden wouldn't raise his 3FG% much (because you don't really get better looks than wide open 3s) but he'd get more of the 3s he's comfortable taking. |
|
Back to top |
|
|
levon Franchise Player
Joined: 11 Oct 2016 Posts: 10602
|
Posted: Wed Jan 17, 2018 12:32 pm Post subject: |
|
|
tox wrote: | BigGameHames wrote: | GoldenThroat wrote: | durden-tyler wrote: | Thanks for the info. I'm no data expert, but he's saying that "individuals have next to no control of their opponent’s 3-point percent", I'm starting to read more about that, what do you (GT and others) think about that ? That really surprises me... |
I'm not a math guy, so I can't explain it from that angle, but this isn't the first time I've heard that. It makes sense to me that a player wouldn't have a BIG impact on an opponents 3PT% simply as a function of opportunity, but I'd figure that they'd have a bigger impact than what the studies seem to suggest. |
It’s hard to believe that length, properly closing out etc don’t have an affect on your opponents 3 pt %. I don’t think I buy that no matter what the studies say. |
Because if you have length and close out better, offensive teams take fewer 3s (because you're running them off the line). When they do shoot 3s, they're still taking the same caliber of looks and so 3FG% isn't much different for a good defense vs. a bad one.
It might sound counterintuitive, but it's been shown in several academic basketball papers. Science is often counterintuitive. |
Think it might be worth factoring in defender distance on 3's taken in that case. I still think that teams that defend the perimeter well will force more contested 3's than ones that don't. Opponent shot selection and closeout attacking skills might have a big effect. |
|
Back to top |
|
|
levon Franchise Player
Joined: 11 Oct 2016 Posts: 10602
|
Posted: Wed Jan 17, 2018 12:35 pm Post subject: |
|
|
tox wrote: | levon wrote: | GoldenThroat wrote: | levon wrote: | [I like what this stat is trying to do but disagree a lot with its derivation. |
Can you expound on that? |
The way I understand it, it punishes 3 point shot creators by regressing it to the average. I'm not entirely sure why the assumption is to take the mean of a player's on/off in terms of 3pt% without providing some kind of prior (like maybe the player's just good at creating 3 pt opportunities), but I suppose that's a lot of the issue with on/off metrics.
The obvious one is it also doesn't seem to account for lineups.
The way you're using it is perfectly informative though, which is within one team and with supplementary knowledge of roles on the Lakers. |
Lineups are definitely the real issue with this stat. If you play next to Brewer, Ingram, Nance, and Randle, the only thing unlucky about having a bad teammate 3FG% is your coach's lineup choices. It wouldn't be hard to find how many 3FGA each teammate had next to you vs. without you (the data is available on NBAwowy so clearly you can pull it from NBA.com programmatically).
I think you are overstating the effects of shot creators though. As mentioned above, the issue tends to be more that shot creators create a higher volume of 3s, not necessarily more open ones. This is most obvious when you take a look at a guy like Ingram. He basically only takes open 3s. So playing next to LeBron or Harden wouldn't raise his 3FG% much (because you don't really get better looks than wide open 3s) but he'd get more of the 3s he's comfortable taking. |
Your second point is nested in your first point though. Shot selection is dependent on players played with. What if instead of Ingram you have a sufficiently trigger happy high volume 3 pt shooter, and having Lonzo on the floor gets that shooter more open looks than off?
I didn't consider that angle though, and based on the amount of 3's taken that might all just be noise and taking a mean might be fine. I guess my two points of contention are related then. |
|
Back to top |
|
|
tox Franchise Player
Joined: 16 Nov 2015 Posts: 17876
|
Posted: Wed Jan 17, 2018 12:37 pm Post subject: |
|
|
BigGameHames wrote: |
Now that actually makes a lot of sense. They take less 3s because they have fewer good opportunities. How trustworthy do you consider the stat as a whole? |
It's an interesting stat without doubt, but I do have my grievances. I mentioned my above qualms with how it calculates teammate 3FG%/FT% effects. I suspect this stat mixes up "luck" with "teammate caliber." Defensively there's a similar thing -- maybe starters are generally "unlucky" on defense because opposing teams play their best shooters to start the game (so naturally the opposing team will have a better 3FG%). It wouldn't be hard to adjust offensively based on on/off data, but adjusting defensively would be a pain, which is why I suspect they didn't do it.
Of course, you can easily argue this isn't a bad thing, adjusting for both teammate caliber as well as actual bad "luck." But I prefer when my stats do a single thing that they are said to do. |
|
Back to top |
|
|
venturalakersfan Retired Number
Joined: 14 Apr 2001 Posts: 144462 Location: The Gold Coast
|
Posted: Wed Jan 17, 2018 12:45 pm Post subject: |
|
|
durden-tyler wrote: | GoldenThroat wrote: | Here are the Luck-Adjusted Net Ratings for the Lakers, per Nylon Calculus. If you want to know what the hell "luck-adjusted" means, click here.
Lonzo Ball........................+8.1
Kentavious Caldwell-Pope...+4.6
Brook Lopez.....................+3.4
Julius Randle....................+2.4
Kyle Kuzma......................+1.1
Josh Hart.........................+0.4
Larry Nance, Jr..................-0.8
Jordan Clarkson.................-2.8
Brandon Ingram................-3.7
Tyler Ennis........................-6.3
Corey Brewer.....................-8.4
Alex Caruso.....................-16.0 |
Thanks for the info. I'm no data expert, but he's saying that "individuals have next to no control of their opponent’s 3-point percent", I'm starting to read more about that, what do you (GT and others) think about that ? That really surprises me... |
A defensive player only has control of 3 pointers his guy shoots. He has no control over his teammates 3 point defense. In round numbers he has 20% control over 3 pointers and FTs. At least that is my thoughts. _________________ RIP mom. 11-21-1933 to 6-14-2023. |
|
Back to top |
|
|
|