RealGM

I really don't think this stat is as useless as everyone thinks it is this early... Maybe someone should screenshot the player rankings and we'll revisit them at year end. That way we'll have some idea of how far up/down players are likely to shift in the future. By closely evalutating the rankings I'd say defensively it's still quite the crapshoot. But offensively i see a lot of merit in the rankings.

I watch a lot of Hornets games and their lineups feature some good talking points so I'll use them as an example and predict what i
think will happen.

First of all, the Hornets are a good case example because Nic Batum has not played a single minute yet and he comes back tonight. MKG missed roughly half of our games due a death in the family. Lastly, Frank/Zeller have a great prior history together (particularly with kemba) but they are not playing with kemba much right now while dwight is playing almost every minute with kemba. We also have gotten abysmal backup PG production from Monk as he's been forced to play out of position due to injuries. Lastly, we've got lamb who is typically a 2nd unit guy playing in the starting lineup quite frequently early on.


Anyway, the reason the Hornets are a great example here is because they are about to undergo some major lineup shuffles due to guys returning from injuries. They are also likely to undergo some future lineup shuffles due to current underachievement. And being a veteran team with multiple injury prone guys it's likely more guys will miss games in the future. The more the lineups shuffle the more good data we have to work with. If everyone plays with the same guys all the time we really have nothing but guesswork.


One of the limitations of this formula, particularly early in the season is how do you separate guys from the other guys they play with... therefore, the more lineup shuffles the better the data truly is for sorting out individual contributions.

Here are my predictions between now and year end for the Hornets based on the above observations:

First of all, I'm not going to spend too much time discussing defense because I believe it's sample size isn't close to acceptable. Focus on offensive analysis.

Kemba Walker #43 overall (#6 PG), 2.61 off, -0.72 def I think he's being dragged down massively by playing with Dwight on offense. And to a lesser extent MKG or Marv at SF (who needs to play PF at all times at this point in his career) He's currently the 4th rated overall offensive player, so his contributions are obviously being noted, but I think his score will continue to rise as the Dwight effect becomes more pronounced. Defensively, if anything he's probably slightly overrated.

Jeremy Lamb (#20 SG) 0.56 off, -1.05 def. Tough to say but probably Overrated. He's a box score guy so the early projections are likely skewed upwards. He's one of those guys that tends to look good at times but creates a drag on your teammates functionality. There is nothing in Lamb's career data that leads me to believe his weak defensive numbers are inaccurate.

MKG (#28 SF) -1.33 off, 1.15 def. Really hard to say as he's missed enough games that his data is more relevant. I think his defense is for sure a huge plus. Offensively, i think his biggest strengths (cutting and off rebounding) are really neutered by playing with Dwight. He's transformed himself into a really good midrange shooter but he doesn't shoot from range so RPM is a good way to try to figure out how limited he truly is offensively. I'll say he ends up not being quite this negative offensively, but no real conviction behind that. Defensively we have looked quite a bit better when he's in the lineup so i think that will continue to rise.

Marv (#23 PF) 0.34 off, 0.77 def. I think he's being dragged down by the Dwight effect for sure. He has looked good this year. He shot like crap last year (rumored finger injury) but he's shot the ball well so far this year. I refuse to believe a floor spacing PF who shoots well is this bad offensively. He's also played in many great Hornets lineups over the past couple of seasons so I think this initial estimate is very conservative for his offensive value.

Dwight (#18 C) -0.51 off, 2.29 def. I'd bet my house that Dwight's offensive impact is grossly overstated by this estimate. He's playing nearly all his minutes with Kemba, Lamb, Marv who have all shot incredibly well and provided spacing. Dwight is turning the ball over nonstop, missing FT's at an incredible rate, and forcing inefficient post possessions. I think once we see Zeller get minutes with kemba this formula will really start punishing Dwight by recognizing his offensive drag. Defensively, he's been quite good that much is obvious. Is he truly the 13th best defensive player in the NBA? I have no idea, but I have no reason to doubt it. The hornets have a wonderful fg% defense and he completely shuts down the offensive glass.


Bench: Now here are a few guys I feel confident in really expanding on.

Cody Zeller: #59 ranked Center (-1.46 off, 0.63 def). There is ABSOLUTELY NO WAY this is accurate. This is off by an order of magnitude. Our bench has been completely disgusting thanks to not having a backup PG for the first month of the season. Let Zeller get some minutes with Kemba instead of playing with atrocious lineups and his numbers will normalize very quickly. There is simply no way Zeller is this negative of an offensive player. No way in the world. He's one of the best screeners and energy bigs in the game and he's produced massively for too many years now. The other problem cody has had is they've tried him at PF some, where he is significantly less effective clearly. Expect Zeller to spike defensively as well. Right now he's being penalized by the fact that Dwight is a very good defender in his own right and that Monk doesn't know what he's doing on that end.

Frank Kaminsky: #80 PF (-0.78 off, -0.68 def). Frank has actually been worse than this defensively IMO. He's having trouble finding anyone on the floor he can guard at times. This is a case where he's getting too much credit for being 7 feet tall. The formula is asssuming he's not the one to blame, but I can assure you he has been a big part of the problem on defense. His offensive rating however is grossly off due to the horrible bench guard play. And his own poor box scores. There's no way he will end the season as a negative offensive player. Again... let him enjoy some Kemba minutes and the numbers will recuperate quickly.

Malik Monk: #92 SG (-1.34 off, -1.78 def). There is no question Malik Monk has sucked. He's a 19 yo rookie forced to learn to play point guard which he didnt even play in college. That's a trainwreck waiting to happen. On top of that he's been fairly cold from the floor and anything but gunshy. Defensively, he's horrid. I don't expect that rating to improve one bit. Offensively, its a bit more nuanced though. Perhaps we could see a lot of box score improvement as he slides back to SG or gets paired with another playmaker now that Batum is back... Realistically, he's had the benefit of playing with Cody AND frank which should be a very good offensive unit and he hasn't gotten Dwight'ed a lot. I'd say his offensive numbers will continue to drop once this formula figures out that Monk is truly the problem.



There that's just an example to show how we can use CONTEXT to explain some of the limiting factors of the stat and what we expect to happen as more data becomes available. Whether I'm right or not about the Dwight effect is somewhat irrelevant. It gives us things we can look for in what I feel is a pretty obvious case so in the future we dont have to just throw our hands in the air and scream "SMALL SAMPLE SIZE"

HoopsMalone wrote:I posted this in the RPM thread on the GB but it's got a lot of Hornets talk so i figured i'd repost here. Would Love to hear other Hornets fans opinions on these observations or their own personal observations.

I really don't think this stat is as useless as everyone thinks it is this early... Maybe someone should screenshot the player rankings and we'll revisit them at year end. That way we'll have some idea of how far up/down players are likely to shift in the future. By closely evalutating the rankings I'd say defensively it's still quite the crapshoot. But offensively i see a lot of merit in the rankings.

I watch a lot of Hornets games and their lineups feature some good talking points so I'll use them as an example and predict what i
think will happen.

First of all, the Hornets are a good case example because Nic Batum has not played a single minute yet and he comes back tonight. MKG missed roughly half of our games due a death in the family. Lastly, Frank/Zeller have a great prior history together (particularly with kemba) but they are not playing with kemba much right now while dwight is playing almost every minute with kemba. We also have gotten abysmal backup PG production from Monk as he's been forced to play out of position due to injuries. Lastly, we've got lamb who is typically a 2nd unit guy playing in the starting lineup quite frequently early on.


Anyway, the reason the Hornets are a great example here is because they are about to undergo some major lineup shuffles due to guys returning from injuries. They are also likely to undergo some future lineup shuffles due to current underachievement. And being a veteran team with multiple injury prone guys it's likely more guys will miss games in the future. The more the lineups shuffle the more good data we have to work with. If everyone plays with the same guys all the time we really have nothing but guesswork.


One of the limitations of this formula, particularly early in the season is how do you separate guys from the other guys they play with... therefore, the more lineup shuffles the better the data truly is for sorting out individual contributions.

Here are my predictions between now and year end for the Hornets based on the above observations:

First of all, I'm not going to spend too much time discussing defense because I believe it's sample size isn't close to acceptable. Focus on offensive analysis.

Kemba Walker #43 overall (#6 PG), 2.61 off, -0.72 def I think he's being dragged down massively by playing with Dwight on offense. And to a lesser extent MKG or Marv at SF (who needs to play PF at all times at this point in his career) He's currently the 4th rated overall offensive player, so his contributions are obviously being noted, but I think his score will continue to rise as the Dwight effect becomes more pronounced. Defensively, if anything he's probably slightly overrated.

Kemba #19 overall (#7 PG) 4.50 off, -0.64 def He really shot up offensively (nearly doubling from 2.5 to 4.5) as the data figured out Dwight was skewing his offense downward.

Jeremy Lamb (#20 SG) 0.56 off, -1.05 def. Tough to say but probably Overrated. He's a box score guy so the early projections are likely skewed upwards. He's one of those guys that tends to look good at times but creates a drag on your teammates functionality. There is nothing in Lamb's career data that leads me to believe his weak defensive numbers are inaccurate.

Jeremy Lamb (#25 SG) 0.59 off, -0.78 def. About the same guy he was on the initial review. Stats have dropped lately due to poor play and knee problems.

MKG (#28 SF) -1.33 off, 1.15 def. Really hard to say as he's missed enough games that his data is more relevant. I think his defense is for sure a huge plus. Offensively, i think his biggest strengths (cutting and off rebounding) are really neutered by playing with Dwight. He's transformed himself into a really good midrange shooter but he doesn't shoot from range so RPM is a good way to try to figure out how limited he truly is offensively. I'll say he ends up not being quite this negative offensively, but no real conviction behind that. Defensively we have looked quite a bit better when he's in the lineup so i think that will continue to rise.

MKG (#27 SF) -1.08 off, 0.87 def. He's come up a little on offense and down a little on defense. Not too shocking, still virtually the same.

Marv (#23 PF) 0.34 off, 0.77 def. I think he's being dragged down by the Dwight effect for sure. He has looked good this year. He shot like crap last year (rumored finger injury) but he's shot the ball well so far this year. I refuse to believe a floor spacing PF who shoots well is this bad offensively. He's also played in many great Hornets lineups over the past couple of seasons so I think this initial estimate is very conservative for his offensive value.

Marv (#6 PF and #36 OVERALL) 1.71 off, 0.58 def. I nailed this one. Marv has been terrific this year (top 40 overall!) and the numbers finally accurately reflect what he brings to the table offensively as a floor spacer/shooter.

Dwight (#18 C) -0.51 off, 2.29 def. I'd bet my house that Dwight's offensive impact is grossly overstated by this estimate. He's playing nearly all his minutes with Kemba, Lamb, Marv who have all shot incredibly well and provided spacing. Dwight is turning the ball over nonstop, missing FT's at an incredible rate, and forcing inefficient post possessions. I think once we see Zeller get minutes with kemba this formula will really start punishing Dwight by recognizing his offensive drag. Defensively, he's been quite good that much is obvious. Is he truly the 13th best defensive player in the NBA? I have no idea, but I have no reason to doubt it. The hornets have a wonderful fg% defense and he completely shuts down the offensive glass.

Dwight (#48 C). -2.88 off, 1.91 def. Tough to be anymore accurate than this. His offensive rating is now nearly 6x as bad and he's one of the worst offensive players in the NBA. The defensive numbers have been somewhat maintained but have come down a touch also. It was truly only a matter of time before the data figured out who was responsible for the awful offensive drag.

Bench: Now here are a few guys I feel confident in really expanding on.

Cody Zeller: #59 ranked Center (-1.46 off, 0.63 def). There is ABSOLUTELY NO WAY this is accurate. This is off by an order of magnitude. Our bench has been completely disgusting thanks to not having a backup PG for the first month of the season. Let Zeller get some minutes with Kemba instead of playing with atrocious lineups and his numbers will normalize very quickly. There is simply no way Zeller is this negative of an offensive player. No way in the world. He's one of the best screeners and energy bigs in the game and he's produced massively for too many years now. The other problem cody has had is they've tried him at PF some, where he is significantly less effective clearly. Expect Zeller to spike defensively as well. Right now he's being penalized by the fact that Dwight is a very good defender in his own right and that Monk doesn't know what he's doing on that end.

Cody Zeller: #45 ranked center (-1.59 off, 0.78 def). I think Zeller will continue to rise while he's out. That second unit has some really bad data associated with it, so the more time he sits and they can see it wasn't his fault the more I believe he will rise. Also, he needs some minutes with the starters. An injury to Dwight would probably help him rise quite a bit IMO. He hasn't made quite the meteoric rise I thought he would by now but that 2nd unit is a war zone.

Frank Kaminsky: #80 PF (-0.78 off, -0.68 def). Frank has actually been worse than this defensively IMO. He's having trouble finding anyone on the floor he can guard at times. This is a case where he's getting too much credit for being 7 feet tall. The formula is asssuming he's not the one to blame, but I can assure you he has been a big part of the problem on defense. His offensive rating however is grossly off due to the horrible bench guard play. And his own poor box scores. There's no way he will end the season as a negative offensive player. Again... let him enjoy some Kemba minutes and the numbers will recuperate quickly.

Frank Kaminsky (#89 ranked PF, -1.04 off, -1.51 def) Nailed this one on the defensive end. The data finally figured out Frank is a disaster on defense. He wasn't getting nearly his lion share of the blame early on. Looks like I was probably wrong on offense though. I don't see him becoming a net positive any longer. Perhaps, a bit better than this though by season's end.

Malik Monk: #92 SG (-1.34 off, -1.78 def). There is no question Malik Monk has sucked. He's a 19 yo rookie forced to learn to play point guard which he didnt even play in college. That's a trainwreck waiting to happen. On top of that he's been fairly cold from the floor and anything but gunshy. Defensively, he's horrid. I don't expect that rating to improve one bit. Offensively, its a bit more nuanced though. Perhaps we could see a lot of box score improvement as he slides back to SG or gets paired with another playmaker now that Batum is back... Realistically, he's had the benefit of playing with Cody AND frank which should be a very good offensive unit and he hasn't gotten Dwight'ed a lot. I'd say his offensive numbers will continue to drop once this formula figures out that Monk is truly the problem.

Malik Monk #99 SG (better than only shabazz muhammad) (-1.98 off, -2.29 def) At this point Monk's future as a productive NBA player is in slight jeopardy. He needs to show something (ANYTHING!) quickly to prove he's not a draft bust. He's worse than advertised defensively and he's been completely worthless offensively. It's tough to see him carving out anymore than a Jamal Crawford/Lou Williams type ceiling, let alone a good player if he doesn't show something quickly. It's easy to see why he slipped in the lottery now.


There that's just an example to show how we can use CONTEXT to explain some of the limiting factors of the stat and what we expect to happen as more data becomes available. Whether I'm right or not about the Dwight effect is somewhat irrelevant. It gives us things we can look for in what I feel is a pretty obvious case so in the future we dont have to just throw our hands in the air and scream "SMALL SAMPLE SIZE"

HoopsMalone wrote:#41 SG is just outrageous for a guy on basically a max deal