Pop on Lamarcus: He's a better defender than we thought when we signed him

[Chinook]

I'm loving LMA's work on the defensive end this year, I wasn't expecting him to be anywhere near as good as Splitter there, much less far superior.

He's superior to broke-ass Splitter, yes. But pre-TSOB Splitter was the best defender on a team with Kawhi and Duncan on it.

I consider LMA vs Splitter to be like Kobe/Lebron vs Bowen. When locked in LMA, Kobe and James have/had the talent to be elite defenders and show/ed it in big match-ups. But they're just going to completely zone out for possessions at a time (or seasons at a time in Kobe's case). If the Spurs needed a stop and I could have 2013-2014 Splitter or 2015-2016 LMA, I wouldn't have a tough decision at all.

[Dre_7]

The casual fans always look at offensive numbers before defensive effort... If Aldridge was averaging 20 ppg with a 47 FG% and playing Harden-like defense, the Spurs would a much less successful team. Instead he's rebounding and blocking shots at a better rate than his career averages while playing fewer minutes than his career average.

This!


People LOVE to focus on PPG (which can be a very overrated stat).

Aldridge has been a BEAST this year!

[cjw]

Which is why fretting over the Diaw/West combo is pointless. Pop can make it disappear when necessary.

And it hasn't even been that bad ... +39 in 270 minutes / 540 possessions. That's +6.9 points over 48 minutes, or +7.2 points per 100 possessions.

It's not playable against some lineups, but definitely something the Spurs can tread water with during the regular season. Come the playoffs, Aldridge for 35 minutes and Duncan for 29 minutes would allow one to be on the floor at all times, and them to play together for about 20 MPG. And that's probably conservative on the minute count.

[spurs10]

Feeling pretty damn lucky right now. PATFO did a of job this summer! Life is good!

[Mel_13]

And it hasn't even been that bad ... +39 in 270 minutes / 540 possessions. That's +6.9 points over 48 minutes, or +7.2 points per 100 possessions.

It's not playable against some lineups, but definitely something the Spurs can tread water with during the regular season. Come the playoffs, Aldridge for 35 minutes and Duncan for 29 minutes would allow one to be on the floor at all times, and them to play together for about 20 MPG. And that's probably conservative on the minute count.

Exactly.

[tmtcsc]

LMA is as good as Splitter on defense and is a threat on offense. You don't have to subs ute him out and I doubt you'll ever see him airball a FT. He was a great pick up for the Spurs and still has more improvement left in him. If or when he gets going on the offensive side, look out. I wonder if Tim's knee is really sore or if they are trying to get LMA to shoot more without deferring.

No Morris or Bledsoe tonight right?

[Mel_13]

LMA is as good as Splitter on defense and is a threat on offense. You don't have to subs ute him out and I doubt you'll ever see him airball a FT. He was a great pick up for the Spurs and still has more improvement left in him. If or when he gets going on the offensive side, look out. I wonder if Tim's knee is really sore or if they are trying to get LMA to shoot more without deferring.

No Morris or Bledsoe tonight right?

Morris will be available. Bledsoe is done for the season.

[ParadoxEN]

. West>Aldridge is pretty telling of the invalidity of this stat. Any +\- stat is pretty generic and meaningless.

No, it's a valid stat; you just don't understand it.

Facts:

- RPM (DRPM + ORPM) does not measure how well a player has performed, offensively or defensively, this season.

- RPM is based on data from previous seasons.

- RPM is designed to predict how well a player will perform based on their own history.

- RPM doesn't represent large increases in performance well; it looks for trends over time.

[exstatic]

No, it's a valid stat; you just don't understand it.

Facts:

- RPM (DRPM + ORPM) does not measure how well a player has performed, offensively or defensively, this season.

- RPM is based on data from previous seasons.

- RPM is designed to predict how well a player will perform based on their own history.

- RPM doesn't represent large increases in performance well; it looks for trends over time.

So, it's completely useless to show how LMA is performing on either side of the court for just THIS year with the Spurs, no? Because that's how he was trying to use it. The ORtg and DRtg stats I posted were for just THIS season for LMA with the Spurs.

[ParadoxEN]

So, it's completely useless to show how LMA is performing on either side of the court for just THIS year with the Spurs, no? Because that's how he was trying to use it. The ORtg and DRtg stats I posted were for just THIS season for LMA with the Spurs.

Well, essentially, the stat uses past data to predict future impact. The "usefulness" of that is up to interpretation, IMO.

That said, I'd say there's a good chance that the poster you were quoting doesn't understand the stat at all. D-West has likely been a better defender than LMA over the course of his career, which doesn't say much about his current impact with the Spurs. Conversely, LMA has likely seen an increase in defensive impact recently, which DRPM is a poor indicator of.

[exstatic]

Well, essentially, the stat uses past data to predict future impact. The "usefulness" of that is up to interpretation, IMO.

That said, I'd say there's a good chance that the poster you were quoting doesn't understand the stat at all. D-West has likely been a better defender than LMA over the course of his career, which doesn't say much about his current impact with the Spurs. Conversely, LMA has likely seen an increase in defensive impact recently, which DRPM is a poor indicator of.

I think we might agree then that since the thread is about PATFO being surprised at his defense since signing here, the stat that ceperez posted might not be relevant to the issue being discussed.

[GSH]

No, it's a valid stat; you just don't understand it.

Facts:

- RPM (DRPM + ORPM) does not measure how well a player has performed, offensively or defensively, this season.

- RPM is based on data from previous seasons.

- RPM is designed to predict how well a player will perform based on their own history.

- RPM doesn't represent large increases in performance well; it looks for trends over time.

It's only useful in retrospect. And only if you are willing to crow over the times it looks good, and throw out the times it looks bad. In other words, it's every bit as useful as a generic, newspaper horoscope.

At the considerable risk of quoting Billy Madison: "... what you've just said is one of the most insanely idiotic things I have ever heard. At no point in your rambling, incoherent response were you even close to anything that could be considered a rational thought. Everyone in this room is now dumber for having read it. I award you no points, and may God have mercy on your soul."

[ceperez]

No, it's a valid stat; you just don't understand it.

Facts:

- RPM (DRPM + ORPM) does not measure how well a player has performed, offensively or defensively, this season.

- RPM is based on data from previous seasons.

- RPM is designed to predict how well a player will perform based on their own history.

- RPM doesn't represent large increases in performance well; it looks for trends over time.

What are you talking about?

Here's the definition from ESPN:

ORPM: Player's estimated on-court impact on team offensive performance, measured in points scored per 100 offensive possessions

DRPM: Player's estimated on-court impact on team defensive performance, measured in points allowed per 100 defensive possessions

RPM: Player's estimated on-court impact on team performance, measured in net point differential per 100 offensive and defensive possessions. RPM takes into account teammates, opponents and additional factors

In short it takes into account TEAMATES, OPPONENTS and ADDITIONAL FACTORS from likely the SEASON where those numbers are coming from. It would be totally ridiculous if it was for a season where he's playing in a different team against different teams. What a load of B.S.! Show me a reference of your *interpretation* of what DRPM means?

Here's a more detailed description:

The metric isolates the unique plus-minus impact of each NBA player by adjusting for the effects of each teammate, opposing player and coach. ... The RPM model sifts through more than 230,000 possessions each NBA season to tease apart the "real" plus-minus effects attributable to each player, employing techniques similar to those used by scientific researchers when they need to model the effects of numerous variables at the same time.

It is an advanced metric that is nowhere near what you just described. Talk about complete B.S.

[ceperez]

It's only useful in retrospect. And only if you are willing to crow over the times it looks good, and throw out the times it looks bad. In other words, it's every bit as useful as a generic, newspaper horoscope.

At the considerable risk of quoting Billy Madison: "... what you've just said is one of the most insanely idiotic things I have ever heard. At no point in your rambling, incoherent response were you even close to anything that could be considered a rational thought. Everyone in this room is now dumber for having read it. I award you no points, and may God have mercy on your soul."

Yeah right, backing up a definition that is complete B.S. You got zero credibility in my book.

[ParadoxEN]

What are you talking about?

Here's the definition from ESPN:

ORPM: Player's estimated on-court impact on team offensive performance, measured in points scored per 100 offensive possessions

DRPM: Player's estimated on-court impact on team defensive performance, measured in points allowed per 100 defensive possessions

RPM: Player's estimated on-court impact on team performance, measured in net point differential per 100 offensive and defensive possessions. RPM takes into account teammates, opponents and additional factors

In short it takes into account TEAMATES, OPPONENTS and ADDITIONAL FACTORS from likely the SEASON where those numbers are coming from. It would be totally ridiculous if it was for a season where he's playing in a different team against different teams. What a load of B.S.! Show me a reference of your *interpretation* of what DRPM means?

Here's a more detailed description:

The metric isolates the unique plus-minus impact of each NBA player by adjusting for the effects of each teammate, opposing player and coach. ... The RPM model sifts through more than 230,000 possessions each NBA season to tease apart the "real" plus-minus effects attributable to each player, employing techniques similar to those used by scientific researchers when they need to model the effects of numerous variables at the same time.

It is an advanced metric that is nowhere near what you just described. Talk about complete B.S.

True or False: Does RPM use data from previous seasons?

[ceperez]

True or False: Does RPM use data from previous seasons?

It does, but it is not as simplistic as how you make it out to be. I don't think you have the statistical or mathematical education to even begin to comprehend what it means.

Anyway, read: https://cornerthreehoops.wordpress.c...al-plus-minus/

I don't think you comprehend what "Bayesian prior" means so I'm not going to waste time explaining how your over simplified B.S. definition is wrong.

[ParadoxEN]

It's only useful in retrospect. And only if you are willing to crow over the times it looks good, and throw out the times it looks bad. In other words, it's every bit as useful as a generic, newspaper horoscope.

At the considerable risk of quoting Billy Madison: "... what you've just said is one of the most insanely idiotic things I have ever heard. At no point in your rambling, incoherent response were you even close to anything that could be considered a rational thought. Everyone in this room is now dumber for having read it. I award you no points, and may God have mercy on your soul."

Google "how does ESPN calculate real plus minus" and report back. It'll be a good learning experience, IMO.

[ParadoxEN]

It does, but it is not as simplistic as how you make it out to be. I don't think you have the statistical or mathematical education to even begin to comprehend what it means.

Anyway, read: https://cornerthreehoops.wordpress.c...al-plus-minus/

I don't think you comprehend what "Bayesian prior" means so I'm not going to waste time explaining how your over simplified B.S. definition is wrong.

What did I say that was incorrect, specifically?

EDIT: LOL. I read the linked article and it's agreeing with me word-for-word:

Real Plus-Minus does NOT measure how well a player has performed this season

One is that RPM tends to be skeptical of player improvements (or regressions) that exceed what is expected for a player that age.

RPM’s focus on prediction makes it a poor way to determine who should get end-of-season awards.

GSH, get in here bruh.

[T Park]

I'm loving LMA's work on the defensive end this year, I wasn't expecting him to be anywhere near as good as Splitter there, much less far superior.

We just hoped the drop off wouldnt be horrible.

Its been upgraded which to me is still shocking.

It still can get better as well!

[ceperez]

What did I say that was incorrect, specifically?

EDIT: LOL. I read the linked article and it's agreeing with me word-for-word:




GSH, get in here bruh.

The way it works, if you need a f**cking education is that it uses Bayesian statistics to make predictions. So by Bayes rule, it uses a Prior to calculate the Posterior. So it's in the model that it does take information from previous seasons and the current season to calculate the persons +/-. What it is saying is that if there is a massive unexpected improvement of a persons performance then it doesn't accurately reflect that. So in other words the numbers for Kawhi are likely lower than they should be. However it doesn't mean that it doesn't take in account improvements in the game from the current season. It is a complex metric but it is more accurate that the DRtg that you quoted that can tease out the effect of a player outside the usual players he plays with.

Your argument is like saying some a drop of fell into a lake therefore the lake is completely tainted and not fit for drinking. It is an oversimplification by someone who knows nothing about statistical modeling.

[BillMc]

Lamarcus has already played 915 minutes this year through 40% of the season, and that's with Pop managing his minutes. He's on pace for ~2,275 minutes.

Compare that to Splitter's 1,030 last year (only made 52 games). That forced the team to give 1,120 minutes of Baynes and 936 minutes to Bonner last year.

The Spurs were only able to get 18 minutes per game out of Tiago last year in the playoffs, which forced Baynes into action for 40 total minutes during the series. We all know how that turned out. Even though Tiago played solid D, his inability to stay healthy OR play long stretches (or for that matter, stretch the floor while out there) limited his effectiveness. It bit the team in the ass in the playoffs because the team had to find someone else to throw at Blake for 20 minutes a game.

Aldridge played 41 minutes a night in the crazy HOU-POR series a few years ago (three OT games) and still had something in the tank against the Spurs the next round, though his shooting % went down a bit. With less of the burden, imagine having a guy who you can easily pencil in for 2x Tiago's minutes come the playoffs and is more matchup proof. You really can eliminate the fourth big from the rotation at that point.

Great post

[ParadoxEN]

The way it works, if you need a f**cking education is that it uses Bayesian statistics to make predictions. So by Bayes rule, it uses a Prior to calculate the Posterior. So it's in the model that it does take information from previous seasons and the current season to calculate the persons +/-. What it is saying is that if there is a massive unexpected improvement of a persons performance then it doesn't accurately reflect that. So in other words the numbers for Kawhi are likely lower than they should be. However it doesn't mean that it doesn't take in account improvements in the game from the current season. It is a complex metric but it is more accurate that the DRtg that you quoted that can tease out the effect of a player outside the usual players he plays with.

Your argument is like saying some a drop of fell into a lake therefore the lake is completely tainted and not fit for drinking. It is an oversimplification by someone who knows nothing about statistical modeling.

Lol? What's my argument? Since it seems like you've forgotten, this is all I said:

No, it's a valid stat; you just don't understand it.

Facts:

- RPM (DRPM + ORPM) does not measure how well a player has performed, offensively or defensively, this season.

- RPM is based on data from previous seasons.

- RPM is designed to predict how well a player will perform based on their own history.

- RPM doesn't represent large increases in performance well; it looks for trends over time.

These are all TRUE; your link practically quotes me. There's no shame in admitting you were wrong for insulting me bro, it's okay.

Also what Drtg did I quote? Lol, you don't even remember who you're talking to.

(PS: I think it would help your case if you stopped throwing out fancy math terms, IMO. We all know that Bayes Rule is completely unrelated.)

[Mr. Body]

Aldridge stays on the court for much longer.

He's a shotblocking threat. Teams can't pull Tim out on p'n'rs and have a clear path to the rim.

He's a very good defensive rebounder, ending possessions efficiently.

He has good length and speed at his position and gets the rotations.

(His offensive rebounding is a plus, as an aside, coupled with how he's fleet enough to still get back on defense.)

[z0sa]

Its been pretty obvious from day 1 a a spur, tbh. And ppl were calling him out this season already when were annihilating trams on that end, lol. You know who you are

[Chinook]

Well, Paradox just ate #ceperez alive.