Sadly, RPM might prevent Kawhi from winning the MVP.

[midnightpulp]

Media (and society in general) is reliant more than ever on "Big Data" analytics to spoonfeed them what should otherwise be commonsense decisions. I hope Kawhi's comparatively "low" RPM doesn't dissuade voters from giving him a fair shake. Right now it's between LBJ, Westbrook, and Kawhi, the former of which both have "better" metrics.

And yes, this is another veiled criticism of RPM. I think it's an inaccurate stat vis a vis the faith media and fans alike have in it.

[Chinook]

Um, what? RPM loves Kawhi's offense enough to make up for the fact that it doesn't care about his D. He's ahead of Harden for example. That won't be the stat that sinks him. On/Off might. But as I've been saying, he will do the most for his candidacy by winning H2H games.

[Arcadian]

RPM can fuck itself in the ass with a used black dildo.

[dabom]

Kawhi is MCP faggot. :lmao

[midnightpulp]

Kawhi is MVP faggot. :lmao

Correction.

I agree. But the "metrics" like other players better. Same way the metrics overrate your boy House.

[lilbthebasedgod]

Big Data should be used for every decision ever tbh

[TXstbobcat]

Leonard just outplayed Harden in the 4th quarter tonight and made a huge statement to the MVP voters.

[SASdynasty!]

Casual fans still digging through "advanced" metrics.

[100%duncan]

Westbrook isnt winning MVP with the 7th seed.

It's between Bron and Whi

[BillMc]

RPM can fuck itself in the ass with a used dildo.

Hey, don't go quoting Shakespeare on us. :lol

[Keepin' it real]

Leonard just outplayed Harden in the 4th quarter tonight and made a huge statement to the MVP voters.

How so? MVP voters don't watch the Spurs. No one does ... besides us.

[TheGreatYacht]

Another reason why the nerds that created "advanced" statistics should stay out of basketball...

They also rate Fathead as an elite defender

[TheGreatYacht]

How so? MVP voters don't watch the Spurs. No one does ... besides us.
I was watching Undisputed today and Shannon Sharpe (who shits on the Spurs) said he doesn't bother watching them because they put him to sleep, but he KNOWS Kawhi isn't the MVP

Broussard also contributed with the "system" thing :lol

[TXstbobcat]

How so? MVP voters don't watch the Spurs. No one does ... besides us.

they will see the espn highlights. Pretty much all Leonard 4th quarter in the highlights.

[tonight...you]

Media (and society in general) is reliant more than ever on "Big Data" analytics to spoonfeed them what should otherwise be commonsense decisions. I hope Kawhi's comparatively "low" RPM doesn't dissuade voters from giving him a fair shake. Right now it's between LBJ, Westbrook, and Kawhi, the former of which both have "better" metrics.

And yes, this is another veiled criticism of RPM. I think it's an inaccurate stat vis a vis the faith media and fans alike have in it.
You used to be my absolute favorite poster, years ago.
I wouldn't miss a post by you...

Then the Lakers and Kobe went into the sewer and you seem like you haven't had the motivation like you used to: to make those awesome skewerings of Kool and anyone else that got in your way... slathered with all kinds of facts and graphs and what-not...
There's the baseball thread, which was okay, but who cares about that?


I really would like to see you get back to your Prime...
No offense, really, but I think many here would agree with me on this.

You were better than you are.

[spurraider21]

3 of the top 5 players in rpm are CP3, Jimmy Butler, and Nikola Jokic.

i dont think the media is blinding looking at rpm tbh

[99 Problems]

Real MVP. Other teams are going to be mentally destroyed before they take the crt.

[Chinook]

3 of the top 5 players in rpm are CP3, Jimmy Butler, and Nikola Jokic.

i dont think the media is blinding looking at rpm tbh

I'm pretty sure that Mid is just trolling by acting like RPM has horribly misjudged Kawhi, and a lot of fans are buying it hook, line and sinker. If you based MVP solely on RPM and record, Kawhi would win.

[RD2191]

IDK man, I'm sure Kawhi can rev his Tahoe pretty high but whatever.

[SpursFan86]

Like Chinook said, if any metric is going to sink Kawhi, it's raw on/off numbers. Kawhi pales in comparison to guys like LeBron, Westbrook, and Harden in that aspect. And unfortunately, uninformed people will just run with the narrative that the Spurs are still great without Kawhi on the floor and use that against him...despite anyone knowledgeable knowing damn well this team is an 8th seed getting swept by GS if not for Kawhi, yet with him, they're on pace to win well over 60 games and at least be competitive in the playoffs.

[FkLA]

I really doubt any advanced metric will prevent him from winning it. It's still a narrative driven award. Muh triple-double and muh 30 and 13 are by far the biggest reasons why the media has Westbrook and Harden as frontrunners. Will do a lot for Kawhi's narrative if he can annihilate Westbrook and GS in this next week. He already got Harden today.

[emanueldavidginobili]

This game showed a lot of people what KL is about. People are gushing about this kid right now.

[GSH]

I really doubt any advanced metric will prevent him from winning it. It's still a narrative driven award. Muh triple-double and muh 30 and 13 are by far the biggest reasons why the media has Westbrook and Harden as frontrunners. Will do a lot for Kawhi's narrative if he can annihilate Westbrook and GS in this next week. He already got Harden today.


That right there. It's got everything. Kawhi can write his own narrative now, if he can dominate the other contenders, one after another.

[Ditty]

Unless the Spurs have a huge meltdown in April, and the Rockets or Cavs pass the Spurs then Kawhi is going to win it.

[spursistan]

At some point team record has to take precedence, no?..Spurs (likely) going to blow out the Cavs/Rockets' out of water..we could be talking about a ridiculous 8-9 games difference in the standings in the end..that's no joke.....Lebron is still a better player, but you have to reward a lone All Star leading a team to mid 60-win total the year after their greatest player left the scene..

[midnightpulp]

I'm pretty sure that Mid is just trolling by acting like RPM has horribly misjudged Kawhi, and a lot of fans are buying it hook, line and sinker. If you based MVP solely on RPM and record, Kawhi would win.

Definitely an ulterior motive here, yes.

I don't like "inference" metrics. Fine for predicting consumer trends, healthcare treatment, and the like, but totally stupid to use in sports as a be-all/end-all value assessment.

[BillMc]

I really doubt any advanced metric will prevent him from winning it. It's still a narrative driven award. Muh triple-double and muh 30 and 13 are by far the biggest reasons why the media has Westbrook and Harden as frontrunners. Will do a lot for Kawhi's narrative if he can annihilate Westbrook and GS in this next week. He already got Harden today.

This. And grabbing the 1 seed would be a big boost as well.

[Chinook]

Definitely an ulterior motive here, yes.

I don't like "inference" metrics. Fine for predicting consumer trends, healthcare treatment, and the like, but totally stupid to use in sports as a be-all/end-all value assessment.

Again, do you even know if RPM does that? I've yet to see any actual formula for it. I've only seen speculation of what it is and how it's related to its ancestor.

[midnightpulp]

Again, do you even know if RPM does that? I've yet to see any actual formula for it. I've only seen speculation of what it is and how it's related to its ancestor.


RPM reflects enhancements to RAPM by Engelmann, among them the use of Bayesian priors, aging curves, score of the game and extensive out-of-sample testing to improve RPM's predictive accuracy.

If it's using Bayesian methods, it's inferring.


Bayesian inference is a method of statistical inference in which Bayes' theorem is used to update the probability for a hypothesis as more evidence or information becomes available.

E.g., Player A lights up scrubs in garbage time, ergo, his RPM is high, thus there's good "evidence" to hypothesize said performance over 48 minutes against starters.

"Well, it does adjust for opposition via a type of ELO system."

How can there be any legitimate statistical foundation in this sense when end-of-the-bench players might never play against a starter over the season? Even the sample sizes of bench players against starters will be insufficient to really draw any meaningful conclusions. A bench player might see his starting counterpart for only a few minutes per game. Not to mention the fact that usually in that case, the bench player is fresh while the starter might be near the end of his minute allotment for that particular quarter. So if a player like Bertans comes in and lights up Anthony Davis for three quick 3s and the Spurs increase their lead in that time frame, his RPM (for that particular segment) will be sky high. But it would be foolish to extrapolate that impact over 48 minutes, which RPM does, even when it supposedly "adjusts." It's blind to game conditions and context.

Again, this works fine for something like a drug trial: "These 1000 patients randomly selected from the general population showed a 70% improvement in arthritis after the 30 day trial. From this , we can safely estimate the efficacy of the drug to be in the range demonstrated during the trial."

But sports aren't a laboratory environment where variables are kept relatively constant. In the drug trial example, every patient had arthritis and took the same drug for the same amount of time. Of course there's variables like age, disease severity, and so on, but at the end of the day, they [I]all interacted with the drug in the same exact way. Kawhi and Kyle Andersen don't "interact" with Lebron James in any similar way. If Anderson does come in to play against Lebron, the conditions will be vastly different from when Kawhi was playing against him. And furthermore, Anderson's sample size will be insufficient to draw any meaningful conclusion about his "real" value.


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 [B]numerous variables at the same time.

Again, many of those "variables" do not interact with each other enough to even attempt to "model" anything. It's a highly speculative exercise.

[Down Under]

Harden will win it.

[Mal]

Westbrook isnt winning MVP with the 7th seed.

It's between Bron and Whi

Bron isn`t winning MVP with Cavs having record worst than Spurs, Warriors and Rockets. If Spurs pull 1st overall seed it`s Kawhi, if Rockets are close to Spurs it`s Harden. Doubt Curry will win it, even with Warriors getting 1st seed.

[UNT Eagles 2016]

RPM can fuck itself in the ass with a used black dildo.
agreed... it counts useless stats from blowouts against sorry teams the same as clutch happenings like last night and recent games for us. Also pimps up inorganic stats like RW's fake triple doubles.

[UNT Eagles 2016]

Bron isn`t winning MVP with Cavs having record worst than Spurs, Warriors and Rockets. If Spurs pull 1st overall seed it`s Kawhi, if Rockets are close to Spurs it`s Harden. Doubt Curry will win it, even with Warriors getting 1st seed.

Curry has an outside chance if they keep the #1 seed and Durant remains out until the playoffs or worse.

[UNT Eagles 2016]

Westbrook isnt winning MVP with the 7th seed.


agreed... otherwise, you'd have to give it to Kobe in 2006, clearly the best player in the league and literally the only thing that kept them from being a 10-72 team or worse... gave the mighty Suns all they could handle and much more before bowing out due to some choking and some bad luck on a rebound.

But seriously, Kobe's 2006 campaign >>>>>>>>>>>>>>>>> Westbrook's 2017 campaign. Inorganic double and triple doubles are useless and not uncommon or original. But Kobe's 62 points in 3 quarters, 81 points in a game, and 11 game winning shots on the season? That's just absurd. If that didn't win MVP as a non top 4 seed (in a year where Nash won it as a rather unimpressive 2-seed) then there will never be an MVP outside the top 4 in a conference.

[100%duncan]

Bron isn`t winning MVP with Cavs having record worst than Spurs, Warriors and Rockets. If Spurs pull 1st overall seed it`s Kawhi, if Rockets are close to Spurs it`s Harden. Doubt Curry will win it, even with Warriors getting 1st seed.

Yes I agree but they are, and should be if the media wasnt filled with idiots, the leading 2 candidates. Westbrook is a fake, his team isnt going anywhere. And I'd have been fine with Harden but Kawhi 3-1'd him already while outplaying him every time iirc.

[spursistan]

Raw ON/OFF is the only argument i would take Lebron for MVP over Kawhi. To me Russ is out of it, there is no way an MVP should go to a player on 7th seed/sub 50-win team

839015926825639936

[$pursDynasty]

RPM is a bull crap stat, I was checking their defensive ratings and some of the players ranked ahead of Kiwi like Kevin Love, Otto Porter, Jokic prove it. Just like when D.Blair and the Red Mamba would be the best Spurs on the court via +/- but anyone with two eyes and common sense always knew otherwise and only trolls suggested differently.

[bklynspursfan]

Westbrook is averaging 10.0 apg exactly, so if that drops below 10, I would assume most voters would bump him down.

How Kawhi finishes this season will be key, and more importantly the big games against other candidates and top teams. He has a very good chance, but I fear the media is too obsessed with Harden/WB at this point

[SpurPadre]

I think the voters will not vote Lebron because of all the help he constantly asks for and usually gets. Wouldn't be surprised if the fucker gets another big to make up for Bogut, tbh.

[FkLA]

Westbrook is averaging 10.0 apg exactly, so if that drops below 10, I would assume most voters would bump him down.

How Kawhi finishes this season will be key, and more importantly the big games against other candidates and top teams. He has a very good chance, but I fear the media is too obsessed with Harden/WB at this point

It's kind of disturbing how much truth there is to this. Amin El-Hassan straight up said this on Sportscenter this morning. Basically said ' if Westbrook maintains his triple-double average he's my MVP if not it's Kawhi'. A couple of tenths of an average shouldn't determine an award like this smh.

[HarlemHeat37]

Nobody is looking at RPM as a primary measure for MVP voting:lol the only media members that even follow advanced metrics are also the same guys who understand how they work(so they wouldn't view them incorrectly like most internet posters)..

Narrative is the most important criteria for MVP, which has been the case for a long time..Kawhi's "narrative" case is growing with games like last night..if he has 2 huge games vs. the Warriors, he'll actually have a legit shot IMO..

[BatManu20]

Nobody is looking at RPM as a primary measure for MVP voting:lol the only media members that even follow advanced metrics are also the same guys who understand how they work(so they wouldn't view them incorrectly like most internet posters)..

Narrative is the most important criteria for MVP, which has been the case for a long time..Kawhi's "narrative" case is growing with games like last night..if he has 2 huge games vs. the Warriors, he'll actually have a legit shot IMO..

Agreed. But if he slips in either and we lose, I think he has no shot. Esp if WB averages the Triple-dub. Media is more obsessed with them than ever. Spurs have to over the dubs for the 1-seed for Kawhi to have a chance imo.

[BatManu20]

Should also be noted that the Thunder have a pretty favorable schedule down the stretch. Still play the Suns, Twolves, Nuggets twice, Milwaukee, Orlando, Kings, and Nets... WB will likely put up big numbers against these bottom-feeders (minus a decent Nugg to team).

[HarlemHeat37]

I think 2nd and 3rd place votes are going to play a larger role than in previous years IMO..Westbrook is going to be a very polarizing choice, since his historically great season(triple double + by far the most valuable player in the NBA in terms of impact compared to teammates) is going to be negated by the 7th seed in the eyes of many voters(only Moses Malone won MVP with an unimpressive record IIRC)..

If OKC cracks 50 wins(which they won't, since their team is horribly constructed), then I could see Westbrook getting it..

[bklynspursfan]

Is it the same voters for DPOY & MVP, or all in general?

Wouldn't it be something if they don't give Kawhi MVP but decide to vote him DPOY as a consolation prize? The meltdown from Green & co. would be glorious

[HarlemHeat37]

Is it the same voters for DPOY & MVP, or all in general?

Wouldn't it be something if they don't give Kawhi MVP but decide to vote him DPOY as a consolation prize? The meltdown from Green & co. would be glorious

It's a mix of media members..some of them vote for both, others don't..

[gambit1990]

Nobody is looking at RPM as a primary measure for MVP voting:lol
seriously :lol

just another one of mid's shitty takes.

[SpursFan86]

Narrative is the most important criteria for MVP, which has been the case for a long time..Kawhi's "narrative" case is growing with games like last night..if he has 2 huge games vs. the Warriors, he'll actually have a legit shot IMO..


Exactly...I mean, look what last night did for him. I've heard more talk of Kawhi being MVP in the past 24 hours than I had heard the entire season prior :lol

[TD 21]

Nobody is looking at RPM as a primary measure for MVP voting:lol the only media members that even follow advanced metrics are also the same guys who understand how they work(so they wouldn't view them incorrectly like most internet posters)..

Narrative is the most important criteria for MVP, which has been the case for a long time..Kawhi's "narrative" case is growing with games like last night..if he has 2 huge games vs. the Warriors, he'll actually have a legit shot IMO..

:tu

The Cavaliers game is also huge.

[HarlemHeat37]

Exactly...I mean, look what last night did for him. I've heard more talk of Kawhi being MVP in the past 24 hours than I had heard the entire season prior :lol

Yep..

- Best player on the best team
- Head-to-head matchups vs. other candidates
- National TV games, especially with key moments or clutch shots
- New system or loss of a key teammate(both true for Harden and Westbrook)
- Team struggling without player(Lebron)
- Historic statistical achievement(averaging a triple double)

It's usually a combination of those traits, the last one being least relevant, since it doesn't occur very often IMO..

[midnightpulp]

Nobody is looking at RPM as a primary measure for MVP voting:lol the only media members that even follow advanced metrics are also the same guys who understand how they work(so they wouldn't view them incorrectly like most internet posters)..

Narrative is the most important criteria for MVP, which has been the case for a long time..Kawhi's "narrative" case is growing with games like last night..if he has 2 huge games vs. the Warriors, he'll actually have a legit shot IMO..

Here's my breakdown of the flaws with inference stats like RPM/RAPM:


RPM reflects enhancements to RAPM by Engelmann, among them the use of Bayesian priors, aging curves, score of the game and extensive out-of-sample testing to improve RPM's predictive accuracy.

If it's using Bayesian methods, it's inferring.


Bayesian inference is a method of statistical inference in which Bayes' theorem is used to update the probability for a hypothesis as more evidence or information becomes available.

E.g., Player A lights up scrubs in garbage time, ergo, his RPM is high, thus there's good "evidence" to hypothesize said performance over 48 minutes against starters.

"Well, it does adjust for opposition via a type of ELO [me: I'm assuming] system."

How can there be any legitimate statistical foundation in this sense when end-of-the-bench players might never play against a starter over the season? Even the sample sizes of bench players against starters will be insufficient to really draw any meaningful conclusions. A bench player might see his starting counterpart for only a few minutes per game. Not to mention the fact that usually in that case, the bench player is fresh while the starter might be near the end of his minute allotment for that particular quarter. So if a player like Bertans comes in and lights up Anthony Davis for three quick 3s and the Spurs increase their lead in that time frame, his RPM (for that particular segment) will be sky high. But it would be foolish to extrapolate that impact over 48 minutes, which RPM does, even when it supposedly "adjusts." It's blind to game conditions and context.

Again, this works fine for something like a drug trial: "These 1000 patients randomly selected from the general population showed a 70% improvement in arthritis after the 30 day trial. From this , we can safely estimate the efficacy of the drug to be in the range demonstrated during the trial."

But sports aren't a laboratory environment where variables are kept relatively constant. In the drug trial example, every patient had arthritis and took the same drug for the same amount of time. Of course there's variables like age, disease severity, and so on, but at the end of the day, they all interacted with the drug in the same exact way. Kawhi and Kyle Andersen don't "interact" with Lebron James in any similar way. If Anderson does come in to play against Lebron, the conditions will be vastly different from when Kawhi was playing against him. And furthermore, Anderson's sample size will be insufficient to draw any meaningful conclusion about his "real" value. Bench players don't interact with the starters in the same way, and so on.


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 [B]numerous variables at the same time.

Again, many of those "variables" do not interact with each other enough to even attempt to "model" anything. It's a highly speculative exercise.

Criticism of Bayes' Theorem:


As described in the article and in lecture, the theorem is used to calculate the conditional probability of an event usually in light of some newly discovered evidence. We have applied the formula to simple scenarios such as drawing balls from urns or choosing a good restaurant, and in both situations probabilities were given. In these cases, Bayes’ theorem is a definite truth and logical conclusions can be drawn easily from the results. However, the problems with Bayes’ theorem begin to manifest themselves when it comes to real life, complex situations, and this is also where the theorem runs into a fair amount of critics.

When applied to the real world, Bayes’ theorem is entirely based off an initial hypothesis. In order to establish this hypothesis though, one must assign some initial probability to it before they can continue adding new evidence to carry out the formula. As such, many argue that Bayes’ theorem is not as much a definite probability as it is a subjective assumption.

The last sentence raises the central issue. A sporting event simply isn't controlled enough (i.e. like a lab experiment) to even begin to develop a complete enough initial hypothesis to work from. Sample sizes also aren't large enough to effectively evaluate player value. Further issues manifest when you take into consideration that a basketball team is more whole than discreet (i.e whole greater than the sum of its parts). This is why I don't really have a problem with advanced team stats, but I think it's nigh-impossible to tease out individual player value in this context using inference and probabilistic methods. This is why I'm becoming more in favor of "hard" mathematical stats for player evaluation rather than subjective "advanced stats." Even "number stats" like PER are subjective, since it can't really quantify the value of assists.

Now in order to have a solid logical foundation to work from regarding "hard" stats, we need to figure out what wins basketball games (from a mathematical point-of-view). Shooting percentage vis a vis PPG and Usage can be a good indicator. We've seen throughout NBA history that successful teams often have efficient volume scorers (relative to the NBA average).

I also think more value needs to placed on where a player does his scoring. Effective paint scoring has long been a hallmark of great teams, so I think a 25 ppg/50% player who scores 40% of his points in the paint is more valuable than 25 ppg/50% player who scores 40% of his points on mid-range jumpshots. A stat that needs to be created relative this is a "Score off miss" stat. Intuition tells me it's easier to trigger fast breaks off missed jumpshots than it is off missed paint shots. But of course a good team defense can cover up the flaws in the former, which again, speaks to the difficulty of isolating performance in a fluid team sport like basketball.

I also think defense and playmaking will remain largely immeasurable and are better evaluated by the eye test.

[tonight...you]

Here's my breakdown of the flaws with inference stats like RPM/RAPM:



If it's using Bayesian methods, it's inferring.



E.g., Player A lights up scrubs in garbage time, ergo, his RPM is high, thus there's good "evidence" to hypothesize said performance over 48 minutes against starters.

"Well, it does adjust for opposition via a type of ELO [me: I'm assuming] system."

How can there be any legitimate statistical foundation in this sense when end-of-the-bench players might never play against a starter over the season? Even the sample sizes of bench players against starters will be insufficient to really draw any meaningful conclusions. A bench player might see his starting counterpart for only a few minutes per game. Not to mention the fact that usually in that case, the bench player is fresh while the starter might be near the end of his minute allotment for that particular quarter. So if a player like Bertans comes in and lights up Anthony Davis for three quick 3s and the Spurs increase their lead in that time frame, his RPM (for that particular segment) will be sky high. But it would be foolish to extrapolate that impact over 48 minutes, which RPM does, even when it supposedly "adjusts." It's blind to game conditions and context.

Again, this works fine for something like a drug trial: "These 1000 patients randomly selected from the general population showed a 70% improvement in arthritis after the 30 day trial. From this [Bayesian data crunching ensues], we can safely estimate the efficacy of the drug to be in the range demonstrated during the trial."

But sports aren't a laboratory environment where variables are kept relatively constant. In the drug trial example, every patient had arthritis and took the same drug for the same amount of time. Of course there's variables like age, disease severity, and so on, but at the end of the day, they all interacted with the drug in the same exact way. Kawhi and Kyle Andersen don't "interact" with Lebron James in any similar way. If Anderson does come in to play against Lebron, the conditions will be vastly different from when Kawhi was playing against him. And furthermore, Anderson's sample size will be insufficient to draw any meaningful conclusion about his "real" value. Bench players don't interact with the starters in the same way, and so on.



Again, many of those "variables" do not interact with each other enough to even attempt to "model" anything. It's a highly speculative exercise.

Criticism of Bayes' Theorem:



The last sentence raises the central issue. A sporting event simply isn't controlled enough (i.e. like a lab experiment) to even to begin to develop a complete enough initial hypothesis to work off of. Sample sizes also aren't large enough to effectively evaluate player value. Further issues manifest when you take into consideration that a basketball team is more whole than discreet (i.e whole greater than the sum of its parts). This is why I don't really have a problem with advanced team stats, but I think it's nigh-impossible to tease out individual player value in this context using inference and probabilistic methods. This is why I'm becoming more in favor of "hard" mathematical stats for player evaluation rather than subjective "advanced stats." Even "number stats" like PER are subjective, since it can't really quantify the value of assists.

Now in order to have a solid logical foundation to work from regarding "hard" stats, we need to figure out what wins basketball games (from a mathematical point-of-view). Shooting percentage vis a vis PPG and Usage can be a good indicator. We've seen throughout NBA history that successful teams often have efficient volume scorers (relative to the NBA average).

I also think more value needs to placed on where a player does his scoring. Effective paint scoring has long been a hallmark of great teams, so I think a 25 ppg/50% player who scores 40% of his points in the paint is more valuable than 25 ppg/50% player who scores 40% of his points on mid-range jumpshots. A stat that needs to be created relative this is a "Score off miss" stat. Intuition tells me it's easier to trigger fast breaks off missed jumpshots than it is off missed paint shots. But of course a good team defense can cover up the flaws in the former, which again, speaks to the difficulty of isolating performance in a fluid team sport like basketball.

I also think defense and playmaking will remain largely immeasurable and better evaluated by the eye test.

Great info... Wow.

[GSH]

- Historic statistical achievement(averaging a triple double)

It's usually a combination of those traits, the last one being least relevant, since it doesn't occur very often IMO..


Not sure what way you mean that. It's not usually relevant, since it doesn't happen? For sure. But when it does, it's a big deal. Averaging a triple double is really similar to when Sosa and McGwire were fighting to beat the home run record that everyone thought was untouchable. Nobody thought that averaging a triple double could ever be done again. That's easy to hype.

Westbrook is sitting on exactly 10.0 AST/Game. If he slips just a little, it will be like the Spurs going 41-1 at home last year - a footnote. If he hangs on to the triple double, Kawhi is going to have to make a real statement in this series of games to overcome that story.

[midnightpulp]

Also, RPM doesn't seem to have a significant advantage over PER (no, I'm not trying to claim PER is a be-all/end stat) in "predictive value."

http://insider.espn.com/nba/hollinger/statistics

http://www.espn.com/nba/statistics/rpm/_/sort/RPM

Personally, PER seems to be more accurate this season. It rates Leonard 4 spots higher than RPM (which I think has been vastly underrating Kawhi this season). I would take Westbrook over CP3 all day/everyday. Jimmy Butler and Lowry are also more accurately rated. And Raymond doesn't show up until page 2, which I also think is accurate, since what Raymond does is heavily reliant on Golden State's overall team make up and scheme. Raymond isn't a "plug and play" player, meaning you could throw him on the Lakers and expect the same performance.

I know the counter-argument here is that RPM is simply evaluating what value a player has to their particular team, but I've never read such. Its goal seems to be an attempt to quantify absolute player value in a vacuum. I mean, GMs seem to take these metrics seriously as such, since nice contracts are handed out to "undervalued" players all the time.

I'm not saying the stat doesn't have some validity, but it's really not much better than past advanced stats.

Ergo:

https://en.wikipedia.org/wiki/IBM_Award

[Em-City]

At some point team record has to take precedence, no?..Spurs (likely) going to blow out the Cavs/Rockets' out of water..we could be talking about a ridiculous 8-9 games difference in the standings in the end..that's no joke.....Lebron is still a better player, but you have to reward a lone All Star leading a team to mid 60-win total the year after their greatest player left the scene..


Here's my breakdown of the flaws with inference stats like RPM/RAPM:



If it's using Bayesian methods, it's inferring.



E.g., Player A lights up scrubs in garbage time, ergo, his RPM is high, thus there's good "evidence" to hypothesize said performance over 48 minutes against starters.

"Well, it does adjust for opposition via a type of ELO [me: I'm assuming] system."

How can there be any legitimate statistical foundation in this sense when end-of-the-bench players might never play against a starter over the season? Even the sample sizes of bench players against starters will be insufficient to really draw any meaningful conclusions. A bench player might see his starting counterpart for only a few minutes per game. Not to mention the fact that usually in that case, the bench player is fresh while the starter might be near the end of his minute allotment for that particular quarter. So if a player like Bertans comes in and lights up Anthony Davis for three quick 3s and the Spurs increase their lead in that time frame, his RPM (for that particular segment) will be sky high. But it would be foolish to extrapolate that impact over 48 minutes, which RPM does, even when it supposedly "adjusts." It's blind to game conditions and context.

Again, this works fine for something like a drug trial: "These 1000 patients randomly selected from the general population showed a 70% improvement in arthritis after the 30 day trial. From this [Bayesian data crunching ensues], we can safely estimate the efficacy of the drug to be in the range demonstrated during the trial."

But sports aren't a laboratory environment where variables are kept relatively constant. In the drug trial example, every patient had arthritis and took the same drug for the same amount of time. Of course there's variables like age, disease severity, and so on, but at the end of the day, they all interacted with the drug in the same exact way. Kawhi and Kyle Andersen don't "interact" with Lebron James in any similar way. If Anderson does come in to play against Lebron, the conditions will be vastly different from when Kawhi was playing against him. And furthermore, Anderson's sample size will be insufficient to draw any meaningful conclusion about his "real" value. Bench players don't interact with the starters in the same way, and so on.



Again, many of those "variables" do not interact with each other enough to even attempt to "model" anything. It's a highly speculative exercise.

Criticism of Bayes' Theorem:



The last sentence raises the central issue. A sporting event simply isn't controlled enough (i.e. like a lab experiment) to even begin to develop a complete enough initial hypothesis to work from. Sample sizes also aren't large enough to effectively evaluate player value. Further issues manifest when you take into consideration that a basketball team is more whole than discreet (i.e whole greater than the sum of its parts). This is why I don't really have a problem with advanced team stats, but I think it's nigh-impossible to tease out individual player value in this context using inference and probabilistic methods. This is why I'm becoming more in favor of "hard" mathematical stats for player evaluation rather than subjective "advanced stats." Even "number stats" like PER are subjective, since it can't really quantify the value of assists.

Now in order to have a solid logical foundation to work from regarding "hard" stats, we need to figure out what wins basketball games (from a mathematical point-of-view). Shooting percentage vis a vis PPG and Usage can be a good indicator. We've seen throughout NBA history that successful teams often have efficient volume scorers (relative to the NBA average).

I also think more value needs to placed on where a player does his scoring. Effective paint scoring has long been a hallmark of great teams, so I think a 25 ppg/50% player who scores 40% of his points in the paint is more valuable than 25 ppg/50% player who scores 40% of his points on mid-range jumpshots. A stat that needs to be created relative this is a "Score off miss" stat. Intuition tells me it's easier to trigger fast breaks off missed jumpshots than it is off missed paint shots. But of course a good team defense can cover up the flaws in the former, which again, speaks to the difficulty of isolating performance in a fluid team sport like basketball.

I also think defense and playmaking will remain largely immeasurable and are better evaluated by the eye test.

Dope post... Thanks.

[140]

seriously :lol

just another one of mid's shitty takes.
:lol

[bklynspursfan]

It's kind of disturbing how much truth there is to this. Amin El-Hassan straight up said this on Sportscenter this morning. Basically said ' if Westbrook maintains his triple-double average he's my MVP if not it's Kawhi'. A couple of tenths of an average shouldn't determine an award like this smh.

really shouldnt... i bet he's not the only one thinking like that

[GSH]

Westbrook is averaging 10.0 apg exactly, so if that drops below 10, I would assume most voters would bump him down.


It's kind of disturbing how much truth there is to this. Amin El-Hassan straight up said this on Sportscenter this morning. Basically said ' if Westbrook maintains his triple-double average he's my MVP if not it's Kawhi'. A couple of tenths of an average shouldn't determine an award like this smh.

Oh, shit. I got lazy and didn't read the whole thread. I didn't know you'd already said that. :shootme

You're right, a few tenths shouldn't decide. But I think we all know that they won't be able to resist pimping a "historic achievement".

Here's a fun scenario: Imagine what happens if Harden comes up with an average of something like 9.95 AST/Game, and media outlets round up and give him credit for averaging a triple double for a whole season. And THEN they give him the MVP. :lol

SASmith and all the other screaming jackass "analysts" would have a field day - on both sides of the argument.

[spurraider21]

:lol PER in 2017

[100%duncan]

839538383705817092
oh fucking well

[Raven]

don't worry, rpm is not based on Big Data.

[midnightpulp]

:lol PER in 2017

You missed the point.

Again, RPM doesn't seem to be ranking players any more accurately than PER. I'll concede RPM is better for ranking undervalued players who don't fill up the stat sheets (your 3 and D guys and the like), but on/off has done just a good of job there as RPM has for over a decade.

[spurraider21]

You missed the point.

Again, RPM doesn't seem to be ranking players any more accurately than PER. I'll concede RPM is better for ranking undervalued players who don't fill up the stat sheets (your 3 and D guys and the like), but on/off has done just a good of job there as RPM has for over a decade.PER takes raw box score numbers and plugs it into a formula... there's no analytics to it. the only credit you get for defense is steals and blocks :lol

[midnightpulp]

PER takes raw box score numbers and plugs it into a formula... there's no analytics to it. the only credit you get for defense is steals and blocks :lol

It is still considered an "advanced" stat.

If you read the thread, I pretty much concede that defense is largely immeasurable. The eye test is better for evaluation on that end.

And again, missing the point. RAPM/RPM are supposed to be these fancy new data driven stats, and yet they don't seem to have much of any explanatory power over PER or the IBM stat that was created in the 80's. I don't mind RPM being used in conjunction with the 1000 other basketball stats, but it's by no means the stat.

[skin27]

Kawhi needs 5 straight 40 point game to steal the MVP from harden!!

[spurraider21]

It is still considered an "advanced" stat.

If you read the thread, I pretty much concede that defense is largely immeasurable. The eye test is better for evaluation on that end.

And again, missing the point. RAPM/RPM are supposed to be those fancy new data driven stats, and yet they don't seem to have much of any explanatory power over PER or the IBM stat that was created in the 80's. I don't mind RPM being used in conjunction with the 1000 other basketball stats, but it's by no means the stat.i think think PER is complete crap. it's literally plug and chug and based on stats that were arbitrarily chosen to be important indicators with no context...

it's like passer rating in the NFL, but at least passing can largely summarized by those handfuls of stats (completions, attempts, yards, TD's, INT's)... the only things that aren't factored into passer rating are things like drops

basketball stats require MUCH more complex than NFL stats, because it's much less structured

[midnightpulp]

i think think PER is complete crap. it's literally plug and chug and based on stats that were arbitrarily chosen to be important indicators with no context...

it's like passer rating in the NFL, but at least passing can largely summarized by those handfuls of stats (completions, attempts, yards, TD's, INT's)... the only things that aren't factored into passer rating are things like drops

basketball stats require MUCH more complex than NFL stats, because it's much less structured

All stats are.

A stat proves its worth if it can correlate, and PER does correlate with other stats, overall team success, and "perception" (i.e. which players are said to be the best via the eye test).

RPM correlates, too. But not any better than stats we've had for 30 years, which is why I'm no longer enamored with it.

[spurraider21]

win shares work better than PER...

[midnightpulp]

win shares work better than PER...

Per 48, I assume?

A player on great team who had a long career will have racked up a shitload of WS (e.g. John Stockton).

[spurraider21]

Per 48, I assume?

A player on great team who had a long career will have racked up a shitload of WS (e.g. John Stockton).yeah, i mean you can look at win shares to compare specific seasons... w/s per 48 to compare trends of different lengths

[tholdren]

At some point team record has to take precedence, no?..Spurs (likely) going to blow out the Cavs/Rockets' out of water..we could be talking about a ridiculous 8-9 games difference in the standings in the end..that's no joke.....Lebron is still a better player, but you have to reward a lone All Star leading a team to mid 60-win total the year after their greatest player left the scene..

no. mvp is highest scorer on best team, in most cases and levels. this is a problem. mvp should go to best all around player

[tholdren]

All stats are.

A stat proves its worth if it can correlate, and PER does correlate with other stats, overall team success, and "perception" (i.e. which players are said to be the best via the eye test).

RPM correlates, too. But not any better than stats we've had for 30 years, which is why I'm no longer enamored with it.

stats lack context= the problem with stats and especially with trying to use stats objectively when they are created and depend on others.